| The subject
of this paper is neither an exhaustive analysis
of the major trends in sociological theory today,
or a systematic treatment of key issues and debates
in that discipline. It is rather to express, in
somewhat eclectic and idiosyncratic manner, some
tentative views (based on a long-term involvement
with problems in sociological theory during both
teaching and research) on what went wrong in the
development of sociological theory, and whether
or not we can do anything about it at present.
The emphasis will be less on the history or the
sociology of sociological theory, and more on
reasons internal to the logic of the discipline
-- reasons that not only inform us about the nature
of the present unsatisfactory state of affairs,
but also suggest possible ways of overcoming it.
On the Rise and fall of Modern Sociological
Theory:
An important feature of classical sociology was
its relative non-differentiation or separation
between theory and empirical analysis. In so far
as the founding fathers developed their theories
in order to understand the kind of society that
emerged in the aftermath of the industrial and
French revolutions, the connections were very
close between their analyses of industrial societies
on the one hand, and theoretical and methodological
insights on the other. This close link between
theoretical and empirical concerns could not withstand
the growth, during this century, of sociological
theory as a distinct subfield of sociological
enquiry. Some observers consider the relative
divorce between theory and empirical research
an unfortunate development that could not help
but lead to the arid theorizing such as is found
in Parsons' work, the excessive abstractedness
of which renders it 'untestable' (Mills 1959).
Others, including myself, do not believe this
development to be as regrettable as critics of
Parsonian functionalism would imply; they see
it, if not as unavoidable, then as an irreversible
result of the growing division of labor within
sociology. In any case, the charge of abstractedness
and non-verifiability is ill-placed, since the
chief aim of modern sociological theory is not
to fashion substantive statements that can be
tested against reality, but to construct what
Althusser calls Generalities II, and what non-Marxist
sociologists call conceptual frameworks or paradigms:
i.e. sets of logically interrelated conceptual
tools for looking at social phenomena in such
a way that interesting questions are generated
and methodologically proper linkages established
between different levels of analysis. In that
sense, sociological theory does not consist of,
and does not aim directly at establishing empirically
testable hypotheses, it is merely meant to prepare
the ground for an empirical investigation of social
structures and actors.
While, therefore, the emergence of modern sociological
theory has meant a break with the great classical
syntheses, it has not estranged itself from empirical
research altogether. On the contrary, the work
of the first generation of sociological theorists
(I have in mind here writers like Parsons, Merton,
Gouldner,
Lockwood) expresses a clear concern with providing
conceptual tools facilitating empirical investigation.
This is apparent from both the conceptual frameworks
elaborated, and from the fact that such frameworks
were systematically applied in practice -- either
by the theorists themselves and/or by their disciples
-- as a means of generating empirically oriented
accounts of the social world. Using an evolutionist
vocabulary, one can say that the differentiation
between sociological theory and the more empirically-oriented
sub-disciplines was accompanied by integrative
mechanisms that ensured some linkages between
theory and empirical research.
It seems to me that it is in this light that one
should assess Parsons' complex work, which has
contributed most to establishing sociological
theory as a specialized sub-discipline. Despite
its numerous shortcomings, the Parsonian framework
could, and did, lead to considerable empirical
research on both the micro and the macro levels
of analysis (e.g. Almond and Verba 1963; Barber
1952; Bellah 1970; Deutsch 1963; Eisenstadt 1963;
Levy 1949; Lipset 1963; Smelser 1960). However,
although Parsonian sociological theory successfully
advanced the intellectual division of labor within
sociology without breaking the link between theory
and research, it did less well in terms of the
conceptual tools it offered. Applied to empirical
issues, these tools systematically neglected the
voluntaristic dimensions of social life, and this
led more or less directly to a reified, positivistic,
and often teleological treatment of social phenomena.
As numerous critics have pointed out again and
again,
Parsons -- especially when his analysis progressed
from his early theory of social action to a theorization
of social systems, and their long-term evolution
-- over-emphasized the systemic-functionalist
dimensions of social systems at the expense of
agency. His theory either portrays agents as passive
outcomes of a system of core societal values,
or ignores them entirely. More precisely, on the
level of micro-action, Parsons fails to develop
a theory of interaction, his analysis moving abruptly
from a theorization of unit acts to a theorization
of social systems (Turner 1990; Habermas 1987).
As a result of this omission, Parsons' role players
seem to be exclusively guided by normative considerations,
as core values are institutionalized into role
expectations and internalized into need dispositions.
On the macro-societal level of analysis on the
other hand, agents tend to disappear altogether
since Parsons conceptualizes society in terms
of four sub-systems, each consisting of a set
of institutionalized norms geared to the solution
of society's four functional problems (adaptation,
goal-achievement, integration and latency -- AGIL
for short).
Since each sub-system is further divided according
to the same AGIL logic into four sub-sub-systems,
social reality ultimately becomes a complex of
systems within systems within systems. In this
onion-like construct, macro actors such as interest
groups, social movements, etc., do not seem to
have a theoretically worked-out place.
In what follows, I shall argue that post-Parsonian
sociological theory took two directions, both
of which are on the whole negative.
(i) Those who have continued in the Parsonian
tradition (which, in parti pris fashion I shall
call sociological theory 'proper'), by over-reacting
to the ultra-systemic and reifying features of
structural-functionalism, have theorized action
and interaction in a manner which creates insurmountable
obstacles to the study of how micro situations,
on which they were exclusively focusing, are linked
up with macro-institutional structures and actors.
This lopsidedness -- which led to a systematic
neglect of social hierarchies and to an ever widening
rift between micro and macro sociology -- was
not corrected by relevant theoretical developments
outside interpretative micro sociology, such as
structuration theory and rational-choice analysis.
(ii) The other major reaction to Parsonian functionalism
(initially led by politically radical sociologists
incensed by Parsons' neglect of collective action
and struggles) turned its attention first to conflict
theory, then to Marxism and finally, after the
demise of Marxism in the 1980s, to post-structuralism/post-modernist
approaches to the social. Post-modernist theorists
reject, not only Parsons' conceptualization of
system and action, but also his attempt (quite
positive, as I see it) to establish sociological
theory as a specialized sub-discipline of sociology
-- i.e. as an intellectual activity portraying
a specific logic of analysis quite distinct from
that of philosophy or of theorizing in other,
neighboring disciplines.
This second group of critics, rejecting the crude
positivism and theoretical provincialism of Parsonian-influenced,
early postwar sociology, shifted their attention
to philosophical issues and/or theoretical developments
in fields such as linguistics, semiotics, psychoanalysis,
etc. The desire to break through entrenched barriers
and to broaden the horizons of a hitherto inward-looking
discipline was reflected in the very terminology,
sociological theory becoming social theory.
However, as I shall argue extensively below,
the broader outlook was not accompanied by any
theoretically coherent and systematic attempt
to translate insights derived from philosophy
and other disciplines into the framework of sociological
theory proper. Instead, sociological theorizing
gave way to discussions on ontological/epistemological
issues, and to attempts at reducing the study
of complex societies to that of language, signs,
texts, the unconscious, etc. The result of all
this has been a serious disconnection between
theory and empirical research, a disconnection
that is found neither in Parsonian sociology nor
in the micro-sociological writings of the first
group of critics.
All in all, I would say that the present malaise
that many sociologists sense vis-à-vis
'theory' is due to a dual failure: those who tried
to continue the research-oriented, specialized
theorizing that Parsons and his disciples had
initiated have, by dealing exclusively with Parsons'
underemphasize of micro interaction, neglected
macro actors and failed, therefore, to redress
the balance of system and action. In the process,
the already existing linkages between micro and
macro sociology were destroyed. Those, for their
part, who rejected both the specific framework
of Parsons' structural functionalism as well as
the kind of specialized sociological theorizing
of which Parsons himself was the main protagonist
have, for the most part, ended up with theorizations
that are neither good philosophy nor good sociological
theory -- i.e. theorizations that are of little
practical use to research-oriented sociologists.
What this means, of course, is an overall impoverishment
of both the empirical and the theoretical branches
of sociology.
The Impasses of Micro-Sociological Theorizing:
With the middle and late period of Parsons' oeuvre
marked by an under-emphasis of both micro and
macro actors, the various interpretative sociologies
that mushroomed from the 1960s onwards have almost
entirely focused on the former. This disregard
of macro actors has had grave consequences for
the further development of sociological theory
as such.
There are several explanations for this surprising
neglect by theorists whose avowed purpose had
been to redress Parsons' overly systemic analysis
and to 'bring back people' into sociological studies.
One obvious reason is the excessive fear of reification
that haunts hermeneutically and ethno methodologically
oriented sociologists. For them, any reference
to organizations of larger collectivities as having
goals, taking decisions, implementing policies,
etc., smacks of anthropomorphism, of endowing
collectivities with characteristics proper only
to individual human actors. Another reason that
partly explains the reluctance of interpretative
sociologists to deal with macro actors is their
populistic predilection for 'lay persons', 'ordinary'
members of society, 'mundane' encounters. As a
result they ignore not only collective actors
but also what, for convenience, one may call mega
actors -- i.e. individual actors in control of
considerable resources, whose decisions stretch
widely in space and time. In other words, micro
sociologists tend to forget that actors, because
of their very unequal access to the economic,
political and cultural means of production, contribute
unequally to the construction of social reality.
Attempting, however, to explain the symbolic construction
of social wholes by exclusive reference to 'lay
persons' or 'ordinary' members is like trying
to account for the construction of a complex edifice
by reference only to bricklayers, completely ignoring
the contribution of architects, managers, foremen,
accountants, lawyers, etc. (Mouzelis 1991a: 67).
A third reason for the neglect of macro actors
in interpretative sociology is the all-pervasive
association of agency with the micro and institutional
structures with the macro level of analysis. This
association makes itself felt particularly strongly
in respect of face-to-face interactions which,
as a matter of course, are invariably considered
as micro phenomena, as the building blocks out
of which macro-institutional orders are constructed.
Whether one looks at the writings of Garfinkel,
Cicourel, Goffman or their numerous disciples,
one always comes up against the ubiquitous idea
that to study face-to-face interaction is to study
micro-phenomena; and given this, the problem becomes
one of linking face-to-face encounters on the
micro level with institutional structures on the
macro level.
But linking face-to-face interaction with the
micro level becomes absurd when one considers
that face-to-face interactions are not the privilege
of 'ordinary' members of society but may occur
between 'non-ordinary' ones as well. To give an
obvious example: the face-to-face encounter between
Churchill, Roosevelt and Stalin at Yalta in 1945
led to crucial decisions which, among other things,
shaped the postwar map of Europe and profoundly
affected the lives of millions of people. In what
sense can this interaction between these three
men be seen as a micro event?
Another important point that must be stressed
here is that the identification of interaction
with micro, and of institutions with macro, leads
to an underemphasize of social hierarchies, of
the fact that institutionalized positions/roles
and actors are often related not only horizontally
but also vertically. So, for instance, social
actors being part of a multiplicity of hierarchically
organized wholes (corporations, trade unions,
political parties, etc.) deal routinely with other
actors hierarchically both subordinate and superordinate.
(The same can be said, of course, when the focus
is not on actors and interactions but on hierarchically
organized positions.) In view of the above, one
cannot consider the passage from the micro to
the meso and macro levels of analysis without
seriously taking into account how modern societies
are hierarchically organized, and particularly
how micro actors are hierarchically related (through
formal organizations or otherwise) to meso and
macro actors.
Micro sociology ignores all of this, insisting
that actors and face-to-face interactions belong
to the micro, and institutional structures to
the macro level. This absurd but strongly rooted
misconception precludes any study of how micro,
meso and macro actors relate to each other within
specific hierarchically-organized institutional
contexts. It is not at all surprising, therefore,
that the present debate on the links between micro
and macro sociology, as well as the older, related
debate on methodological individualism versus
holism, have led precisely nowhere. Not only do
neither of these debates theories the concept
of social hierarchies, but they also both leave
it out of account altogether as though it did
not exist. Trying, however, to investigate how
micro actors and encounters relate to the constitution,
reproduction and transformation of large collectivities,
without taking into serious account the hierarchical
aspects of social wholes, is like trying to swim
in a pool that has been drained of water.
The above point can be made more concrete by
a brief examination of some recent attempts at
bridge-building between micro and macro levels
of analysis. Randall Collins (1981), for instance,
taking into consideration the current emphasis
on intersubjectivity, argues that the basic unit
of analysis in sociology is not the individual
but the face-to-face micro encounter. What, in
fact, can provide micro foundations of large-scale
social phenomena is not methodological individualism
but 'methodological intuitionalism' -- an approach
that focuses on micro events or encounters, on
individuals interacting within specific micro
situations.
Macro phenomena for Collins 'are made up of aggregations
and repetitions of many similar micro-events'
(1981b: 988), and what we usually call social
structure is 'nothing more than large numbers
of micro-encounters repeated (or sometimes changing)
over time and actors' space' (1987: 195).
Now once one realizes that a single face-to-face
encounter can be a macro event (in the sense that
it can have an impact stretching very widely in
time/space), then the notion that we can move
from micro to macro through aggregative means
is seen to be not always valid. Very often macro
events relate to micro ones via a logic of subsumption
rather than aggregation. This is so particularly
when we focus on hierarchically organized configurational
rather than aggregative wholes.
If Collins tries to constitute macro phenomena
exclusively via aggregative procedures, Knorr-Cetina
(1981, 1988) constitutes them via a logic of representation:
macro phenomena are nothing more than 'summary
representations' -- such representations being
created by lay persons or theorists who try to
make sense of micro events entailing interacting
individuals in myriad micro situations.
For instance,
references to such macro phenomena as the state
or the world market, or to such macro events as
wars, are merely second-order theoretical constructs,
i.e. they exist only in the minds of lay persons
or theorists as rough summaries of a multitude
of micro realities. Again, what is misleading
in the above ethno methodologically-based attempt
at providing linkages between micro and macro
sociology is the idea that all face-to-face encounters
are always micro events. Given that such encounters
are the building blocks of all social reality,
macro phenomena are denied existence as first-order
constructs, and are admitted only as 'representations'
that ordinary people or theorists construct. This
nominalist view of macro phenomena can be easily
refuted by pointing out that face-to-face encounters
are not necessarily micro, that they can constitute
macro events in their own right.
To conclude: in so far as micro sociologists keep
ignoring social hierarchies linking micro, meso,
and macro actors, and view macro phenomena as
mere additions or representations of micro events,
they do not provide micro foundations at all,
but simply a false, reductive picture of the social
world.
It must at last be realized that no amount of
theorizing, or rather philosophizing, about the
ontology of the subject and the nature of interaction
can help to bridge the micro-macro gap as long
as human interaction continues to be conceptualized
in a hierarchical vacuum. The micro-macro issue
can be solved properly only by taking seriously
into account the hierarchical and onion-like system-within-system
nature of all complexes, differentiated social
wholes. The Micro-Macro and the Agency Structure
Issues: Other Attempts at Synthesis
In the previous section I have examined some of
the reasons why interpretatively oriented sociologies,
in trying to deal with the problems created by
the over systemic character of Parsonian theory,
have failed to provide effective bridges between
micro and macro sociology. In this section I shall
refer to some other well known attempts to tackle
the action-system and/or the micro-macro issues,
attempts that derive from quite different theoretical
traditions: to Giddens' structuration theory,
Habermas' theory of communicative action, rational-choice
theory as well as neo-functionalist efforts to
restructure Parsons' theory.
Since I have dealt with most of these approaches
elsewhere (Mouzelis 1991a) and due to space limitations,
I would simply state rather than demonstrate their
problematic character as far as the macro-micro
and agency/structure problems are concerned.
Starting with Giddens, I would like, in the first
place, to point out that Giddens, unlike most
theoretically oriented social theorists refused
to jump into the Marxist/Althusserian bandwagon
in the 60s and 70s and in the Foucault-Derridean
one in the 80s. Although sympathetic to both traditions,
he tried to translate a variety of insights derived
from these and other theoretical traditions into
a sociologically relevant vocabulary. In that
sense (despite his preference for labeling his
theory social rather than sociological) he can
rightly be seen as the major heir of Parsons'
mantle.
On the other hand, Giddens' structuration theory,
as an attempt, among other things, to bring closer
together interpretative and structural sociologies,
presents serious shortcomings. In so far as it
is based on the duality of structure notion, it
is unable to account for cases where actors do
not merely use structures (i.e. for Giddens' rules
and resources) in a taken-for-granted fashion
-- but rather adopt a strategic distance from
them in order to understand them or assess the
chances of their maintenance / transformation.
In such cases one needs a concept of subject/object
dualism rather than duality, for it is only on
the basis of such a concept that one can explain
how and why actors do not merely reproduce but
also can radically transform rules and resources.
Moreover, Giddens, while rejecting the conventional
distinction between agency and structure, is obliged
to bring it in by the back door by proposing the
distinction between institutional analysis and
analysis in terms of strategic conduct (1984:
289).
Given this, it is not surprising that his rejection
of functionalist explanations is merely rhetorical
-- in the sense that he simply avoids functionalist
terminology while retaining the logic of non-teleological,
functionalist explanations (Giddens 1982, ch.
2). Finally, Giddens' attempt to replace the micro-macro
distinction with the social-system integration
one (1984: 64-68) leads him erroneously to identify
face-to-face interactions with actions whose consequences
do not stretch widely in time/space (Mouzelis
1991a: 31-34).
As far as Habermas' theory of communicative rationality
is concerned, his rather wholesale adoption of
Parsons' over systemic AGIL scheme leads him to
an underemphasize of collective actors; and this
in turn seriously undermines his ambitious attempt
to integrate systemic and life-world approaches.
For instance, when Habermas talks about the colonization
of the life-world by the system in modern societies
(by the penetration of systemic, steering media
into the spheres of kinship and public-opinion
formation, spheres where it is communicative rather
than instrumental rationality that should be dominant)
he never shows how this process of 'technicization'
or colonization takes place. Like Parsons, he
rarely asks who-questions. Apart from vague references
to 'new social movements', we are left completely
in the dark as to the groups or interest that
resist, and those that promote this process. Habermas
goes hardly at all into questions of how macro
actors control the major means of economic, political
and cultural production, and how they contribute
to the colonizing process. As a result, his work
provides brilliant descriptions but no explanations
of macro-societal transformation.
Thus the colonization process, entailing the penetration
of the life-world by the system, is as opaque
as Parsons' notion of 'core values' controlling
everything via institutionalization and internalization
(Mouzelis 1991a: 187-189).
Coming now to theorists who try to revive Parsonian
functionalism by introducing concepts referring
to collective actors, class struggles, etc. (Alexander
1985; Alexander and Colomy 1990). As in Habermas'
case, such efforts have not been very successful.
Neo-functionalists have tried to inject a voluntaristic,
collective action dimension into Parsonian functionalism
and evolutionism -- but in so far as they accept
rather than radically change Parsons' AGIL schema,
the articulation between the analysis of institutional
structures and collective actors remains ad hoc.
For instance, Colomy (1985) tries to link processes
of uneven differentiation with the struggle of
strategic groups to bring about or resist institutional
differentiation. But the exact linkages between
institutional incompatibilities and collective
struggles remain unspecified.
Finally, rational choice theory as, among other
things, an attempt to bridge the micro-macro gap
by providing the micro foundations of macro phenomena,
faces the following dilemma: in so far as its
mainly logico-deductive theorizing refuses to
take into account the various socio-historical
contexts within which rationality takes its specific
forms, its statements (as all trans-historical,
universalistic statements) tend to be either wrong
or trivial. On the other hand, when rational choice
theory takes institutional context seriously into
account, it loses its distinctive profile and
its logico-deductive elegance (Elster 1986; Mouzelis
1991a: 146-158; 1993).
|