HORTON, Robin, Patterns of Thought in Africa and the West: Essays on Magic, Religion and Science,

Cambridge: Cambridge University Press, 193, xi, 471pp., US$ 59.95, 0 521 36087 0

Robin Horton's collection of essays offers something to offend everyone. Highly iconoclastic,

and frequently idiosyncratic, some thirty years of Professor Horton's thinking on thought in "Africa" and

"the West" is presented in a comprehensive fashion. Readers of these papers may find themselves

occasionally railing against Horton's perspective (particularly his own critics, whom he rails against),

but often rewarded with some trenchant observations about the pitfalls of much anthropological

theorizing on religion in Africa and elsewhere.

Patterns of Thought in Africa and the West offers nine major theoretical essays written between

1960 and 1987, plus an Introduction and Postscript to the collection. Eight of the papers were published

previously; the ninth is an expanded version of the 1987 Frazer Lecture. The collection is divided into

two principal sections, whose pithy titles "Mainly Critical" and "Mainly Constructive", ably capture

their content and (especially) their tone. The chronological scope of this volume allows the reader to

appreciate the development of Professor Horton's approach to the comparative study of systems of

thought. We see his response to the challenges of his critics, responses which include the revision,

clarification, and even rejection of ideas that he no longer finds satisfactory. On the whole, however,

Horton holds his ground, and remains unpersuaded of the virtues of alternatives to his position.

Horton's dissatisfaction with prevailing interpretive frameworks for assessing African religious

thought (which he labels "Symbolist" and "Theological") lead him to propose an avowedly

Intellectualist perspective. One of the premises of this perspective is the insistence that there is nothing

unique or novel about religious or African systems of thought, but, rather, that these are examples of a

much more general, indeed, universal cognitive capacity. Given such a premise, Horton turns

necessarily to a comparison of thought systems in Africa and the West in order both to demonstrate the

underlying continuity between these systems, and to describe and account for the apparent differences

between them. His analyses, then, concern not only the common intellectual
foundations of African and Western thought systems, but also the transformation, and, indeed, evolution

of what might be called different cognitive orientations.

The cornerstone concept of Horton's Intellectualism, as derived in broad outline from Tylor and

Frazer, is that magic, religion, and science can all be characterized as bodies of theory designed for the

"explanation/prediction/control" of events in the world. It is this fundamental "goal" that constitutes the

"common core" of human rationality at all times and in all places. This focus on these cognitive

concerns leads Horton to devote the preponderance of his analysis to the question of "theory building."

How, asks Horton, can we account for the construction of different explanatory theories, in both
intellectual and sociological terms? Cognitive processes, he asserts, consist of a relationship between

two distinct kinds of theories: "Primary theory" defines the everyday world of common sense, objects,

events, and persons. Its objects are experienced as directly given. "Secondary theory" posits an order of

"unseen" forces that provide the intellectual resources for demonstrating that there is an organization and

regularity underlying the apparent disorder of everyday experience. All such "secondary theories"

attempt both to explain the ordinary experiences of everyday life, but must also rely on that primary

level of theory, drawing its systematic models by analogy with those features of commonsense

experience most often associated with order and regularity. Since, according to Horton, "primary

theory" is more or less the same universally, but the domains of experience associated with order and

systematicity vary from culture to culture, "secondary theory" can be seen as both grounded in a

universal rationality, but subject to variation relative to the social contexts of ordinary experience from

which its analogies are drawn.

These dual-dimensions of theory building provide, to Horton's mind, both the intellectualist

foundations as well as the sociological variability of all explanatory ideas, from the "wavicles" of

quantum physics, to Water-Spirits in Kalabari possession cults. He elaborates on the implications of this

cognitive model most directly, developing his ideas about both the intellectual and sociological sides of

his equation, in the papers described as "Mainly Constructive". The two most important essays in this
section, and indeed, the entire volume, "African traditional thought and
Western science" ("ATWS") and "Tradition and modernity revisited" ("TMR") are companion pieces,

the latter Horton's response to criticisms, and reevaluation of the former effort some fifteen years after

its initial publication. In "ATWS", Horton argues that "explanatory theory" must be clearly

differentiated from "common sense", or "everyday discourse", as all theory departs from commonplace

experience in an effort to place phenomena in a wider causal context than common sense can provide.

Horton subsequently rejects the "antithesis" that this formulation implies, and acknowledges that those

things that seem given to us in everyday experience are no less "theoretical" than the phenomena defined

by either science or religion. He therefore rejects the antithesis, but retains a distinction between, as we
have seen, "primary" and "secondary" theories. Horton, in fact, goes even further in "TMR" in his

description of "primary theory"/everyday experience, by insisting that this level of theory does not differ

very much from culture to culture, and that, indeed, its foundations may be a part of our genetic make-

up.

The second principal, and in many ways more substantial revision of "African traditional thought

and Western science" concerns the sociological dimensions of theory building, specifically the

conditions that give rise to the "secularization" of theory. In "ATWS" Horton, invoking a contrast first

posed by Karl Popper, argues that the discontinuities between African religion and Western science can

be explained by the fact that

in traditional cultures there is no developed awareness of alternatives to the established body

of theoretical tenets; whereas in scientifically oriented cultures it is highly developed. It is this
difference we refer to when we say that traditional cultures are 'closed' and scientifically oriented
cultures are 'open'. (222)

This fundamental contrast generates an extreme conservatism in closed societies, such that any

challenge to received wisdom, which is held to be "sacred", produces an intense "anxiety" in the

members of that society. Open societies, in contrast, see their prevailing ideas as merely one possibility

among numerous alternatives, and, rather than feeling anxious about possible threats to theory, relish the

opportunity to produce new and better ones. Thus, the "cognitive superiority... of the modern, Western

world-view" (383) in terms of its explanatory abilities, can be accounted for by reference to this
willingness to consider all manner of theories in the quest for evermore adequate causal contexts.
 Upon reflection, in "TMR", Horton rejects the "open"/"closed" dichotomy as both to extreme and

too individualistic. African's cannot be said to be completely unaware of alternative theories, and are

certainly more critical and skeptical than the "closed" predicament suggests; moreover, the history of

Western science makes it plain that there is a good deal of conservatism and its attendant anxiety when

well-established theoretical frameworks are challenged. Nonetheless, Horton insists that the germ of

this contrast remains valid, but in place of the "open"/"closed" dichotomy he proposes a distinction

between different cognitive orientations: "traditionalism" and "modernism". "Traditional" societies may

not be unaware of alternatives, but they attempt to assure a consensus about knowledge that has been
validated by its transmission from ancestral sources. "Modern" societies compete in the process of

producing new theories, validated by the presumption that knowledge is progressively improving. The

sociological foundations of "traditional" and "modern", as opposed to "closed" and "open" societies,

then, have shifted only to a degree, and with very little consequence for the kinds of knowledge

produced under these "traditional" as opposed to "closed," or "modern" as opposed to "open" conditions.

Horton is to be commended for his intellectual honesty and self-scrutiny; yet his new approaches seem

to rise phoenix-like from the ash-can of his discarded ideas. Finally, in both "ATWS" and "TMR"

Horton suggests that the breakdown of tradition and the rise of modernity can be attributed to

transformations in "technology, economy and social organization," primarily literacy (a variable whose

relevance Horton calls into question in his Postscript) and increasing cultural heterogeneity (which

Horton claims was not characteristic of "Traditional African communities" (252 ff; 380).

These are Horton's "constructive" arguments - both original and revised - about the intellectual

and sociological dimensions of theory building. His "Mainly Critical" papers repeatedly defend this

argument against his critics, and attempt to demonstrate its superiority to two alternative analytic

perspectives, the "Symbolist" and the "Theological." Both of these perspectives, according to Horton,

begin with the premise that religious thought is fundamentally different from scientific thought. For the

Symbolist, a contrast is drawn between "expressive" and "instrumental" thought, and religion is placed
squarely in the expressive field. Religious statements from this perspective
are not meant to be taken literally, but must be understood as referring to, or "symbolizing" some other

aspects of experience that cannot be directly conceptualized by instrumental/scientific thought.

Expressive thought may serve certain social functions when the experiences they refer to are social

groups, or institutions that need to be supported; alternatively Horton asserts that most symbolists hold

that religious thought is focused on "the production of symbolic images for their own sakes" (6; 119). In

Horton's view the Symbolist position simply flies in the face of ethnographic evidence to the contrary.

When, for example, a diviner diagnoses a client's illness to be the product of sorcery, she is attempting to

explain events in the real world. Intellectualists see no difficulty in taking this procedure at face value;
and from Horton's perspective, the Symbolist offers no evidence to support the claim that religious

adherents and practitioners of magic "really mean something else" by such statements and actions.

Symbolists simply dogmatically assert that religious thought is intrinsically expressive, and fail to

acknowledge the eminently pragmatic concerns of most religious practice. Moreover, the Symbolist

opposition between "expressive" and "instrumental" thought, Horton suggests, is drawn from a

misreading of Durkheim's contrast between the "sacred" and the "profane" as different kinds of thought.

But Symbolists here have "stood [Durkheim's] teaching on its head" (100). A proper reading of

Durkheim, according to Horton, should recognize that he sees a continuity between religion and science

as forms of theoretical reasoning that describe underlying causal relations in order to account for the

vagaries of commonplace experience.

The "Theological" perspective (or in one of Horton's less generous characterization, "The Devout

Opposition") holds that all religious thought, Western and African included, forms a unified discourse

which allows the analyst to interpret or "translate" all religious claims according to a shared set of

concerns. This suggest that all religions have as their primary focus a supreme being, whose

worshippers adopt an attitude of reverential awe in their attempts to achieve communion with this

ultimate spiritual power. Again, argues Horton, such a conception runs roughshod over much African

ethnography, which clearly demonstrates the impossibility of seeing all spiritual beings, from Shango to
neighborhood witches, as different manifestations of a single, benevolent, all-powerful deity. With

respect to the purported pursuit of communion as a common

feature of all religions, Horton boldly recapitulates his Tylorean mantra, "In work after work we find that

explanation, prediction and control are the overriding aims of religious life" (177).

Furthermore, Horton concludes all of his critical essays with ideological critiques - which he

calls "diagnoses"- of the Symbolist or Theological theorists he examines. In each case, he finds a

variety of Romantic Liberalism at work, an agenda that is loathe to allow invidious comparisons

between the cognitive efficacy of the West and the Rest. Rather than seeing African traditional thought

(magical, religious or otherwise) as a form of universal rationality, which leads to the inevitable

conclusion that traditional thought is (at least) less successful than modern thought, both the Symbolists
and the "Devout" insist that religion and science are different kinds of thinking, thereby preserving

(because of their incomparability) the integrity and equality of all thought systems. It is this ideology

which blinds its adherents both to the ethnographic proof of all thought's underlying concern with

"explanation/ prediction /control", and to the strength of Horton's ripostes.

I have gone on at some length with this summary of Horton's perspective in the hopes of giving a

fair and representative account of his views. There are a number of strengths both to his critique of

frameworks that fail to account for the ethnographic data they address, and to his assertion that

pragmatic thinking is not the exclusive concern of "modern" science. And this book will clearly be of

great interest and use to scholars and students of comparative rationality. At the same time, there are a

whole host of objections to Horton's theory and method that are worth raising. To begin with, there are

serious editorial problems with this volume. It is understandable in a collection of essays over 400

pages in length that there should be some overlap in certain areas of discussion and commentary. But

most of the essays in any given section address the same concerns, present the identical structure of

argumentation, and reproduce stock phrase, examples, and entire sentences. As a group of papers this

compilation allows the reader judiciously to select an insightful essay or two; as a synthetic volume it is

often so redundant as to become tedious.

There are more substantial points of contention, as well. While many of Horton's criticisms are
well-founded (aside from being unabashedly polemical and ill-mannered- e.g., some theories are
the "palpably inadequate" products of "the narrowly educated man" [185-187]), these critiques

occasionally offer simpleminded characterizations of his opponents. To assert, as he does, that

contemporary accounts that do address the explanatory dimensions of ritual practice "feel compelled to

'dress up' the subjects of these reports in the latest Parisian semeiological/ semantic fashions in order to

make them respectable" (441) suggests that Horton not only fails to see the relevance, but perhaps fails

simply to understand these theoretical alternatives. It may be glib to dismiss others' analyses as pret-a-

porter theorizing, but these "fashionable" arguments are never actually engaged by Horton.

It should also be clear that Horton's developmental framework is inherently teleological. Western
rationality may not be superior to traditional thought, because, in fact, it is no different from it. But,

there can equally be no denying that "cognitive 'modernism'" yields superior explanatory results (343-4;

382-3). The West may not think any differently from the Rest, but its way of theorizing does bring it

closer to "the truth". Horton goes to great lengths to point out that he is not a positivist on such matters.

And with respect to logical positivism, as it developed in the Anglo-American analytic philosophy of

Russell and Ayer, this is certainly the case. That is, Horton does not argue that true knowledge can only

be the product of induction from empirical observation; rather, he insists that theoretical models, and

unobservable principals are essential for any rational act of cognition. But positivism also has a

complementary intellectual legacy derived from the social theory of St. Simon and Comte. This legacy

presumes a reality that exists prior to and independent of our efforts to understand it, a reality which ever

advancing theory will come closer and closer to revealing. Cognitive yields therefore increase, goes the

saying, as theory becomes more and more "value-free." It is this version of positivism and its attendant

teleology that Horton's intellectual and sociological models most clearly describe.

I would suggest that this teleology also lends insights into Horton's counter-reading of Durkheim.

The core of Horton's discussion of Durkheim which argues that religion and science are both forms of

explanatory theory is insightful and productive. But, while Durkheim's intellectual heirs may have

"stood his actual teaching on its head," Horton misses his point. For Durkheim,
both science and religion are conceptual systems grounded in collective life. They are not merely

alternative intellectual theories, nor cognitive mechanisms, they are collective representations generated

by distinct kinds of social experience. Thus, in recent years scholars interested, as Horton is, in these

continuities see less of a divide between science and ritual, but begin by looking at both science and

ritual (to say nothing of the rituals of science) as historically constructed forms of social practice and

knowledge. This is clearly not the way in which Horton combines the cognitive and sociological

dimensions of theory building. Rather, he see cognitive efforts "underpinned" by "technology, economy

and social organization." That is, social life provides a set of material conditions that constrain (or even
distort) cognitive processes, but the emerging recognition and promotion of alternative theories, made

possible by modern social experience, allows for the increasingly untrammeled exercise of reason.

Most of these difficulties emerge from Horton's failure to develop or address issues of culture,

the means by and through which people understand and act on (or "explain/predict /control") themselves

and the world they inhabit. It is probably the case, as Horton ceaselessly insists, that all peoples are

interested in explaining illness, misfortune, and other problematic events. But how far does that insight

alone get us into understanding just how different peoples, at different times and places propose

radically different means of doing just that? Horton's inattention to culture makes it impossible for him

to address anything beyond a common underlying cognitive faculty that social conditions progressively

permit to expand its horizons; surely this leaves unanswered essential anthropological questions of how

people define their world and its constituents, and what specific practices and authorities make

knowledge of that world possible .

College of William and Mary, USA BRAD WEISS