Manifesto for a Relational Sociology1
Mustafa Emirbayer
New School for Social Research
Sociologists today are faced with a fundamental dilemma: whether
to conceive of the social world as consisting primarily in substances
or processes, in static "things" or in dynamic, unfolding relations.
Rational-actor and norm-based models, diverse holisms and structuralism~,and
statistical "variable" analyses continue implicitly or
explicitly to prefer the former point of view. By contrast, this "manifesto"
presents an alternative, "relational" perspective, first in broad,
philosophical outlines, then by exploring its implications for both
theory and empirical research. In the closing pages, it ponders some
of the difficulties and challenges now facing relational analysis, taking
up in turn the issues of boundaries and entities, network dynamics,
causality, and normative implications.
Sociologists today are faced with a fundamental dilemma: whether to conceive
of the social world as consisting primarily in substances or in processes,
in static "things" or in dynamic, unfolding relations. Large segments
of the sociological community continue implicitly or explicitly to
prefer the former point of view. Rational-actor and norm-based models,
diverse holisms and structuralisms, and statistical "variable" analysesall
of them beholden to the idea that it is entities that come first and relations
among them only subsequently-hold sway throughout much of the
discipline. But increasingly, researchers are searching for viable analytic
alternatives, approaches that reverse these basic assumptions and depict
social reality instead in dynamic, continuous, and processual terms.
The purpose of this "manifesto" is to lay out the essential features of
this latter point of view. It begins by presenting the relational perspective
in broad outlines (by way of a comparison with competing substantialist
'I would like to thank Ronald Breiger, David Gibson,Jeff Goodwin,Michkle Lamont,
Ann Mische, John Mohr, Jeffrey Olick, Shepley Orr, Mimi Sheller, Charles Tilly, and
Harrison White for their many helpful comments on earlier drafts. I would also like
to thank participants in seminars at Princeton University and at the New School for
Social Research, and in the Monthly Discussion Group on Theory and Culture in New
York City, for their stimulating criticisms and suggestions. Direct correspondence to
Mustafa Emirbayer, Department of Sociology, New School for Social Research, 65
Fifth Avenue, New York, New York 10003.
O 1997 by The University of Chicago. All rights reserved.
0002-9602/98/10302-0001$02.50
AJS Volume 103 Number 2 (September 1997):281-317 281
American Journal of Sociology
approaches), then explores its implications for both theory and empirical
research. In the closing pages, it also ponders some of the difficulties and
challenges now facing relational analysis. (The essay focuses throughout
upon ontology, largely-but certainly not exclusively-bracketing associated
questions regarding epistemology.) I do not claim to have been the
first to elaborate any of these themes; on the contrary, many distinguished
sociologists, from a wide range of theoretical and empirical perspectives,
have long been working within a relational framework of analysis. Indeed,
this framework itself has long been propounded by social thinkers and
philosophers, arguably going as far back as the pre-Socratics. Nor do I
claim to be exhaustive in either my textual discussions or bibliographic
references. "You can be assured that, for each idea, quite a number of
substantial, and often independent, discussions and implementations
could be cited: Ideas that have any importance, any impact, do, after all,
come in company, not as isolates, and the essayist is mostly a transcriber
of ideas abroad in his networks" (White 1994a, p. 4). What I have done
here is merely to bring together the various lines of reasoning in this perspective
(philosophical, theoretical, and empirical); to clarify how they
present an overarching challenge to reigning assumptions; and to seek
thereby to prevent the sort of eclecticism, the easy mixing together of substantialist
and relational assumptions, that renders even many innovative
studies today partially problemati~.~The key question confronting sociologists
in the present day is not "material versus ideal," "structure versus
agency," "individual versus society," or any of the other dualisms so often
noted; rather, it is the choice between substantialism and relationalism.
SUBSTANTIALIST AND RELATIONAL THINKING
The relational point of view on social action and historical change can
most usefully be characterized by comparing it with its opposite, the substantialist
perspective. The latter takes as its point of departure the notion
that it is substances of various kinds (things,beings, essences) that constitute
the fundamental units of all inquiry. Systematic analysis is to begin
Given these purposes (together with severe limitations of space),it has been necessary
to suppress two other sorts of analysis that would be integral to any fuller scholarly
treatment: careful,judicious consideration of the substantialistalternatives (of promising
developments in substantialist lines of theorization, enduring strengths of these
approaches, and so on)and discussion of internal differences among relational thinkers
themselves: the issues, in other words, that divide them as well as bring them together.
In other writings (Emirbayer and Mische 1995; Emirbayer 1996; Emirbayer and
Sheller 1996; Emirbayer and Goodwin 1996, 1997),I attempt to develop one specific
mode of relational theorizing, which I term relational pragmatics, and to argue for
its strengths vis-a-vis alternative relational as well as substantialist approaches.
Manifesto
with these self-subsistent entities, which come "preformed," and only then
to consider the dynamic flows in which they subsequently involve themselves.
"Relation is not independent of the concept of real being; it can
only add supplementary and external modifications to the latter, such as
do not affect its real 'nature' " (Cassirer 1953, p. 8). One social theorist,
Norbert Elias, points 'out that substantialist thinking corresponds closely
to grammatical patterns deeply ingrained in Western languages. An extended
quotation regarding these modes of speech and thought serves
nicely to introduce this perspective in general terms:
Our languages are constructed in such a way that we can often only express
constant movement or constant change in ways which imply that it has the
character of an isolated object at rest, and then, almost as an afterthought,
adding a verb which expresses the fact that the thing with this character
is now changing. For example, standing by a river we see the perpetual
flowing of the water. But to grasp it conceptually, and to communicate it
to others, we do not think and say, "Look at the perpetual flowing of the
water"; we say, "Look how fast the river is flowing." We say, "The wind is
blowing," as if the wind were actually a thing at rest which, at a given point
in time, begins to move and blow. We speak as if a wind could exist which
did not blow. This reduction of processes to static conditions, which we
shall call "process-reduction" for short, appears self-explanatory to people
who have grown up with such languages. (Elias 1978, pp. 111-12)
In a little-known but important discussion, John Dewey and Arthur F.
Bentley (1949) distinguish between two varieties of substantialist approaches.
The first they term the perspective of self-action; it conceives
of "things . . . as acting under their own powers" (Dewey and Bentley
1949,p. 108),independently of all other substances. The relational matrices
within which substances act provide, in this view, no more than empty
media for their self-generating, self-moving activity. Dewey and Bentley
see such a perspective as most characteristic of ancient and medieval philosophy.
"Aristotle's physics was a great achievement in its time, but it
was built around 'substances.' Down to Galileo men of learning almost
universally held, following Aristotle, that there exist things which completely,
inherently, and hence necessarily, possess Being; that these continue
eternally in action (movement) under their own power-continue,
indeed, in some particular action essential to them in which they are engaged"
(Dewey and Bentley 1949, p. 1 The Christian doctrine of the
"soul," culminating in the theological system of Thomas Aquinas, epitoOne
commentator provides further explication: "In Plato's concept of the soul, the
capacity for self-motion or self-action is the essential feature of the soul, and Aristotle
tells us quite explicitly in his Physics that of those things that exist by nature . . .
'each of them has within itselfa principle of motion' " (Bernstein 1966, p. 81).
American Journal of Sociology
mized this ~iewpoint.~But after Galilee's time, advances in physics and
in the natural sciences eliminated most such traces of self-action from the
study of inorganic matter.
In modern philosophy, however, the notion of self-action lives on in
various doctrines of "the will" and in liberal political theory (sinceHobbes,
Locke, and Kant), while in the social sciences it retains surprising vigor
in the form of methodological individualism. "All the spooks, fairies, essences,
and entities that once had inhabited portions of matter now [take]
flight to new homes, mostly in or at the human body. . . . The 'mind' as
'actor,' still in use in present-day psychologies and sociologies, is the old
self-acting 'soul' with its immortality stripped off" (Dewey and Bentley
1949, p. 131). One increasingly prevalent approach begins with rational,
calculating actors but assumes the givenness and fixity of their various
interests, goals, and "preference schedules." Rational-choice theory takes
for granted, as Jon Elster puts it, that "the elementary unit of social life
is the individual human action. To explain social institutions and social
change is to show how they arise as the result of the action and interaction
of individuals" (Elster 1989, p. 13). When actors become involved with
other actors whose choices condition their own, yielding results unintended
by any one of them taken alone, rational-choice theorists have
recourse to game theory. Here again (in at least some of its versions), pregiven
entities are seen to generate self-action; even as actors engage in
game playing with other actors, their underlying interests, identities, and
other characteristics remain unaltered. Game theory assumes that "there
are two or more players. Each of them has the choice between two or
more strategies. Each set of choices generates a set of rewards. The reward
of each player depends on the choices made by all others, not only on his
own decision. The players are assumed to make their choices independently
of each other, in the sense that they cannot make binding agreements
to coordinate their decisions" (Elster 1989, p. 28). With its analytic
sophistication and rigor, the rational-actor approach (together with complementary
versions of game theory) is fast becoming the major alternative
to the relational approach that I shall outline below.
Another popular approach, only apparently the chief rival to rationalchoice
models, takes norm-following individuals, or more specifically, the
vital inner forces driving them, as its basic unit of analysis. It depicts
individuals as self-propelling, self-subsistent entities that pursue internalized
norms given in advance and fixed for the duration of the action sequence
under investigation. Such individuals aspire not to wealth, status,
or power, but rather, to action in conformity with 'the social ideals they
For Aquinas, God Himself was the most perfect self-acting substance.
Manifesto
have accepted as their own. Nonrational action thus becomes the special
province of this mode of analysis, long a staple of sociological inquiry. To
mark itself off from economics, which endorsed the rational-actor approach
early on, sociology had from its beginnings "a fundamental need
of a theory of action that defined different types of action on the basis of
their specific difference from rational action. It required a theory of society
as a complex of interrelated actions that was more than the unintended
interconnecting of self-interested actions. . . . As a safeguard against the
utilitarian dangers of the theory of rational action, the founding theorists
of sociology [had to] have recourse to Kant and his notion of free, moral
action" (Joas 1993, pp. 246-47). To this day, the norm-following, neoKantian
perspective lives on in various forms of critical theory, value
analysis, and microsociology.
In a very different way, the idea of self-action also insinuates itself into
social thought by means of holistic theories and "structuralisms" that
posit not individuals but self-subsistent "societies," "structures," or "social
systems" as the exclusive sources of action. Proponents of these approaches,
from neofunctionalists and systems theorists to many historicalcomparative
analysts, all too often fall back upon the assumption that it
is durable, coherent entities that constitute the legitimate starting points
of all sociological inquiry. Such entities possess emergent properties not
reducible to the discrete elements of which they consist. Not individual
persons, but groups, nations, cultures, and other reified substances do all
of the acting in social life and account for its dynamism. In some cases,
even sequences of action may discharge such a function: social movements
or nationalist struggles, for example, are seen as propelling themselves
along trajectories "that repeat . . . time after time in essentially the same
form" (Tilly 1995a, p. 1596). Processes as well as structures thus appear
as self-acting entities in many concrete instances of social inquiry.
The second key category of substantialism that Dewey and Bentley consider
is that of inter-action.' This approach, which is frequently confused
with more truly relational points of view, posits "thing [as] balanced
against thing in causal interconnection" (Dewey and Bentley 1949,p. 108),
where entities no longer generate their own action, but rather, the relevant
action takes place among the entities themselves. Entities remain fixed
and unchanging throughout such interaction, each independent of the ex'My
use of hyphenation here follows Dewey and Bentley's own practice and underscores
that I am employing their terminology differently from how such words as
"interaction" and, as we shall see, "transaction" are used in everyday language. (Often,
e.g., interaction and transaction are employed as synonyms, whereas here they represent
very specific-and distinct-philosophical positions.) This point should be borne
carefully in mind even when these words reappear without hyphenation later in the
essay.
American Journal of Sociology
istence of the others, much like billiard balls or the particles in Newtonian
mechanics. Indeed, it was Sir Isaac Newton who actually gave the interactional
perspective its consummate expression. "For many generations, beginning
with Galileo after his break with Aristotelian tradition, and continuing
until past the days of Comte, the stress in physical inquiry lay upon
locating units or elements of action, and determining their interactions.
Newton firmly established the system under which particles could be chosen
and arrayed for inquiry with respect to motion, and so brought under
definite report. . . . The inter-actional presentation had now been perfected"
(Dewey and Bentley 1949, pp. 105, 111).
The idea of interaction finds its home today in a viewpoint that explicitly
or implicitly dominates much of contemporary sociology, from survey
research to historical-comparative analysis. This is the so-called "variablecentered
approach," which features, as Andrew Abbott (1988, p. 170)
points out, a compelling imagery of fixed entities with variable attributes
that "interact, in causal or actual time, to create outcomes, themselves
measurable as attributes of the fixed en ti tie^."^ What decidedly do not do
the acting in this perspective are the substances in question; all of the
relevant action takes place among them-they provide merely the empty
settings within which causation occurs-rather than being generated by
them. If anything, it is the variable attributes themselves that "act," that
supply initiative, in interactional research; "disadvantaged position leads
to increased competitiveness," for example, without any particular actor
engaging in competitive behavior. "It is when a variable 'does something'
narratively that [analysts] think themselves to be speaking most directly
of causality," notes Abbott. "The realist metaphysics implicit in treating
variables (universals)as agents was last taken seriously in the age of Aquinas
. . .but in this [approach] the 'best' causal sentences are clearly realist
ones in which variables act" (Abbott 1992a, p. 58). Variable-centered researchers
employ a variety of quantitative methods to test their causal
hypotheses, including multiple regression, factor analysis, and event history
approaches.
Fundamentally opposed to both varieties of substantialism is the perspective
of trans-action, "where systems of description and naming are
employed to deal with aspects and phases of action, without final attribution
to 'elements' or other presumptively detachable or independent 'entities,'
'essences,' or 'realities,' and without isolation of presumptively
detachable 'relations' from such detachable 'elements' " (Dewey and
In Abbott's (1988,p. 181)portrayal, the interactional approach further assumes "that
these attributes have only one causal meaning at a time; [and]that this causal meaning
does not depend on other attributes, on the past sequence of attributes, or on the
context of other entities."
Manifesto
Bentley 1949,p. 108).In this point of view, which I shall also label "relational,"
the very terms or units involved in a transaction derive their
meaning, significance, and identity from the (changing) functional roles
they play within that transaction. The latter, seen as a dynamic, unfolding
process, becomes the primary unit of analysis rather than the constituent
elements themselves. 'Things "are not assumed as independent existences
present anterior to any relation, but . . . gain their whole being . . . first
in and with the relations which are predicated of them. Such 'things' are
terms of relations, and as such can never be 'given' in isolation but only
in ideal community with each other" (Cassirer 1953, p. 36). Although it
can be traced back to the writings of Heraclitus, this transactional mode
of theorizing first becomes widely influential with the rise of new approaches
in physics, mathematics, and the natural sciences. Most strikingly,
Einstein's theory of relativity "brought space and time into the investigation
as among the events investigated [and] prepared the scene for
the particle itself to go the way of space and time. These steps were all
definitely in the line of the transactional approach: the seeing together,
when research requires it, of what before had been seen in separations
and held severally apart" (Dewey and Bentley 1949, p. lli?).'
Relational theorists reject the notion that one can posit discrete, pregiven
units such as the individual or society as ultimate starting points
of sociological analysis (asin the self-actional perspective). Individual persons,
whether strategic or norm following, are inseparable from the transactional
contexts within which they are embedded; as Michel Foucault
puts it, the "soul is not a substance; it is the element in which are articulated
the effects of a certain type of power and the reference of a certain
type of knowledge" (Foucault 1979, p. 29).' By the same token, structures
In his masterful history of modern science, Substance and Function (1953), Ernst
Cassirer distinguishes between the "relation-concepts" pertinent to this transactional
approach and the "thing-concepts" central to substantialist ways of thinking since
Plato and Aristotle. He charts the rise of relational theorizing in a multiplicity of
problem-areas, including the theories of space and number, geometry, and the natural
sciences.Today the most sustained and philosophically subtle exploration of the transactional
perspective can be found in the writings of Margaret Somers (1993, 1994,
1995; see also Somers and Gibson 1994).
Relational thinkers are relentless in their critiques of such concepts as "the soul" or
"person." Consider, e.g., Harrison White (1992, p. 197, n.21): "Person should be a
construct from the middle of the analysis,not a given boundary condition. Personhood
has to be accounted for. . . . But in most present social science 'person' is instead taken
as the unquestioned atom. This is an unacknowledged borrowing and transcription of
the soul construct from Christian theology. . . .The ultimate fixity of the soul, carried
over to hobble social science, was a Pauline theological imperative." Or consider Pierre
Bourdieu (Bourdieu and Wacquant 1992, pp. 106-7): "When I talk of [any given]
field, I know very well that in this field I will find 'particles' (let me pretend for a
moment that we are dealing with a physical field) that are under the sway of forces
American Journal of Sociology
are empty abstractions apart from the several elements of which they are
composed; societies themselves are nothing but pluralities of associated
individuals. Transactional theorists-starting, in fact, with some of the
"founders" of sociology-are in virtually complete agreement on this
point. Karl Marx (1978, p. 247) argues, for example, that "society does not
consist of individuals, but expresses the sum of interrelations, the relations
within which these individuals stand." Near the end of Capital, volume
1, he further observes that "capital is not a thing, but a social relation
between persons which is mediated through things" (Marx 1977, p. 932).
Georg Simmel, the classical sociologist most deeply committed to relational
theorizing, notes that "society is the supra-singular structure which
is nonetheless not abstract. Through this concept, historical life is spared
the alternatives of having to run either in mere individuals or abstract
generalities. Society is the generality that has, simultaneously, concrete
vitality" (Simmel 1971, p. 69). And even mile Durkheim, the "founder"
most identified with substantialist ideas, acknowledges that "the force of
the collectivity is not wholly external. . . .Society can exist only in and by
means of individual minds" (Durkheim 1995,p. 2 11;see Emirbayer 1996).9
Variable-based analysis (as in the interactional perspective) is an
equally unviable alternative; it, too, detaches elements (substances with
variable attributes) from their spatiotemporal contexts, analyzing them
apart from their relations with other elements within fields of mutual determination
and flux. As Margaret Somers and Gloria Gibson (1994, pp.
65,69)point out,"While a social identity or categorical approach presumes
internally stable concepts, such that under normal conditions entities
within that category will act predictably, the [relational, transactional]
approach embeds the actor within relationships and stories that shift over
time and space and thus precludes categorical stability in action. . . .The
classification of an actor divorced from analytic relationality is neither
ontologically intelligible nor meaningful" (seealso Somers 1994;Bates and
of attraction, of repulsion, and so on, as in a magnetic field. Having said this, as soon
as I speak of a field, my attention fastens on the primacy of this system of objective
relations over the particles themselves. And we could say, following the formula of
a famous German physicist, that the individual, like the electron, is an Ausgeburt des
Felds: he or she is in a sense an emanation of the field."
In the present day, Niklas Luhmann (1995, pp. 20, 22) suggests in similar fashion:
"There are no elements without relational connections or relations without elements.
. . . Elements are elements only for the system that employs them as units and they
are such only through this system." And Harrison White (1992, p. 4) turns to material
science to make a similar point, likening sociocultural reality to dense, partly fluid
matter within which no self-subsistent entities ever fully crystallize: "There is no tidy
atom and no embracing world, only complex striations, long strings reptating as in
a polymer goo."
Manifesto
Peacock 1989; Breiger 1991; for a summary of this line of critique, see
Wellman [1988], pp. 31-33). A corollary to this view calls into question
attempts by statistical researchers to "control for third variables"; all such
attempts, too, ignore the ontological embeddedness or locatedness of entities
within actual situational contexts.1°Even as statistical models grow
ever more complicated, even alert to the "interaction effects" among variables,
these problems, rooted as they are in fundamental assumptions, fail
to go away.
What is distinct about the transactional approach is that it sees relations
between terms or units as preeminently dynamic in nature, as unfolding,
ongoing processes rather than as static ties among inert substances."Previously
constituted actors enter [transactions]but have no ability to traverse
[them] inviolable. They ford [them] with difficulty and in [them] many
disappear. What comes out are new actors, new entities, new relations
among old parts" (Abbott 1996, p. 863).11The imageries most often employed
in speaking of transactions are accordingly those of complex joint
activity, in which it makes no sense to envision constituent elements apart
from the flows within which they are involved (and vice versa). Dewey
(1929, p. 142),for example, states in Experience and Nature that "the import
of. ..essences is the consequence of social interactions, of companionship,
mutual assistance, direction and concerted action in fighting, festivity,
and work." Dewey and Bentley (1949, p. 133) further point out that
"no one would be able successfully to speak of the hunter and the hunted
as isolated with respect to hunting. Yet it is just as absurd to set up hunting
as an event in isolation from the spatio-temporal connection of all
the components." In Social Organization, Charles Horton Cooley (1962)
supplies a vivid analogy to "joint music-making" to convey similar insights,''
while Norbert Elias (1978, p. 130)invokes "game playing" to explicate
his key concept of fluid "figurations": "By figuration we mean the
lo "Nothing that ever occurs in the social world occurs 'net of other variables.' All
social facts are located in contexts. Sowhy bother to pretend that they aren't?"(Abbott
1992b, p. 6).
"Actually, social actors, the parties to ongoing transactions, can just as well be communities,
firms, or states as they can individual persons (although the latter is what
one typically has in mind). Indeed, they can even be events: any actor, after all, is
but a series of "events that keep happening in the same way," events with "stable
lineages" (Abbott 1996, pp. 873, 863).
""Whether, like the orchestra, [social organization] gives forth harmony may be a
matter of dispute, but that its sound, pleasing or otherwise, is the expression of a vital
cooperation, cannot well be denied. Certainly everything that I say or think is influenced
by what others have said or thought, and, in one way or another, sends out an
influence of its own in turn" (Cooley 1962, p. 4).
American Journal of Sociology
changing pattern created by the players as a whole . .. the totality of their
dealings in their relationships with each other."I3One might just as well
speak here of negotiations or conversations; the underlying idea would
remain the same-regarding the primacy of contextuality and process in
sociological analysis.14
To be sure, the transactional viewpoint-like its two main substantialist
rivals-rarely corresponds with exact precision to any one school
of thought or individual's life work; what is often of great interest, in fact,
is exactly how these ideal-typical approaches, on the one hand, and actual
authors, texts, or research traditions, on the other, crisscross one another:
how, for example, a given thinker shifts back and forth (often implicitly)
among several different points of view." This is evident in the case of
both classical and contemporary thinkers. Marx, for instance (as the earlier
quotations from him suggest), was a profoundly relational thinker;
this is clear from his early analyses of alienation (Ollman 197I), his discussion
of commodity fetishism, his keen insights into the internal relations
among production, distribution, exchange, and consumption, and, indeed,
his understanding of the capitallwage-labor relation itself.16And yet, even
Marx exhibits substantialist tendencies, most notably in his reification of
class interests, in his assumption that actors within the same class category
(to the extent that they are a "class-for-itself") will act in similar ways
even when differentially situated within flows of transactions or "relational
settings" (Somers 1994). Meanwhile, from the opposite direction,
the now classic theorist Talcott Parsons exhibits tendencies seemingly in
line with substantialism. His theory of action strongly tends in the direction
of self-actional, norm-based reasoning, while his ideas about order
often suggest holism and the reification of system goals. Yet, even Parsons,
l3 Zygmunt Bauman (1989)underscores the "remarkable affinity" that this imagery of
gamelike figurations bears with Anthony Giddens's "structuration theory." "Gameplaying"
clearly means something quite different here, of course, from the self-actions
of game theory (although see the paragraph that follows).
l4 The preeminently agentic nature of the above-mentioned analogies, their emphasis
upon transformation as well as iteration, might be noted; this opens up questions, too
seldom addressed, regarding the normative dimensions of social action. I shall take
up such issues below.
'j Perhaps the theorist who most firmly grasped this distinction between the analytical
and the concrete was Talcott Parsons (1937)."At no point . . . [did] Parsons claim
that theories which actually existed in history are identical to the different logical
positions he has outlined. Most of them are in fact opaque versions or combinations
'
of several of these logical alternatives" (Joas 1996, p. 12). 

l6 "Within a relational view [such as Marx's], the working class is defined by its quali- 

tative location within a social relation that simultaneously defines the capitalist class. 

. . . Classes as social forces are real consequences of social relations" (Wright 1979, 

p p 6-7). 

Manifesto
especially after the development of his later '(interchangemodel" and theory
of "generalized media" (Parsons 1953, 1969),moves decisively in the
direction of a relational, transactional point of view.17And finally, among
contemporary theorists and research traditions, game theory is emblematic
of rational-actor (i.e., self-actional) models, as mentioned above. And
yet it, also, often sounds emphatically transactional themes, as in studies
of repeated games within which temporally embedded actors engage with
each other in sequences of mutually contingent action (see Kreps 1990;
Macy 1991).
Despite these various concrete examples, however, the point still holds
that substantialism (in both its self- and interactional forms) and relationalism
(or transactionalism) represent fundamentally different points
of view on the very nature and constitution of social reality.
THEORETICAL IMPLICATIONS
The implications of the transactional approach are far-reaching. In what
follows, I shall specify these implications from two different points of view
in turn: those of key sociological concepts and of levels of inquiry-from
"macro" to "micro."
To begin with, the central concepts in sociological analysis-for example,
power, equality, freedom, and agency (to mention several of the most
pervasive)-are themselves open to extensive reformulation in terms of
relational thinking. Take the key concept ofpower, which is typically seen
in substantialist terms as an entity or a possession, as something to be
"seized" or "held."18In the transactional approach, "the concept of power
[is] transformed from a concept of substance to a concept of relationship.
At the core of changing figurations-indeed, the very hub of the figuration
process-is a fluctuating, tensile equilibrium, a balance of power moving
l7 ASJeffrey Alexander points out, it becomes quite illuminating to see how such discordant
trends in the classical theorists-e.g., Marx and Parsons-later manifest
themselves in their followers' writings: "The members of a sociological school change
the founder's thought as much as they faithfully articulate it, and . . . they change
it, moreover, in a manner that can be systematically related to the analytic tensions
in the original theoretical position" (Alexander 1983, p. 277).
This conceptualization is still popular to the extent that Weberian understandings
of power retain (explicitly or implicitly) their influence in political sociology. "Max
Weber proceeds from the teleological [i.e., self-actional] model of action in which an
individual or a group has a set purpose and chooses the means suitable for realizing
it. The success of the action consists in bringing about a condition in the world that
fulfills the purpose set. To the extent that this success depends on the conduct of
another subject, the agent has to dispose of the means that instigate the desired conduct
on the part of the other. This manipulative power over the means that afford
influence on the will of another Max Weber names power" (Habermas 1983, p. 173).
American Journal of Sociology
to and fro. . . . This kind of fluctuating balance of power is a structural
characteristic of the flow of every figuration" (Elias 1978,p. 131).Contemporary
social-network analysts define power in similarly relational terms,
as an outgrowth of the positions that social actors occupy in one or more
networks (Knoke 1990).So too do theorists such as Michel Foucault and
Pierre Bourdieu. "Relations of power," in Foucault's words, "are not in a
position of exteriority with respect to other types of relationships (economic
processes, knowledge relationships, sexual relations), but are immanent
in the latter; they are the immediate effects of the divisions, inequalities,
and disequilibriums which occur in the latter, and conversely they
are the internal conditions of these differentiations" (Foucault 1990,p. 94).
Bourdieu similarly argues for a relational view: "By field of power I mean
the relations of force that obtain between the social positions which guarantee
their occupants a quantum of social force, or of capital, such that
they are able to enter into the struggles over the monopoly of power"
(Bourdieu and Wacquant 1992, pp. 220-30). Far from being an attribute
or property of actors, then, power is unthinkable outside matrices of force
relations; it emerges out of the very way in which figurations of relationships-as
we shall see, of a cultural, social structural, and social psychological
nature-are patterned and operate.
The idea of equality can also be recast in transactional terms. Typically,
equality (like inequality)is defined essentialistically as a matter of individual
variations in the possession of "human capital" or other goods.'' The
primary causes of equality (and inequality), moreover, are located in the
orientations and actions of entities such as groups or individuals, rather
than in the unfolding relations among them: in attitudes such as racism,
sexism, and ethnic chauvinism. Yet, in Charles Tilly's (1995b, p. 48)
words, "bonds, not essences, provide the bases of durable inequality." Inequality
comes largely from the solutions that elite and nonelite actors
improvise in the face of recurrent organizational problems-challenges
centering around control over symbolic, positional, or emotional resources.
These solutions, which involve the implementation of invidious
categorical distinctions, resemble "moves" in a game, or perhaps even attempts
to change the rules of the game. Members of a categorically
bounded network, for example (such as recently arrived immigrants), acl9
"Faced with male-female differences in wages, investigators look for average human-capital
differences among the individuals involved. Encountering racial differences
in job assignments, researchers ask whether across categories individuals distribute
differently with respect to residential location. Uncovering evidence of sharp
ethnic differences in industrial concentration, analysts only begin to speak of discrimination
when they have factored out individual differences in education, work experience,
or productivity" (Tilly 19953, p. 9).
Manifesto
quire control over a valuable resource (e.g., information about employment
opportunities), hoard their access to it (e.g., by sharing it only with
others in their personal networks), and develop practices that perpetuate
this restricted access (e.g., by staying in touch with their places of origin
through frequent correspondence and visits home). Hard, durable differences
in advantages and disadvantages then crystallize around such practices.
Unfolding transactions, and not preconstituted attributes, are thus
what most effectively explain equality and inequality.
Third, the notion offreedom is also open to far-reaching reformulation
in relational terms. A very common way of thinking about freedom (or
liberty) is in substantialist fashion, as a possession, a "legal status represented
in laws" (Stinchcombe 1995, p. 126), and as contrasted with the
equally essentialist status of slavery. But, as Arthur Stinchcombe (1995,
p. 126)indicates, freedom is best described pragmatically as a set of "liberties
. . . in fact enjoyred], whether or not they are defended in the law."
Drawing upon John R. Commons (1924, p. 19),who cautioned that ideas
such as liberty refer not to the "thing itself but [to] the expected 'uses' of
the thing, that is, [to] various activities regarding the thing," Stinchcombe
(1995,p. 126)regards freedom not as a fixed, pregiven attribute, but rather
as what we can do under given circumstances: "A liberty creates an exposure
of others to the different consequences of different choices by the free
person. . . . The definition, then, is a sum of practically available liberties,
including in particular the social capacity to get others to suffer the consequences
of [one's] practical . . . freedom . . . to decide." Freedom, in other
words, means nothing apart from the concrete transactions in which individuals
engage, within cultural, social structural, and social psychological
contexts of action; it derives its significance entirely from the ongoing interplay
(akin to a game) of decision, consequence, and reaction. "I . . .
argue," asserts Stinchcombe, "that this tack toward understanding variations
in freedom helps get us out of the box of defining freedom, or slavery,
by its essence. Defining things by their essences is always troublesome in
an explanatory science" (Stinchcombe 1995, p. 130).'O
Finally, and relatedly, the idea of agency can also be reconceptualized
20 A similar conception of freedom can be found in Simmel's work, especially in The
Philosophy of Money (1990) and Soziologie (1950):"[The category of freedom] implies
not a mere absence of relationships but rather a very specific relation to others" (Simme1
1990,p. 298; quoted in Breiger 1990,p. 457)."Freedom itself is a specific relation
to the environment. . . . It is neither a state that exists always and can be taken for
granted, nor a possession of a material substance, so to speak, that has been acquired
once and for all. . . . [It] is not solipsistic existence but sociological action. . . . This
is suggested by the simple recognition of the fact that man does not only want to be
free, but wants to use his freedom for some purpose" (Simmel 1950,pp. 120-21.) For
a discussion of Simmel on freedom, see Breiger (1990).
American Journal of Sociology
from a transactional perspective. (Of course, this exercise in reconceptualization
could just as well be extended to other key terms in the sociological
lexicon.) Agency is commonly identified with the self-actional notion of
"human will," as a property or vital principle that "breathes life" into passive,
inert substances (individuals or groups) that otherwise would remain
perpetually at rest. By contrast, the relational point of view sees agency
as inseparable from the unfolding dynamics of situations, especially from
the problematic features of those situations. As it is conceived of elsewhere
(Emirbayer and Mische 1998),agency entails the "engagement by actors of
different structural environments [which] both reproduces and transforms
those structures in interactive response to the problems posed by changing
historical ~ituations."~~Viewed internally, agency involves different ways
of experiencing the world, although even here, just as consciousness is
always consciousness of something (James 1976; Husserl 1960), so too is
agency always "agency toward something," by means of which actors can
enter into relationship with surrounding persons, places, meanings, and
events. Viewed externally, it entails concrete transactions within relational
contexts (cultural, social structural, and social psychological) in
something much like an ongoing conversation. Agency is always a dialogic
process by which actors immersed in the dure'e of lived experience engage
with others in collectively organized action contexts, temporal as well as
spatial. Agency is path dependent as well as situationally embedded; it
signifies modes of response to problems impinging upon it through sometimes
broad expanses of time as well as space.
In addition to these general concepts, the transactional point of view
allows for the reconceptualization of distinct, sui generis levels of inquiry,
on a continuum from "macro" to "micro."22At the most macroscopic level,
for example, society is often interpreted as an autonomous, internally organized,
self-sustaining "system." Sociological thinkers often assume that
inquiry ought to begin with such naturally bounded, integrated, sovereign
entities as national states or countries. And indeed, this approach is not
entirely implausible, for not only has the European state system been ascendant
for two or more centuries, but also, throughout much of world
history, disparate interaction networks-economic, military, political, and
''Such a conception corresponds closely to classical pragmatist thought: "The subject
is that which suffers, is subjected and which endures resistance and frustration; it is
also that which attempts subjection of hostile conditions; that which takes the immediate
initiative in remaking the situation as it stands" (Dewey, quoted in Colapietro
1990,p. 653).I shall discuss this theoretical connection to pragmatism in greater detail
below.
''For a discussion of analytic levels similar (but not identical) to that which follows,
see Wiley's [1995,chap. 61 fourfold distinction among self, interactional, socialorganizational,
and cultural levels.
Manifesto
the like-have frequently converged (or been "institutionalized" [Mann
19861)in coherent, distinct societies. Yet, boundaries of national states do
overlap unevenly with populations, territories, production and consumption
patterns, cultural identities, collective emotional commitments, and
so on, while "interstitial interactions," both within and across these
bounded units, also repeatedly belie visions of the latter as unproblematic,
unitary entities: "Societies have never been sufficiently institutionalized
to prevent interstitial emergence. Human beings do not create unitary
societies but a diversity of intersecting networks of social interaction. The
most important of these networks form relatively stably. . . . But underneath,
human beings are tunneling ahead to achieve their goals, forming
new networks, extending old ones, and emerging most clearly into our
view with rival configurations" (Mann 1986, p. 16). As Michael Mann
(1986, p. 1) observes, "societies" are best seen as "constituted of multiple
overlapping and intersecting sociospatial networks of power."23Somers
(1994, p. 72) goes even further in replacing the term "society" with "relational
setting," which she defines as "a patterned matrix of institutional
relationships among cultural, economic, social, and political practice^."^^
Such a conclusion does not invalidate the historical-comparative study of
national states or "countries" (see Goodwin 1995), but it does prescribe
considerable caution in assuming their primacy as units of sociological
analysis.
At the "meso-level" as well, the relational perspective leads to significant
reconceptualizations. To the extent that they had been theorized,face-toface
encounters were most typically seen (at least prior to Erving Goffman)
in self-actional or interactional terms, as a question of the mutual
interplay among preconstituted, self-subsistent actors. It is precisely this
framework that Goffman explodes in his celebrated studies of "copresence"
and the '(interaction order." Of paramount importance in "the
proper study of [face-to-face] interaction," he argues, "is not the individual
and his psychology, but rather the syntactical relations among the acts of
different persons mutually present to one another" (Goffman 1967, p. 2).
A "sociology of occasions" is called for that takes as its unit of analysis a
gamelike, unfolding, dynamic process, one developing within cultural, so-
23 I shall focus in the following section upon several such network contexts.
'"'For most practicing social science research, a society is a social entity. As an entity,
it has a core essence-an essential set of social springs at the heart of the mechanism.
This essential core is in turn reflected in broader covarying societal institutions that
the system comprises. . . . [By contrast,] the most significant aspect of a relational
setting is that there is no governing entity according to which the whole setting can
be categorized; it can only be characterized by deciphering its spatial and network
patterns and temporal processes. As such, it is a relational matrix, similar to a social
network" (Somers 1994, pp. 70, 72; see also Somers 1993;Somers and Gibson 1994).
American Journal of Sociology
cia1 structural, and social psychological matrices. (In what could well
serve as an epigraph for this entire manifesto, Goffman [1967, p. 31asserts:
"Not, then, men and their moments. Rather moments and their men.")2s
The study of face-to-face encounters thereupon becomes a matter of locating
regularities across such transactional processes, of specifying recurrent
mechanisms, patterns, and sequences in meso-level "occasions."
At the microscopic level of analysis, similarly, the notion of individual
can also be significantly reworked from a relational point of view. Individual
identities and interests are not preconstituted and unproblematic; parties
to a transaction do not enter into mutual relations with their attributes
already given. By reinterpreting Hobbes's construct of the "state of nature,"
Alessandro Pizzorno (1991) shows how such arguments are internally
self-contradictory; they regard as self-subsistent entities what are in
fact actors lacking in stable, durable identities to begin with. Selves, not
to mention the interests corresponding to them, require mutual trust and
reciprocal recognition to come into being, conditions that happen to be
absent in the state of nature, where a "war of each against all" prevails.
"Individuals threatened by nature to impermanence" achieve "preservation
of self, real formation of self" (Pizzorno 1991, pp. 220, 2 18) only
through transactional processes of recognition and what Pizzorno terms
mutual "name-giving": "The fiction of individuals not yet involved in social
relations but originally knowing what their interests are and what
the consequences of their choices can be is discarded in favor of a view
in which the interaction between persons mutually recognizing their right
to exist is the only originally conceivable reality. No preestablished interests
are imagined. The individual human agent is constituted as such
when he is recognized and named by other human agents" (Pizzorno 1991,
p. 220). Individual identities are thus constituted within "circles of recognition,"
while interests (a secondary construct)"grow out of different positions
in the[se] networks and circles" (Pizzorno 1991,p. 2 19).26Such circles
of recognition can include "virtual" circles with cultural ideals and fanta-
25 Goffman (1967,p. 2) characterizes these moments as "shifting entit[ies], necessarily
evanescent, created by arrivals and killed by departures." As Arthur Stinchcombe
(1991, p. 373) further explains it: "The same people act differently if they are inside
the temporal and spatial and communicative boundaries of [such] situation[s] than if
they are outside those boundaries." See also White (1973).
26 For a discussion of the relational dimension of collective identity that develops (albeit
implicitly) along similar lines to that of Pizzorno, see Melucci (1996).In true transactionalist
fashion, Melucci emphasizes the fluid, processual nature of collective identification,
a point of view, however, that sits uncomfortably with his own substantialist
terminology, as he himself acknowledges: "We should. . . take notice of the fact that
the term 'identity' remains semantically inseparable from the idea of permanence and
may, perhaps for this very reason, be ill suited for the processual analysis for which
I am arguing" (Melucci 1996, p. 72).
Manifesto
sized objects, as well as circles of interpersonal, social relationships. (Recent
feminist theorists have developed quite similar insights; indeed, gender
theory in general might well be the approach that in recent decades
has most widely disseminated transactional modes of reasoning within the
social sciences. See, e.g., Scott [1988]; Benhabib [1992]; see also Somers's
important work on "narrative identity" [e.g., Somers and Gibson 19941.)
Delving more deeply still into the individual level, relational perspectives
also make possible the recasting of long-established lines of theorization
regarding intrapsychic processes. In the psychoanalytic literature, for
example, the standard view is that of drive theory, which conceptualizes
the individual actor as a separate, monadic entity with physically based
urges that seek out psychical expression in the form of sexual and/or aggressive
desires. These desires, which are preconstituted, conflict with the
demands of both human civilization and the natural world; psychical life
builds around "compromises" between them and the defenses that control
and channel them. By contrast, a new school of thought in psychoanalysis
emphatically rejects this essentialist perspective and proposes instead a
theory of "relational individualism,"one that sees transactions with others,
and not pregiven drives, as the basic units for psychological investigation.
"Object-relations theory does not need to idealize a hyper-individualism;
it assumes a fundamental internal as well as external relatedness to the
other. The question is then what kind of relation this can and should be.
The relational individual is not reconstructed in terms of his or her drives
and defenses but in terms of the greater or lesser fragmentation of his or
her inner world and the extent to which the core self feels spontaneous
and whole within, rather than driven by, this world" (Chodorow 1989a,
p. 159).27Freudian gender theorists (e.g., Chodorow 1978, 1989b; Benjamin
1988) utilize such insights with great effecLZ8
''These very same assumptions are crucial to the new "relational school" of psychoanalytic
theory. In this approach, as Stephen Mitchell (1988, p. 3) summarizes it, "We
are portrayed [as]shaped by and inevitably embedded within a matrix of relationships
with other people. . . . Desire is experienced always in the context ofrelatedness, and
it is that context which defines its meaning. Mind is composed of relational configurations.
The person is comprehensible only within this tapestry of relationships, past
and present."
'*Transactional thinking appears in nonpsychoanalytic theories as well. For example,
Norbert Wiley argues against both old and new versions of faculty psychology, which
posits pregiven, innate properties in human nature, in favor of a more "dialogic" perspective
influenced by C. S. Peirce and George Herbert Mead. For him, the self is a
structure consisting of three elements, the I, you, and me, in continual interaction
with each other and with other selves in an ongoing "semiotic flow" of meaning. From
Wiley's (1994,p. 72) transactional perspective, the self is "a kind of public square . . .
the members of which are in constant conversation." Also relevant here are the writings
of Kenneth Gergen (e.g., 1994).
American Journal of Sociology
RESEARCH DIRECTIONS AND TECHNIQUES
In addition to having far-reaching implications for theoretical inquiry, the
transactional perspective also opens up many new, exciting directions for
substantive research. In this section, I shall consider several of these more
empirical lines of investigation, using as my main organizing principle the
idea of three transpersonal, relational contexts within which all social action
unfolds: social structure, culture, and social psychology. Each of these
"environments" (Alexander 1988),I assume, operates according to its own
partially autonomous logic, intersecting with the others in varied and interesting
ways that require empirical study (for a fuller discussion, see
Emirbayer and Goodwin [1996, 19971; Emirbayer and Sheller [1996]).I
shall also discuss new research at the individual level, research that likewise
builds upon transactional assumpti~ns.~~
The best developed and most widely used approaches to the analysis
of social structure are clearly those of social-network analysis. This perspective
is not primarily a theory or even a set of complicated research
techniques, but rather a comprehensive new family of analytical strategies,
a paradigm for the study of how resources, goods, and even positions
flow through particular figurations of social ties. Eschewing self-actional
approaches that begin with preconstituted individuals or groups, as well
as interactional approaches such as statistical (variable) analyses, network
analysts pursue transactional studies of patterned social relationships
(Breiger 1974). They adhere to what has been termed an "anticategorical
imperative" (Emirbayer and Goodwin 1994), rejecting the
primacy of attributional categories and other substantives in favor of
dynamic, "observable processes-in-relations" (White 1997, p. 60)-notice
the felicitous combination here of temporal and spatial imageriesthrough
which pass money, friendship, information, and other elemenk30
Z9 Although I shall range widely across a number of different empirical literatures, I
wish to underscore here the desirability of eventually elaborating a common relational
language within which to theorize about all these various contexts and levels of anal-
ysis.
30 Peter Bearman supplies a vivid illustration of the comparative utility of networkanalytic
and categorical approaches for understanding elite social action in England
from 1540to 1640:"Consider the tortuous debate over the appropriate bases for classifying
the gentry that arose because historians came to recognize that the gentry as a
group acted neither coherently nor uniformly with respect to interests. A primary
solution to the problem has been a continued subcategorization, and we now have as
salient groups the middling, official,rising, falling, court, and country gentry, to name
only the most prominent. These subcategories do correspond to attributes that real
gentry had. . . . But it was as likely that court gentry would rise as fall; that Puritan
gentry would be at court or in the country; and that official gentry would be as mere
as the middling or as grand as the rising. Gentry assigned to the same category frequently
acted at cross-purposes. . . .The solution to understanding elite social action
[thus]cannot be further subclassification from received categories. Categorical models
Manifesto
(This does not mean that in certain cases network analysts do not unduly
privilege synchrony over diachrony, spatial over temporal figurations, a
point to which I shall return below. Nor does it mean that they never
import into their explanations self- or interactional assumptions, most
problematically a rational-choice approach to action;for a critique of such
eclecticism, which occurs in other lines of transactional research as well,
see Emirbayer and Goodwin [1994].) Important here is the notion that
social networks cut across discrete communities and other entities-are
"interstitialn-even though in certain cases they may also congeal into
bounded groups and clusters (network researchers speak in such instances
of "CATNETS" [CATegories +NETworks; see White 1966;Tilly 19781).
Significant as well is the insight that transactions unfolding within social
networks are not always symmetrical in nature: flows are often "directional"
in content and intensity, with significant implications for actors'
differential access to resources. Finally, often important is the patterning
of invisible relations among actors-"relations visible only by their absence"
(Burt 1992, p. 181)-since such "structural holes" or gaps in networks
usually mean disparities in access to both information and control
benefits.
Network analysts draw heavily upon the methodologies of sociometry
and graph theory (the mathematical study of structural patterns in points
and lines) to formally represent social figurations. They also draw upon
the techniques of "multidimensional scaling" to map out sometimes complex
patterns in such ties, and upon set theory to model role structures
algebraically (for an overview, see Scott [1991]). These methods figure
prominently in cluster analyses as well as in studies of structural equivalence,
the two central currents in social-network research.31The former
approach focuses attention upon actors' direct and indirect ties, explaining
social processes through the very fact of connectivity itself, as
well as through the strength, density, and so forth, of the ties that bind.
It seeks to specify the relative centrality of actors within networks, the
"prominence" of such actors, and the subgroupings to which they belong.
"Network structure is described in terms of the typical relations in which
individuals are involved and the extent to which actors are connected
within cohesive primary groups as cliques" (Burt 1980, p. 81). The latter
approach emphasizes more the patterning of actors' ties, not with one
alone rarely partition people in a way that confirms with observed action, because
individual activity in the world is organized through and motivated not by categorical
affiliations but by the structure of tangible social relations in which persons are embedded"
(Bearman 1993, pp. 9-10).
3L In Emirbayer and Goodwin (1994),these were referred to as "relational" and "positional"
currents, respectively-the term "relational" having a different meaning there
than in the present context.
American Journal of Sociology
another, but with third parties; the relevant question here is the specific
position or role that a set of "structurally equivalent" actors occupies
within a given network. "Network structure is described as interlocked,
differentially prestigious status/role-sets,in terms of which actors in a system
are stratified" (Burt 1980,p. 81).An algebraic procedure called "blockmodeling"
partitions overall populations into such equivalency classes.32
Relational approaches to the sociological study of culture are not nearly
so well developed as those concerned with networks of social relationships.
Yet they share many of the same basic assumptions, beginning with
the notion that cultural formations entail, not individual "attitudes" or
"values," much less disembodied "systems," but rather bundles of communications,
relations, or transactions. Relational methodologies come into
play when analyzing the meaning structures that order or organize these
patterns. Such methodologies are transactional or relational precisely because
they involve "a shift away from thinking about a concept as a singular
categorical expression to regarding concepts as embedded in complex
relational networks that are both intersubjective and public. . . . That is,
concepts cannot be defined on their own as single ontological entities;
rather, the meaning of one concept can be deciphered only in terms of its
'place' in relation to the other concepts in its web. What appear to be
autonomous categories defined by their attributes are reconceived more
accurately as historically shifting sets of relationships that are contingently
stabilized" (Somers 1995, p. 136). Such insights, of course, are familiar
from Saussurean linguistics and structural anthropology (Saussure 1959;
LCvi-Strauss 1963),both of which stress in eminently relational terms that
meaning derives not from the intrinsic properties of signs (understood dyadically
as "sound-image" and "concept," "signifier" and "signified"), but
rather from their differences from all other signs within a semiological
system.33Such ideas are also central to two other sources of semiotic theory
that are perhaps less familiar, but even more congenial to transactional
thinking: Peircean semiotics and Bakhtinian dialogism. Charles Sanders
Peirce, the founder of both pragmatism and semiotics, agrees on the relational
embeddedness of symbols but diverges sharply from the Saussurean
tradition in taking as his unit of analysis not dyadic structures or snap-
32 Expressed more accurately, blockmodeling tests the hypothesis that a set of relations
breaks down into a set of equivalency classes specified a priori. 

33 Also relevant here is the seminal work of Roman Jakobson: "Jakobson was drawn 

to the fact that for Einsteinian physics, as for Cubism, everything is based on relation- 

ship. . . .The artist's credo 'I do not believe in things, I believe in their relationship' 

(Georges Braque) thus joined the mathematician's motto 'It is not things that matter, 

but the relations between them.' . . .Or as Jakobson himself put it much later, 'Atten- 

tion must be paid not to the material units themselves but to their relations' "(Waugh 

and Monville-Burston 1990, p. 5). 

Manifesto
shots, but rather a triadic process of "sign," "object," and "interpretant":
"A sign. . .is something which stands to somebody [interpretant] for something
[object] in some respect or capacity. It addresses somebody" (Peirce
1931-58, p. 228). As one commentator observes, for Peirce "a sign only
has meaning in the context of a continuing process of interpretation. Because
[of this], his theory is intrinsically processual and thus incompatible
with Saussure's dyadic and intrinsically static theory" (Rochberg-Halton
1986,p. 46). A similarly processual approach to language and culture can
be found in the writings of Mikhail Bakhtin, who takes the "utterance"
as his point of departure. Words, concepts, and symbols derive their meaning
only from their location within concrete utterances, but these in turn
only make sense in relation to other utterances within ongoing flows of
transactions: "The utterance is filled with dialogic overtones. . . . Each
individual utterance is a link in the chain of speech communication. . . .
Like Leibnitz's monad, [it] reflects the speech process, others' utterances,
and, above all, preceding links in the chain" (Bakhtin 1986, pp. 92-93).
Cultural analysts are increasingly turning toward such relational approaches
for methodological guidance in studying conceptual or symbolic
networks. Much empirical research in anthropology, history, and sociology,
for example, draws upon the semiotic theories mentioned above in
analyzing the internal logics and contradictions of cultural languages, idioms,
and meaning systems. Research programs articulated by Jeffrey Alexander
and Philip Smith (1993), Eugene Rochberg-Halton (1986), and
Marc Steinberg (1996)exemplify these Saussurean (cum LCvi-Straussian),
Peircian, and Bakhtinian perspectives, respectively. In a more formalized
fashion, researchers are also drawing upon sociolinguistics in analyzing
"speech communities," or sets of transactions marked by shared "knowledge
of the communicative constraints and options governing a significant
number of social situations" (Gumperz 1972,p. 16).34Finally, certain analysts
of culture are drawing, not upon semiotic or linguistic analysis, but
rather upon the relational social-network strategies outlined above. One
such line of work features cluster or connectivity analysis, measuring the
ties between "focal concepts" and other symbols within "semantic networks"
in terms of their "density," "conductivity," and "consensus" (Carley
and Kaufer 1993; Carley 1994).Another line of work pursues a structural
equivalence approach, using blockmodeling to determine the formal structure
of "discourse roles" or positions within classification schemas, rhetorics,
or other sets of cultural practices (Mohr 1994).Together, these lines
of inquiry provide a wealth of techniques for modeling the internal struc-
3 3 e ealso Halliday (1976, 1978)on "speech registers," or shared modalities of language
use within a speech community, and Hanks (1990) on "discourse genres"; the latter
is deeply influenced by Bakhtinian dialogic theory.
American Journal of Sociology
ture of cultural formations, techniques that yield important results at both
micro- and macroscopic levels of investigati~n.~~
Relational studies of social psychology are perhaps the least well developed
of the three categories of research (but see Ritzer and Gindoff 1992).
Here the objects of analysis are psychical patterns that constrain and enable
action by channeling flows and investments of emotional energydurable,
transpersonal structures of attachment and emotional solidarity,
as well as negatively toned currents of hostility and aggression. The emotions,
from this point of view, inhere not in " 'entities' that have been
located in individuals, such as 'personalities' or 'attitudes,' " but rather
in "situational ways of acting in conversational encounters" (Collins 1981,
p. 1010),that is, in transactional dynamics. Some of the most significant
research done in this vein appears in the "new structural social psychology."
Randall Collins, for example, argues that enhanced levels of the
physical density of transactions, together with the increasing ecological
boundedness of a group, raise the group's focus of attention and the intensity
of common emotions. For any actor, social life consists in a long series
of such interaction rituals, "an intermittent chain of IRs" (Collins 1993,
p. 2 lo), with levels of emotional energy accumulating gradually across
them. Collins envisions an emergent market of such rituals, with certain
figurations yielding higher emotional energy rewards than others, yet requiring
higher entry costs in the coin of emotional "investment." In this
way, he opens up possibilities for mapping the structure of emotional flows
across the broader social psychological environment. Structures of energyenhancing
ritual interactions would parallel in such a mapping the social
structural and cultural figurations of the other two relational contexts of
action. Collins is currently developing new computer software to chart
these emotional flows with greater specificity.
Finally, relational analyses of self-dynamics and individual psychology
have become increasingly prevalent. (Such investigations appear at a different
level of analysis than those discussed above, the former being personal
and the latter transpersonal in nature.) Relational studies of personality
conceptualize it, not as fixed traits or dispositions that endure across
times and contexts, but rather as a stable configuration of "distinctive gthen
situation-behavior" transactions, or characteristic patterns of behavior
within distinct but meaningfully similar circumstances. While the traditional
substantialist perspective regards situations as "error" and aims
to average out their effects by aggregating across them, the transactional
approach takes the actor's pattern of engagement with "evoking situations"
as its basic unit of analysis. "This shift call[s] attention away from
35 Mention should also be made here of the work of Bruno Latour (1987), which combines
relational thinking at both the social structural and cultural levels of analysis.
Manifesto
inferences about what broad traits a person has, to focus instead on what
the person does in particular conditions in the coping process" (Mischel
1990, p. 116). One especially rich line of substantive research on such
"behavioral signatures" is that of Walter Mischel and his associates in a
series of studies of children in a residential summer camp. Through analysis
of observations and computer codings of the children's behavior (undertaken
by camp counselors as well as independent judges), these researchers
find that "there are characteristic intraindividual patterns in
how [the children] related to different psychological conditions and that
these patterns form a sort of behavioral signature that reflects personality
coherence" (Mischel and Shoda 1995,p. 251).Mischel and his associates do
not map out these behavioral patterns sociometrically, but their findings
certainly do point them in precisely such a direction (see White 1994b).
PERPLEXITIES, DIFFICULTIES, CHALLENGES
Thus far, I have discussed the considerable promise of transactional theory
and research, saying relatively little about the issues that it has not
yet fully resolved. Despite its many important contributions, this perspective
still confronts a number of unanswered questions. In the section that
follows, I shall survey the most significant of these problems, taking up
in turn the issues of boundaries and entities, network dynamics, causality,
and normative implications.
The problem of boundary specification, of moving from flows of transactions
to clearly demarcated units of study, from continuity to discontinuity,
is perhaps the most frequently encountered of all challenges to relational
analysis. Social-network researchers, for example, continually
grapple with the question of where to draw lines across relational webs
possessing no clearcut, natural boundaries. Analogous questions arise in
respect to cultural and social psychological matrices. No definitive solution
has been found to such diffi~ulties.~~According to one set of network
investigators, two basic strategies exist for demarcating boundaries: "realist"
and "nominalist." The first takes the point of view of the actors involved,
treating a network "as a social fact only in that it is consciously
experienced as such by the actors composing it." The second proceeds
from the concepts and purposes of the social-scientific observer instead,
36 For studies of boundary drawing, see Silverstein (1979), Zerubavel (1991), and Lamont
and Fournier (1992). Ronald Breiger notes (personal communication, July 11,
1996) that one "approach . . .is to internalize, or endogenize, or 'make reflexive' this
problem within the frame of a relational analysis.(Substantialist)boundaries are [then]
played off against boundaries defined by relations." For a substantive example, see
Breiger (1981).
American Journal of Sociology
taking the correspondence between "the investigator's analytically drawn
boundaries and the subjective awareness of these boundaries by participants
[as] an empirical question rather than an assumption" (Laumann,
Marsden, and Prensky 1983, pp. 20-2 Thus, Bourdieu's concept of a
"field of practice" falls squarely within the "nominalist" side of this distinction:
its boundaries are drawn in accordance with the observer's (and
not the participants') frame of reference. The overall intractability of the
boundary problem, on the other hand, is eloquently attested to by the
apparent circularity in Bourdieu's own "solution" to it: "We may think of
a field as a space within which an effect of field is exercised. . . . The
limits of the field are situated at the point where the effects of the field
cease" (Bourdieu and Wacquant 1992,p. 100).When beginning with ramifying
webs of relations rather than substances, it becomes notoriously difficult
to justify the empirical boundaries that one draws.
Behind the analytical and methodological difficulties in setting boundaries,
moreover, stand fundamental questions of an ontological nature.
Once one defines the boundaries around a given matrix of transactions,
how does one characterize what obtains inside those boundaries? When,
if ever, does a set of relations actually count as a "thing," a substance, or
an entity? In Identity and Control, Harrison White (1992)points out that
some (although not all) clusters of events or transactions may be deemed
entities. An entity or "identity," he asserts, is "any source of action not
explicable from biophysical regularities, and to which observers can attribute
meaning. . . . Clusters can come to be perceived as and act as identities,
if they reappear repeatedly or in a variety of other contexts" (White
1992, pp. 6-7). Andrew Abbott (1996) further makes clear that not all
"zones of difference" within a social process or social space become entities;
the "entitativity" (Campbell 1958)of a recurrent phenomenon consists
in its exhibiting two key features: "coherence," some structure of causes
internal to it that permit it to recur, to "keep happening in the same way,"
and "causal authority," an "independent standing as a site of causation,
as a thing with consequences . . . [an ability] to create an effect on the
rest of the social process that goes beyond effects . . . merely transmitted
through the causing entity from elsewhere" (Abbott 1996, p. 873). The
difficulty, of course, is that neither element in Abbott's definition can be
seen as given in a processual ontology; both must be explained. Identities
3T Laumann et al. (1983) add that one can further distinguish three approaches (each
with its attendant methodologies) that cut across the two just mentioned strategies
for determining whether or not given actors will be included inside a boundary: possession
of a specified attribute; involvement in a specified type of relationship; or participation
in a given event or activity.
Manifesto
and entities are problematic in transactional approaches in ways that they
never can be in a substantialist mode of analysis.38
Closely related to the vexing issues of boundaries and identities is that
of network dynamics. Paradoxically (for a mode of study so intently focused
upon processuality), relational sociology has the greatest difficulty
in analyzing, not the structural features of static networks, whether these
be cultural, social structural, or social psychological, but rather, the dynamic
processes that transform those matrices of transactions in some
fashion. Even studies of "processes-in-relations," in other words, too often
privilege spatiality (or topological location) over temporality and narrative
unfolding.39(The reason for this state of affairs can probably be traced
back to the same hegemony of substantialism that makes self- and interactional
approaches so very prevalent.) Again, the most promising "ways
out" come from the relational analysis of social structures. In his inquiry
into "structural holes," for example, Ronald Burt (1992) surveys the circumstances
that maximize "structural autonomy," the capacity to exploit
entrepreneurially whatever information and control benefits a network
affords.40Burt's insight is that what best explains action are not the attributes
of social actors themselves, but rather the complex figurations of
relations (especially relations of absence, or "holes")to which they respond.
(In principle, parallel arguments could also be developed in respect to
cultural and social psychological figurations.) In a major network study,
There is, in other words, an unavoidable trade-off in attendant "problems" when
shifting from substantialist to relational points of view: "By making change our constant,we
also exchange our explananda. It becomes necessary to explain reproduction,
constancy, and entity-ness, rather than development and change." More to the point,
Abbott continues, "the central reason for making this [shift] is practical. It is possible
to explain reproduction as a phenomenon sometimes produced by perpetual change;
it is not possible to explain change as a phenomenon sometimes produced by perpetual
stasis" (Abbott, 1997, p. 98). It is difficult to imagine a more concise and persuasive
articulation of the relative advantages of transactional vis-a-vis substantialist perspec-
tives.
j9For a classic case in point, see the celebrated article on blockmodeling by White,
Boorman, and Breiger (1976),which provides no more than a mere succession of static
representations (or "snapshots") of social structure. (For a critical analysis, see Emirbayer
and Goodwin [1994].)More recent approaches (e.g.,the special issue on network
dynamics published by the Journal of Mathematical Sociology [1996])provide exciting
technical innovations but do not satisfactorily resolve the problem. In the area of
cultural analysis, meanwhile, much the same can be said of the structuralism of
Claude LCvi-Strauss (1963), which itself rests quite explicitly upon a conceptual dichotomization
of structure and time.
"According to his definition, "Players with relationships free of structural holes at
their own end and rich in structural holes at the other end are structurally autonomous"
(Burt 1992, p. 45).
American Journal of Sociology
John Padgett and Christopher Ansell (1993)combine this structural autonomy
notion with a focus upon "robust action," the "flexible opportunism"
made possible by certain actors' "structurally anomalous" locations within
social networks marked by deep structural holes. Still other analysts (e.g.,
Gould and Fernandez 1989)examine the pivotal roles that "brokers" play
in dynamic processes-in-relations. These researchers all converge upon the
idea that network dynamics are dialectically related to network structures,
each of these "moments" partially conditioning the other.
Yet another inquiry into dynamic processes is currently being carried
out by Harrison White, whose major contribution has been to shift the
analytic focus of attention away from social networks alone toward sociocultural
units which he terms "network-domains," or "netdoms." (White
speaks relatively little about social psychological formation^.)^' Actors
continually "switch" between netdoms as they negotiate the subtle transitions,
sometimes barely perceptible, among the various modes of social
interaction that mark everyday life. "Publics," a simple example of which
is the momentary "coffee break," facilitate such transitions; in them, social
times and meanings pertaining to specific netdoms are suspended and network-domains
themselves are "decoupled" from one another in a sort of
"temporary bubble" of "continuing present time."42Switches also occur
within netdoms, among the various "stories" or accountings of social reality
that together are borne by and serve to constitute each networkdomain.
"In ordinary, everyday social relations," writes White, "multiple
alternative accounts are being carried along until temporary resolutions at
disjunctions which I call switches. At a switching,the continuing juggling
among a set of stories is resolved into the account from which the next
phase of reality constructing takes off, among relations cohering through
that there and then" (White 1996,p. 1049).Here White invokes from statistical
science the idea of "Bayesian forks" to capture the uncertainty or
ambiguity-"What's going on here?"-that arises and is resolved at network
switchings. By means of such phenomenological categories, White
helps to enhance our understanding of the mechanisms driving network
processes. What still remains missing from his account, on the other hand
(as from other approaches to network dynamics outlined above), is any
41 "Networks catch up especially the cross-sectional patterns of connection and resonance
in interaction. Domains catch up especially the meanings and interpretations
which are the phenomenology of process as talk. These two, networks and domains,
come together for the type of tie and . . . for construction of social meanings and
times" (White 1996, p. 1038).
42 Publics ease switches because "it is easier to evolve a mere 2n ways to enter and
exit a common public state from n distinctive network-domains than it is to evolve
the much larger (n times n - 1) number of ways to switch from one to another of
the network-domains" (White 1996, p. 1056).
Manifesto
systematic, normative consideration of the role of what Dewey termed
"intelligence" in the intentional guidance and direction of human affairs.
Social actors' reflexive engagement with the problems confronting them
in their everyday lives remains significantly undertheorized in recent studies
of network processes (more on this below as well).
A third related challenge facing the relational approach concerns causality.
How are network dynamics to be accounted for? How are shifts
in the content and direction of transactional flows to be explained? One
persistent problem, yet another outgrowth of substantialist ways of reasoning,
is the tendency even of many relational thinkers to depict causes
as immaterial phenomena. Entities such as "forces," "factors," and "structures"
are said to "impel" social substances, including persons and groups,
down the causal path. Even figurations or patterns of relations (including
structural holes) become substantives that putatively "move" action along
in some fashion. Can causality be conceptualized in a more satisfactory
way? One writer begins to address this problem by calling for a new "action
language" for the social sciences. Roy Schafer (1976),a psychoanalytic
theorist, proposes that we "regard each psychological [and, by extension,
social] process, event, experience, or behavior as some kind of activity,
henceforth to be called action, and . . . designate each action by an active
verb stating its nature and by an adverb (or adverbial locution) when
applicable, stating the mode of this action. Adopting this rule entails that,
insofar as it is possible to do so sensibly, we shall not use nouns and adjectives
to refer to [transactional] processes, events, etc." (Schafer 1976, p.
9). A sentence such as "their disadvantaged position led to heightened
competitiveness" should thus be translated as "they responded to their
disadvantageous situation by acting more competitively." Social actors
themselves, in gamelike transactions within ever-changing contexts, do
all of the acting in social life, not some imaginary entities within or without
them, as in the substantialist worldviews of self- or interaction.
Of course, such a methodological prescription remains useful only insofar
as one bears in mind that not all social actors are individual persons:
in certain contexts, organizations or other identities might also be deemed
actors for purposes of transactional analysis. More important, social actors
are always embedded in space and time; they respond to specific situations
(opportunities as well as constraints)rather than pursuing lines of conduct
in purely solipsistic fashion. The narratives of their responses (together
with the situations within which these occur) help to explain how causes
actually produce effects in history. Action language, in other words, clears
the ground for causal analysis by eliminating reified structures as "causal
factors," yet it needs to be supplemented by an explicit concern for the
"situational mechanisms" (Stinchcombe 1991)that actually channel flows
of events. Such mechanisms provide an answer to those who would sug-
American Journal of Sociology
gest (as variable-centered analysts often do, e.g.) that action language
abandons the search for causal generalities in social life. Tilly provides
an example of the sort of mechanism one might now be looking for: "The
relationship among an activity, the set of agents that control the means
that might make that activity possible, and the bargaining that goes on
between the agents of the activity and those who hold the resources, produce
unexpected sets of structures that themselves constrain the next
round of action. . . . The causal mechanism lies in the bargaining that
comes out of resistance to release the resources that are already committed
to other ends. . ..The general cause lies in that struggle over control over
wanted resources" (Tilly 1993,p. 6).This approach to causation retains the
action orientation that Schafer advocates, yet avoids the latter's tendency
toward pure voluntarism by insisting upon a search for robust explanatory
processes that operate across a multiplicity of social situations.
Finally, the normative implications of relational inquiry come into
question when one considers what sorts of moral and practical leverage
such inquiry might provide in respect to social realities. We can distinguish
here between "critical" and "reconstructive" implications. Preeminent
among the former is the capacity afforded by transactional thinking
to "unfreeze" static, substantialist categories that deny the fluidity-hence,
the mutability-of figurational patternings. In present-day cultural studies,
for example, "essentialist" modes of thinking all too often see individuals
and collectivities as possessing singular, unitary "identities" rooted in
race, class, gender, or sexuality; putatively, such fixed attributional features
explain both "interests" and action unpr~blematically.~~Nor is such
essentialism limited to cultural analysis alone; contemporary "identity politics"
largely revolves around the attempt similarly to thematize (as well
as to legitimize) particular collective identities (e.g., "African-American,"
"Latina woman") long devalued in mainstream culture. More clearly pernicious
essentialisms (e.g., racism, [hetero-]sexism, xenophobic nationalism)
share in such reasoning, if not in its progressive ideological agenda.
Now, categories certainly do matter a great deal in social life; as W. I.
Thomas famously pointed out, "If men define situations as real, they are
real in their consequences" (Thomas and Thomas 1928, p. 572). Moreover,
categorical thinking can have crucial critical implications, such as
normatizing previously devalued categories or redistributing resources by
means of social policy. Yet, transactional thinking contests the intrinsically
43 Versions of cultural studies that speak in more nuanced terms of the intersection
of multiple identities, as in the category of "black lesbians,"add considerable complexity
to the picture, yet still do not escape the difficulties that pertain to all categorical,
substantialist thinking.
Manifesto
reified nature of all categories: it shows how they "totalize" identities that
are in fact often multidimensional and contradictory; prescribe modes of
thought and action against which alternatives can only be labeled "deviant";
naturalize rigid distinctions that suppress possibilities for creative
(self-) transformation;and, most generally, accept rather than contest the
historically variable relational matrices that serve to constitute invidious
distinctions and categorizations in the first place (Somers and Gibson
1994, pp. 55-57).44 Transactional thinking, in a word, deconstructs a
taken-for-granted moral universe; in so doing, it attacks in moral and
practical life the very same tendencies toward "process-reduction" that
Elias so forcefully underscored at the level of ~ognition.~'
In a more reconstructive vein, relational reasoning also speaks directly
to the problem of what is to count as a "better" or "worse" line of conduct
(or end point of action), from an ethical or political point of view. If values
are not preconstituted substances, then whence do they arise and how are
they to be evaluated? What kinds of transactions ought to be most highly
valued in a processual, relational view of the world? Such issues, which
have long vexed transactional theorists, are perhaps most provocatively
addressed by classical American pragmatists such as John Dewey-with
whom we began-and George Herbert Mead (although they were also
engaged with in similar ways by Bakhtin [I9931 and later by Jiirgen Habermas
[1984-871). For the pragmatists, normative implications flow naturally
out of the central concept of transaction itself: "Values are not simply
objective givens, which are independent of human existence. They
are, however, also not merely the product of the subjective evaluation of
objects which are essentially neutral with regard to this evaluation. Rather
. ..the evaluation is the result of an 'interaction' [or transaction] of subject
and object" (Joas 1985, p. 131). Thus, values are by-products of actors'
engagement with one another in ambiguous and challenging circumstances,
which emerge when individuals experience a discordance between
the claims of multiple normative commitments. Problematic situa"See
n. 38, above. In practical politics and cultural struggle, social actors often invoke
collective identities strategically, even as they question categorical thinking theoretically.
"The point . . . is not to stop studying identity formation, or even to abandon
all forms of identity politics, but rather . . .to rely on notions of identity and identity
politics for their strategic utility while remaining vigilant against reification" (Epstein
1994, p. 197).
45 For examples of self-consciously deconstructionist, "poststructuralist" thinking in
respect to questions of identity, see Scott (1988) and Sedgwick (1990). Relationalism
by no means entails adoption of all aspects of deconstructionism, even if it does share
important features with it. The continuities (and discontinuities) between deconstructionism
(Derrida 1974),on the one hand, and contemporary relational perspectivesfrom
neopragmatism (esp. Rorty 1982)to social-network theory (esp. White 1992),on
the other-would make for a valuable comparative study.
American Journal of Sociology
tions of this sort become resolved only when actors reconstruct the
relational contexts within which they are embedded, and in the process,
transform their own values and themselves: "The appearance of. ..different
interests in the forum of reflection [leads to] the reconstruction of the
social world, and the consequent appearance of the new self that answers
to the new object" (Mead 1964, p. 149).
Appealing visions of both action and order follow directly from this line
of reasoning. The rational resolution of problem-situations-a matter of
action-requires capacities for prudential reasoning and practical judgment,
which the pragmatists (beginning with Dewey) termed "intelligence."
In this mode of action, "the individual . . .take[s] into account all
of those interests [that are implicated in a given situation] and then
make[s] out a plan of action which will rationally deal with those interests"
(Mead 1934, p. 388; see Emirbayer and Mische, in press). The actor puts
himself in the place of the other, broadens his own point of view through
argumentative engagement with the latter, and attains in this manner to
ever more comprehensive, cosmopolitan, universalistic perspectives. In
parallel fashion, the ideal mode of mutual engagement or transactiona
matter of order-entails a free and open communication of actors in a
universal community, a relational matrix within which both cooperation
and conflict are rationally regulated. This "mode of associated livingnin
a word, democracy-embodies moral intelligence on a transpersonal
scale; it involves "conjoint communicated experience" in which practical
reasoning is undertaken in common, through inquiry into moral and political
problems on the model of an experimental science (Dewey 1980, p.
93). Such a point of view-as it pertains to both action and order-does
resolve many problems concerning the normative implications of relational
reasoning, while also remaining well within a transactional frame
of reference. What it does not adequately address, however, is the question
of whether the relevant standards of normative judgment are to be substantive
or merely procedural: Does the idea of free and open communication
in transactional processes mean nothing other than a formal method
of intelligent reasoning, or does it lead to a view of moral character and
collective social arrangements that is more contentful? This is a question
that relational thinkers continue to be troubled by (to the extent that they
address it at all), and that they have not yet-and may never-satisfactorily
resolve.46
" One of the most serious shortcomings of relational sociology to date is its relative
neglect of normative concerns, despite the profound interpenetration (in true transactionalist
fashion) of all questions of "is" and "ought" in social-scientific analysis.
Manifesto
CONCLUSION
Such unresolved issues and problems notwithstanding, of course, the
promise of a relational mode of sociological inquiry remains considerable.
Philosophically, theoretically, and empirically, it offers a compelling alternative
to the currently ascendant perspectives of self- and interaction, and
to their present-day champions in rational-choice, neo-Kantian, structuralist,
and variable-based sociologies. New and exciting approaches to cultural,
social structural, and social psychological analysis already exemplify
this vast potential; ongoing efforts at the reconstruction of major theoretical
concepts further attest to it. Moreover, social thinkers from a wide
variety of disciplinary backgrounds, national traditions, and analytic and
empirical points of view are fast converging upon this frame of reference,
often without even grasping its full significance. A set of assumptionssome
might call it a paradigm-that first attained to systematic expression
in sociological theory, as in other fields such as physics, near the turn of
the century (its antecedents reach much farther back in time, of course)
is finally now, at the close of the century, starting to receive the widespread
attention it so richly deserves. Despite the sharp attention accorded to so
many other debates, dualisms, and oppositions in sociology, the choice
between substantialist and relational modes of inquiry, a choice of bedrock
assumptions regarding the very nature of social reality itself, is fast
becoming the most important and consequential dividing line in sociological
investigation.
Many challenges lie ahead; the preceding section discussed only a few
of the most significant. Relational theorists and researchers must now focus
upon several tasks. One is to explore ever more aggressively the analytical
levels of culture and collective emotions, importing into these problem
areas many of the same insights and research techniques already
elaborated by network analysts but also exploiting, for example, the exciting
new approaches developed by sociolinguists and social psychological
researchers. (Analysts have been moving into the culture field in ever increasing
numbers but as yet the study of transpersonal emotional flowsthe
social psychological dimension-has remained seriously underdeveloped.)
Second, and relatedly, transactional researchers must strive resolutely
to maintain theoretical consistency across levels of analysis, not only
in their more case-specific explanations, but also, and especially, in their
general efforts at theory-building. Often the wariness of social thinkers
today regarding comprehensive theorization (an outdated carryover from
earlier battles against the Parsonian legacy?)leads to an all too easy acceptance
of hybrid models (e.g., juxtapositions of rational-actor with network-analytic
approaches). The richness and breadth of relational ways
American Journal of Sociology
of thinking allow one to avoid such ad hoc reasoning and to develop
causal explanations more self-consciously within a unitary frame of reference.
And finally, transactional thinkers must begin to systematize
some of the alternative ways in which central issues and problems have
been thematized from within their own tradition. Internal debates will
be lifted to a much higher plane-and theory-building facilitated-once
analysts begin to see differences, for example, between Bourdieu and Foucault
on "power," between Tilly and Somers on "culture," or between
Dewey and White on "intelligence," as alternative ways of proceeding
from the selfsame philosophical premises regarding processes-in-relations.
Only then will transactional sociologists be able fully to grasp the possibilities
and choices confronting them; only then will they (and the sociological
discipline more generally) finally arrive at the theoretical clarity and
reflexivity that they have long been capable of attaining.4'
REFERENCES
Abbott, Andrew. 1988. "Transcending General Linear Reality." Sociological Theory
6:169-86.
, 1992a. "What Do Cases Do? Some Notes on Activity in SociologicalAnalysis."
Pp. 53-82 in What Is a Case?Exploring the Foundations of Social Inquiry, edited
by Charles C. Ragin and Howard S. Becker. Cambridge: Cambridge University
Press.
. 19923. "Of Time and Space: The Contemporary Relevance of the Chicago
School." Sorokin Lecture, delivered at the meeting of the Southern SociologicalSo-
ciety.
. 1996. "Things of Boundaries." Social Research 622357-82.
, 1997. "On the Concept of Turning Point." Comparative Social Research 16:
85-105.
Alexander,Jeffrey C. 1983. TheoreticalLogic in Sociology. Vol. 4, The Modern Reconstruction
of Classical Thought: Talcott Parsons. Berkeley and Los Angeles: University
of California Press.
, 1988. Action and Its Environments: Toward a New Synthesis. New York:
Columbia University Press.
Alexander, Jeffrey C., and Philip Smith. 1993. "The Discourse of American Civil
Society: A New Proposal for Cultural Studies." Theory and Society 22:151-
207.
Bakhtin, Mikhail M. 1986. "The Problem of Speech Genres." Pp. 60-102 in Speech
Genres and Other Late Essays, edited by Caryl Emerson and Michael Holquist.
Translated by Vern W. McGee. Austin: University of Texas Press.
. 1993. Toward a Philosophy of the Act, edited by Vadim Liapunov and Mij7Lest
this "manifesto" seem incomplete without one, I sha!l offer one final rallyingcry:
Entities of the World-Relate! Both descriptive and prescriptive, with implications
both causal and normative, this statement (despite slight interactional overtones)
nicely captures the rich possibilities inherent in the relational, transactional vision of
social reality.
Manifesto
chael Holquist. Translated by Vadim Liapunov. Austin: University of Texas
Press.
Bates, Frederick L., and Walter Gillis Peacock. 1989."Conceptualizing Social Structure:
The Misuse of Classification in Structural Modeling." American Sociological
Review 54:565-77.
Bauman, Zygmunt. 1989. "Hermeneutics and Modern Social Theory." Pp. 34-55
in Social Theory of Modern Societies: Anthony Giddens and His Critics, edited
by David Held and John B. Thompson. Cambridge: Cambridge University
Press.
Bearman, Peter S. 1993. Relations into Rhetorics: Local Elite Social Structure
in Norfolk, England, 1540-1640. New Brunswick, N.J.: Rutgers University
Press.
Benhabib, Seyla. 1992. Situating the Self: Gender, Community and Postmodernism
in Contemporary Ethics. New York: Routledge.
Benjamin, Jessica. 1988.The Bonds of Love: Psychoanalysis, Feminism, and the Problem
of Domination. New York: Pantheon.
Bernstein, Richard J. 1966.John Dewey. Atascadero, Calif.: Ridgeview.
Bourdieu, Pierre, and Loi'c J. D. Wacquant. 1992.An Invitation to Rej?exive Sociology.
Chicago: University of Chicago Press.
Breiger, Ronald L. 1974."The Duality of Persons and Groups." Social Forces 53:181-
90.
. 1981."The Social Class Structure of Occupational Mobility." American Journal
of Sociology 87578-61 1.
. 1990. "Social Control and Social Networks: A Model from Georg Simmel."
Pp. 453-76 in Structures of Power and Constraint: Papers in Honor of Peter M.
Blau, edited by Craig Calhoun, Marshall W. Meyer, and W. Richard Scott. Cambridge:
Cambridge University Press.
. 1991. Explorations in Structural Analysis: Dual and Multiple Networks of
Social Interaction. New York: Garland.
Burt, Ronald S. 1980. "Models of Network Structure." Annual Review of Sociology
6x79-141.
. 1992. Structural Holes: The Social Structure of Competition. Cambridge,
Mass.: Harvard University Press.
Campbell, D. T. 1958."Common Fate, Similarity, and Other Indices of the Status of
Aggregates of Persons as Social Aggregates." Behavioral Science 3:14-25.
Carley, Kathleen M. 1994. "Extracting Culture through Textual Analysis." Poetics
22:291-312.
Carley, Kathleen M., and David S. Kaufer. 1993. "Semantic Connectivity: An Approach
for Analyzing Symbols in Semantic Networks." Communication Theory 3:
183-2 13.
Cassirer, Ernst. 1953.Substance and Function, translated by William Curtis Swabey
and Marie Collins Swabey. New York: Dover.
Chodorow, Nancy J. 1978. The Reproduction of Mothering: Psychoanalysis and the
Sociology of Gender. Berkeley and Los Angeles: University of California Press.
. 1989a. "Toward a Relational Individualism: The Mediation of Self through
Psychoanalysis." Pp. 154-62 in Feminism and Psychoanalytic Theory. New Haven,
Conn.: Yale University Press.
. 1989b. Feminism and Psychoanalytic Theory. New Haven, Conn.: Yale University
Press.
Colapietro, Vincent M. 1990."The Vanishing Subject of Contemporary Discourse: A
Pragmatic Response." Journal of Philosophy 87:644-55.
Collins, Randall. 1981."On the Microfoundations of Macrosociology." American Journal
of Sociology 86:984-1014.
American Journal of Sociology
. 1993."Emotional Energy as the Common Denominator of Rational Action."
Rationality and Society 5:203-30.
Commons, John R. 1924. Legal Foundations of Capitalism. New York: Macmillan.
Cooley, Charles Horton. 1962.Social Organization: A Study of the Larger Mind. New
York: Charles Scribner's Sons.
Derrida, Jacques. 1974.Of Grammatology, translated by Gayatri Chakravorty Spivak.
Baltimore: Johns Hopkins University Press.
Dewey, John. 1929.Experience and Nature. La Salle, Ill.: Open Court.
. 1980. Democracy and Education. The Middle Works of John Dewey, 1899-
1924, vol. 9. Edited by Jo Ann Boydston. Carbondale, Ill.: Southern Illinois University
Press.
Dewey, John, and Arthur F. Bentley. 1949.Knowing and the Known. Boston: Beacon
Press.
Durkheim, mile. 1995.The Elementary Forms of Religious Lge, translated by Karen
E. Fields. New York: Free Press.
Elias, Norbert. 1978. What Is Sociology? translated by Stephen Mennell and Grace
Morrissey. New York: Columbia University Press.
Elster, Jon. 1989. Nuts and Bolts for the Social Sciences. Cambridge: Cambridge
University Press.
Emirbayer, Mustafa. 1996. "Useful Durkheim." Sociological Theory 14:109-30.
Emirbayer, Mustafa, and Jeff Goodwin. 1994. "Network Analysis, Culture, and the
Problem of Agency." American Journal of Sociology 99:1411-54.
. 1996. "Symbols, Positions, Objects: Toward a New Theory of Revolutions
and Collective Action." History and Theory 35:358-74.
. 1997."Positions, Symbols, Objects: A Relational Theory of Social Action and
Historical Change." Manuscript. New School for Social Research, Department of
Sociology.
Emirbayer, Mustafa, and Ann Mische. 1998. "What Is Agency?" American Journal
of Sociology, in press.
Emirbayer, Mustafa, and Mimi Sheller. 1996. "Publics in History: A Programmatic
Statement." Paper presented at the meeting of the Social Science History Association,
New Orleans.
Epstein, Steven. 1994. "A Queer Encounter: Sociology and the Study of Sexuality."
Sociological Theory 12:188-202.
Foucault, Michel. 1979. Discipline and Punish: The Birth of the Prison, translated
by Alan Sheridan. New York: Vintage Books.
. 1990.The History of Sexuality. Vol. 1,An Introduction, translated by Robert
Hurley. New York: Vintage Books.
Gergen, Kenneth J. 1994.Realities and Relationships: Soundings in Social Construction.
Cambridge, Mass.: Harvard University Press.
Goffman, Erving. 1967. Interaction Ritual: Essays on Face-to-Face Behavior. New
York: Pantheon.
Goodwin, Jeff. 1995."A Case for Big Case Comparison."Comparative Historical Sociology
Newsletter 8:1, 6.
Gould, Roger V., and Roberto M. Fernandez. 1989."Structures of Mediation: A Formal
Approach to Brokerage in Transaction Networks." Sociological Methodology
19189-126.
Gumperz,John J. 1972.Introduction to Directions in Sociolinguistics, edited by John
J. Gumperz and Dell Hymes. New York: Holt, Rinehart, & Winston.
Habermas, Jurgen. 1983."Hannah Arendt: On the Concept of Power." Pp. 173-89 in
Philosophical-Political ProJiles, translated by Frederick G. Lawrence. Cambridge,
Mass.: MIT Press.
Manifesto
. 1984-87. The Theory of Communicative Action, 2 vols. Translated by Thomas
McCarthy. Boston: Beacon Press.
Halliday, M. A. K. 1976. System and Function in Language: Selected Papers, edited
by G. R. Kress. London: Oxford University Press.
. 1978. Language as Social Semiotic: The Social Interpretation of Language
and Meaning. London Edward Arnold.
Hanks, William F. 1990.Referential Practice: Language and Lived Space among the
Maya. Chicago: University of Chicago Press.
Husserl, Edmund. 1960. Cartesian Meditations: An Introduction to Phenomenology,
translated by Dorion Cairns. The Hague: Martinus Nijhoff.
James, William. 1976.Essays i n Radical Empiricism. Cambridge, Mass.: Cambridge
University Press.
Joas, Hans. 1985.G. H. Mead: A Contemporary Re-examination of His Thought, translated
by Raymond Meyer. Cambridge, Mass.: MIT Press.
. 1993.Pragmatism and Social Theory. Chicago: University of Chicago Press.
. 1996. The Creativity of Action, translated by Jeremy Gaines and Paul Keast.
Chicago: University of Chicago Press.
Knoke, David. 1990. Political Networks: The Structural Perspective. Cambridge:
Cambridge University Press.
Kreps, David M. 1990. Game Theory and Economic Modelling. New York: Oxford
University Press.
Lamont, Michkle, and Marcel Fournier, eds. 1992. Cultivating Differences: Symbolic
Boundaries and the Making of Inequality. Chicago: University of Chicago
Press.
Latour, Bruno. 1987. Science in Action: How to Follow Scientists and Engineers
through Society. Cambridge, Mass.: Harvard University Press.
Laumann, Edward O., Peter V. Marsden, and David Prensky. 1983."The Boundary
Specification Problem in Network Analysis." Pp. 18-34 in Applied Network Analysis:
A Methodological Introduction, edited by Ronald S. Burt and Michael J. Minor.
Beverly Hills, Calif.: Sage.
LCvi-Strauss, Claude. 1963. Structural Anthropology, translated by Claire Jacobson
and Brooke Grundfest Schoepf. New York: Basic Books.
Luhmann, Niklas. 1995. Social Systems, translated by John Bednarz, Jr., with Dirk
Baecker. Stanford, Calif.: Stanford University Press.
Macy, Michael W. 1991."Chains of Cooperation: Threshold Effects in Collective Action."
American Sociological Review 56:730-47.
Mann, Michael. 1986. The Sources of Social Power. Vol. 1,A History of Powerfrom
the Beginning to A.D. 1760. Cambridge: Cambridge University Press.
Marx, Karl. 1977. Capital, vol. 1. Translated by Ben Fowkes. New York: Vintage
Books.
. 1978. The Marx-Engels Reader, 2d ed. Edited by Robert C. Tucker. New
York: Norton.
Mead, George Herbert. 1934. "Fragments on Ethics." Pp. 379-89 in Mind, Self,and
Society: From the Standpoint of a Social Behaviorist, edited by Charles W. Morris.
Chicago: University of Chicago Press.
. 1964. "The Social Self." Pp. 142-49 in George Herbert Mead: Selected Writings,
edited by Andrew J. Reck. Indianapolis: Bobbs-Merrill.
Melucci, Alberto. 1996.Challenging Codes: Collective Action in the Information Age.
Cambridge: Cambridge University Press.
Mischel, Walter. 1990."Personality Dispositions Revisited and Revised: A View after
Three Decades." Pp. 111-34 in Handbook of Personality: Theory and Research,
edited by L. A. Pervin. New York: Guilford Press.
American Journal of Sociology
Mischel, Walter, and Yuichi Shoda. 1995. "A Cognitive-Affective System Theory of
Personality: Reconceptualizing Situations, Dispositions, Dynamics, and Invariance
in Personality Structure." Psychological Review 102:246-68.
Mitchell, Stephen A. 1988. Relational Concepts in Psychoanalysis: An Integration.
Cambridge, Mass.: Harvard University Press.
Mohr, John. 1994. "Soldiers, Mothers, Tramps, and Others: Discourse Roles in the
1907 New York City Charity Directory." Poetics 22:327-57.
Ollman, Bertell. 1971. Alienation: Marx's Conception of Man in Capitalist Society.
Cambridge: Cambridge University Press.
Padgett, John F., and Christopher K. Ansell. 1993. "Robust Action and the Rise of
the Medici, 1400-1434." American Journal of Sociology 98:1259-1319.
Parsons, Talcott. 1937.The Structure of Social Action: A Study in Social Theory with
Special Reference to a Group of Recent European Writers. New York: Free Press.
. 1969. Politics and Social Structure. New York: Free Press.
Parsons, Talcott, with Robert F. Bales and Edward A. Shils. 1953. Working Papers
in the Theory of Action. New York: Free Press.
Peirce, Charles Sanders. 1931-58. The Collected Papers of Charles Sanders Peirce,
vol. 2. Edited by C. Hartshorne and P. Weiss. Cambridge, Mass.: Harvard University
Press.
Pizzorno, Alessandro. 1991."On the Individualistic Theory of Social Order." Pp. 209-
31 in Social Theoryfor a Changing Society, edited by Pierre Bourdieu and James
S. Coleman. Boulder, Colo.: Westview; New York: Russell Sage Foundation.
Ritzer, George, and Pamela Gindoff. 1992."Methodological Relationism: Lessons for
and from Social Psychology." Social Psychology Quarterly 55:128-40.
Rochberg-Halton, Eugene. 1986.Meaning and Modernity: Social Theory in the Pragmatic
Attitude. Chicago: University of Chicago Press.
Rorty, Richard. 1982.Consequences of Pragmatism: Essays, 1972-1980. Minneapolis:
University of Minnesota Press.
Saussure, Ferdinand de. 1959.Course in General Linguistics, edited by Charles Bally
and Albert Sechehaye, in collaboration with Albert Riedlinger. Translated by Wade
Baskin. New York: McGraw-Hill.
Schafer, Roy. 1976. A New Language for Psychoanalysis. New Haven, Conn.: Yale
University Press.
Scott, Joan Wallach. 1988. Gender and the Politics of History. New York: Columbia
University Press.
Scott, John. 1991. Social Network Analysis: A Handbook. London: Sage.
Sedgwick, Eve Kosofsky. 1990.Epistemology of the Closet. Berkeley and Los Angeles:
University of California Press.
Silverstein, Michael. 1979. "Language Structure and Linguistic Ideology." Pp. 193-
247 in The Elements: A Parasession on Linguistic Units and Levels, edited by Paul
R. Clyne, William F. Hanks, and Carol L. Hofbauer. Chicago: Chicago Linguistic
Society.
Simmel, Georg. 1950. The Sociology of Georg Simmel, edited and translated by Kurt
H. Wolff. New York: Free Press.
. 1971. On Individuality and Social Forms, edited by Donald N. Levine. Chicago:
University of Chicago Press.
. 1990. The Philosophy of Money, 2d ed. Edited by David Frisby, translated
by Tom Bottomore and David Frisby, from a first draft by Kaethe Mengelberg.
London: Routledge.
Somers, Margaret R. 1993. "Citizenship and the Place of the Public Sphere: Law,
Community, and Political Culture in the Transition to Democracy." American Sociological
Review 58587-620.
, 1994. "Rights, Relationality, and Membership: Rethinking the Making and
Meaning of Citizenship." Law and Social Inquiry 19:63-112.
Manifesto
. 1995. "What's Political or Cultural about Political Culture and the Public
Sphere? Toward an Historical Sociology of Concept Formation." Sociological Theory
13:113-44.
Somers, Margaret R., and Gloria D. Gibson. 1994. "Reclaiming the Epistemological
'Other': Narrative and the Social Constitution of Identity." Pp. 37-99 in Social
Theory and the Politics of Identity, edited by Craig Calhoun. Oxford: Blackwell.
Steinberg, Marc W. 1996."Toward a Dialogic Analysis of Social Movement Culture:
Reframing Frame Analysis." Paper presented at the meeting of the Social Science
History Association, New Orleans.
Stinchcombe, Arthur L. 1991. "The Conditions of Fruitfulness of Theorizing about
Mechanisms in Social Science." Philosophy of the Social Sciences 21:367-88.
. 1995.Sugar Island Slavery in the Age of Enlightenment: The Political Economy
of the Caribbean World. Princeton, N.J.: Princeton University Press.
Thomas, William I., and Dorothy Swaine Thomas. 1928. The Child in A,nerica: Behavior
Problems and Programs. New York: Knopf.
Tilly, Charles. 1978. From Mobilization to Revolution. Reading, Mass.: Addison-
Wesley.
. 1993."Mechanisms of State Formation and Collective Action." Interview by
Ruud Koopmans and Sophie de Schaepdrijver. CSSC Working Papers Series no.
171. New School for Social Research.
. 1995a. "To Explain Political Processes." American Journal of Sociology 101:
1594-1610.
. 1995b. "Durable Inequality." CSSC Working Papers Series no. 224. New
School for Social Research.
Waugh, Linda R., and Monique Monville-Burston. 1990. "Introduction: The Life,
Work, and Influence of Roman Jakobson." Pp. 1-45 in Roman Jakobson: On Language.
Cambridge, Mass.: Harvard University Press.
Wellman, Barry. 1988. "Structural Analysis: From Method and Metaphor to Theory
and Substance." Pp. 19-61 in Social Structures: A Network Approach, edited by
Barry Wellman and S. D. Berkowitz. Cambridge: Cambridge University Press.
White, Harrison C. 1966."Notes on the Constituents of Social Structure." Manuscript.
Harvard University, Department of Sociology.
. 1973. "Everyday Life in Stochastic Networks." Sociological Inquiry 43:43-
49.
. 1992. Identity and Control: A Structural Theory of Social Action. Princeton,
N.J.: Princeton University Press.
. 1994a. "Network Switches, Domination, and Grammars: Is It Soul or Ties
that Engender Language?" Manuscript. Columbia University, Department of Soci-
ology.
. 1994b. "Social Networks Can Resolve Actor Paradoxes in Economics and in
Psychology." Pre-Print Series, Columbia University, Center for the Social Sciences.
. 1996. "Network Switchings and Bayesian Forks: Reconstructing the Social
and Behavioral Sciences." Social Research 62:1035-63.
. 1997."Can Mathematics Be Social? Flexible Representations for Interaction
Process and Its Sociocultural Constructions." Sociological Forum 12:53-71.
White, Harrison C., Scott A. Boorman, and Ronald L. Breiger. 1976."Social Structure
from Multiple Networks: I. Blockmodels of Roles and Positions." American Journal
of Sociology 81930-80.
Wiley, Norbert. 1994. The Semiotic Self: Chicago: University of Chicago Press.
Wright, Erik Olin. 1979.Class Structure and Income Determination. New York: Academic
Press.
Zerubavel, Eviatar. 1991. The Fine Line. New York: Free Press.