2

A SCIENCE THAT MAKES TROUBLE

Q. Let's start with the most obvious questions. Are the social
sciences, and in particular sociology, really sciences? Why do you feel
the need to claim scientificity?
A. Sociology seems to me to have all the properties that define a science.
But to what extent? That's the question. And the answer that can be given
will vary greatly from one sociologist to another. I would simply say that
there are many people who say and believe that they are sociologists and
whom I find it hard to recognize as sociologists. In any case, sociology long
ago emerged from its prehistory, the age of grand theories of social
philosophy, with which lay people often still identify it. All sociologists
worthy of the name agree on a common heritage of concepts, methods and
verification procedures. The fact remains that, for obvious sociological
reasons, sociology is a very dispersed discipline (in the statistical sense), in
several respects. That's why it gives the impression of being a divided
discipline, closer to philosophy than to the other sciences. But that's not the
problem. If people are so pernickety about the scientific nature of
sociology, that's because it's a troublemaker.

Q. Aren't you led to ask yourself questions that arise objectively for
the other sciences although there the scientists don 't have to raise them
concretely for themselves?
A. Sociology has the unfortunate privilege of being constantly con-
fronted with the question of its status as a science. People are infinitely more
demanding than they are towards history or ethnology, not to mention
geography, philology or archaeology. Sociology is constantly called into
question and constantly calls itself, and the other sciences, into question.
And that makes people imagine there's a sociological imperialism: just
what is this science, still in its infancy, that takes upon itself to question the
other sciences? I'm thinking, of course, of the sociology of science. In fact,
however, sociology does no more than ask the other sciences the questions
that arise particularly acutely for itself. If sociology is a critical science,
that's perhaps because it is itself in a critical position. Sociology is an
awkward case, as the phrase goes. We know, for example, that it is said to
have been responsible for the events of May 1968. What people object to is

Interview with Pierre Thuillier, La Recherche, 112, June 1980: 738-43

 A SCIENCE THAT MAKES TROUBLE 9

not just its existence as a science, but its right to exist at all--especially at the
present time, when some people who unfortunately have the power to
succeed in doing so, are working to destroy it, while at the same time they
build up an edifying 'sociology'. at the Institut Auguste Comte or Sciences
Po. 1 All this is done in the name of science, with the active complicity of
some 'scientists' (in the trivial sense of the word).

Q. Why is sociology particularly a problem?

A. Why? Because it reveals things that are hidden and sometimes
repressed, like the correlation between educational achievement, which is
identified with 'intelligence', and social origin, or more precisely, the
cultural capital inherited from the family. These are truths that the
epistemocrats - that's to say a good number of those who read sociology
and those who finance it - don't like to hear. Another example: when you
show that the scientific world is the site of a competition, oriented by the
pursuit of specific profits (Nobel prizes and others, priority in discoveries,
prestige, etc.), and conducted in the name of specific interests (interests that
cannot be reduced to economic interests in their ordinary form, which are
therefore perceived as 'disinterested'), you call into question a scientific
hagiography which scientists often take part in and which they need in
order to believe in what they do.

Q. Right: so sociology is seen as aggressive and embarrassing. But

irhy does sociological discourse need to he 'scient(fic'? Journalists ask
embarrassing questions too. hut they don't claim to he scient(fic. Why
is it crucial that there should he a frontier between sociology and
critical journalism?
A. Because there is an objective difference. It's not a question of vanity.
There are coherent systems of hypotheses, concepts and methods of
verification, everything that is normally associated with the idea of science.
And so, why not say it's a science, if it is one? And then, something very
important is at stake: one of the ways of disposing of awkward truths is to
say that they are not scientific, which amounts to saying that they are
'political', that is, springing from 'interest', 'passion', and are therefore
relative and relativizable.

Q. ff sociology is asked the question of its own scient(ficity, is that

also because it dei,eloped rather later than the other sciences?
A. Certainly. But that ought to show that this 'late development' is due to
the fact that sociology is an especially difficult, an especially improbable,
science. One of the major difficulties lies in the fact that its objects are stakes
in social struggles- things that people hide, that they censor, for which they
are prepared to die . That is true of the researcher himself, who is at stake in
his own objects. And the particular difficulty of doing sociology is often due
10 SOCIOLOGY IN QUESTION

to the fact that people are afraid of what they will find . Sociology confronts
its practitioner with harsh realities~ it disenchants. That's why, contrary to
what is commonly thought, both inside and outside the discipline, it offers
none of the satisfactions that adolescents often seek in political commit-
ment. From that point of view, it is at the opposite end of the scale from the
so-called 'pure' sciences which, like art, and especially music, the 'purest'
art, are no doubt to some extent refuges into which people withdraw in
order to forget the world, universes purged of everything that causes
problems, like sexuality or politics. That's why formal or formalistic minds
generally produce wretched sociology.

Q. You show that sociology intervenes on socially important

questions. That raises the question of its 'neutrality', its 'objectivity'.
Can the sociologist remain above the fray, in the position of an
impartial observer?
A. The particularity of sociology is that it takes as its object fields of
struggle - not only the field of class struggles but the field of scientific
struggles itself. And the sociologist occupies a position in these struggles:
first as the possessor of a certain economic and cultural capital, in the field
of the classes; then, as a researcher endowed with a certain specific capital in
the field of cultural production and, more precisely, in the sub-field of
sociology. He always has to bear this in mind, in order to try to allow for
everything that his practice, what he sees and does not see, what he does and
does not do (for example, the objects he chooses to study), owes to his social
position. That's why, for me, the sociology of sociology is not one
'specialism' among others, but one of the primary conditions for a scientific
sociology. It seems to me that one of the main causes of error in sociology
lies in an unexamined relationship to the object -· or, more precisely, in
ignorance of all that the view of the object owes to the point of view, that is,
to the viewer's position in the social space and the scientific field .
One's chances of contributing to the production of truth seem to me to
depend on two main factors, which are linked to the position one occupies -
the interest one has in knowing and making known the truth ( or conversely,
in hiding it, from oneself and others), and one's capacity to produce it. As
Bachelard so neatly put it, 'There is no science but of the hidden.' The
sociologist is better or worse equipped to dis-cover what is hidden,
depending on how well armed he is scientifically - how well he uses the
capital of concepts, methods and techniques accumulated by his pre-
decessors, Marx, Durkheim, Weber and many others - and also on how
'critical' he is, the extent to which the conscious or unconscious intention
that impels him is a subversive one, the degree of interest he has in
uncovering what is censored and repressed in the social world. And if
sociology does not advance more quickly than it does, like social science in
general, that's perhaps partly because these two factors tend to vary in
inverse ratio.
 A SCIENCE THAT MAKES TROUBLE 11

If the sociologist manages to produce any truth, he does so not despite the
interest he has in producing that truth but because he has an interest in
doing so - which is the exact opposite of the usual somewhat fatuous
discourse about 'neutrality'. This interest may consist, as it does every-
where else, in the desire to be the first to make a discovery and to
appropriate all the associated rights, or in moral indignation or revolt
against certain forms of domination and against those who defend them
within the scientific world. In short, there is no immaculate conception.
There would not be many scientific truths if we had to condemn this or that
discovery ( one only has to think of the 'double helix') on the grounds that
the discoverers' intentions were not very pure.

Q. But in the case of the social sciences, can't 'interest', 'passion ' and
'commitment' lead to blindness, as the advocates of 'neutrality' would
argue?
A. In fact - and this is what makes the particular difficulty of sociology -
these 'interests' and 'passions', noble or ignoble, lead to scientific truth only
in so far as they are accompanied by a scientific knowledge of what
determines them and of the limits that they set on knowledge. For example,
everyone knows that resentment stemming from failure produces lucidity
about the social world only by inducing blindness to the very principle of
that lucidity.
But that's not all: the more advanced a science is, the greater is the capital
of knowledge accumulated within it and the greater the quantity of
knowledge that subversive and critical strategies, whatever their 'motiva-
tions', need to mobilize in order to be effective. In physics, it is difficult to
triumph over an adversary by appealing to authority or (as still happens in
sociology) by denouncing the political content of his theory. There, the
weapons of criticism have to be scientific in order to be effective. In
sociology, on the other hand, every proposition that contradicts received
ideas is open to the suspicion of ideological bias, political axe-grinding. It
clashes with social interests: the interest of the dominant groups, which are
bound up with silence and 'common sense' (which says that what is must be,
or cannot be otherwise); the interest of the spokesmen, the 'loud speakers',
who need simple, simplistic ideas, slogans. That is why sociology is asked to
provide infinitely more proof (which is no bad thing, actually) than is asked
of the spokesmen of 'common sense'. And every discovery of science
triggers off an immense labour of conservative 'critique', which has the
whole social order working for it (budgets, jobs, honours ... and therefore
belief), aimed at re-covering what has been dis-covered.

Q. A moment ago, you cited in the same breath Jfarx, Durkheim and
Weber. You seem to imply that their respective contributions are
cumulative. But infact their approaches are different. How can there
be one single science behind that diversity?
12 SOCIOLOGY IN QUESTION

A. In more cases than one, to enable science to progress, one has to

establish communication between opposing theories, which have often
been constituted against each other. It's not a question of performing the
kind of eclectic pseudo-syntheses that have been so popular in sociology. (It
should be said, in passing, that the denunciation of eclecticism has often
served as an excuse for ignorance - it is so easy and comfortable to wrap
oneself up in a tradition. Marxism, unfortunately, has often been used to
provide this kind of lazy security.) Synthesis presupposes a radical
questioning that leads one to the principle of the apparent antagonism. For
example, in contrast to the usual regression of Marxism towards econ-
omism, which understands the economy only in the restricted sense of the
capitalist economy and which explains everything in terms of the economy
defined in this way, Max Weber broadens economic analysis (in the
generalized sense) to areas that are generally abandoned by economics,
such as religion. Thus, in a magnificent formulation, he characterizes the
Church as the holder of the monopoly of the manipulation of the goods of
salvation. He opens the way to a radical materialism that seeks the
economic determinants (in the broadest sense) in areas where the ideology
of 'disinterestedness' prevails, such as art and religion.
The same goes for the notion of legitimacy. Marx breaks with the
ordinary representation of the social world by showing that 'enchanted'
relationships - such as those of paternalism - conceal power relations.
Weber seems to contradict Marx radically: he points out that membership
of the social world implies a degree of recognition of legitimacy. Sociology
teachers - this is a typical effect of position - note the difference. They prefer
contrasting authors to integrating them. It's more convenient for designing
clear-cut courses: part one Marx, part two Weber, part three myself ... But
the logic of research leads one to move beyond the opposition, back to the
common root. Marx evacuated from his model the subjective truth of the
social world, against which he posits the objective truth of that world as a
system of power relations. Now, if the social world were reduced to its
objective truth as a power structure, if it were not, to some extent,
recognized as legitimate, it wouldn't work. The subjective representation of
the social world as legitimate is part of the complete truth of that world.

Q. In other words, you are trying to integrate into a single conceptual

system theoretical contributions that have been arbitrarily separated
by history or dogmatism.
A. Most of the time, the obstacle standing in the way of concepts,
methods or techniques of communication is not logical but sociological.
Those who have identified themselves with Marx (or Weber) cannot take
possession of what appears to them to be its negation without having the
impression of negating themselves, renouncing their identity (it shouldn't
be forgotten that for many people, to call themselves Marxist is nothing
more than a profession of faith - or a totemic emblem). The same is true of
 A SCIENCE THAT MAKES TROUBLE 13

the relations between 'theoreticians' and 'empiricists', between the sup-

porters of what is called 'fundamental' research and what is called 'applied'
research. That is why the sociology of science can have a scientific effect.

Q. Does it follow that a conservative sociology is bound to remain

supe~ficial?
A. Dominant groups always take a dim view of sociologists, or the
intellectuals who stand in for them when the discipline is not yet
constituted, or cannot function, as in the USSR today. Their interests are
bound up with silence because they have no bones to pick with the world
they dominate, which consequently appears to them as self-evident, a world
that goes without saying. In other words, I repeat, the type of social science
that one can do depends on the relationship one has to the social world, and
therefore on the position one occupies within that world.
More precisely, this relation to the world is translated into the function
that the researcher consciously or unconsciously assigns to his practice and
that governs his research strategies - the objects chosen, the methods used,
and so on . You may make it your goal to understand the social world, in the
sense of understanding for understanding's sake. Or you may seek
techniques that make it possible to manipulate it, in which case you place
sociology in the service of the management (f!he established order. A simple
example will make this clear: the sociology of religion may amount to
research for pastoral purposes that takes as its objects laymen, the social
determinants of church-going or abstention; it then becomes a kind of
market research making it possible to rationalize sacerdotal strategies for
the sale of the goods of 'salvation'. Alternatively it may aim to understand
the functioning of the religious field, of which the laity is only one aspect,
studying for example the functioning of the Church, the strategies through
which it reproduces itself and perpetuates its power- strategies that include
sociological studies (initially carried out by a canon).
A good number of those who describe themselves as sociologists or
economists are social engineers whose function is to supply recipes to the
leaders of private companies and government departments. They offer a
rationalization of the practical or semi-theoretical understanding that the
members of the dominant class have of the social world. The governing elite
today needs a science capable of (in both senses) rationalizing its
domination, capable both of reinforcing the mechanisms that sustain it and
of legitimizing it. It goes without saying that the limits of this science are set
by its practical functions: neither for social engineers nor for the managers
of the economy can it perform a radical questioning. For example, the
science of the Managing Director of the Compagnie Bancaire, which is
considerable, much greater in some ways than that of many sociologists or
economists, is circumscribed by the fact that its sole and unquestioned goal
is the maximization of the profits of that institution. Examples of this
partial 'science' would be the sociology of organizations, or 'political
14 SOCIOLOGY IN QUESTION

science', as taught at the Institut Auguste Comte or Sciences Po, with their
favoured instruments, such as opinion polls.

Q. Doesn't the distinction you dra»' between theoreticians and social

engineers put science in the position of an art for art's sake?
A. Not at all. Nowadays, among the people on whose existence sociology
depends, there are more and more who are asking what sociology is for. In
fact, the likelihood that sociology will disappoint or vex the powers that be
rises to the extent that it successfully fulfils its strictly scientific function .
That function is not to be useful for something, that is to say, for someone.
To ask sociology to be useful for something is always a way of asking it to
be useful to those in power - whereas the scientific function of sociology is
to understand the social world, starting with the structures of power. This
operation cannot be socially neutral, and undoubtedly fulfils a social
function. One reason for that is that all power owes part of its efficacy-and
not the least important part - to misrecognition of the mechanisms on
which it is based.

Q. I'd now like to turn to the question of the relationship between

sociology and the neighbouring sciences. Your book Distinction opens
with the sentence: 'Sociology is rarely more akin to a social
psychoanalysis than when it con.fronts an object like taste.' Then come
statistical tables, and accounts of surveys - but also analyses of a
'literary' type, such as one finds in Balzac, Zola or Proust. How do
these two aspects fit together?
A. The book results from an effort to integrate two modes of knowledge-
ethnographic observation, which can only be based on a small number of
cases, and statistical analysis, which makes it possible to establish
regularities and to situate the observed cases in the uni verse of existing
cases. So you have, for example, the contrasting descriptions of a working-
class meal and a bourgeois meal, each reduced to their pertinent features.
On the working-class side, there is the declared primacy of function, which
appears in all the food that is served: the food has to be 'filling', ' body-
building', as sports are expected to be (weight-training, etc.), to give
strength (conspicuous muscles). On the bourgeois side, there is the primacy
of form, or formality, which implies a kind of censorship and repression of
function, an aestheticization, which is found in every area, as much in
eroticism, functioning as sublimated or denied pornography, as in pure art
which is defined precisely by the fact that it privileges form at the expense of
function. In fact, the analyses that are described as 'qualitative' or, more
pejoratively, 'literary', are essential for understanding, that's to say fully
explaining, what the statistics merely record, rather like rainfall statistics.
They lead to the principle of all the practices observed, in the most varied
areas.
 A SCIENCE THAT MAKES TROUBLE 1.5

Q. To come back to my question, what is your relationship to

psychology, social psychology, etc.?
A. Social science has always stumbled on the problem of the individual
and society. In reality, the divisions of social science in to psychology, social
psychology and sociology were, in my view, constituted around an initial
error of definition. The self-evidence of biological individuation prevents
people from seeing that society exists in two inseparable forms: on the one
hand, institutions that may take the form of physical things, monuments,
books, instruments, etc., and, on the other, acquired dispositions, the
durable ways of being or doing that are incorporated in bodies (and which I
call habitus). The socialized body (what is called the individual or the
person) is not opposed to society; it is one of its forms of existence.

Q. In other ivords, psychology seems to be caught bet1veen, on one

side, biology ( l·vhich provides the.fundamental invariants) and, on the
other, sociology, ivhich studies the way these invariants develop - and
which is therefore entitled to talk about everything, even what is called
private 1(/e,.friendship, love, sexuality, etc.
A. Absolutely. Contrary to the common preconception that associates
sociology with the collective, it has to be pointed out that the collective is
deposited in each individual in the form of durable dispositions, such as
mental structures. For example, in Distinction, 1 try to establish empirically
the relationship between the social classes and the incorporated systems of
classification that are produced in collective history and acquired in
individual history - such as those implemented by taste (the oppositions
heavy/light, hot/cold, brilliant/dull, etc.).

Q. But then , what does the biological or the psychological represent

for the sociologist?
A. Sociology takes the biological and the psychological as a 'given'. And
it tries to establish how the social world uses, transforms and transfigures it.
The fact that a human being has a body, that this body is mortal, raises
difficult problems of social groups. I'm thinking of the book by
Kantorowicz, The King's Two Bodies, which analyses the socially approved
subterfuges resorted to in order to assert the existence of a royalty
transcending the king's real body, which suffers imbecility, sickness,
weakness and death. 'The King is dead, long live the King. ' Ingenious.

Q. You yourself talk qf ethnographic descriptions . ..

A. The distinction between ethnology and sociology is a perfect example
of a spurious frontier. As I try to show in my latest book, Le Sens pratique
[The Logic of Practice], it's a pure product of history (colonial history) that
has no kind of logical justification.
 16 SOCIOLOGY IN QUESTION

Q. But aren't there some very marked d(fferences of attitude? In

ethnology, one has the impression that the observer remains external
to his object and that he can even record appearances whose meaning
he does not knmt' - whereas the sociologist seems to adopt the point of
view of the subjects he studies.
A. In fact, the relation of externality that you describe, which I call
objectivist, is more common in ethnology, probably because it corresponds
to the vision of the outsider. But some ethnologists have also played the
game - the double game - of participation in native representations: the
bewitched or mystical ethnologist. Your remark could even be inverted.
Because they mostly work through the intermediary of interviewers and
never have direct contact with the respondents, some sociologists are more
inclined to objectivism than ethnologists (whose first professional virtue is
to be able to establish a real relationship with their respondents). To that
has to be added class distance, which is no less powerful than cultural
distance. That's why there is perhaps no more inhuman science than that
\ produced in Columbia under the direction of Lazarsfeld, where the
\ distance produced by questioning and by the buffer of the interviewer is
\reinforced by the formalism of blind statistics. You learn a lot about a
science. its methods and its content, when, as in the sociology of work, you
do a kind of job description. For example, the bureaucratic sociologist
treats the people he studies as interchangeable statistical units, subjected to
closed questions that are identical for all, whereas the ethnologist's
informant is a person of standing, sought out for long and detailed
discussions.

Q. So you are opposed to the 'objectivist' approach that substitutes

the model for the reality; but also to Michelet, who wanted to
'resurrect' the past, or Sartre, who rvants to grasp meanings through a
phenomenology that you see as arbitrary?
A. Exactly. For example, since one of the functions of social rituals is to
relieve the agents of everything that we put under the heading of 'subjective
experience' [le vecu], it is particularly dangerous to put in 'subjective
experience' where there is none, in ritual practices, for example. The idea
that there is nothing more generous than to project one's own ·subjective
experience' into the consciousness of a 'primitive', a 'witch' or a
'proletarian' has always seemed to me somewhat ethnocentric. The best the
sociologist can do is to objectify the inevitable effects of the objectification
techniques that he has to use- writing, diagrams, maps, calendars, models,
etc. For example, in The Logic o.f' Practice I try to show that, having failed
to appreciate the effects of their situation as observers and of the techniques
they use to grasp their object, ethnologists have constituted the 'primitive'
as a 'primitive' because they have not been able to recognize in him what
they themselves are as soon as they cease to think scientifically, that is, in
 A SCIENCE THAT MAKES TROUBLE 17

practice. So-called ·primitive' logics are quite simply practical logics, like
the logic we implement to judge a painting or a quartet.

Q. But isn't it possible to rediscover the logic of all that and at the
same time preserve 'subjective experience'?
A. There is an objective truth of the subjective, even when it contradicts
the objective truth that one has to construct in opposition to it. Illusion is
not, as such, illusory. It would be a betrayal of objectivity to proceed as if
social subjects had no representation, no experience of the realities that
science constructs, such as social classes. So one has to rise to a higher
objectivity, which makes room for that subjectivity. Agents have a
subjective experience that is not the full truth of what they do but which is
part of the truth of what they do. Take for example the case of a chairman
who says 'The meeting is suspended' or a priest who says 'I baptize you'.
Why does that language have power? It's not the words that act, through a
kind of magic power. But the fact is that, in particular social conditions,
certain words do have power. They derive their power from an institution
that has its own logic - qualifications, ermine and robes, the professorial
chair, the ritual formulae, the participants' belief, etc. Sociology points out
that it is not the words, or the interchangeable person who pronounces
them, that act, but the institution. It shows the objective conditions that
have to be fulfilled to secure the efficacy of a particular social practice. But
the analysis cannot stop there. It must not forget that, in order for it all to
work, the actor has to believe that he is the source of the efficacy of his
action. There are systems that run entirely on belief and there is no system,
not even the economy, that does not depend to some extent on belief in
order to work.

Q. From the standpoint ofscience, I can see very clearly what you 're
doing. But the result is that you devalue people's 'lived experience'. In
the name of science, you're liable to take away people's reasons for
living. What gives you the right ( (f I can put it that way) to deprive
them of their illusions?
A. I too sometimes wonder if the completely transparent and dis-
enchanted social universe that would be produced by a social science that
was fully developed (and widely diffused, if that could ever be the case)
would not be impossible to live in. I think, all the same, that social relations
would be much less unhappy if people at least understood the mechanisms
that lead them to contribute to their own deprivation. But perhaps the only
function of sociology is to reveal, as much by its visible lacunae as through
its achievements, the limits of knowledge of the social world and so to make
more difficult all forms of prophetic discourse, starting, of course, with the
propheticism that claims to be scientific.
 18 SOCIOLOGY IN QUESTION

Q. Let's turn to the relationship with economics, and in particular

with certain neo-classical analyses like those of the Chicago school. In
fact, the confrontation is interesting because it shows how two different
sciences construct the same objects - fertility, marriage and, more
especially, educational investment.
A. That would be an enormous debate. What may mislead some people is
that, like the neo-marginalist economists, I refer all ~91:ial beha....iou.rs.to .a
?-l ~£~(~sJorm-cl.in.t~x.e.s.t..-o.Liox~.~J.ment. But we are only using the same
words. The interest I am talking about has nothing to do with Adam
Smith's notion of self-interest, an a-historical, natural, universal interest,
which is in fact simply the unconscious universalization of the interest
engendered and presupposed by the capitalist economy. It is no accident
that, to escape from this naturalism, the economists have to appeal to
sociobiology, as Gary Becker does in an article entitled 'Altruism, egoism
and genetic fitness'. For Becker, not only 'self-interest' but also 'altruism
with regard to descendants' and other durable dispositions are to be
explained by the selection over time of the most adaptive features.
In fact, when I say that there is a form of interest or function that lies
behind every institution or practice, I am simply asserting the principle of
sufficient reason which is implied in the very project of 'explaining' (rendre
raison as we say in French) and which is intrinsic to the notion of science.
{This principle postulates that there is a cause or reason making it possible to
'.explain or understand why a given practice or institution is rather than is
•not, and why it is as it is rather than otherwise. This interest or function is in
jno way natural or universal, contrary to what is supposed by the neo-
classical economists, whose homo economicus is simply a universalization of
homo capitalisticus. Ethnology and comparative history show us that the
specifically social magic of institution can constitute almost anything as an
interest and as a realistic interest, i.e. as an investment (in both the economic
and the psychoanalytic senses) that is objectively rewarded, in the more or
less long term, by an economy. For example, the economy of honour
produces and rewards economic dispositions and practices that are
apparently 'ruinous' - because they are so 'disinterested' - and are
consequently absurd, from the point of view of the economics of the
economists. And yet, even the behaviours that are the most wildly irrational
from the standpoint of capitalist economic reason are based on a form of
,.,,. enlightened self-interest (e.g. the interest there is in being 'above suspicion')
and can therefore be studied by an economics. Investment is the disposition
~ to act that is generated in the relationship between a space defined by a
game offering certain prizes or stakes (what 1 call a field) and a system of
dispositions attuned to that game (what I call a habitus) - the 'feel' for the
game and the stakes, which implies both the inclination and the capacity to
play the game, to take an interest in the game, to be taken up, taken in by the
game. You only have to think of the importance, in western societies, of
educational investment - which in France finds its extreme form in the
 A SCIENCE THAT MAKES TROUBLE 19

classes preparing for the grandes ecoles - to realize that the institution is
capable of producing the investment and, in this case, the hyper-
investment, which is the condition of the functioning of the institution. But
the same could be demonstrated about any form of the sacred. The
experience of the sacred presupposes, inseparably, the acquired disposition
which causes sacred objects to exist as such, and the objects which
objectively demand a consecrating approach (that is the role of art in our
societies). In other words, investment is the historical effect of the harmony
between two realizations of the social - in things, through institution, and
in bodies, through incorporation.

Q. lsn 't the kind ofgeneral anthropology that you put forward a way
ofachieving the philosophical ambition ofa system, but with the means
of science?
A. The aim is certainly not to remain eterna11y locked in the totalizing
discourse that was expounded by social philosophy and which is still
common currency nowadays, especially in France, where prophetic
pronouncements still enjoy a protected market. But I think that, in trying to
conform to a very one-sided representation of scientificity, sociologists
have moved into premature specialization. One could cite countless cases in
which artificial divisions of the object, generally following the lines of
administrative demarcations, are the main obstacle to scientific under-
standing. To give an example from an area I know well, there's the
separation of the sociology of culture and the sociology of education; or the
economics of education and the sociology of education. I also think the
science of man inevitably appeals to anthropological theories; that it can
make real progress only on condition that it makes explicit these theories
that researchers always bring in, in a practical form, and which are
generally no more than the transfigured projection of their relation to the
social world.

Note

The Paris Institut d'Etudes Politiques (Fondation Nationale des Sciences Politiques)
[translator].

Further reading

For further discussion see Bourdieu P. ( 1975) 'The specificity of the scientific field and the
social conditions of the progress of reason', Social Science Information, 14 (6): 19-47 [also in
Lemert, C. (ed.) (1981) French Sociology, Rupture and Renewal Since 1968, New York:
Columbia University Press, 257-92; (1975) "Le langage autorise: note sur Jes conditions de
J'efficacite sociale du discours rituel', Actes de la recherche en sciences sociales, 5- 6: 183-90;
( 1980) 'Le mort saisit le vif: les relations entre l'histoire reifiee et l'histoire incorporee', Actes de
la recherche en sciences sociales, 32-3: 3-14.
 3
THE SOCIOLOGIST IN QUESTION

Q. Why do you use a specialjargon- an especially difficult one- that

often makes your texts inaccessible to the lay reader? Isn't there a
contradiction between denouncing the scientist's self-assigned
monopoly and re-creating it in the text that denounces it?
A. Often one only has to let ordinary language speak for itself, to give way
to linguistic laissez-faire, in order to accept unwittingly a whole social
philosophy. The dictionary is charged with a political mythology (I am
thinking for example of all the couples of adjectives: brilliant/serious, high/
low, rare/common, etc.). Devotees of 'common sense', who move in
ordinary language like a fish in water and who, in language as in other
things, have the objectified structures working for them, can (except for
their euphemisms) speak a language that is clear as crystal and freely
denounce the jargon of others. But the social sciences have to win all that
they say against the received ideas that are carried along in ordinary
language and have to say what they have won in a language that is
predisposed to say something quite different. To try to disrupt verbal
automatisms does not mean artificially creating a distinguished difference
that sets the layman at a distance; it means breaking with the social
philosophy that is inscribed in spontaneous discourse. Using one word in
place of another often means effecting a decisive epistemological change -
which may well pass unnoticed.
But it is not a question of escaping from the automatisms of common
sense in order to fall into the automatisms of critical language, with all the
words that have too often functioned as slogans or rallying cries, all the
utterances that serve not to state the real but to paper over the gaps in
knowledge. (That is often the function of concepts with capital letters and
the propositions that they introduce, which are very often no more than
professions of faith whereby one believer recognizes another.) I am
thinking of the 'basic Marxism', as Jean-Claude Passeron calls it, that
flourished recently in France: an automatic language that runs all on its
own, but in neutral, and makes it possible to talk about everything with

The questions discussed in this chapter are those which seemed most
important among those that were most often put to me in various discussions
in the early l980s in Paris (at the Ecole Polytechnique), Lyon (the Universite
Populaire), Grenoble (the Faculte des Lettres), Troyes (the Institut
Universitaire de Technologie) and Angers (Faculte des Lettres)
 THE SOCIOLOGIST IN QUESTION 21

great economy, with a very small number of simple concepts, but without
much thought. The mere fact of conceptualization often exerts an effect of
neutralization and even denial.
Sociological language cannot be either 'neutral' or 'clear'. The word
'class' will never be a neutral word so long as there are classes: the question
of the existence or non-existence of classes is a stake in struggle between the
classes. The work of writing that is necessary in order to arrive at a rigorous
and controlled use oflanguage only rarely leads to what is called clarity, in
other words, the reinforcement of the self-evidences of common sense or
the certainties of fanaticism.
In contrast to the search for literary quality, the pursuit of rigour always
leads one to sacrifice a neat formula, which can be strong and clear because
it falsifies, to a less appealing expression that is heavier but more accurate,
more controlled. Thus the difficulty of a style often comes from all the
nuances, all the corrections, all the warnings, not to mention the reminders
of definitions and principles that are needed in order for the discourse to
bear within itself all the possible defences against hijacking and
misappropriations. Attention to these critical signs is no doubt directly
proportional to the reader's vigilance and therefore his competence - which
means that the warnings are most clearly seen by the reader who needs them
least. At least one can hope that they discourage phrase-mongering and
parroting.
But the need to resort to an artificial language is perhaps more compelling
for sociology than for any other science. In order to break with the social
philosophy that runs through everyday words and also in order to express
things that ordinary language cannot express (for example, everything that
lies at the level of it-goes-without-saying), the sociologist has to resort to
invented words which are thereby protected, relatively at least, from the
naive projections of common sense. These words are secure against
hijacking because their 'linguistic nature' predisposes them to withstand
hasty readings (for example, habitus, which refers to acquired properties,
capital) and perhaps especially because they are inserted, locked, into a
network ofrelationships that impose their logical constraints. For example,
allodoxia, which well expresses something that is difficult to explain or even
think in a few words (the fact of taking one thing for another, of thinking
that something is other than it is, etc.), is bound up in the network of words
from the same root: doxa, doxosopher, orthodoxy, heterodoxy, paradox.
Having said that, the difficulty of transmitting the products of sociol-
ogical research is due much less than people think to the difficulty of
language. An initial cause of misunderstanding lies in the fact that even the
most 'cultured' readers have only a very approximate idea of the conditions
of production of the discourse that they are trying to appropriate. For
example, there is a 'philosophical' or 'theoretical' reading of works in social
science that consists in noting the 'theses', the 'conclusions', independently
of the process of which they are the outcome (i.e. in concrete terms, the
 22 SOCIOLOGY IN QUESTION

empirical analyses, the statistical tables, the indications of method, and so

on). If you read like that, you are reading another book. When I 'condense'
the opposition between the working class and the dominant class into the
opposition between the primacy given to substance (function) and the
primacy given to form, people see a philosophical disquisition, when the
point I'm making is that one group eats beans while the other eats salad,
that differences in consumption, which are nil or very low for underclothes,
are very strong for outer clothes, etc. It is true that my analyses arise from
applying very abstract schemes of thought to very concrete things, the
statistics for the purchase of pyjamas, underpants or trousers. It's not easy
to read statistics on pyjamas while thinking of Kant. .. Everything people
learn at school tends to discourage them from thinking of Kant apropos of
pyjamas or thinking of pyjamas when reading Marx (I say Marx because
you will readily allow me Kant, though in this respect it's much the same
thing).
Add to that the fact that many readers do not know, or reject, the very
principles of the sociological mode of thought, such as the intention of
'explaining the social by the social', in Durkheim's phrase, which is often
perceived as an imperialist ambition. But, more simply, ignorance of
statistics or, rather, lack of familiarity with the statistical mode of thought
leads people to confuse the probable (e.g. the relationship between social
origin and educational achievement) and the certain, the necessary. This
leads to all sorts of absurd accusations, such as the charge of fatalism, or to
misplaced objections, such as the scholastic failure of a proportion of the
offspring of the dominant class, which is in fact a central element in the
statistical mode of reproduction. (A 'sociologist', a member of the Institute,
has devoted a great deal of energy to showing that not all the sons of
Polytechnicians become Polytechnicians!)
But the main source of misunderstanding lies in the fact that, ordinarily,
people hardly ever talk about the social world in order to say what it is, but
almost always to say what it ought to be. Discourse about the social world
is almost always performative: it contains wishes, exhortations, reproaches,
orders, etc. It follows that the sociologist's discourse, though it tries to be
descriptive, has every likelihood of being perceived as performative. Ifl say
that women respond less often than men to questions in opinion polls -and
that the difference becomes more marked as the questions become more
'political' - there will always be someone to complain that I exclude women
from politics. That's because, when I say what is, people hear 'and it's fine
that way'. Similarly, if you describe the working class as it is, you're
suspected of wanting to lock it into what it is, as a destiny, of trying either to
push it down or to exalt it. For example, the observation that, most of the
time, men (and even more so, women) in the culturally most deprived
classes entrust their political choices to the party of their choice, and, as it
happens, to the Communist Party, has been understood as an exhortation
to abandon oneself to the Party. In fact, in ordinary life, people will
 THE SOCIOLOGIST IN QUESTION 23
describe a working-class meal only in order to express wonderment or
disgust, never in order to understand its logic, to explain it, in other words
to secure the means of taking it for lvhat it is. Readers read sociology
through the spectacles of their habitus. And some people will find a
reinforcement of their class racism in the same realist description that
others will suspect of being inspired by class contempt.
In that lies the principle of a structural misunderstanding in the
communication between the sociologist and his reader.

Q. Don't you think that, given the way in which you express yourself,
your only possible readers are intellectuals? Isn't that a limit on the
effectiveness of your work?
A. The sociologist's misfortune is that, most of the time, the people who
have the technical means of appropriating what he says have no wish to
appropriate it, no interest in appropriating it, and even have powerful
interests in refusing it (so that some people who are very competent in other
respects may reveal themselves to be quite obtuse as regards sociology),
whereas those who would have an interest in appropriating it do not have
the instruments for appropriation (theoretical culture, etc.). Sociological
discourse arouses resistances that are quite analogous in their logic and
their manifestations to those encountered by psychoanalytical discourse.
The people who read that there is a very strong correlation between
educational level and museum-going have every likelihood of being
museum-goers, of being art lovers ready to die for the love of art,
experiencing their encounter with art as a pure love, a love at first sight, and
of setting countless systems of defence in the way of scientific object-
ification.
In short, the laws of diffusion of scientific discourse mean that, despite the
existence of relays and mediators, scientific truth is very likely to reach
those who are least disposed to accept it and very unlikely to reach those
who would have most interest in receiving it. Yet one may think that one
would only have to provide the latter with a language in which they
recognized themselves, or rather, in which they felt recognized, that's to say
accepted, justified in existing as they exist (which they are necessarily
offered by all good sociology, a science that, as such, explains things), in
order to induce a transformation of their relationship to what they are.
What needs to be made available to people is the scientific gaze, a gaze
that is at once objectifying and understanding, and which, when turned
back on oneself, makes it possible to accept oneself and even, so to speak,
lay claim to oneself, claim the right to be what one is. I'm thinking of
slogans like the American blacks' 'Black is beautiful' and the feminists'
assertion of the right to the 'natural look'. I have been accused of sometimes
writing pejoratively about those who impose new needs and of thereby
putting forward an ideal image of humanity that is reminiscent of the 'noble
savage' but in a socialized version. In fact, it's not a question of locking
24 SOCIOLOGY IN QUESTION

agents into an 'original social being' treated as a destiny, a nature, but of

offering them the possibility of taking on their habitus without guilt or
suffering. That can be seen clearly in the area of culture, where the sense of
inadequacy often comes from a dispossession that cannot recognize itself as
such. What probably emerges through my way of speaking about
beauticians, dietitians, marriage guidance counsellors and other purveyors
of needs is indignation against that form of exploitation of people's
deprivation which consists of imposing impossible norms and then selling
the means - generally ineffective ones - of bridging the gap between these
norms and the real possibilities of achieving them.
In that area, which is completely ignored by political analysis, although it
is the site of objectively political action, the dominated groups are left to
their own weapons; they are absolutely bereft of weapons of collective
defence in order to confront the dominant groups and their 'poor man's
psychoanalysts'. Yet it would be easy to show that the most typically
political kind of political domination also runs through these channels.For
example, in Distinction, I wanted to start the chapter on the relationship
between culture and politics with a photograph - which, in the end, I didn't
use, fearing that it might be misread - in which two trade union leaders are
seen sitting on Louis XV chairs facing Giscard d'Estaing, who is himself
seated on a Louis XV sofa. That picture pointed out, in the clearest possible
way, through the ways of sitting, of placing the hands, in short, all the body
language, which participant has on his side not only all the culture, that's to
say the furniture, the decor, the Louis XV chairs, but also the ways of using
it, inhabiting it - which one is the possessor of that objectified culture, and
which ones are possessed by that culture, in the name of that culture. If,
when face to face with the managing director, the trade-unionist 'feels
small', that's at least partly because he only has instruments of analysis, and
self-analysis, that are too general and too abstract, which give him no
possibility of understanding and controlling his relation to language and
the body. And this state of abandonment in which the available theories
and analyses leave him is particularly serious - although the state of
abandonment of his wife, in the kitchen of their council flat, faced with the
simpering condescension of the presenters on the commercial radio
stations, is not unimportant - because lots of people speak through him,
because the speech of a whole group passes through his mouth and his
body, and because his reactions, generalized in this way, may have been
determined, without his realizing it, by his horror of long-haired hippies or
intellectuals wearing spectacles.

Q. Doesn't your sociology imply a deterministic view of man? What

role, if any, is left for human freedom?
A. Like every science, sociology accepts the principle of determinism
understood as a form of the principle of sufficient reason. The science which
must give the reasons for that which is thereby postulates that nothing is
 THE SOCIOLOGIST IN QUESTION 25

without a reason for being. The sociologist adds: 'social reason' - nothing is
without a specifically social reason for being. Faced with a statistical table,
he postulates that there is a social factor that explains that distribution, and
if, having found it, there is a residue, then he postulates the existence of
another social factor, and so on. (That's what makes people sometimes
imagine a sociological imperialism: but it's fair enough - every science has
to use its own means to account for the greatest number of things possible,
including things that are apparently or really explained by other sciences.
It's on that condition that it can put real questions to the other sciences -
and to itself - and destroy apparent reasons or raise clearly the problem of
overdetermination.)
Having said that, people are often referring to two quite different things
under the term 'determinism' - objective necessity, implied in reality itself,
and 'experiential', apparent, subjective necessity, the sense of necessity or
freedom. The degree to which the social world seems to us to be determined
depends on the knowledge we have of it. On the other hand, the degree to
which the world is really determined is not a question of opinion; as a
sociologist, it's not for me to be 'for determinism' or 'for freedom', but to
discover necessity, if it exists, in the places where it is. Because all progress
in the knowledge of the laws of the social world increases the degree of
perceived necessity, it is natural that social science is increasingly accused of
'determinism' the further it advances.
But, contrary to appearances, it's by raising the degree of perceived
necessity and giving a better knowledge of the laws of the social world that
social science gives more freedom. All progress in knowledge of necessity is
a progress in possible freedom. Whereas misrecognition of necessity
contains a form of recognition of necessity, and probably the most
absolute, the most total form, since it is unaware of itself as such,
knowledge of necessity does not at all imply the necessity of that
recognition. On the contrary, it brings to light the possibility of choice that
is implied in every relationship of the type 'if X, then Y'. The freedom that
consists in choosing to accept or refuse the 'if' has no meaning so long as
one is unaware of the relationship that links it to a 'then ... '. By bringing to
light the laws that presuppose non-intervention (that's to say, unconscious
acceptance of the conditions of realization of the expected effects), one
extends the scope of freedom. A law that is unknown is a nature, a destiny
(that's true, for example, of the relationship between inherited cultural
capital and educational achievement); a law that is known appears as a
possibility of freedom.

Q. Isn't it dangerous to speak of laws?

A. Yes, undoubtedly. And, as far as possible, I avoid doing so. Those who
have an interest in things taking their course (that's to say in the 'if'
remaining unchanged) see the 'law' (when they see it at all) as a destiny, an
inevitability inscribed in social nature (which gives the iron laws of the
26 SOCIOLOGY IN QUESTION

oligarchies of the neo-Machiavellians, such as Michels or Mosca). In fact, a

social law is a historical law, which perpetuates itself so long as it is allowed
to operate, that's to say as long as those whose interests it serves (sometimes
unknown to them) are able to perpetuate the conditions of its efficacy.
What we have to ask is what we are doing when we state a social law that
was previously unknown (such as the law of the transmission of cultural
capital). One may claim to be fixing an eternal law, as the conservative
sociologists do with the tendency towards the concentration of power. In
reality, science needs to know that it merely records, in the form of
tendential laws, the logic that is characteristic of a particular game, at a
particular moment, and which plays to the advantage of those who
dominate the game and are able, de facto or de Jure, to define the rules of the
game.
Having said that, as soon as the law has been stated, it may become a
stake in struggles - the struggle to conserve, by conserving the conditions of
the functioning of the law, the struggle to transform, by changing these
conditions. Bringing the tendential laws to light is a precondition for the
success of actions aimed at frustrating them. The dominant groups have an
interest in the law, and therefore in a physicalist interpretation of the law,
which pushes it back to the state of an infra-conscious mechanism. By
contrast, the dominated groups have an interest in the discovery of the law
as such, that is, as a historical law, which could be abolished if the
conditions of its functioning were removed. Knowledge of the law gives
them a chance, a possibility of countering the effects of the law, a possibility
that does not exist so long as the law is unknown and operates unbeknown
to those who undergo it. In short, just as it de-naturalizes, so sociology de-
fatalizes.

Q. Isn't an ever more developed knowledge of the social world likely

to discourage all political action aimed at transforming the social
world?
A. Knowledge of the most probable outcome is what makes it possible, by
reference to other ends, to bring about the least probable. By consciously
playing with the logic of the social world one can bring about possible
outcomes that did not seem to be implied by that logic.
True political action consists in making use of knowledge of the probable
so as to strengthen the chances of the possible. It differs from utopianism
which, like magic, attempts to act on the world through performative
discourse. Political action, often more unconsciously than consciously,
expresses and exploits the potentialities inscribed in the social world, in its
contradictions or immanent tendencies. The sociologist - and this is why
people sometimes deplore the absence of the political in his discourse -
describes the conditions that political action has to take account of and on
which it depends for its success or failure (for example, nowadays, the
collective disenchantment of young people). In so doing, he warns against
 THE SOCIOLOGIST IN QUESTION 27

the error that leads one to take the effect for the cause and to see effects of
political action in the historical conditions of its efficacy - though one
should not ignore the effect that political action can have when it
accompanies dispositions that it does not produce and that pre-date it, and
intensifies them by expressing them and orchestrating their manifestation.

Q. I am rather worried about the conclusions that might be drawn,

perhaps based on a misreading.from what you demonstrate about the
nature of opinion. Jsn 't that analysis liable to be somewhat
demob ifiz ing?
A. Let me clarify. Sociology reveals that the idea of personal opinion (like
the idea of personal taste) is an illusion. From this it is concluded that
sociology is reductive, that it disenchants, that it demobilizes people by
taking away all their illusions.
Does that mean that one can only mobilize on the basis of illusions? Ifit is
true that the idea of personal opinion itself is socially determined, that it is a
product of history reproduced by education, that our opinions are
determined, then it is better to know this; and if we have some chance of
having personal opinions, it's perhaps on condition that we know our
opinions are not spontaneously so.

Q. Sociology is both an academic activity and a critical, even a

political, activity. lsn 't there a contradiction there?
A. Sociology as we know it was born, in France at least, from a
contradiction or a misunderstanding. Durkheim was the one who did all
that needed to be done to make sociology exist as a universally recognized
science. When an activity is constituted as a university discipline, the
question of its function and the function of those who practise it no longer
arises. One only has to think of the archaeologists, philologists, medieval
historians, historians of China or classical philosophy, who are never asked
what use they are, what their work is for, who they work for or who needs it.
No one calls them into question and they consequently feel completely
justified in doing what they do. Sociology is not so lucky ... The question of
its raison d'etre is asked increasingly the more it moves away from the
definition of scientific practice that the founders had to accept and impose,
that of a pure science, pure as the purest, most 'useless' and 'gratuitous' of
the academic sciences (papyrology or Homeric studies), those that the most
repressive regimes allow to survive and that serve as a refuge for specialists
from the 'hot' sciences. You know how much work Durkheim had to do to
give sociology this 'pure', purely scientific, 'neutral' image - ostentatious
borrowings from the natural sciences, countless signs of a break with
external functions and politics, such as preliminary definition, etc.
In other words, from the very beginning, sociology has been an
ambiguous, dual, masked science; one that had to conceal and renounce its
own nature as a political science in order to gain acceptance as an academic
 28 SOCIOLOGY IN QUESTION

science. It's no accident that ethnology raises many fewer problems than
sociology.
But sociology can also use its autonomy to produce a truth that no one -
among those who are in position to command or commission it - asks of it.
Through skilful use of the institutional autonomy that it has as a university
discipline, it can find the conditions for epistemological autonomy and try
to offer what no one really asks of it, truth about the social world. It's not
surprising that this socio-logically impossible science, capable of revealing
what ought socio-logically to remain masked, could only arise from false
pretences as to its ends, and that anyone who wants to perform sociology as
a science must constantly reproduce this initial misrepresentation. Larvatus
prodeo.
Truly scientific sociology is a social practice that, socio-logically, ought
not to exist. The best proof of this is that as soon as social science refuses to
be locked into the forced choice between pure science, which can
scientifically analyse objects that have no social importance, and false
science, which manages and caresses the established order, its very social
existence is threatened.

Q. Can't scientific sociology count on the solidarity of the other

sciences?
A. Indeed it can. But sociology, the newcomer among the sciences, is a
critical science, critical of itself and the other sciences and also critical of the
powers that be, including the powers of science. It's a science that strives to
understand the laws of production of science; it provides not means of
domination but, perhaps, the means of dominating domination.

Q. Doesn't sociology try to give a scientific answer to the traditional

problems of philosophy and, to some extent, to eclipse them by a
dictatorship of reason?
A. I think this was true at the beginning. The founders of sociology
explicitly made that their objective. For example, it's no accident that the
first object of sociology was religion. The Durkheimians immediately
attacked what was (at a particular time) the primary instrument for
constructing the world, and especially the social world. I also think that
some traditional questions of philosophy can be re-posed in scientific terms
(that's what I tried to do in Distinction). Sociology as I conceive it consists
in transforming metaphysical problems into problems that can be treated
scientifically and therefore politically. On the other hand, sociology, like all
sciences, constructs itself in opposition to the totalizing ambition of
philosophy, or rather of prophecies, discourses which, as Weber indicates,
claim to offer total answers to total questions, especially on 'life and death'
questions. In other words, sociology was built up with the ambition of
stealing some of the problems of philosophy, but without the prophetic
project that philosophy often set for itself. It broke with social philosophy
 THE SOCIOLOGIST IN QUESTION 29

and all the ultimate questions in which the latter indulged itself, such as the
questions of the meaning of history, progress and decadence, the role of
great men in history, and so on. The fact remains that sociologists
encounter those problems in the most elementary operations of practice,
thro.ugh the way a question is asked, by presupposing, in the very form and
content of their enquiry, that practices are determined by the immediate
conditions of existence or by the whole previous history, and so on. Only if
they are aware of them, and direct their practice appropriately, will they
avoid slipping unawares into the philosophy of history. For example, if you
question someone directly about the social class to which he belongs, or
alternatively if you try to determine his place 'objectively' by questioning
him about his income, his job, his level of education, etc., you are making a
decisive choice between two opposing philosophies of practice and history.
This choice cannot be settled, if it is not addressed as such, by asking both
questions at the same time.

Q. Why are you so harsh on theory, which you almost always seem to
identify with philosophy? You yourself theorize, even if you deny it.
A. What is called theory is generally verbiage fit for manuals.
Theorization is often just a kind of 'manualization', as Raymond Queneau
once put it. To make the play on words clearer I can quote Marx by way of
commentary: 'Philosophy is to the study of the world as onanism is to
sexual intercourse.' If everyone in France knew that, social science would
make a 'great leap forward', as someone once said. As for whether I
theorize, it depends what one means by the word. A theoretical problem
that is converted into a machine for research is set in motion, it becomes in a
sense self-propelling, it is driven as much by the difficulties it brings up as by
the solutions it provides.
One of the secrets of the craft of sociology is to know how to find
empirical objects about which one can really raise general problems. For
example, the question of realism and formalism in art, which, at certain
times and in certain contexts, has become a political question, can be raised,
empirically, in connection with the relation of the working class to
photography, or through analysis of reactions to various types of television
programmes, and so on. But it can equally well be raised, and simul-
taneously too, in connection with frontality in Byzantine mosaics or the
depiction of the Sun King in paintings or historiography. Having said that,
the theoretical problems raised in that way are so profoundly transformed
that the friends of theory would no longer recognize their offspring in them.
The logic of research is this intermeshing of problems in which the
researcher is caught up and which drags him along, often despite himself.
Leibniz constantly complained to Descartes in his Animadversiones that he
expected too much of intuition, insight and intelligence and did not rely
enough on the automatisms of 'blind thought' (he was thinking of algebra,
which would make up for the intermissions of intelligence). What is not
30 SOCIOLOGY IN QUESTION

understood in France, the land of the brilliant essay, the cult of originality
and intelligence, is that method and the collective organization of research
work can produce intelligence, intermeshings of problems and methods
that are more intelligent than the researchers (and also, in a world in which
everyone seeks originality, the only true originality, the one that is not
looked for - I am thinking for example of the extraordinary exception
represented by the Durkheimian school). To be scientifically intelligent is to
place oneself in a situation that generates real problems and real difficulties.
That is what I have tried to do with the research group that I run. A research
group that works is a socially instituted interlocking of problems and ways
of solving them, a network of cross-checks, and, at the same time, a whole
set of productions which, without any imposition of norms or any
theoretical or political orthodoxy, have a family resemblance.

Q. What is the relevance of the distinction between sociology and

ethnology?
A. That distinction is unfortunately inscribed, and probably irreversibly
so, in university structures, that's to say in the social organization of the
university and the mental organization of academics. My work would not
have been possible if I had not tried to hold together some problematics
traditionally regarded as ethnological and other problems traditionally
regarded as sociological. For example, for a number of years ethnologists
have addressed the problem of taxonomies, classifications, a problem that
arises at the crossroads of several traditions in ethnology. Some are
interested in the taxonomies applied in classifying plants, or diseases, etc.
Others are interested in the taxonomies used to organize the social world,
the taxonomy par excellence being the one that defines kinship relations.
This tradition has developed in areas where, because of the relatively
undifferentiated nature of the societies in question, the problem of classes
does not arise. Sociologists, on the other hand, deal with the problem of
classes but without addressing the problem of the systems of classification
used by the agents and their relationship to the objective division of classes.
My work has consisted in bringing together, in a non-scholastic way - I say
that because otherwise it could sound like one of the academic cross-
fertilizations that are performed in lectures - the problem of social classes
and the problem of classification systems, and in asking questions such as
this: don't the taxonomies that we use to classify things and people, to judge
a work of art, a pupil, a hairstyle, clothes, etc. - and thereby to produce
social classes - have some connection with the objective classifications, the
social classes understood (crudely speaking) as classes of individuals linked
to classes of material conditions of existence?
What I am talking about is a typical effect of the division of scientific
labour: there are objective divisions (the division into disciplines, for
example) which, having become mental divisions, function in such a way as
to make certain thoughts impossible. This analysis is an illustration of the
 THE SOCIOLOGIST IN QUESTION 31

theoretical problematic that I have just outlined. Institutional divisions,

which are a product of history, function in objective reality (for example, if
I put three sociologists on an examining board it must be a sociology thesis,
etc.) in the form oflegally sanctioned objective divisions, inscribed in career
paths, etc., and also in people's heads, in the form of mental divisions,
logical principles of division. The obstacles to knowledge are often
sociological obstacles. Having crossed the frontier between ethnology and
sociology, I have been led to ask ethnology all sorts of questions that
ethnology does not ask, and vice versa.

Q. You define social class in terms of volume and structure of capital.

How do you define the kinds of capital? For economic capital, it seems
that you depend entirely on the statistics provided by INSEE, 1 and/or
educational capital on qualifications. Can you really construct classes
on that basis?
A. It's a long-standing debate. I explain my position in Distinction. You
are faced with a choice between either a pure (and simple) theory of social
classes, which is based on no empirical data (position in the relations of
production) and which has practically no capacity to describe the state of
the social structure or its transformations; or empirical studies, like those of
INSEE, which are based on no theory but which provide the only available
data for analysing the division into classes. Personally I have tried to move
beyond what has been treated as a theological opposition between theories
of social classes and theories of social stratification, an opposition that goes
down well in lectures and in 'dialectical materialist' thinking, but which is in
fact merely the reflection of a state of the division of sociological labour. So
I have tried to put forward a theory that is both more complex (taking into
account states of capital that are ignored in classical theory) and
empirically better grounded, but which is obliged to resort to imperfect
indicators such as those provided by INSEE. I am not so nai·ve as to be
unaware that the indicators provided by INSEE, even for example
concerning share ownership, are not good indicators of the economic
capital possessed. That is clear to everyone. But there are cases where
theoretical purism is an alibi for ignorance or the abdication of practice.
Science consists in doing what one does while knowing and saying that it is
all one can do, making clear the limits of validity of what one does.
Having said that, the question you have asked me in fact conceals
another problem. What do people mean when they say, or write, as they
often do: 'What ultimately are social classes in so-and-so's theory?' In
asking a question like that, one is sure of winning the approval of all those
who are convinced that the problem of social classes is resolved and that
one only has to consult the canonical texts (which is very convenient and
economical, when you think about it), and who cast suspicion on all those
who, by continuing to look, betray that they think that not everything has
been found. This strategy of suspicion, which is particularly likely to be
32 SOCIOLOGY IN QUESTION

generated by certain class habitus), is an unbeatable one, and gives much

satisfaction to those who practise it, because it comforts them in their self-
satisfaction. For that reason it seems to me scientifically and politically
obnoxious.
It is true that I have constantly insisted on going back to basics on things
that were regarded as settled. Capital, for example: we all know what that is
.... You only have to read Capital or, better, to read Reading Capital 2 (and
so on). Fine, if it were true, but in my view it isn't, and if there has always
been this gulf between theoretical theory and empirical descriptions (a gulf
that means that people who have only their Marxist spectacles to aid them
are totally helpless when it comes to understanding the historical originality
of the new forms of social conflict, for example those that are linked to the
contradictions resulting from the functioning of the educational system), if
this gulf has always existed, perhaps it is because the analysis of the different
kinds of capital still remained to be done. To get beyond it, it was necessary
to shake up some self-evidences, and not for the pleasure of performing
heretical, and therefore distinctive, readings.
To return to the question of the kinds of capital, I think it's a very difficult
question and I realize, when I tackle it, that I am moving outside the charted
area of established truths, where one is sure of immediately attracting
approval, esteem, and so on. (At the same time, I think that the scientifically
most fruitful positions are often the most risky ones, and therefore the
socially most improbable.) As regards economic capital, I leave that to
others; it's not my area. What concerns me is what is abandoned by others,
because they lack the interest or the theoretical tools for these things,
cultural capital and social capital. Very recently I've tried to set out in
simple terms for didactic purposes what I mean by these notions. I try to
construct rigorous definitions that are not only descriptive concepts, but
means of construction, which make it possible to produce things that one
could not see previously. Take social capital, for example: one can give an
intuitive idea of it by saying that it is what ordinary language calls
'connections'. (It often happens that ordinary language designates very
important social facts; but it masks them at the same time, by the effect of
familiarity, which leads one to imagine that one already knows, that one
has understood everything, and which stops research in its tracks. Part of
the work of social science consists in dis-covering what is both unveiled and
veiled by ordinary language. This means running the risk of being accused
of stating the self-evident, or, worse, of laboriously translating, into a
heavily conceptual language, the basic verities of common sense or the
more subtle and more agreeable intuitions of moralists and novelists.
When, that is, people do not accuse the sociologist of saying things that are
simultaneously banal and untrue, which just goes to show the extra-
ordinary resistances that sociological analysis arouses.)
To return to social capital: by constructing this concept, one acquires the
means of analysing the logic whereby this particular kind of capital is
 THE SOCIOLOGIST IN QUESTION 33

accumulated, transmitted and reproduced, the means of understanding

how it turns into economic capital and, conversely, what work is required
to convert economic capital into social capital, the means of grasping the
function of institutions such as clubs or, quite simply, the family, the main
site of the accumulation and transmission of that kind of capital, and so on.
We are a long way, it seems to me, from common-sense 'connections',
which are only one manifestation among others of social capital. The
'social round' and all that is related in the high-society gossip columns of Le
Figaro, Vogue or lours de France cease to be, as is generally thought,
exemplary manifestations of the idle life of the 'leisure class' or the
'conspicuous consumption' of the wealthy, and can be seen instead as a
particular form of social labour, which presupposes expenditure of money,
time and a specific competence and which tends to ensure the (simple or
expanded) reproduction of social capital. (It can be seen, incidentally, that
some ostensibly very critical discourses miss what is essential, because
intellectuals are not very 'sensitive' to the form of social capital that
accumulates and circulates in 'society' gatherings and tend to sneer, with a
mixture of fascination and resentment, rather than analyse.)
So it was necessary to construct the object that I call social capital -
which immediately brings to light that publishers' cocktail parties or
reciprocal reviewing are the equivalent, in the intellectual field, of the 'social
work' of the aristocracy - to see that high-society socializing is, for certain
people, whose power and authority are based on social capital, their
principal occupation. An enterprise based on social capital has to ensure its
own reproduction through a specific form of labour (inaugurating
monuments, chairing charities, etc.) that presupposes professional skills,
and therefore an apprenticeship, and an expenditure of time and energy. As
soon as this object is constructed, one can carry out genuine comparative
studies, talk to historians about the nobility in the Middle Ages, reread
Saint-Simon and Proust, or, of course, the work of the ethnologists.
At the same time, you are quite right to ask the question. Since what I do
is not at all theoretical work, but scientific work that mobilizes all the
theoretical resources for the purposes of empirical analysis, my concepts
are not always what they ought to be. For example, I constantly raise the
problem of the conversion of one kind of capital into another, in terms that
do not completely satisfy even me. It's an example of a problem that could
not be posed explicitly- it posed itself before one knew it - until the notion
of kinds of capital had been constructed. Practice is familiar with this
problem. In certain games (in the intellectual field, for example, in order to
win a literary prize or the esteem of one's peers), economic capital is
inoperative. To become operational it has to undergo a transmutation.
That's the function, for example, of the 'social work' that made it possible
to transmute economic capital - always at the root in the last analysis - into
nobility. But that's not all. What are the laws governing that conversion?
What defines the exchange rate at which one kind of capital is converted
34 SOCIOLOGY IN QUESTION

into another? In every epoch there is a constant struggle over the rate of
exchange between the different kinds of capital, a struggle among the
different fractions of the dominant class, whose overall capital is composed
in differing proportions of the various kinds of capital. Those who in
nineteenth-century France were called the 'capacities' have a constant
interest in the revaluing of cultural capital with respect to economic capital.
It can be seen - and this is what makes sociological analysis so difficult-that
these things that we take as our object- cultural capital, economic capital,
etc. - are themselves at stake in struggles within the reality that we are
studying and that what we say about them will itself become a stake in
struggles.
Analysis of these laws of conversion is not complete- far from it - and if
there is one person for whom it's a problem, it's myself. Which is fine.
There's a host of questions, very fertile ones I think, that I ask myself or that
are put to me, objections that are raised and that were only made possible
because these distinctions had been made. Research is perhaps the art of
creating fruitful problems for oneself - and creating them for other people.
Where things were simple, you bring out problems. And then you find
yourself facing a much more sticky reality. Of course, I could have
produced one of those courses of Marxism-without-tears on the social
classes that have sold so well in the last few years, under the name of theory,
or even science. Or even sociology - you find yourself dealing with things
that are both suggestive and worrying (I know the effect that what I do has
on the guardians of orthodoxy and I think I also have some idea why it has
that effect, and I'm delighted that it does). The idea of being both suggestive
and worrying is one that suits me fine.

Q. But isn't there something static about the theory of the social
classes that you put forward? You describe a state of the structure
without saying how it changes.
A. What statistical analysis can grasp is a moment, a state of a game with
two, three, four or six players, or whatever. It gives a photograph of the
piles of tokens of various colours that they have won in the previous rounds
and which they will play in the rounds to come. Capital apprehended
instantaneously is a product of history that will produce more history. I'll
simply say that the strategies of the different players will depend on their
resources in tokens, and more specifically on the overall volume of their
capital (the number of tokens) and the structure of this capital, that's to say
the composition of the piles (those who have lots of red tokens and few
yellow ones, that is, a lot of economic capital and little cultural capital, will
not play in the same way as those who have many yellow tokens and few red
ones). The bigger their pile, the more audacious they can be (bluff), and the
more yellow tokens (cultural capital) they have, the more they will stake on
the yellow squares (the educational system). Each player sees the play of the
others, that is, their way of playing, their style, and he derives clues from
 THE SOCIOLOGIST IN QUESTION 35

this regarding their hand, tacitly hypothesizing that the former is a

manifestation of the latter. He may even have direct knowledge of part or
all of the capital of the others (educational qualifications play the role of
calls in bridge). In any case, he uses his knowledge of the properties of the
other players, that is, their strategy, to guide his own play. But the principle
of his anticipations is nothing other than the sense of the game, that is, the
practical mastery of the relationship between tokens and play (what we
express when we say of a property - a garment or a piece of furniture, for
example - that it's 'petit-bourgeois'). This sense of the game is the product
of the progressive internalization of the immanent laws of the game. It's
what Thibaut and Riecken grasp, for example, when their respondents,
questioned about two people who give blood, spontaneously assume that
the person of higher class is free, the person oflower class forced (although
we do not know, and it would be very interesting to know how the
proportion of those who make this assumption varies between upper- and
lower-class respondents).
Obviously the image I have just used is only valid as a didactic device. But
I think it gives an idea of the real logic of social change and gives a sense of
how artificial it is to oppose the static to the dynamic.

Notes

Jnstitut National de la Statistique et des Etudes Economiques [translator].

2 By Louis Althusser [translator].