Vol. 7 - Evolutionary Psycology and Economic Theory - Koppl, R.

All rights reserved.
May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or applicable
Copyright @ 2004. JAI Press Inc.
copyright law.
EBSCO : eBook Collection (EBSCOhost) - printed on 9/12/2018 1:04 PM via CONACYT - CONSEJO
NACIONAL DE CIENCIA Y TECNOLOGIA
AN: 166726 ; Koppl, Roger.; Evolutionary Psychology and Economic Theory
Account: s2037900.main.ehost
All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or applicable copyright law.
AND ECONOMIC THEORY

EVOLUTIONARY PSYCHOLOGY
Series Editor: Roger Koppl

Associate Editors: Jack Birner and
Peter Kurrild-Klitgaard
ADVANCES IN AUSTRIAN ECONOMICS
ADVANCES IN AUSTRIAN ECONOMICS VOLUME 7
EVOLUTIONARY
PSYCHOLOGY AND
ECONOMIC THEORY
EDITED BY
ROGER KOPPL
Department of Economics and Finance,
Fairleigh Dickinson University, USA
2004
Amsterdam – Boston – Heidelberg – London – New York – Oxford

Paris – San Diego – San Francisco – Singapore – Sydney – Tokyo
ELSEVIER B.V. ELSEVIER Inc. ELSEVIER Ltd ELSEVIER Ltd

Radarweg 29 525 B Street, Suite 1900 The Boulevard, Langford 84 Theobalds Road
P.O. Box 211 San Diego Lane, Kidlington London
1000 AE Amsterdam CA 92101-4495 Oxford OX5 1GB WC1X 8RR
The Netherlands USA UK UK
© 2004 Elsevier Ltd. All rights reserved.
This work is protected under copyright by Elsevier Ltd, and the following terms and conditions apply to its use:
Photocopying
Single photocopies of single chapters may be made for personal use as allowed by national copyright laws. Permission of the Publisher and
payment of a fee is required for all other photocopying, including multiple or systematic copying, copying for advertising or promotional
purposes, resale, and all forms of document delivery. Special rates are available for educational institutions that wish to make photocopies
for non-profit educational classroom use.
Permissions may be sought directly from Elsevier’s Rights Department in Oxford, UK; phone: (+44) 1865 843830, fax:
(+44) 1865 853333, e-mail: permissions@elsevier.com. Requests may also be completed on-line via the Elsevier homepage
(http://www.elsevier.com/locate/permissions).
In the USA, users may clear permissions and make payments through the Copyright Clearance Center, Inc., 222 Rosewood Drive,
Danvers, MA 01923, USA; phone: (+1) (978) 7508400, fax: (+1) (978) 7504744, and in the UK through the Copyright Licensing Agency
Rapid Clearance Service (CLARCS), 90 Tottenham Court Road, London W1P 0LP, UK; phone: (+44) 20 7631 5555; fax: (+44) 20 7631
5500. Other countries may have a local reprographic rights agency for payments.
Derivative Works
Tables of contents may be reproduced for internal circulation, but permission of the Publisher is required for external resale or distribution
of such material. Permission of the Publisher is required for all other derivative works, including compilations and translations.
Electronic Storage or Usage

Permission of the Publisher is required to store or use electronically any material contained in this work, including any chapter or part
of a chapter.
Except as outlined above, no part of this work may be reproduced, stored in a retrieval system or transmitted in any form or by any
means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission of the Publisher.
Address permissions requests to: Elsevier’s Rights Department, at the fax and e-mail addresses noted above.
Notice
No responsibility is assumed by the Publisher for any injury and/or damage to persons or property as a matter of products liability,
negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein.
Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made.
First edition 2004
British Library Cataloguing in Publication Data

A catalogue record is available from the British Library.
ISBN: 0-7623-1138-X
ISSN: 1529-2134 (Series)
⃝ The paper used in this publication meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). Printed in The
∞
Netherlands.
CONTENTS
LIST OF CONTRIBUTORS ix
ADVISORY BOARD xi
ECONOMICS EVOLVING: AN INTRODUCTION TO THE

VOLUME
Roger Koppl 1
ECONOMICS AND EVOLUTIONARY PSYCHOLOGY

David Friedman 17
MOGUL GAMES: IN DEFENSE OF INEQUALITY AS AN

EVOLUTIONARY STRATEGY TO COPE WITH MULTIPLE
AGENTS OF SELECTION
Deby Cassill and Alison Watkins 35
AUSTRIAN ECONOMICS, EVOLUTIONARY PSYCHOLOGY

AND INDIVIDUAL ACTIONS
Geoffrey M. Hodgson 61
HAYEK AND MODERN EVOLUTIONARY THEORY

Paul H. Rubin and Evelyn Gick 79
HAYEK’S THEORY OF THE MIND

Brian J. Loasby 101
IS HOMO ECONOMICUS EXTINCT? VERNON SMITH,

DANIEL KAHNEMAN AND THE EVOLUTIONARY
PERSPECTIVE
C. Athena Aktipis and Robert O. Kurzban 135
v
vi
AUSTRIAN ECONOMICS, EVOLUTIONARY PSYCHOLOGY

AND METHODOLOGICAL DUALISM: SUBJECTIVISM
RECONSIDERED
Viktor J. Vanberg 155
THE NEW FABLE OF THE BEES: MULTILEVEL

SELECTION, ADAPTIVE SOCIETIES, AND THE CONCEPT
OF SELF INTEREST
David Sloan Wilson 201
SYMPOSIUM ON GROUP SELECTION
GROUP SELECTION AND METHODOLOGICAL

INDIVIDUALISM: COMPATIBLE AND COMPLEMENTARY
Douglas Glen Whitman 221
ON GROUP SELECTION AND METHODOLOGICAL

INDIVIDUALISM – A REPLY TO DOUGLAS GLEN
WHITMAN
Elliott Sober and David Sloan Wilson 251
COMMENT ON “GROUP SELECTION AND

METHODOLOGICAL INDIVIDUALISM: COMPATIBLE AND
COMPLEMENTARY” BY DOUGLAS GLEN WHITMAN
Richard N. Langlois 261
RECONCILING GROUP SELECTION AND

METHODOLOGICAL INDIVIDUALISM
Todd J. Zywicki 267
INDIVIDUALISM
Adam Gifford Jr.
Douglas Glen Whitman

INDIVIDUALISM: REPLY TO COMMENTS
GROUP SELECTION AND METHODOLOGICAL
LEVELS OF SELECTION AND METHODOLOGICAL
297
279
vii
LIST OF CONTRIBUTORS
C. Athena Aktipis Department of Psychology, University of

Pennsylvania, USA
Deby Cassill Department of Environmental Science and
Policy, University of South Florida St.
Petersburg, USA
David Friedman School of Law, Santa Clara University,
USA
Evelyn Gick Department of Economics, Dartmouth
College, USA
Adam Gifford Jr. Department of Economics, California State
University, USA
Geoffrey M. Hodgson The Business School, University of
Hertfordshire, UK
Roger Koppl Department of Economics and Finance,
Fairleigh Dickinson University, USA
Robert O. Kurzban Department of Psychology, University of
Pennsylvania, USA
Richard N. Langlois Department of Economics, The University
of Connecticut, USA
Brian J. Loasby Department of Economics, University of
Stirling, UK
Paul H. Rubin Department of Economics and School of
Law, Emory University, USA

Elliott Sober Department of Philosophy, Stanford
University, USA
Viktor J. Vanberg University of Freiburg and Walter Eucken
Institute, Freiburg, Germany
ix
x
Alison Watkins Information Systems and Decision

Sciences, University of South Florida St.
Petersburg, USA
Douglas Glen Whitman Department of Economics, California State
University, USA
David Sloan Wilson Departments of Biology and Anthropology,
SUNY Binghamton, USA
Todd Zywicki School of Law, George Mason University,
USA
ADVISORY BOARD
Don Bellante Uskali Mäki

University of South Florida, USA University of Helsinki, Finland
James Buchanan Ferdinando Meacci
George Mason University, USA Università degli Studi di Padova, Italy
Stephan Boehm Mark Perlman
University of Graz, Austria University of Pittsburgh, USA
Peter J. Boettke John Pheby
George Mason University, USA University of Luton, UK
Bruce Caldwell Warren Samuels
University of North Carolina, USA Michigan State University, USA
Jaques Garello Barry Smith
Université d’Aix-Marseille, France State University of New York, USA
Roger Garrison Erich Streissler
Auburn University, USA University of Vienna, Austria
Jack High Martti Vihanto
George Mason University, USA Turku University, Finland
Masazumi Ikemoto Richard Wagner
Senshu University, Japan George Mason University, USA
Richard N. Langlois Lawrence H. White
The University of Connecticut, USA University of Missouri, USA
Brian Loasby Ulrich Witt
University of Stirling, UK Max Planck Institute, Germany

Ejan Mackaay
University of Montreal, Canada
xi
ECONOMICS EVOLVING: AN
INTRODUCTION TO THE VOLUME
Roger Koppl
The modern revival of “Austrian economics” dates to the South Royalton

conference of 1974 (Vaughn, 1994, pp. 103–111). At that time, neoclassical
orthodoxy excluded evolutionary concepts. It was, in Ludwig Lachmann’s
memorable phrase, “late classical formalism” (1977, p. 35). Opposition to
neoclassical orthodoxy was part of the definition of Austrian economics. It formed
part of our identity. Today it is no longer clear what “orthodoxy” is or whether
current mainstream economics is “neoclassical” at all (Colander et al., 2004).
One of the more salient changes in mainstream economics over the last 30 years
is the introduction of evolutionary ideas. Mainstream economics is rich with
evolutionary concepts. Evolutionary game theory, for example, is certainly a part
of today’s standard toolbox. Thirty years ago, it did not even exist.1 Some of
the evolutionary ideas entering mainstream economics are similar or identical to
ideas from the Austrian tradition. In this situation, it is no longer clear what the
Austrian differentiae are. I hope this volume will help to sort out some of the issues
relating to Austrian economics and one group of evolutionary ideas, namely, those
of evolutionary psychology.
The growth of evolutionary ideas in economics is a challenge and an opportunity
for Austrian economists. The challenge is to identify what we have to offer
to the new, more evolutionary mainstream. The opportunity is to find in this
new mainstream an audience more open to Austrian ideas. The contributors

to this volume help us to understand the opportunities and challenges for
Austrians considering the modern literature in evolutionary psychology. It is
Evolutionary Psychology and Economic Theory

Advances in Austrian Economics, Volume 7, 1–16
Copyright © 2004 by Elsevier Ltd.
All rights of reproduction in any form reserved
ISSN: 1529-2134/doi:10.1016/S1529-2134(04)07001-2
1
2 ROGER KOPPL
probably true that only a minority of contributors the current volume, however,
are Austrian economists. Some might be labeled “ambiguous Austrians.” But
all are seriously engaging issues in the crossroads where Austrian economics,
evolutionary economics, and evolutionary psychology meet.
The term “evolutionary psychology” has a strict meaning and a broad meaning.
In the strict sense, “evolutionary psychology” refers to the writers who take
as their manifesto Barkow, Cosmides and Tooby (1992). For them, “the mind
is a set of information-processing machines that were designed by natural
selection to solve adaptive problems faced by our hunter-gatherer ancestors”
(Cosmides & Tooby, 1997). Notice that the mind consists of multiple “information-
processing machines” rather than being one big, all-purpose calculating machine.
It is a hodgepodge of different cross-communicating “modules” or “decision
algorithms.” Different modules are “specialized to process different kinds of
information” (Barkow et al., 1992, p. 599).
In the strict sense, the contributions of David Friedman, Geoffrey Hodgson,
C. Athena Aktipis & Robert Kurzban, Paul Rubin & Evelyn Gick, and Viktor
Vanberg draw on evolutionary psychology. Brian Loasby expresses doubts about
evolutionary psychology in the strict sense. David Sloan Wilson is a pioneer of
evolutionary psychology in the broad sense, but opposed to important elements
of evolutionary psychology in the strict sense. In the broader sense, the term
evolutionary psychology applies to any theory or argument that draws conclusions
about human psychology from man’s evolutionary history. In this sense, all of the
contributions in this volume concern evolutionary psychology.
There is an even wider sense of the term employed by Heyes (2000) who
favors an “evolutionary psychology” of all sentient life. In this view evolutionary
psychology in the strict sense is narrow in its focus on humans. She coins the term
“evolutionary psychology in the round” to capture this idea. It is quite normal,
however, to use the word “psychology” to refer to human psychology. A university
may have a department of psychology, but I am not aware of any departments of
“human psychology.”2 The real point of Hayes’s phrase, “evolutionary psychology
in the round,” seems to be its potential as a blunt weapon with which to attack
evolutionary psychology in the strict sense. This sense of the term will not distract
us further.
Hayek was an evolutionary psychologist in the broad sense. His book The
Sensory Order relies on the concepts of natural science to explain how, in principle,
mind could emerge from matter. (He thus commits the phenomenological sin of
“naturalizing consciousness.”)3 Mises, by contrast, merely tipped his hat to the
evolutionary origins of human reason (1949, pp. 32–36). He seems to have thought
that evolution produced in man an animal more rational than the beast described
by evolutionary psychologists in the strict sense. As I note below, evolutionary
Economics Evolving: An Introduction to the Volume 3
psychology contradicts this view, while supporting other vital aspects of Mises’
position.
The contributions of Friedman and Vanberg both contain useful primers on

evolutionary psychology in the strict sense. Cosmides and Tooby have posted their
own primer on the web (http://www.psych.ucsb.edu/research/cep/primer.html).
As Friedman notes, evolutionary psychology gives us “some ability to predict
what utility functions individuals will have.” This knowledge is limited and
indirect. As Friedman notes, evolutionary biology lets us know, metaphorically,
“the objective of genes,” namely, “reproductive success, getting as many copies
of themselves as possible into future generations.” Combining this central idea
with knowledge of our evolutionary past helps us form guesses about what sorts of
dispositions people are armed with. As Friedman makes clear, these guesses cannot
be too particular. Our guesses are weak in part because “we have found ways . . . to
sabotage the objectives of our genes.” We are not endowed with a generalized skill
of reproduction, whether of genes or organisms. We are endowed with specialized
preferences and dispositions that helped the genes of our ancestors to reproduce.
The desire for status is an example. Such desires and dispositions can be satisfied
without serving their original evolutionary function. Friedman illustrates by noting
that using birth control and keeping pets allow us to have sex and dependent love
objects without reproducing. Friedman’s point is consistent with Hayek’s insight
that cultural evolution partly overrides the behavioral dispositions produced by
biological evolution (Hayek, 1979).
Evolutionary psychologists might object to putting Hayek’s point in just
this way, which seems to contrast “nature” with “nurture,” and “genes” with
“culture.”4 Genes and culture co-evolve. The actions implied by our “nature”
vary, sometimes radically, with the particular environments we experience, and
thus depend on “nurture.” An evolutionary psychologist would therefore probably
express Hayek’s point differently. Because the “modern” conditions of civilized
life have “existed for an evolutionarily insignificant amount of time,” Donald
Symons explains, “we have no adaptations specifically designed to deal with” them.
A given psychological adaptation “may be less adaptive (i.e. less reproduction-
maximizing) today than it once was, or perhaps maladaptive” (Symons, 1992, p.

138). Our natural psychology is not well suited to many of the rules we are subject to
in modern society. Only in this somewhat attenuated sense does culture “override”
biology. Although the supposed dichotomy between “culture” and “biology” is
simplistic, the substance of Hayek’s point is good. As Cosmides and Tooby (1997)
put it, “Our modern skulls house a stone age mind.”5
Friedman points out that evolutionary psychology also gives us a theory of
error. It tells us that people “will make those mistakes that would have led to
reproductive success in the environment in which the psychological characteristics
4 ROGER KOPPL
leading to those mistakes evolved.” This insight deserves some attention. Recently,
Gigerenzer and others have put forward the theory of “fast and frugal heuristics,”
which builds in part on evolutionary psychology (Gigerenzer & Selten, 2001).

They study simple heuristics whose out-of-sample performance beats that of linear
regression and other seemingly more rational techniques. At points they seem to
take it for granted that people in society use high-performance algorithms. Todd
(2001) says, for example, “the vision of ecological rationality,” characterizing the
literature on fast and frugal heuristics, “emphasizes that humans use specific simple
heuristics because they enable adaptive behaviors, by exploiting the structure of
information in natural decision environments” (pp. 52–53, emphasis in original).
But as Henrich et al. (2001) note in the same volume, we have clear examples
in which cultural evolution produces maladaptive heuristics (pp. 355–357).
Evolutionary psychology tells us about the fabric of human nature; social theory
tells us what different societies make with that fabric. Different institutions produce
different results.
Deby Cassill provides a radically new insight into the propensity of people and
other animals to fall into hierarchical structures. Her notion of “skew selection”
(Cassill, 2002, 2003) is a radically new model of dominance hierarchies. Most
traditional models can be lumped into one of two classes: those assuming
essentially selfish individuals and those relying on some notion of altruism, whether
based on kinship, group selection, or reciprocity. Cassill points out that if group
membership has both costs and benefits, and if those costs and benefits vary with
one’s position in the hierarchy, then natural selection should tend to favor organisms
that know when to support the group and when to exploit their position within the
group. In periods of safety and relative abundance, the strong support the weak.
In periods of danger or relative want, the agents of death attach themselves to the
weak and pass over the strong. This arrangement is clearly good for the strong.
They have a buffer between themselves and the agents of death. The weak benefit
too, however, since they can reproduce if the good times last long enough. The
strong, then, have a selfish incentive to make clever investments in the welfare of
others in the group. Natural selection exploits love and its feelings of altruism to
serve the selfish genes of the group’s strongest members.

Cassill’s simulations suggest that in human societies hierarchy may be natural
in at least two senses. First, we may be programmed for life in hierarchical
systems. If we are, then our psychology is oriented to groups and may include a
deep-seated desire to feel solidarity with others and a natural disposition to rank
oneself in relation to others. Second, even in historical time, hierarchies may
emerge “naturally” with little planning or foresight by the very individuals who
emerge as “moguls.” Both points fit nicely with Hayek’s social theory. The first
point resonates with Hayek’s notion of the “atavism of social justice” (Hayek,
1978) and expands upon it. We are at once selfish and altruistic. We have genuine
feelings of solidarity for those within our group. These noble feelings, however,
may correspond to fear or hatred of those outside the group. Hayek’s political
philosophy seems to have little room for such feelings of solidarity except,
perhaps, to suggest we mortify them. The second point suggested by Cassill’s
paper, that the emergence of hierarchies and moguls is “natural,” resonates with
Hayek’s argument that the very idea “social justice” is incoherent because the
distribution of income is not planned in a spontaneous order (Hayek, 1976,
pp. 96–97).
Geoffrey Hodgson draws attention to an important issue Austrians need to
confront. As an expression of thoroughgoing Darwinism, evolutionary psychology
invokes the “principle of evolutionary explanation,” which Hodgson defines as “the
injunction that any behavioral assumption, including in the social sciences, must
be capable of causal explanation in evolutionary terms, or at least be consistent
with a scientific understanding of human evolution.” Hodgson’s call to place greater
emphasis on habit and to place habit “between instinct and reason” resonates nicely
with Hayek’s ideas on cultural evolution. Hayek (1979) places culture, including
morality, between instinct and reason. Significantly, Hodgson cites Hayek’s remark
that “Man is as much a rule-following animal as a purpose-seeking one” (Hayek,
1973, p. 11).
Hodgson is right to note that many Austrians have made “psychologizing” taboo:
Economic theory must not consider the content of individual preferences. Hodgson
cites passages in Mises, Hayek, Lachmann, and Shackle supporting this position.
If evolutionary psychologists are mapping out a universal human psychology that
includes “domain-specific cognitive specializations” (Tooby & Cosmides, 1992,
p. 71) and “content-specific” problem-solving “mechanisms” (pp. 112–113), then
Mises was wrong to associate the content of action with “history” and the pure,
content-independent form of action with “theory.”
The Misesian project of “praxeology” was built on the idea that the only
universals in human action concern its categorical form (Mises, 1949, pp. 32–36).
As Hodgson argues, this sort of distinction is inconsistent with Darwinism and
evolutionary psychology. Mises modeled man as “rational” in a rather strong sense.

“That reason has the power to make clear through pure ratiocination the essential
features of action is a consequence of the fact that action is an offshoot of reason”
(Mises, 1949, p. 39). Modern psychology, including evolutionary psychology in
the strict sense, shows that man is not rational in the strong sense invoked by
Mises. Human rationality in the social world is not a set of pure categorical forms,
but a collection of pre-programmed decision algorithms supplemented by the
idiosyncratic results of individual learning. We can mentally isolate the universal
aspects of human thought and action, but only by empirically identifying the design
6 ROGER KOPPL
architecture of human thought. Austrians would do well to revise pure praxeology

along more Darwinian lines.
Hayek’s approach to human reason was far more empirical and skeptical than
that of Mises. Indeed, Hayek wrote a famous book in “theoretical psychology,”
The Sensory Order, which attempted to explain how an earthly central nervous
system could produce thought sufficiently rich to generate the sort of “sensory
order” each of us knows from direct acquaintance (Hayek, 1952). In other words,
Hayek worked out an empirically driven theory of mind.
Brian Loasby’s contribution to this volume emphasizes the connectionist
aspects of Hayek’s theory, which he shared with Adam Smith and Alfred
Marshall. In The Sensory Order, Hayek discusses the formation of “connexions”
among brain cells. He explicitly eschews any attempt to sort out what part of the
process “takes place in the course of the development of the single individual”
and what part “takes place in the course of the development of the species” (1952,
p. 102). Evolutionary psychologists in the strict sense emphasize phylogeny,
the development of the species. They do not deny the importance of ontogeny,
the development of the individual, or of individual differences. Their research
project, however, is to seek out the universals of human cognition and trace their
operation in the diverse cultures. Loasby is interested in the very differences
neglected by evolutionary psychology and in our ability to think up new ways
to classify events. He emphasizes “the ability of ‘the development of the single
individual’ in some circumstances to override – though not to replace – the
results of ‘the development of the species.’ ” Indeed, for Loasby, the “significance
of such a capability can hardly be overestimated.” History matters, and so
does imagination.
Loasby warns us of the dangers of “neoDarwinism,” including its potential
to converge with traditional equilibrium theory. He draws our attention to an
important criticism of evolutionary psychology in the strict sense, namely, that
of Annette Karmiloff-Smith (2002). She criticizes Steven Pinker’s discussion of
Williams’ syndrome and his call for “cognitive genetics.” This discussion is useful
for any economist interested in evolutionary psychology. Regardless of which
party has the better argument, following the dispute helps one to sort out the
issues and understand the nature of the sometimes strikingly virulent opposition
to evolutionary psychology (Kamiloff-Smith cites Pinker, 1997, 1999; see also
Pinker, 2001).
Loasby says, “all knowledge is a structure of selected connections.” This
insight applies to individual minds. It also applies, if somewhat metaphorically, to
super-individual systems, whether orders or organizations (McQuade & Butos,
2004). Thus, “The creation, distribution and selective connection of domain-
specific modules within the economy is a central issue for explaining economic
development and for effective policy at the level of firms and governments.” In
other words, “Structures matter.”
Loasby shows that Hayek, Adam Smith, and Alfred Marshall all had
evolutionary theories, even though he interprets those theories in ways unfavorable
to “neoDarwinism.” Rubin and Gick compare Hayek’s work with modern
evolutionary theory. Hayek’s work holds up well, according to Rubin and Gick.
“Hayek got the basic outline of the problem right.” Nevertheless, “evolutionary
science and particularly its applications to human beings have advanced since
Hayek wrote, and some of the details of his analysis are no longer consistent with
current theories.”
Two such details may be worth mentioning here. First, Hayek wrote too early
to know of Boehm’s (1999) discovery of “reverse dominance hierarchies,” in
which, Rubin and Gick report, “subordinates cooperate to limit the power of
dominants.” The existence of reverse dominance hierarchies is clearly important
to our understanding of politics as illustrated by Rubin’s study in “Darwinian
Politics” (Rubin, 2002). Second, Hayek’s theory of cultural evolution was less
subtle and empirically grounded than that of Boyd and Richerson, whom Rubin
and Gick describe as “the major authorities on this issue.”
Some view the models of Boyd and Richerson as the current industry standard
for cultural evolution, including cultural group selection. Their most important
work might be Boyd and Richerson (1985). Recently, the Journal of Economic
Behavior and Organization devoted a special issue (Vol. 53, No. 1, 2004) to an
extensive symposium on a recent product of the Boyd and Richerson tradition,
namely, Henrich (2004). Participants included Axelrod, Gintis, Güth, Hodgson,
Houser, Khalil, Kliemt, McCabe, Sethi, V. Smith, Tullock, and D. S. Wilson.
I discuss this important class of models below. Given the importance of group
selection in Hayek’s work, this is a literature Austrians should engage. Douglas
Glen Whitman’s contribution to this volume is an example.
Evolutionary psychologists recognize that their perspective casts doubt on
“economic man” in at least some of its standard incarnations (Cosmides &
Tooby, 1984). Kurzban and Aktipis ask whether, indeed, homo economicus is
extinct. They survey experimental results that challenge the standard theory of
expected utility. They look at the work of two groups of scholars, namely, the
group around Vernon Smith and the group around Daniel Kahneman and the
late Amos Tversky. Their survey includes an evolutionary account of Kahneman
and Tversky’s prospect theory. They note, correctly in my view, that “under very
particular conditions, such as impersonal exchange,” homo economicus “is a
reasonable model.”
The contribution of Aktipis and Kurzban may suggest the need for a theory
that would help economists to know when to employ homo economicus and when
8 ROGER KOPPL
to employ other models of man. In this spirit, Glen Whitman and I (Koppl &
Whitman, 2004) have argued for a “rational-choice hermeneutics” that would
introduce a kind of realism to our models of economic actors, but without

abandoning the strong points of rational-choice theory. I have argued that homo
economicus is most suited to competitive markets under conditions of “atomism”
and “stability” (Koppl, 2002). I concur with the remark of Aktipis and Kurzban,
who favor “acknowledging that our actual species, Homo sapiens, is an amalgam
of Homo reciprocans, Homo prospectus, Homo nepotismis, and no doubt many
other interesting subcomponents.”
Viktor Vanberg also wants economists to forsake narrow models of rationality
in favor of a more empirically grounded understanding of human action (2004).
In his contribution to this volume, Vanberg argues that if we adopt an evolutionary
perspective and view action as “program-based behavior,” the Austrian doctrine of
methodological dualism must be abandoned. Vanberg rejects the dualist position
of Mises and of hermeneutical Austrians such as Lavoie and Boettke. He defines
methodological dualism as “the claim that the nature of its subject matter,
namely purposeful and intentional human action, requires economics to adopt a
methodology that is fundamentally different from the causal explanatory approach
of the natural sciences.” In Vanberg’s interpretation, methodological dualism
requires us to apply methods in the human sciences that are “fundamentally
different from the causal explanatory approach of the natural sciences.” In
this sense, Vanberg is surely right to claim that Hayek’s methodology violates
methodological dualism.
Vanberg’s criticism of hermeneutics is directed at the “universal hermeneutics”
of Gadamer and others. Vanberg correctly notes the reliance of most hermeneutical
Austrians on Heidegger and criticizes Lavoie’s “reinterpretation of what
‘subjectivism’ is about,” away from the view that meanings are internal to
individual minds and toward the view that they are inter-subjective. I share
Vanberg’s perplexity when he says “it is by no means easy to decipher what such
reorientation of subjectivism is exactly meant to be about.” The “hermeneutics” I
have defended is the more modest “classical hermeneutics,” which does not make
the extravagant and perplexing claims of universal hermeneutics (Koppl, 2002;

Koppl & Whitman, 2004). In its classical version, hermeneutics “claims only that
human actions can be understood in much the same way we understand a poem
or the instructions on a tube of toothpaste: we attempt to understand the purpose
of the act (whether written or otherwise) in terms of the internal perceptions and
beliefs of the person who performed it” (Koppl & Whitman, 2004).
Caldwell’s excellent recent intellectual biography of Hayek raises some
of the same issues Vanberg raises (Caldwell, 2004). In this important work,
Caldwell, like Vanberg, repudiates hermeneutics and holds that Hayek was
not a methodological dualist. I have already suggested that the repudiation of

hermeneutics is appropriate only for universal hermeneutics. Similarly, a more
mild definition of “methodological dualism” might be called for. If the doctrine is

meant only as a counter to extreme forms of reductionism, then Hayek probably
was a methodological dualist. Indeed, the final chapter of Hayek’s The Sensory
Order, “Philosophical Consequences,” can be read as a scientific defense of
methodological dualism in the mild sense of the term. Hayek defends “verstehende
psychology” and says we “will never be able to bridge the gap between the realm
of the mental and the realm of the physical” (1952, p. 192). For the “understanding
of human action” the “last determinates to which we can penetrate” are “familiar
mental entitities” not “physical facts” (1952, p. 192). This is not a rejection
of causal reasoning and is not, therefore, methodological dualism in Vanberg’s
sense. It is instead a perfect example of Vanberg’s claim that “Hayek advocates
what one may describe as ‘naturalistic subjectivism.’ ” Caldwell employed the
term “scientific subjectivism” to convey the same general idea (Caldwell, 1994).
In my view both terms are elegant expressions that capture vital aspects of
Hayek’s thought.
Regarding the existence of somewhat Kantian “categories” of thought, Hayek
said “in so far as we have been led into opposition to some of the theses traditionally
associated with empiricism, we have been led to their rejection not from an opposite
point of view, but on the contrary, by a more consistent and radical application of
its basic idea” (1952, p. 172). Similarly, Hayek was led to a rigorous defense of
“verstehende psychology,” not from an anti-scientific point of view, but on the
contrary, by a more consistent and radical application of scientific ideas to the
understanding of human action.
Hayek was a subtle thinker who found ways to consistently combine elements
of seemingly contradictory theories. Hayek’s treatment of empiricism and
rationalism is one example. His treatment of science and Verstehen is another.
His treatment of group selection and individualism is a third example. Outside
the orbit of Austrian economics, group selection is associated with somewhat
anti-individualist opinion. Political and methodological individualists tend to
reject group selection. Hayek might easily seem to have fallen into contradiction
by advocating both individualism and group selection. If Whitman is right,
however, there is no contradiction. In biology, the theory of group selection was
revised and reasserted by D. S. Wilson and Elliott Sober (see especially Sober
& Wilson, 1998). The rehabilitation of group selection has been a major event
in biology.
Wilson’s contribution to this volume draws conclusions for political economy
from his defense of group selection. If humans and other organisms exist in
groups, natural selection may operate on groups as well as individuals. If these
10 ROGER KOPPL
groups are subject to periodic dissolution and reformation, one of the results of
natural selection may be “altruism,” making sacrifices for others. (Biologically,
the organism’s actions raise the fitness of others and lower its own fitness.) In
humans altruism may be accompanied by feelings of solidarity. “Selfishness” may
be variously defined, but reasonable definitions should probably exclude the sort
of altruism that group selection may produce. But then, Wilson suggests, standard
economic notions of selfishness should probably be rejected.
If we take an evolutionary approach to economics and accept the elements
of Wilson’s “multi-level selection theory,” then, Wilson urges, we should give
up the individualism of standard theory. For many species, including us, groups
evolve into adaptive units and natural selection produces a psychology that fits
them to such groups. This psychology sometimes includes altruistic impulses.
Wilson cites Bowles and Gintis (1998), Henrich (2004), Hodgson (1993, 2001),
and Rubin (2002) as examples of the social theory that takes multi-level selection
seriously. Wilson claims, “multilevel selection theory is sufficiently general
as an accounting method for evolutionary change, and sufficiently focused on
the question of how groups evolve into adaptive units, to provide a theoretical
framework for the study of human society.”
Bowles (2004) is a recent synthesis of the sort of work Wilson is pointing
to. Austrian economists will be interested to note that Bowles makes several
favorable references to Hayek. In his highly Austrian treatment of the socialist
calculation debate, Bowles says, “by focusing attention on which institutions more
effectively utilize the information that is available [to economic actors], Hayek’s
paper (Hayek, 1948), like Mandeville’s Fable, counts as a landmark work in the
theory of economic institutions” (Bowles, 2004, p. 476, emphasis suppressed).
At the beginning of the Austrian movement thirty years ago, the profession was
mostly deaf to the very Austrian arguments that Bowles has now woven seamlessly
into his important new synthesis. This event marks an opportunity for Austrians
to make more frequent and substantial contributions to the mainstream economics
literature.
Whitman suggests a way for Austrians to retain their specific difference within
the profession while making contributions that non-Austrians may value. He

suggests that the evolutionary framework of Wilson and others can be integrated
with the “Schutzian methodological framework of Koppl and Whitman (2004).”
If this is possible, as I believe, then Austrians can contribute to literature in
the Boyd and Richerson tradition, for example, while explicitly drawing on the
Verstehen tradition of Wilhelm Dilthey, Max Weber, and Alfred Schutz. Drawing
on evolutionary psychology, I have outlined an evolutionary explanation for
“Verstehen,” defined as the “human ability to understand human meanings” (Koppl,
2002, p. 207). Recently, functional imaging studies, which use brain scans, such
as McCabe et al. (2001) are giving support to the view that humans have a
biologically programmed “theory of mind mechanism,” defined as “our ability to
explain and predict other people’s behaviour by attributing to them independent

mental states, such as beliefs and desires” (Gallagher & Frith, 2003, p. 77).
In the language of classical hermeneutics, this mechanism is our capacity for
“understanding.”
I believe Whitman’s paper is important, however, for a different reason.
Whitman gives us an argument that is sufficiently important and innovative to
warrant the symposium on it contained in this volume. Whitman shows that
methodological individualism and group selection are, as his title says, “compatible
and complementary.” This conclusion is probably uncomfortable for advocates
of both positions. The most common view in the secondary literature on Hayek
is probably that Hayek’s advocacy of group selection was a deviation from
methodological individualism. Whitman’s argument to the contrary is a major
statement of a claim that has, until now, been a distinctly minority opinion.
In their comment on Whitman, Elliott Sober and D. S. Wilson express
carefully qualified agreement. They give reasons other than group selection for
doubting that methodological individualism is an appropriate methodological
principle. They suggest that his argument may not carry implications for their
previous work.
In his comment on Whitman, Todd Zywicki suggests why Hayek, a
methodological and political individualist, would be led to group selection. In
Hayek’s theory, the objects filtered by group selection are rule sets. Contrary to
common opinion, the rules of just conduct were not chosen by powerful men.
They are “chosen” by an anonymous process of group selection. It makes sense,
therefore, to contrast the rule of law with the rule of men.
Whitman’s achievement is no meager thing. He has taken on in one paper two
issues that are separately the object of great controversy. As Richard Langlois
correctly says, Whitman argues that “methodological individualism is not what
you think it is.” Hayek scholars should take seriously Langlois’ comment on group
selection, as well. In current theories, group selection creates altruism and solves
the problem of too little solidarity. “For Hayek, ” however, “the problem is too
much group solidarity, not too little.”
Our prehistoric ancestors had a problem of too little solidarity. Group selection
solved this problem by selecting for altruism. With a relatively large fraction of
altruists in its population, biologically modern humans thrived as a species of
hunters and gatherers living principally in small loosely formed bands. Civilization,
however, required that we learn to live by a rather different set of rules. A different
process of cultural group selection sufficiently overrode our innate fear of strangers
to permit the evolution of an extensive division of labor.
12 ROGER KOPPL
Modern theories of multi-level selection do not always distinguish between

the evolutionary processes that formed our basic psychology, which is suited to
band-level society, and the evolutionary processes that later partially overrode
our innate psychology and fitted us for cooperation with strangers. Henrich, for
example, focuses on the problem of “prosocial emotions” from an imagined
baseline without them. He argues that “By systematically altering the social
environment in favor of prosocial phenotypes, cultural processes create the
conditions for natural selection to favor prosocial genes that could not otherwise
be favored in mammalian social species because non-humans lack the requisite
social learning capacities” (Henrich, 2004, p. 20).
Wilson recognizes the distinction between biological and cultural evolutionary
processes. Citing his recent book on religion (Wilson, 2002) he makes the Hayekian
claim that “Human history and current events can be regarded as an ongoing process
of cultural multilevel selection, whose dynamics are influenced by psychological
traits that evolved by genetic multilevel selection in the distant past.” It is probably
fair to say, however, that Sober and Wilson have not considered the Hayekian
problem of excess solidarity to which Langlois alerts us. As David Lahti notes,
“The thesis that morality did not evolve to curb the selfish course of human behavior
is perhaps counterintuitive” (Lahti, 2003, p. 648).
Austrian economists should consider adapting models in the Boyd and Richerson
tradition to the Hayekian problem of two-stage group selection. In so doing,
however, they should be aware of a possible difficulty not shared by the multi-
level selection theory of Sober and Wilson. Henrich and others argue that cultural
group selection and gene-culture interaction have produced prosocial emotions in
humans. These models depend on phenotypes such as “conformist transmission,”
which Henrich defines as “a psychological propensity to preferentially copy high
frequency behaviors” (2004, p. 22). Such mechanisms are supposed to help explain
“large-scale cooperation,” but large-scale cooperation requires a division of labor
and diversity in who does what. In this sense, we are not conformists in society, but
individualists. It is not clear in what dimensions we are supposed to be programmed
for conformity and in what dimensions we are psychologically free to innovate or,
at least, act out our unique roles in the social production process. It is not clear
what mechanism would keep us conformist in only the right dimensions. Perhaps
the answer has to do with language and ideology, but as far as I know, the issue
has not been raised in this literature.
Whitman’s colleague, Adam Gifford, gives an extended commentary on
the issues raised by Whitman’s paper. Gifford argues that methodological
individualism is consistent with the new group selection of Sober and Wilson,
but not with the sort of group selection model Hayek seems to have had in mind.
Gifford points out that Hayek’s model of group selection was not well worked out.
He calls it “folk” group selection, and notes that it “does not provide, apart from
positing the group as the unit of selection, much detail about the actual evolutionary
process involved.”
Gifford’s paper raises an issue for historians of economic thought. To what
extent are Hayek’s written comments on group selection consistent with the new
group selection and to what extent do they commit him to old-fashioned versions
of group selection, versions that have been definitively repudiated? Gifford’s paper
raises a second question as well. Can we substitute new group selection for old
group selection in Hayek’s system and come out with a consistent system? The
answer to the first question is probably that Hayek had in mind old group selection.
After all, he did not cite any literature in the new group selection, which begins
with Wilson (1975). The answer to the second question is probably yes, although
modifications of the new theory seem required, as I have suggested above. Austrian
economists should view the need for such modifications as an opportunity.
This volume places Austrian economics in the context of modern developments
in evolutionary psychology and, therefore, current debates about group selection,
altruism, and cultural evolution. These modern developments from outside
economics have been having an increasingly important influence on economics,
as Bowles (2004) illustrates. Indeed, some economists are involved in research
and debate in evolutionary biology and psychology (see, e.g. Robson, 2001). As
I have argued above, Austrians have a comparative advantage in bringing the
Verstehen tradition to evolutionary thinking. To do so, however, they will have to
study Hayek’s scientific defense of methodological dualism, which he first clearly
articulated in the concluding chapter of The Sensory Order (see Koppl, 2002,
pp. 68–69, 204–207). Today, Austrian economists have an opportunity to bring a
somewhat different set of evolutionary ideas to the table, a set derived largely from
the work of F. A. Hayek. They also have the opportunity to modify existing Austrian
theory with results from modern evolutionary theory. Thirty years ago, Austrian
economists were so many voices, crying in the wilderness. If exchange is the
foundation of society, the reciprocal exchange of ideas with modern evolutionary
thought promises to return Austrian economists from the wilderness to the society
of the sciences, natural and social.
NOTES
1. Maynard-Smith (1982, p. 2) suggests that the first contribution to evolutionary game
theory was Maynard-Smith and Price (1973). Note that this is the same Price of the Price
equation discussed below.
2. A Google search reveals only a department of “Human Psychology & Philosophy” at
the Russian Christian Institute for the Humanities, founded in 1989.
14 ROGER KOPPL
3. Roughly, you “naturalize consciousness” when you study the human mind as you
would study the behavior of insects or falling rocks. See Husserl (1965), especially pages
188–190.
4. I thank Robert Kurzban for alerting me to the problem at issue.
5. They thank William Allman for suggesting the phrase, “which is a very apt summary
of our position.”
ACKNOWLEDGMENTS
For helpful comments I thank Robert Kurzban, Brian Loasby, Thomas McQuade,
Glen Whitman, and Leland Yeager.
REFERENCES
Barkow, J., Cosmides, L., & Tooby, J. (1992). The adapted mind: Evolutionary psychology and the
generation of culture. New York and Oxford: Oxford University Press.
Boehm, C. (1999). Hierarchy in the forest: The evolution of egalitarian behavior. Cambridge: Harvard
University Press.
Bowles, S. (2004). Microeconomics: Behavior, institutions, and evolution. Princeton and Oxford:
Princeton University Press.
Bowles, S., & Gintis, H. (1998). Is equality passe? Boston Review, 23, 4–26.
Boyd, R., & Richerson, P. J. (1985). Culture and the evolutionary process. Chicago, IL: University of
Chicago Press.
Caldwell, B. (1994). Hayek’s scientific subjectivism. Economics and Philosophy, 10, 305–313.
Caldwell, B. (2004). Hayek’s challenge: An intellectual biography of F. A. Hayek. Chicago and London:
University of Chicago Press.
Cassill, D. (2002). Yoyo-bang: A risk aversion investment strategy by a perennial insect society.
Oecologia, 132, 150–158.
Cassill, D. (2003). Skew selection: Nature favors a trickle-down distribution of resources in ants.
J. Bioeconomics (in press).
Colander, D., Holt, R., & Rosser, J. B. (2004). Introduction. In: D. C. Colander, R. Holt & J. B. Rosser
(Eds), The Changing Face of Economics: Conversations with Cutting Edge Economists. Ann
Arbor: University of Michigan Press.
Cosmides, L., & Tooby, J. (1984). Better than rational: Evolutionary psychology and the invisible hand.
American Economic Review, 84, 327–332.
Cosmides, L., & Tooby, J. (1997). Evolutionary psychology: A primer. http://www.psych.ucsb.edu/
research/cep/primer.html.
Gallagher, H. L., & Frith, C. D. (2003). Functional imaging of ‘theory of mind’. Trends in Cognitive
Science, 7, 77–83.
Gigerenzer, G., & Selten, R. (2001). Bounded rationality: The adaptive toolbox. Cambridge, MA: MIT
Press.
Hayek, F. A. (1948). The use of knowledge in society. In: Individualism and Economic Order
(pp. 77–91). Chicago: University of Chicago Press.
Hayek, F. A. (1952). The sensory order. Chicago: University of Chicago Press.

Hayek, F. A. (1973). Law, legislation and liberty, Volume 1: Rules and order. Chicago: University of
Chicago Press.
Hayek, F. A. (1976). Law, legislation and liberty: The mirage of social justice (Vol. 2, pp. 153–176).
Chicago: University of Chicago Press, .
Hayek, F. A. (1978). The atavism of social justice. In: F. A. Hayek (Ed.), New Studies in Philosophy,
Politics, Economics and the History of Ideas (pp. 57–68). Chicago: University of Chicago Press.
Hayek, F. A. (1979). The three sources of human values. In: F. A. Hayek (Ed.), Law, Legislation and
Liberty: The Political Order of a Free People (Vol. 3, pp. 153–176). Chicago: University of
Chicago Press.
Henrich, J. (2004). Cultural group selection, coevolutionary processes and large-scale cooperation.
Journal of Economic Behavior and Organization, 53, 3–35.
Henrich, J., Rapporteur, W. A., Boyd, R., Gigerenzer, G., McCabe, K., Ockenfels, A., & Young, H. P.
(2001). Group report: What is the role of culture in bounded rationality? In: G. Gigerenzer &
R. Selten (Eds), Bounded Rationality: The Adaptive Toolbox (pp. 343–359). Cambridge, MA:
MIT Press.
Heyes, C. M. (2000). Evolutionary psychology in the round. In: C. M. Heyes & L. Huber (Eds), The
Evolution of Cognition. Cambridge, MA: MIT Press.
Hodgson, G. M. (1993). Economics and evolution. Cambridge, UK: Polity Press.
Hodgson, G. M. (2001). How economics forgot history. New Brunswick, NJ: Routledge.
Husserl, E. (1965). Phenomenology and the crisis of philosophy. New York: Harper & Row.
Karmiloff-Smith, A. (2002). Elementary, my dear Watson, the clue is in the genes . . . or is it? In:
Proceedings of the British Academy (Vol. 117, pp. 525–543). Oxford University Press for the
British Academy, Oxford.
Koppl, R. (2002). Big player and the economic theory of expectations. London and New York: Palgrave
Macmillan.
Koppl, R., & Whitman, D. G. (2004). Rational-choice hermeneutics. Journal of Economic Behavior
and Organization (forthcoming).
Lachmann, L. (1977). Austrian economics in the present crisis of economic thought. In: W. Grinder &
L. Lachmann (Eds), Capital, Expectations, and the Market Process: Essays on the Theory of
the Market Economy (pp. 25–41). Kansas City, KS: Sheed Andrew and Mcmeel.
Lahti, D. (2003). Parting with illusions in evolutionary ethics. Biology and Philosophy, 18, 639–651.
Maynard-Smith, J. (1982). Evolution and the theory of games. Cambridge: Cambridge University Press.
Maynard-Smith, J., & Price, G. R. (1973). The logic of animal conflict. Nature, 246, 15–18.
McCabe, K. D., Houser, L. R., Smith, V., & Trouard, T. (2001). A functional imaging study of
cooperation in two-person reciprocal exchange. Proceedings of the National Academy of
Sciences, 98, 11832–11835.
McQuade, T., & Butos, W. (2004). The sensory order and other adaptive classifying systems. Journal
of Bioeconomics, 6, 1–24.
Mises, L. (1949). Human action: A treatise on economics. New Haven, CT: Yale University Press.
Pinker, S. (1997). How the mind works. New York: W. W. Norton.
Pinker, S. (1999). Words and rules: The ingredients of language. New York: Basic Books.
Pinker, S. (2001). Talk of genetics and vice versa. Nature, 413, 465–466.
Robson, A. J. (2001). The biological basis of economic behavior. Journal of Economic Literature, 39,
11–33.
Rubin, P. H. (2002), Darwinian politics: The evolutionary origin of freedom. New Brunswick, NJ:
Rutgers University Press.
16 ROGER KOPPL
Sober, E., & Wilson, D. S. (1998). Unto others: The evolution and psychology of unselfish behavior.
Cambridge, MA and London: Harvard University Press.
Symons, D. (1992). On the use and misuse of Darwinism in the study of human behavior. In: J. Barkow,
L. Cosmides & J. Tooby (Eds), The Adapted Mind: Evolutionary Psychology and the Generation
of Culture (pp. 137–159). New York and Oxford: Oxford University Press.
Todd, P. M. (2001). Fast and frugal heuristics for environmentally bounded minds. In: G. Gigerenzer
& R. Selten (Eds), Bounded Rationality: The Adaptive Toolbox (pp. 51–70). Cambridge, MA:
MIT Press.
Tooby, J., & Cosmides, L. (1992). The psychological foundations of culture. In: J. Barkow, L. Cosmides
& J. Tooby (Eds), The Adapted Mind: Evolutionary Psychology and the Generation of Culture
(pp. 19–136). New York and Oxford: Oxford University Press.
Vanberg, V. (2004). The rationality postulate vs. program-based behavior – alternative theoretical
approaches and their relevance for the study of instutions. Rationality and Society, 14, 7–54.
Wilson, D. S. (1975). A theory of group selection. Proceedings of the National Academy of Sciences,
72, 143–146.
Wilson, D. S. (2002). Darwin’s cathedral: Evolution, religion and the nature of society. Chicago:
ECONOMICS AND EVOLUTIONARY

PSYCHOLOGY!
David Friedman
Economics is built on a simple assumption – that individual behavior can best be

predicted by assuming that each individual will take the actions that best achieve his
objectives. The justification for that assumption, somewhat misleadingly labeled
“rationality,” is that we have no good theory of mistakes, no way of predicting
what particular irrational action an individual will take. That leaves the rational
element as the best – although imperfect – way of predicting behavior.1
Evolutionary psychology offers, among other things, a theory of mistakes – an
alternative to the rationality assumption. In this essay I sketch out the nature of that
theory, describe some puzzles that economics has a difficult time explaining, and
try to show how modifying economics with the aid of evolutionary psychology
might help explain them.
EVOLUTIONARY PSYCHOLOGY: THE SHORT VERSION

Evolutionary psychology2 starts from two simple assumptions:
The human mind is best understood not as a general purpose computer but as a set of specialized
software modules, each designed to deal with a particular subset of problems.
! An earlier version of this article was published, in English, Catalan, and Castilian in the 4/2001
issue of the webbed journal Indret. It is a first pass at a large problem. I hope others will find it
sufficiently interesting to want to both extend the analysis to other economic puzzles and sharpen the
tests it generates so as to provide something more like evidence and less like anecdote to help us
choose between the alternative approaches to human behavior offered by economics and evolutionary
psychology.

© 2004 Published by Elsevier Ltd.
ISSN: 1529-2134/doi:10.1016/S1529-2134(04)07002-4
17
18 DAVID FRIEDMAN
Those programs have been designed by Darwinian evolution to produce reproductive success
in our environment of evolutionary adaptiveness – the hunter-gatherer environment in which
our species spent most of its species history.
Researchers in evolutionary psychology, starting with these assumptions, have

generated and tested predictions ranging from differences in male and female
special abilities to the timing of morning sickness.
Three important points are worth making about the second assumption in order
to avoid misunderstanding. The first is that the assumption is not that individuals
seek reproductive success – if we were doing that, the population of developed
countries would be increasing much faster than it is – but only that we have
those psychological characteristics that produced reproductive success in the
environment we evolved in. The second is that reproductive success is an objective
for the individual, not the group or species. Most scholars in evolutionary biology
accept the view that traits which benefit group or species at the cost of the individual
who carries them will be selected out.3
The third point is that we are adapted not to the world we now live in but to
the environment in which our species spent most of its history. Agriculture is a
recent development. We would expect most of our characteristics to be designed to
produce reproductive success in the environment in which our species spent most
of its evolutionary history – believed to be an environment of small hunter-gatherer
bands.4
WHAT EVOLUTIONARY PSYCHOLOGY ADDS

Evolutionary psychology adds two modifications to the rationality assumption.
The first is an increase in its precision. Economists assume that individuals have
objectives. But economic theory does not tell us what those objectives are, although
observation and introspection provide at least a rough idea of what they are likely
to be. Evolutionary biologists, on the other hand, know the objective of genes5 –
reproductive success or, more precisely, inclusive fitness, getting as many copies
of themselves as possible into future generations.6

It follows that the assumptions of evolutionary psychology give economists
some ability to predict what utility functions individuals will have. That ability is
limited by our ignorance of the opportunity sets facing the genes – what sorts of
organisms it is possible for them to construct. If, for example, there were a way of
constructing a (philoprogenitive) superman, a human being much stronger, faster,
healthier, smarter, than existing humans and capable of surviving on practically
anything, the gene that pulled off the trick would be a big winner in the Darwinian
Economics and Evolutionary Psychology 19
race. The absence of such supermen suggests that it cannot be done, or at least that
doing it is so difficult that no gene has yet had managed it.
A less obvious example of the same point is the observed limit to how
philoprogenitive real human beings are. We are designed to seek reproductive
success through a variety of traits – desire for sex that leads us to reproduce,
parental love that leads us to care for our offspring, and many others. But despite
those traits human beings, in the environments of recent centuries, produce far
fewer children than they could produce and successfully rear – in part because we
have found ways, ranging from birth control to pets, to sabotage the objectives
of our genes in order to better serve our own objectives.7 A true philoprogenitive
gene, one that made reproductive success a high priority of every individual,
would confer an enormous reproductive advantage on its carriers and rapidly
spread through the population.8 The absence of such a gene is presumably due
to the difficulty of such precise programming of an organism as complicated
as a human being, plus the short time that has passed since the developments
that make that tactic for reproductive success so much superior to less direct
approaches.
Knowing the objective of our genes is not sufficient to tell us, with confidence,
the objectives of the human beings that those genes build. But it is enough to
suggest hypotheses – characteristics that would lead to increased reproductive
success and that it might be possible for genes to give to the organisms they
construct. Having formed such hypotheses, we can then test them by comparing
their predictions to what we observe. That is a methodology routinely used in
evolutionary biology – including, but not limited to, evolutionary psychology. It is
the same as the methodology of positive economics save for a different and more
explicit procedure for forming hypotheses.
One way in which evolutionary psychology modifies the rationality assumption
is by predicting what objectives individuals are likely to have. A second is by
providing a theory of mistakes.
Compared to rational thinking, Darwinian evolution is a slow process. While
we expect economic man to choose the actions that achieve his objectives in the
environment he observes around him, we expect evolutionary biological man to

be designed to achieve his objectives – more precisely, his genes’ objectives –
in the environment in which his species evolved. It follows that individuals are
likely to be irrational – designed to act in ways not well designed to achieve their
objectives – when the relevant features of the environment have changed rapidly
enough so that evolution has not yet had time to catch up. The theory predicts
not merely that individuals will make mistakes – that we already knew – but what
mistakes they will make. They will make those mistakes that would have led to
20 DAVID FRIEDMAN
reproductive success in the environment in which the psychological characteristics

leading to those mistakes evolved.9
ECONOMIC PUZZLES
Economists sometimes observe people acting in ways that appear difficult or

impossible to make consistent with the economic approach to understanding
behavior. In this section I first consider a group of such economic puzzles that
I believe can all be explained by a common characteristic of human psychology
– the belief in just prices – itself explainable on evolutionary grounds. I then go
on to consider two more puzzles – inconsistent time preferences and endowment
effects – each of which I argue has an evolutionary explanation.
Behavioral Consequences of the Belief in Just Prices
The first puzzle is the existence of predictable lines. Consider a restaurant whose
patrons know that if they come for dinner on Friday or Saturday they will have
to wait forty-five minutes for a table. The line does not increase the number of
people the restaurant can serve10 but does impose an additional cost on customers
in waiting time, raising the total cost of the meal enough to reduce quantity of
meals demanded to the quantity the restaurant is capable of producing.
Suppose the wait is the equivalent, from the standpoint of the customers,11 to a
ten-dollar increase in price. If the restaurant simply raised its price for the nights
it was busy by ten dollars the line would shrink to close to zero. Customers would
be no worse off – they would be paying the extra price in money instead of time –
and the restaurant would be better off by ten dollars per diner. In the longer run,
the increase in the amount restaurants could charge on busy nights would increase
the supply of restaurants, bringing down the price and transferring at least some
of the benefit back to the customers.
Restaurants do, to some extent, vary their price in this way – usually by
announcing special discounts for low-demand nights rather than special surcharges
for high-demand nights.12 Nonetheless, predictable long lines are a familiar feature
of the restaurant world, which suggests some significant constraint limiting the
degree to which they can vary their prices. A similar pattern is observed in other
contexts – concerts, opening nights of popular films, and the like. Producers
frequently follow pricing polices that lead to wasteful competition for under-priced
goods. Doing so appears to make the producer worse off, contrary to what we would
expect from the assumption of rationality.
One response consistent with casual observation is that a rock group or movie
theater that routinely charged a price sufficient to ration demand down to supply
for high-demand events would offend its customers and thus lose more in the long
run than it gained in the short. But this explanation raises a second puzzle – the
behavior of the customers. The average customer is no worse off in the short run
as a result of such a policy, since it merely converts cost in time into cost in cash.
And he is better off in the long run. So why should customers be offended? Why
should they choose not to patronize producers who price their goods in the way
that economic theory suggests they should?13
A second and less striking puzzle is why firms that sell the same product at
different prices at different times almost invariably describe their policy as a normal
price and a discount rather than a normal price and a surcharge. This is less puzzling
than the existence of predictable lines because there is no strong argument against
doing it that way. But there is no strong argument for, either. Given that we already
know what firms do, it is easy to argue that doing it that way makes them look
good – they are giving their customers a special deal. But one could just as easily
argue that the alternative policy makes them look good because it implies a lower
ordinary price.
Next consider the history of price control. A law fixing a legal price below
the price that would emerge from the market has a variety of consequences. In
the long run, it is likely to make almost everyone worse off.14 In the short run,
however, it may well benefit purchasers at the expense of producers – and for some
products, most notably rental housing, the short run may be long enough to produce
substantial transfers. Voters are, for familiar reasons, rationally ignorant, and long-
run costs are often less obvious than short-run benefits. So it is not astonishing that
imposing price control is sometimes politically profitable.
What is surprising is the pattern of when it is politically profitable. The costs and
benefits of holding a price 10% below its market level do not depend on whether
the restriction prevents a price rise or forces a price reduction.15 Yet the former case
is very much more common than the latter. Price control almost always arises in a
context where it is intended to prevent prices from rising, very rarely in a context
where a price is stable and the control is designed to push it below its current level.
All of these puzzles can be explained by a single assumption: Individuals believe
that the proper price for a good is the price at which they are used to buying the
good, resent being charged more than that price, and therefore attempt to punish
those charging the higher price. That assumption is consistent both with casual
observation of reactions to sharp price increases and with the history of ideas such
as the scholastic philosophers’ doctrine of the just price.16
Seen from the standpoint of economic rationality, the assumption makes little
sense. Most people have no clear idea what determines the prices of the goods
22 DAVID FRIEDMAN
they buy, so no way of knowing whether yesterday’s price was fair, or just, or cost
justified, or whether today’s higher price fails any of those criteria.17
The behavior associated with the belief – the attempt to punish those charging
“unjust” prices-also makes little sense. Suppose I really do know that a particular
price is in some meaningful sense too high – say higher than the economically
efficient price. Why is that a reason for me not to buy the good, assuming it is
still worth more to me than it costs? Why is it a reason for me to be angry at the
seller and express that anger by avoiding future transactions with him even if they
would benefit me? In a world where goods and services are sold to large numbers
of anonymous customers I cannot reasonably expect my refusal to buy, however
justified, to induce the seller to lower his price.
Evolutionary Psychology and the Just Price
Now shift the analysis back twenty thousand years. As a member of a hunter-
gatherer band, you engage in a variety of transactions with your fellows, trading
goods and services – food, sex, support in intra-group conflict, and the like. While
money has not yet been invented, prices – the amount of food you must give in
exchange for sex, the favors you must do someone if you want him to do a favor
for you – are a familiar part of your environment. In this world all markets are thin
– it is, after all, a small band – so the typical transaction is a bilaterial monopoly
bargain.
Assume an environment sufficiently stable so that, for some transactions, there
are “usual prices.” Those prices must be within the bargaining ranges of both buyer
and seller,18 since otherwise the transactions would not occur. The environment
is not, however, perfectly stable. Sometimes the circumstances of one party or
another shift his bargaining range – the range of terms for which the transaction
is in his interest.
You are a buyer whose current circumstances make the good much more
valuable to you than usual, widening the bargaining range. If you could somehow
commit yourself not to pay more than the usual price, you, rather than the seller,
would get the increased benefit from this transaction. One way to do so is to be
emotionally programmed to resent any increase above the usual price – resent it
enough so that the humiliation of being “cheated” will outweigh the gain from the
transaction.
As in any bilateral monopoly game, the argument works both ways. If the seller
could somehow commit himself not to accept less than your reservation price,19 he
would be the one to pocket the gains from the trade. There is, however, an important
difference between your situations. You know the usual price and, assuming the
special circumstances affect only you, know that it is probably within the seller’s
bargaining range. So your commitment strategy is unlikely to commit you to a price
outside the bargaining range – which would make the transaction impossible. The
seller does not know your reservation price, so if he commits himself to his guess
at what you are willing to pay he may choose a price at which the transaction can
not occur.
A second difference is that the usual price is common knowledge of both parties.
If the seller believes that the buyer is committed not to pay more than the usual
price, he knows how much he can ask. But if the buyer believes that the seller is
committed not to accept less than the buyer’s reservation price, he still has to guess
what the seller believes that is.
The strategy works symmetrically when the seller is, for special reasons, willing
to accept a much lower price than usual – on a barter market, this is the same
situation seen from the other side. And, returning briefly to the present, we observe
that people resent not only unusually high prices when they are buyers but unusually
low prices when they are sellers – giving us, among other things, one explanation
for why wages are sticky downwards.
What about the situation where the seller’s costs are unusually high, making
him unwilling to sell at the usual price? If the result is to eliminate the bargaining
range, no transaction will or should take place. But if the seller’s cost is lower
than the value to the buyer, either because the special circumstances affect both
in similar ways or because the increased cost is still within the usual bargaining
range, a buyer’s commitment not to pay more than the usual price results in an
inefficient bargaining breakdown.
There is a solution to this problem. A seller charging an unusually high price
can defend himself against the buyer’s commitment strategy by offering to show
the buyer that his costs really are unusually high, that he is really, and not only
strategically, unwilling to sell at the usual price. From that we get the conventional
view of pricing that economists find so frustrating and wrongheaded – as the
outcome of bargaining between buyer and seller, with each required to justify any
deviation from past prices.
It follows that, in the context of a hunter-gatherer society (more generally, a

society where most transactions take place on thin markets) belief in a just price
– defined as the usual price – can be understood as a commitment strategy that
benefits those who adopt it. The benefit depends on an environment sufficiently
stable so that knowledge of past prices provides a simple rule for identifying a
price that is probably within the other party’s bargaining range. It works better
if bargainers can to some degree identify situations where the rule would result
in one party demanding a price outside of the bargaining range and treat them as
special cases.
24 DAVID FRIEDMAN
We now have a possible solution to the set of puzzles described above. Human
beings have a functional module in their minds that deals with exchanges with
other human beings. One feature of that module, hard-wired in by evolution, is

that human beings regard the usual terms of exchange as right and any deviation
from those terms that makes them worse off as a presumptively wicked act by the
other party. This feature resulted in human beings that possessed it getting better
terms in bilateral monopoly bargains in the environment in which we evolved, so
having more resources and achieving greater reproductive success.
The belief in just prices and the associated commitment strategy continue to
benefit those who hold it under some circumstances – markets with the features
I have described. They injure those who hold it in the anonymous mass-market
settings in which most modern economic transactions occur. And the fact that
other people have that characteristic makes me worse off, both because it leads
to inefficient policies by firms (long lines) and governments (price control) and
because it leads to those others getting better terms from me when we happen to
be in a bilateral monopoly transaction.
Primitives, Prices and Exchange
I began the previous section with the claim that hunter gatherers are familiar
with prices and exchange. The obvious context is individual trade – usually
and somewhat confusing described in the evolutionary literature as reciprocal
altruism.20 Thus, for example, we observe a pattern among both humans and
baboons where one individual assists another in intragroup conflict in exchange
for later assistance in the other direction.21 The existence of such exchanges implies
an implicit price – how much assistance must be given in exchange in order that
one party not regard the other as cheating on the relevant social rule. Similarly,
the exchange of food for sexual favors has been observed among both humans and
closely related primates.22
For a less obvious form of exchange, consider meat sharing among the Ache of
eastern Paraguay, who were until very recently full time nomadic hunter-gatherers
and still engage in extended foraging trips. Hunting is done by the men. The
individual hunter who kills a particular animal has no ownership rights over it; the
cooked animal is shared out among the group. On the face of it, this looks like
a situation of pure communism – individual production, communal consumption,
no exchange.
Anthropologists Kim Hill and Hillard Kaplan, in their study of the Ache,23
discovered an interesting pattern: Individuals were identified as good or bad hunters
and good hunters had substantially greater reproductive success, more surviving
offspring, than bad hunters. Their explanation was that, despite the apparent
egalitarianism of the sharing of meat – which provided the bulk of the calories
consumed – good hunters were still rewarded. Membership in foraging bands was
fluid. In order to make sure that the good hunters went with a particular band it
was necessary for other band members to offer them rewards for doing so – in the
form of better care for their children, more sexual access to women in the band,
and the like.
While the individual foraging band lacks the formal hierarchical structure of
a modern corporation, the situation is in other respects one familiar to us. Most
employees in a modern society, like hunters in an Ache foraging band, have no
ownership claim over the particular goods or services they produce. Their reward
still depends on their productivity, but through a more indirect mechanism. Firms
that pay employees less than they are worth risk losing them to other firms. Foraging
bands that under reward good hunters may find that, next time around, the good
hunters go with other bands. Just as in the case of individual exchange, there is an
implicit price, just one linked to the services provided rather than directly to the
goods produced.
To see how the logic of just prices might play itself out in this situation, consider
the formation of a foraging band at a time when it happens that no other bands are
being formed. Members of the band could propose to good hunters they wished to
recruit that they be satisfied with the same treatment as everyone else – that being
preferable to staying home. Good hunters committed to a norm of “just payment”
would decline that proposal – and if they were known to be committed to that norm
would end up with their usual favorable treatment.
A Bird in the Hand: Evolutionary Psychology and Preferences Over Time
Intertemporal choice provides a second example of puzzling behavior that may

perhaps be explained by evolutionary psychology. The usual economic model,
going back at least to Marshall24 and given its present form by Samuelson,25
assumes that an individual’s preference between present and future utility can be
described by a discount rate. The preferred alternative is the one that maximizes
the present value of utility, discounted at that rate. Mathematically:
U1 + U2 U3
U= + +···
(1 + r) (1 + r)2
where r is the individual discount rate on utility, U is the utility that the individual
maximizes, U1 is the utility received in year 1, U2 in year 2, and so forth.
26 DAVID FRIEDMAN
While it is not clear that this particular structure is required by rationality, it

does have some attractive properties. In particular, it guarantees that the preference
between two future alternatives at different dates does not change as we approach
them; individual choice is, in that sense, consistent over time.
While this seems a plausible description of rational choice, it does not appear
to describe real world behavior.26 A variety of experiments show that many
individuals, faced with (say) the choice between a thousand dollars today or eleven
hundred dollars next week will prefer the former, yet faced with the choice between
a thousand dollars a year from now and eleven hundred dollars a year plus a week
from now will prefer the latter. The usual pattern appears to be a very high discount
rate for choices in the near future and an increasingly low discount rate as the
alternatives become more distant.27
Evolutionary psychology suggests a straightforward explanation for such a
pattern. In the experimental setting, subjects are told that they are choosing between
two certain payoffs at different times. In the world in which we now live, that is
a believable story; modern financial institutions make possible secure promises
of future performance, such as those embodied in certificates of deposit or U.S.
government bonds.
The world in which our species evolved did not have such institutions. In that
world it was rational to discount promises of future performance. One meal today
was worth a great deal more than one meal next week because today’s meal was
there to be taken; next week’s might not be. One meal a year from now, on the
other hand, was not worth much more than one meal a year plus a week from now;
both were promises that might well fall through, and the chance of their falling
through was not greatly altered by the additional week.28
Not only was the behavior rational twenty thousand years ago, it was, to a
considerable degree, rational a good deal more recently, sufficiently so to become
proverbial. “A bird in the hand is worth two in the bush.”29
ENDOWMENT EFFECTS
A professor purchases lots of school mugs, selects at random half the members of
a group of students, and gives one to each of them. He then asks each student with
a mug to state the price at which he would be willing to sell it and each student
without a mug to state the price he would be willing to pay for one. Finally, he
calculates the market clearing price – the price at which there is exactly one seller
for each buyer – and reallocates mugs and money accordingly.
At the end of this process, the mugs should be in the possession of whichever
students most value them. Since they were originally handed out at random, we
would expect that about half of the students who most valued them would have
gotten them and half would not, hence that about half of the mugs should change
hands.
In fact, almost none of them did. The conclusion from this classic experiment30
was that, on average, people value a mug more when they have it than when they
do not – the lowest price at which someone who has a mug will sell it is, on
average, higher than the highest price that someone who does not have a mug will
be willing to pay to buy it. That was supported by the actual prices the students
gave. The median seller required about twice as much to be willing to sell as the
median buyer was willing to offer.
This result – applied to mugs and much else – is known as an endowment
effect. On average, for many but not all sorts of things, someone who owns
something values it more than someone who does not. While the results of a single
experiment along these lines can usually be explained away as due to something
else, the experiment has been repeated enough times in enough different ways
to justify considerable confidence in the conclusion. One thus has the apparently
paradoxical result that someone choosing between one package containing a
mug and ten dollars and another containing no mug but fourteen dollars reveals
inconsistent preferences – whichever package he starts with, he prefers it to
the alternative.
The explanation of this pattern of behavior starts with the observation that it is
not limited to humans.
Territorial Behavior
It has long been known that some species of animals exhibit territorial behavior.
An individual fish, bird, or mammal in some way claims a particular territory for
his own and attacks other members of his species that trespass on it. Even if the
trespasser is somewhat larger and stronger than the claimant, the claimant usually
wins such conflicts – at some point the trespasser retreats.31
The logic of the situation is straightforward. Unless the trespasser is much

stronger a fight to the death is a losing game for both parties, since even the winner
risks substantial injury. The claimant has somehow committed himself to fight more
fiercely the closer the trespasser gets to the center of the territory. The trespasser,
recognizing that commitment, eventually backs down and retreats. Presumably
the commitment is accomplished through a behavior pattern hard-wired into the
psychology of a territorial species.
A different way of putting this is that territorial animals exhibit an endowment
effect – each individual will fight much harder to keep his territory than he will
28 DAVID FRIEDMAN
to conquer someone else’s territory. The effect is not limited to real estate. It is a
familiar observation that a dog will fight harder to keep his own bone than to take
another dog’s bone.

Now consider the same logic in a hunter-gatherer society with no external
institutions to enforce property rights. Imagine that each individual considers
every object in sight, decides how much each is worth to him, and then tries to
appropriate it, with the outcome of the resulting Hobbesian struggle determined by
some combination of how much each wants things and how strong each individual
is. It does not look like a formula for a successful society, even on the scale of a
hunter-gatherer band.
There is an alternative solution, assuming that humans are at least as smart
as dogs, robins, and Siamese fighting fish. Some method, possibly as simple as
physical possession, is used to define what belongs to whom. Each individual
commits himself to fight very hard to protect his property – much harder than he
would be willing to fight in order to appropriate a similar object from someone
else’s possession – with the commitment made by some psychological mechanism
hard-wired into humans. The result is both a lower level of (risky) violence and a
more prosperous society.
The fact that the result is attractive does not guarantee that it will occur –
evolution selects for the (reproductive) interest of the individual, not the group.
But in this case they are the same. To see that, imagine a population in which some
individuals have adopted the commitment strategy described above, some have
adopted no commitment strategy, and some have adopted different commitment
strategies – for example, a strategy of fighting to the death for whatever they see
as valuable. It should be fairly easy to see that individuals in the first group will,
on average, do better for themselves – hence have (among other things) greater
reproductive success – than those in the second and third.
How do I commit myself to fight very hard for something? One way is by
perceiving it as very valuable. So the same behavior pattern that shows up as
territorial behavior in fish and ferocious defense of bones in dogs shows up in
Cornell students as an endowment effect. Just as in the earlier cases, behavior that
was functional in the environment in which we evolved continues to be observed,

even in a context where it now serves no useful purpose.32
FILLING IN THE UTILITY FUNCTION

So far I have been discussing the use of evolutionary psychology to solve puzzles –
explain situations where the conclusions of economic analysis appear inconsistent
with observed behavior. Evolutionary psychology can also be used to make sense
of behavior which, while not inconsistent with conventional economics, is also

not implied by it – to explain why the individual utility function has charcteristics
needed to explain observed behavior.

One obvious example is parental altruism towards children. In some
environments it makes sense as a means to narrowly self-interested ends –
productive children are better able to take care of their parents in their old age
in a society where that is the chief form of old age insurance. But the behavior
appears more general than that – as we would expect if our utility functions were
shaped by evolution to maximize reproductive success. Children who die young
do not produce grandchildren; children who grow up to be able and productive
individuals, at least in most past societies, can and do produce and rear more
grandchildren as a result.
A less obvious example is concern with status. As Robert Frank has convincingly
argued,33 humans appear to care about both real income and relative income. While
a concern with relative income is not inconsistent with economic rationality, neither
is it implied by it. My ability to get most of what I want depends on how much my
income can buy, not on how much yours can.
In a hunter gatherer band, however, there is one resource which is in fixed supply
and of enormous importance for male reproductive success – women. How many
children I can feed depends on my real income. But my ability to persuade one
or more women to produce children with me depends on my resources – material
and otherwise – relative to those of the other men against whom I am competing.
Similarly, the ability of a woman to persuade a man to produce children with her
and help support them depends in part on her status vis-a-vis the other women on
whose children that man might spend his limited resources. So we would expect
both relative status and real income to play important roles in the individual utility
function produced by evolutionary selection.
This explanation has an interesting implication. If it is correct, men ought to
be primarily concerned with their status relative to other men, women with their
status relative to other women. I do not know whether or not that prediction is
empirically confirmed.
CONCLUSION
I have offered three examples of patterns of behavior apparently inconsistent with
the usual account of economic rationality and shown how each may be explained,
perhaps even predicted, by evolutionary psychology. Behavior associated with
deeply held beliefs about just prices makes sense as a hard-wired commitment
strategy designed to give its holder an advantage in the bilateral monopoly bargains
30 DAVID FRIEDMAN
that must have been common in the hunter-gatherer societies in which our species
evolved. A pattern of apparently inconsistent choices over time makes sense as
reflecting the lack of reliable mechanisms for guaranteeing the performance of

future obligations in that same society. Endowment effects are the predictable
result of commitment strategies that make possible a system of property in a world
without public law enforcement.
In all three cases, evolutionary psychology plays the second of the two roles that
I described at the beginning of this article – it functions as a theory of mistakes.
That is not surprising. Insofar as evolutionary psychology tells us what our utility
function is, it improves upon the economic model of rationality but does not
contradict it. It is only when it tells us what actions we will take that do not
serve our interests that it contradicts the conventional approach and so makes it
possible to find evidence for one and against the other.
NOTES
1. The argument is summarized in Friedman (1996, pp. 3–9). A webbed version can
be found in Friedman (1990), Price Theory, http://www.daviddfriedman.com/Academic/
Price Theory/PThy Chapter 1/PThy CHAP 1.html.
2. For a description and defense of evolutionary psychology, along with a number of
interesting applications, The Adapted Mind see Barkow, Cosmides, and Tooby (1992). For
a discussion of its relevance to social sciences in general, including economics, see Zywicki
(2000).
3. It is logically possible for traits that benefit a group at the expense of the group
members who carry them to produce an increase in success for the group large enough to
make up for the decreased success of the member relative to other members of the group,
but that is usually viewed as an unusual special case.
4. Owen Jones uses the term “time shifted rationality” to express this idea.
5. “Objective of genes” is, of course, a metaphor. Genes do not have minds, hence do
not have objectives. But the organisms we observe are constructed by those genes that
succeeded, in past generations, in constructing organisms that got those genes passed
down. Hence genes are shaped, as by an invisible hand, to construct organisms whose
characteristics result in reproductive success for the genes that constructed them.
6. “Inclusive fitness” includes both increasing the frequency of your genes through your
own reproduction and increasing their frequency by aiding the survival and reproduction of
kin, who share some of your genes.
7. A point dramatically made by Richard Dawkins (1990); he describes human beings as
the real world equivalent of science fiction robots who have revolted against their makers.
The clearest evidence that couples in modern societies could produce and successfully rear
substantially more than two children is that many do so, despite having no more resources
available to them than the much larger number who do not.
8. Reproductive success includes both producing offspring and successfully rearing them
– successfully enough to give them, in turn, the opportunity for successful reproduction.
While it is an important way of increasing the frequency of your genes in future generations
it is not the only way, so would not be the sole objective of an organism perfectly designed
for that purpose. My full siblings are as closely related to me as my children, so keeping a
brother or sister alive and able to reproduce is another way of increasing the frequency of
my genes in future population, hence of achieving inclusive fitness.
In developed societies most couples have about two children but could produce and
successfully rear eight – if doing so were their highest priority – as demonstrated by some
who do. A gene that resulted in its carrier doing so would increase its frequency in the
population a billion fold in a mere fifteen generations.
9. Readers interested in a much longer and more detailed exposition of the basic ideas of
this and the preceding section of this essay – what evolutionary psychology is and what it
implies about rationality – will find it in Jones (2001, pp. 1161–1173). The later parts of the
article in part overlap with ideas of this chapter as well as offering additional applications
of the evolutionary approach, in particular ones relevant to the design of legal institutions.
10. A short line would increase the number served, since it provides an inventory of
customers that allows the restaurant to produce at full capacity in the face of unpredictable
demand. My analysis is of the effect of any additional wait beyond that necessary to achieve
that effect.
11. Patrons are not identical, but for my present purposes it is not necessary to discuss
the differing effects on patrons with differing value for time.
12. This pattern fits the explanation I propose below.
13. There have been a variety of attempts by economists to explain lines within a
conventional economic framework, none of which appears to me entirely satisfactory. See,
for instance, Becker (1991).
14. For an analysis of reasons this is true, see (Friedman, 1990) Price Theory Chap. 17
(webbed at http://www.daviddfriedman.com/Academic/Price Theory/PThy Chapter 17/
PThy Chapter 17.html) and Hidden Order Chap. 17.
15. This is a slight oversimplification; there are costs associated with the mechanics of
pricing, such as revising price tags and prices in advertisements, which are incurred if price
control forces prices down but avoided if it prevents prices from going up.
16. For a discussion of that doctrine and the function it served, along lines related to the
argument of this essay, see Friedman (1991) “In Defense of Thomas Aquinas and the Just
Price.” For some evidence of similar attitudes in modern consumers, see Richard H. Thaler,
Mental Accounting and Consumer Choice, Marketing Science, 1985. For a more extensive
discussion of modern behavior patterns inconsistent with the usual economic account of
rationality, along with a good deal of evidence, see Quasi-Rational Economics by Richard
H. Thaler.
17. “For one thing, the recent tripling of oil prices followed a sharp drop. In real terms,
prices are still one-third below their level in 1990, when Iraq invaded Kuwait, and half their
level in 1981.” The Economist, Sept. 9, 2000, p. 17.
18. Strictly speaking the distinction is meaningless here, since we are talking about barter
transactions, but it is still useful for expository purposes.
19. More realistically, to insist on a price a little below your reservation price, so as to
reduce risks from error. Since the distinction is not important to my general argument I will
ignore it below in order to simplify the discussion.
20. The original article is Trivers (1971). The author speculates about rates of exchange
among humans engaged in reciprocal altruism (p. 46) but offers no evidence.
32 DAVID FRIEDMAN
21. Buss (1999, pp. 259–264). Packer (1977) describes the observed pattern of reciprocal
support among olive baboons; his data are not adequate to determine whether there is
an implicit one for one exchange. De Waal (1983) is primarily concerned with behavior
(among chimpanzees) more nearly analogous to human political behavior than to human
market behavior. Nonetheless, he describes behavior that appears to be “direct payment
for services rendered” (p. 203) and writes: “For the time being I should like to sum up as
follows: Chimpanzee group life is like a market in power, sex, affection, support, intolerance
and hostility. The two basic rules are ‘one good turn deserves another’ and ‘an eye for an
eye, a tooth for a tooth’ ” (p. 207).
22. Buss (1999, pp. 174, 177). Symons (1979, pp. 158–162, 253–261).
23. Kim Hill and Hillard Kaplan,“Tradeoffs in male and female reproductive strategies
among the Ache: Part 1.” In: Laura Betzig, Monique Bogerhoff Mulder & Paul Turke
(Eds), Human Reproductive Behaviour: A Darwinian Perspective (pp. 283–284), where
the authors also discuss alternative explanations for their observations. Triver (1977, p. 47)
writes: “An individual in a subgroup who feels that another member is subtly cheating
on their relationship has the option of . . . Attempting to join another subgroup . . . There is
evidence in hunter-gatherers that much movement of individuals from one band to another
occurs in response to such social factors as have just been outlined.”
24. Alfred Marshall (1920), Principles of Economics, Bk III Chap. V §3, 4.
25. Samuelson (1937).
26. For an extensive discussion of the experimental evidence and various attempts to
interpret it, see Loewenstein and Elster (1992).
27. The observed pattern is sometimes described as “hyperbolic discounting,” since it
fits a hyperbolic function better than it fits the exponential implied by the conventional
economic model of intertemporal choice.
28. This would not be true if we modeled contract non-fulfillment as the result of a
stochastic process where each week there was a certain probability that the other party
would decide to renege on his obligation. That would yield the same result – expected value
declining exponentially with time – as the conventional model. But there are at least two
more plausible models that work better:
Fulfillment depends on the relations between the two parties at the time the obligation
comes due. Relations each week have some probability of switching from friendly to
unfriendly, and also some probability of switching from unfriendly back to friendly. They
are friendly when the agreement is made. As time goes on, the expected state of relations
moves towards its equilibrium level.
Fulfillment depends not on future events but on the present plans of the other party. The
longer it is until the obligation comes due, the easier it is for him to get out of it – to be
somewhere else, or to have acquired allies to use against any retaliation by your for non-
performance, or to claim to have forgotten it (this is, after all, a world without writing). Hence
the fact that the offer is well into the future signals that he probably does not intend to fulfill it.
29. An explanation along generally similar lines can be found in Jones (2001,
pp. 1177–1179).
30. Kahneman, D., Knetsch, J., & Thaler, R. (1991). Experimental Tests of the
Endowment Effect and the Coase Theorem. Journal of Political Economy, 98 (1990),
1325–1348; Reprinted in Richard Thaler (1992, pp. 167–188).
31. For an early description of such behavior, see Konrad Lorenz (1963, Chap. 3,
pp. 23–48).
32. A willingness to fight unreasonably hard to defend what is yours is still sometimes
useful even in a society where the government plays a role in protecting property. For a
general discussion of commitment strategies as a foundation for social order in general and
property in particular, see Friedman (1994).
Jones (2001, pp. 1183–1185) also links endowment effects to territorial behavior, but
offers a different explanation.
33. Frank (1987). The book is an extensive discussion of the economic consequences of
the fact that individuals value status. On page 19 the author mentions competition for food,
mates and other desiderata as one explanation of that taste but does not point out the special
significance of the fixed supply of potential mates for the importance of relative rather than
absolute outcomes.
REFERENCES
Barkow, J., Cosmides, L., & Tooby, J. (1992). The adapted mind: Evolutionary psychology and the
generation of culture. New York and Oxford: Oxford University Press.
Becker, G. (1991). A note on restaurant pricing and other examples of social influences on price.
Journal of Political Economy, 99(5), 1109–1116.
Buss, D. M. (1999). Evolutionary psychology: The new science of the mind. Needham Heights, MA:
Allyn & Bacon.
Dawkins, R. (1990). The selfish gene. Oxford University Press.
De Waal, F. (1983). Chimpanzee politics: Sex and power among apes. Baltimore: Johns Hopkins
University Press.
Frank, R. (1987). Choosing the right pond: Human behavior and the quest for status. Oxford University
Press.
Friedman, D. (1991). In defense of Thomas Aquinas and the just price. In: M. Blaug (Ed.), St. Thomas
Aquinas (pp. 1225–1274). Elgar. Reprinted from History of Political Economy, 12(2) (1980).
Friedman, D. (1994). A positive account of Property Rights. Social Philosophy and Policy, 11(2,
Summer) 1–16 (http://www.daviddfriedman.com/Academic/Property/Property.html).
Friedman, D. (1996). Hidden order: The economics of everyday life. New York: HarperCollins.
Friedman, D. (1990). Price theory: An intermediate text. South-Western (http://www.daviddfriedman.
com/Academic/Price Theory/PThy ToC.html).
Jones, O. (2001). Time-shifted rationality and the law of law’s leverage: Behavioral economics meets
behavioral biology. Northwestern University Law Review, 95(4).
Loewenstein, G., & Elster, J. (1992). Choice over time. New York: Russell Sage.
Lorenz, K. (1963). On aggression. M. K. Wilson (Trans).

Marshall, A. (1920). Principles of economics. Macmillan.
Packer, C. (1977). Reciprocal altruism in Papio anubis. Nature, 265, 441–443.
Samuelson, P. (1937). A note on measurement of utility. Review of Economic Studies, 4, 155–161.
Symons, D. (1979), The Evolution of human sexuality. New York: Oxford (check pages).
Thaler, R. H. (1992). Quasi-rational economics. New York: Russell Sage.
Trivers, R. L. (1971). The evolution of reciprocal altruism. Q. Rev. Biol., 46, 35–57.
Zywicki, T. (2000). Evolutionary psychology and the social sciences. Humane Studies Review, 13(1)
(http://www.theihs.org/libertyguide/hsr/hsr.php/36.html).
MOGUL GAMES: IN DEFENSE OF

INEQUALITY AS AN EVOLUTIONARY
STRATEGY TO COPE WITH MULTIPLE
AGENTS OF SELECTION
Deby Cassill and Alison Watkins
. . . this is the oldest story in America: the struggle to determine whether ‘we, the people’ is a
spiritual idea embedded in a political reality – one nation, indivisible – or merely a charade
masquerading as piety and manipulated by the powerful and privileged to sustain their own
way of life at the expense of others.
Bill Moyers, 4 June, 2003
Take Back America Conference
Sponsored by the Campaign for America’s Future.
ABSTRACT
In this paper, we propose that the “powerful and privileged” sustain their way
of life through greed and they sustain the lives of others through trickledown
sharing. Greed provides the powerful and privileged a buffer against famine.
Trickledown sharing provides them a buffer against predation or war.
The inspiration for this integration of greed and trickledown sharing as
self-preservation strategies is a multi-selection model called skew selection.

According to skew selection, when perennial organisms are subjected to
cycles of famine and predation, greed and trickledown sharing increases

ISSN: 1529-2134/doi:10.1016/S1529-2134(04)07003-6
35
36 DEBY CASSILL AND ALISON WATKINS
the organism’s survival relative to a greed-only strategy. Skew selection is

extended to explain greed and trickledown sharing among humans through
the introduction of mogul games. The results of mogul games reported herein
suggest that inequality is an emergent property of self-organizing systems
and potentially an essential precursor to the evolution of social behavior. In
the future, it is our hope that mogul game simulations will be employed by
others to explore the effect of variation in cycles of predation and resource
abundance on the rules of greed (resource acquisition) and trickledown
sharing (resources redistribution).
INTRODUCTION
In modern humans, greed is ubiquitous. Every family has its wealthy relative,
every religion its wealthy spiritualist, every corporation its wealthy CEO,
every bureaucracy its wealthy supervisor, every court its wealthy judge, every
government its wealthy leader. Paradoxically, sharing is ubiquitous too – a person
offers a seat in a crowded bus to a stranger, another cooks a meal for a sick
neighbor, another donates money to help the poor in a continent on the other side
of the world, yet another stops to help a stranger change a flat tire. If greed is
essential for survival and reproduction (Darwin, 1859), why do so many humans
share resources so often in the form of time, money, food and shelter?
A large number of mathematical models has been constructed to address the
redistribution of wealth among kin – in small increments through sharing or in large
increments through altruism (Bergstrom, 1995; Dawkins, 1976; Hamilton, 1964;
Hofbauer & Sigmund, 1998; Maynard-Smith, 1964; Trivers, 1985; West-Eberhard,
1975). In addition, a number of game theory and other group selection1 models have
been published to address the redistribution of wealth among strangers – through
sharing or altruism (Axelrod, 1984; Bergstrom, 2003; Bergstrom & Stark, 1993;
Lewontin, 1970; Ridley, 1996; Sober & Wilson, 1998; Trivers, 1971; Wade, 1985).
In our opinion, there are several problems that limit the applicability of kin
selection, game theory and group selection models to explain sharing or altruism.
First, although most games provide mechanistically efficient models for greed
or sharing among players within the same generation, none that we know of
explains the persistence of greed and sharing among players among generations
over evolutionary time. Second, most models lack a sound heritable mechanism,
either biological or cultural, by which sharing or altruism would be reproduced
into the next generation (for exceptions, see Hofbauer & Sigmund, 1998; Packer
et al., 1990; Weibull, 1995). Third, most models assume maximal or optimal
efficiency rather than effectiveness (e.g. Alcock, 1979; Axelrod, 1984; de Waal,
Mogul Games 37
1996; Drickamer & Vessey, 1982; Ghiselin, 1974; Hirshleifer, 2001; Hölldobler
& Wilson, 1990; Noë et al., 2001; Sober & Wilson, 1998; Stearns, 1992; Stephens
& Krebs, 1986; Wilson, 1975). In a brilliant accounting of selection, Hull et al.
(2001) noted that one striking feature of selection processes was that they were
“incredibly wasteful.” In this paper, we outline a model and a game that focus on
the effectiveness of selection (maintaining diversity) rather than the efficiency of
selection (maintaining optimality).
The model we employ to explain the social behavior of humans is skew
selection (Cassill, 2003). Skew selection integrates greed and trickledown sharing
as complementary strategies for self-preservation against famine and predation. In
the following sections, we describe skew selection, and then apply the model in
the form of simulated mogul games to test the patterns of greed and trickledown
sharing against the patterns of famine and war.
SKEW SELECTION
Skew selection was first proposed to explain the evolution of sterile workers in ants.
By directing the allocation of resources among their offspring in a trickledown
pattern, fire ant queens produce sterile offspring and fertile offspring (Cassill,
unpublished data). Sterile offspring (workers) are small, cheap and therefore
disposable (Cassill, 2002). The job of sterile offspring is to shield the queen and
her more expensive fertile offspring from agents of death such as winter famine,
spring floods and summer territorial battles (Adams & Tschinkel, 2001; Hölldobler
& Wilson, 1990; Macom & Porter, 1996; Porter, 1988; Porter & Tschinkel, 1993;
Tennant & Porter, 1991; Tschinkel, 1993). Allocating a significant portion of their
resources toward the production of disposable offspring is the way queens ensure
the next generation of queens. We are applying skew selection to humans because,
at a basic level, ants and humans share a number of social traits. Ant societies
are families consisting of a single parent and her offspring; or, a tribe consisting
of a number of unrelated parents and a multitude of cooperating kin and non-kin
offspring. Ants construct stable domiciles in which the young are reared. Ants
gather and hunt in the surrounding territory, then return to the domicile to feed the
young and other nestmates. Ants send soldiers to defend territories surrounding
the domicile from invading neighbors. Ants increase the carrying capacity of
their territories through division of labor and economies of scale (Hölldobler &
Wilson, 1990). Most importantly, ants employ greed, competition, cooperation and
philanthropy to survive a complex, multi-selection environment (Cassill, 2003).
Skew selection extends Darwin’s natural selection model (1859) by integrating
greed, competition, cooperation and philanthropy into an organism’s fitness
Fig. 1. Conditions that Favor Greed, Sharing or a Combination of Greed and Trickledown
Sharing. (a) Greed produces the greatest good – largest body size for surviving famine – but
not the greatest number. (b) Sharing produces the greatest number – largest group size for
surviving predation – but not the greatest good. (c) A combination of greed and trickledown
sharing produces the greatest good for the greatest number when both famine and predation
are selective agents.
equation. According to skew selection, greed, competition, cooperation and

philanthropy represent multiple social strategies that co-evolve when players in
the game of life are faced with multiple agents of selection such as famine and
predation (Fig. 1). Cycles of famine (for example, winter scarcity) select for greed.
Greed consolidates resources into stockpiles. Stockpiled resources attract other
players. The cluster of players attracts predators. Players who support freeloaders
at a marginal level as bait for predators do better than players who do not support
freeloaders. What may seem “wasteful” from one point of view (Hull et al., 2001)
may in fact be rational. Freeloaders (which could include the young, inexperienced,
sick, disabled, aged) might turn out to be the least expensive way to gain safety in
numbers (Carr & Landa, 1983; Hamilton, 1971; Landa, 1998). Hence, predation
pressure selects for philanthropy.
Another benefit to supporting freeloaders is that they stabilize predators at low-
levels of ability. Given a choice between a low-ability freeloader and a high-ability
donor, predators will select the easy prey – the freeloader (Zahavi & Zahavi, 1997).
If the predator need only evolve to be fast enough to catch a freeloader, it follows
Mogul Games 39
then that philanthropy might be one of nature’s clever countermeasures to runaway

selection (i.e. the ratchet effect) in which faster predators select for faster prey
which then select for even faster predators (Pianka, 1988).

The mechanism for maintaining a large number of freeloaders and also
maintaining a central position within the group of freeloaders is trickledown
sharing (Cassill, 2003). By retaining more than he donates to others, the donor
maintains his position of status in the center of the group, away from the vulnerable
edge. Trickledown sharing is also called competitive cooperation. In the final
analysis, a shrewd combination of greed and trickledown sharing provides four
Fig. 2. Why do Moguls share? By sustaining players of lesser ability, moguls not only gain
efficiencies from division of labor and economics of scale, they also gain a buffer against
agents of death such as predation or war. In addition, by sustaining marginal individuals,
moguls keep predators at relatively low levels of virulence. Source: Figure reprinted with
permission from the Journal of Bioeconomics.
types of indemnity against famine and predation: (1) stockpiled reserves; (2) safety
in numbers; (3) safety in position; and (4) low-ability predators.
The mechanisms for the inheritance of inequality are through parents and
offspring. Nature will select for parents who invest unequally in their offspring,
producing many disposable, low-ability offspring to buffer the parents and their
high-ability offspring from predation. Hence, skew selection integrates individual
selection (the parent) and group selection (the offspring) into a single model of
self-preservation. The next generation of surviving offspring become parents who
invest unequally among their offspring.
Mogul games, introduced for the first time in this paper, extends the skew
selection model by framing the social behavior of humans into a natural selection
context. Skew selection predicts that, when faced with cycles of resource scarcity
and predation, moguls who employ a combination of greed (resources acquisition)
and trickledown sharing (resource redistribution) are more likely to survive than
moguls who employ only greed. According to skew selection, if a mogul shares
a portion of his resources to feed the masses during periods of famine, then the
masses will aggregate around the mogul, buffering him from death during periods
of high predation (Fig. 2).
We think that future mogul games will demonstrate that moguls do not sustain
their own way of life “at the expense” of others. To the contrary, it is the moguls who
sustain the masses at a modicum of expense to themselves. Because the expense of
sharing comes from vast excesses, there is no appreciable short-term or long-term
cost to a mogul’s survival or reproduction.
SIMULATED MOGUL GAMES

There are five ingredients to the mogul games: (1) players; (2) resources such as
currencies; (3) territories; (4) rules for resource acquisition (greed); and (5) rules
for resource redistribution (trickledown sharing). There are five phases to these
mogul game simulations. The first three phases explore how inequality in the form
of moguls and masses might emerge from self-organizing processes. The final two
phases explore why inequality might be sustained to benefit moguls and masses
given cycles of famine and predation. In this paper, we play the game through the
first four phases of resource acquisition. Future games can focus on the benefits
of certain patterns of resource redistribution under different conditions of famine
and predation (i.e. war).
Phase I: Initial Conditions. Our mogul game simulations consisted of twenty
players, a hundred units of currency (100 U.S. dollars), a territory consisting
of four-hundred (400) cells and a set of rules by which each player could
Mogul Games 41
acquire currency. Players began the game with no currency. Each simulation
ended when all currencies had been collected. The initial distribution of dollar
bills within the territory was random, uniform or clustered. At the beginning of
each game, players were placed randomly within the territory. Thereafter, players
traveled within the territory according to the rules in the next section. When
a player landed in a cell containing a dollar bill, he collected it and carried it
with him.
Phase II: Rules for Resource Acquisition: Players moved within the territory
according to three parameters: (1) the direction of each step; (2) the number of
steps each turn; and (3) the duration of sleep between turns.
Direction: A random direction was selected each move. If a random direction
could not be made (i.e. a player was at the edge of the grid), then another random
direction was selected until a move could be made.
Number of Steps Per Turn: The number of steps per turn per player was either
randomly assigned per game (1–10 steps) or was constant per game (5 steps).
When the numbers of steps was randomly assigned, that number was fixed per
player until the end of the game. For example, a player who was assigned 9 steps
would move 9 steps each turn until the game ended. Players who moved more
steps per turn were considered more “able” than players who moved fewer steps
per turn.
Duration of Sleep Per Turn: Players were assigned either a random duration
of sleep between turns (10–250 minutes) or a constant duration of sleep between
turns (50 minutes). In each case, once the sleep was assigned, it was fixed for each
player for the duration of the game. Players who slept shorter periods of time were
considered more motivated than players who slept longer periods. Players started
their turns in random order. Players who slept less, had more turns.
Phase II ended when the last of the dollar bills in the territory had been collected.
In these games, the number of dollars a player needed to survive one day was $1.
Any amount above that number was considered a reserve for surviving days when
resources were scarce or luck was bad. Other rules are possible.
Phase III: The three players who accumulated the largest number of dollar
bills during Phase II were designated moguls. Once the moguls were identified,
the remaining seventeen players were programmed to cluster around the mogul
nearest to them. Other rules are possible.
Phase IV: After the players had clustered around the moguls, dollars were tallied
within and among mogul groups. Tallies included each mogul’s personal capital,
each mogul’s group size (group capital) and each mogul’s gross capital – the sum
of his capital and that of the other players in his group. For example, if a mogul
collected $10 and his group consisted of five other players each of whom had
collected $1, then the mogul’s personal capital was $10, his group size was 6; and
his gross capital was $15. Depending on the goal of the game, any one of these
tallies could be used to determine the winning mogul.
Phase V: Mogul survival over time depends on his capital investments (personal
and group capital) and the agents of selection under study. A mogul’s gross
capital at the end of a game would depend on the amount of resources within the
territory, how often those resources are replenished, the rules by which players
collect dollars, the rules for player aggregation and the rules for redistributing
dollars to maintain mogul status and group size. Depending on the game, dollars
can trickle up from player to mogul to win entrance into a different mogul’s
group, or dollars can trickle down from mogul to player to increase group size.
The intent of Phase V is to test which mogul strategy results in the greatest
good for the greatest number in a given environment in which cycles of resource
abundance (summer glut and winter paucity) and predation (war) vary over
time.
METHODS
These simulated mogul games were developed in Java using the Math class
random number generator and multithreading. Each game was replicated 10 times
to determine the robustness of the final distribution of dollars among players.
Other programs and other rules might produce different results from those reported
below.
Player clusters in Phase III were formed using Euclidean distance. There was
no overlap of players or dollars in the initial distributions – random or clustered.
We began with twelve games generated by three distributions of dollar-bills in the
territory × four sets of player rules.
Analysis: Regression and nonparametric tests (Wilcoxon-Kruskal-Wallis Tests
with chi square approximations for the p values) were used to analyze trends or
differences in player income based on player acquisition rules or on the distribution
of dollars in a territory. Data were analyzed using a software program by Sall
and Lehman (1996). The initial order of play was not a significant predictor
of the final distribution of income among players (Regression: R 2 = 3.4%,
p = 0.588).
RESULTS
Regardless of the initial conditions, the distribution of dollars among players at the
end of the games was highly unequal or skewed (Fig. 3). In other words, regardless
Mogul Games 43
Fig. 3. Player income. At the conclusion of twelve different kinds of simulated games
player income was, in every case, unequally distributed (skewed). Note: Even when the
initial conditions were uniform (Game 3), the final distribution of dollars among players
was skewed.
of the randomness or uniformity of conditions, a few players gained significant

wealth relative to the other players. It should be noted, however, that the degree of
inequality in the distribution of dollars was less for games with uniform parameters
than for games with one or more randomized parameter (Fig. 4a–d ). In games with
uniform parameters, moguls averaged $10.20 and the masses averaged $4.60. In
games with random parameters, moguls averaged $17.30 and the masses averaged
$3.15. Of note is the fact that clustered resources generated the greatest skew
in the distribution of resources (Fig. 4c). Cycles of famine and predation would
determine the winning distribution over time. It is the goal of future mogul games
to explore the evolution of different types of distributions of wealth from different
patterns of famine and predation.
Four games were analyzed in detail to show how future mogul games might
be played. When the distribution of dollars was random and the rules of resource
acquisition were random (Fig. 5), Mogul 1 acquired $18, Mogul 2 acquired $14
and Mogul 3 acquired $12. Mogul 1 had the greatest personal wealth and therefore
would be the least vulnerable to famine than the other moguls and non-mogul
players. However, Mogul 1 had the smallest group size and therefore would be the
most vulnerable to predation. Conversely, although Mogul 3 would be the most
vulnerable to famine relative to the other moguls, he would be the least vulnerable
to predation because of his large group size. Mogul 2 had moderate wealth
and moderate group size and might improve his survival by merging or warring
with another mogul. Patterns of resources scarcity (famine) and predation would
Fig. 4. Four Games were Detailed. These Histograms Show the Degree of Variation within
Games (10 Replicated Runs Per Game), and the Descriptive Statistics for Each Game. Note:
The descriptive statistics are the means of the ten replicates. For example, the maximum
income acquired by a player was the mean of ten replicates. The greatest inequity (skew)
of income among players was in the clustered distribution of dollars with random rules of
play (c). The least inequity (skew) was in the uniform distribution with uniform rules of
play (d).
determine which hierarchical pattern of resource acquisition and redistribution

would be most likely to survive long-term – more personal wealth with smaller
group size or less personal wealth with larger group size.
When the distribution of dollars was uniform and the rules of resource
acquisition were random (Fig. 6), Mogul 1 acquired $17, Mogul 2 acquired $14
Mogul Games 45
Fig. 4. (Continued )
and Mogul 3 acquired $9. In this game, Mogul 1 had a group of intermediate size;
Mogul 2 had the largest group and Mogul 3 the smallest group. The only player
to join Mogul 3 had acquired no dollars during the search. This player would
therefore be dead the next day unless Mogul 3 was willing to donate a dollar to
keep him alive. In this case, the mogul and the player might both do better by
joining another mogul.
When the distribution of dollars was clustered and the rules of resource
acquisition were random (Fig. 7), Mogul 1 acquired $18, Mogul 2 acquired
$17 and Mogul 3 acquired $11. In this game, Mogul 1 had the largest group,
Mogul 2 the intermediate group size and Mogul 3 the smallest group. Hence,
Fig. 5. Random Distribution of Dollars and Random Rules of Play. Note: Grids show
the initial distribution of dollars and location of players (Phase I), the final distribution
of income among players (Phase II) and the cluster of subordinates about the mogul
(Phase III).
Mogul 3 and his group might be doomed depending on the intensity and
duration of famine and predation. Their best option might be to join forces with
another mogul.
When both the distribution of dollars and the rules of resource acquisition were
clustered (Fig. 8), moguls acquired less money relative to other game parameters.
Mogul 1 acquired $9 and a group size of 2. Mogul 2 acquired $8 and a group size
of 4. Mogul three acquired $7 and a group size of 14. Relative to moguls in the
other games, these moguls might all die if there was an extended famine. Hence,
Mogul Games 47
Fig. 6. Uniform Distribution of Dollars and Random Rules of Play. Note: Grids show
(Phase III).
these data suggest that equality might leave groups more vulnerable to extinction
than inequality!
Player Ability versus Motivation: Recall that number of steps per player per turn
and the duration of sleep time per player between turns were proxies for player
ability and player motivation respectively. In addition, we assumed that variation
in player income at the end of the game that was not accounted for by motivation
or ability was attributed to player luck.
Fig. 7. Clustered Distribution of Dollars and Random Rules of Play. Note: Grids show
(Phase III).
We found that player luck accounted for 59% of the observed variation in player
income at the end of the game, player motivation accounted for 25% and player
ability accounted for 16% (Multiple Regression: unexplained variation, R 2 = 59%,
p < 0.0001; sleep duration, R 2 = 25%, p < 0.001; number of steps, R 2 = 16%,
p < 0.001). This finding supports the idea that motivation is a better predictor of
success than raw talent.
Mogul Games 49
Fig. 8. Uniform Distribution of Dollars and Uniform Rules of Play. Note: Grids show
the initial distribution of dollars and location of players (Phase I), the final distribution of
income among players (Phase II) and the cluster of subordinates about the mogul (Phase III).
FUTURE MOGUL GAMES

In this section, we describe how future mogul games can be played to explore ef-
fective survival strategies (i.e. the rules for resource acquisition and redistribution)
given different patterns in the cycles and intensity of famine and predation.
Mogul Survival: The heuristic element of this game begins with Phase V where
moguls anticipate cycles of famine and predation and act accordingly. Moguls
balance the benefits of group size gained through trickledown sharing with the
benefits of their own personal capital gained through greed. We think mogul games
offer an opportunity to explore the underlying conditions that shape inequalities

among humans facing complex environments. In the following three sections, we
describe other ways to play mogul games.

Experimental Mogul Games: Experimental mogul games are best imagined as
a television game show with two moguls on stage and an audience of players on
which the moguls bid to establish their group. Each mogul begins the game with
the same amount of personal capital (for example, $1,000). The game consists of
“n” number of rounds in which “m” number of players from the audience will be
introduced to the moguls. There is no Phase I–IV in this game as players begin
with a pre-set amount of dollars. The goal of Phase V is to test the rules for the
redistribution of dollars among moguls and players against different agents of
selection.
Each mogul spins an income wheel or presses an income button to determine the
amount of income he receives for that round (the range and frequency distribution
of income amounts is set by the experiment designer). The starting income of a
mogul is not revealed to the other mogul and vice versa. Thereafter, as moguls earn
income, the amounts are visible to each other. A mogul can keep the income as
personal capital, or spend any portion of it to “buy” a player to gain group capital.
A player is introduced to the moguls simultaneously. Players signal their
economic status directly by displaying their value or indirectly through other
signals such as clothes, hairstyles, jewelry, grooming, and posture. Initially, no
player has higher economic status than the moguls. For example, in this game,
the personal capital of players could range from $10 to $990. Each mogul bids
for a player. The bids can be open and repeated – like an auction. Or bids can be
one-time and silent. In either case, the mogul with the higher bid wins the player.
The player moves from the audience and sits behind the successful mogul. The
winning bid is debited from the winning mogul’s income (or personal capital if
the mogul wishes to spend more than the amount of income he just received) and
is credited as personal income for the player. The game continues with multiple
rounds of spinning for income, and bidding for players. The personal capital of
each player (their initial value plus the bids) becomes part of the mogul’s group
capital.
At the conclusion of the bidding rounds, the personal capital of each mogul
and all players in each mogul’s group is revealed and their economic ranking
determined. If any player’s personal capital is greater than that of the mogul, that
player usurps the current mogul and emerges as the new mogul. Hence, a mogul
competes first with a potential “mogul from within” for leadership of the group
before he competes with the opponent mogul. A mogul might get usurped if he has
depleted his personal capital during the bids to gain large group size in anticipation
of a predatory threat from another mogul.
Mogul Games 51
After within-group rankings are tallied and the reigning mogul for each group
is determined, the personal capital and group capital investments of each of the
two moguls are displayed to the other mogul. At this point, the income wheel
disappears and a selection wheel appears. The selection wheel contains good times
and bad times. The ratio of each would vary according to the experimental question.
Different agents of selection such as disease, war, predation, accident, famine,
drought, natural disasters and so on would cost the mogul either personal capital or
group capital. For example, if the agent of selection is disease, the disease deducts
“x” amount of dollars from the mogul. The mogul can pay the amount himself or
he can designate a player (or players) within his group to pay the amount. When a
player’s personal capital goes to zero, that player dies and group size shrinks. Each
mogul takes a turn spinning the selection wheel and losing dollars or players in
his group. When the game ends, the mogul with the largest gross personal capital
wins (a total of all personal capital including the mogul and the remaining players
in his group).
Because experiments must come to an end, this game has a constructed end
point – the final spin of the selection wheel. In real life, moguls have two choices in
the disposal of the “spoils” of the losing mogul – he can compete to gain personal
capital or cooperate to gain group capital. If the winning mogul decides to compete
with the losing mogul, 100% of personal capital is debited from each member
in the losing group and credited to the mogul. The mogul can keep 100% of the
losing group’s total person capital or he can systematically distribute the losing
group’s personal capital among his group members (a fixed or disproportional
distribution). One example of a distribution pattern would have the individual with
the most personal capital in the winning group (the winning mogul) acquiring
the personal capital of the individual with the most personal capital in the losing
group (the losing mogul) and so on down the economic rank ladder until all
paired debit-credits have been made. In this case, if group size is larger for the
winning mogul, then the members with the lowest economic rank in his group
will not gain personal capital from the losing group. If group size is smaller for
the winning mogul, then the personal capital of the remaining members of the
losing group is credited to the mogul’s personal capital. When players lose 100%
of their personal capital, they “die” and are removed from the game permanently.
Hence, when a winning mogul competes against a losing mogul, the losing group
goes extinct.
If the winning mogul decides to cooperative with the losing mogul, the groups
merge to form a larger unit. Each member of the losing mogul’s group has “x” % of
his personal capital debited; this amount (which can be fixed or variable according
to economic rank) is then credited to the personal capital of each member of the
winning mogul’s group. The merger is systematic, according to economic rank.
Hence, there is a major reshuffling of economic rank among group members as

they merge. Once the merger occurs, there is the possibility that the losing mogul
can usurp the winning mogul if his personal capital is greater of the two. The
“mean personal capital” of a group or “group size” can be used to break a tie
between two moguls with the same “gross personal capital,” although other rules
are possible.
Experimental versus Simulated Mogul Games: Experimental versus simulated
mogul games test the investment decisions of human players against the decisions
of simulated players. There are a number of ways to play this game. For example,
the human player is designated as Mogul 1. The simulated player is designated
as Mogul 2. In addition, there are “m” number of simulated players. A player
is randomly selected and is introduced to the moguls. Moguls can see, but not
converse, with the player. Mogul 1 is given an amount of income that he can keep,
divide equally with the player or give it all to the player. If Mogul 1’s investment
decision agrees with Mogul 2’s prediction of what Mogul 1 will invest, then Mogul
1 wins and advances to another player. The predictions of Mogul 2 follow the
investment rules set by the simulation designer. The game ends when Mogul 1
completes “n” number of rounds or when the decision by Mogul 1 mismatches the
prediction of Mogul 2 “x” number of times (i.e. three strikes and he’s out). The
modeler or experimenter determines the allowed number of rounds or the allowed
number of mismatches before the game ends. The magnitude of income per turn,
and whether income is variable or fixed with each turn, is determined a priori
depending on the question the game is designed to address.
Complex Simulated Mogul Games: Following are examples of more complex
mogul games. This list is by no means comprehensive. As long as currencies and
visual assessment of player status are maintained, game rules and parameters can
be modified any number of ways to illuminate the nature of social decisions –
greed or sharing.
Multiple Group Members: In complex games, moguls and players can join more
than one group. For example, let us say a player’s personal capital is $100. Mogul
1 bids $15 to attract him; Mogul 2 bids $20 to attract him. If the player joins both
moguls, his personal capital is greater and therefore each mogul’s group capital is
greater, than if the player joined only one mogul. Moguls can become members
of other mogul groups as well. A mogul might pay a premium price in personal
capital as membership dues to gain a portion of the group capital belonging to
another mogul.
Two-Way Income Flow: In complex games, income can flow, not only from
mogul to player, but from player to mogul. Moguls can purchase players for their
groups; or players can pay to join a mogul’s group. Player payment is debited from
his personal capital and credited to the mogul’s personal capital. In this version of
Mogul Games 53
the game, the economic rank of players counts. For example, if a player’s personal
capital is $100 and he pays the mogul $20 to join the group, the player’s personal
capital is now $80. However, the player gains a portion of the group capital of
players in the group who are lower in economic rank than he. Hence, high-status
players have more incentive to join multiple groups than low status players. In
addition, high-status players have more incentive to join large groups with lower-
status individuals than small groups with higher-status players. As in simple games,
the economic motives behind investment decisions are always self-preservation2
over time.
Income for Players and Moguls: Players can spin to acquire income at the
same time that moguls spin for income, increasing the possibility of a “mogul-
within” emerging into the leadership role or a “lieutenant mogul” forming a group
within the group by purchasing members himself. With players acquiring income,
positions of economic rank (status) among players will shift on a regular basis.
Rules on the magnitude of income per player would be determined by the question
the game designer is addressing.
Super Complexity: The more complex a game, the more the games might
resemble the economic exchanges of modern humans. Moguls are allowed to
visually assess each player’s status before an investment decision is made. Hence,
each mogul can estimate a player’s gender (Bergstrom, 1996), age, race and other
variables (e.g. health, symmetry, height, weight) and his or her potential economic
value (i.e. personal capital) before the mogul decides to invest. Signals of a player’s
economic value are assumed to be honest.
DISCUSSION
In the abstract, inequality (i.e. diversity) is not a bad thing (Hull et al., 2001).
Low-ability individuals with great motivation are productive people; and high-
ability individuals are productive people. Hence, those with high motivation or
high ability are likely to be the greatest sources of innovation and productivity in a
changeable environment. Highly productive people aggregate resources faster and

more effectively than others. Once wealth is accumulated, it becomes a significant
resource for the less-abled or the less motivated by way of jobs, consumption and
charity (Demsetz, 1988; Hill & Cassill, 2004).
Even though, in the abstract, inequality is not a bad thing, there are examples
throughout history of abuses by the “powerful and privileged” (Ehrlich & Lui,
1999; Haile et al., 2003). Therefore, it is important to differentiate between
benign greed and malignant greed (Hill & Cassill, 2004; Vehrencamp, 1983).
Benign greed involves stockpiling when resources are super-abundant – that is
when resource availability exceeds the survival and reproductive needs of all
organisms within the local environment. On the other hand, malignant greed is
the acquisition of resources during periods of resource scarcity. Malignant greed

reduces the stockpile of capital collected by others within the local environment –
it is an anti-Robin Hood strategy of “taking from the poor and giving to the rich.”
Whereas malignant greed increases the probability of survival at the expense of
others, benign greed increases the probability of survival, but not at the expense
of others.
In nature, malignant moguls are probably the exception rather than the rule
which may explain why war is a relatively rare event in human history worldwide.
Eventually, malignant moguls are executed by the suffering masses, or they perish
when the suffering masses defect or die off, leaving the mogul vulnerable to
predation. Shrewd moguls can prevent death by rebellion or predation by sharing a
portion of their resources to appease the masses. The degree of trickledown sharing
and the resulting level of inequality that evolves depend on the complexity of the
environment (Fig. 9).
The important point is that, from a mogul’s point of view, sharing is an investment
into group capital to enhance his own survival. Because sharing comes from
gross excesses, it usually does not diminish the mogul’s survival or reproduction
long-term. In some cases however, sharing may incur a small short-term loss of
productivity. But, if the mogul gains in long-term survival, then his total lifetime
productivity is greater than if he did not share. For example, in ants, low-quality
subordinate queens paid a price of ten eggs per unit time to cohabitate with a high-
quality dominant queen (dominants suppress egg production of subordinates by
10–20%). But, because subordinate queens lived twice as long in cooperative nests
with the dominant queen, the lifetime egg production of subordinates was greater
than if they established their own nest independent of other queens (Cassill et al.,
unpublished data). As a second example, if a solitary company makes 100 cars in
one year and then goes extinct, its lifetime productivity is 100. If a cooperative
company makes 50 cars per year for three years, its lifetime productivity is 150.
From a biological “fitness” point of view, the cooperative company is the winner
even though its annual “fitness” is lower than that of the solitary company. A shrewd
combination of greed and trickledown sharing can increase lifetime production in
a complex, multi-selection environment relative to greed only. The important point
here is that game models that measure only short-term production losses or gains
from cooperation (e.g. Bergstrom, 2003; Reeve & Keller, 2001) miss the boat
for perennial species such as ants and humans. The more important parameter of
biological success is lifetime production, not annual production.
The paradox of Darwinian selection is this: evolution requires variation;
natural selection reduces variation (Price, 1995). Skew selection offers a solution
Mogul Games 55
Fig. 9. Environmental Conditions that Favor Greed and Trickledown Sharing (Inequality)
for Perennial Species. (a) When predation is low and resources are homogenous in time,
space and composition (a steadystate), organisms do not share or stockpile resources. (b)
When predation is low and resources cycle in time, space or composition, organisms do
not share, but they do stockpile resources. (c) When predation is high and resources are
homogenous, organisms aggregate (often by cloning), but do not stockpile resources. (d)
When predation is high and resources cycle in time, space or composition, organisms
stockpile and share. Note: The degree and direction of sharing (one-way philanthropy or
multiple-way cooperation) depends on the degree of resource cycles and diversity.
to this paradox by revealing what might be a fundamental law of nature:

inequality is an emergent property of self-organizing systems coping with complex
environments (Fig. 9). In an irony that biologists and economics alike might
appreciate, nature’s “most normal (as in most common) distribution may be
skewed! As we show in Fig. 10, inequality is homologous to diversity. Stating

that complex environments select for offspring inequality is the same as stating
that complex environments select for offspring diversity. Nature generates new
diversity by selecting parents who invest unequally in their offspring (Cassill,
2003). Thereafter, nature maintains or reshapes diversity among offspring as they
mature into parents in their own right. Another way to view skew selection is as a
parental r-K selection strategy (Cassill, 2003). Parents produce many low-ability
offspring (r-selection) as well as a few high-ability offspring (K-selection).
Depending on the species, low-ability offspring may serve as helpers, dispersers,
Fig. 10. Understanding Skew Selection from Three Points of View. (a) From a sociological
point of view, trickledown sharing produces inequality. Like Robin Hood, sociologists are
the middlemen and women who advocate taking from the rich and giving to the poor. (b)
From an animal behavior point of view, trickledown sharing produces a hierarchy. Biologists,
economists and political scientists understand the economic benefits of hierarchies via chain
of command, division of labor and economies of scale. (c) From a bioeconomic point of view,
trickledown sharing produces diversity. Note: The important point is that each shape in this
figure is functionally identical to the other – the only difference is in the emotional response
that each shape elicits which is often based on a cultural/political/educational biases. When
equated to diversity, inequality is not a social evil, it is the engine upon which evolution
runs. Source: Figure reproduced with permission from the Journal of Bioeconomics.
innovators or buffers against predation depending on the complexity of the local

environment.
In conclusion, without the redistribution of wealth by the powerful and
privileged, the masses would never have been born; even if they had been born, they
would not have survived periods of famine on their own. On the other hand, without
the masses, the powerful and privileged would not survive periods of high predation
(i.e. war). Thus, greed and trickledown sharing are reciprocal, not opposing,
strategies by which moguls maintain personal capital and sustain group size at
the same time. Contrary to Rousseau’s “Discourse on the Origin of Inequality”
(translation of the 1755 monograph by Cress & Miller, 1992), in which he states
Mogul Games 57
that “. . . inequality is practically non-existent in the state of nature . . .,” inequality

might be one of nature’s shrewdest strategies for producing the greatest good for
the greatest number in multi-selection environments over evolutionary time.
NOTES
1. Games such as prisoner’s dilemma that offer a payoff for player consensus in addition
to a payoff for independent decisions by players is a group selection game. In other words,
the payoff for the group differs from the sum of payoffs by each individual.
2. Selfishness and self-preservation are not the same things. Selfishness implies never
sharing. Self-preservation often entails sharing.
ACKNOWLEDGMENTS
We thank Brian Woodhouse, Timothy Travis and an anonymous reviewer for
critical comments that vastly improved this work. We thank Roger Koppl for
encouragement and invaluable advice during the submission of this manuscript to
the Advances in Austrian Economics.
REFERENCES
Adams, E. S., & Tschinkel, W. R. (2001). Mechanisms of population regulation in the fire ant Solenopsis
invicta: An experimental study. Journal of Animal Ecology, 70, 355–369.
Alcock, J. (1979). Animal behavior: An evolutionary approach. Sunderland, MA: Sinauer Assoc.
Axelrod, R. M. (1984). The evolution of cooperation. New York: HarperCollins.
Bergstrom, T. C. (1995). On the evolution of altruistic ethical rules for siblings. The American Economic
Review, 85, 58–81.
Bergstrom, T. C. (1996). Economics in a family way. Journal of Economic Literature, 34, 1903–1934.
Bergstrom, T. C. (2003). The algebra of assortative encounters and the evolution of cooperation.
International Game Theory Review, 5, 211–228.
Bergstrom, T. C., & Stark, O. (1993). How altruism can prevail in an evolutionary environment. AEA
Papers and Proceedings, 83, 149–155.

Carr, J. L., & Landa, J. T. (1983). The economics of symbols, clan names, and religion. Journal of
Legal Studies, 12, 135–156.
Cassill, D. L. (2002). Yoyo-bang: A risk aversion investment strategy by a perennial insect society.
Oecologia, 132, 150–158.
Cassill, D. L. (2003). Skew selection: Nature favors a trickle-down distribution of resources in ants.
Journal of Bioeconomics, 5, 83–96.
Cress, D. A., & Miller, J. (1992). Jean-Jacques Rousseau: Discourse on the origin of inequality.
Indianapolis, IN: Hackett Publishing.
Darwin, C. (1859). The origin of species (1968 ed.). Penguin Books.
Dawkins, R. (1976). The selfish gene. Oxford, UK: Oxford University Press.
de Waal, F. (1996). Good natured: The origins of right and wrong in humans and other animals.
Cambridge, MA: Harvard University Press.

Demsetz, H. (1988). Ownership, control, and the firm: The organization of economic control. Oxford,
UK: Basil Blackwell.
Drickamer, L. C., & Vessey, S. H. (1982). Animal behavior: Concepts, processes, and methods. Boston,
MA: Willard Grant Press.
Ehrlich, I., & Lui, F. T. (1999). Bureaucratic corruption and endogenous economic growth. Journal of
Political Economy, 107, S270–S293.
Ghiselin, M. T. (1974). The economy of nature and the evolution of sex. Berkeley, CA: University of
California Press.
Haile, D., Sadrieh, A., & Verbon, H. A. A. (2003). Self-serving dictators and economic growth. CEFifo
Working Paper No. 1105 www.CESifo.de.
Hamilton, W. D. (1964). The genetical evolution of social behavior, I & II. Journal of Theoretical
Biology, 7, 1–52.
Hamilton, W. D. (1971). Geometry for the selfish herd. Journal of Theoretical Biology, 31, 295–311.
Hill, R. P., & Cassill, D. L. (2004). The naturological view of the corporation and its social responsibility:
An extension of the Frederick model of corporation-community relationships. Business and
Society (in press).
Hofbauer, J., & Sigmund, K. (1998). Evolutionary games and replicator dynamics. Cambridge, UK:
Cambridge University Press.
Hölldobler, B., & Wilson, E. O. (1990). The ants. Cambridge, MA: Harvard University Press.
Hull, D. L., Langman, R. E., & Glenn, S. S. (2001). A general account of selection: Biology,
immunology, and behavior. Behavioral and Brain Sciences, 24, 511–573.
Landa, J. T. (1998). Bioeconomics of schooling fishes: Selfish fish, quasi-free riders, and other fishy
tales. Environmental Biology of Fishes, 53, 353–364.
Lewontin, R. C. (1970). The units of selection. Annual Review of Ecology and Systematics, 1, 1–18.
Macom, T. E., & Porter, S. D. (1996). Comparison of polygyne and monogyne red imported fire
ant (Hymenoptera: Formicidae) population densities. Annals of the Entomological Society of
America, 89, 535–543.
Maynard-Smith, J. (1964). Group selection and kin selection. Nature, 201, 1145–1147.
Noë, R., van Hooff, J., & Hammerstein, P. (2001). Economics in nature. Cambridge, UK: Cambridge
University Press.
Packer, C., Scheel, D., & Pusey, A. E. (1990). Why lions form groups: Food is not enough. American
Naturalist, 136, 1–19.
Pianka, E. R. (1988). Evolutionary ecology. New York, NY: HarperCollins.
Porter, S. D. (1988). Impact of temperature on colony growth and developmental rates of the ant,
Solenopsis invicta. Journal of Insect Physiology, 34, 1127–1133.

Porter, S. D., & Tschinkel, W. R. (1993). Fire ant thermal preferences: Behavioral control of growth
and metabolism. Behavioral Ecology and Sociobiology, 32, 321–329.
Price, G. (1995). The nature of selection. Journal of Theoretical Biology, 175, 389–396.
Reeve, H. K., & Keller, L. (2001). Tests of reproductive-skew models in social insects. Annual Review
of Entomology, 46, 347–385.
Ridley, M. (1996). The origin of virtue: Human instincts and the evolution of cooperation. Auckland,
New Zealand: Penguin Books.
Sall, J., & Lehman, A. (1996). JMP start statistics: A guide to statistical and data analysis using JMP
and JMP IN software. Albany, NY.
Mogul Games 59
Stearns, S. C. (1992). The evolution of life histories. NY: Oxford University Press.
Stephens, D. W., & Krebs, J. R. (1986). Foraging theory. Princeton, NJ: Princeton University Press.
Tennant, L. E., & Porter, S. D. (1991). Comparison of diets of two fire ant species (Hymenoptera:
Formicidae): Solid and liquid components. Journal of Entomological Science, 26 , 450–465.
Trivers, R. L. (1971). The evolution of reciprocal altruism. Quarterly Review of biology, 46 , 35–57.
Tschinkel, W. R. (1993). The fire ant (Solenopsis invicta): Still unvanquished. In: B. N. McKnight
(Ed.), Biological Pollution: The Control and Impact of Invasive Exotic Species. Indianapolis,
IN: Indiana Academy of Science Press.
Vehrencamp, S. L. (1983). A model for the evolution of despotic versus egalitarian societies. Animal
Behaviour, 31, 667–682.
Wade, M. J. (1985). Soft selection, hard selection, kin selection and groups selection. American
Weibull, W. J. (1995). Evolutionary game theory. Cambridge: MIT University Press.
West-Eberhard, M. J. (1975). The evolution of social behavior by kin selection. Quarterly Review of
Biology, 50 , 1–33.
Wilson, E. O. (1975). Sociobiology: The new synthesis. Cambridge, MA: Harvard University Press.
Zahavi, A., & Zahavi, A. (1997). The handicapped principle: A missing piece of Darwin’s puzzle. NY:
Oxford University Press.
AUSTRIAN ECONOMICS,
EVOLUTIONARY PSYCHOLOGY
AND INDIVIDUAL ACTIONS
Geoffrey M. Hodgson
ABSTRACT
Friedrich Hayek, Ludwig von Mises, Ludwig Lachmann and George Shackle
upheld that investigations of the causes of purposes, preferences, beliefs or
behaviors by the social scientist were unwarranted. Shackle proposed that
human agency is an “uncaused cause.” Others admitted that human volitions
and actions are caused, but ruled out explanations of these causes from
social science. By considering Darwinian insights from modern evolutionary
psychology, this essay criticizes the view that causal investigations of human
volitions and actions are beyond social science. These insights also point to
the role of habit and instinct in human behavior.
1. INTRODUCTION
This paper critically examines a theme that has been prominent, but not universal, in
Austrian School writing, and scrutinizes it the light of some of the key theoretical
underpinnings of modern evolutionary psychology. The theme is found in the
writings of Friedrich Hayek, Ludwig von Mises, Ludwig Lachmann, George

Shackle and others, and it involves the claim that it is not the concern of the
economist to examine the causes of individual motives, purposes or beliefs.

ISSN: 1529-2134/doi:10.1016/S1529-2134(04)07004-8
61
62 GEOFFREY M. HODGSON
Evolutionary psychology draws from the well of Darwinism. Some of the

Darwinian underpinnings of evolutionary psychology are examined here, including
the insistence that all phenomena should be susceptible to causal explanation. The
upshot of this critique does not undermine all Austrian theoretical arguments, but
shows that the aforementioned attempts to restrict the purview of economic enquiry
are misguided. It is argued that principles underlying evolutionary psychology
point to the role of both habit and instinct in human behavior, in a manner redolent
of earlier instinct-habit psychologists, such as William James (1892).
The paper is organized in four sections. Section 1 reviews the arguments of
Hayek, von Mises, Lachmann and Shackle to the effect that causes of individual
motives or behavior are not within the purview of social science. Section 2 briefly
discusses the foundations of Darwinism including, by contrast, the unrestricted
Darwinian commitment to causal explanation. Section 3 explores some of
the implications for social science of insights from evolutionary psychology
concerning the human mind. Section 4 extends these insights by adding the role of
habit within the same evolutionary framework and arguing for its relevance within
social science. Section 5 concludes the essay.
2. ON AUSTRIAN ECONOMICS AND

INDIVIDUAL PURPOSES
The internal diversity within the Austrian School, in both theoretical and policy
terms, is widely recognized. We here consider an aspect of the writings of Hayek,
von Mises, Lachmann and Shackle, and no claim is made that their stance is
necessarily representative of the Austrian School as a whole.
What does unite members of the Austrian School, and indeed many other
economists, is a recognition of the reality of human agency and the fact that
individual action is powered by purposes and framed by perceptions that are seated
in the mind of the agent. Accordingly, to understand human action it is vital to
understand this subjectivity, and processes by which he or she may make choices
that are consistent with subjective goals and perceptions.
Personally, I endorse the standpoint in the preceding paragraph.1 Elsewhere I
have argued additionally that the social and institutional enablers of individual
cognition have also to be taken into account (Hodgson, 1988). I do not want to go
into the details of my auxiliary argument here; I firstly wish to raise the general
question of the causal determination of the goals and perceptions of each individual.
How are individual purposes and cognitive processes caused? I first examine some
selected Austrian answers to this question.
Austrian Economics, Evolutionary Psychology and Individual Actions 63
Hayek alluded to this issue in an essay of 1943, reprinted in his Individualism and
Economic Order. Hayek (1948, p. 67) here wrote of a “common misunderstanding”
concerning the individual in the social sciences:

The misunderstanding is that the social sciences aim at explaining individual behavior . . . If
conscious action can be ‘explained,’ this is a task for psychology but not for economics . . . or
any other social science.
Here Hayek separated the social sciences from psychology: psychology should
not aim at explaining individual behavior, while the social sciences can and should
do that job. It is not that Hayek denied that human characteristics or behavior
are determined by circumstances. In another essay of 1946, reprinted in the same
volume, Hayek (1948, p. 6) wrote of “men whose whole nature and character is
determined by their existence in society” but then goes on to assert that “there
is no other way toward an understanding of social phenomena but through our
understanding of individual actions directed toward other people and guided by
expected behavior.” He omitted the explanation of individual actions and expecta-
tions themselves. To summarize, Hayek freely admitted that individual character,
goals and expectations are caused and variable, but regarded the explanations of
such phenomena as unnecessary for, and beyond the purview of, social science.2
Ludwig von Mises gave slightly different reasons for neglecting the explanation
of human motives. While, Hayek establishes his position with a dogmatic
demarcation of the social sciences from psychology, von Mises pointed more
reasonably to the complexity of the causal relations involved. Von Mises (1966,
p. 18) argued that:
Concrete value judgments and definite human actions are not open to further analysis. We
may fairly assume or believe that they are absolutely dependent upon and conditioned by their
causes. But as long as we do not know how external facts – physical and physiological –
produce in a human mind definite thoughts and volitions resulting in concrete acts, we have
to face an insurmountable methodological dualism. . . . Reason and experience show us two
separate realms: the external world of physical, chemical, and physiological phenomena and
the internal world of thought, feeling, valuation, and purposeful action. No bridge connects –
as far as we can see today – these two spheres. . . . Human action is one of the agencies [that]
bring about change. . . . As – at least under present conditions – it cannot be traced back to its
causes, it must be considered as an ultimate given and must be studied as such.
The argument here is that we do not know the highly complex causes of human
action and therefore we “must” take value judgments and actions as given. While
von Mises was right in surmising the complexity of the causal process involved, his
argument would begin to dissolve once any significant single causal explanation
of some judgments or actions were established.
Another position is found in the works of Ludwig Lachmann. For him,
“methodological individualism” meant “that we shall not be satisfied with any
type of explanation of social phenomena which does not lead us ultimately to a

human plan.” But he insisted, on reaching this point, that we should be led no
further into exploring the causes of the plans themselves. As Lachmann (1969,
p. 94) put it:
Such analysis of observed phenomena in terms of pre-existent plans has nothing to do with
psychology. We are here concerned with purposes, not with motives, with plans, not with the
psychic processes which give rise to them, with acts of our conscious minds, not with what lies
behind them.
This stance is very similar to that of Hayek: both agreed that it is not the task
of the social sciences to examine the causes that lie behind purposes and plans.
In a nearby passage, Lachmann distances himself from behaviorist psychology,
consequently proclaiming: “Spontaneous mental action is not a ‘response’ to
anything pre-existent” (p. 93). But it is not clear whether Lachmann was saying that
“spontaneous mental action” was literally uncaused – which is one interpretation
of the word “spontaneous” – or it is not caused by a “response to stimulus” in the
particular manner described by behaviorist psychology.
While Hayek and von Mises did not propose an “uncaused cause,” and
Lachmann was unclear on this point, George Shackle adopted the notion. Shackle
is often regarded as an Austrian economist, despite his attempts to synthesize
Austrian insights with Keynesian theory. In particular, Shackle and Lachmann
had a particularly close intellectual relationship. In a Festschrift for Lachmann
on his eightieth birthday, Shackle (1986, pp. 281–282) argued that if choice
is “predetermined” then the “notion of choice would be empty and the act of
choice sterile.” Consequently, we are required “to suppose that choice can be in
some respects exempt from governance by antecedent thought or contemporary
circumstance, that choice can be in some respects an uncaused cause.”
This indeed was a persistent theme in Shackle’s writing. In a preceding and more
complete account, Shackle (1976, pp. 33–34) wrote of “randomness, of inspiration”
and that they “would free us from the fetters of complete determinate causation
of thoughts” and “from the abolition of inceptive choice, of choice in the sense
of an origin of the new.” For Shackle, “the new” was “the unforeknowable” and it
was beyond the bounds of reason. “Reason is analysis” and “there must be some
practical limit to its useful pursuit.” From this Shackle drew the conclusion that
“the notion of the uncaused” has a “place in scholarly discourse.” This argument
was further developed elsewhere (Shackle, 1979).
Many, including myself, have been seduced by Shackle’s beguiling prose, only
to see later its flaws. He conflated knowledge of cause with existence of cause; to
say that there are limits to knowledge, reason, understanding, or explanation has no
bearing on the existence or non-existence of any causal relation or determination.
Phenomena may be causally determined without being knowable or predictable.

He conflated doctrines of “determinism” in the sense of claiming predictability,
with notions of “determinism” in the sense of proposing causal determination.

Furthermore, neither “randomness” nor “inspiration” undermines an ontology of
causal determination. Random or stochastic causation is itself a form of causal
determinacy (Bunge, 1959). And inspiration is itself a cause, and arguably itself
caused.
A problem with Shackle’s position is that it involves an investigatory closure.
Once we affirm an “uncaused cause” we say that science should explain this much,
but no more. We may move so far down the causal chain, but no further. We arrive
at a causal and explanatory roadblock, policed by the adherents of the “uncaused
cause.” Admittedly, all ontological commitments involve dogma in the sense that
they cannot be directly verified by experience. But the principle of determinacy
is preferable to the “uncaused cause” because it does not place dogmatic bounds
on the scope of scientific enquiry and explanation. Accordingly, the preferable
ontological commitment is one that rules out miracles and denies any no-go zones
for science. The roadblock must be opened, even if the road ahead is treacherous
and complex.
Hence a remarkable contrast is present within this segment of the Austrian
tradition, between Hayek at one extreme and Shackle at the other. When Hayek
(1948, p. 6) admitted “men whose whole nature and character is determined by
their existence in society” he suggested that individual purposes and preferences
could themselves be molded or caused. Similarly, Von Mises (1966, p. 18) accepted
that judgments and actions are “dependent upon and conditioned by their causes.”
In contrast, Shackle argued repeatedly that choices were themselves “uncaused.”
Significantly, both authors denied that social scientists should investigate the causes
of choices, purposes or behaviors, but on entirely different grounds. While Hayek
ruled out such investigations because they were beyond the domain of social
science, Shackle’s “uncaused cause” meant that such causal explanations were
impossible. I shall argue below that the shared conclusion, plus the two very
different reasons for reaching it, are all incompatible with the letter and spirit of
modern evolutionary psychology.

Before we move towards this goal, however, I wish to reaffirm that I am not
attempting to describe any stance that is characteristic of the Austrian School
as a whole. To do this I quote from the eminent Austrian economist Eugen von
Böhm-Bawerk (1890, p. 247 n.) who wrote:
the ‘isolating’ method . . . its essence consists in, first of all, observing different aspects of
complex phenomena separately, but let it be understood, what is often forgotten, that it should
not content itself with a partial view, or mistake the part isolated in thought for the whole, but,
as far as possible, should construct out of the clearly grasped parts a complete whole.3
Böhm-Bawerk here referred to what some methodologists call “abstraction” and

others “isolation,” meaning that for the purpose of science it is impossible to deal
with every aspect and causal relation, and it is necessary to focus on those which
are deemed to be typical or most important. Accordingly, in the social sciences, it
is quite legitimate in many circumstances to take the goals or preferences of the
agent as given, thus isolating them within the theory from their complex and often
elusive causes. Böhm-Bawerk insisted, however, that we should never be satisfied
with any abstraction or isolation, and we should look further in an attempt to build
up a more complete picture. This would imply that the causes of choice, purposes
or behaviors should themselves come under investigation. As Anthony Endres
(1997) has shown, and in contrast to Hayek, von Mises, Lachmann and others,
Austrian School members such as Friedrich von Wieser and Böhm-Bawerk saw it
as legitimate for economists to explore the formation of preferences, using insights
from psychology if needed. Furthermore, they included endogenous preference
changes precipitated by learning as part of their theory.
Just as significantly, Carl Menger, the founder of the Austrian School,
based his theory of value on determinate physiological and other needs rather
than strictly subjective values. This standpoint is clear in his Grundsätze der
Volkwirtschaftslehre (Principles of Economics) of 1871. In working on later
editions of this work, Menger turned increasingly to the study of biology and
physiology, with the aim of formulating a theory of human needs as a foundation
to his theory of value.
Hence the propositions under critical scrutiny are not found in the writings
of all members of the Austrian School. They concern both matters of causality
and the contested domain of legitimate enquiry for the social sciences. The claim
that evolutionary psychology has a bearing on these issues is established by first
considering its Darwinian underpinnings and then its impact on considerations of
individual purposefulness or intentionality.
3. ON THE DARWINIAN UNDERPINNINGS

OF EVOLUTIONARY PSYCHOLOGY
Charles Darwin did not only proclaim that species had evolved, but also pointed to
the processual, causal mechanisms of evolution. He extended the realm of causal
explanation into areas that were deemed taboo by religious doctrine. He rejected
explanations of natural phenomena in terms of design, to focus instead on the
detailed causes that had cumulated in the emergence of elaborate phenomena over
long periods of time. He believed that relatively simple mechanisms of cause and
effect could lead to amazingly complex and varied results; complicated outcomes
could be explained in terms of a detailed succession and accumulation of step-

by-step causal mechanisms. This doctrine applied to the most sophisticated and
complex outcomes of evolution, such as the eye and human consciousness.

Accordingly there were neither sudden nor miraculous leaps in the evolution of
human intentionality. Like all human attributes, they must have been prefigured
in the species from which humans are descended. In this way the causal origin of
these features is susceptible to explanation. Darwin (1859, p. 208) thus wrote: “A
little dose . . . of judgement or reason often comes into play, even in animals very
low in the scale of nature.”
His devoted follower, Thomas Henry Huxley, had similar views concerning
causality and the aims of science. For Huxley the idea of uncaused and
spontaneous event was absurd and unacceptable. Science was nothing less than an
ongoing endeavor to reveal the causes behind phenomena. Huxley (1894, Vol. 1,
pp. 158–159) opined that the progress of science meant “the extension of the
province of what we call matter and causation.” Similarly, George Romanes (1893,
p. 402) – an academic friend of Darwin and Huxley – argued that Darwinism
seeks to bring the phenomena of organic nature into line with those of inorganic; and therefore
to show that whatever view we may severally take as to the kind of causation which is
energizing in the latter we must now extend to the former . . . the theory of evolution by natural
selection . . . endeavours to comprise all the facts of adaptation in organic nature under the same
category of explanation as those which occur in inorganic nature – that is to say, under the
category of physical, or ascertainable, causation.
The causal mechanisms of evolution identified by Darwin concerned variation,

inheritance and selection. He argued that natural populations were characterized
by slight variations in key features. However, he was largely unaware of the
biological mechanisms through which variety is created and sustained, other than
upholding a belief that characteristics were likely to be inherited from parents
to offspring. Such inheritance created the possibility that some characteristics
might survive for significant periods of time. If so, their viability in enhancing
the fitness of the organisms involved would be tested by natural selection in their
environment.
These ideas of variation, inheritance and selection are quite abstract, and might
apply to a number of systems in which there is a variety of replicating units. Hence
Darwin (1859, pp. 422–423; 1871, Vol. 1, pp. 59–61, 106) himself proposed that
they might apply to the evolution of language, as well as organisms. When he
formulated these principles, the detailed mechanisms immediately became matters
of further scientific exploration and controversy.
Darwinism involves an unrelenting commitment to causal explanation. At the
same time, Darwinism does not provide a complete theory of everything, from cells
to human society. Darwinism provides an over-arching framework of explanation,

but without claiming to explain every aspect or detail. Explanations additional
to natural selection are always required to explain any evolved phenomenon. For
example, natural selection alone cannot explain why some birds have dull, and
others colorful, plumage. Different auxiliary explanations such as camouflage or
competition for mates are required.
The theory of natural selection was Darwin’s supreme achievement.
Nevertheless, the underpinning of this theory is an ontology where every event
has a cause, and he upheld the role of science to search out causal explanations.
This applies to human motivation, as well as to other phenomena. Hence
Darwinism brought not only human evolution, but also the human mind
and consciousness within the realms of science (Richards, 1987). Evolutionary
psychology applies these Darwinian principles to the human brain, and to the mind
it sustains. Darwin himself predicted that his theory would have major implications
for psychology. Darwin (1859, p. 488) wrote: “Psychology will be based on a new
foundation, that of the necessary acquirement of each mental power and capacity
by gradation.” In this domain Darwinism implies that instinct and reason are,
to use Darwin’s phrase, differentiated “by gradation.” It is a task of psychology
to explain characteristic aspects of the human psyche, including the acquisition
of reason and deliberative behavior, in terms of natural selection. Darwinism
thus brought the frontier of scientific enquiry to the inner workings of the
human mind.
In contrast, Hayek, von Mises, Lachmann and Shackle declared that it is not the
role of the social scientist to examine such matters. Shackle’s “uncaused cause” is
inconsistent with Darwinism. While Hayek and von Mises admitted the causation
of beliefs, they declared investigation of such causes as beyond social science. Of
course, even if causation is present, the social scientist does not have to investigate
all causes, such as the laws of physics, for example. But such abstraction would
be valid only if knowledge of the particular evolution of the human mind had
no impact on our understanding of social phenomena. On the contrary, the
next section argues that there is a significant cost to social science for such an
exclusion, in terms of understanding how individuals reason and behave in social

contexts.
4. ON EVOLUTIONARY PSYCHOLOGY AND

INDIVIDUAL REASON
Evolutionary psychologists Leda Cosmides and John Tooby (1994a, p. 68) argued
that human intentionality must be studied in an evolutionary context: “The human
brain did not fall out of the sky, an inscrutable artefact of unknown origin, and
there is no longer any sensible reason for studying it in ignorance of the causal
processes that constructed it.” This has led to a critique of prevailing versions of
rationality and intentionality in the social sciences. Among these is the separation
of thought from its neural and material context. As Denise Cummins (1998, p. 31)
put it: “The Cartesian fantasy is that mind is pure intellect, the engagement in pure
thought for its own sake. But evolution doesn’t work that way.”
Cosmides and Tooby (1994b, p. 327) rejected the widespread assumption “that
rational behavior is the state of nature, requiring no explanation.” They went on to
criticize what they call the Standard Social Science Model, where the mind harbors
general cognitive processes that are “context-independent” or “context-free.” The
key argument in this modern literature is that postulates concerning the rational
capacities of the human brain must give an explanation of their evolution according
to established Darwinian principles of evolutionary biology (Cummins & Allen,
1998).
However, some evolutionary psychologists have gone too far by arguing
that psychology must be reduced to evolutionary biology and suggesting that
psychological and social phenomena are explicable largely in biological terms. In
contrast, another group of evolutionary psychologists has upheld that psychology
must be consistent with our understanding of human biological evolution and have
developed evolutionary psychology in an explicitly non-reductionist manner.4
The “principle of evolutionary explanation” is the injunction that any behavioral
assumption, including in the social sciences, must be capable of causal explanation
in evolutionary terms, or at least be consistent with a scientific understanding of
human evolution (Hodgson, 1998, 2004a). Other sciences are not mere extensions
of biology: but they must be consistent with an acceptable version of it. In
particular, if there are biological constraints or influences on human capacities
or behaviors, then they should be neither contradicted nor negated by assumptions
at the psychological or social levels.
I am responsible for giving the principle of evolutionary explanation a name, but I
am not the originator of the idea. It is found in the writings of Darwin, and suggested
by others, including Thorstein Veblen (1898). Modern evolutionary psychologists

have revived the idea. For example, as noted below, they have provided evidence
that human beings are not particular good at abstract logical arguments. Reflecting
the fact that humans have evolved in social groups, our rational capacities are
greater when logical rules are placed in a social context. More generally, our
knowledge makes use of “modular” intelligence or “fast and frugal” heuristics,
rather than extended, intricate computations that consume as much as possible of
the available information. The human mind bears the marks of its evolutionary
context and origins.5
The principle of evolutionary explanation requires that theories and assumptions

in psychology and the social sciences should be consistent with our knowledge
and understanding of human evolution. The operationalization of this principle is

controversial, but it does seem to point to some key results.
For example, the celebrated Wason tests suggest that humans are much better
at logical problems when they are concretized in terms of social rules (Cosmides,
1989). In the Wason experiments, subjects were given the rule: if a card has a
vowel on one side, then it must have an even number on the other. It was found
that many people are unable to find the best way of checking for violations of this
rule. In contrast, given the problem: if you are under 21 years old then you may
not drink alcohol in a public bar; significantly more people can readily determine
who or what has to be checked to see if the rule is being followed. However,
the two problems have a logically identical structure. Our capacity to reason
logically improves greatly when the logical problem is situated in a social context,
particularly concerning the adherence to social rules. The argument is that humans
have evolved to deal with social rules and social rule violations over millions of
years of existence as a social species.
These arguments, even if from psychology, would seem to have enormous impli-
cations concerning the assumptions that social scientists might make concerning
human rationality. When making assumptions concerning human rule-following
capacities we should take account of insights from evolutionary psychology.
This would be especially the case in the later works of Hayek (1982, Vol. 1, p. 11)
who wrote: “Man is as much a rule-following animal as a purpose-seeking one.” Is
it relevant to ask: what sustains the rule and gives it some durability through time?
Hayek did not provide a sufficiently clear answer, but in discussing the process
of cultural transmission he put emphasis on the role of imitation (Hayek, 1967,
pp. 46–48; 1982, Vol. 3, pp. 155–157; 1988, pp. 21, 24). This might help to explain
how behavioral regularities are reproduced but we still lack a causal explanation
of imitation and rule-following itself. What are the mechanisms involved in the
genesis of action: the transformation of a rule into an act? Hayek (1967, p. 69)
wrote vaguely of the “external stimulus” and the “internal drive,” without giving
us much more to go on. There is a gap in his theory, which could usefully be filled
partly by using insights from evolutionary psychology.
If there are “internal drives” – or “instincts” as others would say – then
some knowledge of them is entirely relevant for the social scientist. While fully
accepting that any such inherited predispositions do not themselves determine
human behavior, and for any individual the weight of enculturation is paramount,
it is nevertheless important to understand the what those predispositions are, and
how ceteris paribus they might bias human behavior in one direction rather than
another. Human purposes are a result of our individual enculturation and unique
personal history, on the one hand, and the mental structures and basic drives that
we have inherited over millions of years of human evolution.
The relevance of this perspective is beginning to be appreciated by some

economists. Experimental economists have long faced allegedly “irrational”
results, such as in trust games and ultimatum games, where human subjects do
not play the optimal strategies predicted by game theory. Consequently, some
leading experimental economists have attempted to explain some “irrational”
cooperative behavior in terms of inherited dispositions towards social reciprocity.
Elizabeth Hoffman, Kevin McCabe and Vernon Smith (1998) cite arguments
from evolutionary psychologists, that humans have evolved mental algorithms for
identifying and punishing cheaters in social exchange, to explain some “irrational”
outcomes in experimental settings.
Consequently, while we must accept that many of these arguments are
controversial, it should be clear that the Darwinian and evolutionary perspective
on the human mind and behavior does have major implications for the social
sciences. Any hermetic compartmentalization of social science from psychology
means a rejection of knowledge that is significant for our understanding of social
and economic phenomena.
5. PLACING HABIT BETWEEN

INSTINCT AND REASON
Understanding how the mind works is vital for the social scientist in building up
a picture of human agency, while at the same time we must acknowledge human
individuality and diversity. Consider, for example, the role of habit. I have argued
elsewhere that habits are central dispositions, and are indispensable for human
action and reason (Hodgson, 1988, 2003, 2004a).6 This perspective dovetails with
modern evolutionary psychology and complies with the principle of evolutionary
explanation.
At the present time, modern evolutionary psychology is in its infancy, and it has
stressed the evolution of instincts rather than habits. The burden of its argument is
that the human psyche is characterized by instincts that have evolved over the last
few hundred thousand years or more. In contrast, while habits can be formed as a
result of instinct-driven behavior, habits themselves are not inherited genetically,
and depend critically on the social and cultural environment. But the rehabilitation
of the concept of instinct has also led to the return of an understanding of habit.
Modern evolutionary psychologists widely acknowledge that their approach is
highly redolent of that of William James (1892). James saw instinct and habit as
complementary, and both as the foundation of human deliberation and reason. In
Britain, Conwy Lloyd Morgan (1896) developed a similar approach. In the modern
context, there is a literature in several respects allied to evolutionary psychology
that links acquired dispositions or habits with instincts.

For instance, Arthur Reber (1993) has argued that implicit learning of an
unconscious character is ubiquitous even in humans and higher animals. This
is partly because higher levels of deliberation and consciousness are recent
arrivals on the evolutionary scene and came after the development of more basic
mechanisms of unconscious learning in organisms. Reber (1993, p. 85) proposed
that: “Once successful forms are established, they tend to become fixed and serve
as foundations for emerging forms.” In addition: “earlier appearing, successful, and
well-maintained forms and structures will tend towards stability, showing fewer
successful variations than later appearing forms.” In other words, the more basic
structures, once established, stabilise and become less changeable than the layers
that are built upon them.7
Accordingly, there was no cause or possibility for evolution to dispense with
habits and instincts once human reasoning emerged. It built upon them, just
as human bipedal physiology built upon the modified skeletal topology of a
quadruped. Earlier structures and processes, having proved their evolutionary
success, are likely to be built upon rather than removed. Hence earlier evolutionary
forms can retain their use and presence within the organism. They will do this when
they form the building blocks of complex further developments. That being the
case, we retain unconscious mental processes that can function independently of
our conscious reasoning. Our layered mind, with its unconscious lower strata,
maps our long evolution from less deliberative organisms. Consistent with the
evolutionary doctrine of continuity, habits and instincts are highly functional
evolutionary survivals of our pre-human past.
A central principle of evolutionary psychology is that assumptions concerning
the human psyche have to be at least consistent with, and if possible grounded
on, our theoretical and empirical knowledge of human evolution. Accordingly, if
the focus of evolutionary psychology were widened from instincts alone to the
psychological intermediations between instinct and reason, then the principles
in the previous two paragraphs come into play. Just as evolutionary psychology
insists that we should consider how our inherited dispositions or instincts have
evolved over generations, social scientists have a similar obligation to consider
how acquired dispositions or habits have evolved in individuals and groups in a
social setting. Evolutionary psychology points to, and is consistent with, a social
science that takes habit as well as instinct into account.8
A heuristic test of this broader perspective has been carried out using a computer
simulation (Hodgson & Knudsen, 2004). The simulations show that strength
of habit and processes of habituation can play a vital role alongside rational
deliberation and selection pressure. In our model we chose one of the most
straightforward of conventions: whether the rule is to drive on the right or on
the left of the road. In this model, artificially intelligent “drivers” in “cars” are
programmed to negotiate a circular road configuration along with a number of
other, similar vehicles. To negotiate the road and avoid collision, it would seem
to be rational for each driver at least to consider conformity with the perceived
distribution of traffic to the left and right, and avoidance of cars that are close
ahead. In the model these particular dispositions are fixed at birth, like instincts or
a given preference function. Instincts vary from driver to driver, but are fixed for
life. To these factors, our model adds habit.
All drivers in turn make a (subjective) decision based on the (objective)
information about the traffic ahead. The driver looks a bounded distance ahead
and counts the number of cars in each lane and the number of cars going in each
direction. The decision algorithm combines decision elements that vary according
to the cognitive personality of the driver and the global parameter weights. Based
on this information, and given its behavioral and cognitive dispositions, the driver
decides on which side of the ring to drive in its next move.
The final element to be taken into consideration is the possibility of error. An
error probability variable is fixed at the beginning of the simulation. A random
number generator determines whether each car makes the move opposite to its
subjective evaluation. At this final stage, the left or right inclination of the car in
the upcoming move is determined. Obviously, error can disturb this process of
convergence to a left/right convention.
If two cars collide, then they are removed and replaced with two new cars
with randomly chosen characteristics in a random location on the circular road.
Accordingly, the instinct profile of the surviving cars plays no part in determining
the instincts of the replacement cars. While there is evolutionary selection of both
habits and instincts in our model, the lack of generational transmission means that
instinctive evolution is limited. The reason for this is to isolate the role of habit –
rather than instinct – in aiding convergence in the model.
Many simulations were performed, with many different parameter weights.
Generally, as the strength of habit increases from zero, mean convergence levels
improve, for all levels of error. Increases in the levels of error have a significant
and opposite effect. No other variable emerged to generally improve convergence
in our simulations. In these boundedly rational situations, where drivers are unable
to see the whole ring, habit emerged as the single most significant factor improving
convergence.
The most important result of these simulations concerns the effect of introducing
processes of habituation into the modeling of agent behavior. In a substantial region
of parameter space, strength of habit can increase the systemic rate of convergence
towards a left/right convention. In some circumstances it can also enhance systemic

resistance to error. In short, habit helps agents to deal with uncertainty, complexity
and change.
The model suggests that a crucial role played by habit is to build up and reinforce
an enduring disposition in each agent, concerning the appropriate side of the road
on which to drive, especially in a situation where information concerning the traffic
ahead is limited. The development of habits amounts to an element of endogenous
preference formation. A sequence of similar and repeated behaviors creates in each
agent a habitual predilection, which can stimulate a “belief” or “conviction” that
a particular behavior is appropriate.
Again this is reminiscent of the arguments of the pragmatists, who saw acquired
habits as the basis of firmly held beliefs. For Charles Sanders Peirce (1878, p. 294)
the “essence of belief is the establishment of habit.” Pragmatist philosophy and
psychology see action as forming preferences and purposes, at least as much as
preferences and purposes lead to action (Joas, 1996; Kilpinen, 2000).
Similarly, in our model, habit differs from mere inertia, in that it creates stubborn
“beliefs” in the appropriateness of an action, that weigh heavily in the decision-
making process of each agent. The evolution of an equilibrium convention
depends largely on one set of stubborn “beliefs” triumphing over the other. Once
a stable convention forms, it is encoded in the dispositions of the majority and it
can resist the intrusion of a substantial amount of erratic behavior. Accordingly,
habit is more than a means of economizing on decision-making for individuals,
it is a means by which social conventions and institutions are formed and
preserved.
These arguments suggest that for the social scientist is it not good enough to
take the purposes, preferences or behaviors of individuals as given. Although such
an assumption is legitimate in some contexts, it cannot be adequate for economics
as a whole. The full understanding of central social phenomena such as rule-
following and the formation and institutions – to which Austrian writers from
Menger to Hayek have made a substantial analytical contribution – requires some
understanding of human psychology and the mechanisms of human calculation
and cognition.
6. CONCLUSION
The above arguments undermine the proposition upheld by Hayek, von Mises,
Lachmann and Shackle that the explanation of individual purposes and behaviors is
beyond the scope of social science. Shackle’s more extreme position is undermined
by a Darwinian denial of an “uncaused cause.” The positions of Hayek and von
Mises are very different and cannot be dismissed in such a manner. Instead,
what undermines the exclusion arguments of Hayek, von Mises and Lachmann
is the demonstration that evolutionary psychology has something relevant to say

concerning particular features of the human mind that are relevant to understanding
individual actions and interactions in a social context.
To repeat, this does not mean that abstraction from explanations of human
purposes or preferences is unwarranted in some contexts. What it does mean is
that fuller explanations may require the crossing of the boundary between the
social sciences and psychology that Hayek and others have been so keen to police.
Accordingly, the position proposed here, in the light of modern evolutionary
psychology, is consistent with the general methodological argument of Böhm-
Bawerk. This involves temporarily “isolating . . . different aspects of complex
phenomena” from one another, but without being “content . . . with a partial view,”
without mistaking “the part isolated in thought for the whole,” and moving “as far
as possible” towards a view of the “complete whole.”
In particular, not being “content . . . with a partial view” it is incumbent on the
social scientist to look further into the structures and recesses of the human mind,
appreciating the long evolution of this mind in an interactive social setting over
millions of years, bringing in insights how the mind is mould in specific cultural
contexts, and developing explanations that are relevant for the understanding of
social and economic behavior.
NOTES
1. However, this does not mean that I endorse methodological individualism, or an
exclusively subjectivist approach. See Hodgson (1988, 2004a).
2. See my dispute with Caldwell on this point. Caldwell fails to acknowledge my
appreciation that Hayek accepted that tastes and preferences were variable. But the causal
explanation of such formation or variation was for Hayek beyond any social science; it is
on this point of disciplinary demarcation that I differ with Hayek. Causal explanations of
tastes and preferences are not only within economics, but also essential to it (Caldwell,
2001, 2004; Hodgson, 1993, 2004b).

3. This is redolent of Lawson (1997, p. 236) who wrote that “there is literally a world of
difference between leaving something (temporarily) out of focus and treating it as though it
does not exist. The achieving of an abstraction and treating something as though it existed in
isolation are not the same thing at all.” However, Lawson prefers to use the word abstraction
rather than isolation in this context.
4. For instance, Plotkin (1994, p. 176) wrote: “What saves intelligent behaviour from
such a reductionistic account is the presence of selectional processes in the mechanism of
intelligence.” In sum, “intelligent behaviour . . . cannot be reductively explained by genetics
or genetics and development.” Plotkin and other evolutionary psychologists recognize a
major role for culture, as well as genetic make-up, in human development. Ben-Ner and
Putterman (2000) also argue that evolutionary psychology does not lead us to the biological
reductionism that some of the popularizers have promoted.
5. See Buss (1999), Cosmides and Tooby (1994a, b), Cummins (1998), Gigerenzer et al.
(1999), Plotkin (1994, 1997), Sperber (1996), Todd and Gigerenzer (2000), Weingart et al.
(1997).
6. My position on this issue has affinities with earlier writers such as James (1892),
Morgan (1896), Veblen (1914) and Dewey (1922), and later writers such as Margolis (1987)
and Joas (1996). James, Dewey and Veblen reflect the American pragmatist tradition and
the later writers are part of its modern revival.
7. See also Schank and Wimsatt (1987).
8. Boyd and Richerson (1985) are pioneers of a “dual inheritance” approach that involves
cultural as well as genetic transmission. This is an example of such a joint consideration of
both instinctive and cultural evolution.
ACKNOWLEDGMENTS
The author is very grateful to Roger Koppl and an anonymous referee for remarks
on an earlier version of this paper. This article makes use of some material from
Hodgson (2004a).
REFERENCES
Ben-Ner, A., & Putterman, L. (2000, September). Some implications of evolutionary psychology for
the study of preferences and institutions. Journal of Economic Behavior and Organization,
43(1), 91–99.
von Böhm-Bawerk, E. (1890). The historical versus the deductive method in political economy. Annals
of the American Academy of Political and Social Science, 1 (October), 244–271.
Boyd, R., & Richerson, P. J. (1985). Culture and the evolutionary process. Chicago: University of
Chicago Press.
Bunge, M. A. (1959). Causality: The place of the causal principle in modern science. Cambridge, MA:
Harvard University Press.
Buss, D. M. (1999). Evolutionary psychology: The new science of the mind. Needham Heights, MA:
Allyn & Bacon.
Caldwell, B. J. (2001, July). Hodgson on Hayek: A critique. Cambridge Journal of Economics, 25(4),
539–553.
Caldwell, B. J. (2004, March). Hayekian evolution reconsidered: A reply to Hodgson. Cambridge
Journal of Economics, 28(2), 301–305.
Cosmides, L. (1989). The logic of social exchange: Has natural selection shaped how humans reason?
Studies with the Wason selection task. Cognition, 31 , 187–276.
Cosmides, L., & Tooby, J. (1994a, April–June). Beyond intuition and instinct blindness: Towards an
evolutionary rigorous cognitive science. Cognition, 50 (1–3), 41–77.
Cosmides, L., & Tooby, J. (1994b, May). Better than rational: Evolutionary psychology and the invisible
hand. American Economic Review (Papers and Proceedings), 84(2), 327–332.
Cummins, D. D. (1998). Social norms and other minds. In: D. D. Cummins & C. Allen (Eds), The
Evolution of Mind (pp. 30–50). Oxford and New York: Oxford University Press.
Darwin, C. R. (1859). On the origin of species by means of natural selection, or the preservation of
favoured races in the struggle for life (1st ed.). London: Murray.
Darwin, C. R. (1871). The descent of man, and selection in relation to sex (1st ed., 2 vols). London:
Murray; New York: Hill.
Dewey, J. (1922). Human nature and conduct: An introduction to social psychology (1st ed.). New York:
Holt.
Endres, A. M. (1997). Neoclassical microeconomic theory: The founding Austrian version. London
and New York: Routledge.
Gigerenzer, G., & Todd, P. M. et al. (1999). Simple heuristics that make us smart. Oxford and New York:
Oxford University Press.
Hayek, F. A. (1948). Individualism and economic order. London: George Routledge; Chicago:
Hayek, F. A. (1967). Studies in philosophy, politics and economics. London: Routledge & Kegan Paul.
Hayek, F. A. (1982). Law, legislation and liberty (3 vol. combined ed.). London: Routledge & Kegan
Paul.
Hayek, F. A. (1988). The fatal conceit: The errors of socialism. In: W. W. Bartley III (Ed.), The Collected
Works of Friedrich August Hayek (Vol. I). London: Routledge.
Hodgson, G. M. (1988). Economics and institutions: A manifesto for a modern institutional economics.
Cambridge: Polity Press; Philadelphia: University of Pennsylvania Press.
Hodgson, G. M. (1993). Economics and evolution: Bringing life back into economics. Cambridge, UK:
Polity Press; Ann Arbor, MI: University of Michigan Press.
Hodgson, G. M. (1998, March). The approach of institutional economics. Journal of Economic
Literature, 36(1), 166–192.
Hodgson, G. M. (2003, March). The hidden persuaders: Institutions and individuals in economic theory.
Cambridge Journal of Economics, 27(2), 159–175.
Hodgson, G. M. (2004a). The evolution of institutional economics: Agency, structure and Darwinism
in American institutionalism. London and New York: Routledge.
Hodgson, G. M. (2004b, March). Hayekian evolution reconsidered: A reply to Caldwell. Cambridge
Hodgson, G. M., & Knudsen, T. (2004). The complex evolution of a simple traffic convention: The func-
tions and implications of habit. Journal of Economic Behavior and Organization, 54(1), 19–47.
Hoffman, E., McCabe, K. A., & Smith, V. L. (1998). Behavioral foundations of reciprocity:
Experimental economics and evolutionary psychology. Economic Inquiry, 36(3), 335–352.
Huxley, T. H. (1894). Collected essays (9 vols). London: Macmillan.
James, W. (1892). Psychology: Briefer course. New York: Holt; London: Macmillan.
Joas, H. (1996). The creativity of action. Chicago: University of Chicago Press.

Kilpinen, E. (2000). The enormous fly-wheel of society: Pragmatism’s habitual conception of action
and social theory. Helsinki: University of Helsinki.
Lachmann, L. M. (1969). Methodological individualism and the market economy. In: E. W. Streissler
(Ed.), Roads to Freedom: Essays in Honour of Friedrich A. von Hayek (pp. 89–103). London:
Routledge & Kegan Paul. Reprinted in Lachmann, L. M. (1977). Capital, expectations and the
market process. Introduction by W. E. Grinder (Ed.). Kansas City: Sheed Andrews and McMeel.
Lawson, T. (1997). Economics and reality. London and New York: Routledge.
Margolis, H. (1987). Patterns, thinking and cognition: A theory of judgment. Chicago: University of
Chicago Press.
Menger, C. (1871). Grundsätze der Volkwirtschaftslehre (1st ed.), J. C. B. Mohr, Tübingen. Published
in English in 1981 as Principles of Economics, J. Dingwall (Ed.), B. F. Hoselitz (Trans). New
York: New York University Press.

Mises, L. von (1966). Human action: A treatise on economics (3rd ed.). Chicago: Henry Regnery.
Morgan, C. L. (1896). Habit and instinct. London and New York: Edward Arnold.
Peirce, C. S. (1878). How to make our ideas clear. Popular Science Monthly, 12, January, 286–302.
Reprinted in J. Buchler (Ed.), Philosophical Writings of Peirce (1955). New York: Dover
Publications.
Plotkin, H. C. (1994). Darwin machines and the nature of knowledge: Concerning adaptations,
instinct and the evolution of intelligence. Harmondsworth: Penguin.
Plotkin, H. C. (1997). Evolution in mind: An introduction to evolutionary psychology. Harmondsworth:
Penguin.
Reber, A. S. (1993). Implicit learning and tacit knowledge: An essay on the cognitive unconscious.
Oxford and New York: Oxford University Press.
Richards, R. J. (1987). Darwin and the emergence of evolutionary theories of mind and behavior.
Chicago: University of Chicago Press.
Romanes, G. J. (1893). Darwin and after Darwin: An exposition of the Darwinian theory and a
discussion of post-Darwinian questions (Vol. 1, 2nd ed.). London: Longmans Green.
Schank, J. C., & Wimsatt, W. C. (1987). Generative entrenchment and evolution. In: A. Fine &
P. Machamer (Eds), PSA 1986: Proceedings of the Meeting of the Philosophy of Science
Association (Vol. 7. pp. 33–60). East Lansing, MI: Philosophy of Science Association.
Shackle, G. L. S. (1976). Time and choice. Proceedings of the British Academy, 66, 309–329.
Reprinted in Shackle, G. L. S. (1990). Time, expectations and uncertainty in economics:
Selected essays of G. L. S. Shackle. J. L. Ford (Ed.). Aldershot: Edward Elgar.
Shackle, G. L. S. (1979). Imagination and the nature of choice. Edinburgh: Edinburgh University Press.
Shackle, G. L. S. (1986). The origination of choice. In: I. M. Kirzner (Ed.), Subjectivism, Intelligibility
and Economic Understanding: Essays in Honour of Ludwig M. Lachmann on his Eightieth
Birthday (pp. 281–287). London: Macmillan. Reprinted in Shackle, G. L. S. (1988). Business,
time and thought: Selected papers. S. F. Frowen (Ed.). London: Macmillan.
Sperber, D. (1996). Explaining culture: A naturalistic approach. Oxford: Blackwell.
Todd, P. M., & Gigerenzer, G. (2000, October). Précis of simple heuristics that make us smart.
Behavioral and Brain Sciences, 23(5), 727–741.
Veblen, T. B. (1898, September). The instinct of workmanship and the irksomeness of labor. American
Journal of Sociology, 4(2), 187–201. Reprinted in Veblen, T. B. (1934). Essays on our changing
order. L. Ardzrooni (Ed.). New York: Viking Press.
Veblen, T. B. (1914). The instinct of workmanship, and the state of the industrial arts. New York:
Macmillan.
Weingart, P., Mitchell, S. D., Richerson, P. J., & Maasen, S. (Eds) (1997). Human by nature: Between
biology and the social sciences. Mahwah, NJ: Lawrence-Erlbaum.
HAYEK AND MODERN

EVOLUTIONARY THEORY
Paul H. Rubin and Evelyn Gick
INTRODUCTION
Hayek was a firm believer in the effect of evolution on human behavior. This
was a real advance since he wrote in a time when most social scientists believed
in the “blank slate” (Pinker, 2002) and denied the influence of biology on human
actions. Moreover, Hayek got the basic outline of the problem right. Most of human
existence has been spent in small groups (25–250 members) and many of our tastes
and preferences have evolved in that setting. These tastes and preferences are not
always adapted for modern mass societies with market economies and extensive
division of labor and exchange. However, evolutionary science and particularly its
applications to human beings have advanced since Hayek wrote, and some of the
details of his analysis are no longer consistent with current theories.
In this paper we consider some areas where Hayek wrote and compare his writing
with more modern theories. We consider specifically: the Original Society; the
Evolution of Society; the “New Social Order”; Individual and Group Selection;
Religion; and Political Decisions. In all cases we consider first Hayek’s views, and
then modern views on the same topic.

ISSN: 1529-2134/doi:10.1016/S1529-2134(04)07005-X
79
80 PAUL H. RUBIN AND EVELYN GICK
THE ORIGINAL SOCIETY

Hayek
Hayek’s analysis of the evolution of human societies starts with a historic view of
mental and societal developments. For a better understanding of social phenomena
as described in Hayek’s works we summarize his cognitive theory.1 This theory
shows the relationship between stimulus and response on an individual plane.
Every perceived stimulus or bunch of stimuli must fit into “categories.” These
categories work as a filter in that only categorized or classified stimuli may lead
to an action. After stimuli have passed the process of classification, the human
mind is able to perceive them because they belong to a certain set of categories the
brain processes. Categories, the filters themselves, are formed by “dispositions.”2
Dispositions and categories share the same features: they are genetically inherited
but also the result of individual and societal experience. Dispositions are “general
rules” or “patterns of action” which are usually superimposed by other dispositions
that refer to the current situation of the individual and evoke the individual’s
response. In the very beginning of humankind, as human brains developed a
certain structure and also certain dispositions of behavior, the instincts, developed.
Humans during this stage developed similar dispositions since they shared the
same perception of the environment and its aims, risks, and threats. Dispositions
are closely connected to knowledge: because individuals know their environment,
they know what they perceive (stimuli pass certain categories) and they know
how to react. Thus, dispositions store knowledge. It is necessary to stress that
knowledge is not only knowledge of a specific individual but also knowledge of
the society itself. We observe this kind of knowledge when looking at societal
traditions such as how group members usually act. Dispositions are therefore not
only genetically inherited but also the result of individual and societal experiences
which the individual is part of.
At the time when instincts became genetically fixed, the predominant pattern of
society was a “face-to-face society.” Small groups with 10–15 members, usually
groups of relatives, such as gatherers and hunters, emerged together with strong
links between one another. Instincts played the decisive role in ensuring the best
available adaptation to the prevailing environment.3 Insofar as group members
had similar knowledge about their environment, they shared similar needs and
expectations about the behavior of their group members.4 The dispositions to act
“instinctively” were therefore the best response to the stimuli that reached the
individual. Instincts, the “moral rules” of small groups or the “natural morality”5
also helped control and ensure cooperation among group members. This form of
cooperation had been established by trust; the behavior among the members was
Hayek and Modern Evolutionary Theory 81
dominated by altruism,6 solidarity, sympathy, and group decisions. The regularity

of their actions which were grounded on similar perceptions brought about an
order of the society: the order of the small group.
Modern Theories
Much of Hayek’s argument is congruent with modern theories, but much is not. In
particular, while individuals were more closely genetically related than is true of
societies now, nonetheless, there was more conflict in these societies than Hayek
seems willing to admit. There is both theoretical and empirical evidence of conflict
in early societies. Empirically, Keeley (1996) and Knauft (1991) have documented
high rates of homicide among contemporary societies that are analogous to these
societies. Moreover, many students of human evolution believe that the only
selective force with sufficient power to generate our level of intelligence was
competition with other humans (e.g. Byrne & Whiten, 1988; Humphrey, 1976;
Rubin, 2002; Whiten & Byrne, 1997). This is because such competition can have
positive feedback, and so can generate any level of intelligence that the organism
can support.
Hayek discusses the “herald” or leader. The best evidence is that in early soci-
eties, while there is a leader, he (and it is always a man) is often severely restricted
in power. Boehm (1999) discusses a “reverse dominance hierarchy,” meaning
that subordinates cooperate to limit the power of dominants. If dominants try
to amass too much power (those whom Boehm calls “upstarts”) then subordinates
use tools ranging from ridicule to exit to homicide to constrain him. Decisions
are generally made through consensus, with all adult males participating in the
process. Of course, some dominants will sometimes be successful in obtaining
power. However, it is generally thought that this is an exceptional situation.
Hayek is probably correct in discussing the evolved nature of moral rules and
morality. There is evidence that even chimpanzees, our closest relatives, have
some moral rules (De Waal, 1996) and presumably our common ancestor also had
such rules. They are crudely associated with the common law rules of property,
contract, and crime. For example, in experiments, a chimpanzee will reward another
chimpanzee who helped him acquire some fruit. Monkeys such as baboons do
not behave in this way. Chimps also recognize property rights, and presumably
early humans did as well. Of course there was little property, but, for example, in
chimps an animal that has killed some game has the right to the game, and generally
distributes it in various ways, including giving some to those who helped in the hunt.
A point that Hayek missed is the role of competition over women. Among
early humans, this is an important driving force. Our ancestors were generally
mildly polygynous, based on relative size of males and females. Thus, dominant
males would have more than one mate. This meant that some subordinate
males had no formal access to females. This was a source of tension in early
societies, and remains so in those modern societies that allow more than one wife.
Unlike economic competition, which is positive sum, competition over women
is essentially negative sum, and so this form of competition will not have the
beneficial effects of competition over economic goods. One of the major features
associated with the rise of the West has been the outlawing of polygamy.
THE EVOLUTION OF SOCIETY
Hayek
In a changing environment the old patterns of behavior, the instincts, were not
the best response any more. Gradually trade started playing a crucial role. Trade
developed in the very early stages of mankind (e.g. 30,000 years ago, in the
Paleolithic) even before agriculture (Leakey, 1981, quoted in Hayek, 1988).
Nevertheless, the greatest expansion of trade occurred around 750 to 550 BC
in the Mediterranean region. This expansion would not have been possible
without a certain population density, the key to specialization and division of
labor. The emergence of property rights, which are linked to the division of
labor, was closely connected to trade as the motor of social development. “The
crucial point,” notes Hayek, “is that the prior development of several property is
indispensable for the development of trading, and thereby for the formation of
larger coherent and cooperating structures, and for the appearance of those signals
we call prices” (1988, p. 31). He sets the beginning of property rights in the stage
of human development when hand-crafted tools appeared. Separate ownership on
perishable goods was only a later issue, “as the solidarity of the group weakened
and individuals became responsible for more limited groups such as a family”
(Hayek, 1988, p. 31).
The possibility to trade with non-members of their own group led to the “division
of labor, which implies the mutual adjustment of activities of people who do not
know each other . . . and thus to the division of knowledge” (Hayek, 1979, p. 158).
To explain how establishment of trade changes the original society we return to
Hayek’s cognitive theory and to his concept of knowledge in particular. As we have
seen, members of a society will tend to act in a specific way because of the social
quality of their dispositions. However, final individual actions cannot be predicted
because of the individual quality of the dispositions. Dispositions are therefore
responsible for what is perceived by the individual, and perception and action have
a subjective character. In the age of hunter-gatherers the individual dispositions

are very similar.
We can link this approach to the concept of knowledge. The key element of
Hayek’s theory is that human knowledge is limited. The reason that individuals
have only a limited knowledge about their environment and are able to act only by
referring to a small part of it lies in the fact that the social system, the marketplace
in economic systems, and the nervous system are all complex phenomena. There
are so many variables or circumstances that bring about a result that one cannot
take them all into account (Weimer & Palermo, 1982). But again, what holds for the
dispositions of the members of a small group in early stages of human development
holds also for the knowledge of these individuals: because the environment is not
very complex, individual’s knowledge is very similar. Nonetheless, we should
take in account that dispositions and knowledge could vary from individual
to individual because of the characteristics of dispositions pointed out earlier:
individual experience colors individual dispositions.
It is not just the limited capacity of the brain to perceive and process knowledge
which is important, but also the fact that knowledge is necessarily tied to individual
perceptions and interpretations of how to act: knowledge is memorized in rules of
perception and in rules of conduct. In other words, knowledge will always have a
subjective quality since all information that reaches the mind depends on individual
categories and individual dispositions. Knowledge is also largely dispersed in a
society. Every individual will acquire only so much knowledge as he needs for his
own interpretation of the world and for his own actions. In a more complex world
with contacts to individuals not belonging to their own group, an individual will
build up more individual experience and therefore different knowledge compared
to the other group members. Knowledge, varying from person to person, leads to
a different perception and to different individual cognitive frameworks.
Taking that into account, it follows that trade cannot be based on collective
knowledge.7 It needs the establishment of conditions that permit these activities to
flourish. There is a need for a new mechanism. Private property rights emerge as
a substitution for common ownership. Hayek, following Hume and Smith, points
out that private property is the fundamental principle of cultural evolution (Hayek,
1988, Ch. 2–3). These new phenomena are responsible for a faster growth of
the group, but one of the most important consequences of this process was the
separation of goals for different members. The group and its behavior changed
and a new social order could arise. The gradual replacement of innate responses
by new rules, perceived and learned from outside the group, accounts largely for
human evolution. The instincts themselves did not lead to a more beneficial life.
Their gradual suppression, together with their replacement by new rules of conduct,
led to a new order.
Modern Theories
Hayek gets much of this wrong. He places more weight on economics as

the driving force for social change than it can bear.8 Anthropologists separate
human existence into two fundamental periods. The major division is between
“mobile” and “sedentary” societies, also called “simple” and “complex” (Knauft,
1991) or “egalitarian” and “nonegalitarian” hunter-gatherers, or foragers (Kelly,
1995). Mobile hunter-gatherers are also called “immediate-return” hunter gathers,
meaning that “no surplus is created and resources, especially food, are consumed
on a daily basis” (Kelly, 1995, p. 31). By far the longest period of our existence
as humans was the time spent as mobile foragers. During this period, human
groups were small, there was little social structure, little food storage, and little
division of labor or specialization. The only “occupational specialization” was by
age (Kelly, 1995, Table 8–1, p. 294). However, even in early societies, property
rights were approximately efficient (Bailey, 1992). Such people traveled light, and
did not burden themselves with belongings (Kelly, 1995, p. 296, citing Woodburn,
1980). This was also the pattern of our pre-human ancestors during the EEA, when
they evolved to become Homo sapiens. The study of mobile hunter-gatherers is
germane because this way of life is similar to the lifestyle in which we evolved, and
so we are adapted to this way of living. Thus, insights into the period of nomadic
hunter-gatherers would be applicable for understanding the hardwired pattern of
the mind. Of course, there were many changes during the EEA. For example,
there were alternating periods of glacial advance and retreat. However, the social
environment seems to be more important in explaining human evolution than the
physical environment, and this would have been more constant.
Complex hunter-gatherers differ in many dimensions; specifically, occupational
specialization is “common.” Thus, this distinction (which Kelly attributes to
changing from a mobile to a sedentary life style) is the boundary between
specialized and unspecialized roles for individuals in societies. By the time of
large agricultural societies and the beginning of written history, specialization and
division of labor were common and significant among humans. But this came late
in our existence as humans – probably too late to have left a major mark on our
evolved preferences or intellects.
There are several economically relevant features of the EEA. Societies were
small – about 25–150 individuals. This means that possibilities for specialization
and division of labor were quite limited. Adam Smith’s admonition that “The divi-
sion of labour is limited by the extent of the market” has always been true, and small
societies imply highly limited possibilities for division of labor (Smith, 1776/1904,
Bk 1, Ch. 3). While division of labor by gender is universal among humans (Brown,
1991) there was little in the way of more complex forms of specialization.
Many anthropologists make the same point regarding specialization among mo-
bile hunter-gatherers. With respect to warfare, Keeley (1996, p. 46) indicates that:
“[S]ocieties without specialization in the economic realm were unlikely to develop

specialized warriors or units.” Conflict in ancestral societies was unorganized and
was usually a small raiding party attacking an individual in the rival group (Keeley,
1996; Wrangham & Peterson, 1996). Carneiro (2000, p. 12929) indicates, “[F]ull-
time craft specialists come into being only when the aggregate demand for their
products has reached a certain threshold.” Maynard-Smith and Szathmary (1999,
p. 148) believe that division of labor was uncommon until relatively recent times:
“Populations of, at the most, a few hundred individuals, with little division of
labour, except, probably, that between the sexes, have been replaced by societies
of many millions, dependent on extensive division of labour.” This limited amount
of specialization is consistent with the conclusion of Stiner et al. (1999, p. 193)
that “Low human population densities during most of the Middle Paleolithic
imply that group sizes and social networks were small, which certainly limited the
numeric scope of individual interactions. Under these conditions the possibilities
for evolution of complex sharing and exchange behavior as ways to counter the
effects of unpredictable resource supplies would have also been quite limited.”9
Because societies were mobile, all possessions had to be portable and there
was little capital accumulation. Our ancestors used wood, stone or bone tools
and perhaps some other simple implements, but all of these had to be moveable.
“They tend to use portable, utilitarian, easily acquired, replaceable artifacts . . . and
avoid those which are fixed in one place, heavy, elaborately decorated, require
prolonged manufacture, regular maintenance, joint work by several people, or any
combination of these” (Woodburn, 1980, p. 99, quoted in Kelly, 1995, p. 296).
Since there was little capital, there was little value to developing an ability to
understand the productivity of capital – which may explain the intuitive appeal of
various “labor theories of value” that Marxists and others have adopted.
There was very little technological advance. The pace of technological change
in early societies seems amazingly slow by current standards. For example, the
Acheulean hand axe tradition lasted for more than one million years in Africa, Asia
and Europe. In the Upper Paleolithic, about forty thousand years ago (when Homo
sapiens had evolved), “major technological change” is defined as taking place
when a change in stone techniques transpired over “a few thousand years.” The
Gravettian tradition in Europe lasted from about 27,000 to about 12,000 years ago
(all data from Gowlett, 1992). This slow rate of technological change may have
been due to relatively less intelligent pre-human ancestors in the early periods.
More recently, it can be explained as being due to low levels of population and
hence fewer individuals to create new technologies (Jones, 2001; Kremer, 1993;
Simon, 1981/1996) and perhaps poorly defined property rights in innovations
(Jones, 2001). Given this low rate of technical change and innovation, there was no
benefit to evolving a mental mechanism for understanding or rewarding innovation.
There was little growth – so little that no individual would observe any growth
over his/her lifetime (Kremer, 1993). Each person would live and die in a world
of constant technology and income. Thus, there was no incentive to evolve a
mechanism for understanding or planning for growth.
The crucial change occurred when population density became sufficiently large
to induce humans to cease moving and settle in one location. This had numerous
effects, both economic and political. Formal farming began. Population density
began to increase more rapidly. Specialization became important and the amount
of trade, within and between societies, increased. There was storage of goods.
However, this was also the beginning of the era of kings, and of wars and conquest.
Many of these early rulers engrossed very large numbers of females as wives
and concubines (Betzig, 1986). This period describes most of human “history” –
the time when there were written records. It began about 10,000 years ago. We
are still in this period, although recently (perhaps within the last 500 years) the
power of governments in the west has begun to decrease. But the driving force was
population density, which led to increased trade; trade, specialization, and division
of labor were results; they did not drive the process.
THE “NEW SOCIAL ORDER”

Hayek
The more complex a society became the more it had to replace innate responses
with new rules, which were perceived and learned from outside the group. The
instincts themselves no longer led to a beneficial life, but instead it was their
gradual suppression combined with their replacement by new rules of conduct that
brought about a new order. Hayek (1979, p. 164) makes the differences between
the old and new patterns of behavior clear:
The conduct required for the preservation of a small band of hunters and gatherers, and that
presupposed by an open society based on exchange, are very different. But while mankind had
hundreds of thousands of years to acquire and genetically to embody the responses needed for
the former, it was necessary for the rise of the latter that he not only learned to acquire new
rules, but that some of the new rules served precisely to repress the instinctive reactions no
longer appropriate to the Great Society.10
These new rules on which the evolved society is grounded form the “morality of the
large group” or the “morality of the open society.” They include frugality, fairness,
property rights, trade and competition. Hayek’s writing follows David Hume’s
thoughts about the morality of property and honesty, which, in an extended order,
ensures cooperation. Only over time these rules have had become increasingly
established inside societal tradition and had to be transmitted through the cultural
process of socialization. These new rules of morality therefore shaped a part of
culture and are, in contrast with the rules of morality of the small group, constantly
in change.
At this point Hayek refers to the concept of altruism. He rejects the idea that
altruism in connection with small groups is “morally good.” As he points out
(1979, p. 167), there is no thing such as
‘natural goodness,’ because with his innate instincts man could never have build the civilization
on which the numbers of present mankind depend for their lives. To be able to do so, he had to
shed many sentiments that were good for the small band, and to submit to the sacrifices which
the discipline of freedom demands but which he hates. The abstract society rests on learnt rules
and not on pursuing perceived desirable common objects: and wanting to do good to known
people will not achieve the most for the community, but only the observation of its abstract and
seemingly purposeless rules. Yet this little satisfies our deeply engrained feelings, or only so
long as it brings us the esteem of our fellows.
Instead of speaking about altruism in the sense of doing good to known people
or, to put it differently, acting in a way to induce certain outcomes, Hayek sees
altruism in the developed society as acting according to rules which enhance the
extended order. It is not the result of an action that is important, but rather the
adherence to abstract rules, the morality of the large group. Again, his cognitive
theory can help to explain this view: Since in a world of complex phenomena
it is impossible to know all the circumstances which influence the outcome of
an action, only the observation of rules underlying the extended order which
has already been proven to benefit this order makes an action good. “Observing
these rules,” Hayek (1988, p. 81) argues, “enables us to confer benefits beyond
the range of our concrete knowledge.” Hayek’s critique of socialist reasoning
about “social justice” and therefore redistribution of wealth and income appears
to be straightforward. Socialism seems to have a strong impact on the altruistic
feelings which were predominant in the small groups. Therefore he calls
socialism a form of atavism which in the long run will undermine and destroy the
extended order.11
Modern Theories
Hayek is correct in arguing that evolved moral principles are not always in
agreement with moral principles needed to operate a market economy. However,
in some instances he exaggerates the differences. But it is important to keep in
mind the purpose of an economy or a society. This is to satisfy human tastes and
preferences. It is true that these tastes evolved in different situations, and if we
were to design tastes for a market economy with millions of inhabitants, we would
probably design different preferences. But we do not have that option. We have
an evolved set of preferences, and those are the preferences we strive to satisfy.
Many have made this point with respect to food preferences. We evolved in a world
where fat and sugar were scarce, and our ancestors who ate as much of them as
feasible were more likely to survive to be our ancestors (Burnham & Phelan, 2001).
Today we might do better if we had different tastes, but we do not. The result is
that modern Americans spend as much time and effort discussing and considering
food as did our hunter-gatherer ancestors, but, while they discussed ways to get it,
we discuss ways to avoid it.
Hayek is concerned with our taste for altruism. We did evolve such tastes;
humans are much less selfish than economic theory normally supposes. But
our altruistic preferences are more constrained than Hayek seems to believe. In
particular, it has been argued that the competitive conditions of the evolutionary
environment would have allowed only “efficient altruism” – altruism aimed at
solving a prisoner’s dilemma or similar situation (Rubin, 2002, Ch. 3). Preferences
such as those required for Marxism or for the Rawls system could not have evolved
in this environment, and indeed, we do not have those preferences. For evidence
about Marxist preferences, we need only look at the fate of the Marxist societies,
which seemed to be thriving when Hayek was writing but which we have now seen
to be inconsistent with human preferences.
In addition to generalized altruism towards kin, two specific forms of altruism
would have been particularly important in evolutionary times. One would have
been to provide medical assistance to a person who was hurt but would likely
survive with such assistance. The world in which we evolved was much more
dangerous than the world in which we now live, and injuries would have been
much more common. Nursing someone through an injury or illness would have
been an efficient behavior, in the sense that it would have served to increase the
strength and survival ability on one’s band. Thus, today, we treat medical care as
being different from other goods, and this creates dilemmas in a world where the
amount that can be spent on such care is boundless.
Second, it is likely that people innately dislike those who accumulate wealth.
But the main source may be the general nature of zero-sum thinking, rather than the
specific mechanisms identified by Hayek. The popular view of business is generally
negative (Rubin, 2002). Stein (1979) shows that movies generally portray business
in a negative light. A more recent example is from the 1987 movie, Wall Street, in
which a financier engaged in efficiently moving assets from lower to higher valued
uses is depicted as evil. Jobling (2002) has argued that in nineteenth century novels,
characters did not become successful through business practices because audiences
would have viewed success through business as reflecting harmful social behavior.
If wealth in the evolutionary environment was limited and if the main way to be
wealthy was to avoid one’s obligation to share, then attitudes viewing the wealthy
as evil could be part of our mental structure. This is consistent with some religious
beliefs; consider Jesus’ saying that “It is easier for a camel to pass through the eye
of a needle than for a rich man to enter the kingdom of God” (Matthew 19:24). Ayn
Rand did portray businessmen in a favorable light, but her novels, while appealing
to some, have not been major popular successes.
TWO SELECTION PROCESSES: INDIVIDUAL

SELECTION AND GROUP SELECTION
Hayek
Let us give an insight in the selection processes discussed by Hayek. The important
contribution of Hayek, as already discussed in the historical perspective, is that
the individual predisposition to perceive rules from outside the group (society
or subgroups) allows for a process of individual selection as well as group
selection.
The development of the society starts on an individual level. The first
evolutionary process is an individual selection process that refers to the perception
of rules that are slightly different from already existing ones and hence leads to
the creation of new rules. It is seen in Hayek’s cognitive framework, a process that
operates on a subjective plane. The decision to act according to a certain disposition
is the answer to an individual process of trial and error; it is a process in which the
individual responds to an outside stimulus referring or not to the dispositions that
reflect group behavior.
If the individual deviates from the “common” rules of action because of his
awareness of a better rule or because of his increased subjective knowledge and he
is shown to be more successful than when he acted in common, he will be imitated

by other members of the society or his group. The more individuals follow this
process of problem solving, the faster this rule becomes a part of the group’s
dispositions: a new rule which changes the existing social order has emerged.
Vromen (1995) calls this kind of imitation “within-group-imitation.” Likewise,
moral rules may also be accepted directly from other groups. But there are set
limits to individual deviation. The individual selection process is subject to the
degree to which an individual may deviate from rules, which in turn depends
largely on his reputation.12 If the individual cannot change his behavior inside his
group, he may still migrate to a group that either already practices the new rule
in question or offers more tolerance toward deviators. This process, according to
Vromen (1995), is called “between-group-migration.”

A second step in the historical interpretation of cultural development is the
working of a group selection mechanism. Group selection occurs in a process in
which the set of new rules or, to put it differently, in which the changed order,
may lead to more beneficial outcomes for the group as a whole, enhancing the
group’s fitness, compared to the situation before the individual selection process
took place.
Hence, the key criterion in the process of individual selection is individual
fitness, in the case of group selection it is the fitness and thus the growth and
survival of the group.13
Modern Theories
The issue of individual or group selection is a major theme of modern biology and
sociobiology, or evolutionary psychology. While the standard view has been that
group selection is impossible because of free riding, there are more recent theories
that indicate that it may after all be possible (Sober & Wilson, 1998). Zywicki
(2000) has written a lengthy and thoughtful paper about the relationship between
Hayek and this work, and we will not address it further here.
One issue is worth addressing, however. This is the issue of cultural evolution.
The major authorities on this issue are Boyd and Richerson, who have written
voluminously on this issue and on the relationship between cultural and individual
selection. A recent example is Richerson and Boyd (2002). A key point in their
analysis is the role of multiple equilibria. There are infinite numbers of options
available for many cultural artifacts. Humans are imitative animals; imitation is a
useful tool and we are very good at it. Thus, if everyone around us adopts some
cultural practice, most of us will adopt it as well. Then many cultural options could
each be an equilibrium, and individuals will selfishly adopt the particular culture
in which they live. In this sense, there is no conflict between individual and group
selection. But nonetheless group selection can occur because some cultures are
more “fit” than others. That is, individuals can adapt to the culture in which they
live but some cultures will out compete others. Soltis, Boyd and Richerson (1995)
indicate that such evolution might be quite slow, but it can still exist. Moreover,
their critics believe that their estimates may be low. For example, we have recently
observed that western capitalist culture is superior (in the sense of fitness) to
communist culture. This sort of group selection is important in human culture and
is consistent with Hayek’s notion of group selection.
RELIGION
Hayek
As we have mentioned, there are limits to individual rule-setting in the sense

of individual selection processes. We find such limitations in legal settings that
come with costly enforcement. Limitations may also occur on an informal level
through taboos, totems, or religions. Hayek (1988, p. 136) stresses the importance
of these rules because they ensure that “[c]ommon practices . . . have a chance to
produce their beneficial effects on a group on a progressive scale before selection
by evolution can become effective.” Religions therefore make sure that the actual
societal order cannot be corroded by individual trial and error processes. The fear
for punishment, by humans or by God, is a huge obstacle for many individuals.
This fear turns out to be beneficial because only a relatively stable societal order
can undergo the slow process of group selection. A stable social order, a result of
religious beliefs, enables “beneficial traditions [to be] preserved and transmitted
at least long enough to enable those groups following them to grow, and to have
the opportunity to spread by natural or cultural selection” (Hayek, 1988, p. 136).
Group selection also plays a role for the survival of the religions themselves.
As Hayek points out, only those monotheistic religions survived, which supported
property rights and highlighted the importance of the family. Societies practicing
those religions developed further and prospered. Hayek does not trace back this
impact to religion per se, although he speaks in favor of specific religions.14 Would
he do so, he would accept immutable divine rules. Religions in his view are a
means to ensure societal stability for a sufficiently long period to enable selection
processes. Societies and their underlying rules have naturally evolved; they are not
the product of human or supernatural will. All these rules are in constant change.
Modern Theories
Rubin (2002, Ch. 6) addresses the role of religion in human behavior; see also
Guthrie (1993), Boyer (2001) and Wilson (2002). Since religious theories are
untestable, societies can have any set of beliefs. However, some belief systems
will be more successful than others, in that societies holding some beliefs will do
better than societies holding other beliefs. This is an example of the sort of cultural
group selection discussed above.
Religions were originally tribal. Members of a tribe would share beliefs in the
same set of gods. Some religions taught that gods would punish individuals for
certain behaviors. If these punished behaviors were economically inefficient, then
the tribe would prosper. For example, if a religion taught that theft (violation of
property rights) and contract breach (violation of an oath) were bad, then the tribe
with this belief would do better than a tribe with the opposite set of beliefs, or with
no beliefs on such issues. Over time, religions with relatively efficient teachings
should have come to dominate.
Religion is also a marker for tribal membership. That is, in original religions,
those in the tribe were also members of the religion, and outsiders were looked
down upon. Moreover, the moral and efficient principles discussed above would
generally apply only to fellow tribe members. The next major change in religions
was that some religions became more inclusive. They began allowing (or forcing)
conversions, and treated converts as members of the tribe. The two great surviving
religions, Islam and Christianity, both had this characteristic: both allowed
conversion.15 By expanding the set of those with whom trade and commerce were
possible, these more inclusive religions provided net benefits. Moreover, while
there have been religions wars both between members of the same religion but
different branches (Sunnis and Shiites, Catholics and Protestants) and between
different religions, nonetheless, the greater inclusiveness of the modern religions
has probably net reduced the amount of conflict in the world.
POLITICAL DECISIONS
Hayek
Until now, we have limited our observation to individuals acting with respect to
the development of the societal order. We have described societal evolution as a
result of individual selection as well as of group selection. We have so far largely
neglected the importance of governmental action and collective acting regarding
evolutionary processes. In order to understand the role of government from a
classical liberal viewpoint, we need again to refer to the knowledge problem.
Knowledge is widely dispersed in societies. Every individual, though being able
to use more knowledge than he knows by adhering to traditional rules, utilizes

only the amount of knowledge necessary to achieve his plans.
This raises the question of whether there is a mechanism that gathers all the
dispersed information in a society. There are two answers: the government or
the market process. Hayek strongly rejects the first choice and clearly prefers the
latter. A market is a spontaneous order that through prices delivers to individuals
all the information they need; it helps “to utilize the knowledge of many people
without the need of first collecting it in single body” (Hayek, 1952/1979, p. 177). A
market is therefore a transmitting system for dispersed knowledge – “an instrument
for communicating to all those interested in a particular commodity the relevant

information in an abridged and condensed form” (Hayek, 1952/1979, p. 177).
This statement has following implication: if the market system is an instrument of

knowledge transfer and therefore an instrument for achieving additional subjective
knowledge, it should be protected from any interference which distorts the price
mechanism. Here exactly lays the main task of a government.
In order to fulfill this task properly, a government may provide and protect a
legal framework. Legal frameworks offer the “rules of the game” that are valid for
every member in a society and are intended to last for long periods. The framework
consists of the Rule of Law, general principles laid down beforehand that enable
the members of society “to foresee with fair certainty how the authority will use
its coercive powers in given circumstances and to plan one’s individual affairs on
the basis of this knowledge” (Hayek, 1944/1994, p. 80). These rules determine the
conditions under which the available resources may be used. They do not, however,
tell the individual for what ends they should be used. Examples of such rules are
those governing private and criminal law; they also apply to constitutional codes
and the Bill of Rights.
Moreover, there are rules in a society that are not legally fixed. These are called
the moral rules and customs of a society. All these rules, together with public law,
form the framework for individual and governmental action. The proper task of
government in a classical liberal society consists largely of assisting individuals
to achieve their plans and goals based on individual knowledge. To this end,
a government should seek to improve the legal framework wherever possible.
However, it implies that government should not suppress the selection process that
leads to a new framework. The danger behind is that governments are likely to be
much more in favor of an intervention.
The results of a selection process are largely unpredictable since neither
politicians nor a majority of voters may act without cognitive limitations. Whenever
government is not acting as a “government under the law,” its actions may turn out to
be a means of coercion and of suppression of market processes. Interventionism as a
political principle that pursues well-specified outcomes contradicts the mechanism
of any spontaneous order and has therefore been rejected by Hayek. Though Hayek
speaks in favor of interventions designed for well-defined and exceptional cases,
he fears that such policies will eventually lead to an expansion of governmental
action. Governmental interventions themselves bring forth new situations that need
further intervention, leading to a vicious circle.
Political reality shows that the idea of a “government under the law” is an
illusion. Modern democracies are not bound to the will of the voters but they
are “bound to serve the several interests of a conglomerate of numerous groups”
(Hayek, 1979, p. 99). Hayek argues in favor of limiting governmental power,
whether of democratic nature or not. Democracies are usually prone to become

the playground for the many interest groups. Many of today’s existing democracies
are not subject to the division of powers between legislative and government and
thus there is no “government under the law.” Unlimited power usually helps to
ensure reelection and to buy “the votes of particular interests, including those of
some small groups or even powerful individuals” (Hayek, 1979, p. 101).
Together with these concerns, Hayek speaks about the reemergence of
primordial instincts, already explained in the “morality of the small group.” Those
seem to take over the “morality of the large group” which results from selection
processes. As we have seen before, these instincts have been gradually suppressed
during the different stages of societal evolution. However, since people show
more interest in terms like “social justice,” redistribution, and a secure income
for everybody, connected with a deep distrust of unintended outcomes from a
market system, and especially for the market as a knowledge-transmitting system,
the “morality of the small group” could gain more and more power. The danger
of eroding the social order of developed societies is huge. He criticizes a system
in which individuals earn their income not in the market but are part of large
organizations and did therefore not submit themselves to the rules governing
the extended order. He observes (Hayek, 1979, p. 165) that “to them the market
economy is largely incomprehensible; . . . and its results seem to them irrational
and immoral.” As a consequence they call for a “just redistribution” based on what
“everybody deserves”: the seemingly immoral result of group selection is corrected
by government, which in turn is now empowered to fulfill these expectations.
Modern Theories
Political decision making in large scale societies makes use of many of the
mechanisms evolved in smaller societies. This often leads to political mismatches
(Rubin, 2002, Ch. 7). For only one example, the political process pays excessive
attention to identifiable individuals. For example, we may spend large amounts
saving the life of one identified person when the same amount could save the lives
of many more “unidentified” individuals. In the evolutionary environment, our
ancestors lived in small groups of related individuals who would have known each
other as individuals. In a hunting-gathering economy, there would have been ample
scope for fitness enhancing income transfers. For example, if one individual had a
successful hunt, then there might have been more food than he and his immediate
family could consume. In this case, transfers would have benefited recipients
more than they would have harmed donors. Moreover, in a society where storage
of wealth was difficult or impossible, there would have been few incentives for
accumulation, again increasing the benefits of many transfers. In this environment,

there would have been fitness increasing incentives for charity or contributions to
welfare of others.
Members of the group with whom an individual came into contact would
generally have been relatives, so that any transfer increasing the fitness of
the recipient would have been selected for by kin selection. Moreover, since
individuals would have known each other personally, “reciprocal altruism”
(Trivers, 1971) would have been relevant. Note that any individual with whom one
was familiar would have likely been a relative; there would have been no need to
distinguish familiar strangers from relatives. Both of these factors mean that there
would have been selection pressure for income transfers to known, identifiable
individuals. That is, those of our predecessors who transferred resources to
those they knew within the local group would have left more genes in the gene
pool from which we are now selected.
On the other hand, as Moore (1996) points out, there is no reason to expect that
we would have evolved to perform appropriate calculations to maximize fitness
for a large amorphous group of unknown individuals. Such groups would not have
existed in the EEA, and so there would have been no evolutionary incentive to learn
how to maximize for such a group. The result is that we might expect contemporary
humans to be adapted to providing benefits to recognizable, identifiable individuals
rather than to anonymous or statistical individuals, even if the net benefit of
the latter type of transfer is greater than the benefit of transfers to identifiable
individuals. Moore has made the argument that this explains the emphasis of
modern medicine on patient care rather than on prevention, and this argument
seems correct. There are other implications as well.
Many policies provide concentrated benefits to a small number of citizens but
impose diffuse costs on many. In many cases, the aggregate benefits to the few are
much smaller than the sum of the costs to the many. Tariffs generate large incomes
for import competing firms, but all consumers pay higher prices as a result. Such
programs generate deadweight losses. Such programs have been studied in detail,
but there are still puzzles as to their passage. In particular, there is no good theory
explaining which groups are able to obtain benefits. At least part of the answer is in
terms of the power of identified individuals in the political system. When a tariff is
being debated, there are particular workers who will expect to lose, and they know
who they are. Moreover, others observe them and see that they will lose. Rhetoric
often stresses the benefits to these individuals. Potential gainers from abolition
of the tariff are amorphous and anonymous, and so have less weight in political
decision making. Similarly, when a union organizes, we can observe wages of
union members increasing. These beneficiaries are identifiable individuals. Those
persons who are denied jobs because of the higher wages are again anonymous,
and cannot be seen. Indeed, they themselves probably do not know who they
are. Again, the identifiable individuals have a privileged position in the political
process. This is perhaps a result of our intuitions favoring observed individuals

over numerous anonymous persons.
Although Hayek and many economists stress the dangers of excessive
government, it is important to keep these dangers in context. Throughout most
of human existence (the long period of mobile hunter-gatherers) societies (or at
least men) were relatively free. About 10,000 years ago with the rise of sedentary
societies, kings and other leaders began to engross excessive power. Modern
western societies are the freest societies that have existed since hunter-gatherer
times; indeed, since modern societies give women as much freedom as men, these
societies are the freest that have ever existed. Moreover, even within modern
societies there is no clear trend towards increased government. For example,
although economists stress the evils of tariffs, international trade is now freer
that in the past. In the United States at least, the amount of economic regulation
has recently decreased. Nonwestern societies are often much less free than western
societies, and it is important to keep this in mind.
CONCLUSION
Like all evolutionary systems that deal with humans, Hayek’s system is in two parts.
First is the evolution of humans to the point where culture takes over. This is a strict
Darwinian evolutionary system based on natural selection. Hayek’s “dispositions,”
or tastes and preferences, evolved under this regime and these tastes still exist.
Hayek has correctly stressed that many problems of current economic systems
result because humans still have the tastes which served well in the evolutionary
environment but do not function so well today.
Once humans reached the current level of development, cultural evolution took
over. Although Hayek’s theory of cultural evolution has some Darwinian insights,
he emphasizes the Lamarckian rather than the Darwinian character (Hayek, 1988).
The bottom line is that cultural evolution refers to the evolution of institutions and
traditions, not to the biological development of the individual. Furthermore, such
institutions, traditions, and the behavior which are correlated with them, cannot be
inherited but only learnt through education and experience. From this it follows that
an individual learns how to behave not only through his parents but also through
teachers, peers and other persons. As Hayek (1988, p. 25) stresses, “Cultural
evolution operates largely through group selection”: different institutions compete
with one another and the more efficient ones will survive. Nevertheless, the
outcome of this process is not predictable. This makes his theory falsifiable: Hayek
predicts that the group practicing an efficient tradition will develop further and grow
larger. However, it is only in the long run that it will become clear if an institution
is efficient because it takes a long time to observe any reproductive advantage

generated by adhering to efficient institutions. Moreover, the mechanism of group
selection can easily be distorted. Man’s innate instincts show deep scepiticism
against the outcome of open processes. This article has highlighted that these in-
stincts may desire political settings in which governmental power leads to excessive
redistribution and inefficient attempts at correction of all kinds of outcomes.
NOTES
1. See Gick (2003) for a thorough explanation.
2. Dispositions according to Hayek (1978, p. 40) are [t]he most convenient starting
point . . . which makes an organism inclined to respond to stimuli of a certain class, not by
a particular response, but by a response of a certain kind.”
3. From Hayek (1988, p. 11), “These primitive people were guided by concrete,
commonly perceived aims, and by a similar perception of the dangers and opportunities –
chiefly sources of food and shelter – of their environment. They not only could hear their
herald; they usually knew him personally.”
4. See Gick and Gick (2001).
5. See Hayek (1979, pp. 153–176) and Hayek (1988, p. 12).
6. Hayek (1988, p. 12) compares such a behavior with instincts that apply “to the
members of one’s own group but not to others.”
7. As Hayek (1988, p. 89) points out, “trade and commerce often depend importantly on
confidentiality, as well as on specialized or individual knowledge.”
8. Much of what follows is based on Rubin (2003).
9. Ridley (1997) argues that there was substantial specialization. But societies were too
small to support much specialization until much later, when humans became sedentary.
10. Hayek (1988, p. 11) states “[M]an’s instincts . . . were not made for the kinds of
surroundings, and for the numbers, in which he now lives. They were adapted to life in the
small roving bands or troops in which the human race and its immediate ancestors evolved
during the few million years while the biological constitution of homo sapiens was being
formed.”
11. From Hayek (1978, p. 268), “These inherited instincts demand that man should aim
at doing a visible good to his known fellows (the ‘neighbour’ of the bible) . . . The demand
for ‘social justice,’ for an assignment of the shares in the material wealth to the different
people and groups according to their needs or merits, on which the whole of socialism is
based, is thus an atavism, a demand which cannot be reconciled with the open society in
which the individual may use his own knowledge for his own purposes.” See also Hayek
(1988, p. 100) “Here again, although members of a primitive group may readily concede
superior knowledge to a revered leader, they resent it in the fellow who knows a way to
obtain by little perceptible effort what others can get only by hard work. To conceal and
to use superior information for individual or private gain is still regarded as somehow
improper – or at least unneighbourly. And these primitive reactions remain active long after
specialization has become the only way to make use of the acquisition of information in its
great variety.”
12. The possibility of deviation depends on the degree of reputation (Hayek, 1979, p. 204,
fn. 48). “Though present morals evolved by selection, this evolution was not made possible
by a license to experiment but on the contrary by strict restraints which made changes of
the whole system impossible and granted tolerance to the breaker of accepted rules, who
may have turned out a pioneer, only when he did so at his own risk and had had earned such
license by his strict observation of most rules which alone could gain him the esteem which
legitimized experimentation in a particular direction.”
13. The criteria mentioned show Hayek as a follower of rule utilitarianism. See especially
Yaeger (2001, Ch. 4).
14. For an accurate explanation see Gick (2003).
15. The third major religion, Hinduism, does not seek converts, but has grown through
conquest and through the natural population increase of its adherents.
REFERENCES
Bailey, M. J. (1992). Approximate optimality of aboriginal property rights. Journal of Law and
Economics, 35, 183–198.
Betzig, L. L. (1986). Despotism and differential reproduction: A darwinian view of history. Hawthorne,
NY: Aldine de Gruyter.
Boehm, C. (1999). Hierarchy in the forest: The evolution of egalitarian behavior. Cambridge, MA:
Boyer, P. (2001). Religion explained: The evolutionary origins of religious thought. New York: Basic
Books.
Brown, D. E. (1991). Human universals. New York: McGraw-Hill.
Burnham, T., & Phelan, J. (2001). Mean genes: From sex to money to food, taming our primal instincts.
New York: Penguin.
Byrne, R., & Whiten, A. (Eds) (1988). Machiavellian intelligence: Social expertise and the evolution
of intelligence in monkeys, apes, and humans. Oxford: Clarendon Press.
Carneiro, R. L. (2000). The transition from quantity to quality: A neglected causal mechanism in
accounting for social evolution. Proceedings of the National Academy of Sciences, 97, 12926–
12931.
De Waal, F. (1996). Good natured: The origins of right and wrong in humans and other animals.
Gick, E. (2003). Cognitive theory and moral behavior: The contribution of F. A. Hayek to business
ethics. Journal of Business Ethics, 45, 149–165.
Gick, E., & Gick, W. (2001). F. A. Hayek’s theory of mind and theory of cultural evolution revisited:
Towards an integrated perspective. Mind & Society, 3(2), 149–162.
Gowlett, J. A. J. (1992). Tools – The paleolithic record. In: S. Jones, R. Martin & D. Pilbeam (Eds), The
Cambridge Encyclopedia of Human Evolution (pp. 350–360). New York: Cambridge University
Press.
Guthrie, S. (1993). Faces in the clouds: A new theory of religion. New York: Oxford University Press.
Hayek, F. A. (1944/1994). The road to serfdom. Chicago: University of Chicago Press.
Hayek, F. A. (1952/1979). The counter-revolution of science. Indianapolis: Liberty Fund.
Hayek, F. A. (1978). New studies in philosophy, politics, economics and the history of ideas. Chicago:
Hayek, F. A. (1979). Law, legislation and liberty (Vol. 3). Chicago: University of Chicago Press.
Hayek, F. A. (1988). The fatal conceit. London: Routledge.
Humphrey, N. (1976). The social function of the intellect. In: P. P. G. Bateson & R. A. Hinde (Eds),
Growing Points in Ethology (pp. 303–317). Cambridge, UK: Cambridge University Press.
Jobling, I. (2002). Faultless monsters: Positive biases in self-perception and the nineteenth century
novel hero. Paper presented at the 2002 meetings of the Human Behavior and Evolution Society,
Rutgers University.
Jones, C. I. (2001). Was an industrial revolution inevitable? Economic growth over the very
long run. Advances in Macroeconomics, 1(2), Article 1. <http://www.bepress.com/bejm/
advances/vol1/iss2/art1>.
Keeley, L. H. (1996). War before civilization. New York: Oxford University Press.
Kelly, R. L. (1995). The foraging spectrum: Diversity in hunter-gatherer life ways. Washington, DC:
Smithsonian Institution Press.
Knauft, B. (1991). Violence and sociality in human evolution. Current Anthropology, 32, 391–428,
including comments and reply.
Kremer, M. (1993). Population growth and technological change: One million BC to 1990. Quarterly
Journal of Economics, 108, 681–716.
Leakey, R. E. (1981). The making of mankind. New York: Dutton, quoted in Hayek (1988).
Maynard-Smith, J., & Szathmary, E. (1999). The origins of life: From the birth of life to the origins of
language. New York: Oxford University Press.
Moore, R. F. (1996). Caring for identified versus statistical lives: An evolutionary view of medical
distributive justice. Ethology and Sociobiology, 17, 379–401.
Pinker, S. (2002). The blank slate: The modern denial of human nature. New York: Viking Press.
Ridley, M. (1997). The origins of virtue: Human instincts and the evolution of cooperation. New York:
Viking Press.
Rubin, P. H. (2002). Darwinian politics: The evolutionary origin of freedom. New Brunswick, NJ:
Rubin, P. H. (2003, July). Folk economics. Southern Economic Journal, 70(1), 157–171.
Simon, J. (1981/1996). The ultimate resource. Princeton, NJ: Princeton University Press.
Smith, A. (1776/1904). An inquiry into the nature and causes of the wealth of nations. Methuen
and Co., Ltd. Edwin Cannan (Ed.). Library of Economics and Liberty. 15 November 2002.
<http://www.econlib.org/library/Smith/smWN1.html>.
Soltis, J., Boyd, R., & Richerson, P. J. (1995). Can group-functional behaviors evolve by cultural group
selection? An empirical test. Current Anthropology, 36(3 June), 473–494.

Stein, B. (1979). The view from Sunset Boulevard. New York: Basic Books.
Stiner, M. C., Munro, N. D., Surovell, T. A., Tchernov, E., & Bar-Yosef, O. (1999). Paleolithic
population growth pulses evidenced by small animal exploitation. Science, 283(5399),
190–194.
Trivers, R. L. (1971). The evolution of reciprocal altruism. Quarterly Review of Biology, 46, 35–57.
Vromen, J. J. (1995). Economic evolution. An inquiry into the foundations of new institutional
economics. London: Routledge.
Weimer, W. B., & Palermo, D. S. (1982). Cognition and the symbolic processes. Hillsdale: Lawrence
Erlbaum.
Whiten, A., & Byrne, R. W. (Eds) (1997). Machiavellian intelligence II: Extensions and evaluations.
Cambridge, UK: Cambridge University Press.
Wilson, D. S. (2002). Darwin’s cathedral. Chicago: University of Chicago Press.

Woodburn, J. (1980). Hunters and gatherers today and reconstruction of the past. In: A. Gellner (Ed.),
Soviet and Western Anthropology (pp. 95–117). London: Duckworth. Quoted in Kelly (1995).
Wrangham, R., & Peterson, D. (1996). Demonic males: Apes and the origins of human violence. New
York: Houghton-Mifflin.
Yaeger, L. B. (2001). Ethics as social science: The moral philosophy of social cooperation. Cheltenham:
Edward Elgar.
Zywicki, T. J. (2000). Was Hayek right about group selection? 13 Review of Austrian Economics, 81.
HAYEK’S THEORY OF THE MIND
Brian J. Loasby
INTRODUCTION
“It is now becoming widely recognised that many of the central unresolved prob-
lems in economics turn on questions of knowledge” (Loasby, 1986 , p. 41). Nearly
twenty years after that was written, it may be appropriate to take a (necessarily
selective) look at ideas about human knowledge and to suggest some implications
for the practice of economists. The ideas with which we shall begin long
predate the observation that I have just recalled; and the delay in recognising their
implications indicates how the growth of knowledge is dependent on the formation
of appropriate linkages – which of course are not recognised as appropriate until
they have been formed. Adam Smith, Alfred Marshall and Friedrich Hayek were
all confronted with the uncertain basis of knowledge before they began their study
of economics; and what their responses have in common is not only a theoretical
focus on the process by which people develop what we call “knowledge” but also
a reliance on similar kinds of process, which result in the formation of connections
within particular domains. Each author recognises the impossibility of demon-
strating that any such process can deliver proven truth; instead each envisages
sequences of trial and error within particular contexts, leading to the preservation
of what seems to work – until it no longer does, when a new sequence of trial and
error begins. In other words, they all offer evolutionary theories, Marshall and
Hayek explicitly so, while Smith, directly and indirectly, had a major influence on
the development of Darwin’s ideas.

Raffaelli (2003), who was the first to emphasise the significance of Marshall’s
theory of human cognition, has now demonstrated the pervasiveness of its influence
in Marshall’s economics. In this paper it is appropriate to begin with Hayek, both

ISSN: 1529-2134/doi:10.1016/S1529-2134(04)07006-1
101
102 BRIAN J. LOASBY
because of his status among Austrian economists and because he offers the most
elaborate account by an economist of the neurological basis of thought and action.
In the second section we shall then take note of some similarities, differences
and complementarities with Smith and Marshall. (Comparisons with the work of
Daniel Kahneman, Vernon Smith, and Gigerenzer & Selten, 2001 must be left
for another occasion.) In the third section Hayek’s theory will be related to some
recent neuropsychological accounts of the working of the human mind, and the
final section will suggest some implications of the characteristics of the human
mind discussed in the preceding sections for the methods of economic analysis
and for economic policy.
HAYEK’S SENSORY ORDER

The problem which attracted Hayek’s attention was this. “In order to be able to
give a satisfactory account of the regularities existing in the physical world the
physical sciences have been forced to define the objects of which this world exists
increasingly in terms of the observed relations between these objects, and at the
same time more and more to disregard the way in which these objects appear to us”
(Hayek, 1952, pp. 2–3). Not only have sensory qualities been progressively elimi-
nated from this scientific account; they have not been replaced in a way that allows
them to be mapped onto the new categories, but by a distinctive ordering. Thus
“objects which appear alike to us do not always prove to behave in the same way
towards other objects; . . . objects which phenomenally resemble each other need
not be physically similar to each other, and . . . sometimes objects which appear
to be altogether different may prove to be physically very similar” (Hayek, 1952,
pp. 5–6). Hayek accepts the superiority of the physical order as a representation
of relationships within the physical world, including the physical properties of
the human brain, but he does not raise the question why the physical order within
the human brain did not produce a closer reflection of the wider physical order
instead of an apparently inferior system of representations; he simply asks how
this sensory order came into existence. “How” may be thought a more “scientific”
question than “why,” and in this instance it may also be thought to have logical
priority: indeed Hayek’s analysis provides a basis for explaining why, as we shall
see. This explanation will be one of many illustrations of the value of Hayek’s
analysis as a general theory of the creation of mental orders – an explanation of
how the mind works. At a methodological level, it is also a demonstration of the
distinctive role of psychology in understanding mental events (Hayek, pp. 6–7);
and what makes this role distinctive, as Hayek argues, is that mental events
are not simply the imprint of physical events, but representations; moreover the
Hayek’s Theory of the Mind 103
coexistence of sensory and physical orders within a single brain demonstrates

the possibility of alternative coherent representations of a set of events. There
is much more to human behaviour than responses to data – a proposition that

has potentially far-reaching implications for the study of economic activity,
not least for theories of choice.
We may also note that the issue identified by Hayek illustrates the principle
that problems are defined by differences (Pounds, 1969); we react not to states
but to changes in states or disparities between states (notably between perceptions
and expectations). Disequilibrium, but not disorder, is the necessary context for
choice; this is a fundamental Austrian insight. As Vernon Smith (2003, p. 468)
argues in his Nobel Prize Lecture (which opens with quotations from Hume,
Hayek and Simon), “it is necessary to constantly remind ourselves that human
activity is diffused and dominated by unconscious, autonomic, neuropsychologi-
cal systems that enable people to function effectively without always calling upon
the brain’s scarcest resource – attentional and reasoning circuitry.” Focussing
on differences is an economising device; and a natural consequence is that we
normally attach value to changes of wealth rather than states of wealth (in defiance
of decision theorists), as Kahneman (2003, pp. 1454–1458) reminds us in his
Nobel Prize Lecture.
Since the disparity to be explained is that between a classification which is
based on the effects produced by external events on other external events and
a classification based on their effects on our senses, the focus of inquiry is on
systems of relationships, and the key to Hayek’s analysis is the hypothesis that
“causal connexions” in either classification are linked to “structural connexions”
within the human brain. Both are relational orders (Hayek, 1952, pp. 18–19). It
follows that the sensory and physical orders are closely associated with different
neurological networks, and that networks of the latter kind are of relatively
recent origin. The essential point to note here is that connections within the
brain are selective, and so connections between human perceptions and the
physical world (including the physical world of the brain) are also selective;
moreover, being selected within the human brain, which as a physical system is
capable of sustaining alternative connections, they are “subjective” rather than

“objective.”
The characteristic Austrian emphasis on subjectivity therefore has a psycho-
logical, indeed biological, basis. Subjectivity is perfectly compatible both with an
objectively-existing universe and with the possibility of coming to understand that
universe better, though not with any final proof of empirical truth; there is nothing
improbable about the intellectual alliance between Hayek and Popper. There is
great scope for error, because connections may be false or incomplete, and for
sheer ignorance as defined by Israel Kirzner, because the connections which
104 BRIAN J. LOASBY
could have mitigated that ignorance may never have been made; there is also great
scope for imagination and novelty through the making of new connections. The
link between Hayek and Shackle is well founded on their understanding of human
psychology, though this link will not be explored in these pages. The influences
on the formation of connections, and on the possibilities of aligning them with
the external world, then become an important field of study, the results of which
may be significant for policy. We shall return to these implications later.
Since connections are formed within the brain, and are necessarily highly
selective, it might be supposed that individuals could develop patterns of
connections which are so diverse that they fail to understand each other; and
this is not a possibility that we should ignore. However, subjectivity does not
exclude intersubjectivity (Hayek, 1952, p. 23), and Hayek argues that similarities
of experience promote similarities of patterns and perceptions, at the level of the
individual or the species; as we shall see, there are important differences between
the evolutionary processes at these two levels. Smith’s (1976a [1759]) Theory
of Moral Sentiments also rests on such similarities of patterns and perceptions.
Because of the normal connotations of the word “experience” it may be more
appropriate to speak of “construing the replication of events,” a terminology
introduced by the psychologist George Kelly (1963), to whose work we shall refer
several times. This is indeed an accurate definition of the process that is analysed
in Hayek’s neuropsychological theory, for what events are deemed to constitute
a replication is determined by the interpretative framework that is applied to
them, or in physiological terms by the neural pathways that they share. In what
circumstances people are likely to use similar constructions is an issue that we
shall have to consider later, as is the issue – perhaps of greater importance – of
the possibility of understanding substantially different constructions which are
used by other people, and of using such different constructions within a single
coherent economic or social system. Mutual understanding and the growth of
knowledge both face the Duhem-Quine problem: a failure may be attributable to
any part of the system, and so the “cause” of that failure cannot be simply deduced
from the evidence. Any discussion of such issues must be based on some account
of how these interpretative frameworks are formed; and that is the problem that
Hayek explores.
Because its conceptual basis is that of a selectively-connected system, Hayek’s
theory is to be sharply distinguished from general equilibrium models, in which
every element is connected to every other – indeed the completeness of the
connections (the equivalent of a “field theory”) is the basis both for analyses of
the existence and stability of general equilibrium allocations and for claims about
their welfare properties. Potts (2000) has produced an incisive argument that the
incompleteness of their connections is the crucial fact about all economic systems;
the incompleteness of all cognitive systems is the foundation of Simon’s work

on human decision-making and organisational design, and Simon, like Hayek,
insists on the interactions between the external environment and the “internal
environment” (Hayek, 1952, p. 109) of the human brain. The incompleteness
and variability of connections is, of course, perfectly compatible with a unitary
order, on which Hayek (1952, p. 19) insists; but it is also compatible with
multiple unitary orders which may be applied to a single system. The potential
for multiple selective but unitary orders is the conceptual basis of Hayek’s theory
of the mind: its application to the method and content of economics will be
indicated in the final section. A hint of one application may be supplied by
observing that Hayek’s hypothesis of selective connectivity naturally suggests
the need for a process-theoretic explanation of the development of selective but
systematic connections.
This is what Hayek provides. Suggesting a topological isomorphism between
the neural and phenomenological orders (Hayek, 1952, pp. 37, 40), he argues that
instead of direct connections between particular stimuli and particular sensory
qualities, the effect that is produced by any stimulus depends, first, on how (or
indeed whether) it is translated into an impulse in some nerve fibre (Hayek, 1952,
p. 10) and, second, on the location of this impulse in relation to other impulses
within the network of connections that have already been established within
the brain (Hayek, 1952, p. 53). “The transmission of impulses from neuron to
neuron within the central nervous system . . . is thus conceived as the apparatus of
classification” (Hayek, 1952, p. 52). De Vecchi explores the influence on Hayek’s
thinking of gestalt psychology, to which Hayek makes approving references,
though criticising its restriction to the analysis of perception. Gestalt perceptions
are orders that are derived not from the parts but from the relationships between
them; these relationships are “the result of a process of organization . . . performed
by the nervous system” (De Vecchi, 2003, p. 144), which selects a particular com-
bination of connections. Because they are the result of a process of organisation,
“the qualities which we attribute to the experienced objects are strictly speaking
not properties of that object at all, but a set of relations by which our nervous
system classifies them” (Hayek, 1952, p. 143). Hayek immediately and explicitly
draws on Popper’s language to emphasise that “all we know about the world is
of the nature of theories and all ‘experience’ can do is to change these theories”;
in other words, we create a different set of connections. All knowledge, including
“knowledge how” as well as “knowledge that” (Ryle, 1949), is constituted
by connections; it is a particular set of relationships among many other sets
which are technically possible, and any such set is always potentially subject to
replacement – though major changes are not easily achieved, as has frequently
been demonstrated.
106 BRIAN J. LOASBY
Every theory is a set of relationships which is the outcome of a trial and error
process in which theories, and the patterns of neural connections which embody
them, are tested by the effectiveness of the actions to which they lead – or, as
we shall see almost immediately, of their success in interpreting phenomena.
Hayek (1952, p. 19) refers to “the discriminations that we perform,” which he
associates with Ryle’s (1949) category of “knowing how.” The test, of course,
is of sufficiency, not optimality; and these discriminations may be of very many
kinds and very many degrees of precision (Hayek, 1952, p. 71). The perception
that an established theory is inadequate stimulates a search for a better theory;
since criteria of inadequacy are themselves subjective Hayek provides a basis
for Carnegie-type models in which search is stimulated by a disparity between
achievement and aspiration, and in which aspiration levels themselves require
explanation. (Unfortunately, neither Hayek nor Simon recognised the intersections
of their ideas.) Indeed, it is such a process, Hayek argues, that has gradually led
to the supersession, for some important purposes, of sensory theories by physical
theories, though Hayek fails to apply Ryle’s analysis to indicate that this entails a
move from “knowing how” to “knowing that.” We should, however, note that the
sensory order is still often crucial for “knowing that.”
As Hayek points out, this process gives us some reason – though not a
conclusive reason – to expect a closer fit between these physical theories and
the environment, provided that the environment does not change at a faster rate
than the revision of theories – a point to which we shall return when we come to
consider alternative versions of this evolutionary process. However because, as
Hayek emphasises, we must always use theories to interpret experience before
we can use experience to modify theories, existing theories provide both the
conditions which stimulate, or fail to stimulate, the revision of theories and
the starting point for any such revision; thus history matters (Hayek, 1952,
p. 123) for the physical as well as the sensory order, though we need not assume
that it determines unique paths or unique outcomes. Moreover, since all these
theories “are generalizations about certain kinds of events, and since no number
of particular instances can ever prove such a generalization, knowledge based
entirely on experience may yet be entirely false” (Hayek, 1952, p. 168). This, we
should note, is a restatement of David Hume’s objection to induction as a means
of demonstrating empirical truth, and also an endorsement of Popper’s position.
HAYEK, SMITH AND MARSHALL

It is now appropriate to consider the relationship between Hayek’s analysis and
the psychological theories that, as noted in the introduction, were developed early
in their careers by two other economists who were also concerned about the nature
and foundations of human knowledge. It was the sensory order that Hayek set out
to explain; in his exposition of “the principles which lead and direct philosophical
enquiries” Adam Smith (1980 [1795]) had sought to account for the development
of mental representations of the physical order, taking as his principal illustration
the succession of astronomical theories. Smith was familiar with Hume’s simple
demonstration that there could be no procedure for correctly deriving universal
empirical laws from observation or experiment, and consequently no impregnable
basis for deductive reasoning about phenomena. He was also familiar with Hume’s
recommendation that we should turn our attention to the psychological processes
by which people come to treat some empirical propositions as if they were unques-
tionably true; but he seems to have been dissatisfied with Hume’s (1875, p. 41)
simple argument that people are “excited by nature” to believe in “constant con-
junctions,” and developed an explanation which rested on an interaction between
human emotions and the innovative human mind.
Smith argued that it is characteristic of human nature to be uncomfortable when
unable to make sense of some phenomenon, especially when that phenomenon
is repeatedly encountered, and that people therefore attempt to achieve comfort
by the invention of “connecting principles” which will allow phenomena to be
collected into categories and also provide a satisfactory explanation of relation-
ships between categories. Satisfactory explanations are a source of pleasure,
especially when these explanations are aesthetically appealing, and they are
therefore likely to be widely adopted by those encountering such phenomena.
The discomfort occasioned by a subsequent failure to accommodate some new
phenomenon within an established pattern then provides the stimulus to create
a new interpretative system by a rearrangement of connections, which may also
require a redefinition of categories (for example, the set of “planets”). Viewed
from the perspective of Hayek’s analysis, it appears that Smith is explaining how a
physical order emerges from the sensory order, and showing how sensations guide
this process.
That Smith, like Hayek, had a conception of knowledge as a set of replaceable
theories is most strikingly demonstrated by his insistence that Newton’s theories

were the product of Newton’s imagination, not a direct perception of the truth.
As Smith notes, its general acceptance as necessary truth is to be explained by the
rhetorical appeal of its unifying principle, which is also discussed in his Lectures
on Rhetoric (Smith, 1983 , pp. 145–146). However, because Hayek attempts no
detailed explanation of the development of the physical order it is not surprising
that he does not discuss the role of sensory elements as motivators or selection
criteria in that development, which is such a notable feature of Smith’s account.
Smith (1980 [1795], p. 77) even noticed that the desire for theoretical comfort
108 BRIAN J. LOASBY
might induce people to reject part of their sensory order (such as the overwhelming
sensory evidence of a stationary earth), “to preserve the coherence of the ideas
of their imagination.” Though this supersession of elements of the sensory order

was a major element in the problem which Hayek was trying to resolve, his
explanation is focussed on the physiological mechanisms and does not incorporate
the motivational issues that were so important to Smith. There seems to be no
fundamental obstacle to combining the two, though that task is beyond the scope of
this paper.
Because psychology was at that time closely associated with philosophy, it is
not surprising that Smith did not attempt to provide a physiological underpinning
for what we may now call his evolutionary theory of cognition. He did, however,
extend that theory to explain how the division of labour promotes the growth of
knowledge. First, science, which had begun as a series of local corrections to an
apparently fallible sensory order, emerges as an identifiable category of knowledge,
and then, as scientific knowledge expands, specialisation between the sciences
simultaneously increases the range of study within the scientific community and
the attention to detail within each sector; and this attention to detail accelerates the
perception of anomalies which, by causing intellectual discomfort even when they
appear to have no practical significance, stimulate the invention of new “connecting
principles” that may accommodate them. (That practical significance may play no
part in very powerful motivations is amply demonstrated by Giocoli, 2003.) Then
Smith (1976b [1776]) transferred his theory of the growth of knowledge from
science to the economy through his fundamental proposition that the division of
labour, by encouraging a focus on detail which generates problems and stimulates
the development of both stronger and novel connections within particular domains,
has such powerful effects on the growth of knowledge that it becomes the primary
instrument of economic growth. (For an extended account, see Loasby, 2002).
It was this application, not Smith’s underlying psychological theory, that
attracted Marshall’s attention when he turned to economics. Before then, however,
Marshall had already recognised the possibility of a conjunction between contem-
porary associationist psychology and Darwin’s ideas (which, as we shall see later,
owed much to Smith’s emphasis on the advantages of differentiation), and in the

process provided a physical equivalent of Smith’s cognitive theory. Marshall’s
encounter with the problems of knowledge has been explored by Butler (1991),
Groenewegen (1995), and Raffaelli (2003); it may be sufficient to note here that
this encounter was partially prompted by a major intellectual controversy about
the possibility of demonstrating religious truths, which coincided with Marshall’s
own religious doubts.
His response was clearly shaped by Alexander Bain’s (1864, 1865) major
reorientation of psychology from philosophy towards physiology, which had the
unintended effect of making it readily accessible to Darwin’s ideas, as Marshall

quickly realised; he had read The Origin of Species by March 1867 (Groenewegen,
1995, p. 119). Perhaps prompted by his mathematical training, he wondered

how far the psychological processes of the development of knowledge could be
represented by a mechanical system – Smith (1980, p. 66) had also noticed the
resemblances between mechanisms and systems of thought – and devised the most
elaborate model of his whole life in order to investigate this question (Marshall,
1994). In doing so he was consciously following the example of Charles Babbage,
who in turn had been inspired by the decision of the French mathematician Prony to
organise the production of mathematical tables on Smith’s principles of the division
of labour (see Raffaelli, 2003, pp. 52–53). (It seems that Marshall was not aware
at this time of either Smithian connection.) The possibility of reducing biology to
physics is not a recent idea, and the problematic relationship between mechanical
and biological concepts which pervades Marshall’s economic analysis seems
to have its origin here.
Marshall clearly distinguishes between the evolution of “machines” as a species
and the evolution of specific capabilities within individual members of that
species. In the first stage of evolution each machine comprises a “body” which is
capable of receiving impressions from its environment, which he calls sensations,
and performing actions in that environment, and a “brain” which has no direct
connection with the environment and therefore must operate, as in Smith’s and
Hayek’s theories, by forming selective connections. Marshall indicates this by
restricting the brain to operating with “ideas of sensations” and “ideas of actions”;
it works by linking the idea of an initial sensation received by the body with
the idea of an action which the body performs in response, and then linking the
latter with the idea of an sensation that is interpreted as a consequence of that
action. If the latter linkage produces a pleasurable sensation, then the linkage
from initial sensation to action is strengthened, and if the sensation is unpleasant
it is weakened. The suggested mechanism, possibly inspired by Babbage’s
conceptions of analytic engines and automata, to which Marshall refers, is of
wheels connected by bands, which may become tighter or looser in response to the
sensation experienced.
This cumulative trial and error process, which forms associations of contiguity
or similarity, is consistent with Bain’s account of the physiology of mental
phenomena; and Marshall shows how the process could produce complex patterns
of relationships, which would differ between machines (even of identical design)
because of differences in their environments and in the initial formation of
connections. Over time each machine may develop a range of closely connected
sensations and actions, which we might now call routines; these routines are not
the result of anticipatory choice but of environmental selection among actions
110 BRIAN J. LOASBY
which, by Marshall’s intentional specification of his model, cannot originate

in consequential reasoning. The basic mechanism, including the importance
of sensation as initiator of action and moderator of environmental selection,

also corresponds quite closely with Smith’s mentally-focussed account of the
growth of knowledge, though it would correspond even better with an elaborated
account of the process by which the division of labour fosters the development
of capabilities. Indeed it should be noted that in Marshall’s presentation action
is essential to the formation or dissolution of associations; this was to become an
important element in his theory of economic development (Raffaelli, 2003).
Marshall then postulates the emergence of a second level of control within the
brain, which uses similar mechanisms for different purposes (an early example of
exaptation as a postulated evolutionary mechanism). Ideas of sensations received
which have not been linked to any idea of satisfactory action can now be referred
to this higher level, which may generate the idea of a novel action and associate it
with the idea of a sensation produced by its effects. Expectations appear; but they
appear as conjectures. A pleasurable linkage between contemplated ideas of action
and sensation is then transferred to the lower level, where it directs bodily action;
and if the action produces the anticipated sensation the corresponding link between
initial sensation and action forms a new routine. This is a crucial development:
it introduces imagination and the possibility of trial and error within the mind
which may improve the chances of success in the environment, thus opening the
path to modern practices of research and development. Since both the conjectures
generated at this level and the internal selection processes applied to them are not
random but oriented to problems, the course of development is now influenced by
purposive behaviour. This does not conform to modern neoDarwinian principles
of variety generation; but it does not conflict with the broader Darwinian
principle of selection at the practical level, as in Darwin’s own example of
selective breeding.
Marshall’s formulation has substantial virtues as an evolutionary model
which ascribes complementary roles and different evolutionary mechanisms
to the development of the species and development within the individual, and
demonstrates how the division of labour between the two levels of the brain
accelerates the growth of knowledge. The evolution of the brain is clearly a
biological phenomenon, though the sequence envisaged by Marshall may be
explained by the application of basic economic principles to biology. The second
level, which is much more energy intensive, requires the prior development of the
first as an effective survival mechanism and subsequently as a problem-generator;
with this precondition it becomes an important source of potential improvement
in the machine’s performance, achieved at low overall cost in mental energy by
the separation of levels and specialisation between them. The additional effort of
generating and checking ideas is undertaken only when the existing set of routines
has proved inadequate, and does not disturb those elements in the set which
appear to work well; any improvements in performance are stored at the lower
level, and thus cease to require active supervision. It is an efficient mechanism for
making local adjustments, a precursor of Marshall’s partial equilibrium analysis.
Marshall and Hayek have much in common, but also some interesting
differences. Both explain the growth of knowledge by the creation of selective
connections, and both are concerned with mechanisms that make this possible,
though Marshall is content with what we would now call a constructive existence
proof (explaining how it could be done) while Hayek seeks to explain how it is
actually done. In addition, both postulate two processes, which are built of similar
elements but emerge in sequence and produce different results. This common
source of differentiated outcomes is a feature of evolutionary thinking. However,
there is an apparent difference between the pairs. Whereas Marshall’s processes,
though relying on similar mechanisms, necessarily operate at different levels,
Hayek presents the processes of creating the sensory and physical orders as if they
operate at the same level – though he gives very little detail about the formation
of the physical order. A potentially significant difference between their models
is Hayek’s (1952, pp. 123–125) ascription of some degree of purposiveness –
which is linked, as with Marshall, to mental experimentation before action – to
very simple mental orders, whereas Marshall’s machine acquires the capabilities
for experimentation and purpose only when it has evolved to incorporate a higher
level. Because of the costs of experiment and deliberation, these capabilities must
be used sparingly, and actions are normally governed by routine, though many
of these routines are the result of earlier experiment and deliberation. This is not
incompatible with Hayek’s model, though his presentation seems to underestimate
the costs of “purposive adjustment,” which is not necessarily conscious (Hayek,
1952, p. 82). In Marshall’s sequence, the first level is quite capable of developing
a sensory order, but a physical order, because it depends on experimentation and
purpose, requires the second; and Smith’s theory may be invoked to explain why
the physical order, which has to meet the selection test of comforting the imag-
ination, may develop in a different direction. Alternatively, it would be consistent

with Hayek’s exposition to assign the sensory order to genetic control, in which
purpose is a product rather than a cause, and the physical order to development at
the level of the individual.
Smith, Marshall and Hayek all built their systems on the fundamental economic
principle of scarcity; but what is scarce in their systems is human cognitive
capacity and the energy that is necessary to drive it. These are precisely the
only resources that are assumed to be freely available in most formal models in
present-day economics, which thus ignore the most fundamental of all allocation
112 BRIAN J. LOASBY
problems that human beings face. The Chicago objection to regulation rests on the
assumed abundance of entrepreneurship, while the Austrian objection is based on
the importance of incentives to expand the supply (Audretsch et al., 1999, p. 620).
Smith, Marshall and Hayek also effectively, if unintentionally, provide the basis
for explaining why the assumption that cognition alone has no opportunity costs is
maintained by most economists; it is essential to underpin the concept of rational
choice equilibrium (as Herbert Simon often pointed out), and thus, in Smith’s
(1980, p. 77) words (quoted earlier), “to preserve the coherence of the ideas of their
imagination.” The preservation of established structures is an important economis-
ing device, and it is exemplified by the practice of mainstream economists, and
in many well-established communities and organisations. The evolutionary
sequence of Marshall’s machine conforms particularly well to this fundamental
scarcity, even though it predates his interest in economics; as Raffaelli (2003)
shows, the cognitive advantages of specialisation between levels pervade his view
of economic organisation and economic development. Smith’s, Marshall’s and
Hayek’s psychological systems rely on routines and institutions which economise
on cognition, and so do the economic systems that they later considered and
which are populated by human beings who are equipped with such systems.
NEODARWINISM AND NEUROCONSTRUCTIVISM

Hayek’s theory of the formation and modification of mental orders is explicitly
designed to encompass two distinct processes, one of which “takes place in the
course of the development of the single individual” and one “in the course of the
development of the species and the results of which will be embedded in the struc-
ture of the individual organism when it commences its independent life or when it
reaches maturity” (Hayek, 1952, p. 102). The idea of an embedded framework of
the human mind which (correctly) controlled human knowledge of such basic and
universal concepts as space and time was developed, in a non-evolutionary fashion,
by Kant in response to Hume, and it was Herbert Spencer (now so out of favour)
who proposed an evolutionary interpretation of such embedding which would

preserve Kant’s conception of the mind’s power of structuring perceptions against
the claims of extreme empiricists (Raffaelli, 2003, pp. 31–34). Hayek (1952,
p. 166) extends this interpretation by arguing that “experience does not begin with
sensations or perceptions, but necessarily precedes them . . . and the distinction
between sensory qualities, in terms of which alone the conscious mind can learn
about anything in the external world, is the result of such pre-sensory experience.”
It seems natural to ascribe this evolutionary sequence to the species rather
than the individual.
However, Smith had already gone further in observing how “the ideas of the
imagination” could overthrow “the evidence of the senses,” which we might now
interpret as the ability of “the development of the single individual” in some

circumstances to override – though not to replace – the results of “the development
of the species.” The significance of such a capability can hardly be overestimated.
Though its emergence is susceptible to a strictly biological explanation, its effect is
to make possible a human history that is not determined (though it is conditioned)
by biological mechanisms, and so to create space for social sciences which go
beyond biological models. Adam Smith had explored several regions of this
space. In particular, it allows the fundamental principles of evolution – variation,
selection, and replication – to operate in different ways for different contexts. In
contrast to the neoDarwinian prescription, social and economic evolution may
include many selection processes, deliberate as well as natural (though since
deliberate selection cannot be based on rational expectations it may have unin-
tended consequences); it may also incorporate direct influences of environmental
change on the search for novelty, though such novelties cannot be directly derived
from such changes. The evolution of human societies is in some significant
measure to be explained by psychology.
Since Hayek’s specific objective was to explain how the sensory order could
differ from the physical order, it was reasonable for him to leave open the
application of his unifying principle to the distinctive systems of individual and
species development, as Smith left open the application of his unifying principle
of the division of labour to various forms of economic organisation; but it is now
difficult to ignore the important differences between them. Hayek (1952, pp. 60,
102–103) carefully avoids any discussion of these differences, and his presentation
in terms of individual development, which was – and for many of us still is –
easier to connect with our own established schemes of ordering, presumably
explains why Hayek’s theory of species development is so often overlooked.
Some neoDarwinians, however, are very sensitive to the implications of proposing
two distinctive evolutionary processes. They would argue that Hayek’s theory
of development within the lifetime of an individual gives no reason why any
such developed order should be transmitted across generations, whereas the

neoDarwinian transmission mechanism of genetic inheritance can be comfortably
fitted to a theory of the development of species-specific patterns of behaviour.
Hayek’s account of development within the individual may be interpreted
as driven by experience (in Kelly’s sense of the constructions that are imposed
on a sequence of events), both in providing the stimulus to experiment with
new connections and in supplying the criteria for choosing among these new
connections; but in species development the role of “experience” is not to stimulate
experimental changes in mental ordering but to select among changes which
114 BRIAN J. LOASBY
have occurred by random mutations. The double helix is a device for accurate
reproduction, and so all mutations must be technically regarded as mistakes in
copying; and although environmental factors may be allowed to influence the

frequency of mistakes it is a fundamental principle of neoDarwinism that it cannot
influence the kind of mistakes that are made. There is thus no role for modifications
that are prompted by experience; instead a very small fraction of mistakes turn
out to enhance fitness, and these are preserved by accurate copying to succeeding
generations. Experience-led learning by individuals is regarded with suspicion
by neoDarwinians, and it cannot be inherited; our mental orders are genetically
adapted to some past environment, with the era of hunter-gatherers being a current
favourite (see Cosmides & Tooby, 1994).
Indeed we may now observe an emerging conflict for supremacy in the social
sciences between the rival unifying theories of rational choice equilibrium and
neoDarwinian evolution. The two stand in a curious relationship. Both are theories
about selection between alternatives and the preservation of what is selected; and
in both, selection is based on the consequences of those alternatives which are
presented for selection. However, rational choosers, being equipped with rational
expectations, know these consequences in advance, and having made the correct
choices they naturally have no wish to change them, but remain in their equi-
librium state until there is some shock to the economic system. (Their cognitive
system, being already fully connected and therefore perfect, never changes.)
In the neoDarwinian model, by contrast, no-one knows the consequences of
the available alternatives, and any attempt to design alternatives in order to
produce desirable consequences is a pretence that is unworthy of science; but if
neoDarwinian processes can discover the best answer that is currently available
only after trying all existing (though not all possible) alternatives, nevertheless
the best currently available answer will be discovered, and once discovered it will
be conserved in the genetic code, which may then be indistinguishable from an
equilibrium allocation. By appropriate allowance for the costs of this process one
may even be able to make claims for optimality along similar lines to the claims
for optimality, subject to information and transaction costs, that are sometimes
put forward in economics. Thus assumptions which appear to be polar opposites

can, with a little sleight of thought, support identical outcomes.
Now deriving equilibria from the initial data is analytically simpler than
tracing processes, because the stages of these processes are not full equilibria
and are therefore difficult for the modeller to control in a non-arbitrary fashion.
Partial equilibria can be devised, but any particular partial equilibrium is always
open to objection – particularly by those who believe either in rationality or
in the long-term power of neoDarwinian processes. (The standard isolation
of game theoretic models from the wider environment raises dual questions
about the appropriateness of this assumption of environmental irrelevance and

their applicability in a wider domain, which modellers do not always address.)
So we should not be surprised that some evolutionary theorists are attracted to

equilibrium modelling; and one particularly attractive application is the attribution
of particular medical conditions or behaviour to specific genes. The explanation of
performance by structure is a favourite theoretical principle across the disciplines,
and a direct link between final outcomes and the initial data has the dual appeal of
simplicity and plausibility, especially when the initial data can be identified as a
specific gene sequence.
However, there is some resistance to the dominance of this strategy among
neuropsychologists; and the combination of argument and evidence which they
have produced should have particular resonance among social scientists of an evo-
lutionary inclination, especially those who are impressed with Hayek’s reasoning.
The following account is based on a series of papers, some jointly-authored, by
Professor Annette Karmiloff-Smith, Head of the Neurocognitive Development
Unit at University College London, and an acknowledged leader of her profes-
sion. In a lecture to mark the Centenary of the British Psychological Society
(Karmiloff-Smith, 2002), her starting-point is the use by neoDarwinian geneticists
of evidence from adult neuropsychological patients and children with genetic
disorders to support claims that the human brain is organised into specialised
modules which are directed by specialised genes. She offers a fundamental
methodological criticism that will appeal to all Austrians: an exclusive focus
on the relationship between initial conditions and end-states may lead us astray,
and a better understanding of causation requires attention to the processes by
which these end-states are produced.
Her central example is of a genetic disorder, the Williams Syndrome, which is
clearly associated both with the deletion of 17 specific genes and with a specific
set of physical consequences in adults, including a smaller brain volume, an
abnormal size, orientation and density of neurons, and atypical proportions of
several regions of the brain, together with psychological consequences of low
IQ, and low spatial skills, with the notable exception of proficiency in facial
recognition. This combination appears to supply strong prima facie evidence for
an exclusively genetic explanation, and, as she notes, it has been cited (e.g. by
Pinker, 1997, 1999) in support of a theory of the direct determination of behaviour,
including altruism, aggression, intelligence, spatial cognition and language, by
specific genes or specific sets of genes (Karmiloff-Smith, 2002, p. 526).
Such an exclusive explanation is then confronted with further evidence. First,
patients who lack a subset of these 17 genes do not exhibit corresponding
subsets of the symptoms: there is no simple mapping from genes to outcomes.
(Though the sample size is small, universal claims, such as that for exclusive
116 BRIAN J. LOASBY
and specific genetic determination of end-states, may logically be refuted by a

single counter-example; questions about the sample must be questions about the
experimental procedure which has generated an apparent counter-example, not

about its logical implications.) Second, the claim that the apparently unimpaired
proficiency in facial recognition of people with Williams Syndrome demonstrates
an intact face-processing module is undermined by careful experimentation which
revealed that these people were processing faces feature by feature, whereas the
supposed “face-processing module” relies on configuration. (Of particular interest
is the observation that control subjects are equally reliant on processing by feature
when presented with inverted faces; the implications of this will be considered
in the following section.) Differences were also found in the means of producing
some other supposedly-intact skills; thus the “pattern of intact vs. impaired
modules formed from intact vs. mutated genes,” which a theory of purely genetic
determination requires, is removed by “[d]ifferentiating between superficial
behavioural scores and underlying cognitive processes” (Karmiloff-Smith, 2002,
p. 536). Third, experimentation with infants revealed substantial differences from
the results with adults, while the use of infants with Down’s Syndrome as controls
had the incidental effect of demonstrating notable differences between the infant
and adult states of those affected by this syndrome also; such changes in response
during the course of development, implying a reconfiguration of neural networks,
are inconsistent with nativist claims that directly link impaired modules with adult
states (Karmiloff-Smith, 2002, p. 538).
These results do not, of course, overthrow the conception of a genetically
driven evolutionary process, or indeed the argument that many human physical
and behavioural characteristics are genetically determined; but the modified
theory that is offered by Professor Karmiloff-Smith, in conjunction with other
cognitive neuropsychologists, allows scope for “complex pathways from gene-
to-brain-to-cognitive-processes-to-behaviour” (Karmiloff-Smith, 2002, p. 526).
Genetics, and the neoDarwinian model of which they are the focus, retain a major
role in channelling development (e.g. by the formation of linkages as suggested
by Smith, Marshall & Hayek); but there is nevertheless considerable space for
social scientists to develop evolutionary explanations of a somewhat different

kind, for which genetic constraints may provide an appropriate baseline, such as
all evolutionary explanations need. This kind of permissive linkage between dis-
ciplines appears to correspond to Ziman’s view of science. Though commending
“weak” reductionism – the search for underlying commonalities – as a research
strategy, Ziman (2000, pp. 323, 326) objects to “strong” reductionism – the
unification of knowledge by the universal application of fundamental principles,
precisely because no such principles can explain “the spontaneous emergence
of novel modes of order in complex systems”; and these selective connections
produce “a simplification of nature, and of human cognition as naturally evolved,

that actually makes scientific research possible.”
Explanations of the emergence of order, in human brains and in human

societies, are not confined to random mutations and natural selection, though
neither is excluded, but can incorporate the search for novelty, though making
new connections, and choices that are made for what appear to be good reasons,
because they embody plausible connections. They may go beyond this to suggest
why particular reasons may be thought to be good and why searches may be
undertaken in particular circumstances and may proceed in particular directions.
Thus they are not restricted to explaining how people may get things right, but
may also help to understand how they may go astray – and an understanding of the
reasons for failure may have practical uses. The drastic simplifications of assuming
all economic agents to be hard-wired optimisers who are extremely well-informed
(and if confronted with asymmetric information know precisely what are the
implications of what they do not know), which excludes the need for any process
other than Bayesian updating, will not suffice; but the kind of psychology-
based social science developed by Hayek, and also by Smith and Marshall, is
highly congenial.
In fact, the final sentence of Karmiloff-Smith’s lecture would serve as a present-
day introduction to Hayek’s Sensory Order: “The contrasting view [to the static
model of genetic determination of adult states] presented in this lecture is that
our aim should be to understand how genes are expressed through development,
because the major clue to genotype-phenotype relations is not simply in the
genes, or simply in the interaction between genes and environment, but in the
very process of development itself” (Karmiloff-Smith, 2002, p. 540). In other
papers she argues that “on the gene side, the interaction lies in the outcome of the
interacting, cascading effects of interacting genes and their environments and, on
the environment side, the interaction comes from the infant’s progressive selection
and processing of different kinds of input. . . . The child’s way of processing
environmental stimuli is likely to change repeatedly as a function of develop-
ment, leading to the progressive formation of domain-specific representations”
(Karmiloff-Smith, 1998, p. 390).

In a jointly-written paper advocating “an emergentist solution to the Nature-
Nurture controversy,” she and her colleagues emphasise “the extraordinarily
plastic and activity-dependent nature of cortical specialisation.” Because
“cortical regions are likely to differ from the outset in style of computation,
which means that they will also differ in the variety of tasks they can perform
best,” there may be widespread dispositions to convert domain-relevance into
domain-specificity; nevertheless any particular pattern of domain-specificity
is a consequence of development – a particular organisation of the genetic
118 BRIAN J. LOASBY
endowment (Bates et al., 1998). (The argument that localisation of mental

functions does not imply localisation in any particular part of the cortex, and
that alternative pathways may be developed in response to specific damage,

had already been made by Hayek (1952, pp. 147–148), citing Lashley’s (1929)
account of “vicarious functioning” and “equipotentiality.”) Though much is
genetically determined and the remainder is genetically constrained, nevertheless
in important respects “the brain progressively sculpts itself, slowly becoming
specialised over developmental time” (Karmiloff-Smith, 2002, p. 527).
“The expression of genes through development,” rather than entirely by pro-
gramming, may itself be given an evolutionary explanation, as Karmiloff-Smith
(1998, p. 390) notes: “although evolution has pre-specified many constraints on
development, it has made the human neocortex increasingly flexible and open
to learning during postnatal development. In other words, evolution is argued to
have selected for adaptive outcomes and a strong capacity to learn, rather than
prior knowledge. Within such a perspective, it is more plausible to think in terms
of what we might call domain-relevant mechanisms that might gradually become
domain-specific as a result of processing different kinds of input.” There has been
some evolution away from genetically programmed domain-specificity towards
a genetically-enabled multi-specific capability for creating domain-specific skills
through development. Domain-specificity – the division of labour – is a general
characteristic, but some domains may be genetically specified and others may
become specified during the course of development.
Present-day humans therefore embody a partial shift from “evolution in the
course of the development of the species” towards “evolution in the course of
the development of the single individual” – a shift which has been confirmed by
natural selection, but which entails other forms of selection (for a discussion of
some of these, see Loasby, 2001). We should not be surprised that evolutionary
processes should themselves be subject to evolution; this is a feature of both
Smith’s and Marshall’s psychological theories, and it is the source of Hayek’s
problem of the discrepancy between the sensory and physical orders. Human
knowledge and skills develop through the creation and modification of connec-
tions within the brain, for selective connections are the key to human cognition.
If two stimuli are experienced differently, “this difference must be reflected
somewhere in the brain. Every new piece of learning changes the structure of the
brain in some fashion, however minor” (Bates et al., 1998). This is precisely how
learning is modelled by Hayek, and also by Marshall (1994). The development
of a new system of connections that constitutes a physical order, and which at
first supplements and then increasingly supersedes our sensory order in many
contexts, may be seen as a consequence of this major trend in selection within
the human species. This deserves some further consideration.
That specialisation on a particular range of activities would result in a move-

ment from relatively undifferentiated potential to domain-specific knowledge and
capabilities, which could confer distinctive advantages, was Adam Smith’s great
idea. It was applied to the world of nature by Milne-Edwards (1827, p. 534) to
explain the great variety of species, and this gave Darwin the principle which
governed the direction of evolution. We may now draw on our knowledge of
genetics to argue that the detailed specification of a limited range of behaviour
for each species provided evolutionary space for very many different species to
demonstrate that their evolved domain-specific behavioural regularities conferred
sufficient comparative advantage within their specific environment to allow them
to survive, though any comparative advantage may be extinguished by changes
in this environment, including the evolution of other species. Only in the human
species is this specialisation associated with exchange, though the principle of
complementary specialisation is manifest in social insects and in many specific
inter-species relationships – plants and pollinating insects provide the largest
class of examples – and in a broader sense in ecology.
However, speciation is only the first stage of differentiation. The growth of
the pre-human brain allowed for an increasing range of behaviour within each
individual; but what appears to have been a crucial change resulted from a very
rapid increase in brain size between 500,000 and 100,000 years ago. Because
it followed the change to an upright stance, which inhibited enlargement of the
birth canal, this increase could be accommodated only by the birth of infants at
a very early stage of brain development; this made them extremely vulnerable to
both accident and predation for an exceptionally long period, and could therefore
have been selected for only if it was associated with some great advantage. This
advantage, we may now conjecture, seems to have been precisely the ability to
direct this new genetically-guided cognitive capacity to form better representations
of the environment as it was experienced, and to develop more appropriate skills
to deal with it – which is the kind of adaptation cited by Karmiloff-Smith. For
this purpose “the unusually slow period of human postnatal brain development”
(Karmiloff-Smith, 1998, p. 394) is actually an advantage, for the connections in
the brain are being formed while the child is interacting with the environment.
With an appropriate genetic endowment of programmable rather than programmed
capacity, domain-specific skills can be developed within individuals as well as
through the development of the species; and this has some advantages. This de-
velopmental process of theory-revision can cope with much faster environmental
change than reliance on the selection and diffusion of fortuitous genetic muta-
tions, and also with movement into an environment that has not previously been
experienced by that individual. Hayek’s model of development at the individual
level applies.
120 BRIAN J. LOASBY
This interaction between the growing brain and the environment could not
have happened if the development of this larger brain were strictly genetically
determined; but the extraordinary increase in the size of the brain entailed a
far greater proportionate increase in the number of potential connections, and
therefore a formidable challenge to programming capacity. “On mathematical
grounds, it is difficult to understand how 1014 synaptic connections in the human
brain could be controlled by a genome with approximately 108 genes, particularly
when . . . humans share approximately 98% of their genes with their nearest
primate neighbours” (Bates et al., 1998). (This argument corresponds to Hayek’s
(1952, p. 185) proposition that “the capacity of any explaining agent must be
limited to objects with a structure possessing a degree of complexity lower than
its own.”) Instead, “brain development in the higher vertebrates appears to involve
massive overproduction of elements early in life (neurons, axons and synapses),
followed by a competitive process through which successful elements are kept and
those that fail are eliminated” (Bates et al., 1998); this is a non-genetic application
of neoDarwinian evolution to introduce a different evolutionary process.
The loss of genetic control has allowed cognitive development to be shaped by
interaction with particular environments at the level of the individual, on evolu-
tionary principles of variation and selective preservation. Thus the evolutionary
process has itself evolved, as genetic determination has been supplemented by a
genetically-enabled potential for adaptation on a much shorter time scale than ge-
netic evolution (though even within a human lifetime, as Adam Smith realised, the
development of domain-specific skills and habits of thought may lead to dangerous
reductions of adaptability). The evolution of the evolutionary process (though
not precisely so expressed) is also a feature of Adam Smith’s psychological
theory of the growth of knowledge, in which specialisation between individuals,
in both knowledge and capabilities, is a later development that enhances the
effectiveness of the powerful human motivation to create mental models of
puzzling phenomena (Loasby, 2002). The principle that greater diversity requires
a relaxation of central control is familiar in studies of organisational design
and innovation; and it is, of course, a central principle of Austrian economics.
(It is not good news for economists who rely on general equilibrium modelling.)
That this diversity within the human species should apparently be an unintended
consequence of the increase in brain size (even though to a neoDarwinian all
consequences are unintended) should also appeal to an Austrian mindset.
SOME IMPLICATIONS
Let us now consider some of the implications of Hayek’s psychological theory of
the development of human cognitive powers in the light of neuroconstructivism
and the views of the growth of knowledge and capabilities developed by Smith
and Marshall. We shall also refer to George Kelly’s (1963) theory of personality,
which focuses particularly on the problem of preserving the internal coherence

of an individual’s “interpretative system” while simultaneously maintaining
a satisfactory correspondence with the events encountered by that individual
or precipitated by her actions, and in doing so provides a powerful line of
enquiry into biases in and obstructions to learning. Apparently-relevant evidence
may be ignored, and locally-effective explanations may be dismissed, because they
appear incompatible with ways of making sense that have become indispensable
– even in the hardest of sciences, as Ziman (1978) observes. The pathological
imperative to impose such coherence on the theoretical systems of modern
economics without empirical warrant or even empirical reference, which Giocoli
(2003) has clinically examined, is a particularly suitable case for Kelly’s treatment.
Hayek’s principal application of his proposition about the limits of any appara-
tus of classification is to show that “no explaining agent can ever explain objects
of its own kind, or of its own degree of complexity, and, therefore, that the human
brain can never fully explain its own operations” (Hayek, 1952, p. 185); thus,
although we can hope to understand the principles underlying our own mental
processes, “mind must remain forever a realm of its own which we can know only
through directly experiencing it, but which we shall never be able fully to explain
or to ‘reduce’ to something else” (Hayek, 1952, p. 194). This is his conclusion to
his investigation into the problem of psychological explanation; human cognition
is inevitably bounded, as Simon also insisted. Austrians in particular should
recognise the inescapable limits to introspection. Hayek also draws attention to
the impossibility of achieving a full explanation of the world around us, while
simultaneously supplying a principle of organisation for the human brain and for
human societies; and this is the starting-point for the following discussion.
Hayek’s impossibility theorem warns us that our knowledge is necessarily
fallible and incomplete, but it also suggests how it may be improved and tested,
and what kinds of opportunity costs are likely to be incurred along different path-
ways of attempted improvement. Knowledge is created by selecting connections
which will constitute domain-specific modules; and we may identify two general
principles on which to base this selection, which apply to everyday cognitive
operations, to our conscious attempts to construct interpretative frameworks, some
of which we may choose to call theories, and to the design of formal organisations.
One principle directs us towards fine discrimination in our definition of categories,
at the expense of reducing the breadth of our view and ignoring interactions with
the rest of the universe, thus restricting our pattern-making to a narrow domain
which we may be able to explore in some depth. The second principle points
towards the strategy of aggregating the elements of our universe into expansive
invented categories on the basis of similarities that we suppose are significant
122 BRIAN J. LOASBY
for our particular purpose, while ignoring many differences within each category
which we assume to be of little relevance for that purpose (or which we simply
fail to notice), thus creating a domain which is broad but almost empty. Though
each has a physical counterpart in the human brain, all categories are located in
the space of representations, and may be manipulated without further reference
to what they are deemed to represent. Such manipulations may be enlightening,
or misleading; much depends on how they are used (Loasby, 2003).
Normally, there is some accommodation between these two principles, and
our representations are typically sub-systems which include both a few external
connections and a few subdivisions within their components. To illustrate from
formal economics, near one extreme we find theories in which everything of
interest is bundled into a few composite categories, as when a whole economy is
represented by the combination of undifferentiated labour and capital to produce
undifferentiated output, with a unit price for each category; near the other extreme
we find models of precisely-specified games which are isolated from all external
influences – though in a wider scientific perspective, the agents within a game
represent a high degree of aggregation by comparison with the attempt to identify
fundamental particles. Even this supposedly-ultimate objective, we may note,
retains the principle of aggregation within each category of particle; but since, as
Herbert Simon (1982, 2, p. 142) observed, the purpose of theory is to economise
on fact, some aggregation is unavoidable.
Indeed, as Hayek (1952, p. 176) pointed out, nothing can be recognised unless it
can be assigned to some existing category; “we can know only such kind of events
as show a certain degree of regularity in their occurrence in relation to others.”
These regularities, and the classes that they support, are necessarily problematic.
Perhaps the clearest, and prior, statement of this necessary principle of contextual
similarity, and the implicit dangers of ignoring apparently irrelevant differences
in favour of salient resemblances, was provided by Frank Knight (1921, p. 206);
Hayek (1952, pp. 145–146) also emphasises that all classification must be based
on selected elements, so that the resulting “system of acquired connexions . . . will
give only a very distorted reproduction of the relationships” which it purports to
represent, and “will often prove to be false,” generating misleading expectations.

Simon (1982, 2, pp. 306–307) similarly observes that because of the active
filtering involved in both direct perception and the handling of information “the
perceived world is fantastically different from the ‘real’ world.” What are called
“framing effects” are not a sign of failure but a condition of intelligent behaviour;
but of course every framework has a limited “range of convenience” (Kelly, 1963,
p. 11) and outside that range may be seriously misleading. Systematic advantages
are necessarily accompanied by the potential for systematic error; and the
practical test of any interpretative, heuristic or organisational system is whether
its advantages apply to situations which are important or frequently encountered

and its errors to situations which are unimportant or very rare (Gigerenzer &
Selten, 2001).
Of course, environments may change in a way that upsets this balance. Hence
the importance of a procedure for revising, or even replacing, classifications
which no longer seem to work (Hayek, 1952, p. 145), and of a strong intrinsic (and
therefore genetic) motivation to do so; such revisions are the means by which the
physical order began to emerge from the sensory order, and by which new sciences
come into existence. The possibility of revision implies the ability to conceive of
alternative principles of classification on which to construct representations. What
is distinctive, at least in degree, about the human species is that the multifarious
forms of the division of labour among its members have produced such an
unprecedented variety of these representations and so have enormously increased
the total of human knowledge. Hayek’s account of the functioning of the human
brain and neurocognitive theory both lead to the conclusion that human knowledge
is dispersed and incomplete; furthermore, the combination of the wide-ranging
potential of the human brain and the limited capacity of each particular brain to
realise that potential implies that knowledge can become less incomplete only if it
becomes more dispersed. The division of labour exploits the ability of individuals
to create domain-specific networks, each with its appropriate range of convenience
within which systematic advantage dominates systematic error – if they are given
the freedom to so. In currently-fashionable terminology that implies delegation and
empowerment, or in economic language imperfectly-specified contracts; but the
obverse of such discretion is loss of control (and even some loss of understanding),
which to those concerned with the overall efficiency of allocation, either as analysts
or policy-makers, is a serious deficiency. The fundamental reason for this negative
perception is the illusion that the system can be safely treated as if it were fully
connected (Potts, 2000); this is comparable to the illusion that the connective struc-
tures of the greatly-enlarged human brain can be fully specified genetically (though
this is a very different specification, not of a fully connected system but of disjoint
modules). Both illusions exclude uncertainty; but in doing so they also exclude
endogenous innovation.
Asymmetric information has become a major focus of attention in economics;
but because full specification (at least of all contingencies and their implications) is
necessary for the calculation of system optima it is inevitable, though unfortunate,
that the resulting problems are treated as some kind of “organisational failure,”
rather than being part of the pathology of success. Kirzner (1973), by contrast, has
rightly insisted on differential alertness to opportunities as an essential contributor
to economic progress. One important consequence of this prevalent attitude is
an implicit assumption that the co-ordination of dispersed knowledge is simple
124 BRIAN J. LOASBY
if incentives are entirely compatible, whereas there is abundant evidence of the

major contribution of well-intentioned misunderstandings to many failures: for
those of a generous disposition, economists’ recommendations to the transition

economies of eastern Europe may be so classified. The apparently analysable
problems of information have diverted attention from the more fundamental issue
of interpretation; asymmetric interpretation is at once a threat to co-ordination, a
basis for opportunism and a route to innovation. The recent growth of interest in
“knowledge management” may provide an opportunity for a balanced analysis of
the costs and benefits of the growth of knowledge, related to an understanding of
the processes of this growth – but not if the management of knowledge is treated
as primarily a problem of information technology.
It is no accident that the principles and compromises that are inherent in
the use of human mental capabilities are to be found in the organisation of
social, economic and political systems, for the operation of these systems entails
equivalent cognitive problems, which cause us to rely on abstract systems of
rules for the selection and classification of relevant phenomena. As De Vecchi
(1903) points out, Hayek used this equivalence in his later work, and advocated
the dispersion of both political power and economic decision-making; Kirzner
has pursued the theme of domain-specific entrepreneurial alertness and Marshall
(1919, pp. 647–648), though describing the state as “the most precious of human
possessions,” insisted on the importance of confining it to “its special work,” and
applied his cognitive model of conjectured linkages to industrial organisation
(Raffaelli, 2003). For example, a cluster of small firms (especially in an industrial
district) has greater potential for variety than a large firm, especially if the advan-
tages of large-scale are believed to require conformity to routine, and therefore
inhibit the changes in administrative and cognitive organisation that generate
increasing returns. As Quéré (2003, p. 198) points out, “Increasing returns do not
pre-exist. They are the result of an economic process; that is, a result of the way
co-ordination problems are managed over time.” Marshall recognised the connec-
tion between the management of co-ordination problems in the economy and the
management of co-ordination problems within the brain: both require combina-
tions of routines and novelty, and these combinations are themselves modified by
evolutionary processes of trial and error.
Economic growth and the growth of knowledge both entail the division of
labour in order to achieve an effective allocation of resources to the development
of domain-specific cognitive modules within the economy and within society –
indeed within many kinds of “space.” As Darwin learnt from Smith, perhaps indi-
rectly through Milne-Edwards, these are the advantages of the division of labour
that have led biological evolution towards the variety of species; and they have
led human societies towards the variety of knowledge. The genetic specification
of life forms has created many short-lived inefficient allocations of resources

along the way, for only a very small proportion of all possible genetically-induced
specialisations produce any advantages; and the effects of specialisation on the

generation of new knowledge and new capabilities also create many short-lived
inefficient allocations of resources to unsuccessful novelties along the way,
because new knowledge and capabilities arise from the variety of conjectures
within human brains. The imperfectly-specified structure of the human brain
has similar merits to the imperfectly-specified contract of the Coasean firm and
the imperfectly-specified activities of a Hayekian economy, which indeed are
institutions that reflect that structure and enhance its effectiveness. It is important
that the resulting knowledge-domains should also be imperfectly specified: indeed
“domain-specificity,” though adequate to mark the contrast with notions of general
applicability, is a misleading label. In Nelson and Winter’s (1982) evolutionary the-
ory, the primary units of evolution are skills, including skills of organisation, which
are treated as cognitive programmes of limited scope; but Nelson and Winter take
care to emphasise and to illustrate how ambiguous this scope may be, and use this
ambiguity within their theory. Imperfect specification is also a condition of those
experiments at the margin, inspired by differences of temperament and interpreted
experience, on which Marshall relied for the variations that were “a chief cause of
progress” (Marshall, 1920, p. 355), and it is essential for Penrose’s (1959, 1995)
central notion of the imagination of new services to be obtained from resources
and of new productive opportunities to which these services may be directed.
Since increasing attention is being paid to the knowledge content of capital (of
which Marshall was very well aware), it may be helpful to apply to structures of
knowledge Lachmann’s (1978) analysis of capital goods: they are substitutable
between uses but within each use they are complementary to a particular set of
other capital goods when combined in a specific way. In other words they are
multi-specific, and their performance depends on their orientation. Lachmann’s
warning also applies: just as the value of capital cannot be maintained simply by
perpetuating the current set of combinations, so the value of knowledge cannot
be maintained simply by perpetuating its current uses. It is indeed an important
characteristic of knowledge that it can be reused, but what is most important is

that it can be reused in ways that are not simply deducible from current uses –
a consideration which is not prominent in endogenous growth theory, because
it is not easily accommodated within the system of thought to which that theory
belongs. Imagination (which Lachmann rated almost as highly as Shackle) is the
genetically-derived device by which genetic evolution allows the humans species
to exceed the limits of genetic evolution.
The exploitation of our cognitive capabilities is enhanced by a (genetically-
programmed) motivation to search, which is crucial to Smith’s account of the
126 BRIAN J. LOASBY
growth of knowledge; this increases the rate of “mutation” among networks in

what is likely to be the appropriate neighbourhood, though we should not lightly
assume that it will improve the success rate within this neighbourhood. Incentives
(broadly defined) replace instructions as the search space is increased. Since
these are fundamental principles of human behaviour we should not be surprised
that they apply to organisational design and management. However, although we
may welcome the present interest in incentives among economists, we should
note that this interest is heavily concentrated on the delivery of what is already
specified, at the expense of the search for new possibilities, let alone the search for
unrecognised problems. (Requiring applicants for research funds to specify the
outcomes marks the triumph of auditing over innovation.) In addition the analysis
of incentives is generally restricted to a very narrow view of human motivation
(Frey, 2001); here the contrast with Smith is especially striking.
Discretion is important in two ways: it allows the development of substantially
different cognitive structures for different specialisms, and it also allows local
variation and therefore localised progress within each specialism. Both major
and minor differences in the environment make their distinctive contributions
to such development; and these environments include the size and structure of
organisations, which were of especial interest to Marshall. (This is the context in
which he introduced his “law of increasing return”; and that is why he was not
prepared to attenuate it in order to conform to a particular concept of equilibrium.)
At neither level is it possible to prescribe the “best” direction in which to seek
improved knowledge; human consciousness and human purpose, organised in
a variety of cognitive patterns which suggest a variety of profit opportunities,
may lead to a faster rate of variety generation, and may even raise the average
quality, but the growth of knowledge remains a process of trial and error, as Hayek
continued to argue throughout his life. Despite the waste generated in this process,
that is so obvious in retrospect, the necessary superiority of a centrally directed
search for knowledge is no less an illusion than the necessary superiority of
central planning.
The digital revolution in information processing has diverted attention from
the structural nature of knowledge. The evidence on facial recognition presented

earlier is particularly relevant at this point. The motivation to recognise faces is,
we may presume, a shared genetic endowment – its advantages in the formation of
human society (including its importance in controlling opportunism) are obvious;
but it is not linked to a unique facial module. Recalling that recognition by feature
is always employed by those affected by the Williams Syndrome, but also by
those not so affected when they are presented with inverted faces, we may identify
recognition by feature as the default mode; configural recognition is employed
by those who have both the capability to do so and have also been presented with
the material that is necessary to build patterns. Pattern-making is an inherited

capability, which may therefore be impaired by a genetic disorder; how that
capacity is used depends on the environment and individual attempts to make

sense of it. The use of different procedures for upright and inverted faces is also
a demonstration that domains may become very specific through development
at the level of the individual; few people encounter inverted faces frequently
enough to build appropriate patterns by which to identify them, but experiments
with inverting spectacles have shown that it can be done. (There is also a familiar
economic principle at work here; investment in developing the skill of configural
recognition within a specific domain is not justified if this skill is very rarely used.)
Developed capabilities are configurations that economise on cognition; Marshall
(1920, p. 251) explains how someone who has learnt to skate can employ that
knowledge as a unit in constructing more elaborate figures; few economists have
understood why he chose this method of introducing his analysis of industrial
organisation (Raffaelli, 2003). Ziman (2000, p. 120) points out that “pattern
recognition is deeply embedded in scientific practice,” and that the construction,
use and modification of such patterns within each scientific field is a particular
(we may say domain-limited) application of a universal and inter-subjective
human capability.
Patterns provide a basis for extending similarities by physical and mental
experimentation at the margin; and since cognitive patterns differ somewhat
between people there will be different margins at which to experiment. Marshall’s
recognition that these differences in cognitive patterns and in their corresponding
experimental margins are substantially influenced by interactions with particular
environments explains his profound interest in the “linkage between what people
do and how they are changed by doing it,” which Becattini (1991, p. 16) forecast
would justify “a radical repositioning of Marshall from theorist of price and
equilibrium to theorist of industrialization and development.” (Raffaelli (2003)
presents such a radical repositioning.) There will also be different margins at
which knowledge may be most readily absorbed from other people or from written
or electronic sources, for absorption requires incorporation into or amendment of
some configuration; and what is absorbed is likely to differ slightly even between
those who have undergone standardised training. Creative and absorptive capacity
both depend on the ability to make new connections, and are therefore limited by
the connections that already exist; this ability and its potential domain, we may
suppose, are genetically determined, but its use is not. That use, however, follows
the broad evolutionary principles of variation, selection, and retention.
A characteristic of this cognitive theory, as of all evolutionary theories, that is
often overlooked is the intimate dependence of change on the absence of change.
Systematic development is impossible unless there is a stable baseline from which
128 BRIAN J. LOASBY
to begin and a stable environment against which options may be assessed – and
which, in theories such as Hayek’s that allow for deliberate attempts to generate
conjectures, may give direction to these attempts. Smith’s (1980) psychological

theory was identified in its title as an explanation and illustration of “the principles
which lead and direct philosophical inquiries.” The heavens provided a stable
environment, which was subject to improving techniques of observation, and
the sequence of robust interpretative systems examined by Smith was used, two
centuries later, to illustrate Kuhn’s (1962, 1970) theory of scientific development,
which is remarkably close to Smith’s. Routines stabilise evolved patterns, thus re-
leasing mental energy and providing a basis for experiment; this interplay between
routine and innovation, within an individual, a firm, an industry, and an economy,
is a pervasive theme in Marshall’s economics (Raffaelli, 2003). Thus a sensible use
of the concept of equilibrium is to enquire which elements of a system stand in an
equilibrium relationship to each other; for these equilibrium relationships provide
the foundations of change. A natural consequence of this dependence of innova-
tion on stability (which is also essential to neoDarwinian theory) is a substantial
degree of path-dependency within each cognitive domain – including that of a
whole economy, as is indicated by Marshall’s (1919) surveys of national systems;
but this tendency is partially offset by the variety and the quasi-independence of
domains – another consequence of the combined effects of cognitive limitations
and the division of labour.
The counterpart of this quasi-independence is the problem of co-ordination,
which arises in two forms: the compatability of separately-produced knowledge,
and its comprehensibility to those who have not participated in its production.
The division of labour offers to the innovator the protection of cognitive distance;
the integration of what has been divided requires cognitive proximity, though
perhaps only at the points of contact. We should not overlook the effects of our
shared genetic inheritance, emphasised by Hayek, which extends beyond the
substantial component of programmed behaviour to the shared procedures by
which our interpretative frameworks are formed (Ziman, 2000, p. 121), and the
language in which they are (partially) expressed. Not only does intersubjectivity
play “a fundamental role in science”; it is “as basic to the human condition as

subjectivity itself” (Ziman, 2000, pp. 105, 106). Smith’s (1976a) hopes for a
civil society rested substantially on his argument that most people could both
understand and appraise the behaviour of others in situations that were different
from their own. Since then, social and economic evolution, based on an inherited
capacity to create differentiated patterns, has increased the variety of situations
and increased the possibilities of juxtaposing interpretative frameworks that have
few elements or connections in common. Development within the individual
dilutes the shared genetic inheritance of domain-specific behaviours. Cultural
evolution, in particular, may serve either to reinforce or to override the similarities

of attitudes and behaviour embedded in humans on which both Smith and
Hayek relied.
Kelly (1963) construes human personality in terms of the interpretative
frameworks which guide each person’s understanding and behaviour; and the
failure to maintain acceptably coherent interpretative frameworks in a changing
environment thus becomes Kelly’s construction of a personal breakdown.
The disruption of order makes intelligent decision-making impossible. In an
organisational context, this analysis can easily be extended to include problems of
incompatability between the frameworks which seem to apply in the work envi-
ronment and those with which each worker is comfortable in other parts of life, and
also to problems of incompatability between the changes of framework that seem
to be required in different parts of the organisation to cope with major changes
in the organisation’s environment. “A breakdown of corporate personality” may
be an appropriate way to describe what has happened to many organisations
(including the Soviet Union).
We should remember, however, that compatability may be necessary in only
a few dimensions, and systems may be connected along the dimensions where it
is most readily achieved (Ziman, 2002, pp. 302–306). Success may depend both
on bringing some people together and keeping others apart, exploiting selective
cognitive proximity while preserving the cognitive distance necessary to gain
the advantages of specialisation. This may sometimes require the redefinition
of organisational boundaries, as Allyn Young (1928) argued; adaptability – the
capacity to modify connections – may be preserved, and sometimes enhanced, by
rearranging the connections between units or between levels. Such rearrangements
may increase independence or exploit complementarities, thus redefining the
cognitive unit, though not without creating new problems. Richardson’s (1972)
analysis of capabilities along the dimensions of similarity and complementarity
(which is consistent with Hayek’s and Knight’s analysis) provides a basis for
analysing these relationships, but Richardson also notes the special organisational
problems that arise when dissimilar capabilities (embodied in dissimilar neural
networks) are so closely complementary that they require to be jointly managed.

Organisational design is always an option of difficulties.
Although the division of labour is usually associated with the separation of
activities, Smith (1976b, p. 21) recognised not only the benefits of recombination
but also that the selectivity of connections provides scope for those who “are
often capable of combining together the powers of the most distant and dissimilar
objects”; thus integration, he observes, is itself a specialised activity. It is a
particularly important exemplar of the advantages of the division of labour,
because it encourages novel extensions of the results of more narrowly-focussed
130 BRIAN J. LOASBY
specialisation through the development of new configurations. The skills of in-

tegration across domains are themselves domain-limited, and for each individual
they are normally restricted to particular categories of “distant and dissimilar

objects,” as the careers of “Schumpeterian” entrepreneurs have shown. Witt (1999)
has also pointed to the entrepreneurial role in creating and maintaining compatible
interpretations (a unitary order, in Hayek’s terms) within a firm; because we
all need plausible interpretative schemes, entrepreneurs may thereby crowd out
opportunism. These interpretations must also be, in some degree, differentiated
from those which frame the activities of other firms, and sometimes differentiated
from the shared assumptions of an industrial community; Schumpeter and
Marshall both recognised the importance of the outsider as a source of novelty,
though it was Schumpeter who drew attention to the major co-ordination failures
that may result from the destruction of established knowledge.
CONCLUSION
As Potts (2000) has reminded us, a system consists of elements (which may
themselves be systems) and the connections between them; changes in systems
may therefore be traced to changes in either elements or connections, or to
interactions between these two kinds of change. It is an essential feature of general
equilibrium modelling that any model should be fully connected; consequently
any differences between equilibrium models must be attributed exclusively
to differences in their elements. Unfortunately, complete connections within
these models require significant omissions from the elements of which they are
composed, even in the supposedly complete Arrow-Debreu system. What is
excluded is nothing less than what Arrow called “the costs of running the system”
– everything from the energy costs of cognition, which are a substantial part of
human energy costs, to the creation and operation of markets and organisations.
Not only does this make these models incompetent to handle externalities
and institutions (Coase, 1988, p. 15), both of which are the consequences of
incompletable connections; economising on these costs requires a high degree

of selectivity among connections. The elements and the connections in a model
cannot, in general, be independently chosen by the modeller.
Structures matter. Marshall’s (1920, p. 139) suggestion that organisation
should be reckoned as a distinct factor of production is not an idiosyncratic idea,
especially as he defined organisation very broadly (almost as the equivalent of
“the costs of running the system”). The idea is particularly appropriate if we are
interested, as Marshall certainly was, in the process of economic development.
Just as perfect reproduction excludes genetic evolution, so perfect organisation
(and, of course, perfect competition) excludes novelty; Marshall (1920, p. 318)

and Young (1928) both identified organisational change as the route to increasing
return. Hayek’s theory of the mind is a theory of connection-building: at the level

of the species, it is part of a general theory of genetic evolution; at the level of the
individual it is a theory of the innovative mind; and if one combines the two we
have a theory of the evolution of the evolutionary process itself, along the lines
suggested by Karmiloff-Smith and her co-authors.
Domain-specificity (more accurately, domain-limitation) is a key concept.
The genome appears to be inherently a method of constructing a system of
domain-specific elements, embodying Smith’s principle of the division of labour;
but with the remarkable enlargement of the human brain we seem to have a partial
but significant supplementation of genetically-determined domain-specificity
(which seems likely to include at least a major part of the sensory order)
by genetically-enabled development of domain-relevant capabilities towards
domain-specificity at the level of the individual. As Hayek argues, this requires
novel – and additional – patterns of connections within the brain; and these
patterns are produced, as Smith and Shackle notably emphasised, by the human
imagination. Though the results of genetic evolution are still pervasive, there
are now significant possibilities for development at the level of the individual to
modify, and even sometimes to override, development at the level of the human
species.
This is by no means all. The mental orders that are created by our imagination
and tested in specific domains are themselves forms of organisation, for all
knowledge is a structure of selected connections. Now although much of our “life-
world knowledge . . . is coded organically into our behaviour, genetic make-up
and bodily form” (Ziman, 2000, p. 299), these created mental orders vary greatly
across individuals. Consequently the knowledge that can be made available in
any human society depends on organisation – which means on particular patterns
of connections – of the kinds listed by Marshall, which exploit the advantages
of similarity (local variation within imperfectly-specified patterns) and of
differentiation (different patterns), which can produce new species of knowledge
incomparably faster than genetic evolution. The creation, distribution and selective
connection of domain-specific modules within the economy is a central issue for
explaining economic development and for effective policy at the level of firms
and governments.
However, economists should never forget opportunity costs. There are important
disadvantages of domain-specificity at all levels. Though it is perhaps somewhat
easier to escape from internally-generated patterns of thought and action than
from those that are genetically determined, it is nevertheless true (as the studies
reported by Karmiloff-Smith confirm) that patterns resulting from development
132 BRIAN J. LOASBY
become increasingly rigid. The reconstruction of a personality to match a changed

environment is a formidable challenge. (As a problem for clinical psychologists,
it stimulated Kelly’s (1963) Theory of Personality). Changing the patterns of

all the members of a group in a way that preserves intra-group compatability
while adjusting to a different environment is even more difficult; reconstructing
an organisation, formal or formal, of any size seems to be impossible without
changing the membership. Penrosian firms, like individual entrepreneurs, may
find that nothing fails like success, because success may entrench belief in the
patterns that appear to have produced it. The coherence of larger societies may
depend on moderating the demands for compatability; for although, as Ziman
(2000, p. 121) observes, the world-wide appeal of some soap operas indicates the
similarity of evolved mental and emotional mechanisms, the power of genetics
is being continually attenuated by individual and social developments which,
though genetically enabled, are not genetically controlled.
In the end, there is no escape from Knightian uncertainty; no procedures for
expanding either theoretical knowledge or practical skills can be proved to be
correct in relation to the total system to which they are to be applied, because our
best representations of this system are necessarily incomplete, and likely to be
erroneous in some unrecognised respects. However, Knightian uncertainty is also
a precondition for novelty, as Shackle kept reminding us; and human cognitive
systems have a distinct, if limited, capacity for creating novelty as an integral
part of their cognitive operations. They may well also have a genetically-based
need to search for novelty, and so a degree of uncertainty (balanced with some
perceived stability) may be a necessity for the survival of the human species, as
well as a potential threat to that survival.
ACKNOWLEDGMENTS
I would like to thank Peter Earl, Roger Koppl and Jack Vromen, together with
an anonymous referee, for their comments on an early draft, without implicating
them in those parts of this version that they do not approve.
REFERENCES
Audretsch, D. B., Baumol, W. J., & Burke, A. E. (1999). Competition policy in dynamic markets.
International Journal of Industrial Organization, 19, 613–664.
Bain, A. (1864). The senses and the intellect (2nd ed.). London: Longmans Green and Co.
Bain, A. (1865). The emotions and the will (2nd ed.). London: Longmans Green and Co.
Bates, E., Elman, J., Johnson, M., Karmiloff-Smith, A., Parisi, D., & Plunkett, K. (1998). Innateness
and emergentism. In: W. Bechtel & G. Graham (Eds), A Companion to Cognitive Science
(pp. 590–601). Oxford: Basil Blackwell.

Becattini, G. (1991). Four meetings about Marshall: Reports, impressions and reflections. Marshall
Studies Bulletin, 1, 5–20.
Cosmides, L., & Tooby, J. (1994). Origins of domain specificity: The evolution of functional
organization. In: L. Hirschfeld & S. Gelman (Eds), Mapping the Mind. Cambridge: Cambridge
University Press.
De Vecchi, N. (2003). The place of gestalt psychology in the making of Hayek’s thought. History of
Political Economy, 35(1), 135–162.
Frey, B. S. (2001). Inspiring economics: Human motivation in political economy. Cheltenham, UK
and Northampton, MA, USA: Edward Elgar.
Gigerenzer, G., & Selten, R. (Eds) (2001). Bounded rationality: The adaptive toolbox. Cambridge,
MA, USA and London, UK: MIT Press.
Giocoli, N. (2003). Modeling rational agents: From interwar economics to early modern game theory.
Cheltenham, UK and Northampton, MA, USA: Edward Elgar.
Groenewegen, P. D. (1995). A soaring eagle: Alfred Marshall 1842–1924. Aldershot, UK and
Brookfield, VT, USA: Edward Elgar.
Hayek, F. A. (1952). The sensory order: An enquiry into the foundations of theoretical psychology.
London: Routledge.
Hume, D. (1875). Essays moral political and literary. T. H. Green & T. H. Grose (Eds). London:
Longmans Green and Co.
Kahneman, D. (2003). Maps of bounded rationality: Psychology for behavioural economics. American
Economic Review, 93(5), 1449–1475.
Karmiloff-Smith, A. (1998). Development itself is the key to understanding developmental disorders.
Trends in Cognitive Sciences, 2, 389–398.
Karmiloff-Smith, A. (2002). Elementary, my dear Watson, the clue is in the genes . . . or is it?
Proceedings of the British Academy, 117, 525–543. Oxford: Oxford University Press for the
British Academy.
Kelly, G. A. (1963). A theory of personality. New York: W. W. Norton.
Kirzner, I. M. (1973). Competition and entrepreneurship. Chicago: University of Chicago Press.
Knight, F. H. (1921). Risk, uncertainty and profit. Boston: Houghton-Mifflin. Reprinted Chicago
1971: University of Chicago Press.
Kuhn, T. S. (1962, 1970). The structure of scientific revolutions. Chicago: University of Chicago
Press.
Lachmann, L. M. (1978). Capital and its structure. Kansas City: Sheed Andrews and McMeel.
Lashley, K. S. (1929). Brain mechanisms and intelligence. Chicago: University of Chicago Press.
Loasby, B. J. (1986). Organisation, competition, and the growth of knowledge. In: R. N. Langlois
(Ed.), Economics as a Process. Cambridge: Cambridge University Press.
Loasby, B. J. (2001). Selection processes in economics. In: K. Dopfer (Ed.), Evolutionary Economics:
Program and Scope (pp. 253–276). Boston, Dordrecht and London: Kluwer.
Loasby, B. J. (2002). The evolution of knowledge: Beyond the biological model. Research Policy, 37,
1227–1239.
Loasby, B. J. (2003). Closed models and open systems. Journal of Economic Methodology, 10(3),
285–306.
Marshall, A. (1919). Industry and trade. London: Macmillan.
Marshall, A. (1920). Principles of economics (8th ed.). London: Macmillan.
134 BRIAN J. LOASBY
Marshall, A. (1994). Ye machine: Research in the history of economic thought and methodology.
Archival Supplement, 4, 116–132. Greenwich, CT: JAI Press.
Milne-Edwards, H. (1827). Nerf. In: M. Bory de Saint-Vincent (Ed.), Dictionnaire Classique de

l’Histoire Naturelle (17 vols). Paris: Rey et Gravier.
Nelson, R. R., & Winter, S. G. (1982). An evolutionary theory of economic change. Cambridge, MA
and London: Belknap Press.
Penrose, E. T. (1959). The theory of the growth of the firm. Oxford: Basil Blackwell.
Penrose, E. T. (1995). The theory of the growth of the firm (3rd ed.). Oxford: Oxford University Press.
Pinker, S. (1997). How the mind works. New York: W. W. Norton.
Pinker, S. (1999). Words and rules: The ingredients of language. New York: Basic Books.
Potts, J. (2000). The new evolutionary microeconomics: Complexity, competence and adaptive
behaviour. Cheltenham, UK and Northampton, MA: Edward Elgar.
Pounds, W. F. (1969). The process of problem finding. Industrial Management Review, 11, 1–19.
Quéré, M. (2003). Increasing returns and competition: Learning from a Marshallian perspective. In:
R. Arena & M. Quéré (Eds), The Economics of Alfred Marshall: Revisiting Marshall’s Legacy
(pp. 182–201). Basingstoke and New York: Palgrave Macmillan.
Raffaelli, T. (2003). Marshall’s evolutionary economics. London and New York: Routledge.
Richardson, G. B. (1972). The organisation of industry. Economic Journal, 82, 883–896.
Ryle, G. (1949). The concept of mind. London: Hutchinson.
Simon, H. A. (1982). Models of bounded rationality (2 vols). Cambridge, MA: MIT Press.
Smith, A. (1976a [1759]). The theory of moral sentiments. D. D. Raphael & A. L. Macfie (Eds).
Oxford: Oxford University Press.
Smith, A. (1976b [1776]). An inquiry into the nature and causes of the wealth of nations (2 vols). R.
H. Campbell, A. S. Skinner & W. B. Todd (Eds). Oxford: Oxford University Press.
Smith, A. (1980 [1795]). The principles which lead and direct philosophican enquiries; illustrated by
the history of astronomy. In: W. P. D. Wightman (Ed.), Essays on Philosophical Subjects (pp.
33–105). Oxford: Oxford University Press.
Smith, A. (1983). Lectures on rhetoric and Belles lettres. J. C. Bryce (Ed.). Oxford: Oxford University
Press.
Smith, V. L. (2003). Constructivist and ecological rationality in economics. American Economic
Review, 3, 465–508.
Witt, U. (1999). Do entrepreneurs need firms? A contribution to a missing chapter in Austrian
economics. Review of Austrian Economics, 11, 99–109.
Young, A. (1928). Increasing returns and economic progress. Economic Journal, 38, 523–542.
Ziman, J. M. (1978). Reliable knowledge. Cambridge: Cambridge University Press.
Ziman, J. M. (2000). Real science: What it is, and what it means. Cambridge: Cambridge University
Press.
IS HOMO ECONOMICUS EXTINCT?

VERNON SMITH, DANIEL
KAHNEMAN AND THE
EVOLUTIONARY PERSPECTIVE
C. Athena Aktipis and Robert O. Kurzban
ABSTRACT
The awarding in October of 2002 of the Nobel Memorial Prize in Economics1
to Daniel Kahneman and Vernon Smith might have profound implications for
the survival of Homo economicus, which has long occupied a privileged
place in the minds of economists and decision-making theorists. The species
has endured many challenges and proven quite adaptable, changing to
accommodate a cascade of findings inconsistent with its original conception.
Homo economicus now faces a potentially more serious challenge: the
resurgence of Homo sapiens, a more coherent and biologically grounded
model for human decision-making, informed by theory and data from across
the scientific spectrum.
We use the term Homo economicus – as others have (e.g. Thaler, 2000) – as
a shorthand for the canonical model of humans as self-interested agents. The
historical conception of Homo economicus is of a rational decision-maker with

perfect information and perfectly ordered preferences. This view has proven
fruitful, yielding precise and often accurate predictions of aggregate economic
behavior, which yields prima facie validation of the model. However, a wealth

ISSN: 1529-2134/doi:10.1016/S1529-2134(04)07007-3
135
136 C. ATHENA AKTIPIS AND ROBERT O. KURZBAN
of data, especially from the controlled setting of the laboratory, bring into
question whether these assumptions reasonably reflect human thought and
behavior.2
The work of both Kahneman and Smith played an essential role in rethinking the
assumptions of classical theory. Smith, with his colleagues, for example, showed
that contrary to a model of pure self-interest, people will sacrifice their own
gains for others’ welfare (e.g. Hoffman, McCabe & Smith, 1996a). The Homo
economicus approach to reconciling such observations within a rational choice
framework is to add auxiliary social preferences for such things as aggregate
welfare or fairness to fit the empirically observed departures from the standard
model (e.g. Charness & Rabin, 2002).
Kahneman and Tversky, in turn, undermined the standard model’s assumption
that people’s preferences are stable, by showing preference reversals that cannot be
explained by any reasonable reading of standard theory. These and other challenges
to the standard economic model have been met by incorporating ancillary theories
and constraints (e.g. other regarding preferences, inequity aversion, risk aversion,
and loss aversion): Homo economicus’ only choice was to adapt, and she has indeed
undergone considerable evolutionary change.
Is, then, Homo economicus becoming extinct? The answer up until now has been
no, largely because there was no reasonable alternative to Homo economicus. But
now, with evolutionary psychology providing an alternative framework that can
potentially form the foundation for a satisfying and coherent theory of economic
behavior, the fate of Homo economicus is increasingly uncertain.
EVOLUTIONARY PSYCHOLOGY:
GENERAL PRINCIPLES
Self-Interest in Economics and Evolutionary Psychology
At first glance, the evolutionary perspective and the standard economic perspective
seem to be quite similar, both claiming that some form of self-interest drives
behavior (Economist, 1993). In broad strokes, economists posit that humans are
self-interested, while evolutionary psychologists’ claims rest on construing genes
as selfish agents, with unidimensional “preferences” for their own replication
(Dawkins, 1976/1989).3 Although these claims have superficial similarities, they
differ in important respects (Cosmides & Tooby, 1994).
In contrast to the assumption that humans are self-interested, evolutionary
psychologists endorse the view that genes themselves are the unit of “selfishness.”
Throughout human evolutionary history, our ancestors faced a wide variety of
Is Homo Economicus Extinct? 137
adaptive problems, from finding food, to securing a mate and taking care of
offspring. Individuals carrying genes that enabled them to solve these adaptive
problems were more likely to survive and leave healthy offspring, thus increasing
the frequency in the next generation of genes coding for cognitive mechanisms
that generated such behaviors (Tooby & Cosmides, 1992). Because solving these
adaptive problems often involved securing benefits for oneself, individuals have
cognitive mechanisms that, under a range of conditions, cause them to compute
the costs and benefits to the self associated with possible outcomes and act in ways
that achieve benefits for the self. This selfishness, although not a cornerstone of
evolutionary theories of behavior, is consistent with the fundamental assumptions
in standard economic models of human behavior.
Of course, not all behavior is transparently self-interested; at minimum, people
clearly do not behave as income-maximizers: people routinely incur costs, financial
or otherwise, for the benefit of others. Such other-regarding preferences are not
a problem for the evolutionary view. For example, the theory of kin selection
indeed predicts that people should have strong preferences for sacrificing in favor
of closely related individuals (Hamilton, 1964a, b).
In contrast, economic approaches need to explain why individuals do things
that benefit others, and this fact forced economic models to adapt in order to
explain “anomalous” findings, generally by adding new theories or constraints
(e.g. inequity aversion, other-regarding preferences). On the other hand, such
behaviors and preferences fall naturally out of well developed models derived from
the evolutionary approach (Hamilton, 1964a, b; Trivers, 1971) without recourse
to ad hoc theories.
Moreover, the evolutionary approach provides an ultimate explanation for the
existence of many fundamental human preferences. It is true that classical theory
can be used to infer preferences from observed behaviors (Samuelson, 1948) but
it can not speak to the origins of these preferences. Economists can (and do) claim
that individuals get utility from these activities, leaving the question of the origin of
tastes and preferences to the other behavioral sciences (Cosmides & Tooby, 1994;
Rubin & Paul, 1979). Evolutionary psychology provides answers – or at least a
way to generate possible answers – about these origins: tastes and preferences
that enabled us to better solve adaptive problems were selected for during human
evolutionary history.
It’s not Just the Fitness
Evolutionary psychologists do not claim that individuals simply try to maximize

the number of offspring. Instead, evolutionary psychologists claim that humans
have cognitive mechanisms designed to cause them to put effort toward things
(proximate goals) that would have tended to increase reproductive success during
our evolutionary history: gaining resources, increasing status, establishing social

networks, finding mates, having sex, and investing in their children. Throughout
our evolutionary history, engaging in these behaviors would have been critical
to reproductive success. Although our current reproductive opportunities are far
more numerous and often less costly than our reproductive opportunities in the
environment in which we evolved, we cannot be expected to give up the pursuit of
status, resources, social partners, and mates for the sperm bank and egg donation
clinic (Burnham & Phelan, 2000).
More specifically, this perspective suggests that humans do not have one
“fitness maximization system” but instead a variety of specific mechanisms
designed for particular adaptive functions. A well known and generally accepted
principle is that the boarder the array of tasks to which a tool – whether physical
or computational – can be applied, the worse the tool will be at accomplishing
each of those tasks (Tooby & Cosmides, 1992). As has been discussed at length
elsewhere, only information processing systems that are designed to apply to
a narrow range of problems have sufficient computational power to efficiently
generate accurate solutions to these particular problems (e.g. Chomsky, 1975).
Analogously, because different adaptive problems require different computations
for solving them, selection favors mechanisms designed to perform specific
functions. These include detecting “cheaters” in social situations (Cosmides
& Tooby, 1992), evaluating prospective mates (Buss, 1994; Singh, 1993),
and evaluating the risks and benefits associated with aggressive behavior
(Wilson & Daly, 1993, 1997).
In the remainder of this article, we describe the ways in which the work of Smith
and Kahneman undermined important components of classical economic theory
and discuss how evolutionary approaches provides a paradigmatic framework that
can make sense of findings considered “anomalous” on older views (e.g. Kuhn,
1996, 1962; Levati, 2002; Palfrey & Prisbrey, 1997).
VERNON SMITH
Empirical Challenges to Homo Economicus
Homo economicus was a reasonable first guess, in light of the widespread and
not completely unsuccessful application of classical theory to the real economy.
However, Vernon Smith’s work undermined the connection between observations
of a market that was behaving in a way that matched an equilibrium analysis and
the (plausible) inference that the agents in the economy were making decisions in
the way implied by the theory.
Smith’s early work on auctions is illustrative. In these experiments, subjects in a

laboratory experiment acted as buyers and sellers in a market designed to simulate
markets in the real world. Critically, participants had only very limited information
about the market as a whole, i.e. a description of the rules of the game and a single
price – the value (in dollars) of the object to the buyer or the seller (Smith, 1962).
In these experiments, results conformed well to those predicted by a rational
expectations competitive equilibrium model. This was true under diverse
parameterization of the experimental conditions, including the details of the market
mechanism, and was especially true as subjects repeated play. However, underlying
assumptions of classical theory were not met in these experiments, especially
information conditions – subjects had knowledge of only their own values, and
certainly not perfect information about other agents. However the competitive
equilibrium is achieved in these experiments, it is not through the computations
implied by a traditional equilibrium analysis taking place in the heads of the
subjects, but through the interactions between individuals with limited information
(Hayek, 1988; Smith, 1991a, b).
Smith’s later work (see the edited volumes Smith, 1991a, b, 2000a) extended the
experimental work on auctions to other kinds of markets and institutions. One of the
many important messages to emerge from this corpus of research was the critical
role played by institutions (Smith, 1994). The mechanisms that subjects interact
with in experiments have important effects on the behavior of those subjects. This
is in large part because, not surprisingly, people respond to their perceptions of
incentives, if not always in ways predicted by standard theory (see below). A
significant contribution of Smith and his collaborators’ work was to catalogue
subjects’ responses to various economic environments and institutions.
For example, Hoffman, McCabe and Smith (1996a) illuminated an important
aspect of bargaining psychology that is not captured by classical theory. They used
an Ultimatum Game, in which one subject (the Proposer) offers a second player
(the Responder) $X out of a total of $M. The second player can either accept,
receiving $X, or decline, in which case both get $0. HMS found that if player 1
earns the role of Proposer (as opposed to being chosen at random), Responders
are, under some conditions, willing to accept smaller offers. This is just one piece
of a large set of data which add richness to our understanding of human economic
decision making by taking psychology seriously – in this case, a psychology of
“entitlement.”
In the context of the present chapter, it is worth pointing out that the wealth
of evidence generated by Smith led him and some of his collaborators toward
evolutionary psychology as a source for new theoretical insights.
Evolutionary Perspectives on Smith’s Work

Cooperation. Recent work by Smith and his colleagues, especially Kevin McCabe
(e.g. Hoffman, McCabe, Shachat & Smith, 1994; Hoffman, McCabe & Smith,
1996a, b; Kurzban, McCabe & Smith, 2001), extended research programs showing
that in laboratory settings, people behave in ways that seem to suggest prosocial
motivations (Guth, Schmittberger & Schwarze, 1982). Hoffman, McCabe and
Smith (1998) focused on findings that players in numerous experiments using
Prisoner’s Dilemma and Public Goods games were far more cooperative than
predicted by standard theory. The authors presented an argument that an
explanation for the existence of cooperation is the result of a history of selection
for “certain cognitive abilities that predispose many people toward reciprocity”
(see also Gintis, Bowles, Boyd & Fehr, 2003, p. 655).
Similar results suggesting preferences beyond simple self-interest have been
obtained with other games as well. Smith’s explanation for this is, in some sense,
to say that in the laboratory you can take the person out of the social world, but
not the reverse. That is to say, people enter experiments as social creatures with
adaptations for a social world and experiences in it. In discussing why lab results
don’t match predictions of classical theory, Smith (2000b) suggests: “What may
be wrong is the very idea that instances of human decision interaction can be
construed as without a history or a future” (p. 80).
Responding to incentives: it’s not just the money. An important contribution
of Smith’s was to detail the ways in which subjects in the laboratory responded
to economic incentives. No one seriously doubts that they do, but Smith’s
work systematically examined how people responded to different institutions
– different rules of the game – a body of research with obviously important
implications for the real world, in particular questions surrounding how to structure
markets to generate economic efficiency (for a collection of relevant papers see
Smith, 1991a, b, 2000a).
Impersonal and Personal Exchange
The juxtaposition of two seemingly contradictory patterns of findings from the

laboratory, that: (1) in some experiments people closely follow their financial self-
interest in a way that bears out predictions drawn from classical theory and; that (2)
in other experiments people are considerably more prosocial than non-cooperative
game theory would suggest led Smith to a distinction that he highlighted in his
Nobel address: Impersonal versus Personal Exchange. As Smith puts it, when
experiments are run to test the standard model: “[A]lthough the test results tend
to be confirming in impersonal market exchange, the results are famously and

recalcitrantly mixed in ‘personal exchange,’ notably in a great variety of two-
person extensive form games where some half of the people attempt and frequently
succeed when risking cooperation, even when anonymously paired” (p. 505).
Impersonal exchange refers to the familiar relatively anonymous transactions
with money in markets with which we are all familiar. As the name suggests, this
type of interaction is with individuals with whom we have little history and quite
possibly little future, and, indeed, they may not be individuals at all, but rather
faceless firms. In this context, people seem to be performing computations that
allow them to follow their money-maximizing self-interest, whether or not these
computations are the ones implied by classical theory. Critically, under conditions
of impersonal exchange, Homo economicus is a reasonably predictive model.
Personal exchange, in contrast, refers to transactions with known others, or,
more generally to transactions that take place within a context that is construed
socially. When the experimental context is constructed to cue participants to the
social aspects of the task – for example, by reducing one’s sense of anonymity
(Hoffman, McCabe & Smith, 1996) – participants cooperate at rates well above
that predicted by traditional theory.
Taken together, the results of experiments in these differing environments
suggest that Homo economicus – in the sense of a set of cognitive mechanisms
that are capable of calculating how to pursue one’s financial interest – is one
aspect of human cognition. However, Homo economicus appears only under certain
conditions, especially ones in which there are few cues that one is embedded in a
social world. This resonates with evolutionary views that take human cognition to
be a set of functionally specific modules that are dynamically activated depending
on proximate cues (Tooby & Cosmides, 1990), and, more specifically, the view
that humans have a number of distinct modes of social interaction depending on
the context (Fiske, 1992).
DANIEL KAHNEMAN
Empirical Challenges to Homo economicus
While Vernon Smith’s work challenged the assumptions of Homo economicus

from inside the discipline of economics, psychologists Daniel Kahneman and
Amos Tversky, his collaborator, used the methods of experimental psychology
to challenge traditional notions of human rationality from the outside. Kahneman
and Tversky are best known for their work that chipped away at a cornerstone of
classical approaches to decision making, expected utility theory. Expected utility
theory assumes that individuals have preferences over choices based on the value
of the outcomes and the probability of the outcomes obtaining (von Neumann &
Morgenstern, 1944). Kahneman and Tversky’s prospect theory illustrates several

other factors that can affect the valuation of an outcome.
In their early work, Kahneman and Tversky described the two central aspects
of prospect theory: the non-linearity of decision-weights and loss aversion
(Kahneman & Tversky, 1979/2000). The non-linearity of decision-weights
describes how the perceived value of an outcome does not increase linearly as
its probability increases, as standard theory would suggest. Instead, individuals
tend to undervalue outcomes that are close to certain (but not 100% certain) and
overvalue extremely unlikely outcomes (Kahneman & Tversky, 1992/2000).
Loss aversion refers to the asymmetric value functions individuals have for
losses and gains. A loss tends to carry with it a greater decrease in perceived value
than the increase in value that comes with an equivalent gain, i.e. “losses loom
larger than gains.” In other words, people seem to dislike losing $10 more than they
like winning $10. Tennis great Jimmy Connors captured the notion of loss aversion
when he said, “I hate to lose more than I love to win”(Wall Street Journal, 2004).
In later work, Kahneman and Tversky expanded the notion of loss aversion,
applying it to economic exchanges (Kahneman & Tversky, 1991/2000). Kahneman
and Tversky convincingly showed that human behavior violates one of the
assumptions of standard models of decision-making: “that preferences do not
depend on current assets” (p. 143). According to the standard model, the amount
of money an individual is willing to pay (WTP) for a good should be the same
as the amount that individual is willing to accept (WTA) for giving up the good.
If I am willing to accept no less than $8 to give up a university mug, I should be
willing to pay $8 for that mug if I do not have it. However, individuals are often
willing to pay less to get an item – a university mug, among many other things
(Kahneman & Tversky, 1991/2000) – than they are willing to accept to give it up.
Kahneman and Tversky used prospect theory to explain this disparity, by arguing
that people don’t think of spending money as incurring a loss, but do perceive losing
goods as incurring a loss (see Kahneman, 1991/2000). That is, the loss aversion
entailed by prospect theory increases the magnitude of the perceived loss of the
item, raising the minimum amount an individual would be willing to accept (WTA)
to give that item up, which leads to the difference between WTA and WTP.
Kahneman and Tversky also showed that under some conditions, individuals
display inconsistent preferences (Kahneman & Tversky, 1984). In their
experiments, Kahneman and Tversky showed that framing an option as a cost
versus framing it as an uncompensated loss can affect whether that option is chosen.
For example, one of their questions asks the respondent to choose between two
options for combating a disease that is expected to kill 600 people. When presented
with the following options, most individuals (72%) choose the risk-averse option
(program A)
A: 200 people will be saved

B: There is a one-third probability that 600 people will be saved and two-thirds probability
that no people will be saved.
However, when individuals are asked to choose between logically identical options
that are framed differently, most individuals (78%) choose the more risky option
(program D) (Kahneman & Tversky, 1984).
C: 400 people will die.

D: There is a one-third probability that nobody will die and a two-thirds probability that 600
people will die.
Although the first and second sets of choices are identical in real terms, they
differ with respect to the framing of the two options. In the first version, the
deaths are framed as an uncompensated loss, implying a reference state in which
everybody dies of the disease (“. . . people will be saved”). The second version
frames the deaths as a loss, implying a reference state in which nobody dies
(“. . . people will die”).
Evolutionary Perspectives on Kahneman’s Work
Although Kahneman and Tversky’s prospect theory was not designed as a model
of ecological decision making, there are clear parallels between the preferences
described by prospect theory and the kinds of behaviors that are rational for
organisms in variable or risky ecological circumstances.
Loss/gain asymmetry. The asymmetry between losses and gains, which is a
centerpiece of prospect theory, can be easily reconciled with evolutionary theory.
For any organism, marginal losses are more fitness relevant than marginal gains
because gains and losses are asymmetrical with respect to their effect on expected
fitness. When an organism realizes a gain, it can increase its energy store, which
usually increases the length of time it will live. However, when an organism realizes
a loss, it may cause that organism to die or become more susceptible to death.
So, everything else being equal, under broadly plausible assumptions about fitness
functions, organisms should be designed to be more concerned with avoiding losses
than with realizing gains because of this asymmetry (Stephens & Krebs, 1986).
Note that the marginally decreasing nature of the utility function also makes
sense from an evolutionary perspective. If an individual is close to death, an
Fig. 1. This graph shows a hypothetical value function based on prospect theory
(adapted from Kahneman & Tversky, 1984). Note that the value of a loss is larger than
the value of an equivalent gain (dotted lines).
increase in their energy reserves is much more valuable than the same amount of
increase would be if that organism had much more energy to begin with (Fig. 1).
Prospect theory acknowledges that the variability and probability of a given
outcome can influence the value ascribed to it to a greater extent than predicted by
an expected value computation. Accordingly, an outcome with a less than certain
chance of occurring, say a 0.95 probability of getting $100, could be subjectively
valued at $80, rather than the $95 expected value. In evolutionary terms, optimal
foraging theory suggests that sensitivity to probability and variability is quite
rational (Stephens & Krebs, 1986). If an organism has a 0.95 probability of
acquiring the large prey item that will sustain it and a 0.05 chance of getting
nothing that day (and therefore dying), the organism should not value this option
equivalently to a sure bet at getting a prey item that is 0.95 the size of the large prey.
Because every organism always faces the threat of death (or decreased fecundity)
when it does not acquire enough resources (Caraco, 1983; Caraco & Lima, 1987),
a less than certain gain can be worth much less than its expected value. The same
reasoning applies for variable outcomes. If organisms have a threshold of energy
they need in order to survive, a variable option that entails the possibility of going
below that threshold is worth less than an option with an equivalent expected value
but no chance of going below that threshold (Caraco, 1983; Caraco & Lima, 1987).
In short, optimal foraging theory, a framework derived from an evolutionary
analysis, which has already proven fruitful for understanding decision making in
non-human animals, provides a potentially unifying perspective on the ultimate
origins of the preferences reflected by prospect theory (Caraco & Lima, 1987). It is
true that Kahneman and Tversky (1984) did not require the evolutionary analysis to
develop their theory – their ingenious empirical efforts were sufficient. But optimal
foraging theory links economic decision making in this domain with a proven
productive theoretical background and, importantly, suggests new directions of
inquiry (e.g. Rode, Cosmides, Hell & Tooby, 1999). This changes loss aversion
from an empirically derived ad hoc add-on into a comprehensible component of a
more seamless conceptual structure.
Relative Well-Being. Classical economic theory, sensibly, holds that utility
increases as wealth and consumption increase. Similarly, evolved organisms’
motivational systems should be designed to find it rewarding to have more of
those things that contribute to growth, health, and reproductive success. However,
a particular good can have different value, depending on the consumption history
of the individual: the first doughnut tastes better than the fifth. A well designed
motivational system should clearly be designed to take into account diminishing
returns – a hungry organism should be more motivated to acquire calories than a
sated one.
In addition, some types of motivational systems will be designed to vary not as a
function of some absolute metric (satiety), but rather in comparison with relevant
social others. For example, if mates are evaluated on the basis of some comparison
among candidates, then selection might favor a mechanism that motivates behavior
with the goal of achieving relative superiority along that dimension, rather than
against some objective metric (Price, Cosmides & Tooby, 2002).
There is considerable support for the idea that people’s happiness and satisfac-
tion is computed in this way. While it is true that there is a positive correlation
internationally between income and happiness (or “subjective well-being,”) this
relationship is neither straightforward nor monotonic, and large national increases
in wealth do not cause correspondingly sizable increases in happiness (Diener &
Oishi, 2000). This is somewhat of a puzzle on standard approaches to utility.
State Dependence. The impact of a given change in welfare depends on an
organism’s current state; a small gain to an organism that has large energy reserves
might represent a large gain to a starving organism. Consistent with this, people
indeed respond to risks to changes in welfare as a function of their current financial

or nutritive state (Rode et al., 1999). In the context of prospect theory, this suggests
that there should be some variation in the shape of individuals’ curves. For example,
a starving animal or a poor person is likely to have a more steeply sloped curve
for gains since a morsel of food or dollar has greater (perceived) value, while a
sated organism or rich person would have a shallower curve for gains because the
marginal value of those gains is small. Somewhat less obviously, the loss portion of
the curve is likely to be less steeply sloped for a hungry animal or poor individual
because a small loss is likely to kill or ruin that individual, and further losses are
irrelevant after that point. For a sated animal or for a rich person, the curve is likely
to be steeply sloped curve over a greater range of losses since there is more to lose
before all is lost.

Individuals with differently shaped curves will value the same risky opportunity
in different ways. A rich individual might value a risky option less because the
perceived value of the potential loss associated with the risk is much greater than
the perceived value of the potential gain. However, a poorer individual might be
more willing to undertake that risk because the value of the potential gain is greater
than the value of the possible loss.
Data suggest that humans do behave in such a way: Wilson and Daly (1997)
found that rates of homicide and early pregnancy are higher in areas of lower
socioeconomic status and lower life expectancy. They claim that this difference
in willingness to undertake risks is adaptive because individuals with lower life
expectancies should be more willing to engage in risky behaviors (such as violence)
that might raise status, thereby increasing reproductive success. One interpretation
of this finding is that individuals who perceive themselves to be poorly off may
have a curve with a steeper slope for gains and shallower slope for losses, which
makes them more likely to engage in risky behaviors such as violence and sex
(Wilson & Daly, 1993, 1997).
The foregoing raises the possibility that the differently shaped curves for
organisms in different states can be profitably thought of as the same prospect
theory curve with reference points above (in the case of a sated individual) and
below (in the case of a hungry individual) the origin. Whether state dependence is
conceptualized as differently shaped curves or changing reference points along
a single curve, individuals value outcomes differently, and therefore behave
differently, depending on their current subjective well-being.
THE RESURGENCE OF HOMO SAPIENS

Tastes and Preferences
Although tastes and preferences are central to economic decision making, the
question of the origins of tastes and preferences is of little interest to many
economists. Evolutionary psychology, on the other hand, is critically concerned
with and has conceptual tools to address the origins of preferences (Cosmides &
Tooby, 1994): our literal taste for fat is because of the adaptiveness of consuming
fat in our evolutionary history, we like finding attractive partners because that
increased our reproductive success, and we care about our children because their
survival meant the survival of our genes. Similar arguments can be made for other
sorts of tastes and proximate goals, such as helping kin (Daly, Salmon & Wilson,
1997), developing positive social relationships (Leary & Baumeister, 2000), and
striving for prestige (Henrich & Gil-White, 2001).

One of the ways in which our tastes and preferences helped us survive during
our evolutionary history was by attracting us to fitness-relevant elements of the
environment. As Thornhill (1998) put it: “beauty experiences are unconsciously
realized avenues to high fitness in human evolutionary history. Ugliness defines
just the reverse” (p. 544). In the case of food, our ancestors lived in a world in
which fat and sugar were much scarcer than it is today. Passing up an opportunity
to eat meat or high sugar items (which were then mostly fruits) could have made
the difference between starving or surviving the winter. Because of the nutritional
value of animal fat, genes that coded for a preference for consuming animal fat
were selected. In modern environments, this taste leads to unfortunate outcomes.
Consuming meat and high sugar foods (which are now candy bars and the like)
can make the difference between an early heart attack and surviving to see one’s
grandchildren. The evolutionary approach predicts tastes for proximate goals (fatty
and sugary foods), not ultimate survival and reproductive success (Burnham &
Phelan, 2000).
Time/Energy Constraints
The evolutionary approach highlights the fact that our ancestors – like all other
living organisms – had only limited time and energy to engage in a wide variety
of fitness-enhancing activities, and therefore had to make trade-offs between one
activity and another, for example, how much time to invest in finding a mate
versus finding food, or whether to invest in having offspring immediately or invest
in increasing status so future offspring might have a better chance at survival and
reproduction.
This process itself, allocating time and energy to decision-making, represents
an important adaptive task. Collecting information relevant to a question and
determining the correct weighting of that information can take an arbitrarily

large amount of time and mental resources. Standard economic approaches were
unconcerned with computational limitations, implicitly assuming that time and
mental resources were unlimited and costless.
Simon (1955) introduced the idea that computation was itself a limited resource,
stating that humans exhibit “bounded rationality,” or rationality within certain
computational constraints. Both Kahneman and Smith extended these ideas in
important ways. Smith and Walker (1993), for example, developed a framework
designed to take into account payoffs to the subject beyond the monetary, including
the cost of making the decision itself. Kahneman and Tversky (1979/2000, 1984)
showed that humans use decision rules that are based on neither perfect information
nor straightforward cost/benefit evaluation.

Both of these research programs laid foundations for extremely productive lines
of research that have been taken up by others who have extended and clarified these
issues. Gigerenzer and his colleagues (Gigerenzer, 2000; Gigerenzer & Selton,
2001; Gigerenzer, Todd & The ABC Research Group, 1999) have strengthened
the connections between bounded rationality and evolution/ecology to develop
precisely formulated models that make sense in the context of an organism with
biological constraints. They have developed the idea that humans are “ecologically
rational,” making judgments that are “fast and frugal,” arriving at good decisions
using only very limited information (Gigerenzer et al., 1999). Other researchers
are similarly engaged in productive research programs that take seriously the idea
that human decision makers have limited time and cognitive resources to bring to
bear on any given problem (e.g. Gabaix & Laibson, 2003).
Emotions
Another aspect of human behavior that is informed by the evolutionary perspective

is that of emotions. Historically, emotions have been seen as impediments to
rationality, getting in the way of good decision-making. However, it has become
clear that emotions play a central, indeed critical, role in decision-making.
Emotions imbue situations with an affective valence that is essential for responding
appropriately to opportunities and weighing choices. Damasio (1995) refers
to this as a “somatic marker,” a physiological response that carries affective
information relevant to a situation. Individuals whose emotional capacities have
been damaged by brain injuries seem to lack this anticipatory response. Damasio
and his collaborators (Bechara, Damasio, Damasio & Lee, 1999) have shown
that these patients persist in choosing cards from a deck with more risk and
lower expected return, while normals developed an anticipatory psychological
response and learned to avoid the inferior deck. In a sense, the patients who lacked
the emotional response were behaving less rationally than the individuals whose
emotional systems were in tact.
This view resonates with that of other functionally oriented researchers (Frank,
1988; Ketelaar & Todd, in press; Tooby & Cosmides, 1990) who have discussed the
evolutionary significance of emotions. According to evolutionary psychologists
Tooby and Cosmides (1990), emotions function to coordinate physiological
and psychological responses to fitness relevant situations, focusing attention on
important aspects of the environment and guiding the organism to adaptively
correct responses. Emotions are obviously important for responding to the social
environment as well. Ketelaar and Au have shown that the emotion of guilt seems
to have an important function in motivating commitment (Ketelaar & Au, 2003).

Hence, emotions are more of an aid to decision-making than an obstacle to
be overcome. Without them we would be unable to attend to relevant stimuli,
properly respond to risky situations, or feel emotions such as guilt that facilitate the
formation and maintenance of social relationships. The evolutionary significance
of emotions is transparent; they help us respond to danger effectively and facilitate
the social relationships that influence our reproductive success.
CONCLUSION
Smith, Kahneman and their colleagues amassed a tremendous amount of evidence
that first threatened, then endangered Homo economicus. Instead of a simple
model of human behavior based on rationality and perfect information, Homo
economicus became an amalgam of various sub-models and constraints that explain
deviations from straightforward rationality. In this paper, we sought to answer
the questions: Why is it that humans so often depart from the reasoning, decision
making, and behavior of Homo economicus? When should we expect human
behavior to be consistent with the predicted behavior of Homo economicus?
Where should we look for the ultimate explanation for our species’ actual
preferences?
Evolutionary psychology and experimental economics share more than a
skeptical view of Homo economicus; they also share the common goal of creating
a realistic model of human behavior. Evolutionary psychology can help provide
a framework for experimental economists to understand behaviors observed in
the laboratory and in the real world, as well as an account of the origins of
preferences. Experimental economists have provided a wealth of evidence with
which adaptationist hypotheses about cognitive design can be built.
The evolutionary view sheds light on the ultimate reasons for behavior
surrounding trust and reciprocity (Hoffman et al., 1996b), and gives potential
insights into people’s asymmetrical valuations of equivalent sized losses and gains
(Kahneman & Tversky, 1991/2000). More generally, evolutionary psychology is
potentially valuable to economics because it can help economists better understand
the ultimate reasons for the cognitive mechanisms underlying economic choices,
or even, as in the case of Prospect Theory, the shape of a particular value function.
The marriage of theory from evolutionary psychology and methods from
cognitive science and experimental economics has already borne empirical
fruit (Hoffman et al., 1996b; Ketelaar & Au, 2003; Kurzban et al., 2001).
The emerging field of “neuroeconomics” (Glimcher, 2003; McCabe, 2003), for

example, owes much to the increasing connections between economics and
cognitive psychology that have been forged during the past two decades, many
of which to a greater or lesser degree are descended from research by Smith
and his colleagues. Reciprocally, the tools of experimental economics can help
evolutionary psychologists test hypotheses focal to their theoretical interests,
including kin-directed altruism (Unur & Peters, 2003), how individuals balance
their interests against those of their group, people’s desire for punishment,
intertemporal choices, and so on.
So, is Homo economicus extinct? Not exactly. In hindsight, Homo economicus is
a species that could never have evolved in the first place. Homo economicus tried to
survive by adding auxiliary theories into an ultimately unsalvageable model. The
model of human decision making envisioned by early rational choice theorists
violates what is known about human behavior in the laboratory on the one hand,
and what is known about the products of evolution by natural selection on the
other. It is true that a wealth of evidence suggests that Homo economicus is a
reasonable model for human behavior, but only under very particular conditions,
such as impersonal exchange. But the correct model of human decision making
requires acknowledging that our actual species, Homo sapiens, is an amalgam of
Homo reciprocans, Homo prospectus, Homo nepotismis, and no doubt many other
interesting subcomponents. Homo economicus can be expected to live on in Homo
sapiens as well, showing herself only when no one is watching.
NOTES
1. Technically the Bank of Sweden Prize in Economic Sciences in Memory of Alfred

Nobel, established in 1968.
2. In his Prize Lecture, Smith (2002) was quick to point out that another Smith (Adam)
had long ago argued (Smith, 1759) that “individuals were mischaracterized by the metaphor,
‘economic man’ ” (p. 2).
3. We use teleological language for discursive economy, confident that this language can
be paid out in terms of purely physical causality (see Dawkins, 1976).
ACKNOWLEDGMENTS
This material is based upon work supported under a National Science Foundation
Graduate Research Fellowship. The second author would also like to thank the
Russell Sage Foundation for their generous support.
REFERENCES
Bechara, A., Damasio, H., Damasio, A. R., & Lee, G. P. (1999). Different contributions of the human
amygdala and ventromedial prefrontal cortex to decision-making. Journal of Neuroscience,
19(13), 5473–5481.
Burnham, T., & Phelan, J. (2000). Mean genes: From sex to money to food: Taming our primal instincts.
Cambridge, MA: Perseus.
Buss, D. M. (1994). The evolution of desire. New York, NY: Basic Books.
Caraco, T. (1983). White-crowned sparrows (Zonotrichia leucophrys): Foraging preferences in a risky
environment. Behavioral Ecology and Sociobiology, 12, 63–69.
Caraco, T., & Lima, S. L. (1987). Survival, energy budgets, and foraging risk. In: M. L. Commons,
A. Kacelnik & S. J. Shettleworth (Eds), Quantitative Analyses of Behavior: Foraging (Vol. 6,
pp. 1–21). Hillsdale, NJ: Lawrence Erlbaum.
Charness, G., & Rabin, M. (2002). Understanding social preferences with simple tests. The Quarterly
Chomsky, N. (1975). Reflections on language. New York: Random House.
Cosmides, L., & Tooby, J. (1992). Cognitive adaptations for social exchange. In: J. H. Barkow, L.
Cosmides & J. Tooby (Eds), The Adapted Mind (pp. 163–228). New York: Oxford University
Press.
Cosmides, L., & Tooby, J. (1994). Better than rational: Evolutionary psychology and the invisible hand.
AER Papers and Proceedings (May), 327–332.
Daly, M., Salmon, C., & Wilson, M. (1997). Kinship: The conceptual hole in psychological studies
of social cognition and close relationships. In: J. Simpson & D. Kenrick (Eds), Evolutionary
Social Psychology (pp. 265–296). Mahwah, NJ: Lawrence Erlbaum.
Damasio, A. (1995). Descarte’s error. New York: Avon Books.
Dawkins, R. (1976/1989). The selfish gene. Oxford: Oxford University Press.
Diener, E., & Oishi, S. (2000). Money and happiness: Income and subjective well-being across nations.
In: E. Diener & E. M. Suh (Eds), Culture and Subjective Well-being (pp. 185–218).
Evo-Economics: Biology meets the dismal science (1993, December 2). The Economist, 93–95.
Fiske, A. (1992). The four elementary forms of sociality: Framework for a unified theory of social
relations. Psychological Review, 99, 689–723.
Frank, R. H. (1988). Passions within reason: The strategic role of the emotions. New York: W. W.
Norton.
Gabaix, X., & Laibson, D. (2003). A new challenge for economics: The “frame problem” In: Vol. 1:
Rationality and Well-Being. Oxford University Press.
Gigerenzer, G. (2000). Adaptive thinking: Rationality in the real world. New York: Oxford University
Press.
Gigerenzer, G., & Selton, R. (2001). Rethinking rationality. In: G. Gigerenzer & R. Selton (Eds),
Bounded Rationality: The Adaptive Toolbox, MIT Press.
Gigerenzer, G., Todd, P. M., & The ABC Research Group (1999). Simple heuristics that make us smart.
New York: Oxford University Press.
Gintis, H., Bowles, S., Boyd, R., & Fehr, E. (2003). Explaining altruistic behavior in humans. Evolution
and Human Behavior, 24(3), 153–172.
Glimcher, P. W. (2003). Decisions, uncertainty, and the brain: The science of Neuroeconomics, MIT
Press.
Guth, W., Schmittberger, R., & Schwarze, B. (1982). An experimental analysis of ultimatum bargaining.
Hamilton, W. D. (1964a). The genetical evolution of social behavior I. Journal of Theoretical Biology,
7, 1–16.
Hamilton, W. D. (1964b). The geneticial evolution of social behavior II. Journal of Theoretical Biology,
7, 17–52.
Hayek, F. (1988). The fatal conceit. Chicago: University of Chicago Press.
Henrich, J., & Gil-White, F. J. (2001). The evolution of prestige: Freely conferred status as a mechanism
for enhancing the benefits of cultural transmission. Evolution and Human Behavior, 22,
165–196.
Hoffman, E., McCabe, K., Shachat, J., & Smith, V. L. (1994). Preferences, property rights, and
anonymity in bargaining games. Games and Economic Behavior, 7(3), 346–380.
Hoffman, E., McCabe, K., & Smith, V. L. (1996a). On expectations and the monetary stakes in
ultimatum games. International Journal of Game Theory, 25(3), 289–301.
Hoffman, E., McCabe, K., & Smith, V. L. (1996b). Social distance and other regarding behavior in
dictator games. American Economic Review, 86(3), 653–660.
Kahneman, D., & Tversky, A. (1979/2000). Prospect theory: An analysis of decision under risk. In:
D. Kahneman & A. Tversky (Eds), Choices, Values and Frames (pp. 17–43). Cambridge:
Kahneman, D., & Tversky, A. (1984). Choices, values and frames. American Psychologist, 39(4),
341–350.
Kahneman, D., & Tversky, A. (1991/2000). Loss aversion in riskless choice: A reference-dependent
model. In: D. Kahneman & A. Tversky (Eds), Choices, Values and Frames (pp. 143–158).
Cambridge: Cambridge University Press.
Kahneman, D., & Tversky, A. (1992/2000). Advances in prospect theory: Cumulative representation
of uncertainty. In: D. Kahneman & A. Tversky (Eds), Choices, Values and Frames (pp. 44–65).
Cambridge: Cambridge University Press.
Ketelaar, T., & Au, W. T. (2003). The effects of feelings of guilt on the behavior of uncooperative
individuals in repeated social bargaining games: An affect-as-information interpretation of the
role of emotion in social interaction. Cognition and Emotion, 17, 425–453.
Ketelaar, T., & Todd, P. M. (in press). Framing our thoughts: Evolutionary psychology’s answer to the
computational mind’s dilemma. In: P. S. Davies & H. R. H. III (Eds), The Evolution of Minds:
Psychological and Philosophical Perspectives. Norwell, MA: Kluwer.
Kuhn, T. S. (1996, 1962). The structure of scientific revolutions (3rd ed.). Chicago and London:
Kurzban, R., McCabe, K., & Smith, V. L. (2001). Incremental commitment and reciprocity
in a real time public goods game. Personality and Social Psychology Bulletin, 27(12),
1662–1673.
Leary, M. R., & Baumeister, R. F. (2000). The nature and function of self-esteem: Sociometer theory
(Vol. 32). San Diego: Academic Press.

Levati, M. V. (2002). Explaining private provision of public goods by conditional cooperation:
An evolutionary approach. Unpublished manuscript, Max Planck Institute for Research into
Economic Systems, Jena, Germany.
McCabe, K. (2003). Neuroeconomics. In: L. Nadel (Ed.), Encyclopedia of Cognitive Science (Vol. 3,
pp. 294–298). Nature Publishing Group, Macmillan.
Palfrey, T. R., & Prisbrey, J. E. (1997). Anomalous behavior in public goods experiments: How much
and why? American Economic Review, 87, 829–846.
Price, M. E., Cosmides, L., & Tooby, J. (2002). Punitive sentiment as an anti-free rider psychological
device. Evolution and Human Behavior, 23, 203–231.
Rode, C., Cosmides, L., Hell, W., & Tooby, J. (1999). When and why do people avoid unknown
probabilities in decisions under uncertainty? Testing some predictions from optimal foraging
theory. Cognition, 72, 269–304.

Rubin, P. H., & Paul, C. W., III (1979). An evolutionary model of taste for risk. Economic Inquiry,
17(4), 585–596.
Samuelson, P. A. (1948). Consumption theory in terms of revealed preference. Economica, 15, 243–
253.
Simon, H. A. (1955). A behavioral model of rational choice. Quarterly Journal of Economics, 69,
99–118.
Singh, D. (1993). Body shape and women’s attractiveness: The critical role of waist-to-hip ratio. Human
Nature, 4(3), 297–321.
Smith, V. L. (1962). An experimental study of competitive market behavior. Journal of Political
Economy, 70, 111–137.
Smith, V. L. (1991a). Papers in experimental economics. Cambridge: Cambridge University Press.
Smith, V. L. (1991b). Rational choice: The contrast between economics and psychology. Journal of
Political Economy, 99, 877–897.
Smith, V. L. (1994). Economics in the laboratory. Journal of Economic Perspectives, 8(1), 113–131.
Smith, V. L. (2000a). Bargaining and market behavior: Essays in experimental economics. Cambridge:
Smith, V. L. (2000b). Bargaining theory, behavior and evolutionary psychology: Introduction. In: V. L.
Smith (Ed.), Bargaining and Market Behavior: Essays in Experimental Economics (pp. 79–89).
Cambridge, UK: Cambridge University Press.
Smith, V. L., & Walker, J. M. (1993). Rewards, experience and decision costs in first price auction.
Economic Inquiry, 31, 237–244.
Stephens, D. W., & Krebs, J. R. (1986). Foraging theory. Princeton: Princeton University Press.
Thaler, R. H. (2000). From Homo economicus to Homo sapiens. Journal of Economic Perspectives,
14(1), 133–141.
Thornhill, R. (1998). Darwinian aesthetics. In: C. Crawford & D. L. Krebs (Eds), Handbook
of Evolutionary Psychology: Ideas, Issues and Applications (pp. 543–572). Mahwah, NJ:
Lawrence Erlbaum.
Tooby, J., & Cosmides, L. (1990). The past explains the present: Emotional adaptations and the structure
of ancestral environments. Ethology and Sociobiology, 11, 375–424.
Tooby, J., & Cosmides, L. (1992). The psychological foundations of culture. In: J. H. Barkow, L.
Cosmides & J. Tooby (Eds), The Adapted Mind (pp. 19–136). New York: Oxford University
Press.
Trivers, R. L. (1971). The evolution of reciprocal altruism. The Quarterly Review of Biology, 46(March),
35–57.
Unur, A. S., & Peters, H. E. (2003). Experimental analysis of other-regarding preferences of parents
and children. Unpublished manuscript.
von Neumann, J., & Morgenstern, O. (1944). Theory of games and economic behavior (1953 ed.).
Princeton, NJ: Princeton University Press.
Wilson, M., & Daly, M. (1993). Lethal confrontational violence among young men. In: N. J. Bell &
R. W. Bell (Eds), Adolescent Risk Taking (pp. 84–106). Newbury Park, CA: Sage.
Wilson, M., & Daly, M. (1997). Life expectancy, economic inequality, homicide and reproductive
timing in Chicago neighborhoods. British Medical Journal, 314(1271).
AUSTRIAN ECONOMICS,
EVOLUTIONARY PSYCHOLOGY, AND
METHODOLOGICAL DUALISM:
SUBJECTIVISM RECONSIDERED
Viktor J. Vanberg
ABSTRACT
The methodological individualism and subjectivism of the Austrian tradition
in economics is often associated with a methodological dualism, i.e. the claim
that the nature of its subject matter, namely purposeful and intentional human
action, requires economics to adopt a methodology that is fundamentally
different from the causal explanatory approach of the natural sciences. This
paper critically examines this claim and advocates an alternative, explicitly
naturalistic and empiricist outlook at human action, exemplified, in particular,
by the research program of evolutionary psychology. It is argued that, within
the Austrian tradition, a decidedly naturalistic approach to subjectivism can
be found in F. A. Hayek’s work.
In short, I shall contend that the empirical element in economic theory – the only part which is
concerned not merely with implications but with causes and effects . . . – consists of propositions
about the acquisition of knowledge.

F. A. Hayek (1948a, p. 33).

ISSN: 1529-2134/doi:10.1016/S1529-2134(04)07008-5
155
156 VIKTOR J. VANBERG
1. INTRODUCTION
The combination of methodological individualism and subjectivism is commonly

regarded as constitutive of the research tradition that was originated by Carl Menger
(Kirzner, 1992a, p. 73; Vanberg, 1998)1 and for which the name “Austrian eco-
nomics” has become the modern label (Kirzner, 1992b, p. 122; Vaughn, 1994, p. 3).
Methodological individualism entails the commitment to explain aggregate or
collective social phenomena in terms of behavioral choices made by individuals,
their interaction-effects and unintended consequences.2 This commitment, in turn,
implies that an individualist approach must include as its theoretical core – or its
principal explanatory conjectures – general assumptions about human behavior.
The Austrian research tradition is subjectivist in its insistence on the fact that
subjective preferences and beliefs are essential determinants of human action, and
that an adequate theoretical account of human behavior must pay proper attention
to these determinants.
It is ultimately its thoroughgoing subjectivism that marks the critical difference
between the Austrian tradition and mainstream, neoclassical economics (Streissler,
1972, p. 427). The latter may also be said to adhere, in essence, to methodological
individualism and to account for the subjectivity of human valuations in its concept
of utility (Lachmann, 1977, p. 156). With its fiction of perfect information,
however, neoclassical orthodoxy entirely assumes away the very complexities
that arise from the subjectivity of perceptions, knowledge and expectations (Witt,
1992, p. 216; 2003, p. 25).3 Notwithstanding its various refinements, the orthodox
approach in essence proceeds as if all economic agents would “see” the world
as it truly is (and as the economic theorist supposedly sees it) and were able to
reliably foresee the consequences of their actions. Such perspective rules out the
possibility that inter-personal differences and inter-temporal changes in behavior
may be due to differences and changes in subjective beliefs about the world, and
it affords the economic theorist the luxury of not having to deal with what, in this
regard, goes on in people’s minds.
There has been general agreement within the Austrian tradition that an
economics that has anything of relevance to say about real world economies
cannot afford the luxury of disregarding the subjectivity, and therefore variability,
of human knowledge and beliefs.4 Yet, Austrian economists have been in much
less agreement on what such subjectivism implies for the manner in which
economics can be – or should be – carried out as a scientific enterprise. In
fact, among advocates of different strands of Austrian economics there has been
considerable disagreement on the methodological implications that a commitment
to subjectivism entails (Caldwell, 1991, p. 488; Lavoie, 1991, p. 471, 1994b, p. 55;
Shand, 1981, p. 13).
Austrian Economics, Evolutionary Psychology, and Methodological Dualism 157
The purpose of the present paper is to take a closer look at some of the principal
arguments that have been advanced by Austrians economists on the methodological
implications of subjectivism. More specifically, I shall take issue with claims,

raised by a number of authors within the Austrian tradition, that recognizing the
subjectivist determinants of human action requires economics (as well as the other
social sciences) to adopt a methodology that is fundamentally different from the
causal explanatory approach of the natural sciences. After summarizing these
claims (in Section 2), I shall critically examine (Section 3) and contrast them
with an alternative, explicitly naturalistic and empiricist outlook at human action,
exemplified by the research program of evolutionary psychology (Section 4 ) and
by the theoretical perspectives of biologist E. Mayr and philosopher K. R. Popper
(Section 5). I shall further argue that such an alternative, empiricist research
program has actually been advanced from within the Austrian tradition itself,
namely in F. A. Hayek’s work (Sections 6 & 7).5 My conclusion will be that there
are ample reasons for advocates of the Austrian research program to revise some
of their traditional views on subjectivism, and to part with methodological claims
that tend to foreclose prematurely promising research.
2. SUBJECTIVISM AND METHODOLOGICAL DUALISM

IN AUSTRIAN ECONOMICS
In their efforts to specify what they consider to be the most consistent version of
“subjectivism” Austrian economists have distinguished between a “subjectivism
of preferences” and a “subjectivism of expectations” (Lachmann, 1977, p. 28) or
between “static” and “dynamic” subjectivism (Kirzner, 1992b, p. 122; O’Driscoll
& Rizzo, 1985, p. 22).6 Such distinctions are meant to emphasize that a consistent
subjectivism must go beyond recognizing the subjectivity of human valuations
and take serious the fact that human purposeful action is always based on
subjective expectations and interpretations, i.e. on mental activities such as
“ordering and formulating ends, allocating means to them, making and revising
plans, determining when action has been successful” (Lachmann, 1982, p. 37).
There is no need in the present context to comment on the controversy over
whether or not Carl Menger (Chamilall & Krecke, 2002) and second or third-
generation Austrians, such as Ludwig von Mises, can justly be said to have
already paid sufficient attention to the “subjectivism of expectations” (Lachmann,
1982, p. 36) or adopted a “fully subjectivist treatment of choice” (Kirzner, 1992b,
p. 126).7 Nor is there a need to address the issue of whether it was not until the
“Austrian revival of the 1970s” and its “radical subjectivism”(Lachmann, ibid.)
that the Austrian tradition arrived at a consistent “subjectivism of active minds”
(ibid., p. 37), or whether, as others suspect, the recent “radically subjectivist”

movement marks, instead, a departure from the paradigmatic core of the Austrian
research program (Kirzner, 1995, p. 13). My exclusive concern here will be with the
above noted claim that the subjectivist dimensions of their subject matter require
economists to adopt a methodological approach that is fundamentally different
from the causal explanatory methodology commonly practiced in the natural
sciences. This claim continues to play a prominent role within the Austrian tradition
ever since it was first and most forcefully pronounced by Ludwig von Mises and,
whether directly or indirectly, Mises’ arguments on this issue have clearly been
most influential in shaping the general paradigmatic outlook of modern Austrian
economics (Witt, 1989, p. 411).
According to Mises, what separates the methodology of economics, and the
sciences of human action more generally,8 from that of the natural sciences is the
fact that they deal with “conscious behavior or purposive activity” (Mises, 1990,
p. 19), i.e. with “behavior open to the regulation and direction by volition and
mind” (ibid.), by contrast to mere reactive responses to stimuli “which cannot
be controlled by volition” (ibid., p. 23).9 Explicitly rejecting K. R. Popper’s
falsifiability criterion as the relevant test for empirically contentful scientific
theories Mises (1978, p. 69) declares: “It is obvious that all this cannot refer
in any way to the problems of the sciences of human action.” And by contrast
to Popper’s view that the same general methodological principles apply to all
empirical sciences, including the behavioral and social sciences, he insists on a
“methodological dualism” that “splits the realm of knowledge into two separate
fields, the realm of external events, commonly called nature, and the realm of
human thought and action” (Mises, 1957, p. 1).
Mises’ arguments in support of his dualism claim are somewhat different for
the two branches of the sciences of human action he distinguishes between,
namely praxeology or the praxeological disciplines on the one side and history
or the historical disciplines on the other (Mises, 1978, p. 41). Economics as a
praxeological discipline, so Mises argues, is concerned with “the a priori category
of action” (ibid.), not with the particular circumstances surrounding observable
actions. It deals “with the categorical essence of choice and action as such, with
the pure elements of setting aims and applying means” (1990b, p. 21), not with
“the concrete content of the ends men are aiming at” (Mises, 1978, p. 73) or
with “the quality of the . . . means applied” (Mises, 1990, p. 9).10 It is based on
“aprioristic reasoning” (ibid., p. 29) about “choosing as such,” not on psychological
arguments on “internal processes determining the various choices in their
concreteness.”11
The study of the particularities of concrete acts is, in Mises’ scheme, the
task of the second branch of the sciences of human action, of history. History
“comprehends the totality of what is experienced about human action” (Mises,

1990, p. 43), its scope is “to investigate what ends people aim at and what means
they apply for the realization of their plans” (ibid., p. 24). And such investigation
is, so Mises argues, an exercise in psychological reasoning.12 It is about
“understanding the meaning of action” (ibid., p. 12) in the sense of inferring the
intentions and expectations that motivate actual human choices. By contrast to the
praxeological approach of economics, Mises regards the method of understanding
(Verstehen) as “the specific method of historical research” (ibid.),13 as the method
that not only historians but “all other people always apply in commenting upon
social events in the past and in forecasting future events” (ibid., p. 26).
As noted, the reasons why Mises finds it necessary to insist on the
methodological autonomy of the sciences of human action and their categorical
separation from the methodological standards that apply to the natural sciences,
differ somewhat for the two branches, praxeology and history. As far as Mises’
claim is concerned that the praxeological branch of the sciences of human action
provides a priori knowledge about “the realm of human thought and action,” it
is obviously incompatible with a methodology à la Popper, according to which
only empirically refutable theories qualify as “scientific.” Acceptance of such a
methodological view would, so Mises argues, require one to “deny the cognitive
value of a priori knowledge” (1978, p. 5) and it would disqualify all a priori
theories, such as praxeology, as “unscientific,”14 an implication that, in his eyes, is
in apparent conflict with the fact that “nothing is more certain for the human mind
than what the category of human action brings into relief” (1978, p. 71).
Mises’ apriorism15 has been the subject of extensive debate and criticism, and
there is no need to repeat in detail here the arguments that speak against Mises’
claims.16 It may suffice here to note that the relevant issue is not, as Mises’ phrasing
of the matter tends to suggest, whether a priori theories can have “cognitive
value” (the cognitive value of mathematics is surely undisputed). The issue rather
is whether theories can be a priori certain and non-refutable and, at the same
time, have any empirical content, inform us about the world of our experience.
Popper’s falsifiability criterion is not meant to disqualify non-refutable theories
as “meaningless,” it is meant to separate empirically contentful theories from

theories that, even though they may provide meaningful insights into matters of
logic, do not inform us about matters of fact, including facts of human thought and
action.17
What deserves more attention in the context of this paper are Mises’ substantive
arguments for why the sciences of human action, the historical disciplines
no less than their praxeological counterparts, cannot be measured against the
same methodological criteria that apply to the natural sciences. This is, in the
first instance, the argument that mental phenomena represent a realm that is
categorically separated from the realm of “external” (physical and physiological)

phenomena, and that they escape, therefore, the methods of the natural sciences
that are appropriate only for the latter realm. The fact that action is determined
by ideas, that it “is the mind’s response . . . to the conditions in which nature and
other people’s actions place a man” (Mises, 1990, p. 20), puts it, as Mises asserts,
outside the explanatory scope of the natural sciences.18 – It is noteworthy, and I
shall return to this issue later, that the way Mises puts this argument leaves some
room for interpretation. While most of his comments on this matter appear to
imply that the divide between the two realms can, for principal reasons, never
be overcome, in other comments he explicitly limits his claim to “the present
state of our knowledge” (ibid., p. 25),19 allowing for the possibility that the issue
may have to be reconsidered in light of the knowledge available at future points
in time.
Mises’ argument that “the methods applied in dealing with natural events” (1978,
p. 47) are not applicable to the study of mental phenomena is specifically directed
at positivistic and behaviorist approaches. Their principal shortcoming is in his
view that, by reducing the study of human action to stimulus-response schemes
(ibid., pp. 40, 121), they ignore the anticipatory capacity of the human mind upon
which intentional and purposeful conduct is based, in particular man’s capacity to
anticipate the conduct of his fellow men (ibid., p. 47). The method that acting man
employs in such anticipation is, as Mises argues, the same method that constitutes
the methodological autonomy of the various historical disciplines, namely the
“study of meaning” (ibid., p. 43) or the method of “understanding (Verstehen)”
(ibid., p. 48). It is a method that allows these disciplines privileged access to “the
ultimate principles” governing the phenomena they study (Mises, 1990, p. 9),
namely knowledge “about the meaning which acting men attach to their action”
(ibid., p. 8). Understanding, or the study of meaning, deals with “the reactions of
the mind,” i.e. “with something invisible and intangible that cannot be perceived by
the methods of the natural sciences,” which, so Mises argues, the natural sciences
must, however, recognize “as real also from their point of view, as it is a link in a
chain of events” (Mises, 1978, p. 47).
It is not only the method of understanding but also the “the category of finality”
(ibid., p. 36) that separates, in Mises’ view, the sciences of human action from the
natural sciences.20 As he puts it: “The natural sciences are causality research; the
sciences of human action are teleological” (ibid., p. 7).21 While the former argue in
terms of cause and effect, the latter argue in terms of “final causes” (ibid., p. 37), in
terms of “the ends sought by acting men in pursuit of their own designs” (ibid., p. 7).
It is, so Mises notes, the finality or ends-oriented nature of purposeful human action
that makes it impossible to “apply to the behavior of man the same methods the
natural sciences apply in dealing with the behavior of mice or of iron” (ibid., p. 37).
As an additional, and separate, argument in support of his autonomy claim Mises

finally stresses the fact that the sciences of human action deal with phenomena the
complexity of which precludes the kind of experimental research that he considers

constitutive for the natural sciences. As he puts it: “The experience to which the
natural sciences owe all their success is the experience of the experiment . . . The
social sciences cannot make use of experiments. The experience with which they
have to deal is the experience of complex phenomena” (1990, p. 5).22
Mises is generally recognized as the single most influential inspiration for
modern Austrian economics. Even if not all aspects of his theoretical system –
such as, for instance, his apriorism – find general acceptance, it is, in particular,
his methodological dualism that seems to have become a rarely if ever disputed
core tenet of the Austrian paradigm. I. Kirzner (1978, p. vi) certainly expresses a
commonly shared view when he approvingly notes: “Economics, Mises explained
again and again, is a discipline the character of which differs drastically from that
of the natural sciences.”
More recently the methodological dualism in the Austrian school has been
reinforced and given a new direction by the “radical subjectivism” of Ludwig
Lachmann. Though Lachmann does not agree with Mises’ praxeological apriorism,
he too insists that it is the intentionality and purposefulness of human action
as a “manifestation of the human mind” that commands “the methodological
independence of the social sciences” (Lachmann, 1977, p. 61).23 Taking his lead
from Max Weber24 and the “understanding-approach” (Verstehende Soziologie)
in early twentieth century German sociology, Lachmann argues that the method
of Verstehen “which aims at the discovering the meaning of things” (ibid., p. 49)
is the adequate method for Austrian economists for whom “the thought design,
the economic calculation or economic plan of the individual, always stands in the
foreground of theoretical interest” (ibid., p. 47).25 It is a method, he notes, that is
“closed to the natural sciences” (ibid., p. 58) and that “apparently conflicts with
most methods used in and suitable to the natural sciences” (ibid., p. 49).
While Mises and Lachmann equally emphasize the role of understanding as
the principal method of the sciences of human action, a method that they see in
contrast to behavioristic approaches,26 there is a significant difference in their

respective interpretations of this method. As has been noted above, to Mises
it is an exercise in psychological or – as he prefers to call it – “thymological”
reasoning, it is an inquiry into the intentions and expectations that motivate
human choices. By contrast, and in a somewhat ambiguous manner, Lachmann
insists that as a “principle of explanation” the method of Verstehen “has nothing
to do with psychology” (Lachmann, 1977, p. 155). As an “analysis of observed
phenomena in terms of pre-existing plans” (ibid.), he argues, it is concerned “with
purposes, not with motives, with plans, not with the psychic processes which give
rise to them, with acts of our conscious minds, not with what lies behind them”
(ibid.).27
Lachmann’s version of the Verstehen approach has become the inspiration for
a rather peculiar – and highly controversial (Vaughn, 1994, pp. 127ff) – variety
in modern Austrian economics, described by its advocates as “interpretative turn”
(Lavoie, 1994a, p. 1), a turn that is claimed to be “simply a vindication of the tra-
ditional Austrian positions against positivist neoclassical philosophical attitudes”
(Lavoie, 1994b, p. 54). The “hermeneutical Austrians,” as Don Lavoie (ibid., p. 55)
and other champions of this “turn” call themselves, share the Austrian emphasis on
subjectivism, “methodological dualism” (Lavoie, 1991, p. 477) and the rejection of
a “crude behavioristic stimulus-response scheme” (ibid., p. 486). And they endorse,
in particular, Lachmann’s view of economics as “a subject that needs to deploy the
methods of Verstehen, of the interpretation of meaning, and not think of itself as
a natural science trying to establish causal laws” (Lavoie, 1994a, p. 9). It is their
main tenet, however, that, in order to be true to Lachmann’s “vision of what a truly
interpretive approach to economics can be like” (ibid., p. 11), Austrian economists
must adopt a different – namely “hermeneutical” – methodological foundation for
their analytical efforts than what has been practiced in the Austrian tradition so far.
As Lavoie (1994b, p. 55) phrases it: “The hermeneutical Austrians contend that
none of the Austrians’ usual methodological positions, from Aristotelianism to
Kantianism, to Neo-Kantianism, to Popperianism, adequately captures the virtues
of their own substantive economics, and that contemporary hermeneutics does.”
In embracing the “post-modernist philosophical position” (Lavoie, 1994a, p. 5)
to be found in the “philosophical hermeneutics” of such authors as Martin
Heidegger, Hans-Georg Gadamer, Paul Ricoeur or Richard Rorty (Lavoie, 1994b,
p. 54), the “hermeneutical Austrians” hope to find the appropriate methodology
that, finally, allows the subjectivism of the Austrian tradition to be developed to
its full potential, namely to the “economics of meaning” (Lavoie, 1994a, p. 9)
that Lachmann envisioned. The essential merit of these “contemporary versions
of the Verstehen approach” (Lavoie, 1991, p. 482) is, according to Lavoie, that
they imply a reinterpretation of what “subjectivism” is about, away from the idea
that “meaning is ‘internal’ to a single mind” (ibid., p. 472) towards the notion that
meaning “resides ‘within’ the inter-subjective.”28 Though it is by no means easy
to decipher what such reorientation of subjectivism is exactly meant to be about, it
is apparent that Lavoie sees the distinguishing characteristic of the hermeneutical
approach that he endorses in the fact that it is not about “seeing things from the
agent’s point of view” (ibid.), but is concerned with a kind of “meaning” that is
separable from the intentions of the acting person.
Whether or not such “hermeneutical subjectivism” amounts at all to a research
program that can be meaningfully classified as “Austrian” need not be of concern
in the present context.29 Of principal interest here is the fact that Lavoie and his
fellow “hermeneutical Austrians” not only subscribe to but even reinforce the
methodological dualism that Mises had made a trade-mark of the Austrian school,
and that to them, as for Mises, it is its concern with “the purposes of the individual
human beings who make up an economy” (Lavoie, 1991, p. 474) that requires
economics to adopt a methodology entirely different from that practiced by the
natural sciences.30
3. METHODOLOGICAL DUALISM
CRITICALLY EXAMINED
So far I have mainly recorded the principal arguments that have been put forward
in the Austrian tradition in support of the claim that economics, if it is to take
proper account of the intentionality and purposefulness of human action, must
adopt a methodology fundamentally different from that of the natural sciences.
The purpose of the present section is to critically examine these arguments, while
the remaining sections of this paper will discuss theoretical approaches that, in
contrast to the dualism claim, adopt an explicitly naturalistic outlook at human
action.
Methodological arguments are meta-theoretical arguments or statements about
theories. They come either in the form of normative principles that are supposed to
guide scientific inquiry into fruitful directions, or in the form of factual judgements
about certain properties of theories. If they are meant as normative principles,
methodological arguments can be neither true nor false. They can only turn
out to provide more or less fruitful guidance to scientific inquiry, and they can
prospectively be judged in terms of their likely effects. If they are meant as factual
claims about theories, they can be true or false. They can be critically examined in
terms of the viability of their premises and they can be confronted with the factual
properties of theories.
In the sense of the noted distinction, methodological dualism can be meant
either as a normative principle or as a factual claim. If it is meant as a normative

principle, as the recommendation that the sciences of human action should adopt in
their explanatory endeavors a methodology categorically different from the causal-
explanatory approach of the natural sciences, it has to be judged in terms of its
“fruits,” i.e. the insights it generates, compared to the “fruits” of research programs
that aim at a naturalistic, causal explanation of human action. If it is meant as the
factual claim that human action, for principle reasons, simply cannot be explained
in terms of naturalistic, causal theories, it is either true or false. And whether it
is one or the other has to be judged in terms of the sustainability of the reasons
cited in its support and in light of the factual state of our theoretical knowledge.
It must be regarded as questionable when its premises can be shown to be faulty,
and it must be regarded as falsified when theories become available that allow for
a causal explanation of human action.
The “Austrian” defense of methodological dualism, summarized above, is
essentially based on the factual claim that a causal explanation of intentional
human action is impossible. If this claim is meant to say that we have no theories
that would provide a satisfactory explanation, it must be judged in terms of our
current theoretical knowledge and may have to be revised as new knowledge
becomes available. If it is meant to say that such explanation is “in principle”
impossible, because of the “nature” of human action, it ignores the simple fact that
we have no other knowledge of the “nature” of things, including the “nature” of
human action, than what our theories tell us. To make such “in principle” claims
means, therefore, that one pretends to know more than one possibly can know,
namely more than what our currently available theories tell us. It follows that the
dualism claim deserves serious discussion only if it is meant in the first sense,
i.e. as a judgement about the state of our theoretical knowledge at a given point
in time.
It helps to avoid confusion about the methodological status of the social sciences
to carefully distinguish at the outset between two issues, namely, on the one hand,
the issue of whether the subject matter of the social sciences includes “subjective”
phenomena, by contrast to the “objective” phenomena studied by the natural
sciences, and, on the other hand, the issue of whether the methods applied in
studying these phenomena should be or need to be “subjective,” by contrast to the
“objective” methods of the natural sciences (Boehm, 1982, p. 44). Only the second
issue is of relevance here. That the social sciences deal with subjective phenomena
such as people’s purposes, intentions, expectations, beliefs, etc. is not controversial
at all among defenders of the methodological autonomy of the social sciences and
their opponents. The controversy is about whether their concern with subjective
phenomena requires the social sciences to adopt a methodology fundamentally
different from that of the natural sciences.31
With regard to the latter issue, a major source of ambiguity in the Austrian
rejection of a natural sciences approach to social science is the failure to carefully
distinguish between the rejection of a “mechanistic” approach and the rejection
of a causal explanatory approach to social phenomena. When Mises (1978, p. 3)
charges that the “study of economics has been again and again led astray by the
vain idea that economics must proceed according to the pattern of other sciences”
it makes quite a difference whether this is meant as a critique of attempts to
model economics “after the scheme of classical mechanics” (ibid., p. 39), or
whether it is meant as a rejection of the notion that the logic of explanation
is the same in all empirical sciences, in the social sciences no less than in the
natural sciences. It is the second and not the first issue that is controversial among
advocates of methodological dualism and their opponents. Both sides can easily
agree in their critique of an economics that – like “Walrasian general equilibrium
analysis” (Lavoie, 1994a, b, p. 10) – transfers the mechanical paradigm of
classical physics to the study of human action, interpreting the world of social
interaction in analogy to the world of interacting physical forces (Mirowski, 1989).
What they disagree about is the issue of the methodological autonomy of the
social sciences.
F. A. Hayek has explicitly distinguished between the two issues, and this is
the proper place to look at his arguments, anticipating the later, more general
discussion of his approach (Section 6). As has often been noted, one can find
in Hayek’s writings, especially in The Counter-Revolution of Science, remarks
on “the contrast between the subjectivist approach of the social sciences and the
objectivist approach of the natural sciences” (Hayek, 1979, p. 47) that seem quite
similar to Mises’ or Lachmann’s dualistic outlooks. In particular his critique of the
methodological attitude that he describes as “scientism or the scientistic prejudice”
(1979, p. 24)32 has been cited as evidence of such similarity (Dufourt & Garrouste,
1992, p. 28). Yet, a careful reading of Hayek’s arguments clearly shows that the
target of his criticism of “scientism” is not an empiricist, naturalistic approach
to social phenomena but an approach that seeks to mold the social sciences along
the “physics model”(Hayek, 1982, p. 289) in “slavish imitation of the method
and language of science” (Hayek, 1979, p. 24), based on a “superficial similarity
of procedure with that of the physical sciences” (Hayek, 1978a, p. 31). His
rejection of a “scientistic” approach to the social sciences that mimics the specific
paradigm and analytical apparatus of physics does not imply, however, that he
endorses Mises’ apriorism or Lachmann’s anti-naturalistic hermeneutics (Witt,
1992, p. 222).
In fact, in contrast to Mises’ verdict, quoted above (p. 4), Hayek on a number
of occasions has explicitly endorsed K. R. Popper’s falsificationist methodology
(Hayek, 1973, pp. 29, 146; 1976, p. 43; 1978a, p. 31), for instance, when he notes
about Popper’s “critical rationalism”: “It seems to me the best term for describing
the general position which I regard as the most reasonable one” (1967d, p. 94).
The fact that in the preface to his “Studies in Philosophy, Politics and Economics”
(1967, p. viii) Hayek attributes a “slight change” in the tone of his critique of
“scientism” to the influence of Karl Popper has often been quoted in support
of the conjecture that Hayek’s original views on the methodology of the social
sciences were much more in agreement with Mises’ methodological dualism.
Such interpretation is, however, not only difficult to reconcile with Hayek’s early
call for an empiricist approach to “mental phenomena” (1979, p. 48) to which I
will return below (Section 6). It is also in contrast to Hayek’s own assessment of
the matter.33
While he shares Popper’s general methodological outlook, Hayek emphasizes,

though, that the criterion of falsifiability must be interpreted somewhat differently
in the social sciences than in a science like physics, because of the complexity of
the phenomena that they study. In this sense he is in perfect agreement with Mises’
argument that the “experience with which they (the social sciences, V. V.) have to
deal is the experience of complex phenomena” (Mises, 1990, p. 5). By contrast
to Mises he does not draw the conclusion, however, that this is a reason to claim
methodological autonomy for the social sciences, nor does he conclude that the
complexity of the phenomena they study prevents the social sciences from aiming
at causal explanations. Hayek’s conclusion rather is that, because of the specific
problems that arise “in connection with those essentially complex phenomena
of which social structures are so important an instance” (Hayek, 1978a, p. 32),
our explanatory ambitions must be more modest in the social sciences than in
the physical sciences. While the latter can aim at “complete explanations” and
“specific predictions,” Hayek sees the social sciences generally limited to what
he calls “explanations of the principle” and “pattern predictions” (Hayek, 1967a,
p. 11; 1979, p. 86). Yet, this is, as Hayek points out, a limitation that the social
sciences share with the natural sciences to the extent that the latter deal with
complex phenomena as well.34
The methodological distinction that Hayek emphasizes cuts across the
traditional distinction between the social and the natural sciences. It is a
distinction between sciences that study “relatively simple phenomena” (Hayek,
1967b, p. 25), or “closed systems” with a “sufficiently small” number of
connected variables (Hayek, 1967a, p. 3), and sciences that study “the more
complex phenomena of life, of mind, and of society” (Hayek, 1967b, p. 25),
such as, in particular, evolutionary biology (Hayek, 1967a, p. 11).35 According
to Hayek, what is true for the theory of evolution by natural selection can be said
about theories of complex phenomena more generally, namely that they provide
“an account of a process the outcome of which will depend on a very large number
of particular facts, far too numerous for us to know in their entirety” (Hayek, 1973,
p. 23). Even if we understand the general laws that are in operation in the respective
realm, due to our lack of knowledge of the particular facts that determine the
course of events, we will not be able to come up with more than an “explanation
of the principle” and a “prediction merely of the abstract pattern the process will
follow” (ibid.).
Such limitation to explanations of the principle and to pattern predictions does
not mean, so Hayek argues, that we have to abandon a falsificationist methodology.
Because and to the extent that they exclude conceivable courses of events, theories
of complex phenomena are empirically contentful and refutable (Hayek, 1967b,

p. 28),36 even if we need to recognize that the “degree of falsifiability necessarily
decreases” (Hayek, 1973, p. 29), compared to theories of more simple phenomena.

This inherent “drawback” of theories of complex phenomena37 is in Hayek’s view
neither a reason for the social sciences to claim methodological autonomy, nor
should it be considered a defect that they need to overcome in order to be more
“scientific” (Hayek, 1967a, p. 16). Instead, he argues, for sciences of complex
phenomena such as economics “it may prove necessary deliberately to cultivate
the technique of explanation of the principle” (ibid., p. 21).38
The principal lesson from Hayek’s argument on the complexity issue is that
we must carefully distinguish between two claims, namely, on the one hand,
the claim that the complexities of human action and of social phenomena put
limits on the degree of specificity and falsifiability that we may hope to achieve in
our explanatory efforts and, on the other hand, the claim that these complexities
make it necessary for the sciences of human action to abandon the standard
concept of causal explanation in favor of a different methodology. This distinction
is of immediate relevance for the principal argument that Austrian advocates
of methodological dualism rest their case on, namely that the subjectivity of
intentional human action precludes any causal, “mechanistic” explanation.
If this argument is meant to say that the complexity of the mental processes
from which human actions result make it, as a rule, impossible for us to fully
explain and to predict specific actions, it is perfectly compatible with Hayek’s
view on the methodological status of theories of complex phenomena. In this
interpretation, Austrian subjectivism can well be accommodated within the
standard notion of causal explanation, more broadly understood to include not
only “deterministic” explanations but explanations of the principle as well.39 The
case for methodological dualism can, therefore, not be based on the complexity
argument alone, but needs to be supported by additional arguments for why a causal
explanation of intentional human action should not be possible. The versions of
Austrian subjectivism that I have summarized above (Section 2) appear to rest their
case for methodological dualism essentially on two arguments. This is, on the one
side, the argument that, because human action is based on subjective intentions and
expectations, understanding these intentions and expectations rather than causal
analysis is the appropriate method of inquiry. And this is, on the other side, the
argument that, because human action is purposeful and goal-directed, it must be
explained in teleological rather than in causal terms.
In examining the “understanding” issue one should, again, at the outset separate
what is controversial and what is not controversial among the advocates and the
opponents of methodological dualism. Our ability to successfully interact with
others is, quite obviously, based on our capacity to infer their intentions, beliefs,
and expectations from their observed behavior as well as on our capacity to predict
their behavior, based on assumptions about their motivations. That we can, in this
sense, “understand” other persons’ actions, in our daily lives as well as in our role
as social scientists, is not controversial at all among advocates of methodological
dualism and their opponents. What the latter oppose is the claim that “this fact must
lead to specific methodological problems which require a general methodology for
the social sciences which differs from the methodology of the natural sciences”
(Albert, 1988, p. 575).
When we seek to “understand” other persons’ behavior, we seek to identify
mental states, namely intentions, beliefs, expectations, etc., that would make the
behavior “meaningful” or consistent. In doing this, we form conjectures about
what kind of intentions, beliefs, etc. may have actually motivated the behavior
that is to be explained (Koppl, 1994, p. 72). These conjectures may be true or
false, and the crucial issue is how we can decide among potential “competing
conjectures about subjective meaning” (ibid., p. 73). The heuristic strategy to
imagine ourselves in the position of the person whose behavior we want to explain
may be a useful tool for generating such conjectures, but it clearly does not qualify
as an inter-subjective test for choosing among competing conjectures. The latter
requires us to refer, in some manner, to observable indicators for the, themselves
unobservable, intentions, beliefs, etc. that are imputed to the actor (Witt, 2003,
p. 24). Obviously, if one wants to avoid circular reasoning, the behavior that is to
be explained cannot at the same time serve as the observable evidence from which
the imputed mental states are inferred (Böhm, 1982, p. 47). One needs to refer
to inter-subjectively accessible evidence that is, on the one hand, independent
of the behavior that is to be explained, and that can, on the other hand, be
conjectured to indicate the actual presence of the mental states that are imputed to
the actor.
As soon as one specifies the method of “understanding” in the above manner,
it becomes very doubtful whether understanding actions is something entirely
different from a causal explanation and whether there is, indeed, any reason
for drawing a categorical methodological line between the understanding-based
analysis of social phenomena and the nomological, causal analysis of physical

phenomena (Albert, 1988, pp. 581, 586). Both, the imputation of mental states as
reasons for action as well as the specification of observable indicators for mental
states imply conjectures about causal relations, namely as conjectures about how
mental states are causally connected to revealed behavior and how mental states,
themselves, are connected to, or causally affected by, observable events.40
What has been said above about the issue of “understanding” is of direct
relevance for the second principal argument in defense of methodological dualism,
namely the claim that, because of its purposiveness and goal-directedness, human
action has to be explained in teleological rather than in causal terms. Here again,
it is useful to separate first the controversial from the uncontroversial. What is
not controversial at all is the fact that human action is goal-directed, based on
“planning ahead,” on foresighted anticipation of its consequences. In this sense it
is, if one wants to use this term, “teleological,” aiming at a !"#o$.41 Controversial
is only whether this undisputed fact implies that human action requires a special,
namely a “teleological” type of explanation, an explanation that is about “final
causation,” “backward causation” or “causation by consequences,” by contrast to
the “efficient causation” of the standard causal model.
Upon closer inspection the claim that human action represents a case of
backward causation or causation by consequences, as opposed to ordinary
causation by antecedent causes, turns out to be little more than an instance of
careless and misleading use of language. Its advocates surely do not want to
say that actions are literally caused by the consequences that materialize as their
effects. What they most probably want to say is that human actions are motivated
or “caused” by their anticipated and desired consequences. Yet, to argue that an
action occurs because the actor expects and wishes it to bring about consequences
of a certain kind is clearly not the same than to argue that actions are caused by
the consequences that result from them. By contrast to the latter the anticipation
of consequences and the desire to bring them about antecede the actions that they
motivate. To account for their role as causes of actions does not require us at all to
abandon the standard notion of “efficient causation.”
Opponents of methodological dualism have no reason to deny that such cognitive
factors as intentions, desires, and beliefs play a causal role in human action,
and they can easily agree with the claim that “the sciences of human action are
teleological” (Mises, 1978, p. 7) if this is only meant to emphasize that explanations
of human action must account for the role of cognitive factors, by contrast to
purely physicalist accounts in the natural sciences. They will, however, insist that
causation by such cognitive factors is still causation in the standard sense, namely
causation by antecedent factors, even if cognitive factors are admittedly causal
factors of a special kind (Mackie, 1974, pp. 285, 295; Meyer, 2002, p. 117).
4. EVOLUTIONARY PSYCHOLOGY: A NATURALISTIC

APPROACH TO MENTAL PHENOMENA
As noted above, methodological arguments are statements about theories.
Methodological dualism is a conjecture about the relation between two bodies of
theories, theories about mental phenomena on the one side, and theories about
“external events” on the other. It claims that, because of inherent differences
between these two realms, theories about “the realm of human thought and action”
(Mises, 1957, p. 1) are of a categorically different nature than theories about
external events. Even though this claim is often stated as if it were an ontological
statement about the “nature” of the respective realms, it can in fact, as I have
argued above, never be more than a conjecture about the state of our theoretical
knowledge. After all, we cannot know more about the “nature of things” than what
our theories tell us. Accordingly, it is ultimately not in terms of methodological
arguments per se that the claim of methodological dualism must be judged, but in
terms of our theoretical knowledge. This fact is, even if only in a passing manner,
recognized by Mises when he limits his dualism claim to “the present state of our
knowledge” (Mises, 1990, p. 25), implicitly acknowledging that it may have to be
revised in light of future theoretical developments.
The “behaviorist” psychology that Mises accused of neglecting the cognitive
foundations of human action may have, indeed, been a just target of his fundamental
objections against empiricist, naturalistic approaches to human action. Yet,
psychology has not only changed since Mises made his case, it underwent, in
particular, what has been called the “cognitive revolution,”42 and whether Mises’
verdict is still valid today has to be judged in light of contemporary psychological
theories, theories that take a significantly different outlook at mental phenomena
than the “crude behavioristic stimulus-response scheme” (Lavoie, 1991, p. 484)
that Mises and modern hermeneutical Austrians attack.43
The purpose of this section is to take a closer look at evolutionary psychology as
one of the post-cognitive-revolution approaches in modern psychology. To be sure,
the tenets of evolutionary psychology are by no means undisputed and they are
presented here not as unquestionable wisdom but as conjectures that are subject
to further inquiry and critical examination. Yet, evolutionary psychology is of
particular interest in the present context because it represents, on the one side,
a rigorous effort to provide a purely naturalistic, causal account of the human
mind and of human action, and sees itself, on the other side, as an integral
part of the cognitive reorientation in psychology, a reorientation that explicitly
rejects the “anti-mentalism” (Badcock, 2000, p. 111) of behaviorism. According
to L. Cosmides and J. Tooby (1987, p. 302), two main initiators of this research
program, evolutionary psychology is concerned with “the discovery and principled
investigation of the human psyche’s innate cognitive programs.” “It unites modern
evolutionary biology with the cognitive revolution” (Cosmides, Tooby & Barkow,
1992, p. 2).44
While the principal ambition of evolutionary psychology is to better understand
“the evolved information-processing mechanisms that comprise the human mind”
(Cosmides, Tooby & Barkow, 1992, p. 2), its fundamental tenet is that, because
humans are the product of the evolutionary process, the explanation of their
characteristics, including the “architecture of the human mind,” must be sought in

the evolutionary process (Cosmides & Tooby, 1987, p. 278; 1994c, p. 47). Since
the human mind is a product of evolution – so evolutionary psychologists argue –

inquiring into the conditions of man’s evolutionary history should provide clues for
our understanding of how the mind works. They base their research strategy on the
argument that, “given the long human generation time, and the fact that agriculture
represents less than 1% of the evolutionary history of the genus Homo” (Cosmides
& Tooby, 1992, p. 222), our genetic make-up can safely be assumed to be essentially
the same as that of our ancestors who predated the appearance of agriculture and
modern forms of life. Therefore, evolutionary psychologists conclude, the most
reasonable conjecture is that man’s genetically coded psychological mechanisms
are adapted to the problem environment that our ancestors were exposed to for
thousands of generations, living as hunters and gatherers in small bands (ibid.,
p. 219; Cosmides, Tooby & Barkow, 1992, p. 5). And we should expect to find
the kinds of recurrent problems that such a hunting and gathering way of life
typically posed to be reflected in the evolved architecture of the human mind
(Cosmides & Tooby, 1994b, p. 86).45 Inquiring into these adaptive problems,
this is the fundamental heuristic premise of evolutionary psychology, provides “a
powerful engine of discovery” (Cosmides & Tooby, 1992, p. 221), “a heuristic for
generating testable hypotheses about the structure of the cognitive programs that
solve the adaptive problems in question” (Cosmides & Tooby, 1987, p. 302).
It is noteworthy that, in his The Ultimate Foundations of Economic Science
(1978), Ludwig von Mises has pointed in rather similar terms to the evolutionary
origins of the human mind, without drawing what would seem to be a rather
compelling conclusion, namely that, if the human mind is a product of the natural
forces of evolution, it should be a legitimate subject of naturalistic, explanatory
inquiry.46 One might well have expected von Mises to proceed to a naturalistic
theory of the human mind when he argues:
The concepts of natural selection make it possible to develop a hypothesis about the emergence
of the logical structure of the human mind and the a priori.
Animals are driven by impulses and instincts. Natural selection eliminated those specimens
and species which developed instincts that were a liability in the struggle for survival . . . We
are not prevented from assuming that in the long way that led from the nonhuman ancestors
of man to the emergence of the species Homo sapiens some groups of advanced anthropoids
experimented, as it were, with categorical concepts different from those of Homo sapiens and
tried to use them for the guidance of their conduct . . . Only those groups could survive whose
members acted in conformity with the right categories, i.e. with those that were in conformity
with reality and therefore – to use the concept of pragmatism – worked (Mises, 1978, p. 14).
Yet, instead of embarking on an inquiry into what an evolutionary account may

be able to tell us about the structure of the human mind, Mises cuts off such a
research avenue by simply declaring: “However, reference to this interpretation of

the origin of the a priori categories does not entitle us to call them a precipitate of
experience” (Mises, 1978, p. 15). Instead, he insists that the human mind cannot be
adequately approached, in a “Darwinian spirit,” as a natural phenomenon “without
recourse to finality” (ibid.). It is the categorical difference between phenomena of
the human mind and what goes on in non-human nature that, in his view, makes
the mind’s evolutionary origins essentially irrelevant for our understanding of its
functioning (ibid., pp. 1, 8).
One of the principal conclusions that evolutionary psychologists draw from their
Darwinian outlook is a critique of rational choice theories (Vromen, 2002, p. 188).
A theory that interprets the human mind as a general purpose mechanism with
universal, content-independent problem-solving capacity is in their view extremely
implausible from an evolutionary perspective (Cosmides & Tooby, 1992, p. 164).
Recognizing the evolutionary origins of the human mind imposes, as they argue
(Cosmides & Tooby, 1992, p. 108; 1994b, p. 88), constraints on what we can
assume about its problem-solving capacities, namely evolvability constraints and
solvability constraints. One must be able to show that the faculties that one ascribes
to the human mind could, in principle, have evolved under the conditions that, as
far as we know, characterized the evolutionary history of our species.47 And one
must, secondly, be able to show how these faculties enable the mind to, in fact,
solve the problems in question. As Cosmides and Tooby (1992, p. 110) put it:
“To be a viable hypothesis about human psychological architecture, the design
proposed must be able to meet both solvability and evolvability criteria: It must
be able to solve the problems that we observe modern humans routinely solving
and it must solve all the problems that were necessary for humans to survive and
reproduce in ancestral environments.”
It is in terms of the evolvability and solvability constraints that evolutionary
psychologists find standard rational choice theory to provide an implausible
account of human capabilities. As far as the solvability constraints are concerned,
they argue that “no one has yet been able to specify a general learning mechanism
or general cognitive problem solver that has the power to solve the complex array of
adaptive problems faced by humans, either in principle or in practice” (Cosmides &

Tooby, 1987, p. 298). With regard to the evolvability constraints they point out that
evolutionary biology provides a number of reasons why it is implausible to assume
that “the human mind is an equipotential, general-purpose machine” (Cosmides
& Tooby, 1994b, p. 89), and why it is much more plausible to adopt “the contrary
view that the mind is richly textured with content-specialized psychological
adaptations” (Cosmides & Tooby, 1992, p. 165), adaptations that serve to solve
long-standing evolutionary problems (Cosmides & Tooby, 1994a, p. 329; 1994c,
p. 65). Evolutionary psychologists do not deny the presence of domain-general
cognitive mechanisms, their argument is that not all innate cognitive programs can
be of a general-purpose nature but that some of them must be domain-specific,
“functionally specialized procedures that exploit the recurrent properties of the

corresponding domain in a way that would have produced an efficacious solution
under Pleistocene conditions” (Cosmides & Tooby, 1992, p. 221). Such domain-
specific cognitive mechanisms, Cosmides and Tooby (1994b, p. 89) argue, “can
be expected to systematically outperform (and hence preclude or replace) more
general mechanisms,” because – by contrast to the latter – they exploit the stable
structural features of evolutionary recurring situations.48
As a paradigmatic example of such specialized cognitive mechanisms49
evolutionary psychologists point to Noam Chomsky’s (1980) psycholinguistics
and its principal conjecture that the human capacity of language-learning cannot
be explained on the basis of general-purpose mechanisms alone but requires the
presence of a functionally distinct mental organ, a “language acquisition device”
(Badcock, 2000, p. 246; Tooby & Cosmisedes, 1992, p. 95).50 According to
Cosmides and Tooby (1987, p. 290), Chomsky’s contribution marked a turning
point in modern psychology because it drew attention to the idea that “just as the
body has many different organs, each of which is specialized for performing a
different function . . ., the mind can be expected to include many different ‘mental
organs’.”
An area to which evolutionary psychologists have paid special attention in their
search for domain-specific cognitive mechanisms is cooperation for mutual benefit
and, in particular, social exchange. As Cosmides and Tooby (1994c, p. 48) put it:
The adaptive problems that arise when individuals engage in this form of cooperation have
constituted a long-enduring selection pressure on the hominid line. Paleoanthropological
evidence indicates that social exchange extends back at least two million years in the human
line, and the fact that social exchange exists in some of our primate cousins suggests that it
may be even more ancient than that. It is exactly the kind of problem that selection can build
cognitive mechanisms for solving.
In a world in which mutually beneficial cooperation is a more sustainable

source of gains than unilateral exploitation (Cosmides & Tooby, 1992, p. 207),
cognitive programs that better enable individuals to successfully engage in

mutually beneficial exchange can be expected to outperform other mechanisms.
And here, as in other areas of human problem-solving, evolutionary psychologists
conjecture that specialized programs are likely to be better adapted to handle
recurrent problem-situations than general-purpose algorithms, and that we should,
accordingly, expect the human mind to contain specialized mechanisms designed
for reasoning about social exchange (ibid., pp. 163, 207; 1994a, p. 330; 1994c,
p. 49). Such specialized adaptations to social exchange are, as they argue,
particularly important where the complexities of non-simultaneous trade are
concerned (Cosmides & Tooby, 1992, p. 169) and where, for instance, the capacity
to reliably separate reciprocating cooperators from exploitative cheaters is an
essential prerequisite for lasting success (ibid., pp. 181ff).

Another issue that evolutionary psychologists have focused their research
interest on, and one that is of particular interest in the context of this paper, concerns
the human ability to “read other people’s minds” and to anticipate their likely
behavior. By contrast to methodological dualists, like Ludwig Lachmann, who
take this human ability, the capacity of “Verstehen” or “understanding,” as evidence
supporting their anti-naturalistic methodology, evolutionary psychologists seek to
provide a naturalistic, explanatory account of this very capacity as “an evolved
attribute of human beings” (Badcock, 2000, p. 114).51 As Cosmides and Tooby
(1994b, p. 101) explain:
Another important set of evolutionary long-enduring regularities is the recurrent design features
of other human minds. Evolved domain-specific cognitive specializations are even more
necessary in this area, not only because other minds constitute the single most important
selective force facing any individual human, but also because mental states such as beliefs,
motives, intentions, and emotions cannot be directly observed. To allow a human to represent at
least some of the mental states that generate other’s behavior, special inferential systems must
be available to bridge the gap from the observable to the unobservable. For example, if there is
a reliable correlation over evolutionary time between the movement of human facial muscles
and emotional state or behavioral intentions, then specialized mechanisms can evolve that infer
a person’s mental state from the movement of that persons facial muscles.
According to Cosmides and Tooby (ibid., p. 102), research efforts in the field
of cognitive development provide substantial support for the hypothesis that our
evolved psychological architecture includes a specialized “theory of the mind
module” that enables us to reliably develop models of other human minds.
As noted before, a relatively new research program like evolutionary psychology
is, of course, not undisputed (Badcock, 2000, pp. 17ff, 108). What is important
in the present context is, however, the general research program on which
evolutionary psychology is based, not specific claims that are made by various
researchers associated with this program. Objections that may well be raised with
regard to such particular claims should not distract from the fact that the general
paradigmatic outlook of evolutionary psychology is shared by a variety of modern
approaches in the cognitive sciences, in psychology and in evolutionary biology
that seek to study the human mind from a naturalistic, explanatory perspective
(Tooby & Cosmisedes, 1992, p. 93). This general outlook interprets the human
mind as an adaptive, information-processing system that translates informational
input into behavioral output. It seeks to explain the working principles of this
information-processing system in light of the fact that its structure has been
shaped by evolutionary forces. And it assumes that the human mind, like all
information-processing systems, can be described in mutually compatible and

complementary ways, on the one hand in terms of its physical components, or
its “hardware,” and on the other hand in terms of the programs that govern its
operations, its “software” (Cosmides & Tooby, 1987, p. 283; 1994c, p. 44).
It is in terms of such interpretation of the mind as an information-processing
system that L. Cosmides, J. Tooby and J. H. Barkow (1992, p. 8) note in their
Introduction to The Adapted Mind – Evolutionary Psychology and the Generation
of Culture:
The brain takes sensorily derived information from the environment [including the organism’s
‘internal’ environment, V. V.] as input, performs complex transformation on that information,
and produces either data structures (representations) or behavior as output. Consequently, it
too, can be described in two mutually compatible and complementary ways. A neuroscience
description characterizes the ways in which its physical components interact; a cognitive, or
information-processing, description characterizes the ‘programs’ that govern the operation. In
cognitive psychology, the term mind is used to refer to an information-processing description
of the functioning of the brain . . . An account of the evolution of the mind is an account of
how and why the information-processing organization of the nervous system came to have the
functional properties that it does.52
According to this outlook at the mind, human problem-solving behavior is guided

by programs53 that incorporate “knowledge” of relevant contingencies in man’s
problem environment and that enable humans to respond more speedily and
effectively to typical, recurrent problem situations than an exclusively general-
purpose device, called “rationality,” would allow them to do.
Even though its focus is on the kind of “knowledge” that is the result of natural
selection and that is “stored” in genetically encoded programs (Cosmides & Tooby,
1987, p. 287) evolutionary psychology does not deny the role of general-purpose
devices and the plasticity of human learning. Evolutionary psychologists insist,
however, that such general-purpose devices and human learning can only work
with the aid of domain-specific cognitive mechanisms that selectively pre-structure
the otherwise unmanageable complexity of the world. As Tooby and Cosmisedes
(1992, p. 105) put it: “A mechanism unaided by domain-specific rules of relevance,
specialized procedures, ‘preferred’ hypotheses, and so on could not solve any

biological problem of routine complexity in the amount of time the organism has
to solve it, and usually could not solve it at all.”54
The clue to understanding the impressive problem-solving capacity and
cognitive power of the human mind, this is the principal tenet of evolutionary psy-
chologists, is to be found in the richness of its evolved content-specific mechanisms.
They hold “that the human capacity for adaptive flexibility and powerful problem-
solving is so great precisely because of the number and the domain-specificity of
the mechanisms we have” (ibid., p. 113), and they reject accounts that attribute
such capacity entirely to domain-general, content-independent mechanisms. Not

only are such accounts, in their view, highly implausible in light of accumulating
research-evidence (ibid., p. 103). To attribute man’s cognitive abilities to a globally

defined capacity “rationality” means, they censure, merely to put a label on
a phenomenon instead of actually examining it (ibid., p. 113). As Tooby and
Cosmisedes (1992, p. 122) phrase it rationality “is not an explanation for anything,
but is rather a phenomenon that itself requires explanation.”
5. PURPOSIVE ACTION AS
PROGRAM-BASED BEHAVIOR
Methodological dualism represents what Tooby and Cosmisedes (1992, p. 21)
describe as a “doctrine of intellectual isolationism,” a doctrine that, as they note,
“has been the reigning view in the social sciences” (ibid.). It is one of the more
wider ranging ambitions of evolutionary psychology to oppose the doctrine of
isolationism and to advance, instead, a “conceptual unification” (ibid.), that seeks
to systematically integrate explanatory efforts in the social sciences with insights
that are generated in neighboring fields such as, for example, cognitive science,
evolutionary biology, or neurobiology.55 Evolutionary psychologists see their own
explanatory efforts embedded in a number of theoretical developments in several
disciplines that point towards, as Tooby and Cosmides (ibid., p. 23) call it, an
“Integrated Causal Model” of human behavior, a model that “connects the social
sciences to the rest of science” (ibid., p. 24).
As noted before, evolutionary psychology is, indeed, just one among a number
of recent research efforts that, in mutually compatible ways, seek to provide a
naturalistic account of the cognitive foundations of human behavior (Vanberg,
2002, 2004). The central theoretical notions around which these research efforts
center have been captured, in a particularly pointed manner, in Ernst Mayr’s
concept of teleonomic or program-based behavior and in K. R. Popper’s concept of
conjecture-based problem-solving. Their arguments deserve to be, at least briefly,
discussed here.
Evolutionary biologist Ernst Mayr has suggested an outlook at purposeful
behavior that locates the capacity for goal-directedness and intentionality in the
presence of behavioral programs that are encoded in an organism and enable the
organism to anticipate the consequences of its own actions.56 Mayr refers to
such purposeful, intentional behavior as teleonomic behavior in order to avoid
the ambiguities of the term “teleological”57 as well as to emphasize that his
ambition is to provide a nomological, causal explanation of goal-directedness and
intentionality. As he explains: “A teleonomic process or behavior is one that owes
its goal-directedness to the operation of a program” (Mayr, 1992, p. 127). And

such program-based behavior, he argues, can be explained in causal terms, “there
is no conflict between causality and teleonomy” (Mayr, 1961, p. 1504).

Expressly adopting the concept from information theory, Mayr defines a
program as “a set of instructions” (Mayr, 1992, p. 128), as “coded or prearranged
information that controls a process (or behavior) leading toward a goal”
(ibid., p. 127). Programs can serve to guide goal-directed behavior because
they incorporate conjectural knowledge of the world, knowledge of relevant
contingencies based on which likely consequences of alternative courses of action
can be anticipated. Such programs can, in principle, be stated as instructions or
decision-rules of the form, “if problem of type A is encountered, then action of
type X is a suitable response.”58 An important consequence of looking, as Mayr
suggests, at purposeful action as program-based behavior is that it draws one’s
attention to two principal issues. This is, firstly, the question of how programs are
coded in an organism and how they become adapted to the organism’s respective
problem-environment. Mayr uses the term encoding to refer to this issue. And this
is, secondly, the question of how coded programs can be implemented to provide
guidance in specific choice situations, an issue that Mayr refers to as decoding.
Decoding is about solving the problem of applying programs or decision-rules
that are about types of problem-situations and types of actions to specific choice-
situations, situations that are always unique in their particular constellations of
circumstances.
Encoding can, as Mayr points out, be explained as a feed-back process that
establishes a systematic link between the performance of programs (success
or failure) and their future role in guiding behavior. The two main processes
through which such encoding occurs are evolution, affecting genetically encoded
programs, and learning, affecting memory-coded programs. The encoding that
the evolutionary biologist studies – and that evolutionary psychologists focus
their attention on – is governed by the feed-back process of natural selection
of successful programs, a process through which “knowledge” is incorporated in
genetic codes that allow for adapted behavior. The genetically coded programs,
in turn, provide the foundation on which – or the framework within which – the
learning of memory-coded programs occurs.59
The distinction between encoding and decoding corresponds to Mayr’s
distinction between ultimate and proximate causes of behavior and between the
explanatory focus of the “functional biologist” and the “evolutionary biologist.”
Proximate causes of behavior are the programs that govern the responses of the
individual to the immediate factors of the environment (Mayr, 1961, p. 1503).
Ultimate causes of behavior are the factors that have shaped these programs. The
functional biologist who “deals with all aspects of the decoding of the information
contained in the DNA code” (ibid., p. 1502) is interested in the proximate causes
of behavior. By contrast, the evolutionary biologist who studies “the laws that
control the changes of these codes from generation to generation” (ibid., p. 1502)
is interested in the ultimate causes that “are responsible for the evolution of the
particular DNA code of information with which every individual of every species
is endowed” (ibid., p. 1503).
To explain behavior in terms of proximate causes means to employ conjectures
about the “psychological mechanisms” or “programs” that allow an individual to
cope with problems faced in its environment.60 Looked at from such a perspective,
the method of “Verstehen” or “understanding” can be interpreted as a particular
research strategy for conjecturing about “proximate causes” of behavior, namely
the strategy of imagining oneself in the choice-situation faced by the individual
whose behavior is to be explained. As discussed above, the problem the Verstehen-
approach faces is that, if one wants to escape circular reasoning, one needs to
specify operational methods for how the conjectures generated by “understanding”
can be critically examined. Mayr’s distinction between proximate and ultimate
causes of behavior suggests ways in which conjectures about “psychological
mechanisms” or “programs” can be checked, namely in light of conjectures about
how they themselves have come to be what they are. Conjectures about genetically
coded cognitive mechanisms, for instance, can be critically analyzed in light
of the question of whether they could have evolved in the kind of problem-
environment to which our species was exposed in evolutionary time.61 And
conjectures about individually acquired, memory-coded behavioral dispositions
can be critically questioned as to whether they could have been plausibly learned
in the kind of environment to which the individual in question was exposed. In
other words, conjectures about ultimate causes can provide indirect evidence for
testing conjectures about proximate causes.
As Mayr emphasizes, to claim that teleonomic purposiveness can be explained
in causal terms does not mean to ignore that “causality in biology is a far cry
from causality in mechanics” (Mayr, 1961, p. 1506). It rather means that the
differences that divide these fields are a matter of complexity and not of different
principles of explanation.62 In terms similar to Hayek’s comments on theories

of complex phenomena, Mayr argues that it is because of the individuality and
uniqueness characteristic of the organic world that “explanation will often have
to be so unspecific and so purely formal that its explanatory value can certainly
be challenged” (ibid., p. 1503). Yet, to recognize such limits in the specificity
of the explanations that can be provided does not mean at all to “question the
ultimate causality of all biological phenomena” (ibid.). Nor should the fact that
the complexity of biological phenomena often precludes prediction be mistaken
as a “lack of cause” (p. 1505). As Mayr (ibid., p. 1506) puts it: “The complexities
of biological causality do not justify embracing nonscientific ideologies . . . but

should encourage all those who have been trying to give a broader basis to the
concept of causality.”
A general outlook at purposeful action that is very similar to E. Mayr’s notion of
program-based behavior has been suggested by K. R. Popper who argues that the
behavior of all organisms – including purposeful human action – can be viewed
as conjecture-based problem-solving behavior.63 Problem-solving, Popper (1982,
p. 150) states, is what all behavior – in fact, all life – is about. Like Mayr’s
approach, Popper’s suggested outlook is meant to draw attention to the fact that
the capacity to solve problems presupposes knowledge of relevant contingencies
in the problem-environment. It posits that, as problem-solving agents, all living
beings are guided in their behavior by pre-existing expectations about the world
around them, expectations that constitute the organism’s conjectural knowledge
of the world. And all such knowledge is, so Popper notes, the result of conjecture
and refutation, of “competitive tentative solutions and the elimination of error”
(Popper, 1972, p. 145).64
According to Popper, it is only in the light of its repertoire of conjectural
expectations that an organism can perceive and identify problems. In other
words, perception is always a theory-impregnated act of selective interpretation
(ibid: p. 343). This is, as Popper argues, no less true for our most elementary,
un- or subconscious sensory perceptions than for our most reflected scientific
observations.65 And just as we can only perceive our environment in the light
of pre-existing conjectural expectations it is, he insists, only on the basis of our
conjectural knowledge of the world that we can act and respond to the problems we
face. Using the same language as E. Mayr, Popper speaks of such action-guiding
conjectural knowledge as “action programs” (Popper & Eccles, 1990, p. 134), as
“dispositions to act, or to behave” (ibid., p. 130).66
The principal mechanism that governs the acquisition of expectations and
conjectural knowledge is, as Popper supposes, essentially the same for all kinds of
“knowledge,” whether it is incorporated in sense organs (Popper, 1972, p. 72) or
in explicitly stated scientific theories. As he puts it: “From the amoeba to Einstein,
the growth of knowledge is always the same: we try to solve our problems, and
to obtain, by a process of elimination something approaching adequacy in our
tentative solutions” (Popper, 1972, p. 261). All acquired knowledge – whether
it is acquired in the process of genetic evolution or in the process of individual
learning – consists, this is Popper’s central claim, in the modification “of some
form of knowledge, or disposition, which was there previously, and in the last
instance of inborn expectations” (ibid., p. 71).
Popper’s theory of conjecture-based problem-solving and of the growth of
knowledge through trial-and-error elimination has become a major inspiration
for evolutionary epistemology,67 a research paradigm that has been developed

independently of (and prior to) evolutionary psychology, but is closely related to the
latter (Vanberg, 2002, pp. 33ff).68 The central tenet of evolutionary epistemology is
that all knowledge processes, i.e. processes that lead to an expansion of knowledge
or problem-solving capacity, can be interpreted as instances of the “variation and
selective retention process of evolutionary adaptation” (Campbell, 1974, p. 450),
where the latter is broadly understood to include genetic evolution as well as
cultural evolution and the “variation and selection processes” that constitute
individual learning (Vanberg, 1994, pp. 174ff).69
What makes evolutionary epistemology particularly noteworthy in the present
context is the fact that F. A. Hayek is counted among its principal founders.70 In
the section below I shall take a closer look at Hayek’s approach, an approach that
exemplifies an interpretation of Austrian subjectivism that is markedly different
from the methodological dualism advocated by authors like Mises, Lachmann and
the “hermeneutical Austrians.”
6. F. A. HAYEK: AN ALTERNATIVE AUSTRIAN

RESEARCH PROGRAM
By contrast to the versions of Austrian subjectivism discussed above (Section 2),
F. A. Hayek advocates what one may describe as “naturalistic subjectivism.” He has
developed a research program that seeks to provide economics with a behavioral
foundation entirely compatible with an empiricist methodology. Beyond the few
remarks made above (Section 3), I shall not address here the often discussed issue
of whether there has been a systematic shift in Hayek’s methodological outlook
since the late 1930s and early 1940s “in a direction away from Mises toward
Popper” (Boehm, 1982, p. 50). What is of relevance in the present context is that
already in his programmatic 1936 Presidential Address to the London Economic
Club (Hayek, 1948a) Hayek explicitly outlined an empiricist research agenda for
a subjectivist economics, a research agenda that he has systematically explored
in various directions in his later work on markets as systems of communication

and on cultural evolution as a process of “collective learning” (Vanberg, 1994), as
well as in what is often described as the most unusual part of his life-work, his
contribution to theoretical psychology (Hayek, 1952).
In “Economics and Knowledge,” Hayek takes issue with an economic theory
that fails to address what must, in his view, be a principal concern of an explanatory
economics, namely the knowledge problem, i.e. the problem of how effective
economic coordination is achieved in the real world among agents who possess
only subjective, incomplete and imperfect pieces of knowledge of the complex
contingencies on which the successful pursuit of their plans depends.71 The

principal target of Hayek’s critique is a formal equilibrium analysis which evades
“any real solution of the problem” (1948a, p. 51) by falling “in effect back on the
assumption that everybody knows everything” (ibid.), an assumption that answers
the question of how markets operate in a purely tautological manner but does not
inform us about how they actually function.72 Implicitly, his critique is, however,
also directed at Mises’ praxeological a priorism, as can be concluded from Hayek’s
comments on a “Pure Logic of Choice” which is concerned with what “we know
to be common to all human thought” (Hayek, 1948a, p. 47). The “axioms” of the
“pure logic of choice,” Hayek notes, define the kind of phenomena which we seek
to understand, namely “conscious” as distinguished from “instinctive” action, and
in this sense they are, of course, “universally applicable to the field in which we
are interested” (ibid.). Yet, so Hayek states in no uncertain terms, the “tautological
transformations of the Pure Logic of Choice” (ibid., p. 39) have little to do with
an explanation of social reality. As he puts it: “But the assumptions or hypotheses,
which we have to introduce when we want to explain the social process, concern
the relation of the thought of an individual to the outside world, the question to
what extent and how his knowledge corresponds to the external facts. And the
hypotheses must necessarily run in terms of assertions about causal connections,
about how experience creates knowledge” (ibid.).
Hayek emphasizes, not any less than other Austrians do, the subjectivity of
the valuations and theories that inform human action.73 He explicitly stresses
that “the analysis of what people will do can start only from what is known to
them” (Hayek, 1948a, p. 44),74 and that “not only man’s action toward external
objects but also the relations between men and all the social institutions can
be understood only by what men think about them” (Hayek, 1979, p. 57). Yet,
Hayek’s methodological conclusions from the subjectivist premise are clearly
different from those drawn by authors like Mises or Lachmann. This is not to
deny the often observed fact that some of Hayek’s arguments on the matter – in
particular in his essay on “Scientism and the Study of Society” (Hayek, 1979,
pp. 17ff) – appear to come close to Mises’ views.75 Yet, as has been noted above
(p. 13), a careful reading of these arguments, especially a reading that gives
appropriate weight to Hayek’s later clarifying comments, cannot fail to recognize
that Hayek’s version of subjectivism systematically differs in its methodological
orientation from Mises’ as well as from Lachmann’s subjectivist outlook. By
contrast to them his ambition clearly is to develop an empirical, explanatory
theory of the role of knowledge or mental phenomena in human action and
social interaction.
As far as the method of understanding is concerned on which e.g. Lachmann
puts so much emphasis as the distinctive method of the social sciences, Hayek
acknowledges that we surely “constantly act on the assumption that we can in

this way interpret other people’s actions on the analogy of our own mind” (1948b,
p. 64). Yet, he cautions that such analogies are mere conjectures, “that we can never
be sure” (ibid.), and “that we can understand less and less as we turn to beings
more and more different from ourselves” (ibid., p. 66).76 More importantly even,
Hayek’s own understanding of what “understanding” is about is not in contrast to
the notion of nomological explanation but is, instead, perfectly compatible with it
(Hayek, 1967c, pp. 58ff).
Hayek’s intention in emphasizing the role of subjective knowledge in human
action is, quite apparently, not to claim a special methodological status – whether
praxeological or hermeneutical – for economic and social analysis. To the contrary,
his concern is with “economics as an empirical science” (1948a, p. 44), it is with an
empirical, explanatory approach to “the role which assumptions and propositions
about the knowledge possessed by different members in society play in economic
analysis” (ibid., p. 33). And such an empirical approach, he insists, must be based
on conjectures about how knowledge is acquired and expectations are formed
(Hayek, 1979, p. 57), on “empirical propositions . . . about how people will learn”
(1948a, p. 55). These conjectures, Hayek emphasizes, are empirically testable
“assumptions about causation” (ibid., p. 46) and, as such, are “of a fundamentally
different nature” not only from the propositions of formal equilibrium analysis
(ibid., p. 55), they are also “rather different from the general assumptions from
which the Pure Logic of Choice starts” (ibid., p. 46).
What is, in the present context, even more important than Hayek’s programmatic
methodological arguments is the fact that, from his 1936 lecture on “Economics and
Knowledge” (1948a) on, Hayek systematically pursued an empirical subjectivist
research agenda that focuses on the issue of how imperfect human agents – instead
of fictitious, perfectly rational homines oeconomici – acquire and communicate
the subjective knowledge of the world that allows them relatively successfully to
operate within their respective problem environments, and relatively successfully
to cooperate with, and coordinate their actions with, other agents who are equally
limited in their understanding of the complexities of the world around them.
Hayek’s entire life work can be viewed as a systematic effort to explore different
aspects of this fundamental issue. His theory of the spontaneous order of the market
seeks to explain how the price mechanism operates as a system of communication
by which the fragments of knowledge dispersed among myriads of agents in
an extended network of exchange relations are utilized to effectively transmit
throughout the system information about changes in relative scarcities. His theory
of competition as a discovery procedure seeks to explain how the market functions
as an open experimental arena in which, by way of trial and error, continuously
new and superior solutions to economic problems are discovered (solutions which
may in turn, of course, generate new problems of their own). His theory of
cultural evolution seeks to explain how the inter-personal and inter-group variety
in acquired practices, artifacts, normative beliefs and conjectural knowledge

feeds into a process of competitive selection that results in an inter-generational
accumulation of experience-based problem-solving knowledge. Finally, and of
particular interest in the present context, there is Hayek’s theory on how the human
mind operates as the “repository” of subjective knowledge, an empiricist theory of
mental phenomena that is outlined in detail in his The Sensory Order – An Inquiry
into the Foundations of Theoretical Psychology (1952) as well as in some of his
other contributions devoted to epistemological issues (1967a, b, c, d; 1978b; 1979,
pp. 31ff).
About the theoretical perspective laid out in The Sensory Order W. B. Weimer
(1982, p. 281) has said that it “is compatible with . . . the resurgence of cognitive
psychology (post-Chomsky and psycholinguistics), as well as the methodological
views of science advanced by Karl Popper.” Even though it has found much
less attention than other parts of Hayek’s work, the significance of The Sensory
Order in the context of his overall research program has been well recognized
by Hayek-experts such as Bruce Caldwell (2000, p. 9), while experts in modern
cognitive neuroscience have praised Hayek’s contribution as anticipating, in
essence, ideas that have only more recently gained wider recognition in this
field.77
In The Sensory Order Hayek took up and expanded ideas that he had outlined
several decades earlier as a young student in a manuscript, entitled Beiträge zur
Theorie der Entwicklung des Bewusstseins (Contributions to a Theory of How
Consciousness Develops) (Hayek, 1920). In the abstract of this early paper he
had explicitly stated that it was his ambition to propose a theory of the mind
compatible with the world view of the natural sciences.78 The term “sensory
order” is used by Hayek (1982, p. 288) to refer to an organism’s internal (mental)
representations or “models” of the outer world, models that embody knowledge
of relevant contingencies in the organism’s problem environment.
Of particular significance in the present context is Hayek’s claim that these
internal representations or models provide the clue to our understanding of adaptive

or purposive behavior (Hayek, 1952, pp. 122ff). The fact that these internal models
allow the organism to anticipate the likely consequences from different kinds of
behavior is, in his view, the essential ingredient of purposive, problem-solving
behavior. In this sense, Hayek notes, even machines governed by programs – such
as, for example, “automatic pilots for aircraft” (ibid., p. 126) – may be said to
exhibit “purposive” (problem-solving) behavior. Though such machine-behavior
is, of course, very primitive compared to the complexities of human behavior, it has
in common with the latter that it owes its problem-solving capacity to the existence
of programs in which knowledge of relevant contingencies of the environment is

incorporated.79
In Hayek’s account it is through the mind’s “internal representations” – through

models, rules or dispositions80 – that all human perception as well as human action
is guided, from our pre- or sub-conscious adaptations to our most deliberate and
reflected responses to problems (Hayek, 1952, pp. 86, 145; 1967c, p. 45).81 It is,
he argues, only on the basis of the knowledge already incorporated in such internal
representations that we are able, on the one hand, to single out in our perceptions
those aspects of “the inexhaustible totality of everything” (Hayek, 1979, p. 121)
that are of relevance to us and, on the other hand, to select the appropriate, problem-
solving course of action from an open-ended set of potential responses (Hayek,
1978b, p. 38).82 As Hayek (1973, p. 30) puts it: “We never act, and could never act,
in full consideration of all the facts of a particular situation, but always by singling
out as relevant some aspects of it.” The “singling out,” in our perception as well
as in our action, is guided by “schemata of thought” (ibid., p. 31), by conjectural
expectations or “abstractions”83 that provide the “basis of man’s capacity to move
successfully in a world very imperfectly known to him” (ibid., p. 30).84
Hayek (1967b, p. 23) explicitly agrees with K. R. Popper’s argument that
“observation is always observation in the light of theories,”85 and like Popper
he interprets “theory” in the broadest possible sense, to include the most basic
genetically coded behavioral dispositions as well as the most reflected scientific
conjectures, an interpretation captured in Popper’s statement, “from the amoeba to
Einstein is just one step” (Popper, 1972, p. 246).86 And just like Popper he views
the growth of knowledge in all forms as a process of trial and error-elimination, as
experience-based “modification of previous knowledge” (Popper, 1972, p. 71).
As Hayek (1952, p. 143) phrases it, “all we know about the world is of the
nature of theories and all “experience” can do is to change these theories.”87
More specifically, Hayek interprets the process through which mental models,
rules or dispositions become better adapted to the relevant problem environment
as a process of classification and reclassification that is controlled by success and
failure (Hayek, 1952, p. 147).88 In case of “expectations which will not be borne out
by events” a reclassification or a “change of the frame of reference” is induced,

controlled by, as Hayek puts it, “the pragmatic needs of the individual and the
species” (ibid., pp. 145, 168).89
In Hayek’s account, the evolution of the mental order proceeds as a continuous
reorganization of the classificatory apparatus in light of which external events
are interpreted, at the level of biological evolution as well as at the level of
behavioral learning and at the level of conceptual thinking (Hayek, 1952, p. 107;
1967c, p. 52). At all levels, rules and dispositions better adapted to the actual
contingencies in the world are a fruit not of pre-adapted “foresight” but of tentative
conjectures and ex-post selection by consequences.90 While the “knowledge” that

has been accumulated over the evolutionary history of our species is incorporated,
as genetically coded conjectures, in our sense (and other) organs, the capability
of learning allows an organism to accumulate experience-based problem-solving
knowledge over its lifetime that is incorporated in memory-coded models, rules
or dispositions (Hayek, 1952, pp. 53, 106, 108, 129ff, 166; 1967c, p. 51). At the
level of “conceptual thinking,” in particular in science, knowledge grows by way of
deliberate conjectural reclassifications that progressively replace “the classification
of events which our senses provide” (Hayek, 1979, p. 31), in order to account for
problems we encounter in the world of our experience, “problems which can be
answered only by altering the picture which our senses give us of that world”
(Hayek, 1952, p. 173).
While Hayek’s ambition is to provide a naturalistic, explanatory account of
mental phenomena, he emphasizes that, due to the complexity of the processes
involved, “we shall never be able to achieve more than an explanation of the
principle on which the mind operates, and shall never succeed in fully explaining
any particular mental act” (Hayek, 1952, p. 34). In the impossibility of a “full
explanation” (ibid., p. 439) of particular acts he sees, however, no reason at all to
discard the standard notion of causal explanation. We can, as he argues, hold the
view that every single action of a human being is caused by “the inherited structure
of his body (particularly of its nervous system) and of all the external influences
which have acted upon it since birth” (Hayek, 1967b, p. 37) and, at the same time,
insist that, “in spite of our knowledge of the principle on which the human mind
works” (ibid.), we will in general not be able to explain single actions because it
is impossible for us “to state the full set of particular facts which brought it about
that the individual did a particular thing at a particular time” (ibid.).91
The limits of explanation and prediction in the realm of human action are,
in the sense explained, for Hayek not a matter of fundamental methodological
differences between the natural sciences and the sciences of human action.
They are due to the uniqueness and novelty that characterizes the operation of a
complex rule-governed system such as the human mind, a system that can adapt
to a complex and changing environment through the simultaneous, combined

application of models or rules (Hayek, 1952, p. 130).92 As he puts it: “It is
this determination of particular actions by various combinations of abstract
propensities which makes it possible for a causally determined structure of actions
to produce ever new actions it has never produced before, and therefore to produce
altogether new behavior such as we do not expect from what we usually describe
as a mechanism. Even a relatively limited repertory of abstract rules that can
thus be combined into particular actions will be capable of “creating” an almost
infinite variety of particular actions” (Hayek, 1978b, p. 48).
7. CONCLUSION
My purpose in writing this paper has been to take issue with a tradition in Austrian
economics that claims that recognizing the subjectivist determinants of human
action requires economics (as well as the other social sciences) to adopt a different
methodology than the causal explanatory approach of the natural sciences. In the
preceding sections I have sought to counter this claim in a number of ways. I
have examined the arguments that have been advanced in defense of this claim,
and I have given reasons why these arguments must be doubted (Section 3). I
have discussed the research program of evolutionary psychology as an example of
theoretical developments in modern psychology and cognitive science that seek
to account, in an explicitly naturalistic manner, for the cognitive foundations of
human action (Section 4), and I have cited E. Mayr’s and K. R. Popper’s concepts
of program- or conjecture-based behavior as theoretical paradigms that show how
human purposefulness can be analyzed within a causal explanatory framework
(Section 5). Finally, I have argued that an outlook at subjectivism that is perfectly
compatible with these naturalistic approaches can, indeed, be found within the
Austrian tradition itself, namely in F. A. Hayek’s theory of rule-based perception
and action (Section 6).
To the extent that modern Austrian economics, explicitly or implicitly, adheres
to the methodological dualism advocated by authors like Mises and Lachmann,
the arguments developed in this paper call for a reorientation. They are meant
to show that the subjectivist research program of the Austrian tradition need
not be associated at all with an anti-naturalistic methodology, but can well be
pursued within a causal explanatory framework. They are an invitation to turn “the
Austrian school . . . into an evolutionary school” (Witt, 2003, p. 26) or, to borrow
a phrase from J. H. Barkow (1992, p. 635), to practice Austrian economics as “an
evolutionarily informed and psychology-compatible social science.”
NOTES
1. With reference to Carl Menger, F. A. Hayek speaks of “that methodological

individualism which is closely connected with the subjectivism of the social sciences”
(Hayek, 1979, p. 64).
2. The principle of methodological individualism has been the subject of extensive debate
in the social sciences. For a detailed examination of that debate see Vanberg (1975).
3. U. Witt (1989, p. 410) “There (in neoclassical economics, V. V.) the subjective
imponderabilities . . . are simply assumed away: Perfect information or, more recently,
rational expectation models hypothesize a one-to-one relationship between “objective”
conditions and the individual agent’s perception of these.” –L. Lachmann (1977, p. 158)
“From our point of view, the crucial significance of the emergence of expectations as a
problem rests in the fact that . . . they have thus far proved refractory of all attempts to
incorporate them into the formal apparatus of the late classical economics of our time.”
4. This is, of course, also the central message of H. A. Simon’s critique of the neoclassical
concept of rationality. As Simon (1997, p. 25) puts it: “Neoclassical theory, put in the
simplest terms, proceeds as though the facts of the real world are known . . . Now, if
you think people can deal with the world as it really is, so that you do not have to
worry about their subjective view of it, then in order to predict their behavior you simply
calculate what would maximize utility in that real world . . . If you believe, however, that
our minds are very limited, that we can only form a very approximate picture of the world,
then . . . you need a theory of how we make our decisions: you need a theory of procedural
rationality.”
5. U. Witt (1989, p. 417; 1992, p. 221) has argued in a similar direction, pointing
out that Hayek’s “different attitude towards the subjectivism problem” is much closer to
Carl Menger’s original, “psychologically inspired, empirically oriented” approach than to
Misesian tradition in Austrian economics.
6. On the two forms of subjectivism G. P. O’Driscoll and M. J. Rizzo (1985, p. 22) note:
“The first form is most closely related to the traditional subjective theory of value and we
shall call it ‘static subjectivism.’ In this case, the mind is viewed as a passive filter . . . The
second form, on the other hand, views the mind as an active, creative entity . . . This form
of subjectivism we shall call ‘dynamic subjectivism.’ ”
7. About L. Robbins’ (1935) influential attempt to infuse Austrian subjectivism into
mainstream economics I. Kirzner (1992b, p. 126) notes: The “subjectivism which Robbins
brought to synthesis from Vienna was severely limited to the ‘static’ aspects of subjectivism.”
– On this issue see also J. Wiseman (1985).
8. As he notes, Mises adopts “the rather heavy term ‘sciences of human action’ ” (1978,
p. 9) only reluctantly for want of a better term: “The German language has developed a term
that would have been expedient to denote the totality of the sciences dealing with human
action as distinguished from the natural sciences, viz., the term Geisteswissenschaften.”
9. As Mises emphasizes, to say that human action is purposive is to say that it is
goal-directed, “aiming at a definite end and guided by ideas concerning the suitability
or unsuitability of definite means” (Mises, 1978, p. 34), it is not to say that it is “objectively
rational” as judged by an observing third party. Rejecting common interpretations of the
rationality postulate in economics he notes that “economics does not deal with an imaginary
homo oeconomicus . . . but with homo agens as he really is, often weak, stupid, inconsiderate,
and badly instructed” (1990, p. 24). – I have discussed Mises’ views on human action and
rationality in more detail in Vanberg (2004).
10. Mises (1990, p. 21) “Not what a man chooses, but that he chooses counts for
praxeology” (ibid., p. 21). – What praxeology asserts about “action as such” is, as
Mises (1978, p. 44) argues, meant to be “strictly valid without any exception for every
action.”
11. Mises (1990, p. 21) “Praxeology is not based on psychology and is not a part of
psychology . . . Psychology deals with the internal processes determining the various choices
in their concreteness.”
12. Mises prefers, indeed, to use the terms “thymology” and “thymological” instead
of “psychology” and “psychological” in order to separate the kind of psychology that he
favored – namely a psychology that is concerned with “the mental activities of men that
determine their actions” (Mises, 1978, p. 47) – from what he considered a too narrowly
positivist “experimental psychology” (ibid., p. 47).
13. Mises (1978, p. 50) “Understanding . . . refers not to the field of praxeology and
economics, but to the field of history. It is a thymological category.”
14. Mises (1978, p. 70) “If one accepts the terminology of logical positivism and
especially also that of Popper, a theory or hypothesis is ‘unscientific’ if in principle it
cannot be refuted by experience. Consequently, all a priori theories, including mathematics
and praxeology, are ‘unscientific.’ This is merely verbal quibble.”
15. M. Rothbard (1990, p. 317) has summarized Mises’ apriorism as follows: “He
assumes nothing about the wisdom of man’s ends or about the correctness of his means.
He ‘assumes’ only that men act, i.e. that they have some ends, and use some means to
try to attain them. This is Mises’ Fundamental Axiom, and it is this axiom that gives the
whole praxeological structure of economic theory built upon it its absolute and apodictic
certainty . . . For this Axiom is true for all human beings, and everywhere, at any time, and
could not even conceivably be violated.”
16. I have critically examined Mises’ arguments in some detail in Vanberg (1975,
pp. 85ff). – See also Vanberg (2004, p. 21).
17. A theory or conjecture that – as Rothbard (1990, p. 317) puts it – “could not even
conceivably be violated” cannot tell us anything about the world of our experience because it
allows for all conceivable states of the world. It does not tell us what is factually impossible
but only what is logically impossible and, therefore, inconceivable.
18. Mises (1978, p. 121) “(H)uman reaction is determined by ideas, a phenomenon the
description of which is beyond the reach of physics, chemistry and physiology. There is no
explanation in terms of the natural sciences.”
19. Mises (1990, p. 25) “But as long as we do not know how external (physical and
physiological) facts produce in the human ‘soul’ definite thoughts and volitions resulting in
concrete acts, we have to face an insurmountable dualism . . . Reason and experience show
us two separate realms: the external world of physical and physiological events and the
internal world of thought, feeling, and purposeful behavior. No bridge connects – as far as
we can see today – these two spheres.” – See also Mises (1957, p. 1).
20. Mises (1978, p. 121) “(T)he natural sciences have no intellectual tool to deal with
ideas and with finality.”
21. Mises (1978, p. 43) “The natural sciences do not know anything about final causes.
For praxeology finality is the fundamental category.”
22. Mises (1978, p. 69) “There are in this orbit no such things as experimentally
established facts. All experience in this field is . . . experience of complex phenomena.”
23. Lachmann (1977, p. 169) “The concept of ‘Purpose,’ for example, has long been
discarded by the older natural sciences like physics, and has now even been expunged from
biology. Yet, it remains an indispensable tool of the social sciences. Where human action is
concerned, a purely behavioristic approach can answer none of our questions. It certainly
cannot explain, i.e. make intelligible, a single human act.”
24. Lachmann (1977, p. 95) – Lachmann (ibid., p. 35) notes about the relation between
Max Weber and Ludwig von Mises: “Their personal acquaintance was brief, though,
confined to the few summer months Weber spent at the University of Vienna in 1918.
Both men shared an interest in neo-Kantian philosophy and an aversion to the cruder
brands of positivism and behaviorism.” – I have compared Mises’ and Weber’s versions
of subjectivism in some detail in Vanberg (1975, pp. 85ff, 101ff).
25. Lachmann (1977, p. 153) defines Verstehen as the “method which explains human
action in terms of plans, constituted by mental acts and linking an imagined future to an
active present.”
26. Lachmann (1977, p. 153) “The alternative principle of explanation is, of course, that
of ‘response to stimulus.’ ”
27. A somewhat different emphasis is implied when Lachmann (1977, p. 173) notes:
“Not the psychological causes of human action, but their logical consequences form the
subject matter of the analytical social sciences.” – See also ibid., p. 168.
28. With reference to a Verstehen approach that inquires into the subjective meaning
of actions Lavoie (1991, p. 482) notes: “I see no reason to interpret subjectivism in this
way . . . Meaning can be viewed as something that resides ‘within’ the inter-subjective, and
this is taking place out in the open all the time and all around us.”
29. For a critique of this new, hermeneutical branch of the Austrian tradition see
H. Albert (1988) who concludes that “the anti-naturalistic hermeneutics in the spirit of
Heidegger . . . leads to a methodological subjectivism which has nothing in common at all
with the original subjectivism of the Austrian School” (ibid., p. 593). – Albert’s main
objection is that “nothing can be gained . . . by wiping out the important difference between
the problem of the interpretation of words and sentences and the problem of the explanation
of human action by using the term ‘hermeneutical’ in the loose way which we owe to modern
hermeneutical philosophy” (ibid., p. 595).
30. Though in this regard, too, Lavoie (1991, p. 477) argues somewhat ambiguously
when he notes: “The point is not to abandon the study of causation, but to recognize the
fact that many of the relevant causes we are looking for in economics are themselves
meanings.”
31. It is interesting to note that from C. Menger to L. von Mises the “Austrian” perspective
on this issue has changed. Menger (1963, p. 59, fn. 18) argued: “The contrast between the
theoretical natural sciences and the theoretical social sciences is merely a contrast of the
phenomena which they investigate from a theoretical point of view. It is by no means a
contrast of method.” – Mises (1978, p. 45) voices the opposite view when he states: “What
distinguishes the sciences of human action from the natural sciences is not the events
investigated, but the way they are looked upon.”
32. See e.g. Hayek (1979, p. 78) “The attitude which, for want of a better term, we shall
call the ‘objectivism’ of the scientistic approach to the study of man and society, has found
its most characteristic expression in the various attempts to dispense with our subjective
knowledge of the working of the human mind, attempts which in various forms have affected
almost all branches of social study.”
33. Recollecting his early reading of Popper’s original German version (published in
1935) of The Logic of Scientific Discovery Hayek notes that he had been led, “already, to
the understanding of what became Popper’s main systematic point: that the test of empirical
science was that it could be refuted, and that any system which claimed that it was irrefutable
was by definition not scientific” (Hayek, 1994, p. 51). And he continues: “I was not a trained
philosopher; I didn’t elaborate this. It was sufficient for me to have recognized this, but when
I found this thing explicitly argued and justified in Popper, I just accepted the Popperian
philosophy for spelling out what I had always felt” (ibid.).
34. Hayek (1967a, p. 20) “(A)s the advance of the sciences penetrates further and
further into more complex phenomena, theories which merely provide explanations of the
principle . . . may become more the rule than the exception. Certain developments of recent
years, such as cybernetics, the theory of automata or machines, general systems theory, and
perhaps also communication theory, seem to belong to this kind.”
35. Hayek (1967b, p. 31) “Probably the best illustration of a theory of complex
phenomena . . . is the Darwinian theory of evolution by natural selection.”
36. About the Darwinian theory Hayek (1967b, p. 32) notes: “The theory as such, as is
true of all theories, describes merely a range of possibilities. In doing this it excludes other
conceivable courses of events and thus can be falsified. Its empirical content consists in
what it forbids.”
37. Hayek (1967a, p. 16) “It is undoubtedly a drawback to have to work with theories
which can be refuted only by statements of a high degree of complexity.” – Hayek (1967b,
p. 28) “Such a theory will, of course, in Popper’s terms, be one of small empirical
content.”
38. Hayek (1967b, p. 29) “The advance of science will thus have to proceed in two
different directions: while it is certainly desirable to make our theories as falsifiable as
possible, we must also push forward into fields where, as we advance, the degree of
falsifiability necessarily decreases. This is the price we have to pay for an advance into
the field of complex phenomena.”
39. J. O’Driscoll and M. Rizzo (1985, p. 22ff.) appear to take this view when, in their
discussion of the methodological implications of “dynamic subjectivism,” they note that
Hayek’s non-deterministic concept of pattern explanation is “consistent with dynamic
subjectivism” (ibid., p. 27).
40. Citing the views of A. Schütz and F. Machlup, R. Koppl (1994, p. 72) points out that
an “understanding” approach need not be “anti-theoretical” and need not imply a dualism
“according to which quite different epistemological principles apply to mental and physical
phenomena.”
41. When L. Robbins (1981, p. 3.) notes that explanations in economics “must to some
extent be teleological” he clearly uses the term “teleological” in this sense. He does not want
to deny at all that explanations in economics are subject to “the usual logical requirements
of a science” (ibid., p. 9). He only wants to emphasize “that explanations of economic
relationships must involve considerations of purposes” (ibid., p. 3).
42. H. Simon (1997, p. 79) “About thirty years ago there occurred in psychology a
‘cognitive revolution,’ which resuscitated older methodologies – rejected by Behaviorism
– for studying complex human thinking, problem solving and decision making; and which
introduced powerful new methodologies.” – H. Simon sees in the cognitive revolution the
prospects for a theory of the mind on which a “subjectivist” economics can be based. As he
argues: “Unless we have a theory of how the human mind operates, we have few grounds
on which to build an economic theory that will talk about the kind of uncertain world we
live in . . . I believe that we have made great progress toward building a theory of human
thinking in the last 30 years, and therefore I am optimistic about the opportunities to apply
it to economics” (ibid., p. 26).
43. Modern cognitive psychology explicitly seeks to account for internal, dispositional
variables that intervene between “external events” and behavioral responses. They are,
therefore, hardly a just target of Mises’ (1978, p. 37) comment: “The same external events
produce in different men and in the same men at different times different reactions. The
natural sciences are helpless in face of this ‘irregularity.’ ”
44. L. Cosmides and J. Tooby (1994c, p. 42) “Indeed, a theoretical synthesis between
the two fields seems inevitable, because evolutionary biologists investigate . . . the set
of adaptive information-processing problems the brain evolved to solve, and cognitive

scientists investigate the design of the circuits or mechanisms that evolved to solve them.”
45. Tooby and Cosmisedes (1992, p. 64) “It is, therefore, meaningful to ask what kind
of cognitive design features would have constituted good solutions to adaptive problems
that persisted over many generations. Evolutionary biology and hunter-gatherer studies
supply definitions of the recurrent adaptive problems humans faced during their evolution,
and cognitive psychology describes the information-processing mechanisms that evolved
to solve them.”
46. H. Albert (1988, p. 592) points to the discrepancy between Mises’ evolutionary
arguments and his methodological dualism when he notes: “Mises himself refers to the fact
that the human mind has acquired its structure in the course of evolution . . . This implies
that there are general laws of its functioning which can in principle be discovered.”
47. Cosmides and Tooby (1994b, p. 97) “Cognitive programs that systematically violate
this constraint cannot be selected for . . . Evolvability constraints . . . specify the class of
mechanisms that can, in principle, evolve.”
48. Cosmides and Tooby (1994a, p. 329) “Natural selection could equip humans’
cognitive specializations with design features and problem-solving strategies that exploited
the presence of these problem-specific regularities to solve particular classes of problems
in efficient ways appropriate only to this class. . . . For the problem domains they are
designed to operate on, specialized problem-solving methods perform in a manner better
than rational.”
49. Cosmides and Tooby (1994b, p. 88) “(H)umans are equipped with a diverse range
of adaptations designed to perform a wide variety of tasks, from solicitation of assistance
from one’s parents, to language acquisition, to modeling the spatial distribution of local
objects, to coalition formation and cooperation, to the deduction of intentions on the basis
of facial expression.” – See also Cosmides and Tooby (1992, p. 166).
50. It has to be noted that Chomsky’s theory is itself subject to the kind of objections
that are raised against “modularity theories” in general, including evolutionary psychology.
See e.g. Tomasello (1999, p. 94, 203).
51. It is worth noting that some of F. A. Hayek’s arguments on the issue of “Verstehen”
or “understanding” are very much compatible with the outlook of evolutionary psychology.
See Hayek (1967c, p. 46ff.).
52. Cosmides and Tooby (1987, p. 282) “Behavior is . . . elicited by information, which is
gleaned from the organism’s external environment, and, proprioceptively, from its internal
states. . . . The evolutionary function of the human brain is to process information in ways
that lead to adaptive behavior.” – Cosmides and Tooby (1994a, p. 328): “The brain
is a complex computational device, a system that takes sensory information as input,
transforms it in various ways, stores it, analyzes it, integrates it, applies decision rules
to it, and then translates the output of those rules into the muscular contractions that we
call ‘behavior.’ ”
53. The term “program” is explicitly employed in analogy to its use in information
science. As Tooby and Cosmisedes (1992, p. 66) note: “(A)n information-processing
program, whether in an organism or in a computer, is a set of invariant relationships between
informational inputs and ‘behavioral’ outputs.”
54. Tooby and Cosmisedes (1992, p. 103) “The inexhaustible range of possibilities
latent in behavior, categorization, interpretation, decision and so on, is not just an
abstract philosophical point. It is an implacable reality facing every problem-solving
computational system. . . . Any design for an organism that cannot generate appropriate
decisions, inferences, or perceptions . . . is lost in an ocean of erroneous possibilities.”
55. It could be just as well be directed against “hermeneutical Austrians” when Tooby
and Cosmisedes (1992, p. 22) note in reference to an other advocate of “hermeneutics”: “For
example, Clifford Geertz advocates abandoning the ground of principled causal analysis
entirely in favor of treating social phenomena as ‘texts’ to be interpreted just as one might
interpret literature.”
56. I have discussed Mayr’s approach in more detail in Vanberg (2002, p. 15ff.).
57. E. Mayr (1961, p. 1504) “We biologists have long felt that it is ambiguous to designate
such programmed, goal-directed behavior ‘teleological,’ because the word teleological has
also been used in a very different sense.”
58. Such instructions or decision-rules can, of course, reach any degree of complexity
by additional specifications added to the if-clause and the then-clause.
59. Mayr (1961, p. 1502) “It is characteristic of these genetic codes that the programming
is only in part rigid. Such phenomena as learning, memory, non-genetic structural
modification, and regeneration show how ‘open’ these programs are.”
60. Adopting the distinction between proximate and ultimate causes of behavior
Cosmides and Tooby (1987, p. 281f.) note: “(T)he psychology of an organism consists
of the total set of proximate mechanisms that cause behavior . . . (B)ehavior is an effect
produced by a causal system: proximately by psychological mechanisms.”
61. This is, as explained above, the explicit aim of evolutionary psychology. As Cosmides
and Tooby (1987, p. 283) put it: “Evolutionary psychology . . . relates explanations in terms
of adaptive strategy to explanations in terms of proximate mechanisms . . . Psychological
mechanisms constitute the missing causal link between evolutionary theory and
behavior . . . (I)t is the proximate mechanisms that cause behavior that promise to reveal
the level of underlying order for a science of human behavior.”
62. Mayr (1961, p. 1506) “Causality in biology is not in real conflict with the causality
of classical mechanics.”
63. I have discussed Popper’s arguments in more detail in Vanberg (2002, p. 19ff.).
64. K. R. Popper (1972, p. 145) “The tentative solutions which animals and plants
incorporate into their anatomy and their behavior are biological analogues of theories and
vice versa: theories correspond . . . to endosomatic organs and their ways of functioning.
Just like theories, organs and their functions are tentative adaptations to the world we live
in.”
65. The “conjectures” built into our (sense-)organs can, Popper (1972, p. 146) notes, be
regarded as “the biological predecessors of linguistically formulated theories.”
66. Popper does, in fact, explicitly – and approvingly – refer to Mayr’s concept of
program-based behavior (Popper, 1987, p. 151).

67. The name “evolutionary epistemology” appears to have been coined by Donald T.
Campbell (1974).
68. Surprisingly, even though the research program of evolutionary epistemology
predates the emergence of evolutionary psychology, authors who publish in the latter field
seem to have taken no notice of its existence.
69. M. Tomasello (1999, p. 216) refers to these three levels of evolutionary adaptation
when he notes: “Modern adult cognition of the human kind is the product of genetic
events taking place over many millions of years in evolutionary time but also of
cultural events taking place over many tens of thousands of years in historical time
and personal events taking place over many tens of thousands of hours in ontogenetic
time.”
70. In his survey of the field W. W. Bartley (1987, p. 20f.) lists F. A. Hayek among its
“founders,” besides K. R. Popper, D. T. Campbell, Ernst Mayr and Konrad Lorenz.
71. As Hayek has later stated it in an other context: “But the concrete knowledge which
guides the actions of any group of people never exists as a consistent and coherent body. It
only exists in the dispersed, incomplete, and inconsistent form in which it appears in many
individual minds, and the dispersion and imperfection of all knowledge are the two basic
facts from which the social sciences have to start” (1979, p. 49f.)
72. In his later seminal article on “The Use of Knowledge in Society” Hayek noted about
the prominence of “equilibrium analysis” in economics: “But I fear that our theoretical habits
of approaching the problem with the assumption of more or less perfect knowledge on the
part of almost everyone has made us somewhat blind to the true function of the price mecha-
nism and led us to apply rather misleading standards in judging its efficiency” (1948c, p. 87).
73. Kirzner (1992b, p. 132) recognizes Hayek’s concern with the role of knowledge as
“a decisive, explicit extension of subjectivism in modern economics.”
74. Hayek specifically notes that in our efforts to explain human action we must remain
aware of the difference between what is known by “the observing economist” (1948a, p. 39)
and what is “known to the persons whose behavior we try to explain” (ibid.). And he points
to the obvious – though in standard rational choice accounts often ignored – fact that “no
superior knowledge the observer may possess about the object, but which is not possessed
by the acting person, can help in understanding the motives of their actions” (1948b, p. 60).
– On this issue see also Hayek (1979, p. 51).
75. In fact, in a footnote to his often quoted remark “that every important advance in
economic theory during the last hundred years was a further step in the consistent application
of subjectivism” Hayek (1979, p. 52) states: “This is a development which has probably
been carried out most consistently by Ludwig von Mises.”
76. For a comparison to Mises’ view on the role of “understanding” in the social sciences
see Mises (1978, p. 50, 71; 1990, p. 8f.).
77. J. M. Fuster (1995, p. 87ff.) “The first proponent of cortical memory networks on a
major scale was neither a neuroscientist nor a computer scientist but, curiously, a Viennese
economist: Friedrich von Hayek . . . Although devoid of mathematical elaboration, Hayek’s
model clearly contains most of the elements of those later network models of associative
memory . . . The main reason for dwelling here on Hayek’s model is simply that it has
certain properties . . . that conform exceptionally well to recent neurobiological evidence on
memory.” – Fuster (ibid., p. 89) points to G. Edelman as “one of the theoreticians of the
brain to have acknowledged the importance of Hayek’s contribution” (Fuster ibid., p. 90).
See G. Edelman (1987, p. 25).

78. In retrospect, Hayek (1982, p. 289) has noted about the writing of The Sensory
Order: “When I then, about 1946, began looking at the current psychological literature, I
found to my amazement that my problem seemed to be in exactly the same state in which
I had left it 25 years before.” – Hayek’s principal reference at the time was D. O. Hebb’s
(1949) neuropsychological theory which has been of major influence on the development
of modern cognitive neuroscience. – It is worth noting that J. H. Holland (whose theory
of complex adaptive systems is very much compatible with Hayek’s views in The Sensory
Order) cites Hebb’s theory as the major inspiration for his own work (Holland, 1998, p. 19).
– On Holland’s theory see Vanberg (2004, p. 30ff.).
79. Hayek (1952, p. 126) “Such machines . . . would lack the capacity of learning from
experience. But although for this reason such machines cannot be described as brains, with
regard to purposiveness they differ from a brain merely in degree and not in kind.”
80. Hayek uses the terms “models,” “rules” and “dispositions” alternatively to describe
the mental events that take place “between the input of (external and internal) stimuli and
the output of action” (Hayek, 1982, p. 288). – While in The Sensory Order he mostly speaks
of “models,” in later publications he prefers to speak of “rules of action (or dispositions)”
(Hayek, 1978b, p. 43). As he notes: “(D)ispositions toward kinds of movements can be
regarded as adaptations to typical features of the environment, and the “recognition” of
such features as the activiation of the kind of disposition adapted to them . . . (A)ll the
“knowledge” of the external world which such an organism possesses consists in the action
patterns which the stimuli tend to evoke . . . (W)hat we call knowledge is primarily a system
of rules of action” (ibid., p. 41).
81. On the “connection between the rules governing perception and the rules governing
action” (1967c, p. 56) Hayek notes: “(T)he perception of events can also be regarded as a
subsummation of particular stimuli, or groups of stimuli, as elements of an abstract class to
which a response possessing certain characteristics is appropriate” (1978b, p. 40).
82. Cosmides and Tooby (1987, p. 297) refer to the same issue when they note: “There
are an infinite number of dimensions that could be used to cave the environment into
categories; there is no assurance that a general-purpose information processing system
would ever . . . guide such a system toward the appropriate dimensions.”
83. Hayek (1952, p. 143) “(A)ll sensory perception is therefore in a sense “abstract,” it
always selects certain features or aspects of a given situation . . . Even the so-called elemen-
tary sensory qualities are in this sense ‘abstractions’.” – See also Hayek (1978b, p. 44).
84. Hayek (1952, p. 142) “Perception is thus always an interpretation, the placing of
something into one or several classes of objects.” – On “the general point that all perception
involves a theory or hypothesis” see also Hayek (1967b, p. 23f.; 1967c, p. 53; 1978b, p. 37f.;
1979, p. 119).
85. Hayek (1978b, p. 43) emphasizes the correspondence between his notion of “the
primacy of the rules of action (or dispositions)” and Popper’s argument “that the capacity to
generalize comes first and the hypotheses are then tested and confirmed or refuted according
to their effectiveness as guides to action.”
86. Hayek (1978b, p. 46) “(T)he capacity for abstraction manifests itself already in
the actions of organisms to which we surely have no reason to attribute anything like
consciousness.” – See also Hayek (1952, p. 108).
87. In a footnote Hayek (1952, p. 143) adds: “I owe this way of putting it to my friend
K. R. Popper, who, however, may not entirely agree with this use I am making of his
ideas.”
88. About the general outlook he adopted in The Sensory Order Hayek has noted in
retrospect that he was led “to interpret the central nervous system as an apparatus of multiple
classification or, better, as a process of continuous and simultaneous classification and
constant reclassification on many levels (of the legion of impulses proceeding in it at any
moment), applied in the first instance to all sensory perception but in principle to all kinds
of mental entities, such as emotions, concepts, images, drives, etc., that we find to occur in
the mental universe” (1982, p. 289).
89. Hayek (1979, p. 36, fn. 8) “The classification of the stimuli in our central nervous
system is probably highly ‘pragmatic’ in the sense that it . . . stresses those relationships
between the external world . . . and our body which in the course of evolution have proved
significant for the survival of the species.”
90. Hayek (1978b, p. 42) “It seems to me that the organism first develops new
potentialities for actions and that only afterwards does experience select and confirm those
which are useful as adaptations to typical characteristics of its environment. There will thus
be gradually developed by natural selection a repertory of action types adapted to standard
features of the environment.”
91. Hayek (1967a, b, c, d, p. 37) “The individual personality would remain for us as much
a unique and unaccountable phenomenon . . . whose specific actions we could generally not
predict or control, because we could not obtain the information on all the particular facts
which determined it.”
92. Hayek’s essential argument – namely that an apparatus that builds models by
combining more basic rules “is of much greater efficiency than could be any . . . apparatus
which contained, as it were, a few fixed models of typical situations” (1952, p. 131) is very
similar to the central argument in J. H. Holland’s theory of complex adaptive agents, namely
that such agents use “building blocks to generate internal models” of the ever new problem
situations they encounter (Vanberg, 2004, p. 31).
ACKNOWLEDGMENT
An anonymous referee’s helpful comments are gratefully acknowledged.
REFERENCES
Albert, H. (1988). Hermeneutics and economics – A criticism of hermeneutical thinking in the social
sciences. Kyklos, 41, 573–602.
Badcock, C. (2000). Evolutionary psychology – A critical introduction. Oxford: Blackwell.
Barkow, J. H. (1992). Beneath new culture is old psychology: Gossip and social stratification. In: J. H.
Barkow, L. Cosmides & J. Tooby (Eds), 627–637.
Bartley, W. W. (1987). Philosophy of biology versus philosophy of physics. In: G. Radnitzky & W.
W. Bartley (Eds), Evolutionary Epistemology, Rationality, and the Sociology of Knowledge
(pp. 7–45). La Salle, IL: Open Court.
Boehm, S. (1982). The ambiguous notion of subjectivism: Comment on Lachmann. In: I. Kirzner (Ed.),
Method, Process, and Austrian Economics. Essays in Honor of Ludwig von Mises (pp. 41–52).
Lexington, MA and Toronto: Lexington Books.

Caldwell, B. J. (1991). Comment on Lavoie. In: N. de Marchi & M. Blaug (Eds), Appraising Economic
Theories – Studies in the Methodology of Research Programs (pp. 487–491). Aldershot: Edward
Elgar.
Caldwell, B. J. (2000). The Emergence of Hayek’s ideas on cultural evolution. Review of Austrian
Economics, 13, 5–22.
Campbell, D. T. (1974). Evolutionary epistemology. In: P. A. Schlipp (Ed.), The Philosophy of Karl
Popper (pp. 313–463). La Salle, IL: Open Court.
Chamilall, N. S., & Krecke, E. (2002). Professor Kirzner on Carl Menger: To what extent was Carl
Menger subjectivist? Journal des Economistes et des Etudes Humaine, 12, 299–342.
Chomsky, N. (1980). Rules and representations. New York: Columbia University Press.
Cosmides, L., & Tooby, J. (1987). From evolution to behavior: Evolutionary psychology as the missing
link. In: J. Dupré (Ed.), The Latest on the Best – Essays on Evolution and Optimality (277–306).
Cambridge, MA and London: MIT Press.
Cosmides, L., & Tooby, J. (1992). Cognitive adaptations for social exchange. In: J. H. Barkow, L.
Cosmides & J. Tooby (Eds), 163–228.
Cosmides, L., & Tooby, J. (1994a). Better than rational: Evolutionary psychology and the invisible
hand. American Economic Review, Papers and Proceedings, 84, 327–332.
Cosmides, L., & Tooby, J. (1994b). Origins of domain specificity: The evolution of functional
organization. In: L. A. Hirschfeld & S. A. Gelman (Eds), Mapping the Mind – Domain Specificity
in Cognition and Culture (pp. 85–116). Cambridge and New York: Cambridge University Press.
Cosmides, L., & Tooby, J. (1994c). Beyond intuition and instinct blindness: Toward an evolutionarily
rigorous cognitive science. Cognition, 50, 41–77.
Cosmides, L., Tooby, J., & Barkow, J. H. (1992). Introduction: Evolutionary psychology and conceptual
integration. In: J. H. Barkow, L. Cosmides & J. Tooby (Eds), 3–15.
Dufourt, D., & Garrouste, P. (1992). Criteria of scientificity and methodology of the social
sciences: Menger, Mises and Hayek. In: R. F. Hébert (Ed.), Themes on Economic Discourse,
Method, Money and Trade, Selected Papers from the History of Economics Conference 1991,
Perspectives on the History of Economic Thought (Vol. IX, pp. 18–36). Aldershot: Edward Elgar.
Edelman, G. (1987). Neural Darwinism: The theory of neuronal group selection. New York: Basic
Books.
Fuster, J. M. (1995). Memory in the Cerebral cortex – An empirical approach to neural networks in
the human and nonhuman primate. Cambridge, MA and London, England: MIT Press.
Hayek, F. A. (1920). Beiträge zur Theorie der Entwicklung des Bewusstseins (typescript dated
“September 1920”) Hoover Institution Archives, F. A. von Hayek Collection, Box 93, Folder 1.
Hayek, F. A. (1948a). Economics and knowledge. In: Individualism and Economic Order (pp. 33–56).
Hayek, F. A. (1948b). The facts of the social sciences. In: Individualism and Economic Order
Hayek, F. A. (1948c). The use of knowledge in society. In: Idem, Individualism and Economic Order
Hayek, F. A. (1952). The sensory order – An inquiry into the foundations of theoretical psychology.
Hayek, F. A. (1967a). Degrees of explanation. In: Studies in Philosophy, Politics and Economics
Hayek, F. A. (1967b). The theory of complex phenomena. In: Studies in Philosophy, Politics and
Economics (pp. 22–42). Chicago: University of Chicago Press.
Hayek, F. A. (1967c). Rules, perception and intelligibility. In: Studies in Philosophy, Politics and
Economics (pp. 43–65). Chicago: University of Chicago Press.
Hayek, F. A. (1967d). Kinds of rationalism. In: Studies in Philosophy, Politics and Economics
Hayek, F. A. (1973). Rules and order. Law, legislation and liberty (Vol. 1). London and Henley:
Routledge & Kegan Paul.
Hayek, F. A. (1976). The mirage of social justice. Law, legislation and liberty (Vol. 2). London and
Henley: Routledge & Kegan Paul.
Hayek, F. A. (1978a). The pretence of knowledge. In: New Studies in Philosophy, Politics, Economics
and the History of Ideas (pp. 23–34). Chicago: University of Chicago Press.
Hayek, F. A. (1978b). The primacy of the abstract. In: New Studies in Philosophy, Politics, Economics
and the History of Ideas (pp. 35–49). Chicago: University of Chicago Press.
Hayek, F. A. (1979 [orig. 1952]). The counter-revolution of science. Indianapolis: Liberty Press.
Hayek, F. A. (1982). The sensory order after 25 years. In: W. B. Weimer & D. S. Palermo (Eds),
Cognition and the Symbolic Processes (Vol. 2, pp. 287–293). Hillsdale, NJ: Lawrence
Erbaum.
Hayek, F. A. (1994). Hayek on Hayek – An autobiographical dialogue. In: S. Kresge & L. Wenar
(Eds). London: Routledge.
Hebb, D. O. (1949). The organization of behavior: A neuropsychological theory. New York: Wiley.
Holland, J. H. (1998). Emergence: From chaos to order. Reading, MA: Perseus Books.
Kirzner, I. M. (1978). Foreword to second edition. In: Mises 1978 (pp. v–ix).
Kirzner, I. M. (1992a). Carl Menger and the subjectivist tradition in economics. In: Idem, The Meaning
of Market Process – Essays in the Development of Modern Austrian Economics (pp. 70–85).
London and New York: Routledge.
Kirzner, I. M. (1992b). Ludwig von Mises and Friedrich von Hayek: The modern extension of
Austrian subjectivism. In: Idem, The Meaning of Market Process – Essays in the Development
of Modern Austrian Economics (pp. 119–136). London and New York: Routledge.
Kirzner, I. M. (1995). The subjectivism of Austrian economics. In: G. Meijer (Ed.), New Perspectives
on Austrian Economics (pp. 11–22). London and New York: Routledge.
Koppl, R. G. (1994). Ideal type methodology in economics. In: P. J. Boettke (Ed.), The Elgar
Companion to Austrian Economics (pp. 72–76). Aldershot: Edward Elgar.
Lachmann, L. M. (1977). Capital, expectations, and the market economy – Essays on the theory of the
market economy. Edited with and Introduction by W. E. Grinder. Kansas City: Sheed Andrews
and McMeel.
Lachmann, L. M. (1982). Ludwig von Mises and the extension of subjectivism. In: I. M. Kirzner
(Ed.), Method, Process, and Austrian Economics – Essays in Honor of Ludwig von Mises
(pp. 31–40). Lexington, MA and Toronto: Lexington Books.
Lavoie, D. (1991). The progress of subjectivism. In: N. de Marchi & M. Blaug (Eds), Appraising
Economic Theories – Studies in the Methodology of Research Programs (pp. 470–491).
Aldershot: Edward Elgar.
Lavoie, D. (1994a). Introduction: Expectations and the meaning of institutions. In: Idem (Ed.),
Expectations and the Meaning of Institutions – Essays in Economics by Ludwig Lachmann
(pp. 1–19). London and New York: Routledge.
Lavoie, D. (1994b). The interpretive turn. In: P. J. Boettke (Ed.), The Elgar Companion to Austrian
Economics (pp. 54–62). Aldershot: Edward Elgar.
Mackie, J. L. (1974). The cement of the universe – A study in causation. Oxford: Clarendon Press.
Mayr, E. (1961). Cause and effect in biology. Science, 134, 1501–1506.
Mayr, E. (1992). The idea of teleology. Journal of the History of Ideas, 53, 117–135.
Menger, C. (1963). Problems of economics and sociology. Urbana: University of Illinois Press.
Meyer, W. (2002). Grundlagen des ökonomischen Denkens. Edited by H. Albert & G. Hesse. Tübingen:
Mohr Siebeck.
Mirowski, P. (1989). More heat than light – Economics as social physics, physics as nature’s
economics. Cambridge: Cambridge University Press.
von Mises, L. (1957). Theory and history: An interpretation of social and economic evolution (reprint
1978). New Rochelle, NY: Arlington House.
von Mises, L. (1978) The ultimate foundations of economic science – An essay on method (2nd ed.).
Kansas City: Sheed Andrews and McMeel.
von Mises, L., (1990). Money, method, and the market process. Essays by Ludwig von Mises, Selected
by Margit von Mises, Edited with an Introduction by Richard M. Ebeling. Norwell, MA: Kluwer.
O’Driscoll, G. P., Jr., & Rizzo, M. J. (1985). The economics of time and ignorance. Oxford and New
York: Basil Blackwell.
Popper, K. R. (1972). Objective knowledge – An evolutionary approach. Oxford: Clarendon Press.
Popper, K. R. (1982). The open universe – An argument for indeterminism. London: Routledge.
Popper, K. R. (1987). Natural selection and the emergence of mind. In: G. Radnitzky & W. W. Bartley
(Eds), Evolutionary Epistemology, Rationality, and the Sociology of Knowledge (pp. 139–155).
La Salle, IL: Open Court.
Popper, K. R., & Eccles, J. C. (1990). The self and its brain – An argument for interactionism. London
and New York: Routledge.
Robbins, L. (1935). An essay on the nature and significance of economic science (revised ed.). New
York: Macmillan.
Robbins, L. (1981). Economics and political economy. The American Economic Review (Papers and
Proceedings), 71, 1–10.
Rothbard, M. N. (1990). In defense of “extreme apriorism”. In: S. Littlechild (Ed.), Austrian Economics
(Vol. I, pp. 445–451). Aldershot: Edward Elgar.
Shand, A. H. (1981). Subjectivist economics – The new Austrian school. Oxford: Pica Press.
Simon, H. A. (1997). An empirically based microeconomic. Cambridge: Cambridge University Press.
Streissler, E. W. (1972). To what extent was the Austrian school marginalist? History of Political
Economy, 4 , 426–441.
Tomasello, M. (1999). The cultural origins of human cognition. Cambridge, MA and London: Harvard
University Press.
Tooby, J., & Cosmisedes, L. (1992). The psychological foundations of culture. In: J. H. Barkow, L.
Cosmides & J. Tooby (Eds), 19–136.
Vanberg, V. (1975). Die zwei Soziologien – Individualismus und Kollektivismus in der Sozialtheorie.
Tübingen: J. C. B. Mohr (Paul Siebeck).
Vanberg, V. J. (1994). Cultural evolution, collective learning and constitutional design. In: D. Reisman
(Ed.), Economic Thought and Political Theory (pp. 171–204). Boston, Dordrecht, London:
Kluwer.
Vanberg, V. J. (1998). Menger, Carl (1840–1921). In: P. Newman (Ed.), The New Palgrave Dictionary
of Economics and the Law (Vol. 2, pp. 635–641). London: Macmillan.
Vanberg, V. J. (2002). Rational choice vs. program-based behavior – Alternative theoretical approaches
and their relevance for the study of institutions. Rationality and Society, 14 , 7–54.
Vanberg, V. J. (2004). The rationality postulate in economics: Its ambiguity, its deficiency and its
evolutionary alternative. Journal of Economic Methodology, 11, 19–47.
Vaughn, K. I. (1994). Austrian economics in America – The migration of a tradition. New York:

Vromen, J. J. (2002). Stone age minds and group selection – What difference do they make?
Constitutional Political Economy, 13 , 173–195.
Weimer, W. B. (1982). Hayek’s approach to the problem of complex phenomena: An introduction
to the theoretical psychology of the sensory order. In: W. B. Weimer & D. S. Palermo (Eds),
Cognition and the Symbolic Processes (Vol. 2, pp. 241–285). Hillsdale, NJ: Laurence Erlbaum.
Wiseman, J. (1985). Lionel Robbins, the Austrian school, and the LSE tradition. Research in the
History of Economic Thought and Methodology, 2 , 147–159.
Witt, U. (1989). Subjectivism in economics – A suggested reorientation. In: K. G. Grunert & F. Ölander
(Eds), Understanding Economic Behavior (pp. 409–431). Dordrecht, Boston, London: Kluwer.
Witt, U. (1992). Turning Austrian economics into an evolutionary theory. In: B. J. Caldwell & S.
Boehm (Eds), Austrian Economics: Tensions and New Directions (pp. 215–236). Boston,
Dordrecht, London: Kluwer.

Witt, U. (2003). Evolutionary economics and the extension of evolution to the economy. In: Idem,
The Evolving Economy – Essays on the Evolutionary Approach to Economics (pp. 3–34).
Cheltenham: Edward Elgar.
THE NEW FABLE OF THE BEES:

MULTILEVEL SELECTION, ADAPTIVE
SOCIETIES, AND THE CONCEPT OF
SELF INTEREST
David Sloan Wilson
In 1705 Bernard Mandeville published a humorous allegory in verse portraying

human society as a bee hive in which every individual is motivated by personal
greed but the effect is to make the society hum along as a unit. The following
passage conveys the general tone.
As Sharpers, Parasites, Pimps and Players,
Pick-pockets, Coiners, Quacks, Sooth-Sayers,
And all those, that, in Enmity
With down-right working, cunningly
Convert to their own Use the Labour
Of their good-natur’d heedless Neighbour:
These were called Knaves; but, bar the Name,
The grave Industrious were the Same.
All Trades and Places new some Cheat,
No Calling was without Deceit.
Mandeville’s fable of the bees, along with Adam Smith’s metaphor of the invisible
hand, has long been used to convey the idea that a well-functioning society can
be forged out of individual self-interest.1 This idea has become such a tenet of
modern thought that for many it is an unquestioned axiom and for decades it has
served as the foundation of formal economic theory.

ISSN: 1529-2134/doi:10.1016/S1529-2134(04)07009-7
201
202 DAVID SLOAN WILSON
Since Mandeville wrote his fanciful tale, an authentic scientific theory has
arisen that explains how beehives and other adaptive animal societies evolve. This
theory has much to say about human societies, but it fundamentally challenges
the concept of individual self-interest as we know it. In this essay I will update
the fable of the bees based on modern evolutionary theory. The updated version
retains Mandeville’s emphasis on self-organization, in which an adaptive society
can operate without any centralized intelligence. However, it rejects the concept
of self-interest as an adequate description of either the thoughts or actions that
enable individuals to self-organize into adaptive societies.
WHAT IS MULTILEVEL SELECTION THEORY?

The Fundamental Problem of Social Life and its Potential Solution
Darwin’s theory of natural selection, which explains how individual organisms can
become exquisitely adapted to their environments, does not explain the evolution
of adaptive societies with equal ease. To understand the nature of the problem,
imagine a mutant individual who behaves in a way that increases the survival of
everyone in her society, including herself, to an equal degree. Such a “no-cost
public good” might not appear very feasible (and will soon be amended), but is
useful for illustrative purposes. By increasing the fitness of everyone, the mutant
trait will not increase in frequency within the society (other than by drift, which can
equally cause a decrease in frequency). This example illustrates the elementary fact
that natural selection is based on relative fitness. It’s not enough for a mutant trait
to increase its own survival and reproduction; it must do so more than alternative
traits in the population. The relative nature of fitness makes the evolutionary forces
within a population insensitive to the welfare of the population as a whole.
If providing a public good requires a private cost, which seems reasonable,
then the prognosis for the evolution of adaptive societies becomes even worse.
The public benefits count for nothing, no matter how great, while the private cost,
no matter how small, causes the mutant type to constitute a smaller fraction of the
population. The final outcome is a population devoid of public goods providers,
whose members demonically strive to widen their slice of the pie while remaining
oblivious to the size of the pie. I have called this the fundamental problem of
social life (Wilson, 2002).
Darwin was aware of this problem and proposed a solution. Suppose that a
population consists of not one but many social groups. In this case, our mutant
no-cost public good provider does not increase her relative fitness within her
group, but she does increase the fitness of her group, relative to other groups
The New Fable of the Bees 203
in the total population. The mutant trait can spread through the total population
on the strength of among-group selection, even if it is selectively neutral within
groups. If there is a cost of providing the public good, then the fate of the mutant
trait will depend upon the relative strength of the opposing forces of within- and
among-group selection.
Darwin’s potential solution to the fundamental problem of social life is elegant
and perhaps even obvious in retrospect. After all, if natural selection is based on
relative fitness, it makes sense that group-level adaptation is based on the relative
fitness of groups. The modern version of Darwin’s idea is called multilevel
selection theory (Sober & Wilson, 1998). Very simply, adaptation and natural
selection can potentially occur at any level of the biological hierarchy. Adaptation
at a given level requires a process of natural selection at that level. It is potentially
undermined by lower levels of selection and in turn potentially undermines
higher levels of selection. This theoretical framework has the capacity to explain
the behavior of individuals who demonically work to undermine their groups
(within-group selection), individuals who angelically work on behalf of their
groups (the bright side of among-group selection) and avenging angels who work
on behalf of their groups to destroy other groups (the dark side of among-group
selection). We might not like the dark sides of animal and human nature, but they
exist and require a theory to explain them. Even with this briefest of introductions,
the discerning reader can probably see how multilevel selection theory has the
potential to explain the good, the bad, the beautiful and the ugly.
The Rejection and Revival of Multilevel Selection Theory
In the middle of the 20th century, the likelihood of natural selection above the
level of the individual was rejected with such force that the atmosphere of taboo
still hangs like a stale odor over current discussions. Nevertheless, this categorical
rejection proved to be mistaken, as I and others have documented elsewhere
(Michod, 1999; Sober & Wilson, 1998; Wilson, 1998, 1999). Today multilevel
selection theory is used to study an extraordinary array of topics, from the origin
of life to the nature of religion. I will touch upon a few of the applications to pave
the way for our examination of self interest.
One of the most important applications has been termed the major transitions
of life (Hammerstein, 2003; Maynard-Smith & Szathmary, 1999; Michod, 1999).
Until a few decades ago, evolution was thought to take place entirely by small
mutational change. Now a second evolutionary pathway has been identified, in
which social groups evolve to be so integrated that they become higher-level
organisms in their own right. The single organisms of today, such as you and I,
are the social groups of past ages. Each transition, from groups of organisms to
groups as organisms, requires a shift in the balance between levels of selection,
restricting the within-group processes that undermine the integrity of the group,
making team work the only game in town. For example, imagine a primordial
cell in which the genes are independent agents. Some contribute to the economy
of the cell, producing public goods that can be used by all. Other genes selfishly
replicate, increasing their representation within the cell without contributing to
the common good. The evolution of chromosomes neatly solved this problem, by
binding the genes into a single structure that replicates as a unit. In the absence of
within-cell selection, natural selection becomes concentrated at the between-cell
level. This kind of transition has been proposed to explain the origin of life as
a society of cooperating molecular reactions, the origin of nucleated cells as a
society of bacterial cells, and the origin of multicellular organisms as a society of
single cells.
The social insects, including the bees in their hives that inspired Mandeville,
are among the newest major transitions of life (Camazine et al., 2001; Seeley,
1995; Seeley & Buhrman, 1999). The industrious honeybees can work to increase
the fitness of the colony, relative to other colonies, or to increase their own fitness,
relative to other members of the same colony. The reason that social insect colonies
can justly be called superorganisms is because among-colony selection dominates
within-colony selection. Higher-level selection is strong in part because social
insect colonies are initiated by a small number of individuals (minimally one
queen fertilized by a single male), reducing genetic variation within groups and
increasing it among groups. The theory of kin selection is based on this genetic
relatedness and was originally regarded as an alternative to group selection.
However, kin selection is now interpreted as a special case of group selection in
which high relatedness within groups favors altruism by increasing genetic vari-
ation among groups (Hamilton, 1975). Moreover, many social insect colonies are
initiated by more than one queen who mate with more than one male. Evolutionary
biologists are increasingly turning to social control mechanisms analogous to the
chromosome, rather than genetic relatedness, to explain why social insect colonies
hum along so well as adaptive units. For example, it would genetically “pay” a
honeybee worker to lay her own eggs rather than selflessly raising the offspring
of the queen – were it not for the fact that the eggs would be eaten and the deviant
worker attacked by other workers. These “policing” behaviors, as they are called,
evolved to suppress within-group selection, making between-group selection the
only game in town, just like the evolution of the chromosome. Actually, it is an
exaggeration to say that between-group selection becomes the only game in town.
It is important to stress that none of these policing mechanisms are completely
successful. There are rogue genes that avoid the fellowship of chromosomes
and rogue bees that manage to dodge the police. Within-group selection happens
in these otherwise highly adaptive nonhuman societies, no less than in human
social groups.
Multilevel Selection and Human Evolution
Multilevel selection theory and its many applications in the biological world
provide a panoramic background for the study of human evolution. To the best of
our knowledge, our ancestors lived in nomadic groups of a few dozen to at most
a few hundred individuals. These groups merged and divided into a wide variety
of groupings for specific activities. Just as we participate in family groups, sports
groups, learning groups, fighting groups, decision making groups, and so on, each
organized around a specific purpose, our ancestors participated in family groups,
gathering groups, hunting groups, raiding groups, and so on. Almost everything
was done in a social context; to be alone was to be in grave danger. In every social
context, opportunities existed to increase the fitness of oneself relative to others in
the same group, or to increase the fitness of one’s group relative to other groups.
We evolved the behavioral propensities to capitalize on both options. We also
evolved the propensity to limit the self-serving behaviors of our social partners,
thereby concentrating natural selection at the between-group level. In short, we
have evolved the equivalent of chromosomes in cells and policing in honeybees,
thereby qualifying at least crudely as the newest major transition of life (Boehm,
1999; Sober & Wilson, 1998).
Modern hunter-gatherer societies and indeed most small human groups exhibit
an organization that anthropologist and primatologist Chris Boehm (1993, 1999)
has called reverse dominance. Instead of dominant individuals benefiting at the
expense of subordinants within their groups, the subordinants are capable of
collectively ganging up on would-be dominants. Perhaps this balance of power
was achieved through the use of weapons that made everyone lethal or cognitive
skills that improved the ability to form alliances. In any case, it resulted in a
form of guarded egalitarianism and a quantum jump in the capacity for collective
action. Under the constant gaze of their fellows, eternally vigilant against being
bossed around, our ancestors were largely constrained to behave in ways that
were agreed upon by consensus. The opportunities for widening one’s own slice
of the pie within the group were not entirely eliminated, but they were severely
curtailed. The features that set us apart from all other species, including our
capacities for culture, language, and symbolic thought (all communal activities)
are increasingly being explained in terms of this shift from primarily within-group
to primarily among-group selection.
It might seem that this scenario accords perfectly with the concept of self-
interest, since social control makes it in everyone’s interest to cooperate. This
will soon become the main focus of our attention, but for the moment we need to
appreciate that the mechanisms of social control evolve primarily by among-group
selection. If I reward you for supplying a public good or punish you for failing
to do so, I have provided a second order public good at my own private expense
(e.g. Heckathorn, 1990, 1993). Social control systems that cause groups to
function adaptively evolve because better controlled groups outcompete less well
controlled groups. A good social control system requires time, energy, and risk
on the part of individuals and ultimately requires the provision of public goods
at private expense, as surely as voluntary acts of altruism (Fehr, Fischbacher &
Gachter, 2002; Fehr & Gachter, 2002; Sober & Wilson, 1998).
Why should these speculations about the distant past concern the modern social
theorist? There is a widespread tendency to regard evolution as irrelevant because
our behaviors are guided by culture and learning instead of instinct. This view is as
mistaken and has the same stale odor of taboo as the rejection of group selection.
Learning and the capacity for culture are genetically evolved adaptations that
enable us to adapt to our environments faster than genetic evolution alone. They
are evolutionary processes in their own right, and like the mammalian immune
system, they are elaborately constructed to reach biologically adaptive outcomes.
Cultural evolution is as multilevel as genetic evolution (Boyd & Richerson,
1985; Richerson & Boyd, in press). A culturally acquired trait can spread on the
strength of within- or between-group selection, just like a genetically acquired
trait. Indeed, culture is largely responsible for the human major transition,
since acquired traits are much more likely to become uniform within groups
and vary among groups than genetic traits. Human history and current events
can be regarded as an ongoing process of cultural multilevel selection, whose
dynamics are influenced by psychological traits that evolved by genetic multilevel
selection in the distant past (Wilson, 2002). Close attention to evolution as
a multilevel process is required to understand these complex but ultimately
comprehensible subjects.
At the formal theoretical level, multilevel selection theory has two virtues.
First, it provides a complete accounting system for evolutionary change. To build
a multilevel selection model, one must specify a global population that consists
of a number of local groups. The dynamics within groups, the manner in which
groups arise and disappear, and other parameters that define the structure of the
global population must all be specified. Only then can global evolutionary change
be monitored as a combination of within- and among-group processes. As we
shall see, all models of social evolution must assume a multi-group population
structure but in many cases the assumptions are not made explicit and the models
do not provide a complete accounting system for evolutionary change.
The second virtue is that the distinction between levels of selection captures
the essence of what is required for groups to evolve into adaptive units. If we
merely want to know if a given trait evolves, it is sufficient to calculate its change
in frequency in the global population, without caring whether its advantage
resides within or among groups. If we want to ask the more focused question
of when groups evolve into adaptive units, then it becomes critical to partition
global evolutionary change into within- and among-group components. Multilevel
selection theory carves the evolutionary process at its natural joints for the
particular question that Mandeville was addressing with his fable of the bees.
It is important to stress that this multilevel framework needs to be employed
for all traits associated with human sociality – not just conventional altruism
but also social control, cultural processes involving imitation and learning, and
traits conventionally regarded as self-interested in a benign sense. In all cases
the question to ask is “Does trait x replace alternative traits by virtue of a relative
fitness advantage within groups or between groups?” When this nested series
of fitness comparisons is carefully employed, the set of group-level adaptations
includes but extends far beyond traits that are conventionally regarded as altruistic.
WHAT IS INDIVIDUAL SELF-INTEREST?

Multilevel selection theory explains adaptive groupings as a product of group-
level selection. Mandeville and the tradition that he represents explain adaptive
groupings as a product of individual self-interest. On the surface, it would be
difficult to imagine two positions more opposed to each other.
To proceed further we must understand what is meant by the concept of
individual self-interest. One possibility is that it will provide an alternative theory
that explains adaptive groupings better than multilevel selection theory. Another
possibility is that it provides an alternative perspective, that explains adaptive
groupings only when multilevel selection theory arrives at the same conclusion.
A third possibility is that it simply fails to provide a coherent account of adaptive
groupings.
Unfortunately, there is no single “it” when it comes to the concept of individual
self-interest, but rather a number of concepts that are rarely distinguished and often
incompatible with each other. At least three layers of diversity can be identified.
(1) The evolutionary concept of self-interest is based on survival and reproduction,

but it includes a number of different formulations, as I will show.
(2) The economic concept of self-interest is based on maximizing a utility that
is often unspecified and therefore broader than the evolutionary concept of
self-interest, with many different specific formulations.
(3) Self-interest can be defined in terms of behavioral actions or psychological

motives. In evolutionary biology this distinction is nicely captured by the more
general distinction between ultimate and proximate causation. All adaptations

must be explained in terms of their effects on survival and reproduction
(behavioral actions) but also in terms of the proximate mechanisms that cause
organisms to act as they do (psychological motives). These explanations are
complementary and one can never substitute for the other. Economic models
often ignore this complementarity by including psychological and material
costs and benefits in a single utility function.
In this essay I will concentrate largely on evolutionary concepts of self-interest
(1) and the proximate/ultimate distinction (3). However, I doubt that the expanded
range of utilities considered by economic models will alter my basic conclusions.
Let us begin by considering the diversity of concepts of self-interest that exist
within evolutionary biology.
Self-Interest as Maximizing Relative Fitness Within Groups
Within the framework of multilevel selection theory, it is easiest to associate

self-interest and group-interest with within- and among-group selection respec-
tively. Thus, purely self-interested individuals are a product of pure within-group
selection and are driven exclusively to maximize their relative fitness within
groups. Purely group-interested individuals are a product of pure among-group
selection and are driven exclusively to maximize the fitness of their group, relative
to other groups in the total population. It is obvious that by these definitions, the
claim that adaptive groups can be forged out of self-interest is as wrong as it can
possibly be. The essential insight of multilevel selection theory is that natural
selection within groups is insensitive to the welfare of the group, as we have seen.
If a disposition to behave that evolves entirely by within-group selection benefits
the whole group, it does so only as a coincidental byproduct. To claim that relative
fitness maximization within groups miraculously results in adaptive groups is to
ignore the reasoning that led to the rejection of group selection in the 1960s and
all the subsequent developments that led to its revival.
Self-Interest as Maximizing Absolute Individual Fitness
Despite the importance of relative fitness comparisons in multilevel selection

theory, the everyday concept of self-interest is often framed in absolute terms.
Purely self-interested individuals are driven to maximize their welfare without
regard to others, not in comparison to others. If I help you in the course of helping
myself, that is no better or worse than if I hurt you, since my only goal is my own
welfare, construed atomistically.

At the formal theoretical level, most economic models and many evolutionary
models of self-interest are similarly based on the maximization of an absolute
quantity. In economic models the quantity is a utility function whose details are
often left unspecified. In evolutionary models the quantity is fitness, defined in
terms of survival and reproduction. In both cases, the models axiomatically assume
that the absolute value of the quantity will be maximized. Economists take it as
a given that individuals are utility maximizers. Similarly, when an evolutionary
biologist predicts that an animal will cooperate rather than defect because its
fitness as a cooperator is higher than its fitness as a defector, the evolutionary
biologist is assuming that the animal has evolved to maximize its absolute fitness
(Wilson, 2004).
What justifies this assumption? All evolutionary biologists acknowledge that
natural selection is based on relative fitness, so the absolute fitness criterion must
be justified in terms of relative fitness. This is the case for nonsocial traits that
influence only the fitness of the actor. In a population with two types, A and S, if A’s
absolute fitness is higher than S then its relative fitness is higher as well. However,
the situation becomes more complicated for traits that influence the fitness of others
in addition to the actor, as we have already seen. If the population consists of a
single group and A-types are public good providers that increase everyone’s fitness
(including their own) by x units at a private cost of y units, then their relative fitness
is lower and they will decrease in frequency despite having increased their absolute
fitness whenever x − y > 0. In this case the absolute fitness criterion cannot be
justified in terms of relative fitness and fails to correctly predict what evolves.
Continuing this example, suppose that our A and S types live in an infinite
population randomly subdivided into groups of size N within which the public
goods are shared. A-types are less fit than S-types within every group, but groups
with more A-types contribute more to the total gene pool than groups with fewer
A-types. It turns out that for these particular assumptions, the trait evolves by
among-group selection, despite being selectively disadvantageous within groups,

whenever x − y > 0; that is, whenever A-types increase their own absolute
fitness. Another way to state the same result is by noting that in randomly formed
groups, an A-type’s effect on itself (x − y) causes evolutionary change but its
effect on other members of its group does not because they are just a random
draw from an infinite population (Grafen, 1984; Nunney, 1985).
To summarize, for traits that influence the fitness of others in addition to the
actor, the absolute fitness criterion can be justified in terms of relative fitness, but
only given certain assumptions about the population structure. If there is only one
group or if variation in the frequency of A among groups is either above or below

random, then self-sacrificing or spiteful traits that violate the absolute fitness
criterion can evolve. In addition, very plausible social interactions can cause the
absolute fitness criterion to fail even in randomly formed groups. Since this point
is not widely appreciated, I will illustrate it with two elaborations of our example
involving public good providing A-types and freeloading S-types (see Wilson,
1998, 2004; Wilson & Kniffin, 1999 for extended treatments).
In the first elaboration, instead of just two types consider a range of strategies
of the form “act as an A-type if there are fewer than x other individuals acting
as an A-type in your group,” in which x varies from 0 to N, where N is the size
of the group. This range of strategies includes the original unconditional S-type
(x = 0) and unconditional A-type (x = N), but the inclusion of the intermediate
types alters the outcome of the model by causing behavioral variation among
groups to be less than random, even in randomly formed groups. This happens
because the intermediate strategies tend to compensate for each other, with those
less sensitive to providing the public good “turning off” in the presence of those
more sensitive. A group will be entirely altruistic only when it consists entirely of
the x = N type, because every other type reverts to selfishness in the presence of a
sufficient number of altruists. Similarly, a group will be entirely selfish only when
it consists entirely of the x = 0 type, because every other type reverts to altruism
in the presence of a sufficient number of selfish individuals. The below-random
behavioral variation among groups caused by these social interactions has an
effect on the evolution of public good provision similar to below-random genetic
variation in the original model.
In the second elaboration, suppose that A and S are socially transmitted behav-
iors rather than genetically innate. However, the process of social transmission
is governed by innate rules that evolve by genetic evolution. Groups are formed
at random by individuals bearing their innate transition rules and an acquired
behavior (A or S) from their previous interactions. The behavioral composition
of the groups then changes according to the transmission rules. Under plausible
conditions, transmission rules can evolve that create above-random behavioral
variation, even in randomly formed groups. All of us are familiar with the internal
and external pressures to conform, promoting uniformity within groups and
differences among groups. The above-random behavioral variation among groups
caused by these social interactions has an effect comparable to above random
genetic variation in the original model.
How do these elaborations escape the argument that effects on others in
randomly formed groups have no effect on global evolutionary change? This
argument assumes that no one else in the group will respond to the actions of the
focal individual. If others either withdraw or contribute their own social benefits in
reaction to the focal individual’s action, then the calculation of costs and benefits
that ignores the response becomes invalid. For example, imagine an individual who
decides to provide a public good because his share of it exceeds his private cost
(x − y > 0). That individual will be disappointed if another member of his group
stops contributing to the public good based on his decision, leaving him stuck with
the cost (−y). Similarly, an individual who decides not to provide a public good
because his share does not exceed his private cost (x − y < 0) has made the wrong
choice if others would have provided public goods on the basis of his decision.
In short, the absolute fitness criterion can be justified in terms of relative
fitness only for a narrow range of parameter values. Departures from randomness
at the behavioral level can invalidate the absolute fitness criterion, as surely
as departures from randomness at the genetic level. This fact is not widely
appreciated because the core evolutionary models of social behavior (such as kin
selection) assume that behaviors are coded directly by genes, thereby ignoring the
many plausible mechanisms that can generate nonrandom behavioral variation in
randomly formed groups.
Let us now take stock of the absolute fitness criterion. Models that employ the
criterion must specify the same basic parameters as multilevel selection models –
a global population that consists of a number of local groups, the dynamics within
groups, the manner in which the groups arise and disappear, and so on. Nothing
new has been added – these are properties of the social environment, not any
particular theoretical framework. The absolute fitness criterion is a claim about
what evolves in a particular social environment. As we have seen, the claim is not
general but holds only for a narrow range of parameter values. Earlier I stated that
multilevel selection theory has the virtue of being a complete accounting system
for evolutionary change. The absolute fitness criterion does not possess this virtue.
Even when the absolute fitness criterion can be justified in terms of relative
fitness, it only accounts for what evolves in the total population. More is required
to address the specific question of when groups evolve into adaptive units. Ac-
cording to multilevel selection theory, this happens precisely when among-group
selection outweighs within-group selection. Behaviors that satisfy the absolute
fitness criterion can evolve by either within- or among-group selection – in the

later case, when the individual’s share in the public good outweighs the private
cost. When we use the absolute fitness criterion to ask the central question “can
adaptive groups be formed out of self-interested individuals?”, the answer is
“yes,” but only under the conditions already specified by multilevel selection
theory; i.e. when among-group selection outweighs within-group selection. The
concept of self-interest per se adds nothing.
This point can be illustrated with an example from the social insects – the
real life equivalent of Mandeville’s fable. In the desert fungus-growing ant
Acromyrmex versicolor, colonies are initiated by groups of queens rather than

by single individuals. The queens must do the work of the colony until the first
generation of workers is born. One queen becomes a specialized above-ground

forager, gathering vegetation for the fungus garden, while the others remain below
ground to excavate the tunnels and cultivate the garden. Foraging above ground
is more dangerous than staying below ground so the forager substantially reduces
her fitness compared to the other queens in her colony (Rissing et al., 1989;
Wilson, 1990 ). How can we explain this apparent example of altruistic behavior?
The first impulse of biologists was to assume that the queens are genetically
related, but DNA evidence revealed that the groups are randomly formed. The next
impulse was to assume that the forager was forced to assume her role on the basis
of aggressive interactions, but behavioral observations revealed no evidence for
fighting and the forager is not smaller than the other queens. The correct answer
appears to involve a very strong form of among-group selection. Many colonies are
initiated at the same time and the first to raise a generation of workers succeeds
at becoming established in competition with the other colonies. If a specialized
above-ground forager provides a sufficient competitive edge in among-group
competition, the behavior can evolve despite its selective disadvantage within
groups – even when the groups are randomly formed. Knowing all of this, we
can imagine the specialized forager reasoning about her options as if she were a
person who cares only about her absolute fitness: “If I don’t forage, I will be safer
but my colony will lose in competition with other colonies. If I do forage, I might
die but if I live my colony will prevail against other colonies and I will share their
success. Therefore I will forage based only upon my self-interest. The fact that I
have reduced my fitness compared to the other queens in my colony is irrelevant,
because they are just a random draw from the total population.” This reasoning
correctly predicts that above-ground foraging will evolve, but only because
strong within-group selection is counterbalanced by even stronger among-group
selection. Self-interest does not provide an explanation of the behavior that stands
as an alternative to group selection. Instead, group selection is merely folded
into the definition of self-interest. To conclude, the absolute fitness criterion adds
nothing to the essential insight of multilevel selection theory, which is that adaptive
groupings evolve by among-group selection.
Inclusive Fitness and Other Extended Definitions of Self-Interest
When groups are non-randomly formed, effects on others – now a biased sample
of the total population – produce evolutionary change in addition to effects
on oneself. A common way to model these effects is with a self-term and an
other-term appropriately weighted by a coefficient. In kin selection theory the

coefficient is associated with the probability of sharing genes identical by descent,
but more generally it can be interpreted as a covariance between the individual’s

genotype and the genetic composition of its group (positive for above-random
and negative for below-random variation).
It might seem that these models straightforwardly show how self-interest is not
sufficient to explain the outcome of evolution whenever the covariance term is
nonzero. However, a shift in perspective enables both the self and other terms
to be interpreted as components of an expanded definition of self-interest. In kin
selection theory, individuals are said to maximize their own inclusive fitness,
which is often explained as an interest in their own genes, regardless of the bodies
in which they exist. In selfish gene theory, the self-interested agent is shifted from
individuals maximizing their inclusive fitness to the genes themselves.
Figure 1 shows how any trait that evolves in a multi-group population can be
interpreted as a form of self-interest, no matter how self-sacrificial (from Sober &
Wilson, 1998). The total population consists of two types, A and S, subdivided into
two groups of size n = 100 (smaller circles). Both types have a baseline fitness
Fig. 1. Evolution in a Two-Group Population with an Altruistic A Type (Dark Shading)

and a Non-Altruistic S Type (Light Shading). Note: Altruism is selectively disadvantageous
within each group but favored sufficiently at the group level that the global frequency of a
increases from 0.5 to 0.516.
of 10 and A-types increase the fitness of a single recipient in their group by five
units at a cost to their own fitness of one unit. They decrease their own absolute
fitness, which means that this behavior would not evolve in randomly formed
groups. However, the groups in Fig. 1 are highly segregated, with A at a frequency
of p = 0.2 in group 1 and p = 0.8 in group 2. A-types have a lower fitness than
S-types and decline in frequency within each group (within-group selection).
However, the group with more A-types grows larger and contributes more to the
total gene pool than the group with fewer A-types (between-group selection). The
balance between levels of selection is such that A-types increase in frequency in
the global population, despite decreasing in frequency within each group. Extreme
variation among groups is required to produce this outcome. For real-world
examples we must specify how the variation is created and maintained, especially
at high and low frequencies of the altruist in the global population. Interesting
forms of frequency-dependence can occur resulting in stable polymorphisms of
the two types, as discussed by Sober and Wilson (1998, Chap. 1) for a parasite
called the brain worm. For my present purpose, the important point is to have a
clear numerical example of a trait that is selectively disadvantageous within groups
but nevertheless increases in frequency in the global population by virtue of a
selective advantage between groups.
Knowing all of this, it is possible to imagine an individual deciding to become
an A-type based only on its self-interest: “As an A-type, it’s true that I give up a
unit of fitness but I also stand a higher chance of being in successful group two.
As an S-type, it’s true that I save a unit of fitness but I also stand a higher chance
of being in unsuccessful group one. The benefits of group membership outweigh
the cost of the behavior; therefore I will become an A-type without caring about
anyone else’s fitness.”
This egoistic reasoning process arrives at the correct answer about what evolves
because it includes the multi-group population structure in its calculation. It is
important to stress that in most real-life situations, an individual who decides to
become an A-type is not magically transported into other groups with a probability
governed by its newly adopted behavior. It merely loses a unit of fitness compared
to behaving as a S-type. The perspective of an omniscient egoist, standing on

Mount Olympus and surveying the entire structure of a multi-group population
before making its decision, is a heuristic device for the modeler, not a hypothesis
that can be taken seriously for humans, much less nonhuman species (Sober, 1998).
In its Olympian form, the concept of self-interest can account for anything
that evolves by natural selection, a virtue shared by multilevel selection theory.
However, it must never be forgotten that we are trying to answer a specific
question – when do groups evolve into adaptive units? Multilevel selection
theory’s other virtue is to address this question by distinguishing levels of
selection. The Olympian form of self-interest does not share this virtue. The
statement “everything that evolves is a form of self-interest” is no help at all.
Furthermore, efforts to address the specific question from the Olympian egoistic
perspective do not turn up any new principles but merely recreate the distinctions
that already exist within multilevel selection theory.
A good example of this futile exercise is selfish-gene theory, which averages
the fitness of genes across all contexts and labels whatever evolves as “selfish”
because it replaces that which did not evolve (Dawkins, 1976). By itself, the
concept of selfish genes is silent on the question of whether individuals or groups
qualify as adaptive units. To address this question, a separate concept of “vehicles”
must be invoked, which tells us whether genes evolve by coordinating with other
genes in the same organism (in which case the individual becomes the vehicle
of selection) or the same group (in which case the group becomes the vehicle
of selection). In this fashion, selfish gene theory must become as hierarchical as
multilevel selection theory to address the issue at the center of Mandeville’s fable
(Sober & Wilson, 1998; Wilson & Sober, 1994).
In retrospect, I think that the concept of self-interest in both evolution and
economics owes its popularity to at least two factors. First, natural selection
and rational choice can both often be represented as an optimization process in
which a single variable is maximized, even when it is a very complex function
of other variables. Second, self-interest is a highly simplified and intuitive case
of optimization, in which an individual juggles costs and benefits to maximize a
single utility. Thus, any optimization argument can be made intuitive by employing
the metaphor of self-interest. Unfortunately, the intuitive appeal of self-interest is
obtained at a cost. The details of what constitutes self-interest must change with the
particular optimization model, which means that there can be no single concept of
self-interest. In addition, optimization in general is poorly suited for addressing the
specific question of when groups evolve into adaptive units. Among-group selec-
tion by itself maximizes the relative fitness of groups and within-group selection
by itself maximizes the relative fitness of individuals within groups, but the entire
process of evolution in multi-group populations is a messy combination of these
opposing forces, along with other forces that prevent adaptations from evolving in
any sense. Showing how groups evolve into adaptive units requires a consideration
of opposing forces, not a process in which a single quantity is being maximized.
Psychological Concepts of Self-Interest
All of the preceding concepts of self-interest are based on how individuals act,
regardless of how they think or feel. Another set of concepts is based on how
Table 1. Psychological and Evolutionary Self-Interest are Defined on the Basis

of Different Criteria, Giving Rise to Four Combinations.
Evolutionary
Selfishness Altruism
Psychological
Selfishness
Altruism
individuals think and feel, regardless of how they act. As I mentioned previously,
one way to relate these two sets of concepts to each other from an evolutionary
perspective is with the distinction between proximate and ultimate causation.
Adaptive behaviors exist because they cause organisms to survive and reproduce
(ultimate causation) but also because of a set of psychological mechanisms that
cause the organisms to exhibit the behavior (proximate causation). The question
of whether the proximate psychological mechanism counts as self-interested is
potentially independent of whether the behavior counts as self-interested in terms
of its effects on fitness, which means that all four cells in Table 1 can potentially
be occupied (from Sober & Wilson, 1998, p. 204). For example, one individual
might be motivated to help others (evolutionarily altruistic) because she values the
welfare of others as an end in itself (psychologically altruistic). Such an individual
would occupy the lower right-hand cell of Table 1. Another individual might be
motivated to help others (evolutionarily altruistic) but only as a means to making
herself feel good (psychologically selfish). Such an individual would occupy
the upper right-hand cell of Table 1.
Psychological egoism is the claim that all psychological mechanisms that drive
behavior count as self-interested, in which case the bottom row in Table 1 would
be empty. As with the evolutionary concept of self-interest, we need to ask two
questions: Is the claim of psychological egoism true, and does it shed any light
on our specific question of when groups evolve into adaptive units?
The first question is evaluated in detail in part II of Sober and Wilson (1998).
To make a long story short, the claim cannot be refuted logically but is unlikely
to be true based on evolutionary principles. Any behavior, no matter how
other-oriented, can theoretically be caused by egoistic proximate mechanisms,
but these mechanisms are often like inefficient and error-prone Rube Goldberg
devices compared to simpler and more robust non-egoistic mechanisms. After
all, if behaviors have been selected to be other-oriented, what simpler way to
motivate them than by other-oriented proximate mechanisms? The empirical
evidence for exclusive psychological egoism (bottom row of Table 1 empty) is
nonexistent. On the contrary, the most careful research points to a mixture of self-
and other-oriented mechanisms and the case for exclusive egoism survives only
by postulating ever more remote and difficult to test egoistic hypotheses once the
simpler ones have been eliminated (e.g. Batson, 1991).
Even more important, the answer to the second question is “no,” regardless
of the answer to the first question. Just like the extended evolutionary concepts
of self-interest, psychological concepts suffer from being too ambitious. In their
drive to explain everything as a form of self-interest, they lose their ability to ask
more focused questions. Let us grant for the sake of argument that the bottom
row of Table 1 is empty. Insofar as psychological egoism is sufficiently flexible
to motivate both self- and other-oriented behaviors (both columns of Table 1), it
remains as silent as selfish gene theory on the factors that cause groups to evolve
into adaptive units (the first column vs. the second column).
I will not dwell further on psychological concepts of self-interest, in part
because most proponents of self-interest in evolutionary biology and economics
do not dwell on them either. Even though self-interest is sometimes defended as
the way that people actually think, this position is usually abandoned in favor of
the much looser position that people act “as if” they are self-interested, regardless
of how they actually think, returning us to the realm of behavior that we have
already examined. The impulse to think of genes and organisms that don’t think
at all (such as bacteria and plants) as self-interested makes it clear that the
concept of self-interest is serving primarily as a heuristic device for the human
modeler, not a claim about the mechanisms that cause behavior in the proximate
sense.
A NEW THEORETICAL FRAMEWORK FOR

THE STUDY OF ADAPTIVE HUMAN SOCIETY
I conclude that evolutionary concepts of self-interest do not provide a framework
for explaining how groups evolve into adaptive units. The version based on
relative fitness within groups shows that self-interest is the problem, not the
solution. The more extended versions explain adaptive groupings only insofar
as they overlap with the outcome of among-group selection. The most grandiose
versions of self-interest become synonymous with “anything that evolves by
natural selection” but this is a weakness rather than a strength when it comes to
addressing the specific question at the heart of Mandeville’s fable.
On the other hand, multilevel selection theory is sufficiently general as an
accounting method for evolutionary change, and sufficiently focused on the
question of how groups evolve into adaptive units, to provide a theoretical
framework for the study of human society. Group selection became such a heresy
in the middle of the 20th century that few people know what needs to be retained
and what needs to be rejected in the light of subsequent developments. What

needs to be retained is the fundamental insight that group-level adaptation requires
group-level selection. What needs to be rejected is the notion that among-group
selection is always weak compared to within-group selection. Once we appreciate
that natural selection is truly a multilevel process, especially in human genetic
and cultural evolution, then we can use multilevel selection theory to identify the
factors that cause human groups to function well or poorly as adaptive units.
It is beyond the scope of this essay to outline how evolutionary theory in general
and multilevel selection theory in particular provide a new theoretical foundation
for the study of human society. Fortunately, it is also unnecessary, because a grow-
ing number of economists, political theorists, anthropologists, psychologists, and
evolutionary biologists are already at work building the foundation (e.g. Boehm,
1999; Bowles & Gintis, 1998; Fehr, Fischbacher & Gachter, 2002; Henrich, 2004;
Hodgson, 1993, 2001; Ofek, 2001; Rubin, 2002; Sober & Wilson, 1998; Wilson,
2002). An excellent example is provided by the recent edited volume Genetic
and Cultural Evolution of Cooperation (Hammerstein, 2003), which employs
the same multilevel selection framework to study the evolution of human society
and the evolution of single organisms, with economists playing a key role in
the integration.
This essay contributes to the effort by showing that the concept of self-interest
does not provide an alternative. My parting generality is that adaptive human
groups must be explained in terms of among-group processes – in the ancient past,
resulting in innate psychological mechanisms that evolved by genetic multilevel
selection, and the more recent past, resulting in products of cultural multilevel
selection. Switching from individual self-interest to among-group processes as
our explanatory framework is a paradigmatic change, even before we consider the
details. The new fable of the bees leads in a different direction than Mandeville
ever imagined.
NOTE
1. An historically astute reviewer pointed out that Smith did not agree with Mandeville
in his Theory of Moral Sentiments (III. 1.3). However, Smith’s theory of moral sentiments
was itself neglected in the development of the modern concept of self-interest. Many of the
mechanisms discussed in Smith’s book could have emerged from group-level adaptations
in terms of multilevel selection theory. They are self-organizing but not self-interested in
the sense of evolving by within-group selection.
REFERENCES
Batson, C. D. (1991). The altruism question: Toward a social-psychological answer. Hillsdale, NJ:
Lawrence Erlbaum.
Boehm, C. (1993). Egalitarian society and reverse dominance hierarchy. Current Anthropology, 34,
227–254.
Boehm, C. (1999). Hierarchy in the forest. Cambridge, MA: Harvard University Press.
Bowles, S., & Gintis, H. (1998). Is equality passe? Boston Review, 23(6), 4–26.
Boyd, R., & Richerson, P. J. (1985). Culture and the evolutionary process. Chicago: University of
Chicago Press.
Camazine, S., Deneubourg, J.-L., Franks, N. R., Sneyd, J., Theraula, G., & Bonabeau E. (2001).
Self-organization in biological systems. Princeton: Princeton University Press.
Dawkins, R. (1976). The selfish gene. Oxford: Oxford University Press.
Fehr, E., Fischbacher, U., & Gachter, S. (2002). Strong reciprocity, human cooperation, and the
enforcement of human social norms. Human Nature, 13, 1–25.
Fehr, E., & Gachter, S. (2002). Altruistic punishment in humans. Nature, 415, 137–140.
Grafen, A. (1984). Natural selection, kin selection and group selection. In: J. Krebs & N. Davies (Eds),
Behavioural Ecology: An Evolutionary Approach (pp. 62–84). Oxford: Blackwell Scientific
Publications.
Hamilton, W. D. (1975). Innate social aptitudes in man: An approach from evolutionary genetics. In:
R. Fox (Ed.), Biosocial Anthropology (pp. 133–155). New York: Academic Press.
Hammerstein, P. (2003). Genetic and cultural evolution of cooperation. Cambridge, MA: MIT
Press.
Heckathorn, D. D. (1990). Collective sanctions and compliance norms: A formal theory of
group-mediated social control. American Sociological Review, 55, 366–384.
Heckathorn, D. D. (1993). Collective action and group heterogeneity: Voluntary provision vs. selective
incentives. American Sociological Review, 58, 329–350.
Henrich, J. (2004). Cultural group selection, coevolutionary processes, and large-scale cooperation.
Hodgson, G. M. (1993). Economics and evolution. Cambridge, UK: Polity Press.
Hodgson, G. M. (2001). How economics forgot history. New Brunswick, NJ: Routledge.
Maynard Smith, J., & Szathmary, E. (1999). The origins of life: From the birth of life to the origin of
language. Oxford: Oxford University Press.
Michod, R. (1999). Individuality, immortality, and sex. In: L. Keller (Ed.), Levels of Selection in
Evolution (pp. 53–74). Princeton, NJ: Princeton University Press.
Nunney, L. (1985). Group selection, altruism, and structured-deme models. Am. Nat., 126, 212–230.
Ofek, H. (2001). Second nature: Economics origins of human evolution. Cambridge, UK: Cambridge
University Press.
Richerson, P. J., & Boyd, R. (in press). The nature of cultures. Chicago, IL: University of Chicago Press.
Rissing, S. W., Pollock, G. B. et al. (1989). Foraging specialization without relatedness or dominance
among co-founding ant queens. Nature, 338, 420–422.
Rubin, P. (2002). Darwinian politics: The evolutionary origin of freedom. New Brunswick, NJ:
Seeley, T. (1995). The wisdom of the hive. Cambridge, MA: Harvard University Press.
Seeley, T. D., & Buhrman, S. C. (1999). Group decision making in swarms of honey bees. Behavioral
Ecology and Sociobiology, 45, 19–31.
Sober, E. (1998). Three differences between evolution and deliberation. In: P. Danielson (Ed.),
Modeling Rationality, Morality, and Evolution (pp. 408–422). Oxford: Oxford University
Press.
Wilson, D. S. (1990). Weak altruism, strong group selection. Oikos, 59, 135–140.
Wilson, D. S. (1998). Hunting, sharing and multilevel selection: The tolerated theft model revisited.
Current Anthropology, 39, 73–97.
Wilson, D. S. (1999). A critique of R. D. Alexander’s views on group selection. Biology and
Philosophy, 14, 431–449.
Wilson, D. S. (2002). Darwin’s Cathedral: Evolution, religion and the nature of society. Chicago:
Wilson, D. S. (2004). What is wrong with absolute individual fitness? Trends in Ecology and Evolution,
19, 245–248.
Wilson, D. S., & Kniffin, K. M. (1999). Multilevel selection and the social transmission of behavior.
Human Nature, 10, 291–310.
Wilson, D. S., & Sober, E. (1994). Reintroducing group selection to the human behavioral sciences.
Behavioral and Brain Sciences, 17, 585–654.
GROUP SELECTION AND

METHODOLOGICAL INDIVIDUALISM:
COMPATIBLE AND COMPLEMENTARY
ABSTRACT
Various authors allege that the theory of group selection is inconsistent with
methodological individualism, and therefore analysts must reject at least one
of these principles. The present article argues for their compatibility. The
meaning of methodological individualism is clarified, and the new version of
group selection (articulated by Wilson & Sober, 1994, 1998) is explained. The
two principles are then incorporated into a single methodological approach.
Group selection affects the assumptions made about the kind of individuals
who populate social scientific models, while methodological individualism
requires models’ conclusions to follow from the actions and interactions of
those individuals.
INTRODUCTION
In his theory of cultural evolution, Friedrich Hayek (1991) explicitly adopts
the notion of group selection as an evolutionary force that leads societies to
adopt group-beneficial rules and practices. Group selection is a hypothetical

process in which traits (whether genetic or cultural) evolve because members of
groups that practice them experience greater reproductive success than members

ISSN: 1529-2134/doi:10.1016/S1529-2134(04)07010-3
221
222 DOUGLAS GLEN WHITMAN
of groups that do not. In his earlier work (especially Hayek, 1952), Hayek
emphasizes the importance of methodological individualism in the social sciences.
Methodological individualism is the practice of explaining social phenomena

in terms of the behaviors and interactions of individuals, rather than those of
“collective” constructs such as groups, classes, or ideologies.
In an influential article, Viktor Vanberg (1986) argues that group selection
and methodological individualism are incompatible, a position that has gathered
the assent of other social scientists. Vanberg alleges a shift in Hayek’s thinking,
“from the notion that behavioral regularities emerge and prevail because they
benefit the individual practicing them, to the quite different notions [sic] that
rules come to be observed because they are advantageous to the group” (1986,
p. 83). Although Vanberg argues for scrapping group selection in favor of
methodological individualism, others – notably, Geoffrey Hodgson (1991) –
have lobbied for the opposite path of scrapping methodological individualism
in favor of group selection. Hodgson points to “an inconsistency in Hayek’s
work between, on the one hand, the ideas emanating from his individualist roots,
and, on the other, his growing commitment to an evolutionary perspective. In
an evolutionary context, methodological individualism has to be either redefined
or abandoned” (1991, p. 78). Hodgson agrees with Stephan Boehm that “Hayek
is by no means the champion of methodological individualism that he claims
to be” (Boehm, 1989, quoted in Hodgson, 1991, p. 78). Where Vanberg and
Hodgson agree is that Hayek has erred by embracing both group selection and
methodological individualism.
Some biologists, too, accept the proposition that group selection and
methodological individualism cannot coexist. Wilson and Sober (1994, 1998),
who have argued strenuously for the acceptance of group selection theory
within evolutionary biology, have indicated an opposition to methodological
individualism. After observing the prominence of group-level theorizing in the
early social sciences, they note:
More recently, the group-level perspective has fallen on hard times, although to different degrees
in different disciplines. Social groups are seen not as adaptive units in their own right but as
by-products of individual self-interest (Wilson & Sober, 1998, p. 133).
Wilson and Sober also favorably cite D. T. Campbell’s denunciation of

methodological individualism:
Methodological individualism dominates our neighboring field of economics, much of
sociology, and all of psychology’s excursions into organizational theory. This is the dogma
that all human social group processes are to be explained by laws of individual behavior –
that groups and social organizations have no ontological reality – that where used, references
to organizations, etc. are but convenient summaries of individual behavior . . . We must reject
Group Selection and Methodological Individualism 223
methodological individualism as an a priori assumption, make the issue an empirical one, and
take the position that groups, human social organizations, might be ontologically real, with
laws not derivable from individual psychology . . . (Wilson & Sober, 1994, p. 599, ellipses in
original).
Later, Wilson and Sober indicate a desire to “tear down the edifice of individualism”
(1998, p. 330) by which they seem to mean both methodological individualism (as
distinct from group functionalism) and psychological individualism (as distinct
from altruism). They argue that methodological individualism owes its popularity
in the social sciences to the desire to avoid the seeming mysticism of holistic
approaches:
. . . [T]he reason for individualism’s growing attractiveness was not the confirmation of a novel
empirical theory that entails that individualism must be true (Wright, Levine & Sober, 1992).
The term methodological individualism is therefore apt, in that this version of individualism is
not grounded in any particular scientific theory (ibid. p. 329).
Thus, Wilson and Sober view the evidence in favor of group selection as
evidence against methodological individualism (which they think should be
an empirical rather than methodological proposition). In short, proponents
of group selection and proponents of methodological individualism agree
on one thing: you cannot have both. If you accept one, you must reject
the other.
The ambition of this paper is to put to bed, at long last, the notion that
methodological individualism (hereafter, MI) and group selection (hereafter,
GS) are fundamentally incompatible. It is true that MI places some constraints
on GS. Such conditions are fully consistent with the new GS theory as
articulated by Wilson and Sober (1994, 1998), which resists the tendency of
early GS theorists like V. C. Wynne-Edwards (1962) to assume group-beneficial
evolutionary processes as the norm. But that conclusion is a far cry from
Vanberg’s dismissal of GS as “theoretically vague, inconsistent with the basic
thrust of Hayek’s [methodologically] individualist approach, and faulty judged
on its own ground” (1986, p. 97), as well as Hodgson’s claim that the use of
both concepts constitutes “a major internal inconsistency in Hayek’s work”

(1991, p. 69).
In defending the compatibility of MI and GS, I am greatly assisted – and
anticipated – by Zywicki (2000), which recognizes the viability of Sober and
Wilson’s (1998) revised version of GS and its applicability to Friedrich Hayek’s
(1979, 1992) theory of cultural evolution. I wish to expand on Zywicki’s point by
directly addressing the implications for MI of accepting GS.
The principle of MI requires that in a satisfying social scientific theory, the
choices and behavior of individuals, as affected and constrained by the physical
environment, must build up to the system effects at any given point in time.
While the behavior of individuals is affected and constrained by the institutional
environment as well, MI posits that the institutional environment itself can in

principle be fully explained by the behavior of past and present individuals (again,
as affected by the physical environment). I will contend that MI, understood in this
way, is indeed a necessary condition for a satisfying theory. But it is not sufficient
for a good theory, because it severely underdetermines the assumptions theorists
can make about the preferences and beliefs of agents. A good theorist must also ask
what sort of individuals are likely to populate the system in question – a question
that inevitably brings in an evolutionary standpoint. The evolutionary forces at
work may include group-level as well as individual-level selection.
APOLOGIA: A BRIEF DEFENSE OF METHODOLOGY
It is not without trepidation that I enter a musty methodological debate, and I

therefore offer the following justification for doing so. There is a popular and
respectable view, in economics and many other sciences, that methodology is a
distraction from the hard work of creating and testing hypotheses. Competing
hypotheses should be judged by evidence, not by methodology. With respect to
the GS controversy, for example, one might argue that the proof should be in
the empirical pudding; the justification for GS is its predictive power, if any,
not its methodological credentials. But I must disagree with this perspective,
for two reasons. First, the definitions of terms such as “testing,” “evidence,” and
even “hypothesis” are not self-evident, but methodology-laden. For instance, the
hypothesis that “a species will exhibit characteristic X” assumes that a species
is something that can be observed and studied, and likewise for characteristic X.
Similarly, in the social sciences, the notion that “markets” or “societies” behave
in particular ways relies on definitions for those terms. Methodology affects what
will be considered adequate definitions, and hence what will count as evidence for
the hypotheses in question.
Second, methodology inevitably affects the direction of scientific research. If,

for instance, GS were rejected on methodological grounds, then evolutionary
researchers would be led to seek different sorts of explanations for observed
phenomena. They might spend less time even looking for the kind of phenomena
that might be explained by GS. If, on the other hand, GS were approved on
methodological grounds, then researchers would be more inclined to search for
phenomena that might be explained on those grounds and less inclined to look for
alternative hypotheses. And the approach that says “methodology is not important,
just let the evidence speak” is, itself, a methodology that in some measure approves
of GS – and any other hypothesis, for that matter. If the pursuit of science were
costless, this perspective would be unobjectionable: every potential hypothesis
could be pursued without distracting from any other. But in the real world of
scientific research, scarcity requires trade-offs. Methodology is the discussion
about how our limited scientific resources would be most gainfully deployed.
The question, then, is whether further research on GS is worthy of our effort and
attention.
EXPLAINING AND DEFENDING

The Meaning and Purpose of Methodological Individualism
To understand the meaning and purpose of MI, it will prove useful to consider
the arguments made in its defense by Hayek (1952). Importantly, this means I am
adopting a particular definition of MI that others may not share. It is possible that
some versions of MI are indeed incompatible with GS. However, since it is Hayek
whom Vanberg (1986) and Hodgson (1991) accuse of having strayed from his own
methodological program, it seems fair to consider Hayek’s conception of MI.
Hayek’s defense of MI, which he also refers to as the “individualist and
‘compositive’ method of the social sciences” (1952, pp. 61–76), depends on two
notions about the role of ideas in generating human action.
The first of these is the subjective character of social scientific data. The social
sciences must deal regularly with data that do not “exist” objectively as do the
data of the natural sciences. The reason is that the data themselves are concepts or
ideas. Hayek offers the example of the words “hammer” and “barometer”:
If the reader will attempt a definition he will soon find that he cannot give one without using
some term such as “suitable for” or “intended for” or some other expression referring to the use
for which it is designed by somebody. And a definition which is to comprise all instances of the
class will not contain any reference to its substance, or shape, or any other physical attribute.
An ordinary hammer and a steamhammer, or an aneroid barometer and a mercury barometer,
have nothing in common except the purpose for which men think they can be used (Hayek,
1952, pp. 44–45).
Other examples abound; “money,” “property,” “government,” and so on all have

definitions that require reference to the (shared) perceptions and intentions of
individuals. There do not exist physical tests for their existence.
Hayek understands that “methodological collectivism,” the alternative to MI,
relies upon the apprehension of subjective wholes:
Closely connected with the “objectivism” of the scientistic approach is its methodological
collectivism, its tendency to treat “wholes” like “society” or the “economy,” “capitalism” (as a
given historical “phase”) or a particular “industry” or “class” or “country” as definitely given

objects about which we can discover laws by observing their behaviour as wholes. While the
specific subjectivist approach of the social sciences starts, as we have seen, from our knowledge
of the inside of these social complexes, the knowledge of the individual attitudes which form
the elements of their structure, the objectivism of the natural sciences tries to view them from
the outside; it treats social phenomena not as something of which the human mind is a part
and the principles of whose organisation we can reconstruct from the familiar parts, but as if
they were objects directly perceived by us as wholes (Hayek, 1952, p. 92, emphasis added).
MI does not prevent the social scientist from apprehending the existence of wholes,
but rather cautions against trying to derive the laws of their behavior based on
alleged observations of the whole (note the emphasized part of the quotation
above). Such caution is justified by the fact that we cannot, in fact, observe wholes
directly; we infer their existence through the apprehension of particular events and
phenomena. For instance, a language is never observed in its entirely; instead, it
is observed in innumerable individual acts of communication that appear to draw
on similar vocabulary and rules.
The second notion behind Hayek’s compositive method is the distinction
between constitutive and speculative ideas (Hayek, 1952, pp. 61–65). Constitutive
ideas are those ideas that actually motivate people in their choices and behavior.
They include concepts such as money and law, preferences, and beliefs about how
the world works. Such beliefs may or may not be valid, but they are still relevant
because they affect behavior. For instance, a superstitious belief in the good fortune
brought by rabbits’ feet is real, in the sense that it actually leads people to produce
and buy the feet of rabbits. Such ideas (whether valid, invalid, or neither) are called
“constitutive” because they constitute the social phenomena that a social scientist
wishes to understand and explain.
Speculative ideas, on the other hand, are the ideas people (particularly social
scientists) use to explain social phenomena. They are theories, models, and
more rudimentary concepts used to understand the operation of the social world.
Speculative ideas often take the form of “meta-ideas,” in that they are ideas about
the existence and interaction of other ideas. The truth status of speculative ideas,
unlike that of constitutive ideas, is of the essence for the social scientist, at least if
he wants his theory to have explanatory or predictive power.
The two categories of ideas are not hermetically sealed from each other. Insofar
as one’s speculative ideas about how the social world works affect one’s behavior,
they are also constitutive ideas. And to that extent, social scientists must treat
them as given just like other constitutive ideas. For instance, if consumers believe
an economic theory that predicts a recession, they may alter their consumption;
if voters believe a politician’s claim that protectionist policies will increase their
purchasing power, they may cast their votes for that politician. But the social
scientist must carefully distinguish between those ideas he accepts as constitutive

(he thinks real people believe them) from those he accepts as speculative (he
believes them himself, at least provisionally). Otherwise, he runs the danger of

accepting popular but potentially false ideas as part of his own theory. Just as
it would be wrong to conclude that rabbits’ feet are lucky simply because some
people think so, it would be wrong to conclude that a popular economic theory is
correct because some people think it is.
The danger is a serious one, as it is not uncommon for people to have
anthropomorphic ideas, which attribute person-like characteristics to non-human
entities. What does “society” want? What are the interests of “the state?” Both
questions assume that entities composed of many individuals may be characterized
as having something akin to preferences that they will act to achieve. That
assumption may or may not be justified; it is the job of the social scientist to
discover when and if it is.
Hayek explicitly recognized the avoidance of as-yet-unjustified speculative
ideas as the root of MI:
That he [the social scientist] consistently refrains from treating these pseudo-entities as facts,
and that he systematically starts from the concepts which guide individuals in their actions, is
the characteristic feature of that methodological individualism which is closely connected to
the subjectivity of the social sciences (Hayek, 1952, p. 64).
People have ideas, often shared ones. Those ideas motivate their behavior, and their
behavior results in social phenomena. But the only ideas that actually motivate
behavior are constitutive. Speculative ideas only motivate behavior when they
become constitutive, as people believe them and act on them. MI is a mental
device whose primary function is to force scientists to sharply distinguish (others’)
constitutive ideas from (their own) speculative ideas.
To put it another way, MI is a bulwark against the accidental intrusion of
unproven hypotheses into one’s initial assumptions. Employing some collective
entity X in one’s theory means assuming, implicitly, that X is both real and subject
to certain behavioral regularities. That assumption may not be justified. By forcing
the analyst to carefully distinguish his own ideas from those of agents in his models,
MI prevents the social scientist from relying on speculative conclusions before they
have been justified by an analysis of constitutive premises.
Methodological Individualism, Scientific Reductionism, and Emergent Properties
MI is, strictly speaking, a social scientific perspective. Its central claim regards
the proper mode of explaining human social phenomena, not all phenomena
in all sciences. Admittedly, there exists an affinity between MI and scientific
reductionism, but there is no necessary connection. MI contends that human social

phenomena can be explained in terms of human ideas, but it takes no position
on whether all human ideas can be reduced to more basic phenomena such as
electrical impulses, hormones, and genes.
Indeed, social scientists often pair MI with the doctrine of methodological
dualism, which contends that social scientific methods differ fundamentally from
those of the natural sciences.1 In the social sciences, the analyst may begin with
mental phenomena such as preferences and beliefs, and use these concepts to
explain social phenomena, without necessarily knowing the physical processes that
underlie or generate the preferences and beliefs. As Ludwig von Mises argues, “we
do not know how external events – physical, chemical, and physiological – affect
human thoughts, ideas, and judgments of value. This ignorance splits the realm of
knowledge into two separate fields, the realm of external events, commonly called
nature, and the realm of human thought and action” (1981, p. 1). Adopting this
point of view does not mean that we are uninterested in the physical processes that
(may) generate mental events – only that it is possible for social science to proceed
without waiting for the natural sciences to discover the minute workings of the
brain first. In short, MI neither implies nor denies reductionism. MI says social
phenomena can and should be understood in terms of individual psychology; it
does not take a position on whether individual psychology can be understood in
terms of something even more basic.2
Nonetheless, methodological individualists and scientific reductionists have
something in common: they share a dissatisfaction with explanations of macro
phenomena that do not make explicit reference to the micro units that comprise
them. The central question here is whether higher-order entities (that is, entities
comprised by smaller-scale entities and their relationships) exhibit emergent
properties, and if so, whether all such emergent properties derive from the
properties of the constituent entities and their interaction.
MI admits the existence of emergent properties of social systems (such as the
price system, the state, the institution of money).3 But MI insists that, at least
in principle, those properties can be explained by the actions and interactions of
individuals who take part in those systems. The ideas and behavior of individuals
may be affected by institutional arrangements, such as firms and legislatures;
but those institutions themselves can be explained by the ideas and behavior of
present and past individuals.4 Moreover, the MI theorist is suspicious of alleged
emergent properties that have not yet been so explained. For instance, if someone
asserts that a legislature’s actions manifest a “general will” of the public5 or the
maximization of some well-defined social welfare function,6 the MI theorist will
be skeptical until a process is found by which a “general will” or social welfare
function could emerge from individual choices and behavior. Arrow’s Theorem
(Arrow, 1951) is the celebrated conclusion that no such social welfare function
exists, if we require such a function to satisfy a number of intuitive conditions.
Black’s Median Voter Theorem (Black, 1958), on the other hand, identifies some
narrow conditions under which the preferences of a particular voter may be
regarded as the preference set that guides legislative choices.
The scientific reductionist has similar, though not identical, concerns. He, like
the methodological individualist, wishes to place foundations under emergent
properties. He wishes to explain the behavior of molecules in terms of atoms
and their interaction; he wishes to explain atoms in terms of subatomic particles
and their interaction; and so on. There is, however, an important difference: in
the natural sciences, the existence of many emergent properties can be verified
independently of their explanation.7 A molecule can be isolated and observed under
lab conditions, whether we know the relations of its component atoms (or even that
such component atoms exist) or not. A social welfare function, on the other hand,
cannot be observed. It does not exist in the same sense as a molecule; its reality,
if any, is more akin to that of a logical deduction or a mathematical calculation.
Without logical proof, we have no strong independent reason to believe it
exists at all.
Yet emergent characteristics of social systems do exist. Prices do have a tendency
to reflect scarcity. Black’s Median Voter Theorem does provide a reasonable
explanation for some kinds of legislative outcomes. In certain types of banking
systems, such as those with fiat money, bad money does tend to drive out good
(Gresham’s Law). Far from denying the usefulness of such conclusions, MI
embraces them if they have been justified in terms of the choices and behavior
of individuals. If they have not yet been justified, the MI theorist will remain
skeptical; they might be true, but they have the status of unproven hypotheses.
Methodological Individualism as a Minimal Requirement
MI does not purport to provide sufficient conditions for good social science. Rather,
it guards against one type of error. Other errors must be avoided by means of other
principles, or just careful thought and research. Here, I wish to make four points
about what MI does not require.
First, MI does not specify the content of agents’ ideas. MI says the theorist
should begin with a notion about the constitutive ideas held by the agents being
modeled, including their preferences and beliefs, but it places no constraint on
what type of preferences and beliefs they might have. A theory in which agents
are assumed to love work and despise leisure might be criticized as unrealistic, but
it would not violate MI.
It follows that MI does not specify whether agent preferences will be selfish,
altruistic, or something else entirely.8 This point deserves emphasis, because the
association of MI with selfishness seems to have motivated Wilson and Sober’s

opposition to it.
This image of ‘selfish individuals’ may also [like the ‘selfish gene’] be metaphorical, but it is
more insidious because it can be taken literally. In other words, it is possible to believe that
individuals really are intentional systems motivated entirely by self-interest and this is, in fact,
the individualistic perspective that pervades the social sciences (Wilson & Sober, 1994, p. 601;
emphasis added).
Wilson and Sober’s conflation of MI with self-interest has been noted by Eric
Alden Smith:
. . . MI per se does not entail an assumption of selfishness or self-interest. Put as simply as
possible, MI asserts that the properties of any group result from the actions and interactions
of its individual members (Elster, 1983; Smith & Winterhalder, 1992); this can remain true
even if these individuals are altruistic, dedicated to the collective good, and so on (Smith, 1994,
p. 637).
Second, MI does not specify exactly how individuals make choices. They might
be perfectly rational actors, boundedly rational actors, rule followers, or even
automatons. Naturally, the assumptions made in this regard will have substantial
consequences for the predictions made, and some assumptions might be criticized
as unrealistic. Economists steeped in the rational-choice tradition are apt to
criticize models that place too much emphasis on cultural programming while
ignoring people’s responsiveness to incentives (thus treating people as either
rule followers or automatons). And that criticism might be valid – as might a
criticism of economic approaches that assume rational agents with unreasonably
high calculative abilities. On these matters, MI is silent.
Third, MI does not deny the existence of collective wholes. It simply requires
their behavior to be explained in terms of their components’ behavior, as holistic
explanations run the risk of ascribing agency where it does not exist. “Even where
individual actions and intentions are coordinated to achieve a collective outcome,
there is no justification for attributing agency to the collectivity, as if it had eyes

that see and hands that move” (Smith, 1994, p. 637).
Fourth, MI does not require the exclusive use of spontaneous order or “invisible
hand” explanations. Vanberg (1986) sees MI and spontaneous-order explanations
as closely connected: “[Methodological individualism] is behind Adam Smith’s
notion of the ‘invisible hand’ as it is behind Adam Ferguson’s conception of social
institutions as ‘the result of human action but not of human design’ ” (Hayek,
1967a, b, pp. 96ff; Vanberg, 1986, p. 80). But while spontaneous-order explanations
of social outcomes are generally consistent with MI, they are not required by it.
An explanation that relies on a collective decision-making process, such as voting

or consensus-building, is consistent with MI if the choices of individuals within
the process can be explained.

For instance, consider the question of why people drive on one side of the road.
An invisible-hand explanation would focus on the instability of any other proposed
equilibrium. Rational individuals who know that 51% or more of the population
tend to drive on the right will choose to drive on the right themselves – and as a
result, everyone will eventually drive on the right (or the left, if the initial majority
happens to lean that way). But this is not the only account consistent with MI. An
explicit collective decision could also have done the job: the religious or political
leaders of the society decreed that everyone should drive on the right. This is,
indeed, what happened in Sweden in 1963 (Lucas, 2002). There is nothing un-
MI about this story, as long as we recognize that individuals make the laws and
individuals follow them. The explanation of a collective choice on something more
controversial than selecting the side of the road would need to include some reason
why dissenters are willing (or forced) to go along with the outcome, as well as the
reasons why the designated decision-makers made the choice they did.
Because MI has so often been used in conjunction with particular assumptions
and approaches, including rationality, self-interest, and spontaneous order, we
must be especially careful not to confuse them. MI’s association with other
methodological notions and social scientific perspectives is mostly attributable to
the dominance of MI in economics, but the connection is historical, not definitional.
THE REJUVENATED THEORY OF GROUP SELECTION

The Old Group Selection Theory
In its earliest manifestation, GS theory was little more than an assertion, backed
by some empirical observations, that organisms tended to evolve group-beneficial
traits – that is, traits yielding benefits in terms of long-run reproduction for all
or most other members of a group, even while (possibly) creating costs for the
individual organism expressing the trait. For instance, Wynne-Edwards (1962)
observed a tendency of animal populations to limit their population density to
the optimal level allowed by the physical environment, even when individual
organisms within the population could benefit themselves by reproducing more
at the expense of the rest.
If intraspecific [within-species] selection was all in favour of the individual, there would be an
overwhelming premium on higher and ever higher individual fecundity, provided it resulted in
a greater posterity than one’s fellows. Manifestly this does not happen in practice; in fact the
reproductive rate in many species, and recruitment of adults in others, is varied according to
the current needs of the population (Wynne-Edwards, 1962, p. 19).
Economists will recognize the situation described as a kind of tragedy of

commons;9 in essence, Wynne-Edwards observed that some animal populations
appeared to have overcome the commons problem.
But how could the problem have been solved? Subsequent works in evolutionary
biology, most notably Williams (1966), argued that it would not be solved because
the group-level solution is vulnerable to subversion from within. Any trait that
allows an individual organism to contribute more offspring to the next generation
relative to other organisms without the trait will tend to grow in the population,
and any trait that causes it to leave fewer offspring relative to others will thus tend
to be pushed out. As a result, group-beneficial traits cannot survive in the long
run, despite Wynne-Edwards’s statement that when group-level and individual-
level selection conflict, “group-selection is bound to win” (Wynne-Edwards,
1962, p. 20).
This powerful critique led many evolutionary biologists to reject GS theory for
much of the latter half of the century. Wilson and Sober (1994, 1998) have since
resurrected GS theory. The new theory, while repudiating the notion that group-
selection is “bound to win,” elucidates some conditions under which it can win. I
discuss the new GS below.
Here, however, I wish to observe that even the old GS can coexist, albeit uneasily,
with MI. No one ever doubted that if some or all members of a group displayed
certain types of traits (such as a tendency to limit offspring or to warn of predators),
the entire group could benefit. The claim of those opposed to GS is simply that
such “altruistic” members will cease to exist in the long run, because they will
be weeded out. In the context of cultural evolution (the only context in which MI
is applicable), this would be equivalent to the claim that individuals could not
possess certain kinds of preferences and beliefs, because such individuals would
tend to be driven out by individuals with different preferences and beliefs. Now,
that argument may indeed be true under some conditions, but it is not an argument
required by MI. As discussed earlier, MI does not specify the content of agents’
ideas.
An objection to the old GS based on MI can be constructed by observing a
key difference between biological and cultural evolution. In biological evolution,
traits are “built into” organisms, and therefore a trait will be expressed as long
as there are individual organisms born with it in a given generation; whereas
in cultural evolution, the individual organism may choose to adopt a different
trait. Thus, an MI-consistent account must consider whether individuals will, in
fact, choose to retain group-beneficial traits. When addressing this question, we
should recall that MI does not require a specific decision-making mechanism (like
rationality) for individuals; they could be rule-followers, for instance. Notably,
Hayek’s (1988, p. 21) theory of cultural evolution downplays the role of reason
while emphasizing the human capacity for imitation, which lends support to a rule-
following conception of human nature. This perspective may partially explain why
Hayek willingly embraced the old GS.
The point here is not to defend the old GS. The point is that the error of the old
GS was only tangentially related to MI. The real fault of the old GS was its failure
to explain how a process of group-level selection could ever overcome the process
of individual-level selection. That task was left to the new GS theorists.
The New Group Selection Theory
In explaining and defending the new GS, it will be helpful to consider explicitly the
mathematical reasoning offered by Sober and Wilson (1998). Much of Sober and
Wilson’s discussion focuses on the evolution of altruistic traits, i.e. traits which
are both group-beneficial and create a cost to the individual organism. Although
altruistic traits are not a necessary element of GS theory, they do pose the greatest
challenge to that theory. (Later, I will consider treating altruism as a feature of
social equilibria, rather than as a feature of traits.)
The simplest explanation for how GS could enable the evolution of altruistic
traits relies on an application of Simpson’s Paradox, which says that the proportion
of observations with some property in an aggregate group may be smaller than the
proportion of observations with that property in every subgroup comprising the
aggregate. For instance (to use one popular example),10 hospital A may save a
larger fraction of all its patients than hospital B, even though hospital B saves
a larger fraction of its elderly patients than hospital A and a larger fraction of
its young patients than hospital A. This could occur because hospital B treats a
disproportionately large number of elderly patients, who have a lower survival rate
than young patients.
As applied to group selection, Simpson’s Paradox indicates that an aggregate

population might display an increase in the proportion of its members with an
altruistic trait, even though each subgroup within the population experiences
a decrease in the proportion of members with that trait. This occurs because
the subgroups with more altruists contribute disproportionately to the aggregate
population.
Suppose there exist two subsets of a larger population of some organism. Each
subset has 100 members. Both subsets have some members who are disposed
toward altruism: they will make personal sacrifices in terms of reproduction for
the benefit of the group. But subset A has a larger fraction of altruists than subset
B – say, 80% versus 20%. The overall population thus has 50% altruists.
Each subset will reproduce for one generation before being combined into a
single larger population. In the meantime, the proportion of altruists in each subset
will shrink, because they contribute fewer offspring than do the non-altruists. To be
specific, let us say each non-altruist contributes (on average) 10% more offspring
than does each altruist. As a result, subset A’s second generation will have 78.43%
altruists (versus 80% before), and subset B’s second generation will have 18.52%
altruists (versus 20% before). Thus, the proportion of altruists in each subgroup
declines.
Then the two subsets combine into a single population, but not in equal numbers.
The group with a larger fraction of altruists will contribute disproportionately
because they benefited more from the actions of the altruists. Suppose that the size
of subset A increased by a ratio of 2 (to 200 members), while B increased by a ratio
of only 1.2 (to 120 members). Then the combined population has 320 members,
with 179 altruists (157 from A and 22 from B). The fraction of altruists in the
aggregate population has increased, from an initial 50% before to just over 55%
now. In short, the altruistic fraction of the whole group increased, even while the
altruistic fraction of each subgroup decreased. This is Simpson’ Paradox at work.
Two aspects of this process bear emphasis. First, the result depends crucially on
the numbers chosen. If I had posited a wider disparity in the number of offspring
contributed by altruists and non-altruists, then the fraction of altruists in the whole
population could have decreased. If I had posited a smaller disparity in growth
rates of the subsets, then again, the fraction of altruists in the whole population
could have decreased. The point – one that the new GS theorists have emphasized
– is that the importance of GS depends on the relative strength of within-group
and between-group selection.
Second, without additional conditions being imposed, the process described
only works for one generation. Suppose that, after recombining into a single
population, the population never split up again. Then there would only be one
group, and within-group selection would result in a diminishing fraction of
altruists in each generation. Or suppose that the population did split up again, but
the altruists were distributed equally among subsets, so that each group had the
same fraction of altruists. Without variation in the representation of altruists, the
groups would not contribute disproportionately to the whole population when they
combined again.
In order to make the process work iteratively, we need additional conditions
on the evolutionary process. Specifically, the altruists must periodically be
reconcentrated so that some groups have greater representation of altruists than
others. This can happen through chance if the groups are small enough. As an
extreme example, suppose the whole population fissions into a large number of
two-member groups – that is, mating pairs. Random mating will result in some
fraction of two-altruist pairs, some fraction of a two-non-altruist pairs, and some

fraction of half-altruist pairs. Variation thus results automatically from a random
process. This is the case of kin selection, encompassed as a special case of GS.11
Even if groups are somewhat larger than two members, there will still be random
variation among groups in the representation of altruists. But as the “sample size”
gets larger, groups come to resemble the overall population, with insufficient
variation to cause differential contributions to the whole population due to altruism.
If groups are large enough, then GS requires the existence of a sorting process
whereby altruists have a better-than-random tendency to group together.
Addressing Potential Objections to the New Group Selection Theory
What objections might an MI theorist have to the version of GS theory just

presented? The difficulty with answering this question is that, as presented, the
GS theory is one of natural science, whereas MI is a principle of social science.
Nonetheless, it is not difficult to imagine an application of the argument to cultural
evolution; indeed, Wilson and Sober (1998) do attempt to extend their approach to
the cultural context (pp. 149–158). In this section, I will grapple with some possible
objections from the MI standpoint, keeping in mind the earlier observations about
the actual content of the MI principle.
The first objection to the new GS theory, as with the old GS theory, is that it is
implausible that any individual will be altruistic. This objection is easily dismissed:
MI does not specify what sort of preferences and beliefs individuals must have.
As noted earlier, MI has no necessary connection with the assumption of self-
interest, despite often being used alongside it. Altruism is one sort of preference
an individual might have, perhaps as a result of biological evolution, perhaps as
a result of cultural transmission of values. The critic of GS might argue that such
preferences (however acquired) will tend to be weeded out when they appear.
Whether the weeding process will always be effective, however, is the question at
hand. To assert that it will is to beg the question. And even if the assertion turns
out to be correct, it is not an assertion dictated by MI.
A second objection might be that individual incentives are being ignored. This
turns out, however, to be a variant of the first objection. Incentives are inherently a
matter of preferences, as what constitutes a sufficient incentive for one individual
may not be sufficient for another. An individual that cares about maximizing its own
offspring, and nothing else, will not, if it has a choice, be altruistic toward unrelated
individuals. But MI does not specify that individuals must value producing their
own offspring. To assert that they do – or more precisely, that they have traits
leading them to act as if they do – is again to beg the question. The same rebuttal
would apply to the assertion that individuals must value some specific goal other
than maximization of offspring, such as maximization of a selfish utility function.
A third objection might be that the theory asserts the existence of group-level
emergent properties. First, this is simply untrue, as the new GS theory recognizes
that the conditions for group-level emergent properties will not always exist. A
greater differential in reproduction between altruists and non-altruists, for example,
could reverse the results and lead to the evolution of selfishness. The absence of a
fission-and-regrouping process could also reverse the results. Second, as discussed
earlier, MI does not deny the existence of emergent properties; it merely asks
for their justification in terms of the relations among lower-level entities. That is
precisely what the Simpson’s Paradox version of GS does: it explains how a specific
type of natural selection under special conditions can lead to group functionalism.
A fourth objection might be that the conditions required for the process to work
– a large enough reproductive advantage for subgroups with more altruists, a small
enough reproductive advantage for non-altruists over altruists within subgroups,
a process of fission and regrouping within a larger population – are unrealistic or
too rare to be important. This, however, is an empirical objection, not an objection
based on the principles of MI. One might make a similar objection against Black’s
Median Voter Theorem, on grounds that the assumption of a one-dimensional
policy spectrum is unrealistic. Whether that’s true or not, it does not mean the
Theorem violates MI. MI is a necessary, not a sufficient, condition for a satisfying
social scientific theory.
A Simple Game Theoretic Extension
While GS theory is often deployed to explain the (possible) existence of altruistic

behavior, GS is not defined by altruism. GS may be manifested in cases where
altruism is not required; in such cases, within-group selection may be weak or
non-existent, and thus group-beneficial behavior evolves easily. Two simple games
can be used to make this point.
Public good problems, and other situations in which selfish behaviors can lead
to inferior results for all (and hence for the group), are often represented in the form
of a prisoners’ dilemma, as shown in Fig. 1. In this figure, S is a dominant strategy
for both participants in the game, and therefore S/S is the predicted equilibrium,
even though both players would be better off if the outcome were A/A.
In Fig. 2, we have what is commonly called a coordination game. Just as in
Fig. 1, both players are better off with A/A than with S/S. But in this game,
Fig. 1. Prisoners’ Dilemma.
there are two equilibria: A/A and S/S. Which equilibrium will occur depends
crucially on the history of the game. In a biological context, a population could
end up in the inferior S/S equilibrium simply because a large enough majority of
the founding population’s members had the S trait. In a choice context, no one
would have an incentive to depart (unilaterally) from the S strategy. Absent group
selection, there is no basis for assuming that either equilibrium would prevail. With
group selection, however, we can imagine a process that leads to the emergence
of the superior outcome: subgroups of the overall population could reach different
equilibria, and those groups in the superior A/A equilibrium would outproduce or
outperform those groups in the S/S equilibrium.
Figures 1 and 2 differ in just one respect: the payoff from playing S when the
other player plays A. Anything that reduces the size of this payoff in Fig. 1 moves
it closer to Fig. 2, until eventually Fig. 1 is transformed from a prisoners’ dilemma
into a coordination game.
How does such a transformation occur in the real world? The mechanisms are
well known, even if the processes of transformation are not. In some prisoners’
dilemma situations, such as the overuse of a common resource (an example to be
discussed more fully in the next section), privatization is a potential solution. But
other solutions exist as well, one of the most common being the establishment
of social norms to punish uncooperative behavior. The use of “trigger” strategies
and similar devices can provide a means of enforcing cooperative behavior in
repeated prisoners’ dilemmas, in which case the cooperative equilibrium and the
uncooperative equilibrium of the repeated game are equivalent to the multiple
equilibria of the coordination game.
Fig. 2. Coordination Game.

If prisoners’ dilemmas can be transformed into coordination games, GS might

seem superfluous in the explanation of how group-beneficial outcomes can evolve.
That conclusion is unwarranted, for two reasons. First, as argued above, attaining
the group-beneficial equilibrium in a coordination game is not to be taken for
granted. Some groups might reach the superior equilibrium while others fail.
(Some cartels succeed in enforcing high prices, while others do not.) Second,
obeying norms that transform one game into another may itself sometimes involve
a degree of unselfish behavior. If the members of a society are expected to impose
a punishment on a norm-violator, the act of imposing the punishment can be costly
to the person imposing it. In such a case, the enforcement of norms to solve a public
good problem is itself a kind of public good; this is known as the second-order
public good problem (Wilson & Sober, 1998, p. 144). How can the second-order
public problem be overcome?
The answer suggested by the new GS theory is that agents’ attitudes may
be internal to the model. If some societies have members with inherent
or ingrained preferences that reinforce group-beneficial norms – including
punishment of non-cooperators, toleration and social support of property
claims, and so on – those societies will contribute a disproportionately large
number of members to future generations relative to other societies. In cultural
evolution, this effect can be exaggerated by the ability of individuals to change
their allegiances. Individuals with the ability to choose among societies will
naturally choose those that have managed to solve (or mitigate) their public-good
problems.
Furthermore, some second-order public good problems involve less costly
activities than the first-order public good problems they are intended to solve.
The lower is the costliness of performing group-beneficial activities, the weaker
will be the pressure of within-group selection relative to between-group selection.
Wilson and Sober use this argument to explain the sharing of meat in hunter-
gatherer societies, even though the act of hunting imposes disproportionate costs
on the hunter, including the possibility of death. Such societies typically have a
variety of rewards for hunters who do share their meat, such as privileged mate
selection, and punishments for those who don’t, such as ostracism.
Let’s call hunting and sharing the primary behavior and the rewards and punishments that
others confer on hunters the secondary behaviors (Ellickson, 1991) . . . The evolution of the
secondary behavior (the rewards and punishments) can be analyzed in exactly the same way as
the evolution of the primary behavior. By causing another individual to perform an altruistic
primary behavior such as hunting and sharing, the secondary behavior increases the fitness of
the entire group. At the same time, the secondary behavior is likely to require at least some time,
energy, or risk for the individual who performs it. The secondary behavior therefore requires
group selection to evolve . . . (Wilson & Sober, 1998, pp. 143–144).
The broader point is that within-group selection interacts with between-group

selection. Some genetic or cultural traits may evolve because they weaken or
even eliminate within-group selection, thereby increasing the likelihood of group-

beneficial traits evolving.
INTEGRATING EVOLUTION AND METHODOLOGICAL

INDIVIDUALISM
A Combined Approach
As the foregoing discussion makes apparent, MI leaves a lot of things unspecified.

It does not say what people believe, what people want, or what means they will
use to achieve their goals. That may seem unsatisfying, but one principle cannot
be expected to do all the work. MI is one methodological principle; it is not the
only one.
In this section, I will outline a methodological approach that incorporates both
MI and evolutionary forces – including, at least potentially, group selection. This
approach is not original to this article; earlier statements (or at least hints) of this
approach can be found in the work of Langlois (1983, 1985, 1986), Koppl and
Whitman (2003), and most likely other works unknown to this author. The central
notion of the approach is that any (satisfying) social scientific model will have two
components: a story about what sort of individuals populate the model, and a story
about how individual actions and interactions lead to the overall system effects
predicted or explained by the model.
In the former story, evolutionary considerations take center stage, because
they shape what kind of assumptions the modeler is justified in making about
agents’ preferences, initial beliefs, and mental capacities. Are individuals selfish
or altruistic, or some combination of both? Are they rational, boundedly rational,
or rule-following?12 The answers to these questions depend on the selective
mechanisms – or to use Langlois’s (1986) language, “system constraints” – that
agents face or have faced historically.13 Some mechanisms have operated to

shape all human beings at a genetic level in the distant past, such as the need
to avoid predators, the need to procreate, the need to acquire sustenance, etc.
Other mechanisms are more recent and culturally specific. One example of a
culturally specific selective mechanism is the discipline of profit and loss in
capitalist economies: firms that persistently make losses tend to be liquidated.
As a result, existing firms are more likely to exhibit the characteristic of profit
maximization, or at least loss avoidance, than they would be in the absence of the
system constraint. (Notice that this selective constraint requires the existence of
other social phenomena, such as money and property, whose existence may also
demand social scientific explanation.)
The latter story, in which individual actions and interactions lead to the overall
system effects, is where MI enters the picture. By deciding what sort of individuals
populate his model, the modeler implicitly creates one or more ideal types
(Whitman & Koppl, 2003). The choices of agents in the model must follow from
their assumed characteristics, combined with whatever constraints are imposed by
the physical and institutional environment. Note again that the origin of specific
institutions may also require explanation, though not necessarily by the model in
question – one scientist cannot explain everything!
MI is a minimal condition, satisfied with relative ease. A theorist could
assume a variety of things about agents in his model, creating ideal types whose
behavior follows from whatever characteristics he has assumed. But there is
no guarantee that the predictions or explanations will be accurate, because the
initial assumptions may be wrong. Suppose, for instance, that a theorist begins by
assuming the existence of altruistic agents. If there are in fact not many altruists
in the relevant population, then the model’s predictions are likely to be inaccurate.
MI is thus necessary but not sufficient for the construction of a fully satisfying
model. In order to make a judgment about the likely accuracy of the model, we
would need to know something about the system constraint. Only this kind of
knowledge can indicate what sort of agents actually populate the social world to
be explained.
GS fits into this framework as one kind of system constraint that might be,
or historically might have been, effective in shaping the set of relevant agents.
The crucial question is whether a GS process constitutes a strong or weak system
constraint. The answer will depend on a variety of factors, including: (a) the relative
strength of the advantage conferred on individuals living in groups with members
who perform group-beneficial (altruistic) tasks; (b) the relative disadvantage
suffered by the individuals who perform those tasks; (c) the tendency, or not,
of the population to separate into smaller groups that periodically regroup; and (d)
the tendency, or not, of those people exhibiting group-beneficial characteristics to
form groups with each other.

Inasmuch as GS may operate counter to individual-level selection, these two may
be considered distinct system constraints that can wholly or partially offset each
other. The stronger is one relative to the other, the more effective will be the weeding
process, which allows the theorist to make more specific assumptions about the
agents in his model. For instance, very strong individual-level selection combined
with weak group-level selection would tend to weed out altruism, so in this context
the theorist would be justified in assuming selfish agents. If the strength of the
two forms of selection were reversed, the theorist would be justified in assuming
altruistic agents. However, it is conceivable that the two forces could balance
each other, leading to a polymorphic equilibrium wherein both types (selfish and
altruistic) are substantially represented in the population. Indeed, Wilson and Sober
(1994) suggest this is the most likely outcome: “We should expect a diversity of
motives in the human repertoire that is distributed both within and among individu-
als” (p. 601). In this context, the theorist is compelled to make weaker assumptions:
rather than assuming selfish or altruistic agents, he will have to allow for both.
Incorporating Institutions
The role of institutions in this structure deserves special attention. As noted above,
the institutional environment constrains the actions of individuals. Typically,
institutions involve patterns of actions or strategies adopted by current and previous
members of society, which impose themselves on new members of a society
by punishing (explicitly or implicitly) the adoption of alternative strategies. For
example, imagine a new member of a society in which people drive on the right
side of the road. Although it is within the new driver’s choice set to drive on the left,
he nonetheless will choose to drive on the right. Any individual who chose to drive
on the left would tend be eliminated from the society. And the fact that today’s new
drivers choose to drive right perpetuates the institution, so that tomorrow’s new
drivers will do so as well. Thus, the existing equilibrium creates a system constraint
that affects both who is in the society and how they will choose to behave.
The same may be said, albeit with less confidence, about norms that have evolved
to solve free-rider problems, such as respect for private property. Such norms may
induce individuals to behave in a manner that appears “altruistic” even though the
behavior is in fact motivated by self-interest. In this sense, we may say that altruism
is not just a property of individual preferences (although it may be that as well),
but is rather a property of a social equilibrium. The tendency to act in a group-
beneficial manner may be transmitted from generation to generation not merely by
genes or socialization, but by the incentive for new generations to conform their
behavior to that of overlapping previous generations.

For MI to be satisfied within a single model, it is enough to show that, given the
existing institutions, individuals will act in a way that sustains them (or, possibly,
undermines them). Someone analyzing the behavior of American drivers need not
explain why America is a society that drives on the right, only why new drivers’
behavior will sustain that norm. But the MI approach insists that the emergence
of an institution, such as right-side driving or enforcement of property rights in
land, can in principle be explained in terms of the behavior of individuals in
the past, as constrained by the physical and institutional environment that faced
them. Obviously, evolutionary forces can affect the evolution of institutions; GS,
in particular, would be felt in the differential contribution of groups with different
institutions to the overall human population.

In what follows, I will offer two examples of phenomena whose explanation
might be assisted by the use of GS thinking: profit-maximizing firms, and norms
for avoiding overuse of common resources. For both phenomena, I will offer
speculative arguments about how GS might have been involved in their emergence.
But I do not claim that these are phenomena are in fact attributable to GS;
first, because other explanations may prove superior, and second, because further
research would be required to show that the conditions for GS actually existed in
these cases. The point here is to offer simple GS accounts of economic phenomena
that are consistent with the strictures of MI. These two examples will also help
show how the interplay of within-group and between-group selection can influence
social outcomes.
Example: The Theory of the Firm
To what extent are we justified in assuming that firms act to maximize profits?
Firms are, after all, collectives. A firm that acts to maximize profit will clearly
prosper relative to other firms. But according to the critics of GS theory, that is
not sufficient, as any collective may be vulnerable to subversion from within. As
Zywicki (2000) argues,
Firms are mechanisms for carrying out team production. (Alchian & Demsetz, 1972). The firm
with the best survival prospects, therefore, is the one that maximize[s] the team’s joint product
while minimizing the costs of this team production. But the creation of this team production
structure opens the door for shirkers who seek to free ride on the efforts of the other members of
the team. The hierarchical structure of a firm may be understood as a mechanism for minimizing
these shirking costs so as to prevent free riding (p. 85).
The standard critique of GS is that within-group selection will always outstrip

between-group selection. If this critique were correct, then we might never be

justified in assuming a profit-maximizing firm. But there is no reason to assume,
a priori, that one form of selection is stronger than the other. If opportunities for
shirking and free-riding within the firm are relatively small, and the discipline
of profit and loss is relatively strong, then the outcome could indeed be profit-
maximizing firms. Firms that fail to control their internal incentive problems will
tend to be eliminated by competition with other firms; as a result, most employees
will find themselves working in firms that address incentive problems with some
measure of success.
It might be objected that firms typically implement specific structures designed

to deal with shirking and free-riding, including stock options for managers,
incentive pay, frequent monitoring, and so on. These structures make the
“altruistic” behavior not truly altruistic; e.g. the frequently monitored worker has
an incentive not to slack. Nonetheless, there is a still a free-rider problem, because
the enforcement of norms to solve a public good problem is itself a kind of public
good; this is the second-order public good problem. If there are monitors whose
job is to prevent shirking, why don’t the monitors themselves shirk?
The obvious reply is that the monitors themselves are given incentives by their
superiors, who are given incentives by their own superiors, all the way up the
chain to the top of the organization. At the top we will find residual claimants
– owners – who have a financial incentive to implement whatever cost-effective
structures are needed to overcome shirking problems. Two responses are relevant
here. First, the emergence of a residual claimant structure can itself be seen as
an evolved response to the public good problem that arises from team production
without a residual claimant. By reducing (but not eliminating!) the free-riding
problem, residual claimants reduced the strength of within-firm selection, thus
giving greater relative strength to between-firm selection. Firms with residual
claimants outperformed those without them. Without the selective pressure of
profit-and-loss, which operates at the firm level, residual claimants might not have
been necessary. Second, while sole proprietorships have unitary residual claimants,
what are we to make of corporations and partnerships? When responsibility is
distributed, the possibility of free riding arises again. Since there exist multiple
means of trying to control opportunistic behavior within the firm, and since
there is an external constraint (the discipline of profit and loss) that rewards
firms that control destructive internal behavior, we predict evolution of the most
effective devices for internal control of incentive problems. GS favors practices
that minimize the strength of within-group selection.
The firm is an instructive example, because it shows how emergent properties
of higher-order entities can be used as a kind of shorthand. While the firm is
not actually a single individual, for some purposes it may be usefully treated as an
individual acting with the particular goal of maximizing profit. Profit maximization
is an emergent property of a system in which groups organized for team production
compete in the context of strong profit-loss discipline. But it may not always
be acceptable to make the assumption of profit maximization (consider firms in
communist economies, where the discipline of profit and loss is absent). This is
the constraint imposed by MI: the behavior of individuals within a firm must build
up to the behavior of the firm itself. The willingness of economic theorists to
look within the firm to examine internal incentive problems is the hallmark of the
MI approach. But that is not sufficient for a compelling theory of the firm, as we
should also enquire into the evolution of the firm’s internal culture, which guides
the attitudes and behaviors of the individuals who comprise it. GS theory provides
a reason why firm-beneficial attitudes persist: those firms that failed to inculcate
such attitudes were less likely to survive.
The application of GS to the theory of the firm is far from unproblematic.
One condition for the new GS to work in biology is the differential reproduction
of groups; what is the cultural analog of reproduction in the case of firms? Since
firms are potentially infinitely lived, “reproduction” might have to be interpreted as
persistence. Or reproduction might take the form of firms that imitate the structures
and practices of existing firms. I will not attempt to answer here whether these
conditions are sufficiently similar to the conditions necessary for GS in biological
evolution. Profit-maximizing firms might be seen as the outcome of selection
among groups, but other explanations might perform better. For example, residual
claimants could be the accidental by-product of private property, an institution that
evolved earlier and possibly for different reasons.
Example: The Tragedy of the Commons
The tragedy of the commons (Hardin, 1968) is a classic case in which individually
rational behavior leads to an undesirable outcome for the group (or more
precisely, for each member of the group). The over-exploitation of common
assets is not just a human phenomenon, as animal populations may overburden
their local environments by excessive reproduction. Indeed, this was precisely
the problem addressed by V. C. Wynne-Edwards (1962) in his early version of
GS theory. Wynne-Edwards did not, like Wilson and Sober, ground his theory in
an understanding of the interplay between within- and between-group selection.
Nonetheless, the commons tragedy provides fertile ground for the application
of the new GS theory. To what extent can populations of animals or humans
be expected to display group-beneficial strategies in their use of assets? The
answer to this question bears on the validity of, for instance, Harold Demsetz’s
(1967) theory of property rights, which asserts that “property rights arise when it
becomes economic for those affected by externalities to internalize benefits and
costs” (p. 97).
In the human context, the most commonly recognized solution to the commons
problem is to privatize the asset in question (say, land), so that individual members
of the population have exclusive stewardship over smaller plots. This solution is
not as simple as it might appear, as it typically requires some kind of enforcement
apparatus. Pure self-enforcement might be sufficient, and if it were, privatization
would occur on its own. But if some kind of social enforcement is necessary,
then a second-order public good problem arises: why should individual members
contribute to the costly enforcement process by refraining from good opportunities
to steal, refusing to trade with known property rights violators, and so on?
In this context, an opponent of GS theory might conclude that privatization
will not occur. But that conclusion would only follow from a situation in which
between-group selection is too weak relative to within-group selection. It might
turn out, on the contrary, that the individual incentive to deviate from property-
enforcing norms is relatively small compared to the gains from cooperation. If
so, then societies composed mostly of individuals with an inherited or ingrained
inclination to follow those norms will prosper relative to other societies. They
would contribute a disproportionate number of members to future societies, thereby
spreading the property-enforcing norms.
Of course, privatization is not the only potential solution to the commons
problem. Another possibility is the enforcement of norms that directly limit
individuals’ consumption of the common resource. Such norms are common
in primitive societies. Their enforcement, just like the enforcement of property-
norms, may be subject to a second-order public good problem; and yet they persist.
One possible explanation is that the enforcement of group-beneficial norms is eased
by the inherent tendency of individuals to detect and punish people perceived as
“cheaters” – a tendency reinforced by a GS process.
For an analysis of the tragedy of the commons to satisfy MI, the social outcome
– whether overuse or efficient use of assets – must result from the attitudes of
individuals and the constraints they face. If individuals are entirely selfish, then
we do not predict a solution to the commons problem except in the presence of
small numbers and very low monitoring costs. If individuals are entirely altruistic,
then we predict commons problems to be solved with little effort. If individuals
have both selfish and altruistic attitudes, then we predict commons problems
to be solved (sometimes) through a combination of monitoring and individual
forbearance. But which of these assumptions about individual attitudes is most
appropriate? In other words, what sort of agents should populate our economic
models? Here, evolutionary theory, and GS theory specifically, can be of assistance.
The stronger was between-group selection in the shaping of human psychology, the
more justified we are in assuming some degree of altruism in the current population
of humans.
As in the case of firms, the application of GS to the tragedy of the commons is
not unproblematic. Among other things, there is the historical question of whether
private property did, in fact, emerge at a time when human populations consisted
of small competing groups, and whether those groups exchanged members in a
fashion approximating the fission-and-reconcentration pattern required for group
selection. In addition, even if norms for the avoidance of the tragedy of the
commons (such as private property) could emerge from a GS process, other

processes could possibly explain the same transformation. Such processes could
include deliberate or non-spontaneous decision-making, such as the choice by

a legislature to establish private property rights in assets previously owned in
common. Even if GS explains the emergence of private property in some cases, it
does not explain it in all cases.
CONCLUSIONS
The notion that MI and GS conflict derives from an overextension of both
principles. MI is taken to imply self-interest, rationality, spontaneous order, and
denial of the existence of groups. GS is taken to imply that groups will always
exhibit group-beneficial characteristics. Both characterizations are mistaken. MI
is a relatively minimal requirement that says group properties must be built up
from individual actions and interactions, and GS (at least in its modern form)
says that groups will exhibit group-beneficial characteristics only under specific
circumstances that do not always obtain.
In this article, I have offered speculative discussions about incentive problems
in the context of firms and common-pool resources. I have suggested, in each case,
that GS could be invoked as a (partial) explanation of how such problems are
solved. But the main point of the article – that there is no contradiction between
GS and MI – does not depend on the correctness of these examples.
Once MI and GS are properly defined and constrained, the appearance of
conflict disappears. Both can be incorporated into a single methodological
approach based on individual choice in the presence of system constraints. MI
dictates that individual choices lead to social outcomes; GS is one force (among
many) that determines what sort of individuals are present in the system. To
accept this position, one need not agree that GS is a significant force in biological
or cultural evolution. That is ultimately an empirical question, one that should
not be foreclosed by the rejection of GS theory on erroneous methodological
grounds.
NOTES
1. Later, Hayek (1967a, b) adopts a more nuanced approach, wherein the dividing line is
not between natural and social sciences, but between simple and complex phenomena. As
a result, even some natural sciences might have to adopt a form of methodological dualism,
because their basic elements cannot easily be reduced to more constituent parts. “Where
we have to deal with a complex situation in which observation discloses only very limited
regularities, be it in the ‘applied’ branches of physics or in biology or in the social sciences,
we usually ask to what extent our existing knowledge of the forces at work, or of the
properties of some of the elements of the complex, may account for what we observe. We
endeavor to find out whether this may be derived by deduction from what we know about
the behavior under simpler conditions of some of the factors involved” (Hayek, 1967a, b,
p. 10). However, there are other features of the social sciences that distinguish them from
all natural sciences: “Another and perhaps more important peculiarity of the social sciences
is due to the fact that here the recognition of the different kinds of facts rests largely on
a similarity between the observer and the observed persons” (Hayek, 1967a, b, p. 18,
Note 15).
2. To be more precise, MI says social phenomena should be reduced to individual
psychology (and environmental factors), while MD says no further reduction is required.
Elsewhere (Hayek, 1952), Hayek argues that it may be impossible to reduce human action
to something even more basic. But this position is distinct from both MI and methodological
dualism, neither of which takes any position on this matter.
3. Langlois (1983) refers to the position that emergent properties of social systems do
not exist as “naı̈ve methodological individualism,” to distinguish it from the version of MI
which seeks the explanation of emergent properties. “Far from denying the importance of
emergent phenomena, the economist and philosopher Friedrich A. Hayek . . . reminds us
that the entire objective of the social sciences is to explain how the behavior of individuals
leads to orderly patterns and institutions that none had consciously planned – to explain, in
other words, the emergent results of individual action” (Langlois, 1983, p. 583). However,
as I argue below, MI does not require the absence of conscious planning.
4. Hayek (1952) unfortunately does not make this point explicit. It is clear, however, than
many of the “wholes” that Hayek says require compositive explanations are institutions that
individuals may take as given, such as money and language. “It makes no difference for our
present purpose whether the process [of emergence of an order] extends over a long period
of time, as it does in such cases as the evolution of money or the formation of language, or
whether it is a process which is constantly repeated anew, as in the case of the formation of
prices or the direction of production under competition” (p. 71).
5. See, generally, Rousseau (1762).
6. See Arrow (1951) and subsequent literature.
7. But this is not always true even in the natural sciences. To the extent that they deal
with complex emergent phenomena that cannot be observed directly, some natural sciences
may face a problem similar to that of the social sciences. See Notes 1 and 2.
8. Not to imply that altruism and selfishness are only properties of preferences. They
might also be properties of institutions or social equilibria. See the final section on
“Integrating Evolution and Methodological Individualism.”
9. In fact, Wynne-Edwards illustrated the basic problem with the case of overfishing, an
example also used by Hardin (1968) to illustrate the tragedy of the commons.
10. See, for instance, Zhang (nd).
11. “As we have seen, however, kin selection is a special case of a more general theory –
a point that Hamilton (1975, 1987) was among the first to appreciate. In his own words, ‘it
obviously makes no difference if altruists settle with altruists because they are related . . . or
because they recognize fellow altruists as such, or settle together because of some pleiotropic
effect of the gene on habitat preference’ ” (Wilson & Sober, 1998, p. 134).
12. This sentence is not meant to indicate that rule-following is necessarily irrational,
only that it might be.
13. Langlois (1985) also refers to the notion of a “compositional principle,” such as
selection or an invisible hand principle, that leads from individual actions to social results.
Langlois’s approach is similar, though not identical, to that taken here.
ACKNOWLEDGMENTS
I wish to thank Roger Koppl, an anonymous referee, and all of the commentators
for their helpful analysis and suggestions.
REFERENCES
Alchian, A. A., & Demsetz, H. (1972). Production, information costs, and economic organization.
American Economic Review, 62, 777–795.
Arrow, K. J. (1951). Social choice and individual values. New York: Wiley.
Black, D. (1958). The theory of committees and elections. Cambridge: Cambridge University Press.
Boehm, S. (1989). Hayek on knowledge, equilibrium and prices: Context and impact.
Wirtschaftspolitische Blatter, 36, 201–213.
Demsetz, H. (1967). Toward a theory of property rights. American Economic Review, 57, 13–27.
Reprinted (1998). In: A. W. Katz (Ed.), Foundations of the Economic Approach to Law
(pp. 93–102). New York: Foundation Press.
Ellickson, R. C. (1991). Order without law: How neighbors settle disputes. Cambridge, MA: Harvard
University Press.
Elster, J. (1983). Explaining technical change: A case study in the philosophy of science. Cambridge:
Hardin, G. (1968). The tragedy of the commons. Science, 162, 1243–1248.
Hayek, F. A. (1952). The sensory order. Chicago: University of Chicago Press.
Hayek, F. A. (1967a). Studies in philosophy, politics, and economics. Chicago: University of Chicago
Press.
Hayek, F. A. (1967b). Degrees of explanation. In: F. A. Hayek (Ed.), Studies in Philosophy, Politics
and Economics. Chicago: University of Chicago Press.
Hayek, F. A. (1979). Law, legislation and liberty, volume 3: The political order of a free people.

Hayek, F. A. (1991). The fatal conceit: The errors of socialism. Chicago: University of Chicago Press.
Hodgson, G. M. (1991). Hayek’s theory of cultural evolution: An evaluation in light of Vanberg’s
critique. Economics and Philosophy, 7, 67–82.
Koppl, R., & Whitman, D. G. (2004). Rational-choice hermeneutics. Journal of Economic Behavior
and Organization (forthcoming).
Langlois, R. (1983). Systems theory, knowledge, and the social sciences. In: F. Machlup & U. Mansfield
(Eds), The Study of Information: Interdisciplinary Messages. New York: Wiley.
Langlois, R. (1985). Knowledge and rationality in the Austrian school: An analytical survey. Eastern
Economic Journal, 9, 309–330.
Langlois, R. (1986). The new institutionalist economics: An introductory essay. In: R. Langlois (Ed.),
Economics as a Process: Essays in the New Institutionalist Economics. Cambridge: Cambridge
University Press.
Lucas, B. (2002). Which side of the road do they drive on? Web page accessed 8 April 2003.
http://www.travel-library.com/general/driving/drive which side.html.
Mises, L. von (1981 [1957]). Theory and history: An interpretation of social and economic evolution.
Auburn University, Alabama: Ludwig von Mises Institute.
Smith, E. A. (1994). Semantics, theory, and methodological individualism in the group-selection
controversy. Behavioral and Brain Sciences, 17, 636–637.
Smith, E. A., & Winterhalder, B. (1992). Natural selection and decision-making: Some fundamental
principles. In: E. A. Smith & B. Winterhalder (Eds), Evolutionary Ecology and Human
Behavior. Aldine de Gruyter.
Vanberg, V. (1986). Spontaneous market order and social rules: A critical examination of F. A. Hayek’s
theory of cultural evolution. Economics and Philosophy, 2, 75–100.
Williams, G. C. (1966). Adaptation and natural selection: A critique of some current evolutionary
thought. Princeton, NJ: Princeton University Press.
Wilson, D. S., & Sober, E. (1998). Unto others: The evolution and psychology of unselfish behavior.
Wright, E., Levine, A., & Sober, E. (1992). Marxism and methodological individualism. In:
Reconstructing Marxism – Essays on Explanation and the Theory of History (pp. 107–128).
London: Verso Books.
Wynne-Edwards, V. C. (1962). Animal dispersion in relation to social behavior. Edinburgh: Oliver &
Boyd.
Zhang, J. (nd). What is Simpson’s Paradox? Web page, accessed 8 April 2003. http://www.ma.iup.edu/
∼zhang/simpson.html.
Zywicki, T. (2000). Was Hayek right about group selection after all? Review essay of Unto others:
The evolution and psychology of unselfish behavior, by Elliot Sober and David Sloan Wilson.
Review of Austrian Economics, 13, 81–95.
ON GROUP SELECTION
AND METHODOLOGICAL
INDIVIDUALISM – A REPLY TO
DOUGLAS GLEN WHITMAN
Elliott Sober and David Sloan Wilson
We admire the clarity of Whitman’s paper and agree with its principal thesis –
that hypotheses of group selection, when understood within the framework of
multi-level selection theory that we presented in Unto Others, are compatible with
methodological individualism (MI), when this is understood in terms of Whitman’s
exposition of the views of Friedrich Hayek. Below we briefly describe the reason
for this compatibility. We then turn our attention to the thesis of MI itself, breaking
it into pieces, commenting on the plausibility of those separate elements, and then
referring MI to some recent philosophical work about reductionism.
METHODOLOGICAL INDIVIDUALISM AND GROUP

SELECTION ARE COMPATIBLE
According to Whitman, Hayek’s conception of MI pertains just to the relationship
of individual psychology and the social sciences, and is neutral on broader
questions about reductionism in other scientific domains. Since hypotheses of

group selection frequently concern organisms that are taken to be mindless (e.g.
viruses and social insects), it is clear that they do not come into contact with MI

ISSN: 1529-2134/doi:10.1016/S1529-2134(04)07011-5
251
252 ELLIOTT SOBER AND DAVID SLOAN WILSON
thus construed. And even when group selection hypotheses make claims about
human evolution, as do the hypotheses we discuss in Chapters 4 and 5 of Unto
Others, there is, once again, a relationship of mutual irrelevance. The reason is that
MI addresses what biologists call the question of proximate mechanism, whereas
hypotheses about natural selection are part of the project of ultimate explanation
(Mayr, 1961). An example used in Unto Others helps to illustrate this distinction.
If one asks why sunflowers turn towards the sun, there are two ways in which this
question might be understood. One might wish to understand how the machinery
inside of each plant causes the plant to exhibit phototropism. Or one might want
to understand the evolutionary processes that caused this behavior to evolve. Both
types of understanding are important, and there is no conflict between them. By the
same token, when a human society exhibits some property – e.g. the type of egali-
tarianism among adult males that Boehm (1999) argues is characteristic of nomadic
hunter-gatherers – we might seek both a proximate and an ultimate explanation of
that arrangement. MI constrains the former problem; it asserts that a group’s having
that property must be understood in terms of the beliefs and desires of the individ-
uals in the group (with properties of the physical environment brought in where
necessary). But even if the question of proximate mechanism gets answered in the
way that MI insists, the question is left open as to whether the group phenotype is
the result of natural selection, and if it is, whether group selection was involved. MI
says nothing about the form that an evolutionary explanation must take; it concerns
proximate explanation only.
The first half of Unto Others is concerned to develop and defend hypotheses
of group selection; the second half addresses the debate in psychology between
psychological egoism (the view that all of our ultimate motives are egoistic)
and motivational pluralism (the view that human beings have both egoistic
and altruistic ultimate motives). Although this part of the book addresses a
question about proximate mechanism, once again MI, as Whitman understands
it, has nothing to say about the question discussed. The reason is that MI does
not commit itself on what sorts of ultimate desires people have – they can be
purely egoistic, or purely altruistic, or a mixture of both. For Whitman, MI is
concerned just to make sure that group properties are built up out of the beliefs
and desires of individual human beings, whatever those beliefs and desires turn
out to be.
Before turning to the narrower thesis of MI that Whitman discusses, we
want to explain why we have understood MI in a wider sense, one that has
implications concerning the nature of evolutionary processes. When opponents
of group selection hypotheses describe what sort of selection hypothesis they
view as legitimate, they usually say that they are happy to invoke the process
of individual selection. In one of the central texts of the group selection debate
On Group Selection and Methodological Individualism 253
that raged in the 1960s, Williams (1966) advances a methodological principle

– when an observation is consistent with a group selection hypothesis and also
with a hypothesis of purely individual selection, we should prefer the latter. This
is individualism of a methodological variety, hence our inclination to view the
group selection controversy as pertaining to MI in at least one legitimate sense of
that phrase.
The attack on group selection during the 1960s was mainly given over to
methodological and conceptual arguments against group selection that, upon
reflection, turned out to be entirely irrelevant. One of them is the subversion-
from-within argument that Whitman describes. If a group somehow were to
consist entirely of altruistic individuals, a selfish mutant or migrant would sooner
or later appear, and since that selfish individual would out-compete the resident
altruists, selfishness would gradually sweep through the population and altruism
would go extinct. The conclusion is supposed to be that altruism is inherently
unstable. This argument was invoked repeatedly, as if it were a mantra that wards
off an evil spirit; what purveyors of this argument neglected to mention is the
significance of the assumption that the imagined process takes place within the
confines of a single group. If there is no variation among groups, there is no
group selection. Although it is true that in the absence of group selection, pure
individual selection will drive altruism to extinction, the subversion-from-within
“argument” says nothing whatever about what group selection might be able
to accomplish. One cannot show that a process is powerless to bring about a
given result by imaging a situation in which the process does not occur. Other
arguments from this period were no better – e.g. “group selection can’t be
right because it is the gene, not the individual or the group, that is the unit
of heredity.”
This deplorable battery of a prioristic arguments coexisted along side a better
way of thinking about group selection, one that held that the existence and power
of group selection as an empirical question to be decided on a trait-by-trait and
lineage-by-lineage basis. Here we can mention Williams’ (1966) insight that a
population’s sex ratio provides a test case for group versus individual selection,
and Hamilton’s (1967) development of a model of sex ratio evolution in which

both group and individual selection are represented. We viewed this empirical
line of attack on the problem as the light and the way – as replacing empty and
irrelevant methodological considerations with something substantive. This is
why Sober (1981) argued that the group selection controversy provides a model
for how the debate about MI in the social sciences might be transformed into
something empirical. If all sides agree that properties of wholes are built up from
properties of parts, empirical questions nonetheless remain about the levels at
which different causal processes occur.
WHITMAN’S VERSION OF METHODOLOGICAL

INDIVIDUALISM, AND ITS PLAUSIBILITY
We take it that Whitman agrees that MI can be viewed as a special instance

of reductionism. This means that if there are general arguments against reduc-
tionism – “general” in the sense that they are meant to apply no matter which
specific sciences are considered – that these will bear on MI. Defenders of
MI therefore need to attend to such arguments (one of which we will describe
shortly), even if, for example, they are neutral on the question of whether biology
reduces to physics.
We discern three different theses at work in Whitman’s exposition of MI – there
is a thesis about constitution, about evidence, and about explanation.
Constitution: Broadly speaking, MI holds that human social wholes are “built
up” from individual human beings and their physical environments. We take
this to mean that if the individuals in one group have the same distribution of
properties (including relational properties) as the individuals in another, and if the
individuals in the first group live in a physical environment that is qualitatively
identical with the environment inhabited by the individuals in the second group,
then the properties of the two groups must also be the same. “No social difference
without an individual difference” might be the slogan for MI, broadly construed,
and we have no reason to disagree with it.
However, we sense that Whitman has something more specific in mind here.
Notice that the broad construal of MI does not specify which properties of
individuals are the ones that determine the properties of social wholes. However,
Whitman says in one place that it is the “choices and behaviors” of individuals
and their physical environment that do the building up. Whitman’s discussion
in other passages of the subjective character of social experience (as mediated
by “constitutive ideas”) leads us to suspect that Whitman is thinking that it is
the beliefs, desires, emotions, and sensations that individuals experience that
(along with the physical environment) suffice to determine all properties of social
wholes (or at least the ones that are of interest to the social sciences). This more
specific thesis strikes us as less than obvious. Why should we be certain that
there are no other properties of individuals that need to be taken into account
if the determination thesis is to be true? Even if common sense psychology
is true, why think that it is complete? If psychological theories undreamt of
in common sense psychology are needed to explain individual behavior, why
shouldn’t properties of social wholes depend on the properties recognized in those
new theories?
Evidence: Whitman claims that the social and the natural sciences differ in
an important respect. “In the natural sciences, the existence of many emergent
properties can be verified independently of their explanation,” but this is not

the case in the social sciences. For Whitman, social science generalizations are
plausible only to the degree that we can explain them in terms of individual
psychology; in contrast, the generalizations of individual psychology can be
plausible without our already having worked out the neuroscience details of
how brains give rise to minds. The reason for this asymmetry, Whitman says, is
that “we cannot, in fact, observe wholes directly, we infer their existence [only]
through the apprehension of particular events and phenomena.”
We find this part of MI implausible. First, we believe that the distinction
between having evidence for a generalization and having an explanation for why
the generalization is true is just as important for the social sciences as it is for the
natural sciences. Before Newton, it was well-known that the phases of the moon
are correlated with the tides; the evidence for this correlation was overwhelming,
but it wasn’t explained until Newton developed his theory of gravity. A similar
point seems to attach to Durkheim’s description of the waxing and waning of
suicide rates. Why couldn’t data on suicide rates in one or more societies convince
us that there is a general pattern here, even if we have no very plausible explanation
of why the pattern obtains? Another example may be found in the near-universality
of incest taboos – we know that the pattern obtains cross-culturally, even if we
are very uncertain as to the explanation of such taboos from the point of view of
individual psychology. Whitman argues for the evidential thesis that he takes to
be part of MI by claiming that we cannot observe social wholes directly, but that
point doesn’t seem to justify the evidential thesis – even if the tides and the moon
were inferred entities rather than observables, we still could have evidence for
the generalization without being able to explain its truth. And the fact that suicide
rates are statistically inferred from samples doesn’t prevent us from having a lot
of confidence in claims about those rates. The same holds for a society’s having
an “incest taboo” – this social fact is somehow inferred from the laws, religious
teachings, and moral precepts that exist in a society. We can infer the existence of
taboos without having a full explanation of why they exist.
Explanation: Whitman says that “[w]hile the behavior of individuals is affected
and constrained by the institutional environment . . ., MI posits that the institutional

environment itself can in principle be fully explained by the behavior of past and
present individuals” (again as affected by the physical environment). Whitman
does not explicitly endorse the following assertion, but we suspect that it is part of
the picture he has in mind – that facts about social wholes cannot in principle fully
explain the behavior of individuals. Adding this further claim turns what Whitman
says into a thesis of explanatory asymmetry.
Whitman’s MI is not committed to the claim that the properties of social wholes
are causally inert. The inflation rate in a society at a given time, its class structure,
etc. – all of these can have effects on individuals and also on later temporal slices of
the social wholes themselves. And of course, the beliefs and desires of individuals
at a time have effects on the properties of social wholes and on later time slices
of individuals. However, there is a determination relation running from parts to
wholes that exist at the same time, but not from wholes to part, and it is not
puzzling why this should be the case. The reason is that the relationship between
properties of social wholes and properties of individuals is typically one-many.
For example, if we specify whether each of the individuals in a group is an altruist,
that set of individual facts determines the frequency of altruism in the group, but
the group frequency (unless it is 0% or 100%) doesn’t settle what trait each of the
individuals has. In the language of recent work in philosophy, properties of groups
are multiply realizable – when a group has a given property, this can be realized
by many different assignments of properties to the individuals in the group.
So we agree with Whitman’s claim about in-principle complete explanation –
that this flows from parts to whole, but not in the other direction. We note, however,
that the explanations that scientists construct in practice are almost never complete,
especially not in the social sciences. This means that the asymmetry thesis we are
attributing to Whitman does not show that an individual-level explanation will
always be better than a group-level explanation. Both will be incomplete, so on
what grounds is the former always better?
There is an influential anti-reductionist line of thinking in recent philosophy of
mind and philosophy of science that bears on this issue of explanatory asymmetry.
It began as a claim about explanation in the social sciences (Garfinkel, 1981) and
then got generalized into a thesis about parts and wholes, no matter what the levels
of organization are at which those parts and wholes are located. Putnam (1975)
illustrates the idea at work here by way of a simple example. Suppose a wooden
board has two holes in it. One is circular and has a 1 inch diameter; the other is
square and is 1 inch on a side. A cubical peg that is 15/16ths of an inch on each
edge fits through the square hole, but not the circular one. What is the explanation?
Putnam says that the explanation is provided by the macro-properties just cited of
the peg and the holes. He denies that the micro-properties of molecules or atoms or
particles in the peg and the piece of wood explain this fact. The micro-description
that permits one to deduce the proposition that needs to be explained is long and
complicated and brings in a welter of irrelevant detail. To explain why the peg
goes through one hole but not the other, it doesn’t matter what micro-properties
the molecules have, as long as the peg and board have the macro-properties I
mentioned. According to Putnam, the macro-properties are explanatory, while the
micro-properties that underlie those macro-properties are not. He concludes from
this that reductionism is false.
Notice that the macro- and micro-properties of the peg-board system exhibit
the same one-many relationship that social and individual facts exhibit. Applied
to explanations in the social sciences, Putnam’s claim would be that group-level

explanations of what happens to a social whole are superior to individual-level
explanations of the same phenomenon. An economist might seek to explain why
the inflation rate goes up in a given time period by describing the money supply,
the unemployment rate, and other properties of the economy. Would it be bet-
ter to explain this event by describing the choices and behaviors of each of the
individuals in the economy? Putnam says no because he values generality; the
micro-story about Tom, Dick and Harry, probably won’t apply to other episodes
of inflation in this and other economies, whereas the macro-story may succeed
in doing so.
Although Putnam claims that the macro-story is explanatory and the micro-story
is not explanatory at all, we are inclined to take a more pluralistic view (Sober,
1999). We think that micro-details are often explanatorily relevant; though they
may not be of interest in some lines of inquiry, they may be so in others. Although
generality is often a virtue of explanations, it is not the only one; detail is often
valued as well. The point is that these virtues are in conflict, which means that
there are various good explanations of a given phenomenon and that different good
explanations will exemplify different suites of explanatory virtues. Reductionism
– the claim that explanation in terms of the properties of parts is always superior
to explanation in terms of the properties of wholes – is mistaken, but so is the
opposite view – the claim that explanation in terms of wholes is always superior
to explanation in terms of parts. The point of rejecting reductionism is not to
substitute one monistic thesis for another, but to embrace a pluralistic account of the
explanatory virtues.
The idea that there is a conflict between explanatory generality and explana-
tory detail connects with our earlier discussion of the proximate/ultimate dis-
tinction. Cohan (1984) performed an artificial selection experiment in which he
selected for longer wings in ten isolated laboratory populations from the same
species of Drosophila. There was a response to selection in each population,
with wings at the end of the experiment being longer than they were at the start.
However, the genetic and developmental mechanisms that allowed the flies to
exhibit longer wings varied from population to population. When we ask why
flies at the end of the experiment have longer wings than flies at the beginning,
we can provide ten (or more) distinct proximate explanations, or a single story
at the level of ultimate explanation. Ultimate explanation thus permits a kind
of unification and generality that proximate explanations often fail to provide
(Sober, 1994, Section 7.3).
THE RATIONAL KERNEL OF METHODOLOGICAL

INDIVIDUALISM
In spite of our reservations about some of the details that comprise Whitman’s
conception of MI, we are sympathetic with his desire to prevent “the accidental
intrusion of unproven hypotheses into one’s initial assumptions.” To speak about
the inflation rate, the ideology, the money supply, or the class-structure in a given
society can be problematic if it is not clear what would count as evidence that
would allow these claims to be tested. Explicitly reducing such talk to a detailed de-
scription of the beliefs and preferences of individuals would suffice to render these
claims about social wholes transparent, if attributions of beliefs and preferences
to individuals were not subject to similar problems. However, even supposing
that attributions of beliefs and desires to individuals is unproblematic, we doubt
that this type of reductionist account is always necessary. Surely economists
have methods for measuring various properties of economies (i.e. estimating their
values, subject, as is always the case in science, to error) that do not require them
to assemble descriptions of the beliefs, desires, and actions of each individual in
the population. In just the same way, physicists can measure the temperature in a
chamber of gas without needing to measure the kinetic energies of the molecules
that comprise the gas. Reductionism is one mode of clarification, but it is not the
only one.
Furthermore, it is important to recognize that the beliefs and desires that other
people have are not things that we directly observe; they are inferred from the
behaviors we observe. Even if each person has introspective access to his or her
own beliefs and preferences, it would be wrong to view the beliefs and desires
that the agents in a society have as something that each person knows directly.
Psychological claims that are made in the 3rd person are “theoretical” in the
sense that we evaluate them by seeing how well they predict and explain what we
observe.
In summary, we agree that it is desirable to identify and avoid untested or
untestable assumptions, but we don’t see that this problem always arises in
connection with hypotheses about social wholes, or that the problem attaches
only to such hypotheses. And when it does arise, the solution need not be
reductionistic.
ACKNOWLEDGMENTS
We thank Daniel Hausman and an anonymous referee for useful suggestions.
REFERENCES
Boehm, C. (1999). Hierarchy in the forest – the evolution of egalitarian behavior. Cambridge: Harvard
University Press.
Cohan, F. (1984). Can uniform selection retard genetic divergence between isolated conspecific
populations? Evolution, 38, 495–504.
Garfinkel, A. (1981). Forms of explanation – rethinking the questions of social theory. New Haven:
Yale University Press.
Hamilton, W. (1967). Extraordinary sex ratios. Nature, 156, 477–488.
Mayr, E. (1961). Cause and effect in biology. Science, 134, 1501–1506.
Putnam, H. (1975). Philosophy and our mental life. In: Mind, Language, and Reality. New York:
Sober, E. (1981). Holism, individualism, and the units of selection. In: P. Asquith & R. Giere (Eds),
Proceedings of the Biennial Meetings of the Philosophy of Science Association (Vol. 2,
pp. 93–101). Reprinted. In: Hodgson (Ed.), Economics and Biology (1995, pp. 399–427).
Edward Elgar Publishing.
Sober, E. (1994). Philosophy of biology. Boulder: Westview Press.
Sober, E. (1999). The multiple realizability argument against reductionism. Philosophy of Science,
66, 542–564.
Williams, G. (1966). Adaptation and natural selection. Princeton: Princeton University Press.
COMMENT ON “GROUP SELECTION

AND METHODOLOGICAL
INDIVIDUALISM: COMPATIBLE AND
COMPLEMENTARY” BY DOUGLAS
GLEN WHITMAN
Richard N. Langlois
My compliments to Glen Whitman on a carefully argued paper that makes an

important point: methodological individualism is not what you think it is; and,
contrary to what many have argued, methodological individualism is not in conflict
with the notion of group selection in the theory of cultural evolution. Of course,
part of the reason I like Whitman’s paper so much is that, as some of his footnotes
hint, I have been saying many of the same things for a long time1 (Langlois, 1983,
1985, 1986).
My view – and, I think, Whitman’s view – on methodological individualism is
more-or-less the following. There are three possible claims.
(1) We can construct wholes only from the parts, that is, we cannot use any
information to construct wholes not contained within the parts. Social wholes
are “nothing other” than the behavior of individuals.
(2) We cannot analyze wholes without careful attention to the parts, which
exist separately from the wholes. But to understand wholes we must also
add information – like compositional principles, filtering mechanisms,
institutions, etc. – that are not logically derived from a consideration of the

ISSN: 1529-2134/doi:10.1016/S1529-2134(04)07012-7
261
262 RICHARD N. LANGLOIS
behavior of the parts. This includes the possibility that the interactions among
the parts may lead to “emergent” phenomena, at least so long as we do not
get sloppy and let the idea of emergence become equivalent to Claim 3.
(3) We can and should study wholes directly. Wholes have a “life of their own”
independent of the parts, and (in some formulations) the parts don’t even
exist except in the context of wholes.
Claim 1 is naı̈ve individualism; claim 3 is naı̈ve holism. Claim 2 is obviously

right. I want to call claim 2 sophisticated methodological individualism, or
simply methodological individualism. Others, notably Geoff Hodgson (1999,
pp. 132–133), want to see claim 2 as a kind of methodological holism.2 The
reason I think the label individualism fits better is that the real dividing line lies
between claims 2 and 3. Both claims 1 and 3 are mistakes. But implementing
claim 1 is not actually possible, so it is less likely to lead to error. All models in
social science necessarily adduce some elements that don’t flow directly from the
behavior of individuals: even a simple model of supply and demand requires the
proposition that a market demand curve is the sum of individual demand curves,
and the property of addition is a (very simple) system constraint not logically
contained in individual demand functions or other properties of the agents.
By contrast, implementing claim 3 is all too possible, as the history of social
science testifies.
It follows immediately that there is no conflict between group selection and
methodological individualism understood (correctly) as claim 2. In both the
original Wynne-Edwards version and the newer Wilson and Sober version, group
selection is about how individual behavior interacts, not about some “group”
characteristics independent of individuals.
Precisely because Whitman makes his case so well, I want to deviate from the
text a bit and to devote the remainder of my comment not to Whitman’s thesis
but to the idea of group selection itself. My basic point is that Hayek’s account of
group selection is in fact strikingly different from the notion now current.
What I find striking about the modern discussion of group selection is how
“neoclassical” it is. By this I mean that the discussion seems to be focused

entirely on issues of incentives and equilibrium – rather than on issues of learning,
experiment, and change. This is no doubt appropriate to some extent and in
some contexts. Perhaps it is the case that the principal evolutionary issue facing
hunter-gather societies was the need to solve common pool problems and to avoid
free riding among group members. Thus questions of altruism (appropriately
defined) versus narrow self interest (appropriately defined) naturally come to the
fore. Those groups whose norms, institutions, rules, or behavior patterns solve the
problem of team production well will tend to thrive (in an appropriately defined
Comment on “Group Selection and Methodological Individualism” 263
sense) while those who fail to solve the problem (as) well will tend to decline (in
an appropriately defined sense).
This is all well and good. But it is not principally what Hayek had in mind
when he endorsed the (Wynne-Edwards) version of group selection.3 Rather than
seeing groups as systems of team production that constrain wayward incentives,
he sees groups more generally as systems of rules of conduct. Humans are not
merely maximizers in the face of incentives; they are followers of rules more
generally – rules that are often tacit and inarticulate. Some (but probably not most)
rules may indeed have to do with solving free-rider problems. But, for Hayek,
individuals follow rules in the sense of what Nelson and Winter call routines, not
rules in the sense of game-theoretic strategies.4 Quite typically, people follow rules
unconsciously, unaware of why the rules lead to the results they do. For one thing,
the world is so complex that it is difficult for agents most of the time even to know
what is in their interests.
Moreover, individual rules cannot often be easily disentangled from the overall
system of rules the group follows; an individual rule is effective only in the context
of other rules. This is why group selection is important to Hayek: you have to
choose the whole package, not individual rules in isolation. Since, as Hayek insists,
cultural evolution is Lamarckian, groups do not expand only through relatively
higher rates of procreation.5 More importantly, groups – that is, coherent systems
of rules of conduct – grow through imitation. Sometimes imitation is attendant
on immigration to a more successful group: it is a familiar story that immigrants,
who are attracted to the economic benefits of better-functioning systems, often
struggle as they are forced to alter a whole package of behavior patterns in order
to prosper from their new surroundings (Choi, 1993; Hayek, 1988, pp. 129–130).
But sometimes imitation involves a conscious attempt to copy patterns of behavior
observed elsewhere. Eastern Europe since 1989 is an example.6
The upshot is that, for the most part, cultural evolution does not suffer fun-
damentally from the problem of a conflict between the interest of the individual
and that of the group. Or rather, if it does, it is a conflict exactly opposite to that
envisioned by biologists. For Hayek, the problem is too much group solidarity,
not too little. Hayek agrees that the hunter-gather lifestyle was indeed one whose
success depended on solving certain public goods or free-rider problems, and
perhaps on some form of altruism. But economic growth entailed the evolution of
quite different systems of rules of conduct. And to get from the solidarist hunter-
gatherer rules to the rules of the modern open society, individuals had to break the
rules – perhaps out of self-interest. From the hunter-gatherer stage “practically all
advance had to be achieved by infringing or repressing some of the innate rules
and replacing them by new ones which made the co-ordination of activities of
larger groups possible. Most of the steps in the evolution of culture were made
264 RICHARD N. LANGLOIS
possible by some individuals breaking some traditional rules and practicing new
forms of conduct – not because they understood them to be better, but because
the groups which acted upon them prospered more than others and grew” (Hayek,
1978, p. 161). We might even extend this idea to suggest that breaking the rules
– in the form of entrepreneurship – is still critical for economic growth (Choi,
1993). The group that prospers is not the one that finds a way to substitute altruism
for narrow self interest but rather the group that allows rules to be broken in
wealth-enhancing ways while constraining unproductive rent-seeking behavior.
NOTES
1. Or, more correctly, I used to say many of the same things back in the days when I
spent a lot of time thinking about methodology. Fritz Machlup used to maintain that one
should write about methodology only at the very beginning and the very end of one’s
career. I’ve tried to follow that advice, albeit with a few lapses here and there.
2. It should be clear from my formulation of claim 2, moreover, that Hodgson (1999,
pp. 132, 135) is wrong to classify me as a “reductionist” who believes that “emergent”
properties don’t exist.
3. Hayek’s view seems to be that, although the concept may or may not be appropriate
to biology, it is nonetheless important in the sphere of cultural evolution (Hayek, 1978,
p. 202, Note 37; 1988, p. 25).
4. “A routine is a way of doing something, a course of action. As Sidney Winter and
I have developed the concept (1982), the carrying out of a routine is ‘programmatic’ in
nature, and like a program tends largely to be carried out automatically. Like a computer
program, our routine concept admits choice within a limited range of alternatives, but
channeled choice. Almost always, there will be a set of understandings or beliefs associated
with a particular outline, which explicates or rationalizes why it is appropriate in a
particular context, and often, which provides an explanation of why and just how it works.
But the key operative concept is the routine itself. It is the routine used that determines
what is accomplished, given the context in which it is employed” (Nelson, 2002, p. 269).
5. Although sometimes they do, like Britain during the industrial revolution, where
population growth responded strongly to economic growth. The population of England
went up by about two-thirds over the course of the 18th century and more than doubled in
the first half of the 19th.
6. In The Fatal Conceit, published in 1988, Hayek already saw this kind of group
selection operating. Communism, he wrote, “is a religion which has had its time, and
which is now declining rapidly. In communist and socialist countries we are watching how
natural selection of religious beliefs disposes of the maladapted” (Hayek, 1988, p. 137).
REFERENCES
Choi, Y.-B. (1993). Paradigms and conventions: Uncertainty, decision making, and entrepreneurship.
Ann Arbor: University of Michigan Press.
Comment on “Group Selection and Methodological Individualism” 265
Hayek, F. A. (1978). Law, legislation and liberty. Volume 3: The political order of a free people.
Hayek, F. A. (1988). The fatal conceit: The errors of socialism. Chicago: University of Chicago Press.
Hodgson, G. (1999). Evolution and institutions. Cheltenham: Edward Elgar.
Langlois, R. N. (1983). Systems theory, knowledge, and the social sciences. In: F. Machlup & U.
Mansfield (Eds), The Study of Information: Interdisciplinary Messages. New York: Wiley.
Langlois, R. N. (1985). Knowledge and rationality in the Austrian school: An analytical survey. Eastern
Economic Journal, 9(4), 309–330.
Langlois, R. N. (1986). Rationality, institutions, and explanation. In: R. N. Langlois (Ed.), Economics
as a Process: Essays in the New Institutional Economics. New York: Cambridge University
Press.
Nelson, R. R. (2002). Technology, institutions, and innovation systems. Research Policy, 31, 265–272.
Nelson, R. R., & Winter, S. G. (1982). An evolutionary theory of economic change. Cambridge: Harvard
University Press.
RECONCILING GROUP SELECTION

AND METHODOLOGICAL
INDIVIDUALISM
Todd J. Zywicki
ABSTRACT
Methodological individualism underpins economic analysis. In his paper
in this volume, however, Douglas Glen Whitman demonstrates that group
selection can be reconciled with methodological individualism. This essay
extends Whitman’s analysis in two ways. First, it summarizes and restates
the necessary conditions for group selection to play a role in the evolution
of human preferences and societies. Second, it discusses the role of group
selection in Hayek’s thought, with a particular focus on the role of group
selection in the evolution of legal rules and the rule of law. The viability of
group selection is demonstrated to be an empirical question.
Individuals exist as parts of groups – families, tribes, firms, clubs. Given this
observable fact, it is plausible that under at least some evolutionary conditions,
individual tastes and preferences might evolve so as to contain sympathy for other
members of a larger group in which the individual exists. As Whitman observes
in the core paper of this discussion, “MI (methodological individualism) dictates

that individual choices lead to social outcomes; GS (group selection) is one force
(among many) that determines what sort of individuals are present in the system.”

ISSN: 1529-2134/doi:10.1016/S1529-2134(04)07013-9
267
268 TODD J. ZYWICKI
As I have argued elsewhere (Zywicki, 2000), properly understood, the argument

against group selection cannot be resolved as an a priori matter, as has been
previously argued. As Whitman extends the argument here, for similar reasons, the
argument that methodological individualism and group selection are incompatible
as an a priori issue is similarly flawed. In both circumstances, the question of the
relationship of group selection to individual selection and methodological individu-
alism is an empirical question. As Whitman observes, the argument between group
selection and methodological individualism in the social sciences mirrors in many
ways the argument between group selection and individual selection in biological
and cultural evolution. Rather than rehashing the arguments in Whitman’s insight-
ful and provocative paper, this comment will discuss some additional aspects of the
group selection versus individual selection debate. Whitman’s central conclusions
strike me as correct, thus there will be little direct commentary on his argument but
rather an attempt to supplement his analysis with some discussion of the conditions
under which group selection can operate as an empirical matter, as well as some
further elaboration on the possible compatibility and complementarities of group
selection and methodological individualism. Consistent with the overall focus of
this volume on Austrian economics and group selection, although perhaps outside
the specific issues addressed in Whitman’s paper, I also will use this opportunity
to discuss some unresolved questions related to the role of group selection in
Hayek’s thought.
THE CONDITIONS FOR GROUP SELECTION

Group selection models posit that individuals will sometimes act “altruistically,”
i.e. in ways that are good for others but detrimental to the individual. Groups
that have members that are willing to act altruistically will have a comparative
advantage in any competition with other groups. For instance, a basketball team
that has players who are willing to pass the ball to open teammates or to expend
energy playing defense will have a comparative advantage over teams that lack
this willingness of players to sacrifice for the good of the team, even though these
acts may reduce the number of points the player scores (and the recognition that
results). More dramatic is the example of soldiers who are willing to sacrifice their
lives in order to protect the lives of other members of their unit, such as by jumping
on a live grenade. It is clear that in these situations, a group that has members who
are willing to sacrifice themselves for the greater benefit of the group will have a
comparative advantage against groups whose members lack this willingness to act
altruistically or selflessly. Thus, from the perspective of the group as a whole, it is
beneficial to encourage this sort of behavior.
Reconciling Group Selection and Methodological Individualism 269
On the other hand, with respect to each individual member of the group, altruism
is subject to a free rider problem. While jumping on a live grenade certainly ben-
efits the group, it is anything but beneficial for the soldier who sacrifices himself.
From the perspective of the group as a whole, the group benefits from someone
jumping on the grenade; from the perspective of the individual soldiers within the
group, by contrast, each prefers that someone else jump on the grenade. Group
selection provides one possible explanation for this altruistic behavior between
genetically-unrelated individuals. As Richard Dawkins states the problem, “Even
in the group of altruists, there will almost certainly be a dissenting minority who
refuse to make any sacrifice. If there is just one selfish rebel, prepared to exploit the
altruism of the rest, then he, by definition, is more likely that they are to survive and
have children” (Dawkins, 1989, pp. 7–8, emphasis added). Most biologists have
traditionally adopted this logic that group selection is unworkable as an a priori
matter, because over time the force of individual decision-making will necessarily
undermine the social benefits of altruism and group selection. Economists have
similarly rejected group selection because it too seems to violate the premises of
methodological individualism, in that the set of incentives necessary for robust
group selection appear to lack firm and coherent individual-level foundations.
But it now appears that this rejection of group selection was too hasty (Sober
& Wilson, 1998). The possibility of group selection cannot be disposed of as
an a priori matter. It is an empirical question about the relative influence of
the force of intragroup selection, i.e. competition among different individuals
within a given group, versus intergroup selection, i.e. selection between different
groups. The possibility of group selection arises from the interaction of these
two offsetting forces: intergroup selection versus intragroup selection. Intergroup
selection encourages altruism within a given group because it benefits the group
as a whole in competition with other groups, even though altruistic individuals
contribute more to the group than they personally receive in exchange. Intragroup
selection, by contrast, favors selfishness and free riding – it is better to be on
the receiving end of the public benefits contributed by altruistic individuals
rather than being the “sucker” who contributes more than he personally receives
in exchange.
Given that these two influences offset one another, the question of group
selection is an empirical question, not a priori. It may be that the historical
circumstances that are necessary for group selection to exert an influence in
human affairs are implausible or unlikely; nonetheless, recognizing that the
question is empirical in nature reorients the discussion to an examination of the
actual circumstances that have shaped the evolution of individuals and cultures.
Whitman provides a technical discussion of the circumstances under which
group selection is possible. In practice, however, the conditions for group selection
270 TODD J. ZYWICKI
may be restated as requiring three operative conditions to be satisfied (Zywicki,

2000, pp. 87–88).
(1) First, the genetic trait or cultural rule must promise sufficient benefits to the
group that the members of the group will benefit from adopting it. In other
words, there must be some benefit to the group that results from the practice
when compared with groups that do not adopt the rule or practice, i.e. a social
surplus is generated. Thus, for instance, a group that adopts rules that honor
private property, freedom of contract, and restraints on the use of force and
fraud will tend to prosper relative to groups that adopt the opposite rules
(Hayek, 1988, p. 23). These rules may be invented consciously or may simply
arise by accident.
(2) Second, there must be some mechanism for between-group competition to
occur, i.e. for groups with more superior traits or practices to displace others.
This competition and displacement may occur either through: (a) warfare and
conquest by the more successful group; (b) migration from the less-successful
group to the more successful; or (c) imitation of the more successful by
the less-successful (Hayek, 1972, p. 169, n. 7; Hayek, 1988, n. 121). An
isolated group that never comes into contact with other groups will be unable
to engage in between-group competition, thus group selection will exert no
selection effect on the group’s rules in terms of group selection.
(3) Third, the group must be able to restrain free riders. The existence of an
altruistic trait or practice inevitably gives rise to the possibility of free
riding on those altruistic impulses. If unchecked, the ability of self-interested
individuals to free ride on others’ altruism can dissipate the social surplus gen-
erated by the altruistic act. On the other hand, it is not necessary to completely
eradicate free riding, which will be virtually impossible given the incentives to
free ride. It may be possible to reduce free riding to the point where the overall
benefits to the group are sufficiently large such that the benefits of retaining
the trait or practice are large enough to offset the costs imposed by free riders.
This specification of the third condition is perhaps the least-understood. Earlier
thinkers supposed that it would be necessary to eradicate free riding for group
selection models to be viable. This seems to be because it was thought that
there was no stable social equilibrium between complete altruism and complete
selfishness. But this ignores the fact that there is an intermediate equilibrium
that restrains free-riding without completely eradicating it. It is not necessary to
completely eliminate free riding, it is sufficient to control it to the extent that social
surplus generated by the practice or trait is not completely dissipated by intragroup
competition. Group selection predicts the possible emergence of altruism, not
masochism – successful groups will develop rules that will both generate social
surplus as well as limit its dissipation. Again, it does not matter whether the rule is
initially developed consciously or by accident, so long as selection forces can act
on it. Social norms against antisocial behavior, legal and political institutions such
as police forces that prevent theft, and constitutional institutions that encourage
positive-sum wealth creation activities rather than zero-sum redistributive activities
(or negative-sum rent-seeking activities), can all be viewed as mechanisms to limit
the ability of free riders to dissipate the social surplus (Zywicki, 2000, pp. 90–93).
If these three conditions are met: – (1) a beneficial practice that benefits the
group in intergroup selection competition; (2) a mechanism for between-group
competition (war, migration, or imitation); and (3) a mechanism for preventing
free riding by social parasites seeking to capture some of the newly-created so-
cial surplus for themselves – then the group selection model generates coherent
testable hypothesis. Note, however, that it may be difficult to meet these condi-
tions in practice, precisely because the forces identified by the methodological
individualism model are so powerful. Moreover, identifying circumstance where
group selection has actually operated will be difficult, given that the “losers” will
have been extinguished or absorbed into the more successful group. Nonetheless,
whether these conditions are met is an empirical question.
HAYEK, AUSTRIAN ECONOMICS,

AND GROUP SELECTION
The interest of Austrian economists in the relationship between methodological
individualism and group selection can be traced to Hayek’s interest in the topic.
The interest of Austrian economists in group selection is somewhat incongruous,
in that a cornerstone of Austrian economics is its emphasis on methodological
individualism. In large part, this strong commitment to methodological indi-
vidualism explains the hostility of many working in the Austrian tradition to
Hayek’s invocation of group selection in his writings (Vanberg, 1986). Many
scholars have been puzzled by Hayek’s use of group selection, precisely because
it seems incompatible with his emphasis on methodological individualism and

also because the selectionist model is not well-specified in Hayek’s work (see
for instance, Shearmur, 1996). This puzzle evaporates, however, if Hayek is
modeling groups of individuals as a sort of “super-organism,” as modern group
selection theorists do (Sober & Wilson, 1998). In fact, Hayek’s understanding of
the relationship between group selection and methodological individualism may
have changed over time, further adding to the confusion.
In his early writings invoking group selection, such as in Volume 1 of Law,
Legislation, and Liberty, Hayek resolves the tension by arguing that group
272 TODD J. ZYWICKI
selection of social rules is useful to any given individual because the rules pro-
duced by evolution at the group level enable a given individual to better coordinate
his purposes with other members of the same group (see, for instance, Hayek, 1972,
pp. 17–18). Although Hayek also makes some suggestions that the rules that have
survived the winnowing of the group selection process are better than those that
have not, he seems less concerned with the optimality of the content of the overall
rules, than with the fact that a system of rules will in fact evolve in a given society.
In this sense, Hayek’s use of group selection in Volume 1 of Law, Legislation, and
Liberty seems to be a sort of coordination game, with only a weak inference of
optimality. In this model, there is clearly no real tension between methodological
individualism and group selection – given some set of rules that govern the
behavior of other individuals within the same social group, a given individual best
accomplishes his own purposes by following the same set of rules. Following the
same rules as other members of the society will best enable him to coordinate his
affairs with other members of the society. In this sense, the social rules themselves
become part of the exogenous “environment” within which the individual acts,
just as the climate and presence of predators in a given society similarly create the
relevant environment for natural selection. In this sense, Hayek’s argument was
primarily concerned with proving the hypothesis that a set of rules could actually
emerge that were exogenous to individual choice and amendment, even though they
were initially constructed by individuals and appear at first to be under individual
control to change. Thus, any individual can prosper by following the rules given to
him by tradition within a given society, even if he does not understand the origin or
purpose of those rules.
What was Hayek attempting to accomplish through this demonstration that
rules could emerge through the group selection process, while retaining a focus
on the benefit to the individual from following those rules? I suspect that Hayek’s
purpose was to respond to the criticisms to his articulation of the rule of law in
The Constitution of Liberty. There (and less precisely in The Road To Serfdom),
Hayek argued that the rule of law required that any coercion applied by the state
to individuals must be according to rules articulated beforehand and applied
consistently to all parties (Zywicki, 2003 ). “Freedom” was defined as the absence
of arbitrary coercion by political actors. Hayek argued that compliance with the
rule of law was both a necessary and sufficient condition for freedom to prevail.
If all government application of coercion was structured and controlled by rules
articulated beforehand and applied equally, then the members of that society
were by definition, free. Thus, even if individuals were coerced by the state
(through taxation, imprisonment, even conscripted for military service), they
were still “free” if these restrictions were set out clearly ahead of time and applied
rigorously.
In Hayek’s account, however, the rule of law may be a necessary but not
sufficient condition of a free society. Even though advance articulation and
equal application of the law reduces the threat of arbitrary coercion, it does not
eliminate it because the rules themselves are still created by men (Hamowy, 1961,
1978). Thus, even if it is possible to fully articulate the conditions for government
coercion in advance, those conditions themselves may have improper distinctions
built into them. Human actors (legislators and judges) must still choose the rules,
therefore, the law itself can have human will built into it, thereby seemingly
legitimating undue restrictions on freedom so long as they comport with the rule
of law (Liggio, 1994; but see McDaniel, 2003).
I believe that Hayek wrote Law, Legislation, and Liberty (from this point referred
to as LLL) in response to this critique of his earlier articulation of the rule of law.
Hayek’s goal in Volume 1 of LLL is to argue that it is fundamentally incorrect
to believe that human decision-makers, whether legislators or judges, choose the
rules that govern their society. The rules are “chosen” for them through the process
of group selection. Thus, men simply articulate these pre-existing rules, they do
not create them. This argument, if correct, provides Hayek’s key response to the
Hamowy critique. If the rules themselves are not consciously chosen by political
decision-makers, but rather the political decision-makers merely articulate the
rules that are chosen through the group selection process, then it appears that
Hayek has closed the loop on his argument regarding the rule of law. Tradition
and group selection “choose” the rules, not transient political actors. To the extent
that an individual is “coerced” into performing on a contract, or imprisoned for
burglary, or prohibited from trespassing on another’s property, this coercion does
not constitute an undue infringement on his freedom. For, by definition, “freedom”
is defined as the absence of arbitrary coercion by another person. Here, it is not
the legislator or judge who is coercing the wrongdoer, but the force of tradition
and spontaneously-generated rules produced by the impersonal process of group
selection.
This explains Hayek’s repeated comparison between the legal system and the
market system in LLL. Hayek suggests that it is a logical absurdity to say that a grain
farmer is “coerced” when he has to sell his grain for a price lower than the price
at which he would prefer to sell. Because prices are set by the impersonal process
of the market, it cannot be said that any identifiable individual or individuals have
“coerced” your in deciding the price at which you can sell your grain. If you can
be said to be “coerced” at all, it is by the impersonal process of the market. In
LLL, rules emerge in society in the same way that prices emerge in a market.
Thus, just as it is nonsensical to say that the farmer is “coerced” by the market
into selling his grain for an undesired price, it is equally nonsensical to say that
your freedom is restricted when you are coerced by legal rules that have evolved
274 TODD J. ZYWICKI
spontaneously. To illustrate the point, it would be equally absurd to say that you are
“coerced” into using the word “car” to communicate the idea of a car to someone
else, rather than some other word you may prefer for the same idea, such as
“gooblestopper.” Is your freedom restricted when you are required to use the term
“car” to coordinate communication with others? No, Hayek suggests, because
language is not invented by anyone and so no particular person is forbidding you
from using “gooblestopper” instead of “car.” It is just that the word “car” has
evolved to mean a certain thing, and if you use that term you can coordinate
with others and accomplish your goals. If you do not use that term, you will
be unsuccessful. Thus, when the rules that govern interactions – market prices,
language, customs, legal rules – are generated by impersonal processes that are
controlled by no one, then being forced to comply with those rules cannot be
said to be an improper restraint on your freedom. With respect to the rule of law
specifically, common law rules that develop spontaneously and are articulated by
judges (not “created” by them) can thus be said to embody the rule of law. It is the
rules that coerce, not individuals. And because the rules themselves emerge from
the evolutionary group selection process and are not chosen by anyone, Hayek
argues that it can be said that their application is consistent with the rule of law
and freedom.
At first glance, the emphasis on the common law process, the spontaneous
development of law, and the centrality of common law judges in LLL seems to
represent a repudiation of the argument in The Constitution of Liberty. There, the
emphasis is on the idea of the Rechtsstaat and the discipline of the rule of law is
aimed at legislatures rather than judges. As suggested by this argument, however,
Hayek probably saw the argument of LLL as an elaboration of The Constitution
of Liberty, not a repudiation, in that it explains how the rules exist independent
of the will of individual law-makers. Thus, it is fully consistent to say that rules
may be the result of human action (methodological individualism) but not human
design (group selection). Put differently, Hayek’s thesis is that while individual
reason and individual action are good for introducing variation in to a system of
rules, individual reason and control is too limited to effectively shape the selection
among rules or systems of rules.

Individuals can choose and experiment with different individual rules within an
ongoing spontaneous order, exerting choice over which rules are selected to prevail
within the spontaneous order. The emergent properties of the system, however,
result from the spontaneous interaction of all of these individually created rules,
and thus are not designed by anyone. The spontaneously-generated system of rules,
therefore, can be understood as having group selection properties independent of
the particular attributes of the rules that comprise it. Because of the abstract and
complex nature of the spontaneous order, individuals will generally be unable to
know for certain which set of rules are optimal; it is only through competition
among different systems of rules that we can discover which system is best. Like
natural selection, selection among systems of rules is backward-looking, in that

which set of rules is superior to alternatives can be determined only after the
fact. On the other hand, Hayek is not wholly defeatist – experience, history, and
anthropology can provide some insight as to the attributes of systems of rules that
are most likely to prevail from this selection process.
In this analysis, group selection operates to generate a set of rules that allow
individuals to predict how others will likely act, and thereby to coordinate with
them. To the extent that this is the function of group selection, there clearly is
no conflict between group selection and methodological individualism (see also
Richerson & Boyd, 2001 for a further discussion of related issues). As Hayek
observes (Hayek, 1972, p. 44), “Society can thus exist only if by a process of
selection rules have evolved which lead individuals to behave in a manner which
makes social life possible.” He continues, “It should be remembered that for this
purpose selection will operate as between societies of different types, that is, be
guided by the properties of their respective orders, but that the properties support-
ing this order will be properties of the individuals, namely their propensity to obey
certain rules of conduct on which the order of action of the group as a whole rests”
(Emphasis added).
This reading of group selection in LLL arising in response to the rule of law
debate triggered by Constitution of Liberty is confirmed by its reduced importance
in the next two volumes of LLL. A rough count finds that Hayek mentions the
concept of group selection at least eight times in the text and several additional
times in the footnotes in the first volume of LLL. Many of these discussions are
quite extensive. There are far fewer references in the latter two volumes of the LLL
trilogy and in Volume 3 of LLL (Hayek, 1979, p. 202, n. 37). Hayek observes that
“the conception of group selection may now not appear as important as it had been
thought after its introduction,” although he adds “there can be no doubt that it is
of the greatest importance for cultural evolution.”
Beginning with his 1983 lecture at the Hoover Institution, “The Origin and
Effects of Our Morals: A Problem for Science” (Hayek, 1984), Hayek began to
develop a larger focus on the role of group selection in human affairs. Recall
that the emphasis on group selection in LLL may have been for the relatively
limited purpose of placing his model of the rule of law on a firmer foundation,
by showing that the rules existed independent of a particular individual’s will.
Given this focus, there appears to have been no strong effort in LLL to try to claim
optimality for these rules. In some sense, Hayek’s argument was limited to the
observation that the common law provided a foundation for a free society grounded
in the rule of law. To the extent that Hayek was making an argument about the
276 TODD J. ZYWICKI
optimality of the rules that emerge from the group selection process, those claims
appear to have been fairly weak. By The Fatal Conceit, however, Hayek seems
to be making stronger claims about the optimality of evolved rules and practices
(Whitman, 1998).
The Fatal Conceit frames the issue in a manner anticipated in the 1983 Hoover
lecture and offers a different reconciliation of the tension between individual and
group selection. As Whitman and I have both suggested, there seems to be little
incongruity in treating a business firm as a “individual” for purposes of an evo-
lutionary unit, even though a firm is actually a collection of individuals (Alchian,
1997 [1950]; Zywicki, 2000). Hayek offers a similar argument about the
small-group hunter-gatherer societies in which humans lived for most of their
evolution. Hayek sees most of the challenges facing these groups as being
collective challenges against a harsh physical environment or warring tribe.
To coordinate these small groups against ever-present threats, Hayek argues,
it was natural that humans would evolve genetic traits that helped to build the
solidarity of the small group internally and to oppose external threats. It is
this instinctual desire for solidarity and intragroup commitment that remains
with us today, motivating humans to aid one another to act altruistically toward
one another.
Like Whitman, Hayek suggests that there is no inherent tension between
individual and group selection. Because we have evolved “instincts of solidarity
and altruism” (Hayek, 1988, p. 12), Hayek implies that we ourselves gain
happiness and utility from seeing other members of our group prosper. It is
possible for small groups facing consistent threats to the group as a whole to
develop a high degree of cooperation and agreement on group ends as well
as the means to be used to accomplish those goals (Hayek, 1988, p. 19). In
economic terms, human beings have interdependent utility functions, in the
sense that they instinctively care about one another within their group. This
seems to be the model discussed by Whitman. And, if Hayek’s hypothesis about
the content of human nature is true, obviously there is no incongruity with the
idea of group selection and methodological individualism being consistent with
one another.
On the other hand, while it thus appears that group selection is possible and
consistent with methodological individualism, this simply leads to the next
question of whether group selection is valid empirically as playing an important
role in the evolution of human preferences and of human societies. Answering
this question will require further explanation of why these instincts would have
come into being in the first place and a deeper understanding the historical and
anthropological contexts that could have created an evolutionary environment
conducive to a role for group selection in human preference formation.
ACKNOWLEDGMENT
I would like to thank Eric McDaniel for helpful comments on an earlier draft of
this essay and the Law & Economics Center at George Mason University School
of Law for financial support.
REFERENCES
Alchian, A. A. (1977[1950, June]). Uncertainty, evolution and economic theory. The Journal of Political
Economy, 58(3), 15–36. Reprinted in Economic Forces at Work. Indianapolis, IN: Liberty Press.
Dawkins, R. (1989). The selfish gene. New York: Oxford University Press.
Hamowy, R. (1961). Hayek’s concept of freedom: A critique. New Individualist Review, 1, 28–31.
Hamowy, R. (1978). Law and the liberal society: F. A. Hayek’s constitution of liberty. Journal of
Libertarian Studies, 2, 287–297.
Hayek, F. A. (1972). Law, legislation, and liberty: Vol. 1, rules and order. Chicago: University of
Chicago Press.
Hayek, F. A. (1979). Law, legislation, and liberty: Vol. 3, the political order of a free people. Chicago:
Hayek, F. A. (1984). The origins and effects of our morals: A problem for science. In: C. Nishiyama &
K. R. Leube (Eds), The Essence of Hayek (pp. 318–330). Stanford, CA: Hoover Institution Press.
Hayek, F. A. (1988). The fatal conceit: The errors of socialism. In: The Collected Works of F. A.
Hayek & W. W. Bartley III (Eds). Chicago: University of Chicago Press.
Liggio, L. (1994). Law and legislation in Hayek’s legal philosophy. Southwestern University Law
Review, 23, 507–530.
McDaniel, E. M. (2003). The philosophy of F. A. Hayek’s economics. Unpublished Manuscript,
George Mason University School of Law.
Richerson, P. J., & Boyd, R. (2001). The evolution of subjective commitment to groups: A tribal
instincts hypothesis. In: R. M. Nesse (Ed.), Evolution and the Capacity for Commitment
(pp. 186–220). New York: Sage.
Shearmur, J. (1996). Hayek and after: Hayekian liberalism as a research programme. New York:
Routledge.
Vanberg, V. (1986). Spontaneous market order and social rules: A critique of F. A. Hayek’s theory of
cultural evolution. Economics and Philosophy, 2, 75–100.

Whitman, D. G. (1998). Hayek contra pangloss on evolutionary systems. Constitutional Political
Economy, 9, 45–66.
Zywicki, T. J. (2000). Was Hayek right about group selection after all? Review Essay of Unto others:
The evolution and psychology of unselfish behavior, by E. Sober & D. Sloan Wilson. Review
of Austrian Economics, 13, 81–95.
Zywicki, T. J. (2003). The rule of law, freedom, and prosperity. Supreme Court Economic Review, 10,
1–26. Chicago: University of Chicago Press.
LEVELS OF SELECTION AND

Adam Gifford Jr.
ABSTRACT
Whitman argues that group selection is consistent with methodological
individualism. He begins by defining a weak form of methodological indi-
vidualism in which agents are not necessarily self-interested or rational and
shows that this form is consistent with Sober and Wilson’s model of group
selection. However, Sober and Wilson’s group selection is also consistent
with a methodological individualism in which the individuals are rational
and self-interested, and consistent with individual selection as well. A version
of group selection similar to what Hayek may have had in mind when he
talked about groups out-competing other groups is presented, however, this
is not a version of group selection that is compatible with methodological
individualism.
LEVELS OF SELECTION AND

Whitman (2004) attempts to untangle the somewhat twisted relationship between
group selection and methodological individualism. In part, this issue arises
because of what some have argued is an inconsistency in the work of F. A. Hayek.

In the great majority of Hayek’s writings his methodology is firmly individualist,
but in his explanations of cultural evolution he at times invokes a group selection

ISSN: 1529-2134/doi:10.1016/S1529-2134(04)07014-0
279
280 ADAM GIFFORD JR.
process. In briefly reviewing the literature, Whitman reveals a consensus among

both biologists and economists that methodological individualism and group
selection are mutually incompatible. For example, Hayek’s seeming inconsistency

led Vanberg (1986) to argue that group selection and methodological individualism
are incompatible and that group selection should be rejected as a basis for cultural
evolution. Hodgson (1991) agreed with Vanberg’s incompatibility argument,
but disagreed with his negative view of the role of group selection in cultural
evolution. Whitman asserts that group selection and methodological individualism
form a mutually compatible basis for explaining cultural evolutionary processes
and he presents a strong argument in support of that position.
Whitman makes clear that methodological individualism does not preclude
higher level organization within which individuals interact and which influences
that interaction, it just requires that higher level “emergent properties of social
systems . . . be explained by the actions and interactions of the individuals
who take part in those systems” (Whitman, 2004, p. 228). He also argues that
“. . . methodological individualism does not specify whether agent preferences will
be selfish, altruistic, or something else entirely” (Whitman, 2004, p. 230). Further-
more, “. . . methodological individualism does not specify exactly how individuals
make choices. They might be perfectly rational actors, boundedly rational actors,
rule followers, or even automatons” (Whitman, 2004, p. 230). However, if
methodological individualism is going to be usefully applied in the development
of models that allow testable hypotheses, individuals in a world of scarce
resources must be assumed to display some basic consistencies in their behavior.
The usual assumptions in economics are that individuals are payoff-oriented and
able to discover economical ways of acquiring various payoffs; that is, they are
self-interested and rational – if only boundedly so. Altruism is consistent with this
model if individuals value the payoffs of others. Of course, putting others’ payoffs
into the utility function of the individual merely assumes altruistic behavior – it
does not explain it. Since group selection’s claim to fame is that it can explain
the existence of altruistic behavior, it would be a weak accomplishment indeed
if it explained a behavior by assuming its existence. If we want to understand the
relationship between methodological individualism, group selection, and altruistic

behavior, we should start with selfish individuals who also express some systematic
capability of achieving their goals when confronted by the various constraints that
are the product of scarcity and with changes in those constraints. In the following,
Whitman’s methodological individualism, which is shorn of any explicit assump-
tions about preferences or rationality, will be referred to as weak methodological
individualism, and adding selfish agents who are also at least boundedly rational
will be referred to as strong methodological individualism. These however,
should be thought of as being mere definitions and not taken as a methodological
Levels of Selection and Methodological Individualism 281
argument for one set of assumptions over the other. It will be argued that the new
group selection is not only consistent with strong methodological individualism but
with individual selection as well. Moreover, there exists another version of group
selection, one that is similar to what Hayek may have had in mind when he talked
about groups out-competing other groups. This version is one in which groupings
of organisms “. . . behave as an individual unit (as an organism) [themselves] . . .”
(Sober & Wilson, 1998, p. 145), however it is one that is inconsistent with both
strong and weak methodological individualism.
THE OLD AND THE NEW GROUP SELECTION
When explaining the emergence of altruism in nature, evolutionary biologists from

Darwin until the 1960s took it for granted that it was necessary to assume that the
group is the proper unit of analysis and that the selection process takes place at
the level of the group. Group selection was often used by biologists to explain
cooperative, seemingly “unselfish” behaviors, and individual selection was then
invoked to explain the vastly larger number of “selfish” traits of species. But in
1966 George C. Williams argued that a population of individuals that carried a
gene that resulted in an unselfish reduction in their fitness by directly limiting their
reproduction – or by engaging in any other costly behavior – for the good of the
group, would be invaded and successfully out-competed by individuals with genes
that led to selfish behavior, because the selfish individuals have a higher “fitness”
than the altruists (where an individual with a higher “fitness” transmits more copies
of its genes to the next generation than does its competitors). Selfish individuals
have higher fitness than unselfish ones, and thus leave more offspring. Williams
argued that altruism could not survive because selfish free-riders would come to
dominate any population of altruists. The old group selection position, because it
ignored the free-rider problem, fell before the logic of Williams’ argument, and
initially individual selection carried the day.
In the 1970s a new group selection that recognized and confronted the free-rider
problem was developed, first by David Sloan Wilson and later by Wilson and
Elliott Sober (see, Sober & Wilson, 1998; Wilson, 1975, 1997; Wilson & Sober,
1994). Proposed in 1975, Wilson’s trait group model still forms the basis of the
argument for the new group selection position.
Whitman does a good job of explaining both the difference between the
old and the new group selection and the main features of Wilson’s trait group
model. The trait group model, which is a major contribution to evolutionary
biology, lays out the conditions under which altruistic behavior might evolve.
Though Wilson’s model shows how the evolution of altruism is possible, the
group selection debate has continued with biologist John Maynard Smith and
others arguing that the new group selection is really just an example of individual
selection. In his review of Sober and Wilson’s (from this point forward referred
to as “S&W”) 1998 book, Maynard Smith stated: “I think the argument is largely
semantic, and could not be settled by observation” (Maynard Smith, 1998,
p. 639). In a discussion of this debate, the philosopher Samir Okasha agrees that
the argument cannot be settled by observation, “[b]ut the inference from ‘no
observations can decide the issue’ to ‘the issue is semantic’ seems to me a bit too
fast . . . As all philosophers know, any two objects are similar in some respects
but dissimilar in others. In devising concepts with which to describe the world,
we must therefore select some similarity relations as salient, and ignore others”
(Okasha, 2001, p. 34). I think the trait group model is best described as an
individual selection model, whereas Okasha and Whitman come down on the
other side of the issue. I believe that S&W (probably unintentionally) framed
their argument in a way that misdirects attention from the similarity relations that
place the trait group model in the individual selection camp. Ironically, Okasha
in part agrees, but he still argues that the trait group model is a model of group
selection.
Part of Whitman’s purpose is to use methodological individualism and the
new group selection to understand the process of cultural evolution. To aid in
understanding the difficulties with such an exercise, consider the following quote
from Hayek, taken from a passage discussing the different rules of conduct that
groups may practice: “These rules of conduct have thus not developed as the
recognized conditions for the achievement of a known purpose, but have evolved
because the groups who practiced them were more successful and displaced
others” (Hayek, 1973, p. 18). In other words, more successful groups have, in
some sense, better norms, rules, and other components of culture that allow them
to out-compete other groups. This view might be called a folk group selection
form of cultural evolution for two reasons: (1) it is probably the type of answer
that the average intelligent individual (with little exposure to the literature on
evolution) would give when asked about cultural evolutionary processes; and (2)
it does not provide, apart from positing the group as the unit of selection, much
detail about the actual evolutionary process involved. There is a sense in which
it may not be wrong, but it is incomplete – for example, by not specifying what
comes next. Furthermore, S&W’s and Whitman’s description of what takes place
next in the trait group model is not likely to spring to mind when one is asked
about the details of such a process: next, the parents all die (in some cases) and
the offspring disperse back into the general population, then they reform into
new groups, and the whole process starts over again. Leaving aside the part about
dying, the continual dispersing and reforming is not, in general, what seems to
have taken place in the process of cultural evolution. Additionally, as we will see,
even S&W give a somewhat confusing answer to the question of what comes next.
The following sections will address the issue of what exactly is going on in the
trait group model.
THE TRAIT GROUP MODEL:

INDIVIDUAL OR GROUP SELECTION?
There are several features of the trait group model that, considered together,
suggest that the best description of what is going on is individual selection: (1)
Although the trait group model explains the evolution of altruism, other charac-
teristics or traits must in fact evolve along with altruism for the process to work;
(2) S&W consistently use language such as the following to describe the problem
with the evolution of altruism: “[b]ehaviors that seem to benefit others at the
expense of self . . .” (S&W, 1998, p. 31), or an altruistic individual “. . . reduces its
own fitness and augments the fitness of others” (S&W, 1998, p. 199). Statements
of this sort provide a clue to the difficulties with the evolution of altruism and to
the problem with S&W’s view of the group selection process: they focus on the
individuals at a given moment in time and ignore the dynamic nature of the process
over time. We might just as well say that credit markets operate to the detriment
of lenders since by lending they reduce their own wealth and enhance that of the
borrower. Here we ignore that fact that the borrower must pay back the loan with
interest in the future.1 Although biological evolutionary processes are blind, they
play out over time and can produce, for example, what economists would call a
mutually beneficial non-simultaneous exchange – even though the agents involved
are not aware of the long-term implications of their behavior; and (3) The trait
group model claims to be a model in which groups are selected, that is, one that
allows for levels of selection above that of the individual to operate, in a process
that is similar to, but at a higher level, than individual selection. However, the
common currency payoff of this higher level process in the new group selection
is not that successful groups leave more offspring groups to the next generation
than less successful groups, but rather that individuals in successful groups leave
more offspring individuals to the next generation. These points will be discussed
in order.
The dispersal and regrouping is a necessary feature of the new group selection:
otherwise, selfish individuals who have a higher fitness in each group will
eventually dominate in the group. Whitman makes this point clear (Whitman,
2004), as do S&W, and it is important because it is a necessary trait of
the individuals involved, a behavior of the individual agents, not the group.
Moreover, even though Whitman argues that the regrouping process can be
“through chance” (Whitman, 2004), if the groups are small enough, this is
unlikely to be the case. “If the groupings are formed at random, then altruism is
eventually eliminated . . .” (Okasha, 2001, p. 30). Even with small groups, as the
dispersal and regrouping plays out through time, the result will be the same as
for the large group case and the selfish trait will dominate. Given assumptions
consistent with those of the trait group model Bergstrom (2002, 2003) establishes
this result. Whitman supports his conjecture by arguing that “[r]andom mating
will result in some fraction of two-altruist pairs” (Whitman, 2004, p. 235). The
problem with this is that the mating pair isn’t the relevant group: the relevant
group is the sibling group that results from the mating, or the parent-sibling
group, neither of which is random. If both of the parents are altruistic, then the
resultant sibling group of altruists is not a product of randomness. In fact, in the
trait group model the grouping process is referred to as assortative: “. . . there is a
statistical tendency for altruists to find themselves grouped with other altruists . . .”
(Okasha, 2001, p. 30).
In the trait group model the fitness of the individual depends on the charac-
teristics of the group it is in, and the fact that altruists are more likely to assort
with other altruists allows altruism to evolve as an emergent property, but this
can be explained entirely by the traits or behavior of the individuals and their
interactions. The traits that lead to this result are altruism itself, dispersal and
regrouping, and assortative behavior.2 Once in a given group, an individual
altruist may or may not discriminate between altruists and non altruists in its
behavior, but because of the discrimination that occurred in the regrouping
processes itself, its own altruistic behavior is likely to be reciprocated by other
altruists in the group. When these traits are considered together, it is not the case
that over time the altruist “. . . reduces its own fitness and augments the fitness
of others” (S&W, 1998, p. 199). Statistically, altruists, by assortative grouping
with other altruists, followed by cooperating with members of that group, end up
with a higher fitness level then non-altruists in the total population because of the
increased likelihood that they will be on the receiving end of benefits provided by
other altruists.
Whitman suggests that a critic might object to group selection because
“. . . individual incentives are being ignored” (Whitman, 2004, p. 235). But in
the new group selection incentives aren’t ignored; altruism is clearly in the
self-interest of the individual. Given the multi-trait aspect of the trait group model,
the individual is not ignoring incentives; rather, it is in the interest of the individual
to cooperate, to provide a benefit to another even though it does not necessarily
receive an immediate benefit in return. It is in the interest of the individual to
be altruistic because it is then likely to end up in a group with a high proportion
of other altruists from whom it receives benefits, whereas non-altruists are much
less likely to end up in a group with a high proportion of altruists. Despite the
language used, these statements should not be taken anthropomorphically or

taken to imply teleology – the processes are blind and non-intentional. The trait
group model suggests that, starting with a population of selfish non-cooperators,
a mutation resulting in “altruistic” cooperators could evolve to the point where
the “altruistic” trait dominates the population. That is, the trait group model
coupled with strong methodological individualism can lead to the evolution
of altruism, but it is an individual selection model with group level emergent
properties.
Assortative Processes
How do the subsequent generations of altruists come together with other altruistic
individuals? There are several ways in which assortative behavior can bring
together groups with proportions of altruists large enough to enable altruism to
evolve. With sexual reproduction, for example, random mating of pairs of adults
that gives rise to “isolated sibling groups” can result in the evolution of altruism
if altruism is a dominant trait and the benefit conferred by altruistic behavior is
sufficiently large relative to the cost (see S&W, 1998, p. 63).
A second alternative is that the grouping of altruists results from altruists being
able to identify other altruists. This identification process can be the result of
a tit-for-tat strategy in which an altruistic player makes an initial investment in
identifying other altruists by first grouping randomly, behaving cooperatively on
the first encounter, and then subsequently cooperating only with other altruists
identified by their past behavior. Or the altruistic trait may result in its carriers pos-
sessing a secondary, identifying trait – Dawkins (1989, p. 89) calls this the green
beard effect.3
These two paragraphs have respectively described William Hamilton’s (1964)
model of kin selection, or inclusive fitness, and Robert Trivers’ (1971) model
of reciprocal altruism, coupled with the tit-for-tat strategy for repeated games
that came out of Robert Axelrod’s (1984) computer tournaments. Interestingly,
though S&W (1998) argue that kin selection and reciprocal altruism are examples
of group selection, Hamilton and Trivers both argued that their models showed
how individual selection can lead to the evolution of altruism.
In the kin selection model, an individual may make sacrifices that benefit
another if the following inequality holds: Br > C, where B is the benefit to the
other individual, r is the average degree of relatedness between the two individuals,
and C is the cost to the individual performing the altruistic act. As an example:
for sexually produced siblings, where r = 0.5, an individual will perform an

altruistic act that yields a benefit to a sibling or siblings that exceeds twice the
cost it bears in performing that act. This, of course, is purely selfish behavior from
a gene perspective. Reciprocal altruism coupled with tit-for-tat strategies involves
a group of individuals in long-term relationships where sacrifices made to benefit
another individual today will be more than made up for by reciprocal sacrifices
in the future. A key requirement for this strategy to work is that the players have
sufficient brain power to remember the past performance of the other players.
Consequently, tit-for-tat can be considered a reputation model, where individuals
cooperate with those who have cooperated with them in the past – once again,
showing a self-interested behavior leading to apparently altruistic acts because,
for the individual, the benefits exceed the costs.
Both of these models have assortative grouping. In the sib group case, siblings
are paired with siblings so that an altruistic act will enhance the benefactor’s
fitness since its sibling carries copies of its genes. Reciprocal altruism supported
by tit-for-tat results in individuals who refuse to cooperate with non-altruists
after once being burned. Economists recognize that individuals cooperate within
groups such as the firm or the state – and that individual actions within a group
are profoundly influenced by the nature and structure of that group – without
giving up strong methodological individualism.
That these models were developed to show that individual selection can
result in the evolution of altruism is part of the reason behind Maynard Smith’s
assertion that the new group selection debate is over “semantic” issues, not facts
about the world. “Sober and Wilson have simply chosen to use the expression
‘group selection’ in a non-standard and highly misleading way, he [i.e. Maynard
Smith] claims; so their departure from orthodoxy is purely terminological”
(Okasha, 2001 , p. 34).
[T]here is no overriding reason to re-describe either kin selection or evolutionary game theory
[reciprocal altruism] within the framework of intra-demic selection,4 as Sober and Wilson
admit; but this does not alter the fact that such re-description is possible. However, the claim
that intra-demic selection is itself a type of group selection is controversial, and it is here that the
majority of evolutionary biologists would quarrel with Sober and Wilson’s reasoning (Okasha,
2001 , p. 34).
Recall that, in the trait group model, individuals are members of a group for only
a fraction of their lifespan. According to S&W, the individuals that come together
in reciprocal altruism, even if for a brief moment, form a group.
Once we go beyond the individual organism interests are never perfectly
aligned and fate never perfectly shared. S&W recognize this problem and to try
to get around it they introduce the concept of secondary altruistic behaviors. But
these so-called secondary behaviors, such as the punishment of free-riders along
with the assortative regrouping process, are mechanisms designed to align the
interests of the individuals involved; additionally, these secondary behaviors are
traits of the individuals. In economic terms these behaviors are actions taken to
enforce and facilitate agreements, actions taken to reduce shirking, opportunism,
the principal-agent problem, adverse selection, and other problems associated
with non-simultaneous exchange. By calling these features secondary, S&W
seem to imply that they are simply add-ons, when in fact they are necessary for
the evolution of altruism, and thus should be considered primary features.
Interestingly, models with these “secondary” altruistic behaviors can be
remarkably similar to those used by economists. Reeve and Keller (1997), for
example, essentially reinvent the Coase theorem and throw in some features
from the “Nature of the Firm.” They start with a kin selection model and add
reproductive bribing that they show, not surprisingly, can increase cooperation
in the face of externality-generating selfish behavior when transactions costs
are suitably low.5 Other new group selection models, for example, see Wilson
and Dugatkin’s (1997), like Trivers’ original reciprocal altruism model, can be
construed as reputation models of assortative behavior.6
Under the right circumstances evolution can facilitate cooperative behavior
among individuals in groups, cooperative behavior that increases the relative
fitness of the group members. Is this group selection or individual selection? When
cooperation succeeds and that cooperation increases the fitness of individual
group members, S&W assert that selection is taking place at the level of the group.
But they recognize that self-interest tends to undermine cooperation, requiring
mechanisms to suppress free-riders. Finally, “[m]ultilevel selection approaches
as exemplified by trait-group selection models are not fundamentally different
from ‘classical’ individual selection approaches as represented by generalized
inclusive fitness models. It is possible in every instance to translate from one
approach to the other without disturbing the mathematics describing the net
results of selection” (Reeve & Keller, 1997, pp. S42–S43).
Economists use models similar to the ones discussed to explain how cooperation
among self-interested individuals is possible. It seems likely that Sober and
Wilson would describe the lender-borrower relationship as an example of group
selection, where lenders assort by checking the credit rating of borrowers and
demanding collateral. Economists produce models of individuals interacting
in groups with much of this interaction involving non-simultaneous exchange,
and they do so without giving up the assumption of strong methodological
individualism or arguing that group selection is taking place.
Levels of Selection
Sober & Wilson view selection as operating on different levels and argue that
group selection is possible when the selective forces at higher levels are strong.
They propose a simple three-step program.
Step 1: Determine What Would Evolve if Group Selection Were the Only Evolutionary Force. In
this case, traits will evolve that maximize the fitness of groups relative to other groups . . . Step
2: Determine What Would Evolve if Individual Selection Were the Only Evolutionary Force. In
this case, traits will evolve that maximize the fitness of individuals, relative to other individuals
in the same group . . . Step 3: Examine the Basic Ingredients of Natural Selection at Each
Level. The process of natural selection requires three basic ingredients: (a) phenotypic variation
among units; (b) heritability; and (c) differences in survival and reproduction that correlate with
the phenotypic differences. To determine the balance between levels of selection, we need to
examine these ingredients at each level (S&W, 1998, pp. 103–105).
But what is really going on is that emergent properties at the level of the group
are important in the trait group model, and these properties can be explained by
the behavior of the individuals involved and their interactions. This brings us to
another part of the new group selection debate: the question of what is meant by
levels of selection.
The debate between S&W and Maynard Smith took place in two stages, as
reviewed by Okasha (2001). Natural selection at a given level requires that entities
at that level vary with respect to a given trait, that the trait be heritable – passed
from parent to offspring – and that the trait influence the entity’s fitness. If the trait
increases an individual’s fitness relative to other individuals, individual selection
requires that the individual organism possessing the trait leave more offspring
individuals then other individuals, and in group selection where the entity is a
group, Maynard Smith (1987) argues that a group possessing the trait must leave
more offspring groups then other groups. “For group selection to occur, Maynard
Smith reasoned, we need groups of organisms that exhibit heritable variation in
fitness” (Okasha, 2001, p. 35). This is consistent with the view expressed by S&W
in the extended quote above, but as we have seen, it is not what actually occurs
in the trait group model. In the trait group model, the fitness of the individual
does depend upon group level emergent properties, but the common currency of
the process is individual offspring. S&W’s group selection model “introduce[s]
a curious asymmetry: individual selection requires individual-level heritability,
but group selection does not require group-level heritability . . .” (Okasha, 2001,
p. 36). In their book, S&W (1998) argue that the trait group model is consistent
with group level heritability, but even Okasha, who supports the position the trait
group model is group selection, asserts that “. . . I do not think Sober and Wilson’s
overall position is defensible” (Okasha, 2001, p. 40).7
So why does Okasha claim that the trait group model of intra-demic selection
is group selection, “. . . given that the operative notion of group fitness in the intra-
demic selection model is average propensity to leave individual offspring, rather

than the propensity to leave offspring groups” (Okasha, 2001, p. 43)? It is because
of the often overlooked fact that the old group selection that “. . . involve[ed]
extinction and colonization still measure[ed] a group’s fitness by the number of
individual offspring its members leave . . .” (Okasha, 2001, p. 43), just as is true
of the new group selection and not by the number of offspring groups it leaves. In
other words, the common currency of success in both the old and the new group
selection is the propensity to leave more offspring individual organisms, not more
offspring groups.8 The importance of group-level effects and the use of the same
measure of fitness in both the new and the old group selection leads Okasha to
assert that intra-demic selection is group selection, because of the “similarity
relations” they share. However, it is important to remember what “similarity
relations” they don’t share. Mainly, the new group selection posits mechanisms
to control free-riders, mechanisms that are absent in the old group selection.
And since these mechanisms are expressed as traits of individuals, even though
group level effects are important in the new group selection it is best described as
individual selection in which emergent higher level properties play an important
role. For the same reasons, the new group selection is consistent with strong
methodological individualism.
What about a version of group selection that S&W envision but don’t produce,
one in which individuals stay in groups that are long-lived, groups beget groups,
fitness is measured by the number of offspring groups and groups each “. . . behave
as an individual unit (as an organism) . . .” (S&W, 1998, p. 145)? Since, this
version of group selection would still have to deal with the free-rider problem, its
actual existence is an empirical question and not a semantic one.9 At this level of
description we have folk group selection, and I believe it is something like what
Hayek, Vanberg, and Hodges have in mind when they discuss group selection.
Furthermore, since in this folk group selection the unit of analysis is the group, I
believe (as is also maintained by Vanberg & Hodges) that it is not consistent with
either strong or weak methodological individualism.

The various models of intra-demic selection allow for the effects of higher level
emergent properties, but these can all be explained by the interactions of the agents
involved and the traits of those agents. Additionally, the payoffs of the group
level effects accrue to the individuals involved, not to groups, as heritability is
an individual level feature. “When they [S&W] talk about the ‘differential fitness
of groups’ as a prerequisite for group selection, they do not mean groups which
differ in the number of offspring groups they are expected to leave, but rather
groups which differ in the average number of individual offspring their members
are expected to leave” (Okasha, 2001, p. 41). The trait group model of intra-demic
selection is revealed to be similar to various economic models of cooperation in
which the payoffs to the individuals depend in part on the higher level structures
in which the cooperation takes place. Payoffs in the firm, the market economy,
the common law, all depend on higher level properties – but we still take the
individual as the unit of analysis and assume the strong version of methodological
individualism.
CULTURAL EVOLUTION
As is pointed out by Whitman, methodological individualism technically is appli-
cable only in the case of cultural evolution. Is it possible that S&W’s discussion of
cultural evolution is consistent with a group heritability version of group selection,
a discussion more in line with our folk notions of what group selection must be
like, even if the trait group model is not? The sections in their book on cultural
evolution, like the others, contain some misleading and confusing discussion. In
their review of social norms in 25 randomly selected cultures, S&W state that
“[d]espite the powerful social norms that exist in most cultures our survey makes
it clear that individuals do attempt to violate norms for their own advantage . . .”
(S&W, 1998, p. 168). Violation of norms, and free-riding in general, is subject to
retaliation by others, either by the refusal to cooperate with the free-rider in the
future or by punishment, as is standard in individual selection models.
In citing studies by Christopher Boehm, who has “. . . surveyed hundreds of
egalitarian band-level and tribal societies” (Boehm, 1999, p. 6), studies that have
since been consolidated in Boehm (1999), S&W stress the egalitarian ethic of
these small band-level societies. But this egalitarianism is political, not economic,
involving equal opportunities not equal outcomes, and it was maintained at a high
enforcement cost. Adult band members used elaborate mental accounts and active
monitoring of other band members to maintain their egalitarian social structure.10
Social control involves far more than an outraged group’s suddenly deciding to employ dramatic
sanctions. In any small group anywhere, people keep track of one another’s behavior and try
to read underlying motives. Types of deviance that all human groups watch for, gossip about,
and react to, include murder within the group, heavily self-interested verbal deception, theft,
and stinginess or failure to cooperate when this is appropriate. On the positive side, foragers
talk about generosity, cooperativeness, honesty, and other prosocial behaviors that involve good
will. In effect the band keeps a dossier on every individual, noting positive and negative points
(Boehm, 1999, p. 73).
The bands, did however employ dramatic sanctions, sanctions ranging from
coolness of behavior toward those who violate social norms to ridicule, shunning,
ostracism, and execution. To the extent that the bands had formal or informal
leaders, those leaders were individuals who had a great deal of knowledge and
wisdom, were good at dealing with people, and were persuasive speakers, but not
boastful, arrogant or overbearing (see Boehm, 1999, p. 69). Robert Kelly (1995),
who has examined over a hundred hunter-gatherer societies, states:
The term egalitarian does not mean that all members have the same amount of goods, food,
prestige, or authority. Egalitarian societies are not those in which everyone is equal, or in which
everyone has equal amounts of material goods, but those in which everyone has equal access to
food, to the technology needed to acquire resources, and to the paths leading to prestige. The
critical element of egalitarianism, then is individual autonomy (Kelly, 1995, p. 296).
Sober and Wilson’s discussion of meat sharing is also misleading. Referring

to the meat sharing and secondary behaviors that induce that sharing, they say
that, “[b]y causing another individual to perform an altruistic primary behavior
such as hunting and sharing, the secondary behavior increases the fitness of the
entire group” (S&W, 1998, p. 143). They argue, further, that the facts of this
behavior “. . . don’t fit comfortably within the framework of individual selection
theory” (S&W, 1998, p. 143). In fact, nothing unique is occurring in these
hunter-gatherer societies, as Kelly notes in his observation that “[s]tudents new
to anthropology . . . are often disappointed to learn that these acts of sharing
come no more naturally to hunter-gatherers than to members of industrial soci-
eties . . . Sharing . . . strains relations between people” (Kelly, 1995, pp. 164–165).
Although other food resources are occasionally shared in various circumstances
(e.g. when food is very scarce or some band members are down on their luck), the
meat of large game is always shared (Kelly, 1995, p. 165). There are significant
variations in the daily returns to hunting in the hunter-gatherer environment, and
where food storage is not an alternative, meat-sharing represents an efficient way
of reducing the variance in access to meat facing any one family. In the absence
of adequate storage technology, a successful hunter faced rapidly diminishing
short-run marginal value with the large quantities of meat yielded by large game.
Meat-sharing was a form of insurance that allowed hunters who were successful
on a given hunt to trade lower valued excess meat for claims to high valued meat
in the future when their hunting was not successful. Unlike what S&W seem to
suggest – that secondary behaviors forced some group members to hunt and then
share with the rest – all adult males hunted and meat sharing was simply extended
reciprocity, enforced by various sanctions, which lowered the variance of access
to high valued resources to all members of the band. What is actually occurring
in these hunter-gatherer bands is extended reciprocity coupled with the natural
desire for individual autonomy. That this could be achieved in stateless societies
was, in part, a function of the relatively small size of the bands. Hunter-gatherer
bands contained around 25 members, a number that facilitated voluntary solutions

to various externality problems because transactions costs were low.11 In other
words, nothing is occurring in these bands that is inconsistent with individual

selection and strong methodological individualism.
S&W, in their discussion of the payoffs in the process of cultural group selec-
tion, refer to examples of the “replacement of ‘weak lineages by strong lineages’
either by warfare or sheer economic superiority . . .” (S&W, 1998, p. 173). They
also present an extended discussion of the displacement of the Dinka tribe by the
neighboring Nuer, both of which were African pastoral societies. Warfare was
the primary means of this replacement of the Dinka by the Nuer, and the Nuer
were successful because they had superior warriors who were “known for their
discipline, bravery, and ability to withstand heavy casualties in battle” (S&W,
1998, p. 189). In discussing these examples, S&W, present no new formal model
and no evidence, aside from the mention of warfare, where one group merely
replaces another group, about whether these examples are generally consistent
with a group selection where group level fitness is the common currency.
A major problem here is that we have no good theory of cultural evolution that
we can use to interpret statements like S&W’s “replacement of ‘weak lineages
by strong lineages’ either by warfare or sheer economic superiority . . .” (S&W,
1998, p. 173) or the displacement of the Dinka by the Nuer. A problem with
cultural evolution is that the various components of culture are very different
from each other and are often selected and transmitted in very different ways.
Innovations such as agriculture can be transmitted from group to group without
the replacement of the individuals involved. The same can be said for economic
goods, cultural norms, religion, systems of government, technology and science.
The mechanisms of cultural evolution are many and varied, with warfare and other
forms of apparent replacement representing only part of the story, and even in these
cases most of the individuals involved are not actually replaced. Given that culture
seems to consist of many disparate things and processes, there does not seem to be
a single simple fundamental unit of culture that serves as both a replicator and a
unit of selection. Furthermore, some components of culture seem to evolve without
conscious awareness, following Hayek (1960), whereas others may be consciously

chosen. With these points in mind, we might better describe the process as cultural
evolutions – consisting of more than one fundamental evolutionary process.
CONCLUSION
The trait group model represents an important contribution to our understanding
of the evolution of altruism because of its careful consideration of both the role of
higher level organization as the environment that influences the evolutionary fate
of individual members of the group and the necessity of controlling free-riding.
However, should we think of group selection and group level emergent properties
as the same process? Sober and Wilson and Whitman argue that we should. But
in S&W’s model the outcomes can all be explained by traits of individuals and,
moreover, fitness is measured in terms of the offspring of individuals. This version
of group selection is actually consistent with individual selection and a method-
ological individualism in which individuals are rational and self-interested.
In their general discussions of higher level selection and cultural evolution,
S&W suggest that group selection involves group level heritable traits that vary
among groups and effect group fitness, with group level reproduction in which
groups reproduce by sending out offspring groups. This is a version of group
selection in which groupings of organisms are long-lived, “. . . behave as an
individual unit (as an organism) . . ..” (S&W, 1998, p. 145) themselves, have
group level traits not explained by the traits of the individual organisms, and
where fitness is measured by the number of offspring groups that each group
produces. This version of group selection is more consistent with what Hayek had
in mind when he talked about groups out-competing other groups, but it is not a
version of group selection that is compatible with methodological individualism.
Importantly, this isn’t what goes on in intra-demic selection models, where traits
are properties of individual organisms and fitness is measured by the number of
offspring individuals those organisms produce.
By choosing the S&W version as his model of group selection, Whitman chose
a version that was consistent not just with his weak notion of methodological
individualism, but also with a methodological individualism that includes rational
and self-interested individuals.
NOTES
1. See Gifford (2000).
2. Assortative behavior, which functions to exclude non-altruists from the group, can be
augmented or replaced with behavior that imposes costs on non-altruists and that tends to
induce a switch in their behavior. However, some additional trait is necessary for altruism
to evolve.
3. A green beard effect has been found in polygyn (multiple queen) colonies of fire
ants. The green beard takes the form of a genetic effect resulting in a queen carrying a
recognizable odor that allows the workers to separate kin from non-kin. See Keller and
Ross (1998).
4. The trait group and other similar models are sometimes called models of intra-demic
selection because they involve the selection of the characteristics of individuals in
subpopulations or demes.
5. Reeve and Keller do not use the economic terminology: externalities and transactions
costs.
6. See Gifford (2000).

7. In part this is because the parent groups break up and the next generation of
individuals reform into new (offspring) groups. Assume, as do S&W (1998, p. 111),
that group parenthood is defined as a group contributing at least some members to an
offspring group. Since the assorting process is not perfect, it is possible that every previous
generation group contributes some individuals to each next generation group, so that each
parent group is a parent of each offspring group, making the fitness of each parent group
equal and rendering the notion of group level fitness meaningless. See Okasha (2001,
pp. 41–42).
8. For an early source of this insight, see Arnold and Fristrup (1984).
9. This version of group selection may explain why “. . . genes are found on chro-
mosomes, chromosomes in cells, cells in tissues, tissues in organs, [and] organs in
organisms . . .” (Okasha, 2003, p. 350).
10. On the other hand, it is very likely that nonhuman cooperation among non-kin is
associated with direct reciprocity supported by simple one-on-one mental accounts stored
in implicit memory systems that are not accessible to conscious introspection and that
generate motivational drives.
11. See Gifford (2002) for an extended discussion of this issue.
REFERENCES
Arnold, A. J., & Fristrup, K. (1984). The theory of evolution by natural selection: A hierarchical
expansion. In: R. N. Brandon & R. M. Burian (Eds), Genes, Organisms, Populations:
Controversies over the Units of Selection (pp. 292–319). Cambridge, MA: MIT Press.
Axelrod, R. (1984). The evolution of cooperation. New York: Basic Books.
Bergstrom, T. C. (2002). Evolution of social behavior: Individual and group selection. Journal of
Economic Perspective, 16, 67–88.
Bergstrom, T. C. (2003). Group selection and randomness, reply. Journal of Economic Perspective,
17, 211–212.
Boehm, C. (1999). Hierarchy in the forest: The evolution of egalitarian behavior. Cambridge, MA:
Dawkins, R. (1989). The selfish gene (new edition). Oxford: Oxford University Press.
Gifford, A., Jr. (2000). The bioeconomics of cooperation. Journal of Bioeconomics, 2, 153–168.
Gifford, A., Jr. (2002). The evolution of the social contract. Constitutional Political Economy, 13,
361–379.
Hamilton, W. D. (1964). The genetical evolution of social behavior. I. II. Journal of Theoretical
Biology, 7, 1–16, 17–52.
Hayek, F. A. (1973). Law, legislation and liberty: Vol. I, Rules and Order. Chicago: University of
Chicago Press.
Hodgson, G. M. (1991). Hayek’s theory of cultural evolution: An evaluation in light of Vanberg’s
critique. Economics and Philosophy, 7, 67–82.
Keller, L., & Ross, K. G. (1998). Selfish genes: A green beard in the red fire ant. Nature, 394,
573–575.
Kelly, R. L. (1995). The foraging spectrum. Washington, DC: Smithsonian Institution Press.
Maynard Smith, J. (1987). How to model evolution. In: J. Dupre (Ed.), The Latest on the Best: Essays
on Evolution and Optimality (pp. 119–131). Cambridge, MA: MIT Press.
Maynard Smith, J. (1998). The origin of altruism. Nature, 393, 639–640.

Okasha, S. (2001). Why won’t the group selection controversy go away? British Journal for the
Philosophy of Science, 51, 25–50.
Okasha, S. (2003). Recent work on the levels of selection problem. Human Nature Review, 3, 349–356.
Reeve, H. K., & Keller, L. (1997). Reproductive bribing and policing as evolutionary mechanisms for
suppression of within-group selfishness. The American Naturalist, 150, S42–S58.
Trivers, R. L. (1971). The evolution of reciprocal altruism. Quarterly Review of Biology, 46, 35–57.
Vanberg, V. (1986). Spontaneous market order and social rules: A critical examination of F. A. Hayek’s
theory of cultural evolution. Economics and Philosophy, 2, 75–100.
Whitman, D. G. (2004). Group selection and methodological individualism: Compatible and
complementary. Advances in Austrian Economics, 7, 221–249.
Williams, G. C. (1966). Adaptation and natural selection: A critique of some current evolutionary
thought. Princeton: Princeton University Press.
Wilson, D. S. (1975). A general theory of group selection. Proceedings of the National Academy of
Sciences, 72, 143–146.
Wilson, D. S. (1997). Introduction: Multilevel selection theory comes of age. The American Naturalist,
150, S1–S4.
Wilson, D. S., & Dugatkin, L. A. (1997). Group selection and assortative interactions. The American
GROUP SELECTION AND

METHODOLOGICAL
INDIVIDUALISM: REPLY
TO COMMENTS
I thank all of the discussants for their comments, which have helped me to sharpen
my understanding of the issues. With a few caveats, all the discussants appear to
agree with my central thesis: that methodological individualism (MI) and some
form of group selection (GS) are compatible. Here, I will focus on those aspects of
their comments I found most provocative – in either the positive or negative sense
of that term.
1. RESPONSE TO LANGLOIS
Langlois agrees with my position – or perhaps I should say, given the order of
our publications, that I agree with his. I accept his taxonomy of methodological
positions and agree that position two is correct.
Langlois also makes an important observation about Hayek’s theory of cultural
evolution: that excessive, not insufficient, group solidarity often poses the greatest
threat to the emergence of beneficial institutions like property rights and markets.
In order for an open society to emerge, Hayek argues, cultural evolution must
overcome certain instincts built up by humans’ prior evolution (both biological
and cultural). Interestingly, those instincts may have resulted, in part, from

ISSN: 1529-2134/doi:10.1016/S1529-2134(04)07015-2
297
GS. It is probably an overstatement, however, to say that the actual conflict is

“exactly opposite to that envisioned by biologists” (Langlois, 2004, p. 263). The
institutions of the open society also require some level of individual sacrifice
for group benefit. The institution of private property, for instance, requires that
individuals sometimes forgo an opportunity to help themselves to the fruit of
others’ labor, even though doing so would benefit the taker. The imposition of
sanctions on violators might also require some amount of sacrifice – e.g. forgoing
trade with someone who has been ostracized. Thus, Hayek’s theory does not
really rely on the weakening of GS relative to individual selection, but rather
the replacement of old group-beneficial norms with new group-beneficial norms
by means of GS.
2. RESPONSE TO ZYWICKI
Zywicki also appears to agree with me in most respects. However, two of his
statements give me pause:
! “As Whitman observes, the argument between group selection and methodolog-
ical individualism in the social sciences mirrors in many ways the argument
between group selection and individual selection in biological and cultural evo-
! “Note, however, that it may be difficult to meet these conditions [for GS to

lution” (Zywicki, 2004, p. 268).
work] in practice, precisely because the forces identified by the methodological

individualism model are so powerful” (Ibid. p. 271).
Both of these statements point to a tension between MI and GS – not a logical
tension, but an empirical one. If I read him correctly, Zywicki’s view is that MI
applies in some circumstances and GS applies in others, much as (for example) a
competitive model applies in some markets and a monopoly model in others. And to
extend the analogy, there exist circumstances where both MI forces and GS forces
are at work, just as there exist markets where both competitive and monopolistic
forces are at work. Thus, he resolves the alleged logical contradiction between MI
and GS by transforming it into an empirical one.
That is not, however, the way I attempt to resolve the contradiction. I do not
treat MI and GS as competing models; indeed, MI is not a model at all, but a
methodological condition that models must satisfy. I argue that GS does satisfy
that condition, once MI and GS are understood properly. My argument relies,
in part, on distinguishing between MI and individual-level selection (IS); note
that Zywicki’s first quotation above appears to equate the two. To see why they are
different, suppose we wish to model a human society that includes some individuals
Group Selection and Methodological Individualism: Reply to Comments 299
with altruistic preferences. MI requires that our predicted social outcomes result
from the choices of its members, including the altruists. That alone is enough to
satisfy MI. We might also choose as modelers to recognize the existence of an

IS mechanism that disadvantages the altruistic members of the society, possibly
reducing their representation in future generations. That might be a sensible thing
to include in our model (depending on the environment the society faces), but it
would not be required by MI. In short, MI is not equivalent to the notion that less
“fit” individuals will be weeded out of the population.
I may have misconstrued Zywicki’s position, but I hope the clarification above
will still prove helpful. I also want to thank Zywicki for drawing my attention
to a statement from Hayek (1972, p. 44) that I had not noticed before, and that
indicates Hayek’s understanding of the issues involved was even better than I had
thought. It addresses the relationship between MI and GS more directly than any
other statement by Hayek (that I know of), and it is perfectly consistent with my
own position.
3. RESPONSE TO GIFFORD
Gifford agrees with my position that the new GS (that is, the trait-group model)
is consistent with MI. His critique is that the new GS is not really GS. For those
who agree with Gifford’s position, the moral of my original article could be
restated thus: “The trait-group model of selection, which has sometimes been
misleadingly characterized as group selection, is consistent with methodological
individualism.”
However, I think there is justification for considering the trait-group model
a form of GS. The issue here is not merely semantic, because, as Okasha says,
“In devising concepts with which to describe the world, we must therefore select
some similarity relations as salient, and ignore others” (Okasha, 2001, cited in
Gifford, 2004, p. 282). To put it another way, some definitions gloss over relevant
or useful distinctions, whereas other definitions highlight them.
In the case of GS versus IS, there does exist a way to conceptualize the difference
between them so as to rule out GS from the get-go. Here is how: define IS as any
process by which some traits increase in frequency relative to other traits in a
population over time, where frequency is measured as the fraction of individual
organisms with the trait. That, in essence, is what Gifford does when he refers
to individual offspring as the “common currency” of success and regards this as
evidence of an IS process.
But defining IS in this way elides an important distinction between two
different reasons a trait might prosper relative to other traits. Is it more fit
because it enhances the ability of an individual organism to produce offspring

relative to other individuals in the same group, or because it enhances the ability
of the members of a group to produce offspring relative to members of other

groups? For the evolutionary theorist to ignore this distinction would be as
misleading as for an economist to ignore the difference between private goods and
public goods.
Gifford cites two articles by Bergstrom (2002, 2003), which show that if
intra-group interactions have the form of a prisoner’s dilemma, some kind of
assortative grouping is necessary for an altruistic (cooperative) trait to prevail
over a selfish (defecting) trait. In this context, “assortative” means that any given
cooperator has a greater chance of being grouped with cooperators than does
a defector, which could not be the case if the grouping process were purely
random. Thus, I was incorrect to say that, for small enough group size, chance
alone would produce sufficient variation among groups to allow the survival
of altruism.1
One might conclude that assortative grouping is a necessary condition for
GS. That conclusion is unwarranted, as two cases argue otherwise. First, there
are situations in which intra-group interaction is not a prisoners’ dilemma, but
wherein IS alone would nonetheless eliminate the group-beneficial trait. Suppose
some trait creates a reproductive benefit for each group member of B, and a cost
to the individual organism of C, with B > C. There is no prisoners’ dilemma,
since having the trait increases the organism’s absolute fitness. But the trait would
vanish from the group over time, because it creates a greater relative fitness for
the other group members who do not incur the cost (Bergstrom, 2002, p. 76). Yet
if groups were embedded in a process in which groups periodically dispersed into
the general population from which new groups were formed, then cooperation
could emerge as the dominant trait after all.
Second, there are situations in which multiple, Pareto-ranked equilibria exist.
In such a situation, GS can play a role in determining which equilibrium prevails.
Since a repeated prisoners’ dilemma has multiple equilibria (the Folk Theorem at
work), the applicability of this result is not limited to simple coordination games.
Bergstrom (2002, p. 81) makes essentially the same point.
4. RESPONSE TO SOBER AND WILSON

Sober and Wilson agree with my position that GS is compatible with MI – which I
consider an important achievement – but they resist MI on other grounds. Although
my paper was not intended as a comprehensive defense of MI, their objections to
it are well worth considering.
4.1. Direct Observation of Social Wholes

Sober and Wilson take issue with my claim that “we cannot, in fact, observe
[social] wholes directly; we infer their existence [only] through the apprehension
of particular events and phenomena” (Sober & Wilson, 2004, p. 255). They adduce
some examples of social wholes and generalizations about them that allegedly can
be observed, but their examples strengthen my point. Their first example is suicide
rates and their correlation with factors such as the time of year (or, they might
have added, gender, nationality, occupation, and so on). Note, however, that a
suicide rate is simply a mathematical computation of the number of individuals in
a small group (those who commit suicide) divided by the number of individuals
in a large group. There is thus no suicide rate independent of “particular events
and phenomena” – the specific acts of suicide. More importantly, suicide is not an
unambiguous category. Does it include individuals who kill themselves gradually
with cigarettes or unhealthful diets? Does it include individuals who die as a result
of deliberately chosen high-risk activities such as skydiving? Ultimately, suicide
is defined by reference to the intentions or at least knowledge of the individuals
involved – a problem that afflicted Durkheim’s sociological analysis of suicide
from the start.
Durkheim’s initial effort at such a definition indeed followed common usage, according to
which a “suicide” is any death which is the immediate or eventual result of a positive (e.g.
shooting oneself) or negative (e.g. refusing to eat) act accomplished by the victim himself. But
here Durkheim immediately ran into difficulties, for this definition failed to distinguish between
two very different sorts of death: the victim of hallucination who leaps from an upper story
window while thinking it on a level with the ground; and the sane individual who does the same
thing knowing that it will lead to his death. The obvious solution – i.e. to restrict the definition
of suicide to actions intended to have this result – was unacceptable to Durkheim for at least
two reasons. First, . . . Durkheim consistently tried to define social facts by easily ascertainable
characteristics, and the intentions of agents were ill-fitted to this purpose. Second, the definition
of suicide by the end sought by the agent would exclude actions – e.g. the mother sacrificing
herself for her child – in which death is clearly not “sought” but is nonetheless an inevitable
consequence of the act in question, and is thus a “suicide” by any other name (Jones, 1986,
pp. 82–83).
Eventually, Durkheim settled on the definition, “Suicide is applied to all cases of

death resulting directly or indirectly from a positive or negative act of the victim
himself, which he knows will produce this result” (Jones, 1986, p. 83, emphasis
in original). Yet even this definition requires reference to the mental state of
knowledge. In addition, it leaves unspecified the certainty required (is a 75%
chance of death high enough?). It also excludes attempted suicides of all kinds,
even those that failed only because the knowledge of the attempter was faulty
(“I thought it was poison!”). One way or another, the definition – and therefore
identification and measurement – of suicide is inevitably theory-laden.
Even once we have agreed upon a definition of suicide, we can still only make
generalizations about its relationship with other variables whose definitions are
also theory-laden (“nation,” “society,” “religion,” “divorce,” etc.). The generaliza-
tions might even hold true over time, giving us some degree of confidence in them.
Sober and Wilson therefore ask, “Why couldn’t data on suicide rates in one or
more societies convince us that there is a general pattern here, even if we have no
very plausible explanation of why the pattern obtains?” We can, now that we
have defined our terms. But an observed pattern is not the same as a theory. The
object of social science is to explain the pattern. If we wanted a satisfying theory
of suicide, we would presumably ask about the factors that influence individuals
decisions to end their own lives. The creation of that theory might even lead us to
modify our definitions, to collect data differently, or to collect different data.
Durkheim’s work on suicide is an excellent example of the sort of theorizing
that MI urges us to avoid. Durkheim strove to avoid explanations that relied upon
individual psychology, and as a result he posited the existence of society-level
“suicidogenic currents” whose source and laws of motion were unclear at best.
It is not mere metaphor to say of each human society that it has a greater or lesser aptitude
for suicide; the expression is based on the nature of things. Each social group really has a
collective inclination for the act, quite its own, and the source of all individual inclination,
rather than their result. It is made up of the currents of egoism, altruism or anomy running
through the society under consideration with the tendencies to languorous melancholy, active
renunciation or exasperated weariness derivative from these currents. These tendencies of the
whole social body, by affecting individuals, cause them to commit suicide (Durkheim, 1951
[1897], pp. 299–300, emphasis added).
Durkheim’s most persuasive explanations, ironically, were those making oblique

reference to the mental states of individuals; for example, that some people
commit suicide because they cannot find any goal in the world worth committing
themselves to (what Durkheim called “egoistic” suicide) (Durkheim, 1951 [1897],
p. 152ff).
4.2. The Explanatory Power of Macro Properties
Sober and Wilson argue that “macro properties” are sometimes superior to “micro
properties” in their explanatory power. There is a sense in which this is true. If
workers renegotiating their wage contracts regard the Consumer Price Index (CPI)
as a good measure of the cost of living, then it will assuredly affect the negotiations.
There is nothing un-MI about this. The notion that CPI measures the cost of living
is a constitutive idea held by individuals, and the labor economist can recognize
it without committing himself to the meaningfulness of the CPI. But if he wished
to explain the level of the CPI and show that it does indeed correspond to the
cost of living, the macro property called CPI is no longer part of the explanation,
it is the thing to be explained. The economist would have to ask how the CPI is
constructed and how it is affected by micro phenomena in the economy, such as
changing relative consumer demands for products. He would also have to define
“cost of living,” a term both theory-laden and far from self-evident.
Macro properties of social systems are not self-evident, but require social
scientific explanation themselves. If we employ them without scrutinizing their
micro foundations, we run a serious risk of unconsciously incorporating popular
but wrong theories as part of our explanations.
4.3. Generality and Specificity
Finally, Wilson and Sober conflate the distinction between MI and non-MI with
the distinction between specificity and generality. “Although generality is often a
virtue of explanations, it is not the only one; detail is often valued as well. The point
is that these virtues are in conflict . . .” (Sober & Wilson, 2004, p. ???). It is true
that generality is sometimes preferable to specificity, and sometimes the reverse.
It is untrue, however, that MI is associated only with the latter. MI theories need
not dwell on minute details; indeed, the most powerful MI theories can generate
interesting conclusions from a relatively spare set of assumptions about individual
psychology. For example, microeconomic theory predicts downward-sloping
demand curves under very broad and abstract assumptions about consumer
preferences; specific information about each and every consumer is not required.
MI constrains the locus of our assumptions, not their generality or specificity.
5. GENERAL COMMENTS
Considering all of the comments, if I were to rewrite my original article, I would

place less weight on the secondary public good argument, and more weight on the
game theoretic argument at the end of Section 4. (I presented the two arguments
together, but now I would choose to distinguish them). As Zywicki and Gifford
both emphasize, solutions to prisoners’ dilemmas, such as public good problems
and tragedies of the commons, often rely on mechanisms to limit free riding,
such as reputation, policing, and ostracism of violators. To the extent that such
mechanisms are available, a repeated prisoners’ dilemma is actually a coordination
game with multiple Pareto-ranked Nash equilibria. Given that many Nash equilibria
are possible, what reason (if any) do we have to believe superior equilibria will in
fact occur?
One possible answer is GS. If some societies happen, possibly by accident, to
develop better norms (i.e. find a superior equilibrium), they will grow larger and
faster. As a result, more future individuals will find themselves in societies with
such norms. To the extent that societies with better norms are more likely to attract
immigrants, an IS process reinforces the GS process. The immigrants will fall in
line with the norms of their adopted societies for the same reason that Americans
visiting England drive on the left: given the existence of an equilibrium, they have
an individual incentive to follow it. Other societies with inferior norms will grow
less quickly or even shrink, in part through the loss of emigrants.
The process I’ve just outlined does, in my opinion, constitute a GS story,
because the group-level properties of the superior norms explain the evolutionary
performance of the societies that have them. It is also an MI story, because
individuals act according to their preferences subject to the constraints they face
– including the equilibrium behavior of other individuals. It is even, in Gifford’s
language, a strong-MI story – that is, a story that satisfies MI (as I have defined it)
and the additional requirements of self-interest and (bounded) rationality. While
other forms of GS in cultural evolution might also work, the kind just described
is the one I find most plausible.
NOTE
1. I resisted this conclusion at first, as Bergstrom’s (2002, pp. 71–72) proof appears to
neglect Simpson’s paradox. However, Bergstrom (2003) gives a more explicit proof that
demonstrates the point.
REFERENCES
Bergstrom, T. C. (2002). Evolution of social behavior: Individual and group selection. Journal of
Economic Perspectives, 16, 67–88.
Bergstrom, T. C. (2003). Group selection and randomness, reply. Journal of Economic Perspectives,
17, 211–212.
Durkheim, E. (1951 [1897]). Suicide: A study in sociology. J. A. Spaulding & S. Simpson (Trans.).
Glencoe, IL: Free Press.
Hayek, F. A. (1972). Law, legislation, and liberty: Vol. 1, Rules and order. Chicago: University of
Chicago Press.
Jones, R. A. (1986). Emile Durkheim: An introduction to four major works. Beverly Hills: Sage.
Okasha, S. (2001). Why won’t the group selection controversy go away? British Journal for the
Philosophy of Science, 51, 25–50.

Vol. 7 - Evolutionary Psycology and Economic Theory - Koppl, R.

Загружено:

Сведения о документе

Оригинальное название

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Vol. 7 - Evolutionary Psycology and Economic Theory - Koppl, R.

Загружено:

Авторское право:

Доступные форматы

All rights reserved.

AND ECONOMIC THEORY

Series Editor: Roger Koppl

ADVANCES IN AUSTRIAN ECONOMICS VOLUME 7

Amsterdam – Boston – Heidelberg – London – New York – Oxford

ELSEVIER B.V. ELSEVIER Inc. ELSEVIER Ltd ELSEVIER Ltd

© 2004 Elsevier Ltd. All rights reserved.

Electronic Storage or Usage

First edition 2004

British Library Cataloguing in Publication Data

ECONOMICS EVOLVING: AN INTRODUCTION TO THE

ECONOMICS AND EVOLUTIONARY PSYCHOLOGY

MOGUL GAMES: IN DEFENSE OF INEQUALITY AS AN

AUSTRIAN ECONOMICS, EVOLUTIONARY PSYCHOLOGY

HAYEK AND MODERN EVOLUTIONARY THEORY

HAYEK’S THEORY OF THE MIND

IS HOMO ECONOMICUS EXTINCT? VERNON SMITH,

AUSTRIAN ECONOMICS, EVOLUTIONARY PSYCHOLOGY

THE NEW FABLE OF THE BEES: MULTILEVEL

SYMPOSIUM ON GROUP SELECTION

GROUP SELECTION AND METHODOLOGICAL

ON GROUP SELECTION AND METHODOLOGICAL

COMMENT ON “GROUP SELECTION AND

RECONCILING GROUP SELECTION AND

Douglas Glen Whitman

C. Athena Aktipis Department of Psychology, University of

Law, Emory University, USA

Alison Watkins Information Systems and Decision

Don Bellante Uskali Mäki

University of Stirling, UK Max Planck Institute, Germany

The modern revival of “Austrian economics” dates to the South Royalton

new mainstream an audience more open to Austrian ideas. The contributors

Evolutionary Psychology and Economic Theory

The contributions of Friedman and Vanberg both contain useful primers on

maximizing) today than it once was, or perhaps maladaptive” (Symons, 1992, p.

which builds in part on evolutionary psychology (Gigerenzer & Selten, 2001).

serve the selfish genes of the group’s strongest members.

evolutionary psychology. Mises modeled man as “rational” in a rather strong sense.

architecture of human thought. Austrians would do well to revise pure praxeology

introduce a kind of realism to our models of economic actors, but without

the extravagant and perplexing claims of universal hermeneutics (Koppl, 2002;

not a methodological dualist. I have already suggested that the repudiation of

mild definition of “methodological dualism” might be called for. If the doctrine is

the profession while making contributions that non-Austrians may value. He

explain and predict other people’s behaviour by attributing to them independent

Modern theories of multi-level selection do not always distinguish between

of the sciences, natural and social.

Hayek, F. A. (1952). The sensory order. Chicago: University of Chicago Press.

ECONOMICS AND EVOLUTIONARY

Economics is built on a simple assumption – that individual behavior can best be

EVOLUTIONARY PSYCHOLOGY: THE SHORT VERSION

Evolutionary Psychology and Economic Theory

our species spent most of its species history.

Researchers in evolutionary psychology, starting with these assumptions, have

WHAT EVOLUTIONARY PSYCHOLOGY ADDS

of themselves as possible into future generations.6

environment he observes around him, we expect evolutionary biological man to

reproductive success in the environment in which the psychological characteristics

Economists sometimes observe people acting in ways that appear difficult or

Behavioral Consequences of the Belief in Just Prices

Evolutionary Psychology and the Just Price

It follows that, in the context of a hunter-gatherer society (more generally, a

other human beings. One feature of that module, hard-wired in by evolution, is