Journal of Classical Sociology
http://jcs.sagepub.com/content/8/1/109
The online version of this article can be found at:
DOI: 10.1177/1468795X07084697
2008 8: 109
Journal of Classical Sociology
Sandro Segre
Durkheim on Rationality
Published by:
http://www.sagepublications.com
can be found at:
Journal of Classical Sociology
Additional services and information for
http://jcs.sagepub.com/cgi/alerts
http://jcs.sagepub.com/subscriptions
http://www.sagepub.com/journalsReprints.nav
http://www.sagepub.com/journalsPermissions.nav
http://jcs.sagepub.com/content/8/1/109.refs.html
Durkheim on Rationality
SANDRO SEGRE
University of Genoa, Italy
ABSTRACT
The article explores Durkheim’s different notions of rationality, aims
to formulate a Durkheimian theory of social integration, which should be both
unitary and compatible with a particular version of Rational Choice Theory (RCT),
and to this end refers to the whole Durkheimian corpus. To pursue these theor-
etical goals, it recalls Durkheim’s fundamental distinction between an individual
and social state of consciousness, which may be related to two distinct conceptions
of rationality as may be found in RCT. The article, furthermore, explores the
societal preconditions and consequences which, according to Durkheim, charac-
terize each state of consciousness, especially insofar as solidary rules and norms are
concerned. A Durkheimian theory of social integration is then formulated, and its
relationship with some unresolved issues in RCT is discussed.
KEYWORDS
Durkheim, rationality, social state of consciousness
A few authors have maintained, though with different arguments, the existence of
a relationship between Durkheimian sociology, on the one hand, and Rational
Choice Theory (RCT), on the other. This article aims, first of all, to shed new light
on Durkheim’s concept of rationality. A second aim of this article is to formulate a
Durkheimian theory of social integration, which should be both unitary and com-
patible with a particular version of RCT. To this end, these authors’ conclusions
will be here considered, compared, and integrated with further references to
Durkheim’s works. A number of sociological concepts and theoretical questions
have been deemed of interest to both Durkheimian sociology and RCT. Instances
are: anomie, constraints, group solidarity, institutions, mechanical and organic soli-
darity, normative commitments, norms, obligations, sanctions, and social order
(Benard, 2004: 6, 8–9; Coleman, 1990: 241, 474, 613; Hechter, 1987: 3, 8, 15, 17,
18, 27, 33, 41, 56, 62–3, 105, 160; Heckathorn, 1993: 330; Macy and Skvoretz,
1998: 654). Yet, generally speaking, rational choice theorists and Durkheimian
Journal of Classical Sociology
Copyright © 2008 SAGE Publications Los Angeles, London, New Delhi and Singapore Vol 8(1): 109–144
DOI: 10.1177/1468795X07084697 www.sagepublications.com
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 109
scholars have not produced a debate on Durkheim’s contribution to RCT (see,
however, Benard, 2004). A further aim of this article is to show that the whole
Durkheimian corpus, with all the variety of its themes and theses, may be useful for
the purpose of formulating this theory. Most RCT sociologists have focused on
only a limited number of Durkheim’s works, such as those on the division of labor
in society, the rules of sociological method, and suicide (Durkheim, 1998b, 2002a,
2002b). Other works by Durkheim that may be relevant to RCT have received less
attention. Durkheim’s study of the elementary forms of social life (Durkheim,
1998a) is a case in point. Students of social movements who have shown consider-
ation for this work (see, for instance, Swidler, 1995) are not close to RCT, while
sociologists who are so have not dwelt on it, even though RCT has been applied
to this research area (Klandermans, 1997: 85, 203, 210).
The concept of rationality takes on quite different meanings for Durkheim
according to whether an individual or a social form of consciousness prevails.
Attention will be paid first to critical, and then positive, assessments of Durkheim’s
potential contribution to RCT. Both sets of assessments will be presented and evalu-
ated. A few theoretical statements which have been derived from these assessments
and bear on the conditions and consequences of a social state of consciousness will
then be formulated and commented on. A Durkheimian Theory of Social Integration,
which follows thereupon, specifies the network characteristics relevant for producing
social integration. Finally, some points of convergence are indicated between this
Durkheimian theory and a particular conception of Rational Choice Theory.
Rational Choice Theory and Durkheimian
Sociology: Critical and Positive Assessments
The following criticisms are noteworthy:
•
Durkheim focused on expressive, rather than purposive, actions (Coleman,
1990: 15).
•
Durkheim’s thesis that ‘in general an increase in the opportunities or goods
offered to individuals will in every case bring about [a] . . . higher degree
of dissatisfaction’ is neither tested, nor necessarily true. Rather, dissatisfac-
tion – as indicated by frustration and envy – would be ‘a response to par-
ticular situations generated by definite structures of interdependence’
between competing individuals (Boudon, 1981: 274–5, 277, 281;
Coleman, 1990: 474).
•
By considering the properties of the social system as the only determinants
of action, Durkheim neglected to study systemic properties as the intended
and unintended consequences of action. Accordingly, he did not conduct
an analysis of the rules, norms, constraints, and goals generating purposive
actions, which combine and interact to produce functioning social systems
(Coleman, 1996: 348; Lindenberg, 1996: 302).
•
Durkheim rejected ‘rational theories of solidarity in favor of normative
ones’. In particular, he maintained that public sanctioning ‘reinforces the
JOURNAL OF CLASSICAL SOCIOLOGY VOL 8(1)
110
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 110
group’s commitment to the violated norm’, rather than arguing, in line with
RCT, that it ‘deters potential deviants from future noncompliance’
(Hechter, 1987: 27, 160).
•
Durkheim failed to indicate ‘where social facts come from’, as he did not
locate ‘the source of individual preferences and the means by which individu-
als act upon those preferences within social facts’. As a consequence, his expla-
nation of collective behavior ‘does not offer clear causal mechanisms’ (Benard,
2004: 8, 9).
•
Durkheim neglected to account for variations in the social obligations
flowing from allegiance to different religions, even though RCT may pro-
vide a satisfactory explanation (Hechter, 1987: 56–7).
All these criticisms will be considered in the following section in the same order in which
they have been here presented. Other assessments have instead underlined the useful-
ness for RCT of some Durkheimian concepts and theoretical propositions. Assuming
rationality on the part of social actors, a number of remarks have been made, as follows:
•
A relatively small density of social relations – Durkheim’s moral density –
produces, along with individualism, greater costs in enforcing both pro-
scriptive and prescriptive norms (Coleman, 1990: 814–15).
•
The Durkheimian emphasis on norms is appropriate from a RCT viewpoint
when these ‘enable individuals to make decisions about a host of matters
that defy rational solution’. Among such matters are those circumstances in
which individuals engage in ritual and religious behavior, or exercise
coercion to produce a collective good. In such cases, behavior ‘cannot be
regarded as the outcome of benefit/cost calculation’, and rational egoists
would not therefore perform it (Hechter, 1987: 41, 63, 105).
•
While group homogeneity fosters social solidarity, as in traditional soci-
eties, group heterogeneity prevailing in industrialized societies does not
necessarily have this effect (Heckathorn, 1993: 330).
•
An extended ‘state of anomie’ will follow if actors are not able to preserve or
restore trust. Anomie is prevented if they can maintain in their relations either
a parochial solidarity (cooperating between neighbors only) or a universal
solidarity (cooperating also with strangers). If, on the other hand, some actors
cooperate with neighbors while pretending to be trustworthy with strangers,
their competitive advantage would encourage imitation, so that eventually no
one would cooperate with strangers (Macy and Skvoretz, 1998: 654).
•
A modest level of opposition to public goods, when conducted by deviant actors
and the public goods are desired by non-deviant actors, reduces the latter’s
problems of coordination and free riding, and therefore facilitates their attain-
ment of the critical mass and, ultimately, of the public goods. Too much oppo-
sition, on the other hand, prevents non-deviant actors from exercising social
control, which is necessary to attain the desired public goods (Benard, 2004).
These different and conflicting assessments, which concern Durkheim’s concep-
tion of rationality and its compatibility with Rational Choice Theory, suggest the
opportunity for further investigating this issue.
SEGRE DURKHEIM ON RATIONALITY
111
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 111
Durkheimian Sociology and Rational Choice
Theory: A Reappraisal
A number of interrelated and ordered theoretical statements constitute a unitary
theoretical framework (Turner, 1982: 6–7). Reference will be made here to state-
ments derived from criticisms found in the secondary literature. Their accuracy, as
indicated by consistency with relevant Durkheimian texts, will be assessed, starting
with the contention that Durkheim focused on expressive, rather than purposive,
actions. In fact, Durkheim considered expressive actions (in the broad sense of non-
purposive actions) as relevant to psychology, rather than to sociology, for individual
conditions only may account for them (Durkheim, 2002b: 14–15). Within the
purview of sociology are, on the other hand, both the individual and the social con-
sciousness. Of these two forms of consciousness (‘deux consciences’) (Durkheim,
1998b: 99), the former characterizes those strictly personal elements of conscious-
ness that make an individual different from anyone else, whereas the latter is shared
by all members of a social group, and consists of a set of moral and religious beliefs
and practices. Both forms of consciousness, the individual and the social, may be
found in every person (Durkheim, 1998a: 389; 1998b: 99–101; 2003: 51, 102),
but their investigation is not psychological. Psychological investigations would
prove exceedingly difficult, as Durkheim contended, since individual intentions are
too unfathomable, and human volition is too complex, to be ascertained by exter-
nal observers (Durkheim, 2002b: 4, 144). Still, the question of whether the indi-
viduals pursue private ends or the public good is amenable to sociological inquiry if
attention is paid to the specific traits characterizing the social types of the purely self-
ish and purely altruistic person (Durkheim, 2002a: 78–81; 2002b: 243, 326–7).
Both forms of consciousness may be called rational, but in a different way.
In societies with advanced division of social labor, the only one in which individual
personalities may develop (Durkheim, 1998b: 275–276, 339, 399), the individual
form of consciousness involves a person’s unrestrained self-absorption. In other
words, the person is willing and free to pursue a maximum benefit in social rela-
tions within his or her own sphere of action. Economic bargaining with a business
partner is a proper instance of such a relationship. Selfishness (égoïsme), as a state
of consciousness, involves deliberation and calculation when directing conduct
toward others, and therefore a non-cooperative form of rationality. By way of
contrast, the disregard of considerations of selfish advantages, the prevalence of
morality, hence an orientation to the public good and, therefore, to the social inter-
est, connote the social state of consciousness. This state of consciousness is rational
from society’s viewpoint. It flows as an emergent effect from the interaction
between, and association of, the actors’ own states of consciousness (Durkheim,
1997: 137–8; 1998a: 296–7, 453, 497, 525, 530, 633–8; 1998b: 101, 175,
180–1, 191, 264, 339–2, 401–3; 2002a: 120–2; 2002b: 314, 325, 350, 406–7).
The predominance of an individual state of consciousness engenders the
excessive development of individualism. Individual differences are enhanced, and
JOURNAL OF CLASSICAL SOCIOLOGY VOL 8(1)
112
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 112
a critical attitude is promoted, ultimately making the existence of a harmonious
social organization impossible (Durkheim, 1997: 127). Opportunities to obtain
coveted benefits are unequally distributed, producing widespread dissatisfaction.
This is so, according to Durkheim, because social heterogeneity in excess of the
amount involved by the division of labor weakens social cohesion (Durkheim,
2003: 50). For instance, contractual commitments violate retributive justice if
they are made because of constraint, as some parties are in a position to exert
power over the others. This is also the case when inequalities are due not to merit
or services rendered to society, but rather to accidental circumstances such as
inheritance. Since distribution of economic opportunities does not correspond to
the distribution of merits, distributive justice is violated (Durkheim, 1997:
210–11, 225–7, 236– 42; 1998b: 189–93).
Assuming instead the prevalence of a social state of consciousness, an
increase in the opportunities or goods offered to individuals will cause satisfaction
rather than dissatisfaction. Their satisfaction flows from a sentiment of solidarity.
Inequalities based on a differential distribution of merits do not endanger this sen-
timent as long as human sympathy confers no particular worth to accidental qual-
ities, such as superior intelligence or moral quality. Individuals competing to obtain
the best conditions in contractual relations seek to find a compromise and co-
operate to this end. For they realize that the legal framework of the contract, which
enforces the compromise, serves a public purpose by benefiting all the involved
parties. Individuals must be made aware, accordingly, of the existence of a common
interest overriding their selfish concerns. Ideas and sentiments that maintain and
promote social solidarity correspond to the needs of contemporary societies, in which
solidarity flows from the division of social labor as long as economic exchanges are
conducive to social or moral sentiments. The task of education is to express and
cultivate in the individuals those ideas and sentiments.
If the social state of consciousness prevails, rationality as a social condition is
evidenced both at the social and at the individual levels. At the social level, society
is able to curb overly strong or overly weak reactions to immoral conduct, and to
control individuals’ and groups’ selfish interests, so that the public interest may be
preserved. At the individual level, everyone is in a position to understand by means
of reflection that voluntary submission to social constraints is necessary to follow
society’s moral ends, and that the cooperative attitude it involves on the individuals’
part is ultimately in their own interest. Individual ends are in this case – and in this
case only – compatible with the public interest. In organized societies, morality is
rational, for it directs the individuals’ work toward those occupations they can best
perform, and sees to it that they receive just pay in return. A cooperative attitude,
which presupposes an already existing social and therefore moral milieu, is rational
in that it restrains the pursuit of selfish wants. If unchecked, this pursuit would pre-
vent the attainment of the public interest, and therefore the enjoyment of a full and
satisfactory life. It is apparent, then, that Durkheim gave careful consideration to
purposive actions, and distinguished their consequences according to whether or
SEGRE DURKHEIM ON RATIONALITY
113
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 113
not social actors hold a cooperative attitude (Durkheim, 1997: 151, 243–4; 1998b:
68–70, 192, 195, 262–6, 404; 2002a: 112–13, 122; 2002b: 429; 2003: 110–12).
As Durkheim maintained, an increase in the opportunities or goods offered
to individuals would not necessarily cause more widespread and/or pronounced
dissatisfaction. Dissatisfaction occurs when a selfish state of consciousness predom-
inates and, as a consequence, there is general noncompliance with the principles of
retributive and distributive justice. Individuals competing unsuccessfully for scarce
social goods would not be acquiescent, but rather would question the legitimacy
of social norms that provide benefits according to differences in power or luck,
rather than in merit. In contrast, to the extent that a differential distribution of
social benefits among interdependent actors is considered deserved, inequality is
accepted, and the solidarity flowing from contractual relations is undiminished. In
keeping with Durkheim, actions by interdependent actors generate emergent and
potentially undesirable effects, such as anomie. Systems of interdependent actions
constrain, if not determine, the actors’ conduct, but purposive actors in turn col-
lectively modify the structure of their interdependence, and thereby the social sys-
tem (on Durkheim as an ‘emergent theorist’, see Sawyer, 2002).
By distinguishing between individual and social forms of consciousness,
Durkheim maintained that whether the former or the latter predominate involves
a choice of different ends on the part of the social actors. Distinct consequences, desir-
able or undesirable, ensue. If the individual form of consciousness prevails, norms are
not intimately accepted, and therefore provide no direction to the actors’ ends.
Actors are then free from moral constraints. If, on the contrary, the social form of
consciousness is prevalent, solidary norms and rules harmonize and constrain the
individuals’ ends and interests, and therefore their purposive actions. This is so even
though the ever-greater possibilities of exchanges and communication have expanded
the diversity of individual ideas, beliefs, ambitions, and interests of individuals and
groups, in correspondence with the extended scope of civilization. The increase in
individuals’ rationality corresponds to an increase in the rationality of collective con-
sciousness. Its general principles have become the object of reflection and public
discussion, which is conducted by individuals as members of groups and citizens of
the state to the benefit of society’s moral life (Durkheim, 1997: 46, 122–9, 137– 40;
1998a: 623; 1998b: 275–6; 2002a: 113, 122; 2002b: 434; 2003: 120).
A Durkheimian theory of rationality is normative. From society’s vantage
point, solidarity is rational only if based on norms constitutive of the collective
consciousness that are agreed upon by all those who are concerned, and result
from their deliberations. This is evidenced, as Durkheim maintains, by the social
reaction to crime. As crime violates collective sentiments embedded in the social
form of consciousness, its punishment is in proportion to the seriousness of the
violation and the intensity of the social reaction. Therefore, public sanctioning is
rationally meted out (Durkheim, 1998b: 68–78; 2002a: 64–75; 2002b: 405–11).
Public sanctioning implies the predominance of a social form of consciousness, for
the violation of collective sentiments, rather than considerations of individual
JOURNAL OF CLASSICAL SOCIOLOGY VOL 8(1)
114
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 114
costs and benefits, is prominent in the individual consciousnesses. Deterring
potential deviants from future noncompliance with norms would imply, on the
contrary, individuals’ consideration of their own costs and benefits, and therefore
the prevalence of an individual form of consciousness.
Public sanctioning is a social fact in Durkheim’s sense. Like social facts in
general, it originates from the interaction, association, combination, and reciprocal
influences of many individual consciousnesses. It provides an instance of those
actions, thoughts, beliefs, and sentiments that are external to the individuals, but are
able to impose themselves by the coercion that they exert on every single person
(Durkheim, 2002a: 5). To the extent that social contacts are possible, and do in fact
occur with the advancement of the division of social labor, the increase in the mater-
ial and moral density of society corresponds to an ever-greater number of social facts.
Shared sentiments, common ideas, beliefs and practices, mutual understanding, an
impersonal moral life, collective actions and representations, and in particular cir-
cumstances a condition of collective excitement or effervescence follow therefrom.
In Rawls’s words, ‘myths and practices are both hearable and observable,
told and done, in the assembled group’. Practices, being ‘enacted’ as rites, can be
investigated separately from beliefs (Rawls, 2004: 40–1). By exerting the same effect
on the individual minds, sentiments, and emotions from which they originate and
which they presuppose, these ritual processes merge them, bring them into res-
onance, and promote their development. In turn, society benefits from individuals’
development. Collective representations, and the emotions they embody, are
considered by Durkheim necessary to social communication, and, more generally,
to society’s existence (Durkheim, 1997: 137; 1998a: 326–7, 547, 571–2, 603,
605, 623, 633; 1998b: 46, 64–5, 74, 238, 264, 339–42; 2002a: xvii, 5–7, 101–6,
112–13; 2002b: 173, 213–14, 224; 2003: 56–8, 73; on Durkheim’s notion of
collective effervescence and collective representations, see Alberoni, 1981: 123–7;
Rawls, 2004: 170–1, 178–80; Shilling and Mellor, 1998).
Accordingly, Durkheim did not neglect to specify the mechanisms that
may account for social facts such as norms and sanctions. Mechanisms are ‘state-
ments that purport to explain an event’, in the sense of giving an account of why
this event ‘happened as it happened’ (Elster, 1989a: 4, 6). Explanatory mech-
anisms provide detailed, ‘fine-grained’ descriptions of the relationship between
cause and effect, are applied to ‘a limited range of phenomena’, and specify which
actors are involved and how (Hedstroem and Swedberg, 1996: 298–9; see also
Elster, 1989a: 168–9; Hedstroem, 2005: Ch. 2). As conversational analysis has
shown, the ‘levels of detail actions and practices’ (that is, levels of ‘granularity’)
which actors select in the production of their daily lives are relevant for the inves-
tigators’ own selection of their observational level (Schlegloff, 2000: 718–19).
As for Durkheim, ‘social mechanisms’ would specify, in particular, ‘the
conditions that would allow the division of labor to generate organic solidarity’
(Mackert, 2004: 315). Durkheim’s explanations of social facts draw attention
to the social and cultural consequences, or ‘events’ with specific characteristics,
SEGRE DURKHEIM ON RATIONALITY
115
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 115
which flow from direct interactions between identifiable actors in specific settings.
These interactions are then the sources of social facts. The actors’ mutual influ-
ence, and thereby their possibility to collectively construct social phenomena such
as a marriage, religious ceremony, or professional group, depend on their number,
proximity, possibility of communicating, and socialization (Durkheim, 1997: 47;
1998a: 493, 571–72; 1998b: 341–2; 2002a: 103, 115; 2003: 72–3).
An explanatory mechanism was used by Durkheim to account for different
social obligations flowing from allegiance to distinct religions. As mentioned, an orien-
tation to the public good or the social interest connotes a social state of conscious-
ness. Accordingly, individual ideas, sentiments, beliefs, attitudes, and actions are
rational to the extent that they conform to this orientation. In particular, social obli-
gations involved by particular religions are rational if they work as moral forces that
bring the sentiments and ideas of the individual followers into agreement with those
prevailing in society (Durkheim, 1998a: 461, 571). The constitutive elements of a
particular religion, such as a given set of beliefs and rites, therefore have little rele-
vance, in line with Durkheim’s endeavor to provide a definition of a given social phe-
nomenon – religion, in particular – in which its permanent aspects receive emphasis
(Isambert, 1982: 187). What matters, as Durkheim put it, is that these elements ‘be
suitable to sustain a collective life of a sufficient intensity’ (‘qu’ils soient de nature à
alimenter une vie collective d’une suffisante intensité’) (Durkheim, 2002b: 173). Such
a condition is met when religions are able by means of ceremonies to assemble indi-
viduals, put them into direct contact with each other, and bring forth the same ideas
or sentiments amplified by their reciprocal influences (Durkheim, 2002b: 493, 572).
To the extent that they can produce these socially notable effects, whether of
a benevolent or malevolent nature (Shilling and Mellor, 1998: 206), religions are
rational in Durkheim’s sense. If religion is considered a component of the social
structure, Durkheim posits a two-way relationship between actors and structure:
‘The structure may change actors, their hearts and minds. The actors may change the
structure, its positions and possibilities’ (Hernes, 1998: 98). This two-way relation-
ship may be translated into different social mechanisms, according to whether the
actors and the structure change or not. In particular, structural changes may promote
collective action by enhancing the ‘potential for communication and organization’,
and thereby the possibility to overcome the free-rider problem (Hernes, 1998:
98–9). If advanced division of labor prevails, ceremonies (religious or otherwise) may
sustain a collective state of consciousness to the extent that they make communica-
tion possible between different groups, and also between individuals who belong to
different groups. If this is the case, ceremonies, too, are rational in Durkheim’s sense.
Durkheim on Norms and Rationality
As Durkheim maintained, the enforcement of social norms depends – other con-
ditions being equal – on the moral or dynamic density of social relations, namely, the
total number of solidary relations in a given population. Economic relations do not,
JOURNAL OF CLASSICAL SOCIOLOGY VOL 8(1)
116
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 116
accordingly, necessarily have a moral nature (Durkheim, 1998b: 238–9, 266–7;
2002a: 112–13), for the greater is the density, the lower are the enforcement costs
(Durkheim, 2003: 50–1). However, solidarity is predicated: (1) not only on the
density of social relations, but also on their egoistic or altruistic character; (2) not
only on relations between individuals, but also on relations between groups. An
egoistic group is opposed to an altruistic one, much as single individuals’ conduct
may be oriented toward either exclusively personal sentiments, beliefs and images of
reality, or the collective consciousness, which is then connoted by the extension and
articulation of the individual’s consciousness (Durkheim, 1998b: 64–5, 174–5,
339). As the collective consciousness becomes more comprehensive, it may accord-
ingly accommodate a greater variety of divergent groups, each with its own con-
sciousness and norms. As mentioned, from Durkheim’s point of view, norms are not
rational in the sense of necessarily corresponding to rationally pursued individual
ends, or bringing distinctive benefits. Rather, norms are rational from a societal
point of view. In this sense, behavior on the part of individuals is rational because it
is subservient to society’s general ends, as indicated by and pursued through social
norms. Individual ends which do not conform to norms are irrational from society’s
viewpoint. There are not, in this sense, rational egoists (Durkheim, 1998a: 245,
296–7; 1998b: 37, 74, 276, 401; compare with Tole, 1993: 22–3).
To the extent that the particularistic or segmental groups gradually fade
away, as occurs with territorial distinctions, so also in the long run does their indi-
vidualistic consciousness. On the other hand, as shown by the professional groups’
moral milieu, every group partaking of the division of social labor enmeshed in a
web of cooperative relations with other groups is altruistic. Professional groups con-
stituting a moral milieu preserve their own ethic, rights, and therefore identity. They
gain importance to the extent that the land-based groups fade away, for they are a
part of a larger and encompassing group – society or the state – that exerts a restrain-
ing and regulatory influence on the collective consciousness, and therefore on the
selfish tendencies of its constituent, well-differentiated parts. Groups representative
of particular interests, or ‘secondary groups’, are in turn necessary to counterbal-
ance state action, which could otherwise be equally oppressive (Durkheim, 1997:
51–2, 67, 98–9, 125–30, 137–8; 1998b: 163–6, 189, 194–5, 199–200, 273–6,
283, 286, 289–90, 401; 2002a: 115; 2002b: 439; 2003: 60, 72–3).
Still, individuals are ever more differentiated in terms of their conscious-
ness and social functions, and therefore less amenable to adaptation to their social
milieu by means of general rules. Social heterogeneity, which results from its
advancement, is therefore less and less conducive to a moral condition capable of
providing guidance in all aspects of individuals’ conducts. Collective conscious-
ness may not assert itself in this case, and its rationality may yield before the
encroachments of individuals’ selfish rationality. However, social solidarity, and
therefore altruism, moral rules, and collective consciousness are necessary to the
existence of any society, no matter how rudimentary or advanced is the division of
social labor. Eventually, therefore, a breakdown of social solidarity would bring
SEGRE DURKHEIM ON RATIONALITY
117
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 117
about the end of society (Durkheim, 1998b: 140–1, 174, 207–8, 267–8, 272–6,
283, 319, 339). Solidarity in societies with an advanced division of labor is based
on contracts, and therefore on the ultimate control of social relations on the part
of the state as the representative of society (Durkheim, 1998b: 188–9).
Social cooperation is accordingly predicated on norms shared by particu-
lar groups, and also by society as a whole. Social development involves increasing
communication between these groups, whose boundaries become more perme-
able. Their individual character becomes ever less distinct and more dependent on
society as the most encompassing group. According to Durkheim, professional
groups are a case in point. Each group forms a moral milieu if it has attitudes, spe-
cial forms of knowledge, ideas, usages, and a mentality of its own, as may be the
case for professional groups (Durkheim, 1997: 67; 2003: 95). To the extent that
a professional or occupational group constitutes a moral milieu, it embodies, epit-
omizes, and regulates its members’ rights and duties, in addition to shared life and
interests. All members are constantly reminded of the moral significance of their
activities for their group, and, indirectly, for society as a whole. A community of
spirits, committed to the pursuit of the same goal, follows as a consequence. This
is so, however, provided that frequent and intimate contacts, and also shared ideas
and sentiments, relate the group’s members in such a way that the constitution of
the group is strong, and may therefore exert authority on the members’ con-
sciousness (Durkheim, 1997: 47, 55, 67; 1998b: xvii–xviii).
Insofar as groups are so characterized, relations are imbued with social sen-
timents. As a consequence, groups as well as their members may thrive without
causing any harm to society. If unregulated, however, all groups apart from the
state may pursue their own selfish interest to the potential detriment of society.
Accordingly, they may engage in perpetual conflicts, causing social fragmentation
and, ultimately, a condition of anarchy and anomie. As for the state, it may fail to
provide regulation to economic and social relations, in such a way that no special
interest would prevail or freely operate, and the negative effect of such a condition
would be counteracted. The recurrent conflicts between economic and occupational
groups, which are paralleled by a corresponding disarray and lack of unity in the sci-
ences, provide evidence that both society and knowledge lack organization and
regulation. Given these societal conditions, egoistic groups not only exert an
oppressive pressure on their members, as their specific identity is preserved by means
of social boundaries confining their members’ actions; they also pursue their own
interests to the detriment of the public interest. Individuals, moreover, are oblivious
to their own and the other workers’ contributions if the specialized character of their
tasks makes them unaware that they concur in a common enterprise (Durkheim,
1997: 54–5, 97–9; 1998b: ii–iii, vii, 163, 166–7, 264–7, 348–9, 356–66, 387, 401).
Cooperation with the members of a person’s own group – such as fellow
workers, colleagues, or scientists – and with others who belong to different groups
is necessary to avoid this condition of anomie. For anomie results from the inability
on the part of some overarching groups, like the state, to regulate social life for
JOURNAL OF CLASSICAL SOCIOLOGY VOL 8(1)
118
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 118
the remaining groups and their members. To this end, the superior importance of
solidarity and common interests over the interests of particular groups or individ-
uals must be impressed on them. Accordingly, interrelations must be cooperative,
rather than simply based on convenience. Different economic sectors in a nation
or different nations may exchange their products, and thereby maintain an eco-
nomic division of labor. And yet they bring a social division of labor into existence
only if they constitute a moral unity encompassing all the participants in an
exchange. Contacts between individuals or collectivities that are infrequent and
unregulated make conflicts between them recurrent, as relations between organ-
izations representative of labor and capital show. For sheer force is then the only
means to decide the winner, and neither the occupational groups nor society or
the state is in a position to embody and enforce authority. Authority must be not
only legal, but also moral. Those who enforce it as well as those who are subject
to it must, in other words, accept its prescriptions to render it effective
(Durkheim, 1998b: vii–viii, xi–xii, 262–6; 2002b: 439–40; 2003: 120).
Moral rules are a public good that benefits all society’s members. Provided
that solidarity prevails, their violation produces a sanction proportional to the
intensity of the sentiments that have been hurt: the stronger the sentiments which
have been violated and the more serious their violations, the stronger are the sanc-
tions. Sanctions must not be too weak lest they be insufficient, but not too strong
either lest they be inefficient. Effective sanctions correspond, then, to the serious-
ness of the violations. Violations are uniformly evaluated in their seriousness when
the collective consciousness is strong and well-defined, as is the case when the div-
ision of social labor is not advanced. However, in socially differentiated societies,
rules concern functions exerted by special groups and their violation concerns a
weaker state of the collective consciousness. Reactions are accordingly moderate
and proportionate to what is necessary to re-establish the affected functions
(Durkheim, 1998b: 69–70, 96–102, 124–6).
Durkheimian Sociology and Rational Choice
Theory
A number of theoretical statements will summarize this discussion concerning
Durkheim’s notion of rationality. Each statement will be presented and commented
on, as follows:
•
A social form of consciousness is rational from society’s viewpoint.
•
If a social form of consciousness prevails, a cooperative attitude restrains the
individuals’ and groups’ pursuit of selfish interests, and their competition in
contractual relations.
According to Durkheim, if a social form of consciousness prevails, solidary norms
and rules rather than power differences between individuals or groups constrain
SEGRE DURKHEIM ON RATIONALITY
119
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 119
individual actions and attitudes. The prevalence of norms, and therefore of a social
form of consciousness, presupposes an orientation to the public good. A public
good is by definition non-excludable (no one of a given set of actors can be ex-
cluded from its benefit) and non-rival (its use on the part of one or more actors does
not make it less available to other actors; see, for instance, Heckathorn, 2001: 277;
Kollock, 1998: 189). A cooperative attitude and behavior on the part of the indi-
viduals, which ensues from this orientation to the public good, characterizes – in the
language of game theory – a cooperative game, in which binding and enforceable
commitments are possible (Harsanyi, 1986: 92). As has previously been shown,
Durkheim contended that a cooperative attitude is conducive not only to the public
good of society as a whole, but in the long run also to the public good of societal
components such as groups, and of single individuals. In contrast, an individual
form of consciousness involves a ‘consideration of reciprocal useful or harmful
effects’ (‘considération relative aux effets utile ou nuisibles des uns et des autres’) of
actions, and therefore implies an ‘idea of deliberation and calculation’ (‘idée de
délibération et calcul’) (Durkheim, 1998a: 296–7), and is accordingly conducive to
a non-cooperative, two-person prisoner-dilemma type of game.
Durkheim’s position is worth discussing against the background provided
by statements and debates on the part of different RCT proponents. Some prelim-
inary notions may be in order. Actors who are connoted by an individual form of
consciousness and behave accordingly are rational in a ‘thin’ sense of the term
(Elster, 1983: 2–15; Hechter and Kanazawa, 1997: 194–5). Namely, their action
is instrumental to the effect that ‘it has to be explained by the actors’ will to reach
certain goals’ (Boudon, 1998b: 818). Rationality in this narrow sense implies opti-
mizing the relationship between an actor’s beliefs, desires, evidence, and action
(Elster, 1989b: 4). A number of assumptions, jointly considered, characterize RCT
if narrowly defined:
1.
Actors are concerned only with the consequences of their actions as they
see them.
2.
They have a set of complete and consistent preferences or desires over
alternative choices of actions, which they have reasonable grounds to
believe are available to them.
3.
Their beliefs are also consistent with themselves and with the available
evidence.
4.
Actors are completely informed as to the other actors’ preferences and strate-
gies.
5.
They are also completely informed as to the the feasible outcomes of their
own and the other actors’ actions.
6.
They pursue the goal to minimize the costs, and maximize the benefits,
which they derive from other actors and/or public goods.
7.
In order to do so, they may not contribute to a public good, and exploit
the contributions from other actors.
8.
They expect other actors to do likewise (Boudon, 2003: 3–4; Elster, 1986:
12–16; Heckathorn, 2006: 623).
JOURNAL OF CLASSICAL SOCIOLOGY VOL 8(1)
120
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 120
Rational action in this narrow sense is not the most convenient action if
benefits are considered in the long term, as shown by the well-known prisoner-
dilemma type of game. In this ‘game’ (in the technical sense of game theory), two
interdependent and rational actors who are concerned only with the consequences
of their actions as they see them, and do not communicate with each other, pursue
the goal to minimize costs and maximize benefits they may derive from their inter-
action. To this end they do not contribute to the public good which would benefit
them if they cooperated, but rather exploit contributions that may come from the
other actor. Failure to produce the public good, which follows as a consequence, is
disadvantageous to all actors. Therefore, ‘individually rational actions lead to a
collectively irrational outcome’, as each actor prefers maximization of individual
short-term benefits (‘unilateral defection’) even though ‘universal cooperation is
preferred to universal defection’ (Heckathorn, 2001: 277; 1996: 257). This is the
so-called the free-rider problem, on which the RCT literature has dwelt extensively
(see, for example, Boudon, 2003: 3–4; Heckathorn, 1996: 256–8; Kollock, 1998:
184–92). Its solution requires abandoning the assumption that actors selfishly
pursue an instrumental orientation to obtain short-run benefits. If the assumption of
an instrumental orientation on the part of the actors is maintained, a norm may be
defined as a socially defined right of control of some actor’s action, which is held by
other actors, and involves rewards for its observance and punishments for its violation
in such a way that compliance is in the actor’s interest (Coleman, 1990: 243).
The pursuit of individual interests is compatible with conformance to
norms only if the following stringent conditions obtain:
1.
Individual actors must have an interest in cooperating. They do so if they
gain substantially from the cooperation of everyone (universal cooperation);
their losses are limited when they cooperate while others do not (unilateral
cooperation); they have limited gains when they do not cooperate (unilateral
defection) while other actors do; they believe that all actors, including them-
selves, are interdependent, and expect reciprocity from the other actors.
2.
They must have an interest in exerting intragroup control to the benefit of
themselves and the other members, as all members are interdependent.
Intragroup control may be achieved by exerting on the other members
persuasion and/or control of their actions, in line with the group’s criteria
of dispensing sanctions and rewards. Their cost should not in any case
exceed their benefit. Over-cooperation, which would be collectively irra-
tional since providing an excess of incentive would be too costly in relation
to its benefits, may be prevented by exerting oppositional control so that
over-cooperation would be discouraged, but group cohesion would not be
weakened to the point of encouraging individual defection.
3.
Individual actors must be guided by a consideration of their long-term
interest, to which they must be willing to sacrifice their short-term inter-
est in order to make their threats and promises credible (Coleman, 1990:
273–8; Elster, 1989a: 132–3; 1995: 282– 4; Heckathorn, 1990; 1996:
271; Kollock, 1998: 198–200, 205–6).
SEGRE DURKHEIM ON RATIONALITY
121
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 121
Attitudes and modes of behavior are deemed rational by Durkheim only if
they conform to social norms prescribing cooperation. Such norms are group
norms (or ‘metarules’) in the sense of being shared in the group along with the
corresponding positive and negative sanctions (Lindenberg, 1996: 309–10). Social
solidarity based on norms, as Durkheim pointed out, is in the long run advanta-
geous not only to societies characterized by the division of social labor, but also
to the individuals partaking of it. From a Durkheimian viewpoint, there is there-
fore no necessary conflict between individual and collective orientations, or in-
compatibility between the private and public benefits that accrue from
norm-conformity. This Durkheimian view of rationality is, however, non-standard
because of its emphasis on a broad notion of this concept. Durkheim’s notion of
rationality directs attention to cases in which ‘actions, decisions, and beliefs are
meaningful to the actor in the sense that they are perceived . . . as grounded on
reasons’ which the actor considers strong and valid for him- or herself and others
(Boudon, 1990: 125–6; see also Boudon, 1998a: 184–92; 1998b: 826–7; 2003:
10–11; Hechter and Kanazawa, 1997: 194). Most RCT representatives would
accordingly either reject this Durkheimian position, or concur with it only if
rationality is defined broadly (or ‘thickly’), rather than ‘thinly’, insofar as con-
formance to norms is concerned. Many RCT authors are aware of difficulties inher-
ent to Coleman’s narrow or ‘thin’ notion of rationality, but they have generally
been inclined to qualify its validity rather than to discard the notion altogether.
In keeping with a ‘thin’ definition, RCT has pointed to some instances in
which actors have an interest in norm compliance. In the case of conjoint norms –
in which ‘the interests favoring observance of the norm and those opposing its
observance are contained within the same actors’ – an actor’s acceptance of norms
that are immediately disadvantageous from his or her point of view may well be
perceived as advantageous in the long run (Coleman, 1990: 247, 288). Norm
compliance in small groups provides a further instance, as shown by communities
characterized by network closure and an internal sanctioning network. Small
groups are, in this case, ‘a means by which collective power is made real and con-
sequential for individuals’ (Harrington and Fine, 2000: 314). The risk is, however,
of collectively irrational over-compliance (Coleman, 1990: 275–8, 494; 1996: 347–9).
Finally, cooperative games that take place between the same actors or sets of actors
make social norms credible. In this case, intragroup normative constraints make
collective sanctions enforceable (Elster, 1989a). Inner-group relations can there-
fore ‘bolster social control’ (Heckathorn, 1990: 377).
Research in line with RCT, but also with Durkheim, has shown that what-
ever advantage may accrue to individuals who follow their selfish preferences is
either short-lived and unlikely or unattainable. It has been maintained, in this con-
nection, that ‘the conditions under which people will cooperate out of self-interest
are quite stringent’ (Elster, 1989a: 132). If the selection of goals is dictated by
preferences or the pursuit of ‘utility’, ‘the conscious and deliberate attempt to
maximize utility tends to be self-defeating’, since ‘it can never . . . be brought
JOURNAL OF CLASSICAL SOCIOLOGY VOL 8(1)
122
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 122
about intelligently or intentionally’, but only as by-products of another action
(Elster, 1983: 9, 43). When private interests are pursued, this pursuit can be suc-
cessful only temporarily at most if moral norms regulating social interactions are
absent. In Durkheim’s words, ‘if interests make people closer to each other, they
do so only momentarily’ (‘si l’intérêt rapproche les hommes, ce n’est jamais que pour
quelques instants’) (Durkheim, 1998b: 180–1). RCT would concur with this
Durkheimian position for the following reasons.
Firstly, there is no prisoner dilemma for rational actors in a dyadic game if
the collective good is less valuable than its production cost (because defection
becomes rational also at the collective level), or more than twice this cost (because
actors have in this case a common interest in reaching an agreement, though an
opposed interest in allocating the common good between them) (Heckathorn,
1996: 258). Secondly, repeated two-person games are conducive to cooperation as
long as actors develop an ongoing relationship, can identify each other, and are able
to recall the other actor’s past conduct (Kollock, 1998: 195–6). Thirdly, RCT has
concurred with Durkheim in arguing that norms of reciprocity cannot produce
social solidarity. Reciprocity norms (which Durkheim called ‘mutualisme’) reward
cooperation and punish defection by reciprocating in kind. Durkheim maintained
that these norms are not sufficient to bind the partners tightly (Durkheim, 1998b:
266–7). RCT has argued along the same lines that social cooperation is fragile if
norms of reciprocity are the only norms that regulate the pursuit of private interest.
For no rational actor is willing to bear the cost of norm enforcement if there are no
prospects for future interaction with specific other actors, and the interaction
partners are chosen randomly.
If more than two actors are considered, authors familiar with RCT have taken
issue with the position, represented especially by Coleman, that social norms – such
as the norms of reciprocity and cooperation – owe their existence to the benefits they
bring to the actors (Turner, 2006: 497–8). A number of reasons account for their
disagreement with Coleman and, more generally, with a ‘thin’ notion of rationality.
In particular: the amount of benefits brought by a norm cannot be calculated as
could be the price of economic goods. Norms of cooperation, which are potentially
useful, may lead to over-cooperation in the absence of oppositional strategies, and
may therefore be inefficient if cooperation means contributing to the public good
and punishing those who do not; as research has shown, the strategy of ‘hypocrit-
ical cooperation’ (not contributing, but punishing others who also fail to contribute)
is quite robust in strategic interactions, but its very existence shows that the norm
of full cooperation has limited validity. Private cooperation – when actors contribute
to the collective good but make no attempt to induce others to contribute – is a
rational strategy only if the value of the collective good is so large that other actors
have no incentive to withhold their contribution (to ‘defect’). Finally, norms that
would be both socially useful and efficient may not exist, or may be beneficial to
some collectivities but detrimental to others (Elster, 1995: 147–9; Heckathorn,
1990: 382; 1996: 254, 265–7, 269, 271; Lindenberg, 1996: 307).
SEGRE DURKHEIM ON RATIONALITY
123
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 123
Accordingly, it has been authoritatively stated that ‘most cooperation is due
to non-selfish motivations of one kind or another’ (Elster, 1989a: 133). Conformity
to norms, hence displaying a cooperative attitude, is itself a norm, which may be
called a ‘metarule’ because of its generality (Lindenberg, 1996: 309–10). Precisely,
this metarule is distinctive of a social form of consciousness. Durkheim’s notion of
rationality is broader than the individualistic conception held by most RCT theorists
because of its emphasis on shared ends and norms of moral actions (Wallwork,
1972: 48). As Durkheim argued, the pleasure that often accompanies altruistic acts
cannot account for the pursuit of socially desirable ends (Wallwork, 1972: 157–8).
Social norms and rules are solidary if they are the objects of reflection and public
discussion. This Durkheimian position is reminiscent of, but does not coincide with,
those of other prominent contemporary authors, as will now be shown.
Habermas’s view of undistorted communication involves ‘the peculiarly
constraint-free force of the better argument’ (Habermas, 1984: 26), which is
agreed upon by all rational members of society. However, Durkheim did not con-
tend that: (a) reflection and public discussion should be made mandatory; (b) an
agreement would necessarily ensue therefrom, or, if this were the case, it would
be unanimous; (c) the public good would be necessarily reached even if some indi-
viduals or groups would pursue private ends; (d) interaction with other individ-
uals or groups would necessarily lead to a collective grasp of the public good, and
be in this sense rational; (e) unanimous agreement would necessarily result from
reflection and public discussion, and will always be incompatible with the pursuit
of private ends. All these objections have been raised from an RCT viewpoint
against this view by Habermas (Elster, 1983: 37– 42).
Rather, if the individuals, and also the private groups and public bodies which
represent them, maintain an exchange of sentiments and communication, then reflec-
tion and deliberation lead to the formation of clearer ideas and a social consciousness.
They are a necessary, though not sufficient, condition to reach the public good. In
contrast, public discussion and a community of language and attitudes are not suf-
ficient. Moreover, this Durkheimian position appears to be proximate to Mead’s con-
ception of rationality, consistent with Weber’s ‘value rationality’ (Weber, 1968: 24–6)
and Mannheim’s ‘substantive rationality’ (Mannheim, 1940: 53) and directing atten-
tion to a cognitive, rather than instrumental, notion of rationality as advocated by
Boudon, Elster, and others. A brief comparison between Durkheim, on the one hand,
and Weber, Mannheim, and Mead, on the other, may clarify Durkheim’s own notion
of rationality by pointing to similarities with and differences from these authors. It will
then be maintained that the Durkheimian broad conception of rationality relates to a
notion of rationality that is cognitive and axiological. Instrumental rationality may be,
however, compatible with a cooperative attitude and behavior.
Weber’s twofold conception of rationality, instrumental rationality, and ration-
ality with reference to values (Weber, 1968: 24–6), bears comparison to Durkheim’s
contrast between two orientations, to the obtainment of benefits disregarding
others and to the public good (see, for example, Durkheim, 1998a: 453).
JOURNAL OF CLASSICAL SOCIOLOGY VOL 8(1)
124
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 124
Durkheim’s former orientation designates an instrumentally rational type of conduct,
while the latter is an instance of a value-rational conduct. Moreover, the two
authors maintained that both types of rationality are present in modern market
exchanges. Durkheim laid emphasis upon the social origin of the solidary bond
connecting the parties pursuing the same occupation and engaged in a contractual
relationship, each of whom strives to rationally pursue his or her private interest.
Likewise, Weber laid emphasis on the relevance of ethical norms in economic trans-
actions. Within the bounds provided by these norms, partners’ behavior ‘is influ-
enced by the rational, purposeful pursuit of interest’ (Durkheim, 1997: 67; 1998b:
190–5; Weber, 1954: 192–3; see also 1973: 453). According to both authors,
finally, democratic parliamentary institutions are considered rational with reference
to values which Durkheim and Weber deemed highly relevant to modern society.
Parliamentary democracy presupposes on the part of the electorate ‘deliberation,
reflection, critical spirit’, in line with the prerequisites of rational morality which
characterize modern times (Durkheim, 1997: 123–4, 151). It presupposes ‘the
resolute will of a nation not to let herself be ruled like a flock of sheep’ (Weber,
1989: 270), rests on ‘the proud traditions of peoples which are politically mature
and free from cowardice’ (Weber, 1968: 1461), and is an indispensable instrument
to control the public administration, carry political decisions into effect, and keep
the emotional and irrational influences of the populace in check (Weber, 1956:
871–6; see also Roth, 1968: cv). For both authors, then, action is rational insofar
as it is consistent with the actors’ beliefs, notions, and values.
Like Durkheim and Weber, Mannheim also formulated a broad conception of
rationality (‘substantive rationality’, as he called it). This conception was distinguished
from a ‘thin’ one, which was designated as ‘functional rationality’ and concerns the
proper relation between ends and means. Like Durkheim, Mannheim advocated the
former type of rationality as a potential remedy against the crisis of liberalism and
democracy, which he considered a major predicament of modern society. ‘Substantive
rationality’, if implemented, would achieve values such as ‘identification with the
other members of society, collective responsibility, and the obligation following from
it to possess in common with the others a basis of attitudes and ways of behavior’.
Individuals would receive from ‘experience and science’ the particular education,
milieu, and tasks that may best promote their development and personality, as opti-
mally expressed in small groups. In contrast to Durkheim, however, who entrusted
individual and societal development to unplanned advancements in moral density and
reforms of the educational institutions, Mannheim proposed a coordinated and com-
prehensive set of political and social measures, which he called ‘planning for freedom’
(Mannheim, 1993: 540–1; see in this connection Kettler et al., 1984: 144–9).
Mead’s conception of rationality, which ‘may include all beings speaking the
same language’, refers to ‘a type of conduct in which the individual puts himself in
the attitude of the whole group to which he belongs’ (Mead, 1934: 334–5; see also
Joas, 1997: 140–1; 2001: 94–95). In line with Durkheim, Mead maintained that
individuals’ selves share a social aspect which produces in all of them feelings of
SEGRE DURKHEIM ON RATIONALITY
125
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 125
cooperation and, therefore, interdependence. In such a way the basis is provided
‘for the development and existence of ethical ideals’ in human society as well as
‘in each single individual self’ (Mead, 1934: 321). Both authors sought to over-
come the divide between individual conscience and society and laid emphasis on
mutual influences. Accordingly, they viewed moral education as a task that is
rational from the individuals’ and society’s viewpoint as conducive to their mutual
development (Mead, 1934: 264–5, 319–23; see also Nisbet, 1974: 111–13).
•
If a social form of consciousness prevails, a differential distribution of social
benefits among interdependent actors is considered deserved, and inequality
is therefore legitimate.
A differential distribution of social benefits is legitimate, according to Durkheim,
only if a social form of consciousness prevails. However, actions by interdependent
actors may have consequences – such as the division of labor and anomie – which
escape the actors’ control. Durkheim then argued that unintended and undesirable
consequences may follow from purposive (or strategic) actions, and may concern
both the actors themselves and other actors. Some modern social theorists, and
representatives of RCT in particular, have also called attention to the undesirable
effects of rational actions (Boudon, 1981: 281; 1990: 119, 125; Coleman, 1990:
Ch. 22; Elster, 1984: 18; 2000a: 162–5). They may result, as Merton pointed out
long ago, from a variety of causes, such as: inadequate knowledge; insufficiency of
time and resources devoted to its obtainment; error; emotional bias due to imme-
diacy of interest; value-rational actions that disregard consequences; and predic-
tions that modify the future course of events (Merton, 1936).
However, Durkheim laid emphasis on still different causes. Assuming
advanced division of labor and inadequate social regulation, on the one hand, some
specialized tasks performed by single individuals and professional groups may not
correspond to the individuals’ talents, or be a constitutive part of the collective con-
sciousness, or finally be sufficient to fulfill their social functions. On the other hand,
differences in individuals’ natural endowment do not correspond to social inequal-
ities because of inheritance. There is not, therefore, fair competition. Moreover,
inequalities are not considered legitimate, as they are not based exclusively on
merit. The individuals’ or groups’ selfish pursuit of their own ends or concerns,
and, ultimately, social disintegration, may result from these causes (Durkheim,
1997: 62, 127, 138, 238–9; 1998b: 97, 108, 348, 358, 362–74, 380–1, 383, 387,
396–8; see also Giddens, 1978: 32–3; Poggi, 2000: 73).
•
If an individual form of consciousness prevails, there may be neither crime
nor public sanctioning, as norms reflecting the collective sentiments, which
crime has violated, would be either weak or non-existent.
•
If a social form of consciousness prevails, punishment is proportionate to the
strength of the collective sentiments, which crime has violated, as measured by
the intensity of social reaction to crime.
JOURNAL OF CLASSICAL SOCIOLOGY VOL 8(1)
126
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 126
•
If a social form of consciousness prevails, but collective sentiments have lost
strength as a consequence of the division of social labor, effective deterrence in
restraining crime depends on the intensity of social reaction, but even more on
the interest of those who have been damaged to demand compensation.
These statements presuppose a sufficiently clear distinction between an individual
and a social form of consciousness. This distinction considers whether a person is
aloof from society, or, on the contrary, actively partakes of and participates in it.
The prevalence of an individual state of consciousness may denote widespread
egoism, which is a social condition characterized by a relative lack of social inte-
gration. Assuming its prevalence, individuals are detached from society. They
therefore follow their own interests and neglect their social obligations
(Durkheim, 2002b: 223, 325, 428). It may also designate a social condition of
anomie, or absence of social regulation. If this condition obtains, there is no social
milieu that limits the individuals’ selfishness and social aloofness by indicating and
prescribing which ends individuals and groups ought to pursue, or how they
should relate to each other (Durkheim, 1998b: 348–9, 358–65, 401; 2002b:
280–1, 325, 429, 440).
Rampant egoism and anomie are distinct conditions, according to Durkheim,
for their causes – respectively, inadequate integration or regulation – are different.
The opposite thesis, that ‘egoism and anomie are identical’ (Johnson, 1965: 886),
does not then seem plausible (for evaluations of this thesis, see Giddens, 1971:
84–5; 1978: 46–7; Pescosolido and Georgianna, 1989: 43–4; Thorlindsson and
Bjarnason, 1998: 105–7). Both these conditions, however, point to the prevalence
of an individual state of consciousness, which leaves room for rational calculations
of costs and benefits without regard to the collective good. The ensuing, actual or
potential, conflicts make enduring solidarity impossible, for shared norms cannot
exist if exchanges are based on self-interests only (Durkheim, 1998b: 180–1;
2002b: 406–7; see also Pope and Johnson, 1983: 683). If an individual state of
consciousness prevails, therefore, there cannot be any solidarity, whether based on
norms or rational calculation of selfish interests.
With the exclusive prevalence of an individual state of consciousness there
would be neither crime nor punishment, for there would be no social sentiments of
sufficient intensity to support the existence of social norms (Durkheim, 1998b: 69).
No one, moreover, would be willing to shoulder the costs of repression as a public
service. Theories of deterrence from crime – such as rational choice theories (Akers,
2000: Ch. 2) – apply if there are social or moral sentiments, the offense of which
produces the punishment of the offender and society’s willingness to bear its cost.
They do not apply in traditional societies, in which these sentiments are strong and
compact, and punishment can be hardly graduated according to the seriousness of
the offense. Such theories apply in modern societies, however, for punishment can
be graduated in these societies since moral sentiments are comparatively less strong
and compact (Pope and Johnson, 1983: 686–7), and they are based on altruism
SEGRE DURKHEIM ON RATIONALITY
127
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 127
rather than on religion or other sources. Moreover, in addition to criminal laws,
there are civil laws, the violation of which concerns specific individuals more than
society at large (Durkheim, 1998b: 37–40, 79–83, 141, 146–8, 404).
•
If a social form of consciousness prevails, any social fact is a consequence of the
actors’ number, proximity, possibility of communicating, and socialization,
even when the collective sentiments that express it have lost strength because of
the division of social labor.
•
If a social form of consciousness prevails, even when it has lost strength because of
the division of social labor, allegiance to a religion ensues from actual or poten-
tial interactions with followers of that religion, rather than from other causes.
•
If a social form of consciousness prevails, and the division of social labor is
advanced, the social relevance of any particular group for its members and
for the other groups depends on the exchange of goods, ideas, and sentiments,
both within the group and outside of it.
•
A social form of consciousness prevails completely only if (in the language of
Social Network Theory): (a) all actors are connected directly or indirectly in
a subgroup; (b) all the relations connecting actors are multiplex; (c) all actors
belong to more than one subgroup; (d) all subgroups are connected directly or
indirectly; (e) all the subgroups have dense networks of relations, and, as a
consequence, the large group containing them has clusters; (f) at regular times
social gatherings take place in which all actors are involved, and collective
sentiments and ideas are entertained and reaffirmed.
•
If the division of social labor is advanced, the exchange of goods, ideas, and
sentiments engenders multiplex contents in social relations, which further
strengthens a social form of consciousness. In contrast, single relations based
exclusively on the exchange of goods, hence on reciprocity, are compatible both
with a non-advanced division of social labor and with the prevalence of an
individual form of consciousness.
In keeping with Durkheim, not only the actors’ numbers and their physical prox-
imity account for the production of social facts, but also their possibility and will-
ingness to interact with other actors, both in their own and in different settings.
RCT explanations of collective religious behavior cannot account for allegiance to
any particular religion to the extent that they posit the prevalence of an individual
form of consciousness, for no actor would be willing in this case to bear the costs,
financial or otherwise, involved in participating in religious ceremonies. No religious
allegiance, in general no allegiance to norms, and, ultimately, no society are there-
fore possible if an individual form of consciousness predominates. This form of
consciousness presupposes – it should be recalled – the absence of social norms.
Hence Durkheim’s insistence on the importance of moral education, whose purpose
is to make people aware of the social importance of moral conduct and willing to
abide by it (Wallwork, 1972: 124–7, 142–6, 166–8, 176–7, 189, 192).
The followers’ less extensive obligations to their religious organization, as
compared to others (Hechter, 1987: 57), may possibly account for its superior com-
petitiveness, but only if it is granted that they care enough about their faith to bear
JOURNAL OF CLASSICAL SOCIOLOGY VOL 8(1)
128
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 128
the costs involved in their membership. Their willingness to do so follows from the
prevalence of a social form of consciousness, which implies a high moral density of
social relations. A social form of consciousness may prevail only to some extent. Its
complete prevalence requires the presence of all the conditions indicated above. On
the other hand, the complete prevalence of an individual form of consciousness
requires the presence of individual interest only (‘là où l’intérêt règne seul’)
(Durkheim, 1998b: 181).
Hence, a social form of consciousness is compatible with an individual form
of consciousness when private interest is pursued along with the public interest (or
common good) both of the subgroup and of the whole network of the intercon-
nected subgroups. Norms – whether proscriptive or prescriptive – cannot be
enforced with a low moral density, as occurs if an individual state of consciousness
prevails completely. Norms are complied with mainly not as a matter of individual
convenience, but rather as an instrument to pursue the common good on the part
of individuals who are aware of its existence, and willing to contribute to it.
Individuals and groups may interact, establishing, for example, commercial rela-
tions, only on the basis of previously existent shared moral sentiments, which
would be thereby preserved and strengthened. In other words, moral density is
what matters for integration. Not all relations, irrespective of their content (com-
mercial, religious, or other), are conducive to moral density. If there is no cooper-
ation between persons or groups that remain otherwise separate though interacting
in some respect, then no moral sentiments arise between them, and no organic soli-
darity is therefore present (Durkheim, 1998b: 259–67).
A Durkheimian Theory of Social Integration
In modern circumstances, as collective consciousness becomes less compact and
more compatible with individual differentiation, individuals are less bound to their
original milieux, and more prone to have contacts with spatially and socially distant
environments. Religious as well as non-religious groups, such as those based on pro-
fessions, follow this general trend, since their notions become ever more abstract and
less defined, and their mores and ways of thinking become ever more shared by other
groups. The density and multiplexity of social relations make those groups able not
only to survive and thrive in societies with advanced division of social labor, but also
to foster and strengthen their moral character (Durkheim, 1997: 46–7; 1998b:
264–90; 2002a: 112–15). Durkheim would have then not objected to the statement,
originating from Social Network Theory, that social integration is predicated ‘on the
nature of ties with social groups, rather than simply their presence or absence’. A par-
ticular religion, for example, may or may not produce social integration, depending
on the characteristics of its followers’ social networks (Pescosolido, 1990: 353).
According to Durkheim, the network characteristics relevant for producing
integration are of two kinds. Firstly, there is the multiplexity of ties. Multiplexity
refers to multiple connections in a network, in the sense that actors are connected
SEGRE DURKHEIM ON RATIONALITY
129
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 129
through two or more roles (for example, as friends and colleagues), or through
different contents of the same role (as friends, actors may cooperate on a common
task, and provide advice and help to each other) (Kadushin, 2004: 27–9; Wasserman
and Faust, 1994: 81–3). Ties should not be, for instance, only religious. Secondly,
there is the existence of interconnected clusters, or subgroups, within the same net-
work. As a result, ‘moral life circulates through all its constituent relations’ (‘la vie
morale circule à travers toutes les relations qui la constitue’) (Durkheim, 1998b: 264).
A subgroup in a larger network is defined by the concentration of ties among actors
(Frank and Yasumoto, 1998: 644). According to Durkheim, and the contemporary
literature on social cohesion as well (Friedkin, 2004: 417–18), integration does not
require networks that are small, or contain strong interpersonal ties. Nor is a high
density of relations necessary. Density refers to the proportion of all the possible
connections that may be actually found in a network (Kadushin, 2004: 35;
Wasserman and Faust, 1994: 101).
Density, in this sense, indicates what Durkheim called a society’s material
density or concentration, which he sought to distinguish from its moral or dynamic
density, or, in his parlance, ‘the degree of coalescence of the social segments’
(Durkheim, 2002a: 113). Material density may be empirically related to moral
density, but they are conceptually and empirically distinct. High material density
may obtain in societies in which moral density is not high. For example, England
has a material density at least as high as France, due to the development of eco-
nomic exchanges, but has a lower moral density, since – in Durkheim’s opinion –
the local traditional mentality and way of life are more persistent in England
(Durkheim, 1998b: 266–7; 2002a: 113–14). The social costs involved in norm
enforcement depend, therefore, on society’s moral, rather than material, density.
The predominance of individualism in a society would make norm enforcement
impossible, since no one would be willing to contribute to the costs of enforce-
ment, and everyone would be rewarded for defection, as shown by free riding.
Anticipating contemporary Rational Choice Theory, Durkheim grappled with the
prisoner’s dilemma in collective action. In this social dilemma, unilateral defection
is rational for the individual, but – in Heckathorn’s words – ‘individually rational
actions lead to a collectively irrational outcome’. Durkheim’s solution to this
dilemma emphasized full cooperation as the only feasible strategy to achieve the
public good in a lasting and stable way. In line with this strategy, actors sanction
those who fail to contribute to the public good, while contributing themselves to it.
According to Durkheim, under certain conditions which he took pains to
specify, full cooperation would be possible. There would then be little danger of
defection, hypocritical cooperation, or full opposition. There would also be little dan-
ger of overproduction of the public good. In the absence of oppositional control, full
cooperation involves the danger of overproducing the public good. The production
costs would then exceed its benefits (Heckathorn, 1993: 376–7; 1996: 254, 257,
258–60). As Durkheim contended, reciprocity – and retaliation strategies in particu-
lar – provides no solidarity between the interacting parties, as shown by the existence
JOURNAL OF CLASSICAL SOCIOLOGY VOL 8(1)
130
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 130
of trading relations between enemies. Actors would provide cooperation, further-
more, irrespective of the cooperation offered by other actors. To achieve full cooper-
ation, actors should agree, even if their interests diverge, to create and support an
organization – such as a professional association, or a society of nations – acting as a
recognized moral authority, and able, therefore, to maintain generalized trust. This
aim involves regulating relations between actors. It also involves keeping in check
selfish individual pursuits, as evidenced by free riding and holding exclusive connec-
tions (whereby some actors exploit their central position in the network to exclude
others from benefits) (Durkheim, 1997: 46–7; 1998b: 265–6, 275–6; 2002a: 112–15;
for the concept of exclusive connection, see Willer, 1999).
Accordingly, the regulatory organization must exert normative oppositional
control over actors, including control over their attempts to control other actors.
Actors would then form a morally homogeneous but otherwise heterogeneous
group, in the sense of containing subgroups with a variety of specific contents, and,
in particular, with different economic contents. In the large group, relations would
be both economic and moral. If social conditions most conducive to a social form
of consciousness obtain, every member belongs to several clusters or subgroups,
each of which has one specific content and is endowed with one or more particular
meanings shared by its members, but is linked through its members to other sub-
groups. A complex unitary society then results from this pattern of multiplex and
meaningful relations, within and across subgroups. This occurs as long as relations
are permeated with the value of solidarity and there is not, therefore, disagreement
within and across subgroups concerning the fundamental importance of this value,
irrespective of the particular goals they may pursue (Durkheim, 1997: 51–2, 67;
1998b: v–viii, xi–xii, xv, xvii–xviii, 260–6; 2003: 50, 102; see also Fine and
Kleinman, 1983; Harrington and Fine, 2000: 316–17; Oliver, 1997).
Solidary or moral sentiments, if prevailing in societies with advanced div-
ision of labor, have a few noteworthy consequences. As everybody would benefit
in the long run, individual and collective rationality coincide. This would be so,
however, not because (at least, not only because) ‘the net loss in the value of the
collective good produced after defecting would exceed the costs of contribution’
(Heckathorn, 1996: 259). Individuals do not evaluate costs and benefits to seek
private advantages that may result in the detriment of others. Rather, they are will-
ing to contribute to the collective good, and bear their share of the costs involved
in sanctioning defectors, as a consequence of being enmeshed in a web of multi-
plex, and hence solidary, relations. Norms mandate compliance, but individual
costs would not as a rule exceed the collective gain, since the diversity of the
actors’ interests, and the multiple and various content of their relations, promote
oppositional control and thereby prevent the overproduction of the public good
(Heckathorn, 1990: 377; 1996: 259–60, 271).
Moreover, the absence of structural isolation both within and across groups
prevents its exploitation on the part of some actors for the purpose of playing the
different members and groups against each other, and gaining control over all of
SEGRE DURKHEIM ON RATIONALITY
131
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 131
them (Padgett and Ansell, 1993: 1280). In addition, for a given level of the indi-
vidual members’ and their groups’ heterogeneity of interests and resources,
threshold levels of collective action are more easily met the greater the multi-
plexity of relations connecting actors, and the greater the density of relations
between actors and between their subgroups. The multiplexity and density of rela-
tions promote a social form of consciousness. Actors are better able to grasp their
common and general interests, become aware of their interdependence, and
behave accordingly by forming a critical mass that overcomes the initial costs in
producing the collective action (Marwell and Oliver, 1993; Oliver and Marwell,
2002: 174–8). As a social form of consciousness involves the existence of many
interconnected subsets or clusters, the greater the number of ‘social ties that
bridge the boundaries of local clusters’, the lower the threshold level needed for
a critical mass (Macy, 1991a: 735).
Frequent interactions foster trustworthiness when situated or ‘embedded’
in particular subgroups (Granovetter, 1985: 490–3). Considering Durkheimian
sociology from the viewpoint of Social Network Theory (Segre, 2004), if ‘occa-
sional contacts’ between similar subgroups prevail, dense ties within cohesive sub-
groups, combined with sparse ties and reciprocity transactions between subgroups
(Frank and Yasumoto, 1998: 673), do not provide the most effective strategy to
create general trust and universal cooperation among all subgroups (Macy and
Skvoretz, 1998: 658). On the contrary, anomie is likely to result from the preva-
lence of random interactions (Macy, 1991b: 834). Rather, a web of multiplex rela-
tions between actors who belong to different subgroups may bring about general
trust, and hence a maximum of social capital, universal cooperation, and, ulti-
mately, a rational society in Durkheim’s sense, for, in his words, ‘only what is
rational is universal’ (Durkheim, 1998b: 275). In this sense, ‘rational actors simul-
taneously sustain their subgroups and the linkages between them’ (Frank and
Yasumoto, 1998: 644), provided that at least some of such linkages have multiplex
or non-identical contents, and a non-occasional character.
Actors may achieve a measure of control over uncertainty, which is inher-
ent to any relation, as they are involved in multiplex relations with every partner
(so the tie cannot be very weak), and know that their partners are similarly
involved (so defection would be costly to them). General trust is promoted if
multiplex relations are held with a plurality of partners, rather than with particu-
lar ones. ‘The confines of safe but closed relationships’ (Yamagishi et al., 1998:
189) promote the obtainment of the collective good for small groups (Olson,
1965) but – as Durkheim contended – are not conducive to general trust. In con-
trast, subgroups having a specific content and a dense network of relations
connecting their members, and connected by relations at least some of which are
multiplex, achieve the collective good of both their members and society (namely,
the whole network). Multiplex relations prevent exclusionary practices, and make
it possible to consistently control and sanction uncooperative behavior on the part of
known and unknown interaction partners. Consistent sanctioning of uncooperative
JOURNAL OF CLASSICAL SOCIOLOGY VOL 8(1)
132
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 132
behavior (which is considered deviant if a social form of consciousness prevails) is
necessary to obtain a cooperative equilibrium (Macy, 1991b: 839).
A social form of consciousness predominates, according to Durkheim, only
in a web of multiplex relations. In such a web, at least some multiplex relations are
present and strong, every actor is a member of distinct and interconnected sub-
groups, and all subgroups are connected by common memberships. Strong ties may
connect actors who are distant in terms of rank, status, or age, and ties – whether
strong or weak – may be advantageous not only to the individual actors, but also to
the network as a whole in terms of creation of social capital (Bian, 1997; Lin, 2001:
94–5). This is the case, in particular, of actors belonging to different subgroups and
therefore occupying distant positions in the social network. For example, actors may
belong to a subgroup as followers of a given religion, to another as members of a
family, to a third as constituents of a group of friends, to a further one as occupa-
tional colleagues, to still another one if active in the same firm. If there is advanced
division of labor, the prevalence of a social form of consciousness implies that there
is organic solidarity, namely, awareness of common interests engendered by the divi-
sion of labor is more strongly perceived than awareness of conflicting interests.
Organic solidarity is, accordingly, a particular manifestation of a social form of con-
sciousness characterizing society as a whole, while mechanical solidarity prevails
within the occupational subgroups (Wallwork, 1972: 44).
Every subgroup is homogeneous, to the effect that relations (or ties, or
links) have the same content, but also heterogeneous, as every actor of every
subgroup has a peculiar combination of links with other actors and subgroups.
Heterogeneity is a property not only of the single subgroups, but also of society’s
network, as relations between subgroups have a multiplex content. If a social form
of consciousness prevails, there are for every actor two overlapping public goods:
that of the actors’ subgroups (there are at least two of them for every actor) and
that of the large group encompassing the subgroups. Their combined value might
be sufficient to encourage universal cooperation, but cooperation is mandated
anyway by solidarity norms. These norms are enforced by large group’s institu-
tions regulating relations between the subgroups, such as markets or hierarchies
(Heckathorn, 2001: 278–81), but are upheld by the actors’ awareness of interde-
pendence between themselves and between the subgroups. Actors draw this
awareness from their multiplex links, which would be coextensive with society as
a whole in the ideal-typical case of complete prevalence of a social state of con-
sciousness and (therefore) maximum moral density. In this abstract case, social cap-
ital is coextensive with society, and rests on the closure of the whole social network.
Trustworthiness, which flows from social capital (Coleman, 1990: 315–20), creates
‘an extensive set of expectations and obligations’ (Coleman, 1990: 318) between
the subgroups. In this case, heterogeneity has no ‘atomizing effect’ on the single
subgroups, in the sense of introducing value heterogeneity, and producing divided
loyalties within the group as a result. Society cannot therefore be equated either
to a voluntary system of collective action, in which subgroups are free to contribute
SEGRE DURKHEIM ON RATIONALITY
133
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 133
or not, or to a system of compliant control, in which actors contribute and punish
those who do not. It may be equated, on the other hand, to a balanced system, in
which actors can enforce or oppose compliance norms.
In a balanced system, oppositional control (thereby control attempts on the
part of others are opposed) prevents the overproduction of the public good in the
large group. Overproduction would otherwise result from excessive, and therefore
irrational, control and cooperation between the subgroups. Norm enforcement, on
the other hand, prevents free riding and, therefore, low levels of production of social
solidarity, which is society’s common good. A balanced system results from the
actors’ participation in different subgroups. An actor who belongs to two or more par-
ticular subgroups would comply with the social norm of solidarity if the subgroups
have similar or complementary interests, and would instead exert oppositional control
if interests cannot be reconciled. In heterogeneous groups – such as society as a
whole – heterogeneity may concern not only every subgroup’s interests or value of
the public good, but also the resources it brings to systemic solidarity, considered as a
public good encompassing the particular public goods of the subgroups.
If societal heterogeneity results from different resources, the amount of the
common good brought by collective action may not be lesser when oppositional
control, rather than compliant control, is exerted. Oppositional control is a conse-
quence of a given actor’s multiplex ties, for participation in several subgroups
makes it unlikely that this actor evaluates the public good of each subgroup in the
same way, has the same cost of contribution to each of them, and gives to each the
same amount of resources. Both forms of normative control, compliant and oppos-
itional, produce a higher level of collective action in comparison to a system in
which actors contribute voluntarily (Heckathorn, 1990: 378–81; 1993: 340–7).
The creation and enforcement of norms, whether compliant or oppositional, are
therefore rational if producing a robust level of collective action is a societal goal.
Rationality in this case does not presuppose an interest in norm compliance on the
part of all actors (as with a ‘thin’ conception of rationality), but rather such an
interest on the part of subgroups or coalitions of them.
Solidarity has a further consequence in addition to universal cooperation and
trustworthiness. It produces severe constraints on the spread of both ‘independent
thinking’ (Durkheim’s ‘individualism’) and ‘provincialism’. While individualism
thrives if weak ties are prevalent, provincialism is instead a social condition in which
local clusters of actors are related by strong ties, deal with each other, and exclude
others (Macy, 1991a: 745–6; on the social consequences of weak and strong ties,
see Granovetter, 1995, 1998). As a social form of consciousness involves multiplex
relations, there are no ties with one content, such as non-contractual relations based
on reciprocity only. In addition, weak ties are not very frequent, for ‘most multiplex
ties . . . [are] strong’ (Granovetter, 1973: 1361, n. 3). The strength of ties – as
measured by their emotional intensity, amount of time spent in a given relationship,
intimacy, reciprocal services spontaneously performed – differs between relations
contained in subgroups and those linking them. The proportion of strong ties is
JOURNAL OF CLASSICAL SOCIOLOGY VOL 8(1)
134
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 134
greater, and a high level of multiplexity occurs more frequently, in the former type
of relations. In modern circumstances, characterized by advanced division of social
labor, economic and moral ties link subgroups, such as nations, that exchange goods
on the basis of contractual relations (Durkheim, 1998b: 265).
A higher level of multiplexity may be expected in relations of a non-economic
nature, which are more frequent within subgroups rather than across them. In any
case, multiplexity may prevent polarization in a balanced system of heterogeneous
subgroups between those which favor compliant control and those which support
oppositional control (Heckathorn, 1993: 344). The web of multiplex relations, so
conceived, differs from a system of relations in which actors are identified only by
the position they occupy (as in Burt, 1982), since the actors’ socialization, ideas, and
sentiments facilitate or constrain their opportunity of interacting. It also differs from
a network containing structural holes that put some actors in the position of gate-
keepers, and therefore provide them with information, control, and thus power over
the other actors (as in Burt, 1992). Since all the actors have multiplex relations, and
all their subgroups are connected directly or indirectly, there are no structural holes,
whether primary or secondary, and no gatekeepers. It differs, finally, from a web of
group affiliations, in which multiplex relations interconnect actors, but solidary ties
do not necessarily bind them (as in Simmel, 1955).
Relations linking subgroups may be conducive to social integration not
merely as ‘an unintended by-product of the efforts’ made by individual actors to con-
trol other actors of their own subgroup (Hechter et al., 1992: 83), because nothing
would in this case ensure the obtainment of the collective good, rather than an endur-
ing anomic condition of unregulated competition ‘for resources and members’
between the subgroups. Nor would the qualification that no subgroup should ‘com-
mand control over resources that threaten the unique position of the State’ (Hechter
et al., 1992: 94) be sufficient to solve the problem of social order in heterogeneous
societies, for this qualification is compatible with unregulated competition, and there-
fore anomie, which the regulative powers of the state may be unable to counteract.
The absence of isolated nodes, or actors, and the multiplexity and density
of relations connecting them do not suffice for the existence of a social state of
consciousness. Its creation and periodical renovation by means of ceremonies,
meetings, and other gatherings are also necessary, as Durkheim argued in The Ele-
mentary Forms of Religious Life (1998a). On such occasions social actors are close
to each other, exchange social ideas, and share sentiments and emotions. These
social conditions are conducive to a particular psychic state of arousal, or ‘efferves-
cence’, which flows from the spiritual intensity and the sacred character of the gath-
ering. Those who have taken part in the these gatherings, which have a ritual and
sacred connotation, are reconstituted as members of the same cognitive, cultural,
and moral community, and thus produce or reaffirm a group identity. As they
originate from society, rather than from the individuals, practices, thoughts, and
representations develop according to rules of their own, and exert a strong influ-
ence on the individual spirits.
SEGRE DURKHEIM ON RATIONALITY
135
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 135
This state of effervescence provides individuals and society as a whole with
the constitutive elements – behavioral and symbolic – of the collective conscious-
ness. To the extent that emotions bring about this consequence, they are rational
in Durkheim’s sense (Durkheim, 1998a: 326–7, 453–64, 547, 553–4, 571–2,
582, 603–5, 610–11, 621–38; for different interpretations of this Durkheimian
work, see Collins, 2004: 32–40; Giddens, 1978: Ch. 5; Lukes, 1975: 462–77;
Poggi, 2000: 135–6; Rawls, 1996, 2001, 2004; Wallwork, 1972: 18–19, 37–41).
As Alberoni (1981: 44) has pointed out, Durkheim distinguished the collective
emotional state that characterizes the meetings of participants in collective move-
ments, whose experience makes their lives qualitatively different, from emotional
conditions arising in the course of other gatherings (on the emotional dimensions
of social movements, see Goodwin et al., 2004). This distinction, incidentally,
cannot be found in the works of coeval prominent sociologists such as Michels
(1911: 25–6) and Simmel (1908: 470–1).
Social Integration and Rationality
In keeping with a broad notion of rationality, and a version of Rational Choice
Theory as an explanatory model (Abell, 1992: 203; Elster, 1984: 112) which is
consistent with this notion, the following statements may be made. They indicate
some points of convergence between this Durkheimian theory and a particular
conception of Rational Choice Theory, whereby cooperative behavior has entered
‘into the utility function of the individual’ (Elster, 1984: 146).
•
Actors are assumed to be usually rational (Elster, 1989a: 22), and thereby
capable of mutual understanding and communication.
•
Actors must be connected by multiplex and meaningful links (which are
therefore not only of an economic nature) if the interdependence of the
individuals’ preferences, which originates from the structure of social rela-
tions, may promote exchange, communication, and socialization between
and within subgroups.
•
Structurally produced interdependence between individual actors and
between subgroups promotes (and is promoted by) a social state of con-
sciousness, and, therefore, also general trust and a cooperative attitude. This
is the more so, the greater the number of subgroups and connections
between them.
A few consequences may be derived from these theoretical statements. Firstly, as this
Durkheimian theory formulates the social conditions conducive to cooperative (rather
than non-cooperative) games, and a narrow notion of rationality is not maintained,
the questions of how individual preferences are formed, or what amount of informa-
tion may be deemed adequate (‘satisficing’), are not relevant in this case. As these are
unresolved issues for Rational Choice Theory if narrowly defined, making no refer-
ence to a narrow notion of rationality may yield a number of distinctive theoretical
JOURNAL OF CLASSICAL SOCIOLOGY VOL 8(1)
136
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 136
advantages. In particular, such a notion of rationality is hardly useful when ceasing to
collect information does not depend on decisions made by given actors, or on the
amount of information they possess. Behavior, furthermore, when indicated by
choices, does not reveal the actors’ preferences, as choices should be interpreted with
reference to sensible psychological assumptions, and such interpretations lose plausi-
bility if choices are interdependent and/or the actors are unknown to the observer.
The content and even existence of preferences cannot be accordingly
ascertained, and, for a given quantity of empirical evidence and a set of beliefs,
there may be either several optimal actions or none at all. There is, therefore, no
empirically validated explanation on how preferences are formed. Preferences are
also not necessarily stable, nor do actors have an inborn tendency to maximize the
benefits they expect from their actions. What is more, given a set of pre-existing
desires and beliefs, the amount of empirical evidence one can rationally achieve
may be undetermined, because the marginal costs the search for further informa-
tion would involve, as well as the marginal benefits that would derive from them,
may be neither known nor assessable (Abell, 2000: 233–5; Baert, 1998: 168–9;
Bohman, 1992; Elster, 1983: 123–4, 143–57; 1986: 17–22; 1989a: 30–41; 1989b:
1–35; 2000a: 8–11; Friedman and Hechter, 1988; Harsanyi, 1986: 102; Mooney
Marini, 1992; Sen, 1986; Tversky and Kahneman, 1986).
Secondly, this Durkheimian theory pays attention to the micro-conditions
that make for a social state of consciousness, and therefore considers (as will be
recalled) direct interactions between identifiable actors in specific settings. It con-
forms, accordingly, to recommendations expressed by some students of Rational
Choice Theory. Namely, RCT as a set of verifiable explanatory statements should
concern itself with relating the micro- and macro-analysis levels (Friedman and
Hechter, 1988). It should also provide an interpretation of the actors’ subjective
rationality – whether they are individuals, or rather organizations and other collective
categories – and of emotions as an incentive to action (Baert, 1998: 166–7; Collins,
1988, 1996; Denzin, 1990; Garfinkel, 1984: 262–83; Goldthorpe, 1998; Green and
Shapiro, 1994: 33–46). While the Durkheimian stress on multiplexity of ties and
interconnection of subgroups is compatible with the actors’ cooperative, rather than
individualist, orientation, it should be emphasized that it does not presuppose altru-
ism on the actors’ part, for the Durkheimian rational and interdependent actors pur-
sue their own private interest along with the public interest (Kollock, 1998: 192–3).
Finally, this Durkheimian Rational Choice Theory differs from other ver-
sions of RCT, broadly defined, in that it not only emphasizes shared ends and norms
of moral actions, or widespread agreement on what are deemed valid and strong
reasons for actions, decisions, and beliefs. It also lays stress on the multiplexity
of ties, and on interconnected subgroups, as social facts that have to be taken into
due consideration in order to avoid anomie, and have moral life circulating (to use
Durkheim’s apt expression) in most, if not all, social relations. Morality would in
this case be brought into line with rationality, if rationality is considered a social prop-
erty (Shilling and Mellor, 1998). If this structure of relations obtains, then there is
SEGRE DURKHEIM ON RATIONALITY
137
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 137
enforced (possibly in addition to spontaneous) cooperation and reciprocity
between the members of each subgroup. Actors partake in a plurality of relatively
small subgroups, in each of which information flows freely (a condition for co-
operation) (Elster, 1984: 146), and restricted as well as generalized exchanges
take place to everyone’s advantage (Ekeh, 1974: 51–5). Cooperation is rewarded
and binding and defection is punished in these subgroups. To the extent that they
constitute moral milieux, as may well be the case with professional or occupational
groups, they may develop an identity of their own, originating from ‘a belief in
the interdependencies of group members and expectation of reciprocities among
the members’. Their members’ cooperation, and establishing clearly defined
group boundaries, would then be likely to follow (Kollock, 1998: 194–5, 198,
200, 204).
As each subgroup possesses an identity of its own, and is homogeneous and
compact (to the effect that it forms a dense network in which social ties have a simi-
lar content), finding support and meeting organizing costs for collective action is
a relatively easy enterprise. ‘A critical mass of interested and resourceful’ actors who
can ‘coordinate their efforts’ may produce the critical mass necessary for collective
action, as they would capitalize on ‘crosscutting associational memberships’
(Marwell and Oliver, 1993: 112, 123; see also Oliver and Marwell, 2002: 174–8),
in addition to the group identity of each association. Moreover, group identity and
the particular structure of relations within and across subgroups (that is, the pres-
ence of local monitoring and sanctioning systems in interconnected subgroups
having multiplex ties) facilitate the use of selective incentives and sanctions
(Kollock, 1998: 205). In the ideal-typical case of complete prevalence of a social
state of consciousness, group identity and stability would prove no obstacle; rather,
they would be conducive to the circulation of social solidarity, hence of moral life,
across subgroups and throughout society.
Conclusion
By way of conclusion, the Durkheimian RCT, which has been reconstructed from
Durkheim’s whole corpus, has been shown to be compatible with a version of
RCT, and also with a number of statements bearing on network theory and the-
ories of collective action. In modern circumstances, with advanced division of
social labor, a rational orientation on the part of individuals and their subgroups
involves being aware of, and pursuing, their common good. In turn, this implies
a social state of consciousness, and the general trust and cooperative attitude that
come with it. Ultimately, a social state of consciousness involves network
properties such as multiplexity of ties and interconnection of subgroups, for they
ensure the monitoring and sanctioning of individual and collective behavior. It also
involves the periodical recreation of a social state of consciousness, which concerns
society rather than particular subgroups, as a consequence of social gatherings to
which a sacred character is generally imputed.
JOURNAL OF CLASSICAL SOCIOLOGY VOL 8(1)
138
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 138
This Durkheimian Rational Choice Theory suggests further theoretical ques-
tions, such as: What kind of rituals, if any in particular, may best promote a social state
of consciousness? Do the most effective rituals (in this sense) depend on given net-
work properties? Do rituals promote a social state of consciousness directly (as
Durkheim apparently maintained), or indirectly, by making individual actors aware of
the multiplexity of ties and interconnection of subgroups? Is the monitoring and sanc-
tioning of individual and collective behavior best performed by individuals acting on
behalf and under the mandate of subgroups, such as representatives of professional or
occupational associations, or rather by state representatives? Finally, and as a more
general question, does an inadequate awareness of the network properties on the part
of the individual actors, or rather the relative paucity of multiplexity of ties and inter-
connection of subgroups, pose a greater threat to a social state of consciousness?
References
Abell, P. (1992) ‘Is Rational Choice Theory a Rational Choice of Theory?’,
pp. 183–206 in J. Clark (ed.) James S. Coleman. London: Falmer.
Abell, P. (2000) ‘Sociological Theory and Rational Choice Theory’, pp. 223–44
in B.S. Turner (ed.), The Blackwell Companion to Social Theory . Oxford:
Blackwell.
Akers, R. (2000) Criminological Theories. Los Angeles, CA: Roxbury.
Alberoni, F. (1981) Movimento e istituzione. Bologna: Il Mulino.
Baert, P. (1998) Social Theory in the Twentieth Century. Cambridge: Polity.
Benard, S. (2004) ‘Boundary Patrols in Collective Action: A Computational Model
of Opposition and Solidarity’. Paper presented at the 99th Annual Meeting
of the American Sociological Association, San Francisco, August 14–17.
Bian, Y. (1997) ‘Bringing Strong Ties Back In: Indirect Ties, Network Bridges,
and Job Searches in China’, American Sociological Review 62: 366–85.
Bohman, J. (1992) ‘The Limits of Rational Choice Explanation’, pp. 207–28 in
J. Clark (ed.) James S. Coleman. London: Falmer.
Boudon, R. (1981) ‘Undesirable Consequences and Types of Structures of
Systems of Interdependence’, pp. 255–84 in P.M. Blau and R. Merton
(eds) Continuities in Structural Inquiry. London: Sage.
Boudon, R. (1990) ‘The Two Facets of the Unintended Consequences Paradigm’,
pp. 119–27 in J. Clark, C. Modgil and S. Modgil (eds) Robert K. Merton:
Consensus and Controversy. London: Falmer.
Boudon, R. (1998a) ‘Social Mechanisms without Black Boxes’, pp. 172–203 in
P. Hedstroem and R. Swedberg (eds) Social Mechanisms. Cambridge:
Cambridge University Press.
Boudon, R. (1998b) ‘Limitations of Rational Choice Theory’, American Journal
of Sociology 104: 817–28.
Boudon, R. (2003) ‘Beyond Rational Choice Theory’, Annual Review of Sociology
29: 1–21.
SEGRE DURKHEIM ON RATIONALITY
139
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 139
Burt, R.S. (1982) Towards a Structural Theory of Action. New York: Academic Press.
Burt, R.S. (1992) Structural Holes: The Social Structure of Competition.
Cambridge, MA: Harvard University Press.
Coleman, J.S. (1990) Foundations of Social Theory. Cambridge, MA: Harvard
University Press.
Coleman, J.S. (1996) ‘A Vision for Sociology’, pp. 343–9 in J. Clark (ed.) James
S. Coleman. London: Falmer.
Collins, R. (1988) Theoretical Sociology. Orlando, FL: Harcourt Brace Jovanovich.
Collins, R. (1996) ‘Can Rational Action Theory Unify Future Social Science?’,
pp. 329–43 in J. Clark (ed.) James S. Coleman. London: Falmer.
Collins, R. (2004) Interaction Ritual Chains. Princeton, NJ: Princeton University
Press.
Denzin, N. (1990) ‘Reading Rational Choice Theory’, Rationality and Society 2: 172–89.
Durkheim, É. (1997) Leçons de sociologie. Paris: Presses Universitaires de France.
Durkheim, É. (1998a) Les formes élémentaires de la vie religieuse. Paris: Presses
Universitaires de France.
Durkheim, É. (1998b) De la division du travail social. Paris: Presses Universitaires
de France.
Durkheim, É. (2002a) Les règles de la méthode sociologique. Paris: Presses Univer-
sitaires de France.
Durkheim, É. (2002b) Le Suicide. Paris: Presses Universitaires de France.
Durkheim, É. (2003) Education et sociologie. Paris: Presses Universitaires de France.
Ekeh, P. (1974) Social Exchange Theory: The Two Traditions. London: Heinemann.
Elster, J. (1983) Sour Grapes. Cambridge: Cambridge University Press.
Elster, J. (1984) Ulysses and the Sirens: Studies in Rationality and Irrationality.
Cambridge: Cambridge University Press.
Elster, J. (1986) ‘Introduction’, pp. 1–33 in J. Elster (ed.)Rational Choice. Oxford:
Blackwell.
Elster, J. (1989a) Nuts and Bolts for the Social Sciences. Cambridge: Cambridge
University Press.
Elster, J. (1989b) Solomonic Judgements. Cambridge: Cambridge University Press.
Elster, J. (1995) The Cement of Society. Cambridge: Cambridge University Press.
Elster, J. (2000a) Ulysses Unbound: Studies in Rationality, Precommitment, and
Constraints. Cambridge: Cambridge University Press.
Elster, J. (2000b) ‘Rationality, Economy, and Society’, pp. 21–41 in S. Turner (ed.)
The Cambridge Companion to Weber. Cambridge: Cambridge University Press.
Fine, G.A. and S. Kleinman (1983) ‘Network and Meaning: An Interactionist
Approach to Structure’, Symbolic Interaction 6: 97–110.
Frank, K.A. and J.Y. Yasumoto (1998) ‘Linking Action to Social Structure within
a System: Social Capital within and between Subgroups’, American
Journal of Sociology 104: 642–86.
Friedkin, N.E. (2004) ‘Social Cohesion’, Annual Review of Sociology 30: 409–25.
JOURNAL OF CLASSICAL SOCIOLOGY VOL 8(1)
140
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 140
Friedman, M. and M. Hechter (1988) ‘The Contribution of Rational Choice
Theory to Macrosociological Research’, Sociological Theory 6: 201–18.
Garfinkel, H. (1984) Studies in Ethnomethodology. Cambridge: Polity.
Giddens, A. (1971) Capitalism and Modern Social Theory. Cambridge: Cambridge
University Press.
Giddens, A. (1978) Durkheim. Glasgow: Fontana/Collins.
Goldthorpe, J.H. (1998) ‘Rational Action Theory for Sociology’, British Journal
of Sociology 49: 167–92.
Goodwin, J., J.M. Jasper and F. Polletta (2004) ‘Emotional Dimensions of Social
Movements’, pp. 380–412 in D.A. Snow, S.A. Soule and H. Kriesi (eds)
The Blackwell Companion to Social Movements. London: Blackwell.
Granovetter, M. (1973) ‘The Strength of Weak Ties’, American Sociological
Review 78: 1360–80.
Granovetter, M. (1985) ‘Economic Action and Social Structure: The Problem of
Embeddedness’, American Journal of Sociology 91: 481–510.
Granovetter, M. (1995) Getting a Job. Chicago, IL: University of Chicago Press.
Granovetter, M. (1998) La forza dei legami deboli e altri saggi. Naples: Liguori.
Green, D.P. and I. Shapiro (1994) Pathologies of Rational Choice Theory: A Critique
of Applications in Political Science. New Haven, CT: Yale University Press.
Habermas, J. (1984) The Theory of Communicative Action, Vol. 1. Boston, MA:
Beacon.
Harrington, B. and G.A. Fine (2000) ‘Opening the “Black Box”: Small Groups and
Twenty-First Century Sociology’, Social Psychology Quarterly 53: 312–23.
Harsanyi, J.C. (1986) ‘Advances in Understanding Rational Behavior’, pp. 82–107
in J. Elster (ed.) Rational Choice. Oxford: Blackwell.
Heckathorn, D.D. (1990) ‘Collective Sanctions and Compliance Norms: A Formal-Theory
of Group-Mediated Social Control’, American Sociological Review 55: 336–84.
Heckathorn, D.D. (1993) ‘Collective Action and Group Heterogeneity: Voluntary
Provisions versus Selective Incentives’, American Sociological Review 58: 329–50.
Heckathorn, D.D. (1996) ‘The Dynamics and Dilemmas of Collective Action’,
American Sociological Review 61: 250–77.
Heckathorn, D.D. (2001) ‘Sociological Rational Choice’, pp. 273–84 in G. Ritzer
and B. Smart (eds) Handbook of Social Theory. London: Sage.
Heckathorn, D.D. (2006) ‘Rational Choice’, pp. 620–4 in G. Ritzer (ed.)
Encyclopedia of Social Theory. London: Sage.
Hechter, M. (1987) Principles of Group Solidarity. Berkeley, CA: University of
California Press.
Hechter, M. and S. Kanazawa (1997) ‘Sociological Rational Choice Theory’,
Annual Review of Sociology 23: 191–214.
Hechter, M., D. Friedman and S. Kanazawa (1992) ‘The Attainment of Global Order
in Heterogeneous Societies’, pp. 79–97 in J.S. Coleman and T.J. Fararo
(eds) Rational Choice Theory: Advocacy and Critique. London: Sage.
SEGRE DURKHEIM ON RATIONALITY
141
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 141
Hedstroem, P. (2005) Dissecting the Social: On the Principles of Analytical
Sociology. Cambridge: Cambridge University Press.
Hedstroem, P. and R. Swedberg (1996) ‘Social Mechanisms’, Acta Sociologica 39:
281–308.
Hernes, G. (1998) ‘Real Virtuality’, pp. 74–101 in P. Hedstroem and R. Swedberg
(eds) Social Mechanisms. Cambridge: Cambridge University Press.
Isambert, F. (1982) ‘Réflexions sur la stratégie durkheimienne de la détermina-
tion de l’objet’, L’Année Sociologique 32: 163–92.
Joas, H. (1997) G.H. Mead: A Contemporary Re-examination of His Thought.
Cambridge, MA: MIT Press.
Joas, H. (2001) ‘The Emergence of the New: Mead’s Theory and Its Contemporary
Potential’, pp. 89–99 in G. Ritzer and B. Smart (eds) Handbook of Social
Theory. London: Sage.
Johnson, B. (1965) ‘Durkheim’s One Cause of Suicide’, American Sociological
Review 30: 875–86.
Kadushin, C. (2004) ‘Some Basic Network Concepts and Propositions’, pp. 3–60
in Introduction to Social Network Theory (http://home.earthlink.net/
~ckadushin/Texts/Basic%20Network%20Concepts.pdf).
Kettler, D., V. Meja and N. Stehr (1984) Karl Mannheim. Chichester: Ellis Horwood.
Klandermans, B. (1997) The Social Psychology of Protest. Oxford: Blackwell.
Kollock, P. (1998) ‘Social Dilemmas: The Anatomy of Cooperation’, Annual
Review of Sociology 24: 183–214.
Lin, N. (2001) Social Capital: A Theory of Social Structure and Social Action.
Cambridge: Cambridge University Press.
Lindenberg, S. (1996) ‘Constitutionalism versus Relationalism: Two Versions of
Rational Choice Sociology’, pp. 299–311 in J. Clark (ed.) James S. Coleman.
London: Falmer.
Lukes, S. (1975) Émile Durkheim: His Life and Work. New York: Penguin.
Mackert, J. (2004) ‘Reorganization and Stabilization’, Journal of Classical Sociology
4: 311–36.
Macy, M.W. (1991a) ‘Chains of Cooperation: Threshold Effects in Collective
Action’, American Sociological Review 56: 730–47.
Macy, M.W. (1991b) ‘Learning to Cooperate: Stochastic and Tacit Collusion in
Social Exchange’, American Journal of Sociology 97: 808–43.
Macy, M.W. and J. Skvoretz (1998) ‘The Evolution of Trust and Cooperation
between Strangers: A Computational Model’, American Sociological
Review 63: 638–60.
Mannheim, K. (1940) Man and Society in an Age of Reconstruction. New York:
Harcourt, Brace.
Mannheim, K. (1993) From Karl Mannheim, ed. K.H. Wolff. London: Transaction.
Marwell, G. and P.E. Oliver (1993) The Critical Mass in Collective Action.
Cambridge: Cambridge University Press.
Mead, G.H. (1934) Mind, Self, and Society. Chicago, IL: University of Chicago Press.
JOURNAL OF CLASSICAL SOCIOLOGY VOL 8(1)
142
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 142
Merton, R.K. (1936) ‘The Unanticipated Consequences of Purposive Social
Action’, American Sociological Review 1: 894–904.
Michels, R. (1911) Zur Soziologie des Parteiwesens in der modernen Demokratie.
Leipzig: Klinkhardt.
Mooney Marini, M. (1992) ‘The Role of Models of Purposive Action in Sociology’,
pp. 21–48 in J.S. Coleman and T.J. Fararo (eds) Rational Choice Theory:
Advocacy and Critique. London: Sage.
Nisbet, R. (1974) The Sociology of Émile Durkheim. New York: Oxford University
Press.
Oliver, A.L. (1997) ‘On the Nexus of Organizations and Professions: Networking
through Trust’, Sociological Inquiry 67: 227–45.
Oliver, P.E. and G. Marwell (2002) ‘Recent Developments in Critical Mass
Theory’, pp. 172–93 in J. Berger and M. Zelditch (eds) New Directions in
Sociological Theory: Growth of Contemporary Theories. Lanham, MD:
Rowman & Littlefield.
Olson, M. (1965) The Logic of Collective Action. Cambridge, MA: Harvard
University Press.
Padgett, J.F. and C.K. Ansell (1993) ‘Robust Action and the Rise of the Medici,
1400–1434’, American Journal of Sociology 98: 1259–319.
Pescosolido, B.A. (1990) ‘The Social Context of Religious Integration and Suicide:
Pursuing the Network Explanation’, The Sociological Quarterly 31: 337–57.
Pescosolido, B.A. and S. Georgianna (1989) ‘Durkheim, Suicide, and Religion:
Toward a Network Theory of Suicide’, American Sociological Review
54: 33–48.
Poggi, G. (2000) Émile Durkheim. Bologna: Il Mulino.
Pope, W. and B.D. Johnson (1983) ‘Inside Organic Solidarity’, American
Sociological Review 48: 681–92.
Rawls, A.W. (1996) ‘Durkheim’s Epistemology: The Neglected Argument’,
American Journal of Sociology 102: 430–82.
Rawls, A.W. (2001) ‘Durkheim’s Treatment of Practice: Concrete Practice vs
Representations as the Foundation of Reason’, Journal of Classical
Sociology 1: 33–68.
Rawls, A.W. (2004) Epistemology and Practice. Cambridge: Cambridge University
Press.
Roth, G. (1968) ‘Introduction’, pp. xxxiii–cx in M. Weber, Economy and Society.
New York: Bedminster.
Sawyer, R.K. (2002) ‘Durkheim’s Dilemma: Toward a Sociology of Emergence’,
Sociological Theory 20: 227–47.
Schlegloff, E. (2000) ‘On Granularity’, Annual Review of Sociology 26: 715–20.
Segre, S. (2004) ‘A Durkheimian Network Theory’, Journal of Classical Sociology
4: 215–35.
Sen, A. (1986) ‘Behavior and the Concept of Preference’, pp. 60–81 in J. Elster
(ed.) Rational Choice. Oxford: Blackwell.
SEGRE DURKHEIM ON RATIONALITY
143
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 143
Shilling, C. and P. Mellor (1998) ‘Durkheim, Morality and Modernity: Collective
Effervescence, Homo Duplex and the Sources of Moral Action’, British
Journal of Sociology 49: 193–209.
Simmel, G. (1908) Soziologie. Berlin: Duncker & Humblot.
Simmel, G. (1955) ‘The Web of Group Affiliations’, pp. 128–95 in Conflict and
the Web of Group Affiliations. Glencoe, IL: Free Press.
Swidler, A. (1995) ‘Cultural Power and Social Movements’, pp. 25–40 in
H. Johnston and B. Klandermans (eds) Social Movements and Culture.
Minneapolis, MN: University of Minnesota Press.
Thorlindsson, T. and T. Bjarnason (1998) ‘Modeling Durkheim on the Micro Level:
A Study of Youth Suicidality’, American Sociological Review 63: 94–110.
Tole, L.A. (1993) ‘Durkheim on Religion and Moral Community in Modernity’,
Sociological Inquiry 63(1): 1–29.
Turner, J. (1982) The Structure of Sociological Theory. Homewood, IL: Dorsey.
Turner, J. (2006) Rational Choice Theory, pp. 497–98 in B.S. Turner (ed.) The
Cambridge Dictionary of Sociology. Cambridge: Cambridge University Press.
Tversky, A. and D. Kahneman (1986) ‘The Framing of Decision and the Psychology
of Choice’, pp. 123–41 in J. Elster (ed.) Rational Choice. Oxford: Blackwell.
Wallwork, E. (1972) Durkheim: Morality and Milieu. Cambridge, MA: Cambridge
University Press.
Wassermann, S. and K. Faust (1994) Social Network Analysis: Methods and
Applications. Cambridge: Cambridge University Press.
Weber, M. (1954) Max Weber on Law in Economy and Society, ed. M. Rheinstein.
New York: Simon & Schuster.
Weber, M. (1956) Wirtschaft und Gesellschaft. Tübingen: Mohr.
Weber, M. (1968) Economy and Society. New York: Bedminster.
Weber, M. (1973) Gesammelte Aufsätze zur Wissenschaftslehre. Tübingen: Mohr.
Weber, M. (1989) Zur Russischen Revolution von 1905. Tübingen: Mohr.
Willer, D. (ed.) (1999) Network Exchange Theory. London: Praeger.
Yamagishi, T., K. Cook and M. Watabe (1998) ‘Uncertainty, Trust, and Commit-
ment Formation in the United States and Japan’, American Journal of
Sociology 104: 165–94.
Sandro Segre is Full Professor of Sociology at the University of Genoa. His recent publications include
Gruppi economici e mercati finanziari nella sociologia di Max Weber [A Sociology of Business Groups and
Financial Markets: A Weberian Perspective] (Rome: Carocci, 2006), which is due to be published in English
as A Weberian Analysis of Business Groups and Financial Markets (Ashgate, spring 2008), and articles in
Sociologia e Ricerca Sociale, Quaderni di Sociologia, Max Weber Studies, Sociological Theory and Journal
of Classical Sociology.
Address: Dipartimento di Scienze Politiche e Sociali, Largo Zecca 8/16, 16124 Genoa, Italy. [email:
Segre@unige.it]
JOURNAL OF CLASSICAL SOCIOLOGY VOL 8(1)
144
109-144 JCS-084697.qxd 5/1/08 2:29 PM Page 144
Wyszukiwarka
Podobne podstrony:
Electrochemical properties for Journal of Polymer Science
Huang et al 2009 Journal of Polymer Science Part A Polymer Chemistry
Ionic liquids solvent propert Journal of Physical Organic Che
extraction and analysis of indole derivatives from fungal biomass Journal of Basic Microbiology 34 (
Dannenberg et al 2015 European Journal of Organic Chemistry
POLITICAL and LEGAL THOUGHT OF CLASSICAL ISLAM
Journal of KONES 2011 NO 3 VOL Nieznany
Journal of Chinese Cinemas Volume 2 Issue 1
Journal of Chemical Ecology 31 (2005)
N Grube A Dictionary of Classic Mayan
Lebrini et al 2005 Journal of Heterocyclic Chemistry
Journal of UN October 2014
ekonomika22, Journal of Media Economics, red
influence of classical iranian thought
Hua et al 2009 European Journal of Organic Chemistry
acetylcholinesterase inhibitors British Journal of Pharmacology
Electrochemical properties for Journal of Polymer Science
The Journal of Transpersonal Psychology Vol 26 1 (1994)
więcej podobnych podstron