The Role Of Theory In Comparative Politics 3y1s5q

The Role of Theory in Comparative Politics: A Symposium Author(s): Atul Kohli, Peter Evans, Peter J. Katzenstein, Adam Przeworski, Susanne Hoeber Rudolph, James C. Scott, Theda Skool Source: World Politics, Vol. 48, No. 1 (Oct., 1995), pp. 1-49 Published by: Cambridge University Press Stable URL: http://www.jstor.org/stable/25053951 Accessed: 19/03/2010 11:23 Your use of the JSTOR archive indicates your acceptance of JSTOR's and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/.jsp. JSTOR's and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please the publisher regarding any further use of this work. Publisher information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=cup. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. 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 [email protected].

Cambridge University Press is collaborating with JSTOR to digitize, preserve and extend access to World Politics.

http://www.jstor.org

THE ROLE OF THEORY

IN

COMPARATIVE POLITICS A Symposium ByATUL KOHLI, PETER EVANS, PETER J.KATZENSTEIN, ADAM PRZEWORSKI, SUSANNE HOEBER RUDOLPH, JAMESC. SCOTT, andTHEDA SKOOL*

Introduction:

THE

Atul

Kohli

Center of International Studies at Princeton University orga nized a symposium during 1993-94 on the topic of "The Role of

in Comparative Politics." The Theory symposium interest among participants, considerable enough

seminars generated to lead the editors of

World Politics to publish this edited and condensed version of the pro for ceedings Peter Evans,

a broader Peter

audience.

Katzenstein,

A group of distinguished scholars? Adam Hoeber Susanne Przeworski,

Rudolph, James Scott, and Theda

Skool?responded

to the main

are itemized themes below followed of the symposium. These main scholars. of the conclu the individual Some the of responses by are then summarized at sions that emerge from these diverse responses the end. themes

The field of comparative political analysis is again embroiled in the new to com are debates theoretical hardly no has dominated the (for example, single paradigm

oretical

controversy.

parative

politics

While

field for too long, and both methodological are common

and ideological differences

current claims and scholars), politics comparative one end of the method to At divisive. be appear deeply are the new students in of political there culture, spectrum, ological or a relativistic who of claims, culturally variety postmodern spired by and thus of conven doubt the value of causal explanations altogether in other The social science tional theorizing comparative politics. among

counterclaims

* The symposium was organized under the auspices of the Center of International Studies, Prince ton University. seminars were all held at Princeton University during 1993-94. Thanks Symposium are due to John Waterbury, the director of the Center of International Studies, for his encouragement in and . This edited and condensed version of the proceedings was prepared for publication World Politics.

WorldPolitics 48 (October 1995), 1-49

WORLD

2

POLITICS

extreme is defined claims: because all by nomothetic methodological are rational social actors, including actors, utility maximizers, political deductive and microeconomics and/or by logic inspired modeling, can

game theory, ent chaos of ward

scholars. Too they

the coherence

life. In between

political "center"?remain

the

others,

uncover

help

"social

sciency" are the ones who

these

the appar is, to

the majority of comparative politics for some and too much of "storytellers" for

em informed pursue theoretically mainly on one or more countries, focusing through di a or and utilizing variety of data, contemporary

analysis, lenses conceptual or historical, qualitative quantitative. at Princeton The symposium helped

pirical verse

that underlies two extremes?that

political

the competing oretical debates: If not, why not? should the "center" hold? If so, why? The invited scholars were asked to address these broad issues. Instead elucidate

these

of giving formal papers, theywere additionally asked to (1) characterize their own

to comparative approach to other be may preferable

proach

their ap (2) discuss why most and (3) impor approaches; controversies outlined above, provide as a subfield is headed and/or politics politics;

tantly, in light of the theoretical their views on where comparative to head. A summary of their responses ought

Peter The

central

the eclectic

follows.1

Evans

as I understand of this symposium, it, is whether question core of the center constituted the traditional that has messy

is in danger of being overrun. My bottom study of comparative politics is the study of comparative line response politics will neither simple: to be about into a swamp of impenetrable that descend purports jargon formal mod and symbols nor turn into a desert of ahistorical to describe of behavior. els that purport the microfoundations political core of are reasons is as it the There comparative politics why powerful the of is and those reasons have not disappeared. study pol Obviously discourse

itics must them.

deal with

Likewise,

by game being generated ter will not be overrun. Since 1 The

this

talks were

conviction recorded

ing the talks for publication. vidual participant.

and how people understand symbols use of the tools conceptual currently cen the eclectic messy theorists. Nonetheless,

values

it should

and

make

stems

in part

from my

view

of the generic

and adapt license in condensing and then transcribed. I took considerable The published version has been approved for publication by each indi

3

SYMPOSIUM forces

that shape

the evolution

science

of social

paradigms,

Iwill

begin

then like to explain more clearly what I

my discussion there. Iwould mean

as an center, using my own work by the eclectic messy example. of rational versions I will discuss asocial ahistoric, done, why are to overrun the center. choice or game-theoretic analysis unlikely

This

Finally, Iwill talk about why I see the role of cultural theories as inher ently partial. In thinking about the evolution ful to ask about microfoundations.

of social What

motivates

that we do?Most

kind of research and writing

it is paradigms, us to undertake

science

use the

scholars would agree

care about cases. If Iwere to particular I was for example, it might be in part because it would be because I but also parameters, hap

is that we

that part of the answer do a study of inflation, in its general interested

inflation has run at 30 live in a country where as we care about par percent a month, tearing apart their lives. As long to do to try to understand ticular cases, we are compelled spe history, that cific sequences of events and to acquire the ideographic knowledge pen

to have

friends who

of events entails. understanding specific sequences set of incentives is another, even more primitive and pervasive There all like to pre in most involved social science investigations. We would

dict when something bad or something good is likely to happen to us out how

to increase

of something a par always trying conceptualize good one a as set of similar ticular sequence sequences. The larger example of we are to be able to antic more we know about the set, the more likely can describe we outcome If of the any given sequence. key char ipate we can handle sets and talk about acteristics bigger parsimoniously, outcomes. The desire to is part of numbers of individual predict larger or, better

still, figure happening. We

social tists pen

the likelihood

are therefore

to

social scien but because science, not because we are positivists to to is know what else the desire share with everyone likely hap to to them and how be able improve prospective they might

outcomes.

Our

interest

cases

in particular

thus keeps

pushing

us in the direc

tion of doing history. Our interest in figuring out what is going to hap to us makes

even

and most

limited theorizing is likely to disappear from tremendously in intellectual fashion. of changes the social science scene, regardless of the social science community The agenda, activity, and paradigms are not scholars. Nor are the motivations of individual shaped only by

pen

attractive.

the

Neither

crudest

motivation

they shaped only by discourse within the community itself ? laThomas

4

WORLD

POLITICS

are Agendas powerfully shaped by the demands from our students of consumers and patrons. Everyone, in having us elucidate is interested Science Foundation,

Kuhn.2

and wants

torical

more

to know

sequences of social action. They quences the ideographic understanding

about

and constraints to the National particular

his conse

the

likely both for incentives, persistent provide of and for predic sequences particular

tions. Their pressure fits nicely with the kind of work that is found in center of messy comparative politics. mean center" should be clearer. I by the "eclectic messy By now, what it can but also cares It iswork that draws on general theories whenever cases as outcomes. It sees particular about historical deeply particular the eclectic

the building blocks for general theories and theories as lenses to iden

is interesting and significant tify what theories nor cases are sacrosanct. Cases vindicate

a

single

so scholars

theory,

cases. Neither particular are too to always complicated are in this tradition who work about

likely to draw on am?lange of theoretical traditions in hopes of gaining greater

on the cases

purchase

they

care about. At

the same

time,

a com

pelling interpretation of a particular case is only interesting if it points to ways

of understanding tion are often chastised in their

specialists Since I consider center,

let me

use

other

search

cases

as well,

so scholars

on the historical for "tresing" for broader generalizations.

in this tradi cases of other

a my own work very much part of the eclectic messy it to illustrate the interplay of theory and cases. First

of all, theory has been very important to me because it helps define an ex at and, by extension, provides looking cases are most to examine. of what In interesting mo on Brazil was ideas." work short, it defines My early originally "big of industrialization how the process tivated by a desire to understand was affected actors from other countries were by the fact that economic what

questions ternal definition

are worth

playing a central role in that process.3Why

did I think this question

of eco proponents general made and theorists my particular investiga growth dependency tion seem relevant and worthwhile. Later in the 1980s I began to focus more on the role of the state in industrial transformation.4 specifically

was

important?

Heated

debates

between

nomic

again, the topic were assertions being

Once

seemed made,

in part because big theoretical important most propo prominently by neoliberal

2 is distinctly "un-Kuhnian." Cf. of scientific paradigms In this sense my vision of the evolution of Chicago Press, 1962). Kuhn, The Structure of Scientific Revolutions (Chicago: University 3 State and Local Capital in The Alliance ofMultinational, See Peter Evans, Dependent Development: Brazil (Princeton: Princeton University Press, 1979). 4 States, Firms, and Industrial Transformation See, for example, Peter Evans, Embedded Autonomy: (Princeton:

Princeton

University

Press,

1995).

SYMPOSIUM nents

of modernized

5

versions

and aggressive

of classic

theories

of the

state.

minimalist

I should point out that in both cases, theory, in the sense of big ideas, was defined more course within a particular

of investigators. These community because resonated outside they engaged of scholars. The debates between transnationalism

community

the modern

and around state have

as a source

of tools.

I draw

my own and de

resuscitations

of the theory of the but also by only by academics, see these is who "regular" people have

also been

on

theoretical

general that make the behavior

me describe mechanisms help stitutions causally plausible. of multinational the behavior

theoreti

not

been

driven, and even politicians, policymakers, sues as to their lives. important General theoretical perspectives

work

of a dis

sense

in the Kuhnian

me

cal debates

pendency minimalist

than

broadly

in my important to frameworks of actors

and

in

to trying example, when distinguish firms from that of local entrepreneurs,

For

I

have drawn (parasitically) on a rudimentary set of propositions from microeconomics.5

By

taking

the

conventional

adding a little bit of industrial organization, and

Simon,

legitimacy,

theory

of

the firm,

a little bit of Herbert

some

about issues of nationalism arguments political to mechanisms able specify causally plausible result was eclectic how these actors behaved. The

I was

and that

and explain case. it the but illuminate messy, helped recent work on the state, several theories of in my more Likewise, about how organizations work?for example, simple economic models helped

and what kind of incentive systems bureau to generate in been trying causally again important about when people ideas mechanisms. Some "cultural" generic plausible are to care about what think of them and feel ac other people likely have also been constrained cordingly important. No single ready-made can to the cases model all the tools theoretical necessary provide explain

how

bureaucrats

cracies

I am

function

create?have

interested

to make

in, but

an eclectic

combination

offers

enough

leverage

a start.

or to work to the im attempts investigate through big questions have always been tied to par of plausible causal mechanisms plications of institutional ticular cases. I end up presenting pictures complicated My

that configurations tings. Nevertheless,

are very much I have always

51 say "parasitically" because theory.

Imake

set in particular historical were to be if useful, that, they

rooted felt

no claim to be a contributor

to the development

of this body of

6

WORLD

these

had

configurations

POLITICS

to be conceptualized

in ways

that were

poten

tially separable from the settings inwhich theywere originally derived. at

to sequences deciphering particular historical as cases to to be worthwhile, other well. The general they have speak case may or may not fit other cases, but ideas I derive from a particular at least seem worth they should applying. I found useful in The notion of "dependent which development," some II develop key features of Brazils post-World War summarizing In order

ment

for my

trajectory,

efforts

is a good

example.

After

working

on Brazil

for a num

ber of years, I became interested inEast Asia. I began by examining the applicability of dependent development and found that it did not apply. Figuring East Asia

out but

why also

that was led me

so not only to reformulate

helped my my earlier

understanding ideas of what

of was

distinctive about Brazil.61 would like to think that every time I gain a of a particular understanding some ideas that will also generate to me but to others as well. only Given my sense of the character

better

historical be useful of work

the process will sequence, in other cases, useful not in the eclectic

messy

center,

it should be clearwhy I feel it is unlikely that the center will be overrun. on or models methodological a source to be of vulnerability, but theoretical may appear propositions movement it also makes it easier to absorb new ideas. The persistent cases and theories makes to it hard for any particular between theory a to to it easier for the tradition broad and makes gain dominance speak

The

absence

of well-defined

consensus

audience. in the In my own case, the absence of threat is probably most obvious case of the recent offshoots It is hard to see these of microeconomics. as a my work. To the contrary, undercutting potential on I hope, for that people working parasite, principal-agent example, more theories can come up with powerful ways of describ simpler and use better I behave. could certainly ing how firms and bureaucracies approaches

as

tools. the natural impe Some might say that I am na?vely underestimating en are actor formal models of that rialism models, already strategic on an ever-wider and will behavior, range of political croaching continue

to do

instead

of simply

so until

they constitute a set of tools. Frankly,

an all-encoming paradigm I think there are several reasons

why this is not going to happen. First, it is very difficult to aggregate 6 in in East Asia: Some Lessons for Latin Americanists," See Evans, "Class, State, and Dependence Frederic Deyo, ed., The Political Economy of theNew Asian Industrialism (Ithaca, N.Y.: Cornell Univer sity Press, 1987).

SYMPOSIUM from microfoundations

without

changing

7 your

style of analysis

and ex

position. Indeed, it is not just difficult, but inefficient. Trying to explain institutions by aggregating individual rationalities is like trying to ex reactions by using propositions from subatomic plain chemical physics. It may be possible in principle, but it is neither efficient nor effective. are not is the fact that rational choice theorists Equally important to the pressures

immune

a of providing comprehensible and patrons will Consumers

understanding to do of particular sequences. push them more to involve an increas history. Once they do, their work will begin amount of detailed of institutions and look more ing analysis particular center. As in the eclectic messy like work sophisticated understandings of game-theoretic models become more widely diffused among new of researchers, generations will ing pale. The tendency

bedded

in historically

the novelty of asocial and ahistorical model em to become for game-theoretic elements

and institutionally

complex arguments will

to less threatening and the approach will become progressively the messy eclectic center. is one other reason why I am less worried There about the expan sion of rational choice I than have been some years might theorizing back. Over the last ten years theoretical work at the core of economics, increase

in both

up space for multiple theory,7 has opened new theoretical are These and path dependence. equilibria approaches even in cannot that formal theoretical models saying principle, provide answers without determinative about individual knowing something trade

and growth

of path dependence, of events. There has to be an element sequences are is to say history. So even as the economistic models which making the theoretical their forays into other social science disciplines, redoubts more came are in ways that are much from which they changing to the kind of work which eclectic center. goes on in the messy friendly But what about new forms of cultural analysis? Like rational choice ex have certainly and game-theoretic cultural approaches approaches, overrun a boom in recent years. Might the eclectic they perienced center? who doubted that oriented messy culturally approaches Anyone was cer be used to address core issues of comparative might politics or reassess to his her views when confronted with a recent tainly forced article which 7

veteran of the field, Samuel in Huntington, by that hard-nosed a new wherein he argued that world politics was entering phase,

See, for example, Paul R. Krugman, Rethinking International Trade (Cambridge: MIT Press, Paul Romer, "The Origins of Endogenous Growth,"JournalofEconomic Perspectives 8 (Winter in the Economy," "Positive s and Brian W. Arthur, (February Scientific American 92-99.

1990); 1994); 1990),

8

WORLD sources

the dominant on a

based

of conflict

of identity

politics

POLITICS would

be culturally distinctive

and culturally

than on the politics postulated by conventional

that is, based, values rather

theories of political

make the argument, but his article economy.8 Not only did Huntington was touted in as the "X" article of the war era. post-cold Foreign Affairs

The article, although fascinating, is also a good illustration of the to the ascendance

limits

of culturally based arguments. a "Confucian-Islamic in which

ton up a world conjures confronts "theWest." Imagining for those who

miss

such a world may be a nostalgic trip versus the free of global communism of the contemporary it sim world description

the old clarity

but as a plausible

world,

In it, Hunting connection"

ply does not fly. you

to which

extent

The

look

at the Confucian

is misguided is most obvious when end of this connection. In constructing his were makes much of the fact that the Chinese the vision

connection, Huntington some money to make by selling missiles willing in the Middle sensitive customers East. Iwould tion

to some not very price agree that the transac

end, but it is not about business practice. affinity, sharp are much more about sell forces that are reshaping China today

is indicative

of behavior

at the "Confucian"

it is about

cultural The ing Nike

shoes

to the U.S.

than

they

are about

selling missiles

to the

Middle East. In East Asia it is the politics of growth and accumulation is delivering the goods. Interest-based economy argu political are not more, less, relevant. If we are looking for plau becoming most in the world's and rapidly sible causal mechanisms populous

that

ments

to "round up the usual suspects" of political of and informal bureaucracies, markets, economy. in the mechanisms material gains pursuit of power and conventional sense both to to in the street. and makes leaders political people As long as we live in a world where the normal, interest-based logic to of political remains plausible and to the peo economy policymakers growing

region, Some

it is time

combination

to overrun in the street, cultural analysis is not going the eclectic we use center. of should the discourse, values, messy analysis Obviously, to and contextualize and symbols conventional modify, complement, but we are not likely to get very far by economy arguments, political an alternative that completely old worldview constructing byes ple

power and interests. of cultural the work of Advocates analysis may protest my using as an in since is the of he post hardly Huntington example, vanguard fashioned

8 The

1993).

reference

is to Samuel Huntington,

"The Clash

of Civilizations,"

Foreign Affairs

72 (Summer

SYMPOSIUM thinkers.

modern

discourse

alyzing

that more respond not escape symbols do

Iwould and

raised in relation toHuntington, own.

their

beginning.

ways of an that I have

sophisticated the problems

and they have additional problems of

are indeed

must anticausal, they they fly in the at the of the two generic social constraints that I outlined consumers The and patrons of social science want help in Insofar

face of one

as

9

figuring out what is likely to happen to them if they take one course of action

rather

than

another.

Individual

who

practitioners

refuse

to de

liver predictions and are only willing to discuss theways inwhich social life is inherently inchoate and unpredictable are unlikely candidates for establishing hegemony. One they

caveat

important

are

looking or not,

is in order.

for better

outcomes.

People Persistent

want

because prediction ex whether times, en to theories that claim bad

with plicable bring disillusionment hance predictability. The conventional logic of political failed to deliver for a large share of the world's population

tle prospect of delivering for them in the future. Those other

end of Huntingtons Confucian-Islamic the Islamic end of the supposed

point. On

has economy and has lit

living on the are a case

connection connection,

there

in

is a region

and people that have long suffered political humiliation and, over the last ten to fifteen as well.

and trade years, have suffered from declining growth move circumstances It is hardly that under such surprising

ments based on identity and culturally distinctive values should gain extent strength. To the dations of conventional

that

the causally plausible microfoun become less economy increasingly

they do,

political

compelling.

The global political economy includes regions inwhich

states and

are not. To

are in which and regions analyze delivering they one must kinds such aworld, combine conventional economy political with cul of analysis, which focus on how people get what they want, us the nature of the prefer tural approaches, which understand help ences of conventional of this themselves. course, And, combining on nature that the with focus economy origin and approaches political sort is precisely the of eclectic work that the of preferences themselves a center of time. for has long practiced politics comparative us to incentives individual and societal Let me sum up. First, push examination of individual toward ward the detailed cases, history. At same to future outcomes?that the is, time, they demand guidelines markets

The prediction. fective response theoretic

character to these

approaches

of work

in the messy eclectic center is an ef rational choice and game Second,

pressures. are less threatening

to the messy

eclectic

center

10

WORLD

POLITICS

can be a useful source of seem. they might They causally plausible more mechanisms institutional for explaining Their change. sophisti will cated practitioners tend to get drawn into more institutional and turn in the in which lead them investigations, historically complex than

center. This

of the messy eclectic that recent innovations

direction the fact

have

theory

is reinforced tendency by at the mathematical heart of economic

theoretical

opened

for historical

space

and

institutional

arguments.

The potential hegemony of cultural approaches is limited by the fact that

the

incentives

of conventional

political

to most

"on the ground." As people not most prediction, they will satisfy therefore remain marginalized. When

economy as cultural

sense

still make

eschew

long consumers

analysts and patrons and will and conventional states, markets,

politics fail to deliver the conventional goods, the politics of identity more

In a global polit that consistently delivers for some but not for others, cul a tural analysis must be combined with conventional economy, political to be offered most center combination the eclectic of messy likely by

and culturally ical economy

distinctive

comparative

values

become

salient.

politics.

Peter

J. Katzenstein

Iwish to argue here that neither specific subfields of political science nor

particular

analytical

deserve

perspectives

standing. Instead Iwelcome

any

special

intellectual

the blurring of distinctions. This is true for

the distinction

between and international but it comparative politics, as ra in these for different such holds, subfields, approaches or tional choice, cultural institutionalism. None of these studies, pro fessional useful. You will therefore not "flags" is in my view particularly as well,

or from me about the color of the flag under which we must rally we must am interested the size of the I in when dragons slay. questions: all is said and done, scholars do their best research because of the polit not be ical problems and the intellectual that them, engage puzzles

hear

cause of the sage advice

of prophets

of the profession.

In a similar vein,

I also hold that professors do their best teaching not by initiating stu sense of sect but the students' by honing prob lem-focused research and by preaching the virtues of eclectic I leave to others the giving of advice about what "the theorizing. Hence

dents

to their

favored

field" should do. Working

at the margins

of comparative

and international

politics

has

11

SYMPOSIUM provided

me with

a distinctive

and

international

comparative distinctions

between

is a growing need of which American ical theory. My actor and unitary

political to erase the barriers

between

comparative politics? relations, and polit that the state is not a

is a part?international politics own research has taught me

of political hence preferences?and I have learned that you have importantly, to is and interesting That the hardest questions. thing that

the origin

But most

history?matters. to ask important do and the hardest perspectives,

research on perspective. Contemporary issues increasingly calls for blurring the there economy, security, and culture. And

thing and methods

to teach. We

can teach paradigms, off the professional

them

by taking to the classroom. What

analytical shelf and

we cannot teach so them transporting readily is to how ask important the hallmark and interesting of supe questions, rior research and great teaching. What do we mean by comparative research? To this question I offer answer the I learned in graduate school, for it is still useful and plausi is a focus on analytical ble: comparative research among relationships

variables validated by social science, a focus that ismodified by differ ences

in the context

in which

we

observe

and measure

those

variables.

church of scholarship is a cathedral not a chapel; hence I subscribe My ac to an notion science." The church of scholarship of "social expansive on wars: sects bent often be commodates different many holy fighting and interpretivists havioralists and nonbehavioralists then, rationalists now.

In my view international research and global the including phenomena, structures transnational and processes across societal and state boundaries. at the

focus

should relations that

on

international

between

occur

either

states

and or

outside

of the comparative and in in interna In the field of fields, they appear cycles. tional relations, for example, one cycle started in the 1960s, when much na and imaginative work was done on strategic theory and original in international tional security. In the 1970s the innovative scholarship As

I look

intellectual

relations

vicissitudes

to move

ternational

moved

toward

the field

of political

economy

and eventually

branched off in two directions. The first branch focused on qualitative research,

with

the

state

and other

institutions

as its main

analytical

focus. The second branch developed along rationalist lines by focusing on

international

regimes and international

parative areas of scholarship, enced an intellectual

and other

international

institutions.

As

com

established economy were becoming political in and mid-1980s studies the experi early security renaissance led by a new cohort of scholars. This

12

WORLD removed

than its prede policy engagement to of the 1950s and 1960s, committed standard applying or rationalist to contemporary structural and historical explanations cases of states. With the end of the cold war security conflicts between on as well as their concentrated focus of military questions security, farther

group, cessors

from

POLITICS

active

was

their statecentric

has provoked perspective, vigorous debate. that of will corpus scholarship substantially

a new

ating

It is gener the expand

field of security studies not by replacing strategic studies but by com plementing itwith a broader view of security and by going beyond the of the mainstream. perspectives are now new intersections? to cycles moving productive even not for ifworld revolutions do change the habits of scholars, bore are dom does. First, scholars interested in security or political economy one field to the other. Furthermore, to as the of seeking apply insights our attention to new ques cultural and rationalist direct perspectives one new the fields of security and politi tions?and, hopes, insights?in state-centered

two

These

in the predictable cal economy, there is less interest disagreements realist scholars of security who favor structural explanations between and liberal scholars of political economy who favor rationalist explanations. in recent

work

Comparative

decades

has witnessed

a move

away

from political culture and political sociology and from styles of analysis to

institutions. Critics literature ofthat have disregard as little more a residual out that it treated culture than pointed typically to the role of attention variable. This focus was replaced by a renewed new institutions in politics. The institutionalism took two forms, that

tended

"thick" and "thin." The sis. Derived takes

from

a rationalist style of analy and choice analysis public theory, it as to coordi solutions facilitating seriously

"thin" version

offers

transaction-cost

the role of institutions

states and so problems. The "thick" form is concerned with both one of them, and it looks at social sectors, cial structures, not with just constraints. and ideological coalitions, institutions, political political nation

These

two movements

in comparative somewhat related

and confronted

forcing tional analyses

the problem

was

one

studies were

problems. of disaggregation.

rein mutually For thick institu This

research

tradition does a good job of illuminating a broad variety of political set

at what it has to say about tings, but it is less successful connecting to individual and states, social structures, political coalitions, ideologies thin institutionalism the prob decision makers. Conversely, confronts lem of aggregation. Done tells us much carefully, thin institutionalism about but

the microfoundations it is much

weaker

in

of

individual

generalizing

from

and politics, policy choices the specific to the general.

SYMPOSIUM Because

corresponding

13 to bedevil

continue

problems

both

types

of in

stitutional analyses, scholarsworking in both traditions typically should be cautious

the virtues of their favored touting approaches. of sociological and cultural studies was a second

about

renaissance

The

de

that came

later to the comparative field. With its focus on velopment some of this research was the institutionalization of meaning, evidently too there is a bi work the of influenced the decade. Here by previous furcation jective

I would

what

between

versus

move.

the subjective

call, for lack of a better term, the ob I will put aside the subjective move,

by which I have in mind postmodernism in its various guises. It is deeply influenced by intellectual currents inEurope as they apply in the to a vari studies, as well as in anthropology, a are often directed no to the ety of texts and analysis of very expansive tion of politics. Much the science social postmodern rejects scholarship as I understand it here, submitting truth claims competing enterprise and in cultural

humanities

to

tests rather than empirical simply offering playful illustrations. is concerned with The objective move identities and norms. Norms as areas of are understood in the sociological tradition intersubjective not with that is, as dealing with objective agreement, reality subjective of culture differs from that of the 1960s, which factors. This conception

focused on individually held values. Obviously between ues. But

intersubjectively compared with

norms

there is a relationship

and deeply held subjective val are less well scientists political

psychologists, version of these relationships. This sociological norms for comparative is important because studies analysis two research programs, The often get institutionalized. identities to understand

equipped cultural and

shared

studies, are thus as intimately economics and culturalism,

cultural and objective as are rationalism

institutionalism lated

to one

another

re and

sociology.

The debates between realism and idealism in the 1940s and 1950s and between

and

quantitative-behavioral

ap

qualitative-historical

proaches in the 1960s were followed in the 1980s and 1990s by the "third debate" temological, and liberal

in international

relations.

and methodological scholars uncomfortable

issues

It has

raised ontological, mainstream

that make

epis realist

are not particularly they occur at to handle them. Typically, such discussions well equipped only Scholars of the margins of the field of comparative comparative politics. extreme versions of a nomo politics have by and large shied away from as some spe in thetic and typically rational choice approach, practiced studies. of international subfields cialized comparative Similarly, politics

scholarship

has

taken

because

a centrist

approach

to cultural

studies,

ac

14

WORLD

POLITICS

cepting the insights of both cultural and rationalist styles of analysis. is a very good reason for this centrism. Many strands of inter on methods recon research that tend toward historical pretivist rely or structions excavations rather than genealogical hypothesis and testing. But interpretive formulation social science is particularly

There

helpful for making us think about issues of identity that evidently have more

become

in the contemporary important analysis of comparative were or two than they three decades theories ago. Rationalist politics on the processes to shed have to date been unable significant light by are built up or down. are which identities these approaches Conversely, well

to the analysis is assumed. identity

tailored

whose rational

of choices In brief,

of individual

between

or

the false

actors corporate extremes of both

to ex analysis continues approaches, comparative in which of power and institutions cultural norms and interact with interests.

and cultural

a world

plore identities

in As is true of most other scholarship science, international political events. This does is and comparative driven real-world research, too, by we as not mean that for has re study only the present, Schopenhauer are the second hands of minded us, newspapers history?they always tell the wrong time. But real-world us back to the we past, which ing

Rethinking

events can then

our spark imagination, see in very different

send .

the past in light of the present and conversely rethinking

in light of the past are the present ways of searching for im productive momentous and The of portant political intriguing questions. upheaval the last decade has spurred renewed in both compar interest in history ative and international

studies,

as scholars

in both fields

turn to the past

for insights and categories of analysis thatwill help us get our bearings in a present

that

is in flux.

events are tells me that real-world Furthermore, my intuition driving con in a second direction. With the of scholarship collapse bipolarity aworld a new kind of are us events to of and temporary moving regions area studies that connect and international research. Since comparative area studies has been criticized the late 1960s traditional relentlessly, to and explicit comparisons a its testable propositions for analysis that and for proclivity privileges and insufficiently cultural uniqueness of a particular presumed analyzed new kind of or I envision The studies that country region. regional connect in would and with culture and expertise language familiarity and often

justifiably,

for its lack of attention

history with a cast of mind of different empirical

or

interested in either testing the plausibility

a propositions analytical perspectives against body of a or that evidence with evidence variety of in illuminating

SYMPOSIUM methods. terpretive the coding of cases ses is often

It

is worth

and variables

15 out

that since explicitly the testing of hypothe are the two activities closely

pointing that precedes

itself an act of interpretation,

related. With good aworld

structure of power there are many of the bipolar aworld are of regions. This is not toward moving in of 1930s about the the and categories only thought

the collapse we why

reasons

sensibly For in this world

of regions there exist two noteworthy global two hundred of years of Anglo-Saxon imperial processes?the legacy are no to states. The ism?that anchored longer tied firmly territorially no states first diffuses hence, longer fully through markets; technologies even contest that fact. The control global though they may profits, 1940s.

second diffuses human rights and international law; also politically conse too, has had profound by states, this process, political states will in recent years. Thus, quences rarely give up their sovereign is nevertheless affected that but their by processes rights, sovereignty no turn these processes will reknit the rela control. In they longer fully contested

states in novel ways. This creates between regions and tionships arena for one that did not exist fifty years ago. Although politics, mons to dominate markets their regions and international sought on state the touching prerogatives, encroaching global processes

a new hege were ratio

nalist realm of profit and the cultural realm of identity did not have the then

prominence political and the states that processes. To recognize bridging distinction more

inhabit

and understand

the world between

recently

now. they enjoy Today to them have respond

that

of comparative a nomothetic

between

rationalism

or

sworld

regions

to these

global

these processes interpret studies. and international and

an

idiographic and culturalism, will

requires A sharp social science, do little to help

us recognize and make intelligible this world of regions. For anybody to answer an and interesting important seeking that can be gleaned fool to sacrifice the insights

tive.Although

be a question would from either perspec

it is true that among the blind the one-eyed is king, it is

two eyes. also true that good depth perception requires creates local processes of political mobilization The new regionalism induced po and regionally and demobilization around internationally intersections of litical struggles over power, markets, and identities. The con and the connections between regional global and local processes us a sense of some of the states may texts and national newly give the collapse of some of and structures as we witness processes emerging the old.

16

WORLD

POLITICS

Adam Przeworski I was

not about the world. I resisted approaches, I am a methodological who believes opportunist or use If game I in doing it. If works. using whatever theory works, is called for is a historical I do that. If deconstruction is what , urged this invitation

needed,

Iwill

to

speak because

about

even

no

So I have

try deconstruction. I do have a message

principles. concerns message in comparative I think research; I spread it because for comparative that it has consequences research that are not fully rec is simple: compar The message and sufficiently ognized appreciated. Nevertheless, the role of theory

to share. This

isons necessarily entail counterfactuals. inevitably Comparisons require cannot theories about what we do not observe, because rely exclu they on observations. to argue, moreover, I am about What holds sively cases we are in the his whether research, up to all doing large-sample or one case. What is involved here is tory of the world, just studying we are or case studies. Rather, not whether the issue I statistical doing am about to outline to all is inherent research. comparative

Iwill proceed as follows: First, Iwill explain what Imean by "com I will discuss why Second, parison." on the I will focus Third, relationship

the issue of counterfactuals between

arises. and coun

comparisons

terfactuals. And finally, Iwill outline some ways of coping practically

with

in statistical

counterfactuals

Let me cussion.

introduce It comes

an

research

example from my current

that

throughout on the project

dichotomously regimes?classified economic growth.9 The question on democracy development.

as democracies

torships?on

Say that we observe Chile was

authoritarian

year declined ine whether

in 1985.We

studies.

use

research

political

in case

and

I will

concerns

this dis impact of and dicta

the

impact

of

see that its political regime

in that and that the per capita income of the country rate of 2.26 percent. to exam Since we want

at an annual regimes

affect

growth,

we

need

to know

what

Chile's

growth ratewould have been in 1985 had it been democratic: we need to compare the growth of Chile in 1985 under dictatorship and democ we cannot racy. Obviously, since Chile was servations,

answer

this question with the available ob in 1985 and we do not observe not authoritarian. in 1985 as simultaneously But as com Chile being we know what are told to to do: We parativists proceed quasi-experi 9 This

project is conducted (University of Pennsylvania),

authoritarian

(DePaul University), Jos? Antonio tly with Mike Alvarez and Fernando Limongi (University of S?o Paulo).

Cheibub

17

SYMPOSIUM to look for a case that is exactly like Chile

mentally, other

than

its regime and, possibly, 1985" that is democratic?and

"Chile

its rate of economic then

to compare

in all aspects growth?a the authoritar

ianChile with the democratic "Chile." Ifwe then find that this demo we conclude rate of 1985" has a positive that growth, more we discover is is for If democracy good growth. decay profound, is bad for growth. that democracy cratic

"Chile

And this is indeed what we do.We

try to emulate experiments by

cases

same that are "comparable." The "matches," finding logic is the we have or many. When one such whether just pair doing comparative research, we are told that one should find cases that are as similar as as in as many and then find a crucial differ aspects possible, possible, ence

that can explain what

one wants

to have this notion of explain. We as if were an intrinsic charac things being "comparable" comparability teristic of our objects of investigation rather than a result of our judg ments. A is acceptable. However, of and Sweden Denmark comparison ifwe

we are told compare Sweden with Chile or Kenya with Argentina, that this is like comparing that this is not a legitimate operation, apples of our discipline?the in and pears. Note that the entire organization as of political science departments, of professional and of the Social Science Research Council?is organized

structure

stitutional sociations,

construction is pre lines. I think the underlying geographical on cases that this that is, having cisely quasi experimentation, emphasis are so we can control are common that and then aspects comparable move variable variable. step by step, by along

Let me

the range of examples. We ask questions about ef on eco fects of coups about the impact of trade policies poverty, on nomic performance, of electoral of systems political participation, on social or of class structure on the emergence revolutions of change, same we ap These share the and research questions democracy. logic cases. We same way, that is, we find them the take proach matching broaden

on

Now,

or versus China, England Japan to my this question, apply procedure

versus to

and then we that

is, the

compare. impact

of

regimes on economic growth, Iwill take 139 countries, between 1950 or the year of data are and 1990 or the last year for which independence or as democracies If I I classify available. their regimes dictatorships.

then calculate means, Iwill find that democracies during this period an average annual rate of of 2.44 percent per year, growth a at rate of 1.82 percent per year. authoritarian regimes grew a it do regression will reveal significant, analysis, positive matter for One then conclude that may regimes democracy.

had

whereas If I also effect growth:

of

WORLD POLITICS

18 promotes democracy ceeds in the manner

development. just delineated:

Much what

of comparative analysis pro is this Imean by comparisons

cases. of matching procedure in 1985 the fact that Chile was a dictatorship that however, Suppose, are not is of its economic several possible growth. There independent reasons for this: first, affect the survival of regimes; second, growth may as there may be other underlying such the level of development, factors, and rates of growth; third, there may regime selection errors in measurement) unobservable factors that (including

that affect some

be

both

are common

to both

the underlying variables. Whatever mechanism, on the first, what are the for comparative implications result will be that there will be cases in our sample analysis? The main a match and our inferences will be biased. without and I will

focus

Let me nomic while

explain. Suppose decline of 5 percent survive dictatorships

tions. We

will

then observe

an eco that experience a in year die instantaneously, particular or die of economic condi independently over a full range of exogenous dictatorships over a part of that range. There will only that democracies

but democracies conditions, be dictatorships facing adverse The match democracies. among

conditions, observable

for which population

there will

be no

of democracies

will be a biased sample of all the potential conditions. in this way, or Indeed, I think that we have to think about the world at least suspect the world of being endogenously If we be produced. lieve that the conditions under which we live are somehow created by in pursuit of their ends, then the social world which we observe people us

of our actions, and independent that we try to explain. We thus as must treat the observable world having been produced by "us," that an addi More is, as having been generated endogenously. specifically, successes and elimi intuition is that the world nurtures tional powerful is not given, that is, is not outcomes various of independent

around

not

successes to If so, one suspects that there should be more am want to that I be observed than failures. I immediately emphasize not a not this is about of from here; question talking samples sampling is variable elemen available observations. the Sampling by dependent nates

failures.

is at stake. The issue rather is and it is not what tary textbook material un that the observable world is not a random sample of the possible we want to compare, we must process If the im conditions. derlying of variables?in and my case, democracy pact independent dictatorship, or in other cases, or and presidential systems, revolution parliamentary no

revolution,

landlord

domination

versus

bourgeois

domination?

19

SYMPOSIUM across

some

Let me

But we

range of similar conditions. across the same conditions.

observations

cannot

to find

expect

one

am mak to buttress the point I piece of evidence in which income declines have an ex per capita ing. Dictatorships are life of that growing forty years. Dictatorships pected economically can expect to live an years. Democracies, fifty-three by contrast, have life of eighteen the economy declines and an ex years when expected the life of when Democracies years economy grows. sixty-eight pected provide

are thus much

more

to assess

In order to take

into

to economic

vulnerable

the impact of regimes the fact that these



Given

crises

on

you

growth, observations are

than dictatorships. therefore, you have are not

exoge

to observe

poor poor democracies. particularly Now we the relationship between compar isons and counterfactuals. Let us put together two assumptions. Suppose a difference for economic that on the average do not make regimes a rate is equal growth: for any regime during particular year the growth nously produced. economic performance

endogeneity, in democracies, are in a to discuss position

to 2.09, which minus

unlikely

is the world average during this whole period, plus or

a random

error.We

also already

know

that democracies

are more

likely to die when they do badly. Pause to thinkwhat we are then going to observe

in the real world

if these

two

hold: regimes have assumptions die when they perform poorly. growth as are The answer is that ifwe take the observations given and they and dictatorships, calculate the average rates of growth for democracies we will find that democracies economic do better in promoting growth. no effect

When not

on

but democracies

democracies observe

face bad economic

conditions, if a democracy does

them

anymore: so that in the observed

dictatorship, that democracies

we do they die, and a it becomes poorly,

we are to observe going population this finding results not because democ on economic are but because democracies growth,

do better. Yet

racy has any effect more sensitive to economic

is not exoge crises. The fact that the world biased and inconsistent?inferences. leads to invalid?specifically, we are the statistical literature here is what calls a What observing I am persuaded work "selection bias." Indeed, that all the comparative can we have been from selection bias. We suffer doing may potentially nous

not do good is, until we

research unless we worry comparative ask each time how our observations

our observations by which or not? Unless to what we want explain biased inferences. making mechanism

are we

about are

that

selection,

produced.

Is the

independent produced we will that test

of be

20

POLITICS

WORLD

to fill in the the bias, we need to conjure counterfactuals to of the truncated distribution. We have determine unobserved, part in somehow what economic democracies would have been had growth remove

To

to the they been exposed is why This comparisons

same

of conditions

range

entail

inevitably

counterfactuals.

cannot find amatch for the authoritarian Chile population

we

observe,

we must

dictatorships. Since we

in 1985 in the entire to determine

other ways

find

as

what

the

rate of growth of Chile would have been in 1985 had it been democratic. this is easier said than done. The basic idea is that we always, are model the way in which observations and explicitly produced

As must

then use this knowledge to fill in the missing parts of the observed dis tribution oriented

to generate

counterfactuals. strategy

development

South

that was

an export adopted but successful, why did

Korea

very

South Korea adopt this strategy? And why did Brazil not adopt the same whether

strategy? South

Only Korea

once we would

answer have

these

done worse

can we assess questions and Brazil better had

that South Korea chose they chosen the alternative strategies. Suppose access to the U.S. market: it the export-oriented because had strategy Had Brazil had equal access to the let this be the selection mechanism. and had it chosen the same strategy, what would its per U.S. market formance

have

been?

Since this procedure entails filling tails of distributions, the methods bias are not robust even when many for correcting selection observa are are tions available. But with many observations, there standard sta that we can use: we assume a distribution tistical techniques (and test if it fits), model the selection mechanism, and use the predicted values to we do not observe. The fill in the parts of the distribution counterfac the "matches" we have been looking for: they generate provide from selection bias. We the actual observations then compare the counterfactual with conjectures. The problem we face is the same in small-sample research. Let me one s observation revolutions that do provide just example: Tocqueville not result in social that transformations. revolutions major Suppose tuals we

are freed

occur the

in which only under situations can selection mechanism. We

causes

of revolution conclude

intuitively not to "treatment."

and

its outcomes

it is hard then

guess are not

to

things: this is that the immediately change

and we can independent, are due to selection, outcomes in situations where occurred change

that the conservative Had

revolutions

is easy, they would have generated transformations. revolutionary on the prop that is avail case studies, we cannot When doing rely is large: we cannot use a statisti able when the number of observations

21

SYMPOSIUM to

cal distribution

the counterfactuals.

generate

We

must

less

proceed

formally. But I offer a suggestion that I have found fruitful in playing mental

up your cases, say South Korea and Brazil, experiments: Write use to the "search and replace" function on your word processor and then sense? Brazil the names of countries. Does the story still make transpose

now has

access

strategy? And

to the U.S.

markets.

it adopt

Does

if it does, is it successful? The

the export-oriented

new story may read

and you will be satisfied. But you may discover that?hold smoothly, in Brazil may have been too high to ex it?the level of relative wages too sectors were port even if the U.S. market were open, the protected a so on. to tolerate neutral exchange entrenched regime, and politically

Clearly, this is nothing but amental experiment guided by intuition. But this procedure should be a standard protocol for case studies: we a at least way of coping with it. about selection and this is worry To conclude, if the observed world is not a random sample of the pos

I think must

from

then inferences sible worlds, must invalid. Comparisons

then

about

cases, one or all, will be We must worry counterfactuals. their effects, and correct for them.

the observable entail

selection mechanisms, identify conclusions add up to an antiexperimental posture. Our since we cannot match when the world should not be to match, These

not generate cases may about

does

we

all the pairs need. Indeed, the exacerbate selection bias. just

the mechanisms

then use

effort

this knowledge world.

the observable matching we must theorize Instead, are and which the observations by generated nature to compensate of for the nonrandom

the observable

Susanne Hoeber

Rudolph

as in an essay on the comparative method, some. to contrast But the (and overlap) pro appear stance we shall take to the compar for discussing what more to the social sciences The ative enterprise?and generally. are offered as a scaffold for the report of my adduced below quotations encounter toWeber via Foucault. The with Fou travels from Weber

To

pair Weber I do here, may vides a vehicle

and Foucault odd

cault was more

in the nature

but it ledme to readWeber cause

he was

comparative

a central enterprise

of a sideswipe

than a full-scale

differently.Weber

encounter,

figures in the story be

the form, in defining figure, albeit in mediated in for persons the non-Western world studying

the 1950s and 1960s. I use him and his American progeny as the main

figures

against

comparative

whom

enterprise.

and with

whom

to define

a stance

toward

the

22

WORLD

An

intellectual

stance

POLITICS

is the fruit of our historical

and o.

life situation

on us.When the way the knowledge worlds of our disciplines impinge in the course of the essay several Weber appear, these are quotations to convey not that there is an essential, meant but definitive Weber can understand readings. We him of the modernists, but also as a post important see a text, not as modernist. That puts me in the camp with those who a determinate as a but that emerges having having meaning, meaning as a collaboration between reader and author and that is in turn em rather

that Weber

to different

is subject

as one of the most

in history. Such an approach not is constructed; only phenomenal we within which live are constructed. bedded

world, we cannot a useful Marxism The

following

construct simply for our times. are the quotes

to that the presumes meaning texts but also social categories Yet even as we construct the

any world

we want

to be considered.

to?to

refashion

First Weber:

A product of modern European civilization, studying any problem of universal history, is bound to ask himself to what combination of circumstances the fact should be attributed that inWestern civilization civilization, and inWestern we to which have like cultural (as think) lie in a line only, phenomena appeared of

development

having

universal

significance

and value.10

And Foucault: is one that seeks to reconstitute the overall form of of a total history project a civilization, or of a society, the material the principle, significance spiritual, common to all the for their cohe of a the law that s phenomena period, to sion ... [but] the of a total In of history begins possibility disappear.... place was to some of reason, which the continuous traced back invariably chronology inaccessible there have scales that are sometimes very brief, dis origin, appeared to a tinct from one another, irreducible law.11 single

The

are cited in tandem for an obvious reason. The in quotations was to the Protestant that harbors troduction the Weber Ethic quote These

an ambitious architectural call a edifice, what Foucault would part of as ofWestern total history, tracing the evolution and world civilizations a continuous of the progress?or failure?of rationalization chronology one side ofWeber, in the world. The quote represents that part of him that

still thought

it possible

10 Max Weber, The Protestant Ethic Charles Scribner's, 1958), 13. 11 Michel Foucault, The Archaeology theon Books, 1972), 8-9.

to carry on the kind

and the Spirit of Capitalism, ofKnowledge

of world-encom

trans. Talcott

and the Discourse

on

Parsons

Language

(New York:

(New York: Pan

SYMPOSIUM ing projects

favored

by

so much

of nineteenth-century

especially in. Despite

social

science,

the deep differences between Weber,

a com their projects were energized by on an was the world It moving upward perspective. an inner reason its future, driven that moved by

and others,

Durkheim,

Marx, mon

23

enlightenment toward

trajectory

straight or dialectically toward a climactic condition, with theWest

the vanguard.12 The Foucault tle wholes, terminacies; pieces whose

the consistency on anomalies; tendencies opposing

disrupt to focus

in their own

trable face

is the marker

quote

as anomalies

sumed build

to

in

to disman of a contrary tendency, of de large schemes and powerful to cut up wholes into discontinuous

can be examined, rather than sub to that schemes; prefer representatives a smooth, rather than offering impene

in holistic contestation

to the world.

in theWeber

The tendencies highlighted

quote were picked up, ra

and systematized tionalized, by the structural-functionalism amplified, that animated the comparativists of the 1950s and 1960s. Moderniza tion

theory

elaborated

paradigms

to deal with

political

and economic

development that would allow the field to move from, as Gabriel Al to the put it, "an 'area studies' approach study of foreign a to and one."13 systems comparative genuinely analytical It was Talcott led the new analytic; he more Parsons who

mond

political than

any so in the United the reception ofWeber States. The vari cial world, he affirmed, the could be understood pattern through diads: ascription/achievement; ables, a series of oppositional affectivity/ affective neutrality; orientation/self-orientation; collectivity particular As the usage of the pattern ism/universalism; diffuseness/specificity.

other

occasioned

construc the items on each side of this dichotomous evolved, seen to be related. Thus, systemically affectivity, collectivity were an inner were mu and orientation, ed by logic particularism was to be ori to human be oriented toward beings tually constrained; to care about ented toward collective and obligations particular persons

variables

tion were

or

objects

rather

than general

principles

and

so forth.14 And

affective

12 see on the modern and historical perspective For an illuminating philosophical project, Stephen (New York: Free Press, 1990). Toulmin, Cosmopolis: The Hidden Agenda ofModernity 13 The Politics of the Developing Areas (Princeton: Princeton and James Coleman, Gabriel Almond this quotation was characteristic of most comparative stud University Press, 1960), vii. Parenthetically, ofWestern viewers the project as something carried on by a community ies in the way it conceived in and transnational social science communities looking at the "foreign." The creation of multinational the idea of the the 1990s, as "native" social scientists have broken theWestern monopoly, compromises and has softened the imperialism of categories typical of the 1960s. "foreign" 14 Talcott Parsons and Edward Shils, Toward a General Theory ofAction (Cambridge: Harvard Uni in this essay, saw the that Parsons and Shils themselves, versity Press, 1951), 77. It is not self-evident

WORLD POLITICS

24

were

universalism

self-orientation,

neutrality,

similarly

thought

to be

ed. two

The

one

which

"growing" chotomies

of the pattern variables were within the framework a in which of history, the march could understand process and "developing" meant moving from the left side of the di sides

man cast of the framework to the right. The ideological a necessary of history toward future. The march one a cul-de routes: one the offered only two possible high road and sac. clo drove this agentless process. The analytic promised Nobody action all of the of had been confidence that sure, possible permutations for: "We maintain ed that there are only five basic pattern vari movement

dated

that, in the sense that they are all of the pattern variables so derive, a they constitute system."15 a more more set out to Parsons (and Shils) parsimonious, produce of tradition of the theoretical consistent and ra summary oppositions and

ables, which

embedded

tionality

in the Weberian sense. As

in the Foucaultian

project tee on Behavioralism

at Harvard

aims activity which scholarly all the fields of the behavioral

It was a totalizing told the Faculty Commit "A long-term of program

architecture. Parsons in 1954:

at no

less than a unification

sciences

is now

envisaged."16 could only be disrupted the possibility that so from oppo take elements

were

elements as awhole, cial actors

related, the system systemically not in parts. It did not accommodate

could, as it were, mix and match, to create new mixes. site sides of the variables or institutions

actors way

of theory in Because the

mixed

to the predetermined

features, future or

they

No

were

lamentably

If social hybridities. on their "transitional," it. unable to make

Weber had laid the basis for the theoretical framework that surmised a fatalistic

and ive theory. Weber's

East

lurking

behind

Parsonian-based

modern

on ei dichotomizing project in which humans were the ther side of a great divide: those civilizations instruments and those inwhich humans were the ves and tools of God ization

sels of God;

on

the one

sorted

side

the ethical

religions,

civilizations

which

led to this

the examples related. However, left and right side of pattern variables as displayed above systematically are in the essays?which such a grouping few?suggest (p. 79). Francis Sutton and Fred W. Riggs de that history was moving from one set of char veloped systematic models, complete with presumptions For Riggs, see "Agraria and Industria: Toward a acteristics to another. Sutton is so cited in Almond. inW. J. Siffin, ed., Toward a Comparative Study of Public of Comparative istration," Typology to The Politics of theDevel in his introduction istration 1957), 23-116. Almond, (Bloomington, theoretical polarization" and stresses the em oping Areas (fn. 13), explicitly avoids such "unfortunate structures beddedness of traditional in modern (p. 23). 15 Parsons and Shils (fn. 14), 77. 16 The Behavioral Sciences atHarvard (Cambridge: Harvard Uni Report by the Faculty Committee, versity, June 1954), 114.

SYMPOSIUM

25

to remake the world; asceticism and the accompanying urgency worldly on the other asceticism side exemplary that religion, otherworldly shunned activity and accepted the cosmos, that is, fatalism.

The first "field trip"on which Lloyd Rudolph and I embarked began

to unravel

this

inherited

apparatus.

sheer noisy,

The

shouting

nature

of

the data encountered flew in the face of the dominant dispositions of theWestern

social

tuted. The an

experience

we by which not only created

as observers

science

our first book

of a social

dered

were

science

but

that could

then

consti

also engen so mislead.

suspicion enduring to be critical for "voluntary which we believed associations," Looking we over to the civil for pluralist democracy, stumbled society necessary caste. That most diffuse, particularistic based, hierarchical, ascriptively institution of Indian society, said to be generative of the fatalism of the an social hybrid that sat spawning adaptive spawned/was on the divide between "modern" and "traditional." comfortably was a most I What and encountered inventive mix Lloyd Rudolph

East,

had

ture of pattern their making traditional

re persons who were an version of by working through adapted in other words, their fate their grasping

variables?ambitious social

fate

collectivities?in

lower-caste

caste itself was In its democratic incarnation, deeply disruptive we were we caste of To for the seeing, hierarchy. phenomena nature ofWeberian had to break up the totalistic, and Parson systemic ian structures.17 We of social change elaborated the idea that the world

hands.

it was confined within badly imagined when change seemed gories and systemic theories. Much tations not the oppositions. Adaptations captured by on one or the other side of mixes of features lying line. ables or the activist and fatalist civilizational

was

cate oppositional to proceed by adap were created out of the pattern Dichotomies

vari were

on which structures that suppressed the intervening ground logical most exists. The world of the phenomenal that constituted categories or the "modern" were not systemically and related, asWeber organically were not and would not be uni and Parsons and Marx thought, they had become related in the specific history of theWest and as yet unknown in the future of the "East." Other shape out different of relationships. modernizers would constellations figure a is total The Foucault cited here of quote critique history, exempli versal. They take

would

fied in the hypersystemic side of theWeberian sonian

17 We

variant.

Instead

of

total

histories,

did so in Rudolph and Rudolph, The Modernity of Chicago Press, 1967,1984,1996).

(Chicago: University

project and by its Par

singular of Tradition:

and

ideological,

Political Development

in India

WORLD

26

POLITICS

histories, space for al providing a ternative social critique of the understanding, or was in modernization that embedded theory, change, "development," interest in the Foucault driven by a continuous chronology of reason. My and disruptions is not about suspend ian suggestion of discontinuities

Foucault

and open multiple His is also destinies.

intimates

it is about dismantling and explanation; ing comparison, generalization, us and the non the all-encoming formats that have imprisoned we were to nonholis societies trying explain. Discontinuous, European more to represent nature of it the contested tic projects make possible or sup to to avoid contradictions truths "larger" "reality," subsuming means in interest of the them "consistency." Disrupt pressing altogether a we so as we not tones construct in the do of break and challenge: story truth, of absolute observer, of objective certainty, but in that recognizes we are both partial and partisan and that we ex indeed may be the only truths there are.18 It press half-truths?which as academic means a less enthusiastic embrace of attributes regarded the omniscient

a voice

such as consistency, parsimony. objectivity, a in developing offers us considerable Weber, surprisingly, more stance to The quotation featured different, contingent knowledge.

virtues,

at the beginning

of this essay

embodies

his

ideological, universalizing, dichotomizing

attachment

to a totalistic,

project. But embedded in

assertion of a confident In the midst of that quote there is an anomaly: he emits a kind of episte Western of the universal culture, significance ". . . inWestern and inWestern Civi Civilization, mological hiccup: we like to which have cultural lization {as appeared only, phenomena

think) He in a line of development having universal significance" (em

in confidence, that sudden pause It is that self-doubt, phasis added). to suggest the "other Weber." that begins moves not "there" in any trans in a world in which Weber reality is parent fashion "All knowledge

but

is constructed

out of the

situation

of the observer.

"is always knowledge from reality," he writes, so so is that infinite, comprehensive, particular points of view."19 Reality severe to grasp it at all, and the human mind must make compromises situation: "The way which arise out of the observers these compromises an infinite us in immediate concrete it presents life confronts situations, and disappearing of successively and coexistently emerging multiplicity a "selects from an absolute events."20 And the "scientist" tiny infinity 18 For a discussion

of cultural

of types of voice and their relationship to truth claims, see Donald N. McCloskey, of Chicago Press, 1990). So Smart (Chicago: University IfYoure " 19 The Methodology Max Weber, of the 'Objectivity' in Social Science and Social Policy," inWeber, Social Sciences, trans, and ed. Edward A. Shils and Henry A. Finch (New York: Free Press, 1949), 71. 20 Ibid., 72.

SYMPOSIUM

27

portion with the study of which he concerns himself."21 "Accordingly, in our sense

science

cultural

involves

insofar

'subjective* presuppositions

as it concerns itself only with those components of realitywhich have some

relationship

...

to events

to which

we

attach

cultural

signifi

cance."22

not an unmediated, reality is clearly never tells us there is no reality; he does

con unproblematic not refrain from using and he does not put it in quotation the postmodern marks, to us for signaling reality skepticism.23 Yet by representing comes a to he and ungraspable, similar position: very close

For Weber, cept. He the word; convention as infinite

and disappearing"; "infinite multiplicity." These "coexistently emerging are not certainties. then that He that the suggests phrases only signal acts of construc is in fact way to come to grips with "reality" through

tion.We

cut out pieces of the infinite flow in light of our subjective in

ideal type allows an unnatural of a moment freezing and it. and the of "reality" extracting possibility addressing terests. The

of

Iwould interpret the "subjective"here to refer to the individual who dwells within

social

communities

digm

and historical on

the other

"cultural significance." (InWeber must

choose

contexts that

on the one hand

suggest

the range

and para is of

of what

this "I"?the lonely individual who

from

of any absolute extracted from

without the guidance among contending possibilities And which of the pieces standard?is unconstructed.) is "true"? Weber the flow that contradictory suggests truth value. At the end of the Protestant Ethic, he offers

cuts may have a denial: "But it is, of course, not my aim to substitute for disingenuous a one-sided an in materialistic causal spiritualistic equally one-sided of culture and history." And then he tells us, "Each is terpretation the culturalist Indeed, Weber constantly equally possible."24 commitment. He materialist refuses exclusive interpretations.

invokes

in a more work contingent clearly place Webers to truth than the discussed relationship totalizing projects above?pro to encom all truths in a single systematic formu jects that aspired as rule that subsumed all that the lation and anomalies prove exceptions This

denial would

rather than as ruptures. It would place him of his own more holistic, macrotheoretical

on the other

side not merely self but also of, say, Parsons,

21 Ibid., 82. 22 Ibid. 23 isworth a biography. It is the caution The career of the inverted comma in academic disciplines on the road to cognition, warning the unwary that aword may not be what it seems, telling the light reader that she should walk all around "sovereignty" or "growth" and have a conversation with it about its provenance before incorporating it.Note thatWeber does put inverted commas around "Objectiv article (fn. 19). ity" in the above-cited 24 Weber (fn. 10), 183.

28

WORLD

POLITICS

whose strive to en andWaltz, theoretical formulations Kaplan, a and subsume the relevant world within com single phenomenal

Easton,

system. Let me similated

sum up then we can read Weber, is generally who why as a to the realm of the modernists, The postmodernist.

as case

on emphasizing the multiplicities with which he continually

rests?rst undermines

the universal

he puts

propositions

on his propen

forward,

sity continually to disrupt his own holistic architectonic; secondon his ambivalence profound march of rationalization

about

the

in human

about the project, he sees as a destruc

enlightenment history, which

tion of all that is intentional and sensitive and varied and centrally human, the world as iron cage; and third, his belief in the multiplicity and simultaneity an absolute

with

This

of "truth," his epistemological polytheism?combined that persons must choose. conviction, ungrounded, In the of the deserves face and si ambiguity emphasis.

last point of truths multaneity

that is a significant

part of the postmodern

people tend to be overcome by a fear of falling.Where is there? There is none, certainty solute but convincing inconsistency, choose, must act.

what

project,

all is in flux,

and adds, with ab must of integrity

saysWeber, the person

as a And why should we readWeber and what has it postmodernist, me to do with method? has the opportu story given comparative My as a to consider, who lived through the era of holis nity comparativist social science and was seduced by it, where we are tic, macrohistorical now,

and why

I think

theory

necessarily

power,

although

"now" is better.

or

entailed merely that was an aspect

I do not believe a colonial of it. But

that macrosocial as of knowledge the macrosocial

project because

projects of the 1960s were holistic, theywere under a logical and intel lectual

compulsion

to suppress

alternative

truths,

to

imagine

thatWest

ern history was world history and suppress the possibility of multiple a culturally narcissistic projects prepared a such prepares the ground perspective Disrupting as agents in the field of an open operate history.

histories.

Those

West.

mirror

for the

for scholars

to

James C. Scott I am not very good

at

capturing

in the abstract

how

I practice

theoret

icalwork in comparative politics. I am better at doing it than at dis a figurative pistol to my temple, I did cussing how to do it.With actually

sit down

and write

an

important

maxim

of comparative

work

that derives from the way I do comparative politics. Iwill be happy to

29

SYMPOSIUM mention

that toward

the end. For

the body of my discussion, however, Iwould like to share some of my current research, which, of course, em to be done. bodies my for how comparative preferences politics ought in understanding I am interested disasters. It would development seem

to be to understand matter relatively straightforward why people ethnic and so on decide that nationalities, groups, religions, one another. It is more difficult?but want to murder they considerably were meant understand that for important?to ostensibly why schemes

of different

the betterment

of the human

condition

have

in practice

turned

out

to

be deadly. Let me

an argument about what I call state begin with simplifications, states need to understand in schematic that is, about why societies to I the readers of the ways. hope persuade potential danger of these even are I would also schemes, argue, they necessary. Certain though, a forms of knowledge and control require of vision; this tun narrowing nel vision

offers

into sharp focus limited as of bringing and very complex very unwieldy reality. This at center the phenomenon the of vision far more

the advantage

of an otherwise

pects

makes simplification and hence far more legible on the one hand

culation

use

I will Prussia

the

susceptible and to control

invention

and Saxony

of scientific as amodel and a

This

somewhat

to careful measurement

and cal

on the other. and manipulation in forestry eighteenth-century

metaphor obscure

for state simplifications. at the outset. It is used

example may appear as a for the forms of knowledge and manipulation char metaphor acteristic of of large institutions with very sharply defined interests, not the the outstanding which the state is perhaps example, although here

only forest

one. To

see how

management

and manipulation simplification see us to how a similar optic allows such

other fields of endeavor, including, Iwill in the social sciences. knowledge generation to Prior the invention of scientific forestry, state to view its tended forests pean through

operates operates

in in

suggest, in the process of the early modern the lens of fiscal

Euro needs.

concerns like timber for masts and ship is, aside from particular some construction, wood for sufficient fuel the building, population, crown viewed its interests in forests the and of course hunting rights, a to its them and fiscal lens, reducing single number, by large through

That

that number being the revenue yield of timber that might be extracted

annually.

To

appreciate

how

heroic

this constriction

left out of the states field of vision. Underlying indication representing

of revenue so many

is, notice

what

is

this single numerical

are not so much trees as commercial wood yield so of of feet salable timber, thousands many cords

WORLD

30

POLITICS

a certain

is everything of course, price. Missing, else about hold lit trees, bushes, and plants?that as well are all those tle or no potential for state revenue. Missing aspects of trees, even of revenue-bearing be useful to the pop trees, that might into fiscal I ulation but whose value is not convertible receipts. Here of firewood

fetching the forest?all

in mind

have

for domestic

the uses animals

those

of foliage and people,

as fodder

and

thatched

trees and branches

ing, hop poles, kindling, bark, roots for medicine

as

as food

roofs,

fenc

bedding,

and tanning, sap for

uses its vast number actual tree with of possible resins, and so on. The a certain volume an abstract tree is of fire by representing replaced From the perspective of a naturalist, of course, most of the forest wood. are the vast is also missing. of flora, the grasses, the Gone majority the mosses, the shrubs, the vine; gone also are the s insects. And and from an anthropologist amphibians,

the ferns,

flowers,

reptiles, birds, on human interaction with the perspective, touching nearly everything to forest is also missing. that impinges poaching Except for its attention on the states claim to revenue and wood or its claim to royal game, the the vast and complex negotiated social uses of the typically ignores for hunting and gathering, pasturage, fishing, digging minerals, and charcoal making, food collection, magic, worship, refuge. trapping, see the real state cannot You could say that the utilitarian existing state

forest

trees.

Its partial and abstract view of forest is is the discourse of natural resources; it is the same in into crops and the the plants we are interested

for the commercial

forest

hardly unique. This that makes discourse

in into pests. It is the that ingest the crops that we are interested are we turns in into game or live animals interested logic that and stock and those that compete with or prey on them into predators varmints. However, it is not the abstract and utilitarian which the logic insects

same

state what can

to the forest its officials through applies is distinctive is that the state can make impose

observing

them?at

least to a certain

that

is distinctive;

these

categories the reality

extent?on

rather, stick and that

it is

and manipulating.

invention of scientific forestry in at the end of the eighteenth century was originally motivated by the need to rationalize princely finances. The forests played a very important role in all of this. The earlier system had been quite crude; it entailed simply dividing up the domanial forests and either felling portions or letting the rights to The

of it. Since the trees of great value were never randomly fell portions zones of the forest, the actual distributed, given the varying ecological revenue to There was thus an effort to tended fluctuate enormously. inventors of scientific forestry spent The that from happening. prevent

SYMPOSIUM a

deal

great

a

was selecting what they thought set about and then large plot, conducting

of time a

quite

plot,

31 representative a in complete

ventory of this plot. It involved getting some twenty people abreast in sight of tray with

one

another

with great trays carried around their necks, each each of the bins carrying a known number of nails of a to a certain size class of trees that color, corresponding designated these surveyors had been trained to recognize. At each tree, they would five bins,

a tree. At the end of the ex the size class ofthat put in nail designating one the of nails and arrived at a counted number ercise, remaining trees of each size class were in the for complete inventory of how many est. On

the basis

of assumptions

about

rates of maturity

and growth,

one could then divide up the forest into a set of subterritories thatwere as even out the amount year by year and, as much possible, exploitable of wood that could be taken. There was no way of quite controlling the con at amount of market for but least the wood, wood, assuming price stant be roughly the same from year to year. prices, would com efforts to determine exactly how much in a tree of a given size class. In a series of experi and take one such tree, cut it up into tiny pieces, ments, they would to get the exact vol it into cubes so that it could be measured compress There

mercial

were

wood

also other

was

ume of wood. Or they would throw the wood with could

based

water; determine

achievements to calculate

on

the amount

the exact volume of German

sustainable

of water of wood

into in large bins filled that was

in one

they displaced, of these trees. The

science in standardizing forestry techniques of commercial timber were very impres yield

sive in this respect. is decisive What

for our purposes is the next logical step in forest to and create, management. through careful seeding, planting, a forest that was easier for state foresters to count, cutting, manipulate, science and geometry, and assess. The fact is that forest measure, to transform and the disorderly backed by the state, had the capacity That

was

into a forest more

the istra closely resembling was tive grid of its techniques. Thus the underbrush cleared, a number was done of plant species were reduced, often to monoculture, planting rows for some tracts. went in and large simultaneously straight They a a in Normalbaum these "normal" forest and distance toward creating chaotic

real forest

forests.

The creation of a simplified, legible forest was only the imminent logic

of

practice, isting

its techniques. It was not and could not nature and the human factor because both

topography,

the vagaries

of fire,

storms,

blights,

ever be realized intervened. climatic

The

in ex

changes,

WORLD POLITICS

32 insect populations, in their efforts to

to thwart foresters obviously conspired the actual forest. Also not insignificant were the

and disease shape

insurmountable difficulties of policing large forests; for 250 years the most popular crime in England was poaching. And by popular, Imean senses

in both difficulties

of the word:

most

common

and most

beloved.

of policing

Given

the

human

large forests, therefore, adjacent popula to graze animals, continued and firewood poach typically use and generally make of the forest in ways charcoal, kindling, make the realization of forest management that thwarted plans. tions

also

The

cannot

forest

s

at

be the naturalist

s forest. Even

if all

in the forest were

interactions the ecological known, they would play a so as to and de constitute shorthand defy reality variegated complex to to The intellectual filter reduce this necessary complexity scription. was case in the interest state's dimensions this manageable provided by in commercial

as Iwill

world?or and

wieldy

and

is, in its natural or social this abstraction. without

That

manipulation. cratically

too

in revenue.

You could say that the natural on to too un the social world?is go argue, simply raw form for direct in its istrative complex

timber

indigestible

context I now

reality propose

is bureau to

apply

the forestry metaphor to land tenure, by giving a hypothetical descrip tion of customary forms of land tenure. Though however, hypothetical, is realistic; that is, I have read about or actually seen all this description I am about to describe. of the practices in which families have rights to community season. main growing certain crops Only parcels of cropland during the seven use are be and redistributed years may every rights planted, to size and the number of able-bodied among families according family Let

us

imagine

a

communities of of Russia, Java, portions corporate and Vietnam, this was common. After the harvest of reverts to commons, such that any crop, all cropland even its fowl and livestock, and family may glean, graze plant quick on the margins Edible wild crops. maturing dry-season plants growing adults.

In the closed

Central

America, season the main

of

fields

known

or courses are available water to be Trees gathered. along are to have been the with their fruit, property together planted,

of the family that planted them, not the property of the people on land they happen now to be located. Fruit fallen from such trees, one of its a however, may be gathered by anyone. When family fells

whose

trees or when

it is felled by wind, the trunk belongs to the family, the

to the immediate and the tops, twigs, fronds, and neighbors, carries them off. Certain leaves to any poor villager who plots of land or are for use or lease by widows with children by depen designated

branches

SYMPOSIUM dents

of

conscripted

males.

Use

rights

33

to land can be

let to anyone

in

the village, but only if no one in the village wants to take it can it be let to anyone outside the village. Let us also imagine that fishing

rights

are distributed

so that anyone

may fish by net or by hook and line from canals and streams. In flooded fields, however, while anyone may fish with a hook and line for small fish, the larger fish, taken usually when the field is drained (with wet rice, for example), belong to the owner of the crop growing in the field. I could

on what to land to the dis easily elaborate happens rights and is not necessary. of crops and their proceeds. That, however, I more elaborate because it is, in itself, a this description also make

tribution could

simplification. I also distort by describing these practices as if theywere

in fact they are a living, negotiated tissue of practices that is laws, when never static, and that is to new according ecologi constantly changing I must add, power relations. cal and social circumstances, including, awritten to represent law that attempted system of positive Imagine this complex relations and land tenure; that is, imag skein of property to respect customary ine a state that wanted forms of tenure by codify

ing them and representing boggles that would could

at the number

them in positive

of clauses

be required to represent the result would be codified,

and

law. The mind

subclauses

these practices. necessarily

and

fairly

sub-subclauses

In principle, sacrifice much

if they of the

plasticity and subtle adaptability of practice; itwould freeze a living in the positive code process. Moreover, changes at best, represent reflect evolving practice would,

to time

from

time

a very

jerky mechani

to

cal adaptation. And what of the next village? After this imaginary code giver had finished representing customary land tenure in village A, he or she would

have

to start

from

nearly

zero

in village

B because

it

would have different ecological conditions, a different cropping history, how devilishly structure, and so on. No matter clever, a state was, it simply such and of law meticulous, customary respectful its own particular his could not represent each village with ecology, and economic It tory, cropping patterns, activity. kinship alignments, a different

would

social

require for each village, communities.

a different making

set of regulations, for at least as many

a different legal codes

code positive as there were

But notice whose of course, this is a nightmare. istratively, cus not that of the it is?certainly whose particular people nightmare tenure toms are is in land Local practice completely being represented.

comprehensible to the inhabitants who livewith it day in, day out. Its

details may

often

be contested

and far from

satisfactory

to all of its local

34

WORLD

POLITICS

practitioners, but there is no doubting its familiarity. They know its subtleties and while they may dislike it, they know how to manipulate as well. The it for their own purposes in then, is nightmare question, that of the state officials who want a uniform istrative grid of land a or for that mat tenure that will represent the situation in whole region ter in a whole state is nation. The very concept of the modern hardly conceivable that

is

a

without

vastly

and hence manipulable

legible

and uniform

simplified

property regime the center. And, of course, a

from

major driving force behind this legibility is the question of taxes. next

states is, of course, to measure, step for all modern codify, tenure in much the same way that scientific land simplify forestry the forest, since the state cannot begin to reconceived the incorporate luxuriant variety of customary land tenure. The historic for solution, The

and

the liberal state anyway, has been the heroic simplification of individual freehold tenure. Land is owned by a legal individual who disposes of wide

powers

of use,

inheritance,

or sale and whose

is repre

ownership

sented by a uniform title deed, enforced through the judicial and police state. Just as the flora of the forest are reduced to so were tenure arrangements the complex of customary Normalb?ume, an to freehold reduced transferable title. In practice agrarian setting, the a uniform was blanketed with istrative landscape grid of ho institutions

mogenous hence,

of the

land, each parcel taxpayer. The modern for

equivalent growth

land

and yield. A

tenure central

of which

as owner, legal person and its tax roll were thus the

had

land scientific

of the

consequence

a

forester's

table

of its imposition

of timber was

that

it

radically devalued local knowledge and autonomy, and this specialized knowledge, tween the

backed

state by authority, and the state.

changed

the balance

of power

be

locality like the forest, land tenure never quite measured up to the Again, envisioned the who did the settlement reports and the utopia by people a whole taxation maps and who issued title deeds, because series of on went in spite of what the paperwork resistance, practices, including state is that most said. One of the important about the modern things at least one of its officials are of necessity step, and often several usually the reality and the citizens steps, removed from direct with they ister. They observe and assess the life of their society by a series of simplifications and shorthand fictions that are always some distance are meant to capture. The from the full reality these abstractions func sees the human activity of in tionary of any large organization actually

terest to him or her largely through the simplified approximation of

documents

and statistics

and tax proceeds.

SYMPOSIUM The

simplified

facts

35

are of a certain necessarily of facts that give a time se series of individual, static facts can

of the

are a series

when

type?static?even they A observations. quence of different never tell you what A and observation between observation happened a two is not drawn between the that line observations B; is, straight an accurate events. The to be of facts have representation necessarily are not cannot be if they and as because standardized, they aggregated in a way that allows for an overall view. The facts also have to sembled

be stylized to simplify a reality out there.Thus, if the objective is to di vide the population into the employed and the unemployed, then one a whole statuses that represent series of intermediate the simplify or are in who real life of the many thousands employed unemployed must be sorted that defy easy categorization; ways they complicated

must

into one bin or the other tion be divided

to these

assignments

the exercise

because

into these

categories. know

categories

requires that the popula who actually make the people the fictional and arbitrary quality

The

underlying each of these decisions and know that they hide awealth of variation. problematic are established, these categories Once as if all cases classified together eluctably uniform.

All Normalb?ume of statistical

the purposes

of course, they operate are, in fact, homogenous in a given size range are exactly the same auto workers, if we All manipulation.

classifying by industry, are the same.All Catholics,

in and for are

ifwe are classifying

un are the same. Students of bureaucratic behavior by religious faith, schemes do their work most effec derstand that central coordinating is known and where the task environment conditions tively under it can be treated as a closed I and where system. What unchanging, state want to argue about is that the modern state, simplification its official attempts and with varying success, creates a popula through tion with monitor,

those count,

characteristics

standardized assess,

and manage.

That

it will be easier to because is a little bolder than Iwould

like to state it and itmisses the fact that these efforts fall enormously short. However, or to represent their jurisdiction trate their ability enormous

I think

there

as transformed

is an effort by officials either to transform the population, space, and nature under

into a closed to control

surprises when

system, without it.This and observe do not work

things about the state

the surprises often

provides out as well.

that frus for some

in this context is that it in par important thing to make its categories stick and simplifications ticular has the capacity to its If it can insist on treating because categories. according people are to to defend real property, you your claim you want normally The

WORLD POLITICS

36

obliged to defend itwith a property deed and in the courts and tri bunals created for that purpose. If you wish any standing in law, you must

have

birth

the documents?the

and

so on?that

used

by

state agents

more

as

officials

certificate, port, identity card, of The categories proof citizenship.

accept are not a means merely are also an authoritative

legible, they tion has to dance, at least some

to make tune

their environment

to which

the popula

of the time.

If I had more time, Iwould make the argument that these simplifi cations

become

deadly

in combination

with

other

factors.

most important of these is an ideology of high modernism, nineteenth-century

worldview,

with

its supreme

One

of the

that is, the

self-confidence

about

the inevitability of linear progress, the development of scientific and the expansion of production, and, most knowledge, impor the rational of the social the satisfaction of tant, order, growing design human needs, and, not least, an increasing control over nature, includ nature commensurate with sci (also including ing human eugenics, me seems to entific of It natural that laws). understanding high at its flood tide is an modernism that requires an enormously ideology as state?it has obviously and straightforwardly authoritarian powerful technical

to there is only one answer to pects to it. According high modernism, almost any social problem, and that is determined by the technicians, see these social and who engineers, analysts things correctly according are married to the laws state have devised. When they simplifications

to high modernist plans for the transformation of human life and its and when improvement, weakened by revolution

these plans or war?so

are

on a that is imposed society that it is relatively and prostrate I think, a recipe for the worst dis

easy to push around?then you have, asters of high modernist planners. In conclusion, the parallel I would draw is between how states sim and how scholars the societies they govern plify the societies simplify

but one may well be necessary, types of simplications study. Both never conse also have they may negative forget profoundly a I have never been able to think my way through quences. personally or it in a without fourth-order extraction simplification embedding case. I once concrete cannot understand fourth-order extractions; just or are am in the air, I there four five of them simultaneously essentially

they must

lost because derstand that

I have no empirical I have never been reference. can run I unless them through concepts

abstract

I understand

well.

For

terribly important, proach a field site where they have

able to un

something the anthropologists' ap as a not as a if certainly technique, theory: a lot of time, and co they spend observing this reason

I find

SYMPOSIUM piously

as if were they

recording

a camera,

37 as if were they

and

stupid

na?ve.When Iwas doing my dissertation, I had a colleague who was influenced by Gabriel Almond. He was applying Almond's scheme to and was

Malaysia,

subtle, widely

enormously

a standardized

conducting

read,

scholar was survey. This and clever. He would tell me things

society that he had picked up on the bus or observed

aboutMalaysian

in this little village, but when it came time to write that he knew and had taught me no things

these

his dissertation, all seemed rele longer

that point, he saw Malaysia of the instruments only through his survey. The great thing about anthropology, is that contrast, you by are at work from the moment until you open your eyes in the morning vant. At

you close them at night; everything is grist for the mill. So itmay often to have

be fine

instruments

that will

measure

carefully,

but

if you

see

theworld only through your instruments, then it is likely to be aworld that

to broaden

that may very well be poverty stricken. can reach. instruments define the conclusions you in principle maxims be for that maxim, I'm against although

is hard

These Now

and

cause they have a habit of hardening into doctrines: If half of your read not outside the confines of political science, you are risking ing is in of the notable extinction the rest of the subspecies. Most along with in the discipline have come in the form of insights, perspec novations and elsewhere. tives, concepts, originating Reading paradigms

exclusively within at the very

the discipline is to risk reproducing orthodoxies or,

least, absorbing

innovations

far from

the source. We

would

do well to emulate the hybrid vigor of the plant and animal breeding world.

Theda For my

contribution

Skool

to this symposium,

I draw on my

1994

essay

"Re

flections on Recent Scholarship about Social Revolutions and How to as I revisited the field of the comparative Study them," which Iwrote study of revolution.25 It reallywas a revisiting because I had not done more

than

an occasional

piece

in that

area

since

the publication

of

States and SocialRevolutions (ssr) in 1979.261 used the occasion of gath ering together an opportunity 25 Conclusion

before and after of my own essays written to review the other literature that has accumulated

various

in theModem World Skool, Social Revolutions Press, 1994). Skool, States and Social Revolutions: A Comparative Analysis Press, 1979). (Cambridge and New York: Cambridge University

Cambridge 26 Theda

toTheda

(Cambridge

SSR as since

and New York:

University

oft Russia,

and China

WORLD POLITICS

38 the mid-1970s was

interested

what

I was

in the comparative social revolution. I study of modern or went in how recent built upon scholarship beyond to argue in 1979, and I reentered times the?at trying

critics of my work. These with critics, writing feisty?dialogue number of perspectives?Marxist, rational and what choice, call interpretive historical?have in SSR. lar arguments made They

a

from I would

more than the disagreed with particu have also taken issue with the entire

strategy of analysis that I call macroanalytic comparative history. As I see it, comparative politics people, including comparative polit not ical sociologists, should idea!?and compare?a startling simply that are not America. the purpose Moreover, study places in the world of comparison should be partly variety of theoretical perspectives

to

test from hypotheses explore and and partly to notice and hypothesize

about new causal regularities. This kind of approach, which

a

I call

can be very fruitful when macroanalytic comparative history, applied, or to either by single scholars communities of understand scholars, by of democratization, issues, such as patterns ing substantively pressing the causes

and the origins of na revolutions, or tional systems of social provision the challenges they have faced in concerns the postwar The substantive of of ana communities period. can be am I but the that and that lysts quite varied, approach proposing can in the comparative I illustrate revolutions field be powerful and fruitful

and outcomes

for individuals

of social

and communities

of scholars.

I think it is fair to say that SSRwas an In retrospect agenda-setting on several counts, not because book, everybody simply agreed and that was the end of it, but rather, in a much more way. Some interesting new theoretical to argue with it chose and and method people develop in a self-conscious

or in dialogue disagreement the arguments. ap picked up methodological of Almost who took everyone up proach macroanalysis. comparative the study of revolutions after SSR did things that almost no one had

ological with my

approaches

Other

scholars

it: they put the state as a set of istrative and coercive at the center of at var organizations analysis; and they looked in depth cases. ious historical

done

before

extent and richness of the his I look back, I am struck by the and comparative historical work on the causes of modern social on to na revolutions. Most ofthat work has moved study Third World some now at tions in the mid-twentieth of it is and century; looking situations like Eastern Europe after 1989. Some of the work consists of When

torical

studies that are theoretically informed. But quite in-depth, single-case a bit involves two or more countries, much and juxtaposing comparing

SYMPOSIUM

39

as I did in SSR.In my opinion, there have been two studies about the causes of social revolutionary transformations inThird World nations in the mid-twentieth century that are blockbusters. Both are full com parative

of both

macroanalyses

cases

and negative

positive

(cases where

revolutions happened and cases where they did not); these books are sGuerrillas and Revolution inLatin Amer Timothy Wickham-Crowley ica and Jeff Goodwins State and Revolution in theThird World?7There is not

sufficient

on the designs

to elaborate

space here

and findings

of

these studies, but they arewell worth looking at; they show the fruit fulness

of the macroanalytic

to

approach

comparative

politics

that SSR

helped to launch. The comparative historical macroanalytic work of the last fifteen or so years

about

the causes

of social

revolutions

from

the 1800s

forward

has provided us with a lot of solid knowledge. The research comes from a series

of studies

now know

that built

each

upon

other's

insights

and methods.

are not regimes susceptible particularly certain authoritarian We know that revolutionary challenge. regimes are not to be either to strong revolutionary likely susceptible challenge or to actual of rulers. We have a good idea of the partic displacements ular kinds of authoritarian in which revolu regimes and circumstances

We

that democratic

to

tions have

actually succeeded?ranging to monarchies patrimonial

bureaucratic

from

certain

kinds to

dictatorships

of agrarian (as op directly

posed to indirectly) ruled colonies. We also know quite a lot about which kinds of international pressures and disruptions contribute to the of regimes or to the strengthening constel of revolutionary sets of forces know less about exactly what shape revolu but we do know something about the circumstances outcomes;

weakening lations. We

tionary that promote in almost

centralization, all revolutionized

In turning

now

to more

and often military

authoritarianism, regimes. theoretical

and methodological

matters,

coups the

main purpose here, I should like to focus on challenges that have been posed

to the macroanalytic

approach,

to whose

value

for the field

of the

comparative study of revolutions I have already alluded. I deliberately referred challenges,

to substantive because

achievements comparative causal patterns.

and varying complex eral model and go out

and

apply

before

considering

methodological deals with macroanalysis a gen cannot One simply develop in it universally, place after place.

historical

27 Guerrillas and Revolution in Latin America: A Comparative Study of Insurgents Wickham-Crowley, and Regimes since 1956 (Princeton: Princeton University State and Revolu Press, 1992); and Goodwin, tion in the Third World: A Comparative Analysis of California (Berkeley and Los Angeles: University Press, forthcoming).

WORLD POLITICS

40

is therefore vulnerable history Comparative not theoretical enough. From another point across contexts involves proach comparisons

to the accusation

that

it is

of view, however, this ap that historians often con

can also be on sider to be unique. So comparative history challenged to the richness and uniqueness the grounds that it does violence of each . case. Both were leveled at SSR The types of challenges only appropri ate answer to these to is show that important objections polar-opposite causal theoretical, tive macroanalysis,

are

being asked by those doing compara about substantive in the world. processes that there has been progress substantively,

questions

questions is to show, response answers to those in questions. developing to SSR and other, similar books are of inter The analytical challenges care about the causes and outcomes est not of rev only for people who same sorts of olutions. The appear in many other literatures challenges

The

best

about

matters.

substantive

other

like this to discuss

posium

That

these

iswhy

in a sym

it isworthwhile

challenges.

I shall be brief about the first theoretical-methodological

challenge

I am not sure how many people are likely against SSR, because a to take this challenge seriously. A few years ago Michael Burawoy, a broadside at Marxist leveled Bura sociologist Berkeley, against SSR.2S that SSR and all comparative of its type are work woy asserted analytic not start from a fixed set of theoret flawed do because inherently they to most ical premises. According Burawoy, Marxist theory has been the for understanding situations; he values the the revolutionary powerful directed

a commitment on the it involves because ory, moreover, part of the as well as to understand scholar to promote it. Burawoy crit revolution to SSR icized for "abandoning" Marxist any commitment theory. He holds

that

refutable

as a source of than using Marxist theory empirically to operate within Marxist it is better hypotheses, theory, rather

bringing historical cases to bear on it for the sole purpose of enriching that theoretical Now, ment

even

tradition. from

coming sort of argument such perspectives, criticizes

of this journal were to reject such an argu a in fact the same perspective, Marxist-Trotskyist be from could actually made other kinds of theoretical

if the audience

as the rational

comparative

analysts

choice perspective. further Burawoy such as myself for apparently reasoning

by empirical induction rather than by theoretical deduction and thereby set of theoretical in an entire untestable smuggling propositions. our selection of cases and hypotheses. this, he believes, through 28 Michael

Burawoy,

ety 18 (1989).

"Two Methods

in Search of Science:

Skool

versus

Trotsky,"

We

do

Theory and Soci

41

SYMPOSIUM

the space here to elaborate on either the details of these or on my responses to them in the "Reflections" accusations essay.29 If versus is Trotsky the standard then the kinds of comparisons Skool, I do not have

that

I made

in SSR?particularly

the comparison between 1905 and a 1917 in explain why purely working-class-based theory Russia?help or where a to be able to of revolution is simply not going predict when with revolution will succeed. Marxist other theories, theory, along should be a source of hypotheses for comparative historical explo if the hypotheses do not accord with the comparative evi not it. refines assumed history only Comparative it can also suggest new and combine traditions; hypotheses in a far more powerful way than can from old perspectives

but

rations;

dence, theoretical

so be

hypotheses the essentially

advocated dogmatic by Burawoy. approach now move on to a second ismore pertinent challenge, which a on to because itwas issued by a ra comparative symposium politics, Let me

tional

theorist, Michael

choice

rator, Edgar

a

Kiser).30 Although that has a much more

perspective ence than

it does

"scientific"

unless

way

with a collabo (in conjunction a Hechter sociologist, speaks for in political sci substantial presence

Hechter

in sociology. He and Kiser argue that theorizing is not one can offer a a in model and deduce general logical one is to that in a sense cover the patterns pointing

hypotheses historical

or inductive

through

historical

description.

Their

essay

trum

pets the value of model building, as opposed towhat they argue is the arbitrary

of comparative

approach

history.

They

suggest

that compara

tive history is really just old-fashioned history, with fuzzy concepts and analysis

to

tailored

particular

cases.

After pages of this, you expect the authors finally to tell you how their general theory?rational tions?can revolutions explain

choice better

on microfounda theory, based than SSR or the work that has

flowed from it. But just at the point in the essay where Kiser and Hechter

would

have

stead of offering

to deliver

any general,

the goods, the subject! In they change of revolutions, cover-law explanation they

make a brief allusion to thework ofMichael Taylor.

s essays, in search of the I ran off to the library to get Taylor general on microfoundations.31 I What discov theoretical based explanation ered were

lengthy

essays

containing

thoughtful

comments

on how

one

29

(fn. 25). 30Skool Kiser and Hechter,

in Comparative Historical "The Role of General Theory Sociology," American Journal of Sociology 97, no. 1 (1991). 31 and Revolution See Michael Press, University Taylor, ed., Rationality (Cambridge: Cambridge in the Explanation and Action of Social Change," Politics and 1988); and idem, "Structure, Culture, no. 2 (1989). Society 17,

WORLD POLITICS

42 reconstruct

might

in rational

peasant communities Russian cause,



choice

that part of SSR focusing

on

and how they rebelled during the French and

offers some convincing (be Taylor speculation not know what do those French and Russian really

Revolutions. of course, we

peasants were thinking back in 1789 and 1917) about how peasants overcome

could have

obstacles

and

much

institutions?very cussed in my analysis. Why,

to collective

action

ties through network the sorts of relationships that I had dis ra he asks, would self-seeking supposedly

tional individuals have engaged in the inherently risky business of state is very careful to say that he is authority? Taylor in no way contradicting in the and them. SSR, arguments reconstructing the in And he entire book about the indeed, my presumed analysis state breakdown of the peasant community rebellions power that made rebelling

against

In short, Taylor possible. a to it. complement of interaction This type

s

analysis

is partly

a

gloss

on mine

and partly

is not so differ scholarly argument ent from what I have seen in some other literatures, where there is a di choice theorists historical between rational institutionalists and alogue or So that is exactly where we are in comparative political sociologists. to the rational choice contributions the study of comparative revolu between

frankly, also in rational choice contributions other macroscopic interest that might processes commentaries individual-level intelligent linking

tions?and many been

conflicts and institutional processes. What macroscopic is to create successful theorists have not done, however, such as revolutions

ries of

phenomena authoritarian regimes. ery on this score. And

Hechter

or democratization

The and

final

challenge work subsequent

processes with rational choice theo general or the rise of

are all and no deliv promise that microlevel models will ever

and others

it seems unlikely historical-institutional

completely displace terns of conflict political

to the study of us. There have

and social

analyses

of macroscopic

pat

change.

to the in SSR embedded macroanalytic approach from the opposite direction. On the one

comes

who like Kiser and Hechter hand, we have the rational choice analysts a to return feel that comparative is simply historical work old-fashioned we have critics like the the other unscientific On hand, totally history. that comparative historian William historical Sewell, Jr., who believe a to is of all that is work total violation science, equivalent experimental to say to one another, sacred in historical work. What would they have

andwhy am I caught in the middle?! Prior towriting the critique I focus on here, Sewell had criticized SSR

SYMPOSIUM over

43

the question of the role of ideology and culture in the understand to criticize is now He revolutions.32 its my book revisiting

ing of

methodology.33 Arguing for what he labels "eventful" historical sociol advocates of particular ogy, Sewell interpretation and the tracing out of narratives of events, looking that are in some sense turning points in a narrative

historical

processes for events

especially sequence.

Outlining

this overall approach to historical social inquiry is SewelTs basic goal; in the process, he reviews major books by Charles Tilly, Immanuel Wallerstein, Sewell

and me.

I take to be a uncertainty. flattering that it is a powerful book, he seems perplexed Although about why that is so, since it violates his own theoretical and method SSR for criticizes the social sciences. Sewell ological prescriptions along com first He echoes lines. several, somewhat contradictory Burawoy's ment but he soon abandons this that SSR uses purely inductive methods, SSR is The thrust of attack is to that untenable his instead say position. approaches he argues

flawed

deeply

because

SSRwith

what

it uses

to test and comparisons develop likes the fact that SSR recounts narratives

tory arguments. He Revolutions and Chinese French, Russian, to tives be quite powerful. He nevertheless across those revolutions?and compare with

other

revolutions

instances of sociopolitical did not occur?violates what

and considers

explana about the

those

narra

to argues that my decision also to compare these cases conflict where successful social he calls the unbreakable

unity and the complexity of history. Sewell argues that all historical social sci entists who do comparative work are falsely applying experimental logic to sets the act of comparing takes particular necessarily history, because one to events out context them of of and juxtaposes another, thereby events across world historical of time. the flow breaking answer to Sewell s must "fracture" his charge that comparativists we so or anyone else who to must is it that he must?and yes, tory uses concepts use of the to talk about and language history. The very of is in invocation word the "French Revolution," "revolution," very use of a of The of flow and the history. complexity fracturing escapably My

such and concepts, including by Sewell himself, suggests that there similar 32 The

is something configurations

we want across

to bundle times

"Revolution" together. implies and places; and "French Revolution"

me about culture and ideology in revolutions is reprinted in exchange between Sewell and (fn. 25), pt. 3. Skool 33 an Eventful in Terrence J.Mc William Toward Sewell, Jr., "Three Temporalities: Sociology," ofMichigan Press, forth ed., The Historic Turn in theHuman Sciences (Ann Arbor: University Donald, coming).

44

WORLD

POLITICS

assembles happenings in French history that not allwould agree should be discussed

together. is hard to take seriously. Critics So the mere charge of "fracturing" must move of social science his from a logical charge to the substance

tory arguments; they need to explore which kinds of fracturing yield fruit and which ones do not. To make any kind of powerful critique, Sewell needs to do what he refuses to do in his methodological treatise; in SSRy as well as to engage the substantive analysis advanced the many other powerful substantive forth analyses put by compara tivists about revolutions and other macroscopic phenomena. must in comparative Those of us who do macroanalytic work politics

he needs

justify

our selections

of cases. We

and groupings

must

show

that we

are

dealing with sufficiently similar instances tomake it possible to explore about the phenomena arguments both similarities and differences

in question, and we must our range of cases. To

across

to

attend

the extent

that all of us who

history do pre practice macroanalytic comparative our critics must deal with the substance of our arguments cisely this, and not dismiss the operation incumbent upon logically. It is therefore to show

studies of comparative new to revolutions in the modern have failed world generate powerful about their causes, processes, and outcomes. Not only does arguments unaware Sewell fail to take on this task, but he also seems blithely of Sewell

that

scholars

who

have made

in the comparative-historical scholarship or revolutions other macroscopic phenomena. the best

sciences

social

about

Finally, I find it ironic that Sewell appreciates the narrative aspects of SSR yet would

discount

the book's

parisons.When

this book, I realize by the concomitant

just how

just spring in SSR that made

pened during the old regimes and China.

It was

forth

relate

to narrative

and other

social

science

But

narratives

are do

the comparative precisely analyses to write of what hap the narratives

into the revolutionized not had

ex

use for investiga of presen modes

readers. Narratives

all like stories.

naively. it possible for me the revolutionary breakdowns

If I had

not have

I originally talked about the methods

to historians tation that are convincing stories. We like Children compelling. not

I may

do comparative-historical we how the methods

do not explain adequately tion and for presentation of results

often

were

narratives

those

analyses.

comparative-causal of us who

in SSR.Those

informed

profoundly

plained this adequately when used

com

"scientific"

overly

allegedly

I think back on how I developed the narrative parts of

new

the comparative,

and the reconstructions regimes causal

in , analysis

Russia, to high

of

45

SYMPOSIUM

light the importance of the breakdown of istrative and military state organizations and to highlight the contributions of peasant revolt, Iwould not have been able to select from the mass of events the kinds of things that I ended up highlighting in the narratives that Sewell likes so much. have

And

set out

if I had

covered

to write narratives, I could never merely social revolutions and several other sociopo

three major

litical conflicts in one book. answer

So my second SSR, you cannot comparative

have

to Sewell them without

is to say, if you like the narratives of the comparative Good analysis. are not narratives go together; they

and powerful I think that the kind

analysis

narra of purely interpretive separate enterprises. an to tive is taken from he apart aspiration suggesting?if approach ward causal analysis, hypothesis the of achievement testing, and rigor of and cases?will lead comparative analysis through positive negative students and other scholars who practice "eventful" historical sociology into a hopeless

muddle

of descriptive

may further anthropologically will

not

further

revolutions

efforts

and other

to

explain

important

s.

SewelFs

prescription

rich descriptive work in history, but it the causes, macroscopic

and outcomes of process, in the transformations

modern world. Yet this latter effort iswhat I take to be the brief of comparative politics This exploration ments should

and comparative political sociology. of substantive research and theoretical

develop area of leads to conclusions that I comparative politics a us in For those of who do like to present fashion. summary

in one

comparative

politics,

it pays

to compare.

it is not necessary

And

to get

hung up on grand model building and purely deductive theorizing, on the one hand, or on producing rich narratives of particu interpretively is a middle way, comparative lar times and places, on the other. There in history and historical allows us to notice patterns analysis, which same we can better At theoretical ideas from them. the time, generate use to well-designed comparisons explore and better specify hypotheses as we launch into a derived from whatever theories we have available given

investigation.

This historical macroanalytic approach to comparative politics has flourished in quite a few substantive literatures beyond the one that I have

talked

macroanalysis

today. I feel certain that people into additional substantive areas,

about

Those

carry comparative an further developing about regimes, arguments

yet historically grounded in the modern and sociopolitical transformations in of us who do comparative social sciences the history

alytically powerful, social structures, world.

will

WORLD POLITICS

46 to

have nothing

can take counsel

for!We

apologize

from

and cooperate

with the model builders and the interpretivists. But we need not defer to either

group.

Atul

Conclusion: The

concerns

main

most

tion. While

Kohli

were symposium invited participants

of the of the

in the

outlined addressed

these

introduc concerns

directly, a few were more implicit, choosing to focus instead on their or implicit, explicit of leading comparative politics provide the role of theory in comparative politics. What if any, emerge? Given that the diverse messages research. Whether

ongoing

scholars

the responses of these a rich on of views array common conclusions,

of each of the sympo a in form, relatively abbreviated already presented to resummarize them. What instead is a brief follows

are

sium participants it is not necessary

areas of agreement themes that underline important a to the of scholars representing variety of approaches

discussion

of three

a group of study comparative The first strong

across

politics. conclusion?near

consensus, is very much

I may add?that a problem-driven research are politics

is that comparative emerges politics motivates the best comparative field of study.What of real-world puzzles significance: Under what conditions cies emerge

do revolutions

and stabilize? Why

do democra

occur? What

is the role

of different types of regimes in facilitating prosperity and equity?This problem

orientation

distinguishes

comparative

from

politics

other

so

cial science fields that tend to be driven primarily by theoretical and/or link between the ends. Of course, given the inevitable methodological are in relative. Even and these data, ory comparative politics, emphases of one "approach" over an the claims of superiority debates involving a tend not to endure, however. Given Such debates perennial. scholars tend strong interest in real-world politics puzzles, comparative as tools to to treat theories, and methods mainly approaches, help is best captured This above frame and explain empirical spirit puzzles. other

are

in Adam

of his

characterization

Przeworski's

whatever

"opportunist"?using cultural deconstruction?or

whether

works, in Peter Katzenstein

own

scholarly

among symposium eralizations. Given

participants important

concerns problems

or

game theory, history, s that he is suggestion or by important prob

variable" by the "dependent mainly lems requiring explanation. A second there was a fair amount theme on which

stimulated

as

style

of agreement of causal gen the significance com that require explanation,

47

SYMPOSIUM

to discern rela scholars pursue causal analyses, hoping politics across a cases. an old that hold variety of comparable Though tionships has in recent years been challenged research norm, this approach by those who hold that there are other, equally valid scholarly objectives

parative

for the study an cultures,

of "foreign

cultures

and

societies":

"thick description" of the "other," and critical

empathetic understanding of much-used Some categories. conceptual are into politics. Nevertheless, seeping comparative

construction bilities

of these

intellectual

search

claims as

interests

are notable:

James Scott disasters development

why

de

sensi

the positions

of the two symposium participants likely to be most these

of

sympathetic to his current

defines occur;

and

re

Susanne

Rudolph explicitly suggests that her interest in the "Foucaultian sug is "not about

gestion[s]"

suspending

comparison,

generalization,

and

explanation."

If the problem orientation of the field tends to relegate the role of to that of a tool of research, the quest for causal empirical moves to its the forefront. This healthy role generalizations, by contrast, inclines other with tension, comparative scholarly preferences, along a to aimed at scholars of intellectual pursue variety strategies politics theory mainly

The last concluding scholarship. theoretically significant generating an area of not so much theme thus concerns, among sympo agreement as the sium participants, variety of their preferences. to generate relevant scholarship may have The attempts theoretically more

a macro

of either

sions and utilizing oretical tendencies

Figure 1).Most follows

from

of interest

or a microfocus, to inductive.

of deductive

the dimension

infuse

they may

also vary along these dimen

Dichotomizing 2x2 matrix gives us four

the conventional that

and

comparative

politics

sets of the

scholarship

(see

scholars of comparative politics reside in cell 2. This

observations

to

noted above, namely, that problems are often macro in nature (for exam to state and that attempts performance)

already scholars

comparative revolutions, ple, democracy, lead schol of such phenomena discover or examine causal explanations cases. ars to one or more country study in such macro the role of theory Peter Evans captures eloquently frame and it gen inductive theory helps empirical puzzles, scholarship: one are erates And worth that causal hypotheses examining. plausible may

add that

inductive

Theda

search

new additional for regularities generates advocates similar, something strongly

also

Skool hypotheses. an she labels "macroanalytical." approach how through further still, demonstrating over time. cumulates

In her contribution such research

she goes ac knowledge

48 WORLD POLITICS Inductive

Deductive 1. structural

Macro

2. single or comparative case studies; political institutions and social groups as

fiinctionalism

Marxism

empirical Micro

4.

3. rational choice

local anthropology-inspired social political studies; structural or cultural foci

1 Figure of Theoretical

Dimensions

foci

Scholarship

The macroinductive genre of scholarship, though practiced widely by of comparative

scholars

is hardly without

politics,

detractors.

Among

the symposium participants, both James Scott and Susanne Rudolph their



of such research,

skepticism

their preference

for

instead reside in cell 4. Adam Przeworski also

scholarship that would warns

indicating

some of their standard methods may be bi comparativists ased and urges the use of counterfactuals and thus of a priori theorizing as an intellectual such problems. strategy to deal with that

Deductive, scholars who comparative

a

of course, be the refrain of would, theorizing priori within defined operate paradigms. Over the years strongly rise and decline has witnessed the of several such politics

intellectual movements. During the 1950s and the 1960s proponents of both structural fiinctionalism andMarxism (cell 1) competed for a par status. More

adigmatic

recently,

those who

adhere

to a rational

choice

approach (cell 3) have similarlymade claims for the superiority of their approach.

While

The

symposium

they all welcome

participants,

however,

are not

persuaded.

new hypotheses that emerge from deductive

claims: both Peter frameworks, they also resist their all-encoming of "aggregation" Evans and Peter Katzenstein about the problem worry in the rational choice the former that arguing aggregating approach, numerous is not the most "efficient" intellectual strategy microinsights a macro Adam for generating Przeworski, generalizations; practitioner treats it as that he of the rational choice approach, suggests implicitly one set of tools among others for both researching important problems;

Susanne Rudolph overly

abstract

or

and James Scott distrust frameworks that provide overly

homogenous

s

of different

cultures

and

49

SYMPOSIUM

people; and Theda choice s echoes

Skool views recent contributions

of rational

as

on comparative existing politics mainly building recent rather than offering novel insights. An important essay these and systematizes of criticisms the of rational choice many

theorists

to

too much, that, by claiming suggesting approach, same of has the many proach developed problems lier deductive intellectual movements.34

this deductive encountered

by

ap ear

sum up, this was to take stock of the role of symposium designed in and of theoretical controversies theory comparative politics. A vari were and surprising conclusion ety of views important expressed. An are controversies however?that theoretical contemporary emerged, To

not deeply divisive. In spite of the challenges thrown out by both ratio a diverse group of sym and postmodern cultural approaches, adhered to a loosely defined "core," or to what one posium participants as the "eclectic center" of characterized comparative participant politics. a The minimal of this "core" includes orientation definition problem nal choice

and a commitment theoretical

that, a variety generalizations. Beyond even is considered desirable. Nevertheless,

to causal

of on

approaches frames of and deductive score, while microinductive approaches most reference provide important complements, comparative politics center pursue in this eclectic scholars relevant, macro theoretically on one or more countries, analyses, focusing through diverse empirical a or his and of lenses, data, contemporary variety utilizing conceptual or torical, quantitative qualitative. this

34 and

See Neil J. Smelser, "The Rational 4 (October 1992).

Society

Choice

Perspective:

A Theoretical

Assessment,"

Rationality