Professional Documents
Culture Documents
Searle versus
Durkheim and the
waves of thought
Reply to Gross
John R. Searle
University of California, Berkeley, USA
In the many comments on The Construction of Social Reality that I have seen since its
publication 10 years ago, the single most astounding claim to me is Gross’s assertion that
it is essentially Durkheimian. He repeats this claim over and over. He says my work has
‘Durkheimian roots’, that it exhibits what he calls ‘unacknowledged and unreconstructed
Durkheimianism’; that I am ‘a Durkheimian at heart’; that the account is ‘steeped in
Durkheim’, that it re-expresses ‘what are essentially Durkheimian ideas in an analytic
idiom’. And so on. I think Gross protests perhaps too much and I want to protest equally
on the opposite side. Both historically and as a matter of content his claims are entirely
false. In order to clarify the differences I will contrast my view of institutional facts with
Durkheim’s account of ‘social facts’.
While I was working on The Construction I read some of the classical works on socio-
logical theory, including Durkheim’s chapter ‘Social Facts’ as well as some other sections
of his book The Rules of Sociological Method (1938). It seemed immediately obvious to
me that Durkheim had an inadequate conception of social facts. He fails to distinguish
social facts in general from the special subclass of institutional facts, he thinks that social
facts are essentially coercive, and he thinks that they exist outside of individual minds.
He says that the essence of social facts is that they are ‘external to the individual and
endowed with a power of coercion, by reason of which they control him’ (1938: 3). This
is exactly the opposite of my view. I think we need to start by distinguishing social facts
in general from institutional facts, that human institutions are in large part enabling and
empowering and that the collective intentionality which creates and sustains them is
entirely in individual minds. Because Durkheim’s account seemed so impoverished I did
not read any further in his work. It is, by the way, a common mistake in the tradition
to suppose that social facts are somehow or other essentially constraining or coercive.1
I, on the other hand, want to argue that social institutions are in general enabling and
empowering. This feature is disguised from us by the fact that once we are working inside
social institutions, they can seem to be constraining. Thus, in baseball, I only have three
strikes. Within the institution of money, I cannot spend money I do not have, or at least
57
ANTHROPOLOGICAL THEORY 6(1)
do not have control of. This makes it seem as if baseball and money are somehow
coercive and constraining, but it is important to remember that the institution itself
creates possibilities that life would not have without that institution. Think of money,
government, private property, cocktail parties, and universities, to take some examples
at random. All of these institutions enable us to perform kinds of activities that we could
not possibly perform without them.
I think this mistake in Durkheim derives from other, deeper, inadequacies in his work.
He never, as far as I know, seriously attempts to describe the differences between human
social reality and that of other social animals. All sorts of animals have social relations
and many species have social facts in Durkheim’s sense of external coercions. But the
special feature of human beings is that they have institutions, and once you explore the
logical structure of these institutions, their enabling and empowering features become
obvious. Language is the most important case, because the institution of language gives
us enormous powers not possessed by other animals, not even by other primates. And
it provides the foundation for all other institutions. I believe it is at least in part because
Durkheim fails to explore human social facts within the larger context of animal social
facts that he fails to see the distinction between social facts and institutional facts. His
analysis does not even get up to the point where he is able to pose the questions that I
am posing. For example, how do human institutional facts go beyond mere ‘social facts’?
What is the logical structure of institutions? How does language constitute institutional
reality in a way that it does not constitute brute reality?
Because Gross prompts me to distinguish my theory from Durkheim’s, it is essential
to remind ourselves at the beginning that Durkheim was writing a century ago and
lacked the resources of contemporary philosophy. For example, in a post-Fregean era,
any philosopher will be alerted by the intensionality (with an ‘s’) of institutional state-
ments.2 For contemporary philosophers intensionality is a red flag and prompts deeper
investigation. Durkheim treats all these contexts as if they were extensional and I believe
that is because he lacked the resources of contemporary philosophy. In what follows I
am not belittling Durkheim’s historical significance but questioning the adequacy of his
account for contemporary theory.
How many well-established cases suggest that thought can travel over a distance? The
difficulty which we may have in conceiving so disconcerting an idea is not sufficient
reason for us to deny its reality, and we shall in all probability have to admit the exist-
ence of waves of thought. . . . (Durkheim, 1953: 19)
58
SEARLE Searle versus Durkheim. Reply to Gross
In The Construction and subsequent works, I have made use of certain essential concepts
and presuppositions in trying to understand social reality in general and institutional
reality in particular. These come directly from my philosophy of language and phil-
osophy of mind. I will list several of these central concepts and distinctions, and for the
sake of brevity I will confine the list to a dozen.
How many of these do we find in Durkheim? As far as I can tell, exactly none. I used to
think his idea of ‘conscience collectif ’ was much like my ‘collective intentionality’, but
they are completely different, as I will explain.
With these concepts in mind, let us explore Durkheim’s account of social ontology. I
will list what I take to be some of the major defects in his account, and again for the
sake of brevity, I will confine the list to only 10. I would not wish to beg any questions
against Durkheim, so those who think he is right and I am wrong can think of this list
as giving us 10 major differences or points of dispute between Searle and Durkheim.
59
ANTHROPOLOGICAL THEORY 6(1)
There are several propositions in Durkheim’s writings on ontology that are not easy
to render consistent. Here are three. First, he claims that mental life and social reality
‘surpass’ their substrata; that collective representations are ‘exterior’ to individual minds
in the same way that mental states are exterior to their neuronal substratum. Second, he
also says that collective intentionality ‘is only realized in individuals’. And third, he
thinks that the way for us to understand how collective intentions can be exterior to
individuals is on the model of how we are supposed to understand that mental
phenomena can be external to the brain cells which form their substratum. To give you
a sense of the difficulty I will quote some representative passages from ‘Individual and
Collective Representations’ and from The Division of Labor.
When we said elsewhere that social facts are in a sense independent of individuals
and exterior to individual minds, we only affirmed of the social world what we have
just established for the psychic world . . . If there is nothing extraordinary in the fact
that individual representations, produced by the action and reaction between neural
elements, are not inherent in these elements, there is nothing surprising in the fact
that collective representations, produced by the action and reaction between
individual minds that form the society, do not derive directly from the latter and
consequently surpass them. The conception of the relationship which unites the
social substratum and the social life is at every point analogous to that which un-
deniably exists between the physiological substratum and the psychic life of indi-
viduals . . . the relative externality of social facts in relation to individuals, is even
more immediately apparent than is that of mental facts in relation to the cerebral
cells. (Durkheim, 1993 [1893]: 24–5)
If one can say that, to a certain extent, collective representations are exterior to
individual minds, it means that they do not derive from them as such but from the
association of minds, which is a very different thing. No doubt in the making of the
whole each contributes his part, but private sentiments do not become social except
by combination under the action of the sui generis forces developed in association.
(Durkheim, 1993 [1893]: 25–6, my italics)
The exteriority of mental facts in relation to the cerebral cells is due to the same causes
and is of the same nature. (1993 [1893]: 26)
Again,
the cells certainly help to produce [a mental representation] but do not suffice to
constitute, since it survives them and manifests different properties. (Durkheim,
1953: 24)
It is not easy to make these passages consistent with the last sentence of this passage
from The Division of Labor in Society:
60
SEARLE Searle versus Durkheim. Reply to Gross
The totality of beliefs and sentiments common to the average members of a society
forms a determinate system with a life of its own. It can be termed the collective or
common consciousness . . . By definition it is diffused over society as a whole, but
nonetheless possesses specific characteristics that make it a distinctive reality. In fact
it is independent of the particular conditions in which individuals find themselves.
Individuals pass on, but it abides . . . it does not change with every generation but,
on the contrary, links successive generations to one another. Thus it is something
totally different from the consciousness of individuals, although it is only realized in
individuals. (Durkheim 1993[1893]: 38–9, my italics)
How can we reconcile his view that social ontology is ‘only realized in individuals’ with
his claim that it is ‘to a certain extent . . . exterior to individual minds’?
It took me a long time to figure it out; but here, I think, is a sympathetic attempt to
give a coherent interpretation: Just as chemical alloys and compounds (one of his anal-
ogies) are explained by the association of the elements within them, but nonetheless go
beyond a mere summation of the elements, so collective intentionality is a result of the
association of individual intentionality but goes beyond it. For this reason, as he says
over and over, collective intentionality is not in individual minds. In the same way that
bronze is not in copper and tin individually, the collective representations of my society
are not in my mind or any other individual mind. It is the association of individual minds
that results in collective intentionality. Another analogy he likes is with life (remember,
he is writing a half century before Crick and Watson). Life does not reside in carbon,
hydrogen, nitrogen and oxygen even though these are the basic components of living
matter, rather their association gives rise to life, even though life is not in any of them
by themselves. Individual representations are the result of the actions of neurons but
they surpass the neurons. Thus there can be a science of psychology which is not neuro-
science because there is a sui generis level of individual representations. Now just as the
sui generis level of the mental makes possible a science of psychology, so a sui generis level
of the social makes possible a science of sociology.
Durkheim has various axes to grind. He wants sociology to be a genuine science like
physics with its own ontological domain that it studies, but he does not want it to be
reducible to, and hence a branch of, psychology. So, just as the association of neurons
produces a separate level, that of consciousness, which does not reside in individual
neurons, so the association of individual consciousnesses produces a level of social facts,
which do not reside in individual consciousnesses. Psychology studies individual
consciousnesses and sociology studies collective consciousness and its manifestation in
social facts.
This leads him to make apparently contradictory claims. On the one hand he says
repeatedly that collective representations are not in individual minds, and on the other
hand that they are realized in individual minds. Now maybe these really are contradic-
tory and he changed his mind from the time he wrote The Division of Labor to the time
he wrote ‘Individual and Collective Representations’, but one way of reconciling them
is to say that he takes the analogies seriously. Bronze is realized in copper and tin, but
neither of them contains any bronze. Analogously collective representations are realized
in your mind, my mind and Sally’s mind and so on, but none of us individuals has any
collective intentionality.
61
ANTHROPOLOGICAL THEORY 6(1)
In the end the picture is inadequate for the following reason. The relations in the
chemical alloys and compounds are causal relations. Similarly the mechanisms by which
the replication of DNA produces the cluster of phenomena we call ‘life’ are causal mech-
anisms. In each case there is a set of causes – the behavior of the lower level phenomena
– and an effect – the resultant higher level feature. But this part of the picture will not
work for collective intentionality. Though there are plenty of causal relations between
individual and collective intentionality, the relation of individual minds to collective
intentionality is constitutive and not causal. The collective intentionality is constituted
by what is in individual minds. It could not work otherwise. The explanation of human
behavior requires intentional causation, but intentional causation can only work if it is
internalized in individual minds. Collective intentionality can only function causally if
it is constituted by what is in the minds of individuals. For Durkheim it is impossible
to explain the mechanisms by which social facts ‘constrain’ individuals, because the
collective representations are not in individual minds.
I want to make this point absolutely clear: from my point of view the fundamental
ontology in Durkheim is mistaken. He does not give a consistent or coherent account
of the fundamental unity of the universe. We can make his claim that social ontology
does not exist in individual minds, consistent with the claim that it is realized in indi-
vidual minds, only at the cost of attributing to him a false conception of the relation-
ships between the social and the individual. On my view collective intentionality has to
be in individual brains, because (a) there is nowhere else for it to be and (b) that is the
only place it can be if it is to function causally in shaping human behavior.
62
SEARLE Searle versus Durkheim. Reply to Gross
beings. Specifically, what exactly is the relation between the collective ‘we’ and the
singular ‘I’? If I am acting as part of a collective, say by playing a part in an orchestra or
by acting as a member of a political movement, I act only because of our shared collec-
tive intentionality, but all the same, the collective intentionality has to be able to move
my body. How do we get from We are doing this to I am doing what I am doing as part
of our doing what we are doing?3 Durkheim cannot address these issues because he has
an inadequate account of intentionality and its relation to ontology in general.
63
ANTHROPOLOGICAL THEORY 6(1)
two distinctions, as well as his eagerness to insist that sociology is a science on all fours
with physics or any other natural science.
64
SEARLE Searle versus Durkheim. Reply to Gross
when I make a promise to someone. Durkheim fails to see this point because he lacks
the conceptual resources even to state the distinction. Actually, there are four different
types of social motivation, four different kinds of ‘coercion’ that Durkheim runs
together, but you will not understand social reality unless you see the distinctions
between them. Here are examples of the four sorts of cases: (1) I buy a certain type of
shirt because that type is currently fashionable. (2) I pay back the money you loaned me
because I promised I would. (3) I am swept along in an emotional crowd. (4) I surren-
der my wallet to an armed gang. All of these fit Durkheim’s account of social facts
because they all involve ‘coercions’ that are external to the individual. But they are quite
different from each other, and only (2), the obligation to keep the promise, is an example
of an institutional fact. Proposition (2) describes a deontic power deriving from a status
function. Any adequate theory of social ontology has to be able to state these differences.
Durkheim’s cannot.
65
ANTHROPOLOGICAL THEORY 6(1)
both center their accounts on the notion of representation, both recognize the exist-
ence of something like collective representations, both view a key feature of such
representations as being that of imposing new statuses, powers, and meanings on
objects, and both accept that social institutions are composed of individual actors
who act on the basis of this collective imposition of status functions. (pp. 50–1)
I think this paragraph is, at best, grossly misleading. And I have no choice but to refute
it, so to speak, line by line.
Durkheim does not have a well-defined notion of representation, for reasons I stated
in my points 1 and 2 given earlier. The minimal conditions of adequacy on any
account of representations are that all representations have to have specific intentional
contents that can function by way of intentional causation and they have to be entirely
in individual minds. Durkheim meets neither condition. Indeed, as I pointed out
earlier, his ‘collective representations’ cannot exist in the minds of individuals. In my
view, all collective intentionality exists only in the minds of individuals. And it is not
enough for a theorist to use the word ‘representation’, he has to specify exactly the
propositional contents and various other features of the logical structure. Furthermore,
for Durkheim it is not a matter of new statuses and deontic powers. It is a matter, as
he says repeatedly, of ‘coercion’. And as I said earlier, I think that is a misconception.
Moreover, it is misleading to say that for him social institutions are composed entirely
of individual actors, since he thinks that the collective representations are not in the
minds of individual actors. And finally, in a breathtaking misstatement, Gross says that
Durkheim recognizes that ‘social institutions are composed of individual actors who
act on the basis of this collective imposition of status functions ’ (p. 51, my emphasis). The
concepts of ‘status function’ and the ‘collective imposition of status functions’ are well-
defined technical notions in my theory. The collective imposition of status functions
requires constitutive rules, with their deontology and so on. Durkheim never got close
to having any such notions. The proof that Durkheim has no concept of status func-
tions is that he lumps together things that are, in my sense, status functions (such as
certain social obligations) with things that are not (crowd enthusiasms). And, as
Durkheim says repeatedly, for him the essence of social facts is that ‘they are external
to the individual and endowed with a power of coercion’.
Gross’s attempt to identify my views with Durkheim’s reminds me of Wittgenstein’s
remark: If you wrap up different kinds of furniture – chairs, tables and so on – in enough
wrapping paper, you can make them all look the same shape.
66
SEARLE Searle versus Durkheim. Reply to Gross
the ultimate foundation for much of social reality, if one need be identified, likely
rests not on collective intentionality alone, but as well on something like the capacity
of human beings to grow accustomed – if only for limited stretches of time – to domi-
nation by powerful others. (pp. 52–3)
But why does he think that collective intentionality cannot derive from the fact that
people have grown accustomed to something, even to ‘domination by powerful others’?
Notice the implied contrast between ‘collective intentionality’ and ‘the capacity of
human beings to grow accustomed . . . to domination by powerful others’. But in fact
there is no contrast. Much collective intentionality is simply inherited, and many,
perhaps most, of the institutions we grow up in – families, money, private property, the
stock market and summer vacations – are simply accepted uncritically. Collective
intentionality need not be arrived at by a process of deliberation. And sometimes the
collective intentionality benefits ‘powerful others’.
Actually, I do discuss these issues but, of course, there is much more to be said. In
The Construction, I mention the fact that people often accept systems of status functions
that are manifestly not in their interest. There is a general problem of accounting for
why people accept social institutions at all, and a special problem of accounting for why
people accept social institutions that are not in their interest. The two standard accounts
that I am familiar with seem to me inadequate. The first is the classical economists’
account, according to which the fact that people voluntarily accept institutions shows
that they believe them to be in their interests. That is, people ‘reveal’ their real prefer-
ences by their acceptance behavior. The second argument is that they are oppressed or
deceived in various ways, that some sort of ‘false consciousness’ is responsible and that
people really have ‘adaptive’ preferences. No doubt these analyses will work for certain
special cases, but they do not capture the complexity of the phenomena. I am not at all
confident that we have adequate conceptual tools for posing this question intelligently
for a large number of cases – for example, the tendency of educated white middle-class
American women a few decades ago to accept status roles in society which did not allow
them full scope to attain their potential. It is common among feminists to say that the
situation of white women was forced on them by males, but I doubt very much that this
is the whole story. Many of them accepted their status in society enthusiastically.
I see these as interesting areas for sociological investigation, but it is important to
understand exactly what is the nature of the status functions being investigated, and that
is what my work is addressed to. It is not in conflict with the analysis of particular
patterns of acceptance and rejection. Rather, it is an attempt to describe exactly what is
accepted and rejected. To put this point succinctly: Gross thinks that collective inten-
tionality is in contrast with habitual acceptance. It is not. One (powerful) source of
collective intentionality is habitual acceptance.
Finally, in an argument I find puzzling, Gross takes me to task for not answering a
set of questions that I am not in any sense attempting to answer. These have mostly to
do with the historical character of the development of institutions. He points out
correctly that narratives, metaphors, and what I have elsewhere called ‘dramatic
categories’ play an essential role in social ontology. I entirely agree with this point, but
it is not in conflict with anything that I have said either in the article under discussion
or in The Construction of Social Reality. The point I am making is that you will not have
67
ANTHROPOLOGICAL THEORY 6(1)
any sense of why the dramatic categories are so important unless you can see how they
impinge on social ontology. And in order to see that, you have to have an adequate onto-
logical theory. He writes, ‘But those of us who think that institutions are also composed
of individuals drawing on metaphors, narratives, tropes, and genres might wonder
whether Searle’s attempt to reduce institution-composing beliefs down to constitutive rules
results in a loss of analytic power’ (p. 54, my emphasis). This passage contains a mis-
understanding. Just as in his previous objection he makes a false opposition between
habitual acceptance and collective intentionality, so here he makes a false opposition
between constitutive rules and the modes of their expression. Status functions may derive
from and be expressed by means of metaphors, narratives and all the rest. There is
nothing ‘reductionist’ in my account. But at some point it is necessary to lay bare the
logical structure which results from the metaphors and narratives and that is what I am
trying to do. He cites, for example, historical works by Charles Taylor and Margaret
Somers. But such books do not offer competing accounts to The Construction of Social
Reality. On the contrary, I regard them as complementary, because they are trying to
describe the historical evolution of specific institutional structures. I am trying to
describe the underlying logical structures. Historical accounts tell how these logical
structures developed. He says ‘It is not clear to me that narratives of the kind that interest
Somers and Taylor can be reduced down to a series of constitutive rules taking the form “X
counts as Y in context C” ’ (p. 54, my emphasis). But why should they be so reduced ?
Reductionism, which was never my aim, would require that nothing is going on in the
creation and evolution of human institutions except the imposition and development of
status functions according to constitutive rules, and I make no such claim. Think of the
development and eventual abolition of the institution of slavery in the United States.
All sorts of things went on, including the Civil War. In order to show a genuine disagree-
ment, he would have to show that accounts using narratives, metaphors and so forth are
actually inconsistent with my account and that they can account for facts that I cannot
account for. But he has not done that. I make a strong claim: at the end of the day, when
the historical story has been told, we are left with a framework of constitutive rules,
status functions, and institutional facts. As I see it, the story of how we got there provides
non-competing answers to different sets of questions, not competing answers to the set
of questions that I am asking. You will not understand the role of the historicist account
of the development of institutions unless you perceive the fundamental underlying
ontology of the institutions which develop.
I am, by the way, less optimistic than he about the favorable outcome of ‘building a
bridge between the sociology of culture and literary theory’ (p. 53), because the literary
theory I am familiar with offers very feeble accounts of language. My article ‘Literary
Theory and its Discontents’ (Searle, 1994) articulates some of my misgivings. But I
express doubt about a specific state of development of a subject, not doubts about the
existence of the subject matter or the fruitfulness of its study.
Finally, Gross talks as though there were well-defined disciplines of sociology and
‘analytic philosophy’ and he apparently thinks I am trying to start a ‘dialogue’ between
the two. This is a misconception. Most ‘analytic philosophers’ have no interest whatever
in social ontology. I am fascinated by the subject and see it as essential to an overall
account of ontology. I need all the help I can get, from sociology or elsewhere. At the
level at which we are trying to work, the distinctions between academic disciplines
68
SEARLE Searle versus Durkheim. Reply to Gross
simply do not matter. If sociology is defined as the study of society, and ‘analytic phil-
osophy’ is at least partly defined as the study of logical structures, then there is a necess-
ary overlap between sociology and analytic philosophy, because society has a logical
structure. That is the central claim of The Construction of Social Reality, and, by the way,
I find nothing remotely like it in Durkheim.
Notes
1 See, for example, North (1990: 3); (cited by D’Andrade in this issue).
2 For example, the statements:
(a) Bush was elected the president of the USA
and
(b) the president of the USA is identical with Laura’s husband
do not imply
(c) Bush was elected Laura’s husband.
3 This issue is discussed in Searle (1990).
References
Durkheim, Emile (1938) The Rules of Sociological Method. Glencoe, IL: The Free Press.
Durkheim, Emile (1953) ‘Individual and Collective Representations’, in Emile
Durkheim Sociology and Philosophy, pp. 1–34. Glencoe, IL: The Free Press.
Durkheim, Emile (1993 [1893]) The Division of Labor in Society, trans. G. Simpson.
New York: Free Press.
North, Douglass (1990) Institutions, Institutional Change and Economic Performance
(Political Economy of Institutions and Decisions). Cambridge: Cambridge University
Press.
Searle, John R. (1990) ‘Collective Intentions and Actions’, in P. Cohen, J. Morgan,
and M.E. Pollack (eds) Intentions in Communication, pp. 401–15. Cambridge, MA:
MIT Press. (Reprinted in Consciousness and Language [2002] Cambridge:
Cambridge University Press.)
Searle, John R. (1994) ‘Literary Theory and its Discontents’, New Literary History
25(3): 637–67.
69