Wesleyan University

Some Afterthoughts on Culture and Explanation in Historical Inquiry

Author(s): Chris Lorenz
Source: History and Theory, Vol. 39, No. 3 (Oct., 2000), pp. 348-363
Published by: Wiley for Wesleyan University
Stable URL: http://www.jstor.org/stable/2678016 .
Accessed: 19/06/2014 21:29

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

Wiley and Wesleyan University are collaborating with JSTOR to digitize, preserve and extend access to History
and Theory.

http://www.jstor.org

This content downloaded from 62.122.79.21 on Thu, 19 Jun 2014 21:29:01 PM

All use subject to JSTOR Terms and Conditions
History and Theory 39 (October 2000), 348-363 ? Wesleyan University 2000 ISSN: 0018-2656

FORUMON CULTUREAND EXPLANATIONIN HISTORICALINQUIRY

4.

SOME AFTERTHOUGHTS ON CULTURE AND EXPLANATION

IN HISTORICAL INQUIRY

CHRIS LORENZ

ABSTRACT

I arguehere that the articlesin this forum containbasic agreements.All threereject natu-
ralism, reductionism,and monism while retaining causality as an explanatorycategory,
and all threeemphasizethe role of time and arguefor a view in which cultureis regarded
as both structuredand contingent.
The differences among the explanatoryproposals of Hall, Biernacki, and Kane are as
importantas the similarities:while Hall favors a Weberianapproach,Biernackiarguesfor
a primarilypragmaticexplanationof culture,and Kane for a primarilysemiotic explana-
tion. I argue that all three positions face immanent problems in elucidating the exact
natureof culturalexplanation.While Hall leaves the problem of "extrinsic"ideal-typical
explanationunsolved, Biernacki simply presupposes the superiorityof pragmatic over
other types of cultural explanation, and Kane does the same for semiotic explanation.
Hints at culturalexplanationin the form of narrativeremainunderarguedand are built on
old ideas of an opposition between "analysis"and "narrative." This is also the case with
the latest plea for "analyticnarratves."I conclude that a renewedreflection on this oppo-
sition is called for in orderto come to grips with culturalexplanationand to get beyond
the old stereotypes regarding the relationship between historical and social-scientific
approachesto the past.

When I volunteered to comment on this forum, my main motive was curiosity.

Having been trained as a sociologist some time ago and having turned myself
into a philosopher of history in the meantime, I was curious about how three
prominent American historical sociologists tackled one of the central problems
of philosophy of history and of the social sciences, that is, the problem of his-
torical explanation. This will be the focus of my comments, which I have subdi-
vided into three parts. First, I shall chart some basic agreements between the
methodological positions of the three panelists. Second, I will identify some
important methodological differences. And third, I will formulate some method-
ological problems raised by the papers that are unresolved as yet and that might
stimulate further reflection.

I. BASIC AGREEMENTS

I first want to chart the common ground in the philosophy of history occupied by
Hall, Biernacki, and Kane.

This content downloaded from 62.122.79.21 on Thu, 19 Jun 2014 21:29:01 PM

All use subject to JSTOR Terms and Conditions
 SOME AFTERTHOUGHTSON CULTUREAND EXPLANATION 349

They all argue against all forms of naturalism, reductionism,and monism.

Hall rejects the monistic explanatoryclaims of positivism explicitly (346-347),
and Biernacki does the same for the monistic explanatoryclaims of Geertzian
culturalism.Kane is less explicit on this subject,althoughin an earlierversion of
her papershe too statedexplicitly that"it has been effectively demonstrated"that
reductionism,culturalor otherwise, is invalid.'
The second common characteristicof the papers is that, althoughthey reject
naturalism,they do not reject causality as a major explanatorycategory in the
humansciences, as is often the case with anti-naturalistsin history.Insteadof the
rigid and positivistic covering-lawview of causal explanation,they all subscribe
to comparisonas the only way to validate singularcausal explanations-and in
this respect they all subscribeto Weber'smethodologicalposition.2
The thirdcommon characteristicis that they all emphasizethe role of the time
dimensionin explanation.However,exactly what this role is is less obvious to me
and is in need of furtherclarification,especially because the differencebetween
the historical and the social-scientific approachesis often located in the time
dimension.
The fourthand last common characteristicof the papersis thatthey all explic-
itly argue for explanatory pluralism, that is, for multi-factor explanatory
approaches.They all argue against mono-causalexplanatorystrategies,as most
sensible persons nowadaysdo. Nevertheless,they all arguethatone essential and
irreducibleexplanatoryfactor is the culturalfactor.All three authorshold that
explanationin historical inquiry cannot proceed without taking culture, that is,
meaning,into account.And all subscribeto the view thatthose culturalmeanings
are neithersomehow determined,nor wholly accidental:they hold thatcultureis,
in a fundamentalsense, both structuredand contingent.
These four common characteristicshighlight the commonalities with Max
Weber'sposition-that of good old sinnverstehendeand comparativesociology.
Thereforeit is no accident thatWeberis very present in Hall's and-to a lesser
extent-in Kane's and in Biernacki'spapers.Hall seems even to be advocatinga
returnto Weber'sideal-typicalmethodology,and Kane's remarkson the "causal
significance"of culturealso have a clear Weberianring. Her explicit and repeat-
ed emphasis on the indispensabilityof "meaningconstruction"and "symbolic
structures"is close to Weber'sfocus on "worldviews"(Weltbilder).Since Weber
has been the patronsaint of modernhistoricalsociology for some time now, this
may hardlybe surprising.However this may be, it provides a focal point for me
to analyze and comparethe papersin relationto Weber.Before thatI will firsttry
to identify some interestingdifferences among the three papers.

1. Anne Kane, Ft.Worth-Social Science History-paper (original version), 1. All parenthetical

page referencesin my text are to the versions publishedin this issue of History and Theory.
2. For the problemsof causal explanationin historysee my book: Konstruktionder Vergangenheit:
Eine Einfirung in die Theorie der Geschichte (Cologne, Weimar,and Vienna, 1997), esp. 189-323,
and my article: "Can Histories Be True? Narrativism,Positivism and the 'MetaphoricalTurn,"'
History and Theory37 (1998), 317-320.

This content downloaded from 62.122.79.21 on Thu, 19 Jun 2014 21:29:01 PM

All use subject to JSTOR Terms and Conditions
350 CHRIS LORENZ

II. IMPORTANTDIFFERENCES

One of the most importantdifferences among the papers is their conception of

culture.Most of the time Biernackicriticizes "essentializing"views on culture,
which he attributesto the godfathersof the "culturalturn"in generalandClifford
Geertzin particular.But while Biernackiemphasizes the analyticaland nominal
statusof culturein historicalexplanations,both Hall and Kane subscribeto a less
nominal,more realist view of culture.
Hall's more realist view of culturecan be derivedfrom his characterizationof
his central topic, collective religious suicide. Collective religious suicide, he
says, is "nota singularsocial phenomenon,"but consists of "two alternativecul-
turalstructures"with alternative"culturallogics" (345-346). Culture,according
to Hall, is definitely not merely an analyticalconstructof the sociologist, but a
process in historicalreality itself. Kane defends a similarview of culture,stating
that "culture'sautonomyrests on the metaphoricnatureof symbols and the pat-
ternedrelationshipof symbols within a structure"(314).
This is not the only interestingdifference among the papers.Next to the dif-
ference with regardto the "essence"of culture-real accordingto Hall and Kane,
or nominal according to Biernacki-is an interesting difference between
Biernackiand Kane with regardto the questionwhethercultureand meaningcan
fruitfullybe analyzedby historiansas a semiotic system. While Biernackiinvests
a considerable amount of intelligence and energy in deflating the Saussurean
view for historicaluse, Kane flatly states the opposite view: "Meaningconstruc-
tion must be analyzedin thefirst and last instance in referenceto the internal,or
the semiotic, structureof symbolic systems."Consequentlyshe locates "thelocus
of meaning"in "symbolicstructures"(314, italics mine). Her analysis of the Irish
Land Movement in terms of symbolic structures,later identified as metaphori-
cally structured,identity-forming,and legitimacy-creatingnarratives,is explicit-
ly legitimatedby her semiotic view of culture.
Biernacki, in contrastto Kane, proposes that the semiotic model of cultural
meaning must be supplementedby a pragmatic and contextualmodel of mean-
ing construction.At the same time he opposes the reduction of meaning con-
structionto intended meaning and its unintendedconsequences, as Weber and
Hall would have it. New, and interesting,is his proposal-based on analyses of
BenedictAnderson,KarlMarx, and his own research-to regardthe structureof
practice itself as an unintendedand unconscious locus of meaning-construction
for the actors.His message in this context seems to be that there is always more
meaning-constructiongoing on thanmeets the eye of the actors.This view brings
Biernacki surprisingly close to one of the godfathers of the "culturalturn,"
Michel Foucault,and his concept of discourse.
So the concepts of culture and meaning in the three papers are not the same.
While Hall adheresto a Weberianconcept of culture,in which meaningis locat-
ed in the subjective interpretationsand the intentionsof actors and their world-
views, Kane adheresto a semiotic concept of culturein which meaning derives

This content downloaded from 62.122.79.21 on Thu, 19 Jun 2014 21:29:01 PM

All use subject to JSTOR Terms and Conditions
 SOMEAFTERTHOUGHTSON CULTUREAND EXPLANATION 351
from the mutualrelations of signs in a system and not from actorial intentions
andinterpretations.Biernacki'spragmaticconcept of culturetakes a kind of mid-
dle position between these two, emphasizingboth the semiotic meaning and the
pragmaticuse of systems of signs (302).
This difference among the concepts of culture in the three papers must and
does have consequences for the concept of explanationin each paper.Now it's
time to take a closer look to see what is going on behind the curtainsof the cozy
"culturalconsensus"on explanatoryfundamentals.Let's startwith Hall.

Hall on explanationin sociohistorical inquiry

a. "Intrinsic"and "extrinsic"history
Hall introduces a distinction between the "intrinsic"(that is, actor-oriented)
approachand the "extrinsic"(that is, observer-oriented)approachin historical
inquiry. While he attaches the "intrinsic"approach to "specific history,"he
attachesthe "extrinsic"approachto "configurationalhistory":
The practiceof "specifichistory"can be defined as investigationof an object's "intrinsic
meaning" whereas "configurationalhistory" is directed toward extrinsic, theoretically
constructedobjects. That is, specific historyis based on the verstehendeprojectof identi-
fying sociohistoricalobjects in terms of their meaning to historicallylocated individuals
who participatedin the events being emplotted.The boundarybetween specific and con-
figurationalhistorycan be located by examiningthe point at which it becomes difficultto
arguethat an object of inquiryis constitutedin relationto intrinsicmeanings. (333)

"Configurationalhistory"is thus not limited to the horizon of the historical

actors,while specific historysupposedlyis. Therefore,accordingto Hall, the rela-
tions between events in specific history are relations of meaning that can be
framed in the temporalorder of narrative,while configurationalhistory "must
employ some other structuraldevice of plot, series, or analysis to arrangea dis-
cussion of culturalobjects and practices(for example, culturalhistoriesof priva-
cy in differenthistoricalepochs, sexualityover centuries,dietingin heterogeneous
circumstances)"(334). Now Hall's examples make abundantlyclear that, in the
case of configurationalhistory,he is thinkingof long-termphenomena,which are
identifiedex post by the historianand which could not be experiencedas such by
the actorsin the past nor explainedby theirintentionsbecause the configurational
phenomena simply transcendthe actors' experientialhorizons. Therefore Hall
feels entitled, as many sociologists before him, to contrastmeaning, temporal
order, and narrativeon the one hand with structure,configurationalorder, and
analysis on the other.And he feels entitledto link the firsttrio of concepts to spe-
cific history,and the second trio of concepts to configurationalhistory.
Although these dichotomies may be fine in theory, Hall is sufficiently
acquaintedwith historical studies to distrustthem in practice. On the basis of
Lynn Hunt's Politics, Culture,and Class in the French Revolution (1984), he
immediatelystartsrelativizinghis theoreticaldichotomies.In Hunthe detects "a
series of hermeneuticand structuralanalyses of specific materials"(334). And
while hermeneuticshas been attributedto specific history, and structuralanaly-
sis to configurationalhistory,Hunt'sstudy,which primafacie looks like an indis-

This content downloaded from 62.122.79.21 on Thu, 19 Jun 2014 21:29:01 PM

All use subject to JSTOR Terms and Conditions
352 CHRIS LORENZ

putablespecimen of specific history, is relabeled as "a transitionalcase" (334).

Hall also offers a remarkableexplanationfor its "transitional"status:
Hunt's study thus is a demonstrationthat detailed event history can employ a metastruc-
ture of extrinsic analysis even though the phenomena described might reasonably be
asserted to have intrinsic linkages. Perhaps with richer data, Hunt would have pursued
specific history more concernedwith intrinsiclinkages. In any case, her study employs a
procedurethat addresses a recurrentproblem of specific history. Often, the historical
recordis too thin for inquiryto tell "whathappened"throughnarrative,even though an
intrinsicallyorderedplot could be constructedif adequatehistoricalevidence were avail-
able. (334)

Hall's commentson Hunt are revealingfor several reasons.First, the theoreti-

cal notion that specific history is just a "plain"and chronologically structured
narrativeof subjective meaning immediately collapses as soon as confronted
with historicalpractice. All historicalaccountsof intentionalactions are embed-
ded in accountsof their "structural"contexts-institutional, ideological, mental,
material,and so forth-and these latteraccountsdon't have a temporalstructure
themselves, as Hall himself acknowledges.4This "dual structure"of narrative
history,here exemplifiedby Hunt,was alreadyobservedby the historicalsociol-
ogist PhilipAbramsback in 1982 in his attemptto pinpointthe characteristicsof
historical and of sociological inquiry:"Whattypically passes as narrativehisto-
ry is containedby and itself contains analysis.... So, if Abramsis right, as I
thinkhe is, all historicalnarrativeis "transitional."
Hall could also have found the same type of argumentin the theoreticalanaly-
ses of narrativeprovidedby Aristotle, Johann-GustavDroysen, William Walsh,
Louis Mink, FrankAnkersmit,and HaydenWhite.6All these theoristsof narra-
tive emphasizedthat narrativeas a form of explanationdoes not have a temporal
structure;instead, narrativesare structuredby plots (Aristotle), "centralideas"
(Droysen), "colligatory concepts" (Walsh), "configurationalcomprehension"
(Mink), metaphors(Ankersmit),or tropes (White).7Most theorists of narrative
have emphasizedthatnarrative,contraryto the popularconception, does not fol-
low the time arrow,but constructsa synchronicZusammenhangconnectingtem-

3. Cf. John R. Hall, Cultures of Inquiry. From Epistemology to Discourse in Sociohistorical

Research (Cambridge,Eng., 1999). In this book Hall distinguishesfour basic discourses in the field
of sociohistoricalinquiry:1. The discourse on values; 2. Narrativediscourse;3. Discourses of social
theory; and 4. Discourse of explanation and interpretation.He emphasizes that in the practice of
inquirythese discourses usually form mixes and bricolages.
4. Cf. Hall, Culturesof Inquiry,67: "Atthe extreme,the intrinsicnarrativeis a story told in the first
person, from one or anothersubjective viewpoint in the course of unfolding events."Hall, however,
acknowledges on page 93 that even in intrinsic narrative"at least some narrativeconnections are
extra-intrinsicproductsof inquiry,not derivativesof causal or meaningfulsequencesin unfoldinghis-
tory."This, however,does not lead him to a reconsiderationof the distinctionbetween intrinsicand
extrinsicnarratives.
5. P. Abrams,Historical Sociology (Ithaca,N.Y., 1982), 305.
6. Hall in Culturesof Inquirydrawsheavily on White's andRicoeur'stheoriesof narrative,although
he does not deal with this aspect of their theories.
7. For an overviewof the positions in narrativetheory,see Lorenz,Konstruktionder Vergangenheit,
127-189, and Historyand Theory:ContemporaryReadings, ed. B. Fay et al. (Oxford, 1998).

This content downloaded from 62.122.79.21 on Thu, 19 Jun 2014 21:29:01 PM

All use subject to JSTOR Terms and Conditions
 SOME AFTERTHOUGHTSON CULTUREAND EXPLANATION 353

porally very disparateevents and thus creates its own temporality.That Mink
characterizesthe historical mode of understandingas "configurationalunder-
standing"-and thus uses the very concept of configurationfor history tout court
and not just for a "social-scientific"type of history,as Hall does-is as telling as
can be.8 So synchronicityratherthan diachronicityis typical for narrativeexpla-
nation, as Mink emphasizes explicitly. Therefore,I find Hall's contrastof narra-
tive versus analysis-and its concomitantcontrastof specific versus configura-
tional history-questionable as a basis for genre distinctions.
The second "revelatory"aspect of Hall's comment on Hunt concerns his
explanationof Hunt's use of narrativestrategies. Hall's idea is that Hunt took
"refuge"in configurationalhistory just because "the historical record was too
thin to tell 'what really happened' through narrative"(334); this betrays three
importantquestionablepresuppositions.
The first problematic presupposition is that historians, ideally, produce a
"pure"and "intrinsic"narrativehistorythatconsists of the history as it was expe-
rienced by the historical actors, a history that is temporallystructuredand does
not contain any "configurationalimpurities."This implies that history is basical-
ly the "untoldstory"of the actors,and thatit is the historian'stask to (re) tell this
story. Moreover,since "intrinsic"history is ideal history, "extrinsic"history is
only second best. In my view, however, there is not one ideal model of history;
neither is there one ideal relationshipbetween "intrinsic"and "extrinsic"per-
spectives in narratives.A demographicor economic history is as legitimate as a
biography,and each containsboth intrinsicand extrinsic elements.
The second problematicpresuppositionis thatthereis a "natural"or ideal rela-
tionship between researchand writing in history: a "thick"historicalrecord, in
Hall's view, correspondswith a normalnarrativeform. This presuppositioncon-
flicts with the well-established "relativeautonomy"of the composition phase
(Darstellung)vis a vis the researchphase (Forschung)in history.Therefore,the
form of representation in history is always underdeterminedby historical
research.Researchonly determinesthe limits of historicalrepresentation.9
The idea of a "determinative"relation between research and composition in
history has been discredited by all the abovementionedtheorists of narrative,
who have recognized that narrativeis basically a retrospectiveconstructionand
selection of the historian.Although Hall recognizes, with Schutz, the fact that
"forthe acting subject,even in relationto 'the same' events, meaningis unstable,
contextual,and dependenton temporalperspective"(337), he does not drawcon-
clusions from the fact-although statedby him elsewhere10-that the same holds
8. L. Mink, "Historyand Fiction as Modes of Comprehension,"in History and Theory:Contem-
porary Readings,ed. Fay et al., 121-137.
9. See Probing the Limitsof Representation:Nazism and the "Final Solution,"ed. S. Friedldnder
(Cambridge,Mass., 1992);andC. Lorenz,"YouTellYourStory,I Tell Mine: Some Reflectionson Truth
andObjectivityin History,"OsterreichischeZeitschriftffirGeschichtswissenschaften10 (1999), 563-585.
Interestingly,Hall could have referredto Droysen, who held thata narrativerepresentationof historyis
dependenton the possibilityof reconstructinga continuousdevelopment,andthatcontinuitypresuppos-
es a continuousrecord.See J.-G. Droysen,Historik,ed. P. Leyh (Stuttgart/BadCanstatt,1977), 234.
10. Cf. Hall, Culturesof Inquiry,91: "Narration,both in lifeworldly conduct and in inquiry,is a
narrationof narration,the interpretationof reinterpretationss"

This content downloaded from 62.122.79.21 on Thu, 19 Jun 2014 21:29:01 PM

All use subject to JSTOR Terms and Conditions
354 CHRISLORENZ

true for the historians who constructnarratives.If he had realized that it is the
historianwho determinesthe relationshipbetween the "extrinsic"(retrospective
or anachronistic)perspectives and the "intrinsic"perspectives of the historical
actors,he would not have tried to connect the narrativeperspectivesdirectly to
"thestate of the record."
Hall's thirdproblematicpresuppositionis his idea that the "thickness"of the
historicalrecordis a simple "given,"and thus not a notion relative to the ques-
tions asked and the representationalgoal. This presuppositionis not correct: a
biography,for instance,presupposesa differentkind of recordthan a world his-
tory,andthe same recordthatmay be regardedas "thick"for a world historymay
be very "thin"for a biography.The qualityof the recordis thus wholly question-
andframe-dependent,and the idea of a "natural"relationbetween the state of the
recordand narrativerepresentationa chimera.
b. Configurationalhistory and explanation
Now it's time to take a closer look at Hall's configurationalhistory.This type
of history is characterizedby Hall as analytical,theoretical,and as workingwith
theoreticallyconstructed"sociohistoricalmodels" of an ideal-typical character
(336). It is later exemplified in his two models of collective religious suicide.
According to Hall configurationalhistory aims at the constructionof models or
ideal types that transcendspecific histories and the "intrinsicmeaning"horizons
of the actors.They bring out the "extrinsic"generalcharacteristicsof "intrinsic"
specific histories. Framedin Jon Elster's terminology,Hall holds that configura-
tional history deals with the sub-intentionaland the supra-intentionalaspects of
social action, while specific historydeals with its intentionalaspects11:"Whereas
the balance of inquiryin specific history is tipped towardusing narrativeto ana-
lyze unfolding situationsas they are meaningfulto the actors involved, configu-
rationalhistory tilts towardanalyzing the structuralinterplayof diverse events
and phenomenaand their unintendedconsequences"(335). Configurationalhis-
tory is thus presentedas the core business of "analytic"social scientists,just as
special history appearsas the core business of "narrative"historians.'2
This Weberianview of the interdisciplinarydivision of laborhas a long pedi-
gree, as does the view that the sociological models somehow fulfill an explana-
tory function in relationshipto specific histories. This view, however, contains
several unresolved problems that are inheritedby all who adopt the Weberian
position.Let me specify two of these problemsthatoccur in Hall's paperand that
are centralto the topic of cultureand explanation.
The first problemis in what sense models and ideal types can be regardedas
explanatory.In Hall's case this problemboils down to whetherhis ideal types or
models of collective religious suicide fulfill explanatoryfunctions vis a vis the
cases. Hall's statements regarding this fundamentalquestion remain ambigu-

11. Jon Elster, Explaining Technical Change: A Case Study in the Philosophy of Science
(Cambridge,Eng., 1983), 84ff.
12. In Culturesof Inquiry Hall connects configurationalhistory firmly to social theory. See, for
example, the scheme on 206.

This content downloaded from 62.122.79.21 on Thu, 19 Jun 2014 21:29:01 PM

All use subject to JSTOR Terms and Conditions
 SOME AFTERTHOUGHTSON CULTUREAND EXPLANATION 355

ous.13 On page 346 he characterizesideal types, just as GuentherRoth did, as

"benchmarks. . . useful for comparativeanalysis;"on page 345 he refers to two
of his own ideal types as "interpretivemodels ratherthan causal explanations;"
and on page 342 he states: "This model [of collective religious suicide] comes
close to a generic descriptionof what happenedin Jonestown,. . . Waco, and . .
. Japan."Now the importantquestion is, whether and if so, in what sense
"interpretivemodels"and "genericdescriptions"can be regardedas explanatory.
If the models are derived from the cases under scrutiny,this seems an unlikely
option, unless we presume that historical cases explain themselves.'4 Or is the
relevantpoint for Halljust thathis models show "thepotentialfor theorizingeven
highly contingentsociohistoricalphenomena"?My objectionto thatmove would
be the question whetherall theorizingin the sociohistoricaldomain is not based
on the presuppositionof contingency.And if the point of theorizingis not expla-
nation, what is it then? Since Hall is addressing the logic(s) of inquiry and
whethernarrativeis explanatory,my question is on his table; I am not sure the
paperprovides an answer.15
Reading Hall, I almost felt temptedto play the positivist devil's advocate and
ask whetherit is not the business of social science to formulatetheoreticalinech-
anisms which explain what is otherwise only describedby historians.16 Or is the
distinctionbetween descriptionand explanationjust a supersededfigment of the
positivist imagination?If so, is the hermeneuticview victorious, according to
which descriptionis a form of interpretationand interpretationis a form of expla-
nation,not in need of any theoreticalsupplement?Is Geertz's"thickdescription"
really all there is in social science? Or is Philip Abrams's suggestion perhaps
right "thatMax Weber'scelebrateddivorce of history and sociology was perhaps
premature,indeed, possibly not a necessary separationat all."'17
Hall surely blursthe borderlinebetween explanationand interpretationalmost
completely.'8This "ecumenical"move, however, has its price, one not recog-

13. See for instanceHall, Culturesof Inquity,217: "Theidentificationof the sociohistoricalobjectand

the componentchangestheorizedas its basis yields an analyticframeworkfor the causalor interpretive
investigationof configurationalchangeby throwinginto relief the issues thatneed to be pursued."
14. Hall's remarkabout "avoidinggeneralizingfrom single instances or cases that have mutually
influencedone another"(340) indicates that he is awareof the problem.
15. In Culturesof Inquiry,98-103, Hall deals with several authorswho claim that narrativeper se
is explanatory,but refrainsfrom presentinghis own position, as far as I can see.
16. For a recent overview of the debate and for a defense of the mechanisms view, see Axel van
den Berg, "Is Sociological Theory Too Grandfor Social Mechanism?,"in Social Mechanisms:An
AnalyticalApproachto Social Theory,ed. P. Hedstr6mand R. Swedberg (Cambridge,Eng., 1997),
204-237. See also E. Kiser and M. Hechter,"TheRole of GeneralTheory in Comparative-Historical
Sociology,"AmericanJournal of Sociology 97 (1991), 1-30.
17. Abrams,Historical Sociology, 188.
18. Cf. Hall, Cultures of Inquiry, 150-151, for his view on explanation and interpretation.
Remarkable,given the traditionalpositions and debates, is his almost complete identificationof both
intellectualoperations.The only distinctionhe makes is with regardto theircriteriaof evaluation:"the
differencebetween interpretationand explanationdoes not derive from logic. Rather,the two projects
operateunderdifferentconditions. Under a regimen of interpretation,multiple criteriafor adjudicat-
ing amongaccountscoexist, leaving the validityof any single accountopen to externalchallenge,even
when it is sustainedinternally.By contrast,the regimenof explanationis based on the claim that there

This content downloaded from 62.122.79.21 on Thu, 19 Jun 2014 21:29:01 PM

All use subject to JSTOR Terms and Conditions
356 CHRIS LORENZ

nized by Hall: as a consequenceof this move, the borderlinebetween specific and

configurationalhistory vanishes almost completely too. Basically, there is just
one type of history that is: interpretative= explanatoryhistory.And as most
narrativetheoristsnever tire of pointing out, narrativizationis essentially inter-
pretation,so we end up with the thesis that all history is narrativehistory.This
conclusion will, of course, be welcomed by all narrativetheorists,but can not be
welcomed by Hall, because he has presentednarrativediscourse as fundamen-
tally distinctfrom explanatoryand interpretivediscourse,not to mentionhis dis-
tinctionbetween the discourseof narrativeand the discourseof social theory.Or
does social theorybasically consist of generalizednarratives?This would surely
be a new and originalview, but I doubt whetherHall would subscribeto it. Or is
"social theory"just a vacuous label, covering a whole range of other "discours-
es," as Axel van den Berg recently concluded?19
The second relatedunresolvedproblemHall inheritsfrom Weberis the ques-
tion whether,and if so, in what sense, the comparisonof ideal types with empir-
ical cases yields explanations.20As is well known, one of the functions Weber
attributedto ideal types is that of a "yardstick"to "measure"their "distance"
from reality. This comparison of "impure"empirical reality with "pure"ideal
types may sharpen our eye for the complexities of empirical reality, but it
remains unclearwhy such comparisonswould be regardedas "explanations"of
any kind. The problem is this: ideal types offer some kind of definition of the
phenomenathey describe, but definitions usually are not regardedas explana-
tions (unless one defends the type of narrativetheory of FrankAnkersmit,which
is not the case with Hall2l). One gets the impression that Hall does identify
both as Weberdid occasionally so in any case he is in good company.22

is only one appropriateset of criteriafor evaluatingthe adequacyof alternativeaccountsabouta given

phenomenon.Despite these differentcontexts,however,as I am construingthem, both explanationand
interpretationshare a core enterprisethat tends to blur the line between them."I can't go into Hall's
argumentshere; however, see my Konstruktionder Vergangenheit,83-189, for the numeroustradi-
tionalargumentsagainst the identificationof explanationand interpretation.See further,Hall, Cultures
of Inquiry,98-103 ("TheExplanatoryand InterpretativePotentialof NarrativeAccounts").
19. See Vanden Berg, "SocialMechanisms,"205: "Moreand more, it seems, what passes for 'soci-
ological theory,'or, more ominously, 'general sociological theory,'deals with mattersof epistemolo-
gy, ontology, and philosophy of science, at the expense of the more mundanebusiness of theorizing
about the social world. 'Sociological theory' nowadays seems to be theorizing about theorizing, not
attemptingto formulatecoherentaccounts of things happening'out there."'
20. For a recent overview of the varieties of comparativehistory, see Geschichte und Vergleich:
Ansdtze und Ergebnisse internationalvergleichenderGeschichtsschreibung,ed. H-G. Haupt and J.
Kocka (Frankfurtam Main and New York, 1996).
21. ForAnkersmit'sposition,see my article"Narrativism,
Positivismandthe 'Metaphorical Turn,"'
317-324.
22. JonathanSperberbroughtup this problemin a reviewof recentstudiesof the Germanbourgeoisie.
In some of thesestudies,an idealtypewas usedthatspecifieda democraticpoliticalorientationas one of the
characteristically
bourgeoisvalues.However,empiricalresearchbroughtout thatformuchof the nineteenth
centurythe majorpartof the Germanbourgeoishadotherpoliticalpreferencesthanthe idealtype specified.
He, justifiably,asked what explanatoryuses such comparisonsbetweenan ideal type and empiricalphe-
nomenahave.Theproblemis that"aswas oftenthe case withMaxWeberhimself,conclusionsarebuiltinto
the definitions"and"we areleft witharbitrarilyassignedascriptivecharacteristics."
See J. Sperber,"Burger,
Biirgerturn, BUrgerlicheGesellschaft:Studiesof the German(Upper)MiddleClass and its
Biirgerlichkeit,
SocioculturalWorld,"Journalof ModernHistory69 (1997), 271-297, andesp. 285-286.

This content downloaded from 62.122.79.21 on Thu, 19 Jun 2014 21:29:01 PM

All use subject to JSTOR Terms and Conditions
 SOMEAFTERTHOUGHTSON CULTUREAND EXPLANATION 357

In the end my problemwith Hall's analysis of history and culturalexplanation

is thathis own position remainsunclear:he rejects both the (neo-) positivist and
the historicistview this much is clear (see 346), yet by rejectingthe idea of a
"transhistoricalgenerality"for his own mode of analysis (346-347), his "middle
position" has a relativistring and remains without clear contours. His remarks
that the discourse of narrativemay "shadeoff' into the discourse of social theo-
ry or into the discourseof explanationand interpretationmay be true,but are not
very helpful in this respect.
I find far more insightful his idea that the generalizing approachmay yield
knowledge of the "patternedlogic" of (cultural)developments, showing "path
dependency"(341). I also find his case studies convincing illustrationsof this
phenomenon.Althoughhe is not explicit about it, here we could locate the rela-
tionshipbetween generalizingand individualizingapproachesin sociohistorical
inquiry.Generalizingapproachesaim to uncover a (general) developmentalor
evolutionarylogic, while particularizingapproachesaim at the reconstructionof
a particularchronology.Philip Abrams's analysis of this relationship,based on
the example of the "deviantcareer,"is still as convincing as ever:
the process [of becoming a deviant]has both an analytical logic-the sequence of stages
that in principle must be passed through-and an empirical chronology-[the actual
biographiesof those who become deviant]. The process can be specified both as a for-
mally orderedprogression,such as the corridorof deviation, and as an actualhistory:the
movement throughthe corridorof individual deviants. Both types of sequence must be
examinedif adequateexplanationis to be achieved.But the vitally importantpoint is that
while processes of social becoming have this doubly historicalcharacter-and so can be
describedeither logically or chronologically-the relationshipbetween the logic and the
chronologyis notfixed.23

From this angle, which can also be backedby WolfgangSchluchter'sdevelop-

mentalinterpretationof Max Weber,24it becomes possible to specify the different
relationshipsof generalizingand particularizingapproachesto time. This specifi-
cation is surelyneeded given the frequentand not very precise referencesto "the
indispensabilityof the time factor"in reflectionson sociohistoricalinquiry.

Biernacki on culture and explanation

What aboutcultureand explanationin Biernacki'spaper?The first thing to note
is that, unlike Hall, Biernackidoes not make a distinctionbetween how histori-
ans and social scientists usually go about studyingculture.He drawsno distinc-
tion between "individualizing"and "generalizing"approaches,and no clear dis-
tinctionbetween explanationand interpretationeither.His stance on this issue is
conciliatory,as is Hall's, but also cloudy; as he put the matterin anotheressay:
"Theriddleof what constitutesan adequateexplicationand of how to distinguish

23. Abrans,Historical Sociology, 276.

24. W. Schluchter, The Rise of Western Rationalism: Max Weber's Developmental History
(Berkeley,1981). Cf. J. Habermas,"GeschichteundEvolution,"in Habermas,ZurRekonstruktiondes
historischenMaterialismus(Frankfurtam Main, 1976), 200-260.

This content downloaded from 62.122.79.21 on Thu, 19 Jun 2014 21:29:01 PM

All use subject to JSTOR Terms and Conditions
 358 CHRISLORENZ

causalclaims from interpretativeones has vexed the best minds in philosophyfor

more than a century,"and leaves the question at that.25
The second thing to note in comparison to Hall is Biernacki's critical
drive.As I observedearlieron, Biernackidevelops a powerful argumentagainst
reductionistexplanationsthatpresupposecultureas the ultimategroundandcon-
stituentof humanreality.He effectively dispenses with semiotic essentialismand
reductionism,and rightly argues that showing the indispensabilityof culture as
an explanatoryfactor is never sufficient to establish the priorityof culture over
otherexplanatoryfactors.In this respect his parallellingof the reductivedrive of
the old social and the new culturalhistory is very enlightening.26
However,criticizingthe weaknesses and illegitimateclaims in culturalistposi-
tions, and developinga more solid positionof one's own, are two differentunder-
takings.27And I must admitthatI am moreconvincedby Biernackias a criticthan
by Biernackias a new theoreticianof cultural"practice."The mainreasonfor this
is that, althoughhis argumentfor the pragmaticdimensionsof meaning-produc-
tion is interestingand important,this argumentonly establishesthe thesis thatthe
pragmaticaspectsof significationcannotbe reducedto the semiotic aspects;given
the state of the debate, this is pretty important.But it does no more than this.
Biernacki'sargument,in otherwords,does not yield the conclusionthatthe prag-
matics of sign use is more importantthanits semiotics. But this is the conclusion
he suggests by his "illustrative"use of the three "pragmatic"case studies of Karl
Marx,BenedictAnderson,and himself, in which semiotic analysesare lacking.
My problemhere is that Biernackidoes not follow the methodologicalrecipe
he is handingover to the "culturalist"authorshe is criticizing.Against cultural-
ists he claims that it makes no sense to postulate the fundamentalexplanatory
role of cultural factors because the explanatory role of culture can only be
assessed in comparison with the role of non-culturalfactors. I find myself in
complete agreementwith him on this issue. However,he suddenly "forgets"his
own comparativerecipe when he assesses the pragmaticsof culturein his three
favorite examples (in Benedict Anderson's Imagined Communities, in Karl
Marx's analysis of the fetish-characterof commodity-exchangein the first vol-
25. See R. Biernacki, "Method and Metaphorafter the New CulturalHistory,"in Beyond the
Cultural Turn:New Directions in the Study of Society and Culture, ed. V. Bonnell and L. Hunt
(Berkeley, 1999), 73.
26. See Biernacki,"Methodand Metaphor,"62-68.
27. This does notmeanthatI am convincedthatall of his critiqueaimedat "culturalists" is to the point.
LabelingCarloGinzburgas a "reductionist" on the basis of one sentencein one article,for instance,is
not my cup of hermeneutics.Althoughin the older studiesby Ginzburg,such as TheBenandanti(Turin,
1972), one can easily detect the tracesof Marxistmaterialism,such as his attemptto link culturalphe-
nomena like "witchcraft"to an amorphousmaterialist"low culture,"this is no longer the case in his
later books. In The Cheese and the Worms(New York, 1982) the link between Menochio's ideas and
materialcultureis very thin indeed, and in Ecstasies: Deciphering the Witches'Sabbath (New York,
1991) the link between cultureand "materialcircumstances"has all but disappeared.In the last study
a semiotic analysis of myths is presentedas the majorexplanationfor a whole rangeof symbolic sim-
ilarities in space and over time; only "in the last instance"does Ginzburghypothesize an ancient
shamanisticcultureas the "underlying"explanatoryfactor.Significantly,cultureis now explainedby
cultureand no longer by any materialisticsubstratum.So the "reductionistdrive"of culturalismis less
serious in Ginzburgthan Biernackisuggests.

This content downloaded from 62.122.79.21 on Thu, 19 Jun 2014 21:29:01 PM

All use subject to JSTOR Terms and Conditions
 SOME AFTERTHOUGHTSON CULTUREAND EXPLANATION 359

ume of Das Kapital, and in his own study of the fabricationof labor). Insteadof
treatingthese three analyses of culturalpracticesas theories, and their explana-
tory claims as claims that can only be assessed in comparisonwith the explana-
tory claims of other theories on the same subject (there is no lack of competing
explanationsof, say, nationalismor the marketeconomy), Biernackipresentshis
favorite theories directly as adequate descriptionsand explanationsof what is
going on in "practice"(thatis, in reality).28
Alas, the concept of practiceis as problematicas the concept of "experience"
in this context. Both are frequently used as illicit substitutesfor "reality"and
must be confrontedwith the same critical question, namely the question how a
representationrelates to alternative representationsof the same phenomena.
Since naive realism is damagedbeyond repair,there is no way around"critical"
and pluralversions of realism (such as "internalrealism"29).So if my readingof
Biernackiis correct,one could accuse him on this point of doing exactly what he
criticizes the Geertzianculturalistsfor: favoringone type of culturalexplanation
(pragmaticinsteadof semiotic) over all others without specifying the arguments
for this preferencein the cases underreview.

Kane on cultureand explanation

What about Kane's view of cultureand explanation?Like Biernacki,and unlike
Hall, she makes no distinctionbetween an individualizing,historical approach,
and a generalizing,social-scientificapproachto culture.She simply assumes that
all scholars studying culture are hauntedby the same set of questions, and that
historians will be interested in general sociological questions, theories, and
approaches(311-312).30 Because this point is neitherevident nor arguedfor, it is
worthnoting. Nevertheless,Kane-like Hall is confrontedwith the problemof
how to connect the generalizingand the particularizingapproachesin her inter-
esting study of the Irish Land Movement. And like Biernacki-she is con-
frontedwith the problemof how to connect the "objective"semiotic analysis of
meaningto the "agentic"analysis of meaning(thatis, the pragmatic,"subjective"
side of meaning-construction).Her problem is, in her terminology,how to con-
nect "culturalstructures"and "culturalmodels" to "action"and to "contingent
events."How the contingentevents are explainedby the culturalstructures,how-
ever, is not easy to grasp.

28. In "Methodand Metaphor,"81, however,Biernackiis far more to the point than in his SSHA-
paper,in which the comparativeargumentis completely absent:"If we do not legitimateour concepts
and theoriesby attachingthem to an ultimatefoundationin the objects understudy, we can do so by
adoptingthe proceduresof inquiryand contest best suited for laying bare our self-made construalof
those objects."The research design, therefore, should contain "an attemptto comparativelyassess
opposing explanations."
29. See my "Historical Knowledge and Historical Reality: A Plea for 'InternalRealism'," in
History and Theory:ContemporaryReadings, ed. Fay et al., 342-377.
30. Kane is even prescriptivevis a vis historianswhen she states that "culturalexplanationof his-
toricalprocesses and transformations. . . must uncoverthe constructionand transformationof mean-
ing, and that in this historiansmust begin with the investigationof the structuresof culture"(313, ital-
ics mine).

This content downloaded from 62.122.79.21 on Thu, 19 Jun 2014 21:29:01 PM

All use subject to JSTOR Terms and Conditions
360 CHRISLORENZ

Actually, Kane faces two explanatoryproblems:the first one and by far the
most important is the explanation of the different Irish "discourses,"the
"Discourse of Retribution"and the "Discourse of Conciliation,"followed by
theirlater"fusion"in the "Constitutionaldiscourse aroundthe dominantsymbol
'Land'." The second explanatory problem is the history of the Irish Land
Movementand war itself, in which these discourses functioned.
Her discourse on "discourses"and "semiotic relations"notwithstanding,her
explanationof the genesis of these discourseslooks prettymuch like a tradition-
al historicalnarrative.It is mainly a developmentalhistory,going from one illus-
trativeevent to another(without any concern for the representativevalue of the
illustratedcases, however).It is a history of "how certainevents are given narra-
tive form"and of how "competinggroups often refer to the same event but nar-
rativizeit differentlyto promotetheir own discursiveposition" (318). "The cru-
cial theoreticalpoint,"Kane emphasizes, "is that the narrativeitself is contested
territory"(318). This partof Kane's explanationis in fact a very shorthistory of
Irish ideology, that is, a history of developing and changing ideas.
How these ideas are linked to the actions and events of the Land Movementis
the otherexplanatoryproblem,but Kane barely touches on it. Given her explicit
focus on the explanationof cultural"structuresof meaning,"this limitationis, of
course, legitimateand was to be expected.31 Nevertheless,I have a problemwith
her proposals for the linkage of "symbolic structures"to agency, because she
seems to be defending two positions instead of one.
In her introductoryparagraph,she adoptsthe agency/structure position of social
theoristslike AnthonyGiddensand PierreBourdieu,32emphasizingthe "explana-
tory focus on the recursivityof meaning,agency,and structure-and more specif-
ically, the mutualtransformation of social structure,social action,andculturalsys-
tems-in historicaltransformation" (311). On page 312 she asserts, accordingly,
that"meaningstructureandmeaningconstructiontogetherform the basis for cul-
tural explanationin historicalprocesses."Structureand action here seem to be
equally important.However,in her furtheranalysis she silently moves to a more
"structuralist" semiotic position.She posits thathistoriansof culture"mustbegin
with the investigationof the structuresof culture"(312)-and thus,by implication,
not with its construction. And although she recognizes that these "cultural mod-
els" are (potentially differently) interpretedby groups and individuals when
appliedto their experience,she holds that "these structuresshould be the initial
theoreticaland analytic focus in studying meaning construction"(314, italics
added).
The ambivalenceof Kane's position can be detected in the following passage:
Meaning constructionmust be analyzed in the first and last instance in reference to the
internal,or the semiotic, structureof symbolic systems. I do not claim a causal priorityof
symbolic systems over human agency, contingent events, or other structuralconditions.
31. Nevertheless,she announceson page 311 that she will presentan "empiricalanalysis of polit-
ical alliance duringthe IrishLandWar, 1879-1882."
32. See Van den Berg, "Social Mechanisms,"for a fundamentalcritiqueof Giddens and Bourdieu
and the other modernstructure/agencytheorists.

This content downloaded from 62.122.79.21 on Thu, 19 Jun 2014 21:29:01 PM

All use subject to JSTOR Terms and Conditions
 SOME AFTERTHOUGHTSON CULTUREAND EXPLANATION 361

However,people do refer first to culturalmodels as they try to make sense of situations,

and shape their strategiesfor action. My assertion,then, is that the locus of meaning, and
thereforethe condition for meaning construction,is symbolic structures.(314)

In this passage Kane asserts, first, that although symbolic structureshave no

"causalpriority"over otherfactors in historicalexplanation,they are the "locus"
and a "condition"for meaning-constructionand action. These assertions only
lead to the conclusion that cultural structuresare indispensable in historical
explanation,no more, and no less; they don't establish any hierarchyof explana-
tory factors. However, this is precisely the conclusion to which Kane is silently
heading in two moves. In the first move, she jumps from indispensabilityto pri-
macy: symbolic structuresare needed "in the first and the last instance"when we
explain ("analyze")"meaningconstruction."In the second move, she jumps from
the primacy of symbolic structuresto the primacy of semiotic analysis. Both
moves are unwarranted,and are reminiscentof similar moves in Marxism with
regardto the "determinationof the economic structurein the last instance."
Kane's dilemmais unsolvableas long as she wantsto arguethat semiotic expla-
nationis more importantthan other types of explanation,without arguingopenly
that symbolic structuresare more importantthan other factors. Basically, she
wants to establishthe primacy of semiotic explanationwithoutopenly claiming a
causalpriority of symbolic systems over others (314). This, however,is impossi-
ble, because both claims are interlinked;one cannot cling to one while rejecting
the other.This is, remarkably,exactly what Kane does. The reason for not claim-
ing causal priorityfor cultureis, of course, not difficultto imagine. This type of
claim would lead her directly to the "culturalist"position that was so effectively
demolishedin Biernacki'spaper.As in Hall's andin Biernacki'spapers,in Kane's
papertoo the issue of explanationand interpretationthus remainsthorny.

III. A RETURN TO NARRATIVE?

One of the most remarkableaspectsof both Kane'spaperand of Hall's book is the

complete revaluation of "the narrative approach." Sentences like Kane's
"Historicalstudies can be enrichedby a narrativeapproachto the constructionof
meanings"(329), and "narrativesare the consolidatingcomponentin a theoretical
model of meaning construction and historical process. They also provide the
methodby which to investigatethe recursiverelationshipof action, structure,and
culture"(329) are quite unexpected.In the ears of a philosopherof history such
advice given by a sociologist even soundslike an outrightparadox.Fromthe nine-
teenth centuryonwards,after all, mainstreamsociology has often legitimized its
very birthright,disciplinarydomain, and explanatorypotential by negating and
opposing "thenarrativeapproaches"of historyandcommon sense. Have we come
full circle at the end of the twentiethcenturyalso in this respect?It often looks this
way, but some caution and analytic precision would be welcome in this respect,
because the meanings of "narrative"and of "the narrativeapproach"are often
highly imprecise. Because the condemnationof "narrative"in the past was often

This content downloaded from 62.122.79.21 on Thu, 19 Jun 2014 21:29:01 PM

All use subject to JSTOR Terms and Conditions
362 CHRISLORENZ

based on impressionisticjudgements,we should be carefulnot to make the same

mistakenow thatwe are exhortedto embrace"thenarrativeapproach"again.
I find severalfeaturesstrikingin the recent reevaluationof narrativeby social
scientists.First,as far as narrativephilosophyof historyis noticed at all, this usu-
ally happensin an uncriticalway. HaydenWhite's theory of historicalnarrative,
for instance, is often taken at face value, and social scientists rarely take notice
of the widespreadcriticism of White's original position.33
Second,it strikesme thatas far as social scientistspresentexplicit definitionsof
"narrative," these definitionsusually are lookalikes of definitionsalreadydevel-
oped in the philosophyof history.Considerthe recentattemptsto define narrative
in termsof event-structures andto interpretnarrativeexplanationin termsof causal
linkages between those event structures,such as is proposedby sociologists like
LarryGriffinand RobertoFranzosi.34ArthurDanto alreadydeveloped a similar
proposalin his AnalyticalPhilosophyof History in 1965, and this has been dis-
cussed for some time since.35This does not imply that their proposalswould not
be interestingor unworthyof discussion-far fromit-but only thatthe earlierdis-
cussions of Danto's narrativephilosophyare the logical startingpoint for further
debates of similarpositions. Kane's remarkthat historiansmight profitfrom the
"narrativeapproach"suffers from similar flaws, because historiansand philoso-
phersof historyhave actuallybeen discussingnarrativefor quite some time.
This leads me to my thirdand final remarkon the recent enthusiasmof social
scientists about narrative.The latest fruit hanging from the narrativetree carries
the promising name of analytic narrative.36 Whatever this neologism means, it
surely is a clear indicationthat the disciplinarymaps of the social sciences and
history-and thus of social science history-have been reshuffled completely
since the 1960s. Analysis and narrativeare not only no longer regardedas com-
peting approachesto the past, but are now openly connected to each other.
When, however,one looks behind the veils of this newlywed couple, one dis-
covers some prettyworn-outclauses in theirmarriagecontract:"Ourapproachis
narrative;it pays close attentionto stories, accounts,and context. It is analyticin
that it extractsexplicit and formallines of reasoning,which facilitateboth expo-
sition and explanation,"the authorsstatein theirintroduction.37 "Wherepossible,
we use formal arguments. We . . . do not provide explanations by subsuming
cases under 'coveringlaws,' in the sense of Hempel.38Ratherwe seek to account

33. For this debate see History and Theory:ContemporaryReadings, ed. Fay et al., and my arti-
cle "CanHistories Be True?"
34. See L. Griffin, "Narrative,Event-StructureAnalysis and Causal Interpretationin Historical
Sociology,"AmericanJournalof Sociology 98 (1993), 1094-1133; R. Franzosi,"A Sociologist Meets
History:CriticalReflections upon Practice,"Journal of Historical Sociology 9 (1996), 354-392; New
Methodsfor Social History, ed. L. Griffinand M. van der Linden (Cambridge,Eng., 1999).
35. A. C. Danto, Analytical Philosophy of History (Cambridge,Eng., 1965).
36. R. Bates, A. Greif, M. Levi, J.-L. Rosenthal,and B. Weingast,AnalyticNarratives (Princeton,
1998), reviewed by Brian M. Downing, History and Theory39 (2000), 88-97.
37. Bates et al., AnalyticNarratives, 10.
38. Rejectingthe Hempeliancovering-lawtheoryof explanationhas turnedinto somethinglike a rit-
ual beatingof a deadhorse.See W. Salmon,FourDecades of ScientificExplanation(Minneapolis,1990).

This content downloaded from 62.122.79.21 on Thu, 19 Jun 2014 21:29:01 PM

All use subject to JSTOR Terms and Conditions
 SOMEAFTERTHOUGHTSON CULTUREAND EXPLANATION 363

for outcomes by identifying and exploring the mechanismsthat generate them.

We seek to cut deeply into the specifics of a time and place, and to locate and
trace the processes that generatethe outcome of interest."39
So, analyticnarrativesare mainly characterizedby opposing them to the cov-
ering-lawmodel of explanation,by their interestin formal modeling (especially
rational-choicemodeling), and by theirinterestin time and place. We are thus, in
effect, presentedwith the applicationof "analytic"rational-choicetheoryto "nar-
rative"history.So in the end, the old oppositionof "analysis"to "narrative" pops
up again, and the same goes for their unquestionedtraditionallinkage to "social
science" and "history."In other words, behind the flashy new formula of "ana-
lytic narratives"the old conceptualoppositionsare still in place. Insteadof jump-
ing on this shiny new narrativetrain,it would be betterin my view to reflecta bit
more on the conceptualrails on which these trainshave been moving for the last
decades. Only then will we be able to find out how we ended up in the remark-
able new predicamentin which sociologists are advising historians to follow
"narrativeapproaches"when they are explainingculture.
Free Universityof Amsterdam/Universityof Leiden, Netherlands

39. Bates et al., AnalyticNarratives, 11-12.

This content downloaded from 62.122.79.21 on Thu, 19 Jun 2014 21:29:01 PM

All use subject to JSTOR Terms and Conditions