The three forms of theoretical knowledge (Pierre Bourdieu)

Objectivist knowledge can only grasp the objective structures of the social world, and the objective truth of primary experience from which explicit knowledge ofthese structures is absen

PIERRE BOURDIEU

The social world may be subjected to three modes of theoretical knowledge,

each of which implies a set of (usually tacit) anthropological theses The only

thing these modes of knowledge have in common is that they all stand in

opposition to practical knowledge The mode of knowledge we shall term

phenomenological (or, if one prefers to speak in terms of currently active schools,

&dquo;interactionist&dquo; or &dquo;ethnomethodological&dquo;) makes explicit primary rience of the social world: perception of the social world as natural and self-evident is not self-reflective by definition and excludes all interrogation aboutits own conditions of possibility At a second level, objectivist knowledge (of which the structuralist hermeneutic constitutes a particular case) constructs

expe-the objective relations (e.g economic or linguistic) structuring not only tices but representations of practices and in particular primary knowledge,

prac-practical and tacit, of the familiar world, by means of a break with this primary

knowledge and, hence, with those tacitly assumed presuppositions which conferupon the social world its self-evident and natural character Objectivist

knowledge can only grasp the objective structures of the social world, and the

objective truth of primary experience (from which explicit knowledge ofthese structures is absent), provided it poses the very problem doxic experience

of the social world excludes by definition, namely the problem of the (specific)

conditions under which this experience is possible Thirdly, what we mightrefer to as praxeological knowledge is concerned not only with the system

of objective relations constructed by the objectivist form of knowledge, butalso with the dialectical relationships between these objective structures and thestructured dispositions which they produce and which tend to reproduce

them, i.e the dual process of the internalization of externality and the

exter-nalization of internality This knowledge presupposes a break with the

objec-tivist form of knowledge, that is, it presupposes investigation into the tions of possibility and, consequently, into the limits of the objectivistic view-

condi-point which grasps practices from the outside, as a fait accompli, rather than

construct their generative principle by placing itself inside the process of their

accomplishment.

The praxeological form of knowledge may appear to be a regression to the

phenomenological mode of knowledge, while the implied critique of

objecti-vism is liable to be confused with the critique of scientific objectification mulated by naive humanism in the name of lived experience and the rights

for-of subjectivity This is so because it is the product of a double theoretical

movement of translation : in effect, it carries out a second reversal of the blematic that objective science of the social world, seen as a system of objec-

pro-tive relationships, constituted by posing those problems which practical

expe-rience and the phenomenological analysis of that experience exclude Just

as objectivist knowledge poses the problem of the conditions of possibility

of practical experience, thereby demonstrating that this experience is defined,

fundamentally, by the fact that it does not pose this problem, so gical knowledge sets objectivist knowledge on its feet by posing the problem

praxeolo-of the conditions praxeolo-of possibility of this problem (theoretical, but also social

con-ditions) and, at the same time, makes it apparent that objectivist knowledge

is defined, fundamentally, by the fact that it excludes this problem Being

set up in opposition to practical perception of the social world, objectivist knowledge is distracted from the task of constructing the theory of practical knowledge of the social world Praxeological knowledge does not cancel out

the gains accruing from objectivist knowledge, rather it conserves and

trans-cends them by integrating that which this knowledge had to exclude in order

to obtain them

We must pause for a moment on what is objectivism’s field par excellence,that of semiology Just as Saussure postulates that language is an autono- mous object, irreducible to its concrete actualizations, that is to the speech-

behaviour it makes possible, so Panofsky establishes that what he calls, following Alois Riegl, Kunstwollen, in other words, roughly, the objective

meaning of a work 1, is no more reducible to the artist’s &dquo;will&dquo; than it is to the

&dquo;will of the age&dquo; or to the lived experiences which the work arouses in the

spec-tator In so doing, both Saussure and Panofsky carry out, with regard to

speech, that particular form of behaviour, and to works of art, those

particu-lar products of action, the operation which builds objectivist science by

build-ing a system of objective relations that are as irreducible to the practiceswithin which they are realized and manifested as they are to the intentions

of the subjects, and to any awareness these may have of its constraints or itslogic Saussure shows that the true medium of communication between twoagents is not speech, as an immediate datum grasped in its observable mate-

riality, but language, as the structure of objective relations making both the

1 "That which ’presents itself’, not to us, but objectively, as the ultimate and definitivemeaning of the artistic phenomenon" (E Panofsky, "Der Begriff des Kunstwollens",Zeitschrift für Aesthetik und allgemeine Kunstwissenschaft 14, 1920, 321-339)

production and decoding of speech possible Similarly, Panofsky shows that

iconological interpretation treats the tangible properties of the work of art,with the affective experiences it arouses, as mere &dquo;cultural symptoms&dquo;, whichonly fully yield up their meaning to a reading armed with the cultural codethe creator himself has &dquo;involved&dquo; in his work

Immediate &dquo;comprehension&dquo; presupposes an unconscious decoding ration which can only be perfectly adequate where the competence which one

ope-of the agents engages in his practice or in his works is identical to that

objec-tively engaged by the other agent in his perception of this practice or work;

in other words, in the particular case in which the coding - in the sense of the

transformation of a subjective meaning into a practice or a work - coincideswith the symetrical decoding operation Immediate &dquo;comprehension&dquo;, a de-

coding act that does not recognize itself as such, is only possible (and only really

accomplished) in the particular case where the historical code which makes the(unconscious) act of decoding possible, is completely mastered (as a cultivated

disposition) by the perceiving agent and coincides with the code which has (as

a cultivated disposition) made the production of the perceived practice or

work possible Partial or total misunderstanding is the rule in all other cases,the illusion of immediate comprehension leading to illusory comprehension,

that of ethnocentrism, in the sense of a code interference: in short, when itssole cognitive tool is what Husserl termed the &dquo; intentional transfer into the

Other&dquo;, even the most &dquo;comprehensive&dquo; interpretation is liable to amount

to no more than a particularly irreproachable form of ethnocentrism

As the heirs to an intellectual heritage, that of linguistics, whose conditions

of production they are not always able to reproduce, structuralist

anthropo-logists have all too often contented themselves with literal translations of

linguistic terms dissociated from the structure from which they derived theiroriginal meaning, sparing themselves the trouble of undertaking their own

epistemological reflection on the conditions and the limits of the validity

of the transposition of the Saussurian construction It is noteworthy, forexample, that, with the exception of Sapir, who was predisposed by his dualformation as linguist and anthropologist to raise the problem of the rela-tionship between culture and language, no anthropologist has attempted

to bring out all the implications of the homology (which Leslie White is tually alone in formulating explicitly) between two oppositions, language

vir-and speech on one side culture, and behaviour or works on the other side.Objectivism states that immediate communication is possible if, and only if,

the agents are objectively disposed in such a way that they associate the same

meaning with the same sign (speech, practice or work) and the same sign withthe same meaning or, to put it another way, if they are objectively disposed

in such a way that, in their coding and decoding operations, i.e in their tices and their interpretations, they both refer to one and the same system

prac-of constant relations, independent of individual consciousness or wills andirreducible to their execution in the form of practices or works (code or cipher).

In so doing, objectivism does not deny the phenomenological analysis of

primary experience of the social world and of the immediate comprehension

of speech or actions: it merely sets the limits of its validity by establishingthe particular conditions within which it is possible and which phenomeno-

logical analysis leaves out of account The social sciences have, necessarily,

to quote Husserl, &dquo;a thematics with a consistently dual orientation, a matics consistently linking theory of the scientific field with a theory of theknowledge of that theory&dquo; 2; in other words, epistemological reflection on

the-the conditions of possibility of the anthropological sciences forms an gral part of the anthropological sciences That is so firstly because a sciencewhich has as its very object that which makes the science possible, such as

inte-language or culture, can only constitute itself by the constitution of its own

conditions of possibility; but it is also because complete knowledge of theconditions of the science, that is, of the operations whereby this science acquires symbolic mastery of a language, a myth or a rite, implies the knowledge of prac-tical comprehension: the practical knowledge accomplishes the same oper-ations, though in absolute ignorance of the general and particular conditionswithin which it is possible and which confer its particularity upon it

We have only to examine the theoretical operations whereby Saussure builds

up linguistics as a science, by treating language as an autonomous object,

distinct from its materializations in speech, in order to reveal the sitions implicit in any form of knowledge which treats practices or works

presuppo-as symbolic facts to be decoded and, more generally, which treats them as

accomplished facts rather than as practices Although one could invokethe existence of dead languages or of mutism in old age as demonstrating that

it is possible for speech to disappear while language remains preserved, although language faults reveal language as constituting the objective norms

underlying speech (were it otherwise, any language fault would modify the

language and there would be no language faults), speech appears to be the tion of language, as much from an individual as from a collective point of view,

condi-since language cannot be apprehended outside of speech, because language

is learnt by means of speech, and because speech lies at the origin of vations in and transformations of language But the priority of the two pro-

inno-cesses mentioned is merely chronological; when one leaves the field of dual or collective history, as does objectivist hermeneutics, in order to inquire

indivi-into the logical conditions of decoding, the relationship is turned on its head :language is the condition of the intelligibility of speech, that is the mediation

which, ensuring the identity of the associations of sounds and concepts

ope-rated by the senders and receivers, guarantees mutual comprehension So,

from this point of view, that of intelligibility, speech is the product of guage 3 It follows that, because it is developed from the strictly intellec-

lan-2 E Husserl, Logique formelle et logique transcendentale, Paris, Presses Universitaires de

France, 1965, p 52.

3 F de Saussure, Cours de linguistique générale, Paris, Payot, 1960, 37-38

tualist point of view, that of decoding, Saussurian linguistics gives priority

to the structure of signs, that is, to the relations between them, to the detriment

of their practical functions, which are never reducible, as structuralism tacitly

assumes, to functions of communication or knowledge: those practices

apparent-ly most strictly oriented towards functions of communication for the sake of

com-munication (the phatic function) or communication for the purposes of ledge, such as feasts and ceremonies, ritual exchanges or, in a wholly different

know-field, the circulation of scientific information, are always more or less openly

oriented towards political or economic functions

Structuralist linguistics bases the construction of the structural properties

of the message as such, that is to say, as a system, on the assumption of animpersonal and interchangeable sender and receiver and on the ignorance

of the functional properties that each message owes to its utilization within

a certain social(>, structured interaction In fact, we know well that the bolic interactions within any group depend, not only on the structure of theinteraction group within which they occur 4, but also on the social struc-

sym-tures within which the interacting agents are situated (e.g the class structure):

consequently, it is probable that a measurement of symbolic exchanges whichwould enable us to distinguish, with Chapple and Coon 5, those who only originate, those who only respond and those who respond to the sending ofthe first group while originating with regard to the second group, would reveal,

both on the level of a society in its entirety and inside a circumstantial group,the dependence of the structure of symbolic power relations upon the struc-

ture of political power relations The perfect competition model is just as

unrealistic here as it is elsewhere, the market in symbolic goods also having

its monopolies and its structures of domination

In short, the moment one shifts from the structure of language to the tions it fulfills, that is, to the uses agents really make of it, one sees that know-

func-ledge of the code alone permits only a very imperfect mastery of the linguisticinteractions actually carried out; as Luis Prieto observes, the meaning of a

linguistic element depends at least as much on extra-linguistic as on linguistic

factors, that is, on the context and situation in which it is employed It is

as if, in the class of significates abstractly corresponding to a speech sound,

the receiver &dquo;selected&dquo; the one that seemed to him to be compatible with the

circumstances, such as he perceives them 6 Which is another way of saying

that the reception (and doubtless the emission too) largely depends on the

objective structure of the relations between the objective positions in the

so-cial structure of the interacting agents (e.g competitive relations, objectively

4 S Moscovici and M Plon, "Les situations-colloques : Observations théoriques et rimentales", Bulletin de psychologie, jan 1966, pp 701-722.

expé-5 E D Chapple and C S Coon, Principles of anthropology, London, Jonathan Cape,

1947, p 283.

6 L J Prieto, Principes de noologie, Paris, Mouton, 1964, and J C Pariente, "Vers un

nouvel esprit linguistique", Critique, 1966, 334-358.

antagonistic relations or power and authority relationships, etc.), for it is this

structure which determines the form assumed by the interactions observedwithin a particular conjuncture.

Nothing demonstrates better the inappropriateness of the theory of tice haunting linguistic (and also anthropological) structuralism than itsinability to integrate, into this theory, all that pertains to execution, as Saus-

prac-sure puts it The foundations of this inability reside in the incapacity to

think of speech and, more generally, of practice otherwise than as execution 1

Objectivism constructs a theory of practice (as execution), but only as a

nega-tive sub-product or, one might say, as a refuse immediately thrown away,left over from the construction of language or culture as systems of objective

relations So, with the aim of delimiting, within language facts, the &dquo;field

of language&dquo; and of isolating &dquo;a well defined object&dquo;, &dquo;an object capable of

being studied seperately&dquo;, &dquo;with a homogeneous nature&dquo;, Saussure rejects

the &dquo;physical aspect of communication&dquo;, that is, speech as a pre-constructed object, liable to obstruct the construction of language; then within the &dquo;speech circuit&dquo;, he isolates what he terms the &dquo;executive aspect &dquo;, that is, speech as a

constructed object, defined as the actualization of a certain meaning within

a particular combination of sounds, which he finally eliminates by stating that

&dquo;execution is never carried out by the collectivity&dquo;, but is &dquo;always individual&dquo;

Thus, the same concept, that of speech, is divided by theoretical constructioninto a preconstructed datum, which is immediately observable and the very one

against which the operation of theoretical construction is carried out, and a

constructed object, the negative product of the operation whereby language as

such is constituted or, better, which produces the two objects by producing

the conflicting relationship within which and by which they are defined It

would be easy to show that the construction of the concept of culture - inthe sense of cultural anthropology -

or of social structure (in Brown’s sense and that of social anthropology) also implies the construction

Radcliffe-of a notion of conduct as execution which coexists with the primary notion

of conduct as simple behaviour taken at face value The extreme confusion

of debates on the relationship between &dquo;culture&dquo; (or &dquo;social structures&dquo;)

and conduct usually arises out of the fact that the constructed meaning ofconduct and its implied theory of practice lead a kind of clandestine existenceinside the discourse of both the defenders and the opponents of cultural

anthropology In fact, the most virulent opponents of the notion of

&dquo;cul-7 "Neither is the psychological part of the circuit wholly responsible: the executive side

is missing, for execution is never carried out by the collectivity Execution is always

indivi-dual, and the individual is always its master: I shall call the executive side speaking (parole)"

(F de Saussure, Course in general linguistics, New York, Philosophical Library, 1959, p 13)

The most explicit formulation of the theory of speech as execution is certainly found in the work

of Hjelmslev, who clearly reveals the various dimensions of the Saussurian opposition between

language and speech, the former being institutional, social and "rigid", the other beingexecutive, individual and "non-rigid" (L Hjelmslev, Essais linguistiques, Copenhagen,Nordisk Sprog-og Kulturforlag, 1959, 79)

ture&dquo;, such as Radcliffe-Brown, can only set over a naive realism against

the realism of the ideas which turn &dquo;culture&dquo; into a transcendent and autonomous

reality, which obeys only its own internal laws e The implicit state of its theory

of practice is what protects objectivism against the only really decisive cism, that which would be aimed at its theory of practice, the generator of allthose metaphysical aberrations on the &dquo;locus of culture&dquo;, on the mode ofexistence of the &dquo;structure&dquo; or on the unconscious finality of the history

criti-of systems, not to mention the too famous &dquo;collective consciousness&dquo; 9.Short of constructing practice other than negatively, that is, as execution,

objectivism is condemned either only to record regularities, ignoring the whole

8 "Let us consider what are the concrete, observable facts with which the social

anthro-pologist is concerned If we set out to study, for example, the aboriginal inhabitants of a

part of Australia, we find a certain number of individual human beings in a certain natural environment We can observe the acts of behaviour of these individuals, including of course

their acts of speech, and the material products of past actions We do not observe a

"cul-ture", since that word denotes, not any concrete reality, but an abstraction, and as it is

com-monly used a vague abstraction But direct observation does reveal to us that these humanbeings are connected by a complex network of social relations I use the term "social struc-

ture" to denote this network of actually existing relations" (A R Radcliffe-Brown, "On

social structure", Journal of the Royal Antropological Institute of Great Britain and Ireland

70, 1940, pp 1-12) The reason for the extreme confusion surrounding debates on the notion of culture probably lies in the fact that most authors place — if only in order to oppose them — concepts of very different epistemological status, such as culture and society or the

individual or conduct, etc., on the same level The imaginary dialogue on the notion of

culture presented by Clyde Kluckhohn and William H Kelly (cf C Kluckhohn and W H.Kelly, "The concept of culture", pp 78-105 in: R Linton (ed.), The science of man in the world crisis, New York, Columbia University Press, 1945) gives a more summary, though

livelier image of this debate than that to be found in A L Kroeber and C Kluckhohn’swork, Culture: A critical review of concepts and definitions, Cambridge, Mass., Harvard

University Press, 1952, Papers of the Peabody Museum of American Archaeology and

Ethno-logy 67 (1) Leach has observed that, despite their apparent opposition, Malinowski and Radcliffe-Brown at least agree to consider each "society" or each "culture" (in their respec-

tive vocabularies) as a "totality made up of a number of discrete, empirical ’things’, of rather diverse kinds —

e g groups of people, ’institutions’, customs" or also as "an rical whole made up of a limited number of readily identifiable parts", the comparison

empi-between different societies having the purpose of examining whether the "same kinds of

parts" are to be found in all cases (E R Leach, Rethinking anthropology, London, Athlone

Press, 1961, p 6)

9 If we except those rare authors who confer on the notion of conduct a meaning that

is rigorously defined by the operation constituting it as opposed to "culture" (for example,

H D Laswell, who states that "if an act conforms to culture then it is conduct, if not, it is

be-haviour", H D Lasswell, "Collective autism as a consequence of culture contact", Zeitschriftfür Sozialforschung 4, 1935, pp 232-247) without drawing any conclusions from it, most ofthose who employ the opposition propose epistemologically discordant definitions of culture

or of conduct, opposing a constructed object to a preconstructed datum, leaving the place

of the second constructed object, namely practice, in the sense of execution, empty: thus —

and this is far from the worst example — Harris opposes "cultural patterns" to "culturallypatterned behaviours", as "what is constructed by the anthropologist" and "what members

of a society observe or impose upon others" (M Harris, "Review of selected writings ofEdward Sapir, language, culture and personality", Language 27 (3), 1951, 288-333)

question of the principle of their production, or to reify abstractions, by

treating objects constructed by science - be they &dquo;culture&dquo;, &dquo;structures&dquo;,

&dquo;social classes&dquo;, &dquo;modes of production&dquo;, etc - as autonomous realities,

endowed with social efficacity, capable of acting as subjects responsible forhistorical actions or as a power capable of constraining practices Although

it has the merit of rejecting the coarser forms of the realism of ideas, the

hypothesis of the unconscious nonetheless tends to mask the contradictions

arising out of the uncertainties of the theory of practice which &dquo;structuralanthropology&dquo; accepts, if only by omission, and even worse, it may permitthe restoration - in the apparently secularized form of a structure that isstructured without the aid of any structuring principle - of the old entelechies

of social metaphysics Unless, of course, one assumes, along with Durkheim,

that none of the implicit rules constraining subjects &dquo;are to be found intheir entirety in their applications by individuals, since they may even existwithout actually being applied&dquo; 10, and consequently that the rules havethe transcendent and permanent existence that Durkheim ascribes to all col-lective &dquo;realities&dquo;, it is impossible to escape the coarsest naiveties of legalism,

which believes practices to be the product of obedience to norms, except by playing on the multiple meanings of the word rule: most often used in the sense

of a social norm, expressly stated and explicitly recognized, as the moral or

juridical law, sometimes in the sense of a theoretical model, a construction

developed by science in order to explain practices, the word is also used,

exceptionally, in the sense of a scheme (schème) (or a principle) that is

im-manent in practice, which should be considered implicit rather than cious, merely in order to signify that it exists in a practical state, in thepractice of agents, and not in their consciousness

uncons-One has only to re-read the following paragraph, from the preface of thesecond edition of Structures elenrerrtaires de la pat-eiiti (Elementary structures

of kinslrip) dealing with the distinction between &dquo;preferential&dquo; and

&dquo;pres-criptive systems&dquo;, in which one may assume that the terms norm, rule or model

are used with particular care: &dquo;Conversely, a system which recommends

mar-riage with the mother’s brother’s daughter may be called prescriptive even

if the rule is seldom observed, since what it says must be done The question

of how far and in what proportion the members of a given society respect

the norm is very interesting, but a different question to that of where this

society should properly be placed in a typology It is sufficient to

acknow-ledge the likelihood that awareness of the rule inflects choices ever so little

in the prescribed direction, and that the percentage of conventional marriages

is higher than would be the case if marriages were made at random, to be able

to recognize what might be called a matrilateral ‘oper-ator’ at work in this

society and acting as a pilot: certain alliances at least follow the path which it

10 E Durkheim, Les règles de la méthode sociologique, Paris, Presses Universitaires

de France, 1956,

charts out for them, and this suffices to imprint a specific curve in the

genea-logical space ~( No doubt there will be not just one curve but a great number of

local curves, merely incipient for the most part, however, and forming closedcycles only in rare and exceptional cases But the structural outlines whichemerge here and there will be enough for the system to be used in making

a probabilistic version of more rigid systems the notion of which is completely

theoretical and in which marriage would conform rigorously to any rule thesocial group pleases to enunciate.&dquo; 11 This passage, as indeed the whole preface,

is written in the language of norms, while Structural anthropology is written inthe language of models, or if one prefers, of structures; this vocabulary is not

entirely absent here, since the system of physico-mathematical metaphors

on which the central passage is founded (&dquo;operator&dquo;, &dquo;certain alliances&dquo;

&dquo;follow the path which it charts out for them&dquo;, &dquo;curvature&dquo; of the gical space&dquo;, &dquo;structures&dquo;) evokes the logic of the theoretical model and the

&dquo;genealo both declared and repudiated -

equivalence of model and norm : &dquo;A

pre-ferential system is prescriptive when envisaged at the model level, a tive system must be preferential when envisaged on the level of reality.&dquo; 12

prescrip-But for those who remember the passages in Structural anthropology on the

relationship between language and kinship (e.g &dquo; ’Kinship systems’, like

’phonemic systems’, are built by the mind on the level of unconscious thought &dquo;13)and the imperious flatness with which &dquo;cultural norms&dquo; and all the &dquo;rationali-zations&dquo; or &dquo;secondary arguments&dquo; produced by the natives were rejected infavour of &dquo;unconscious tructures&dquo;, not to mention those passages where

the universality of the rule lying at the origins of exogamy is affirmed, theconcessions made here to &dquo;awareness of the rule&dquo;, and the dissociation from theserigid systems, whose notion is completely theoretical, may come as a surprise,

as may this other passage taken from the same preface: &dquo;It is nonetheless true

that the empirical reality of so-called prescriptive systems only takes on itsfull meaning when related to a theoretical mode worked out by the nativesthemselves prior to ethnologists.&dquo;11; or again: &dquo;Those who practise them

know fully that the spirit of such systems cannot be reduced to the tautological proposition that each group obtains its women from ’givers’ and gives its

daughters to ’takers’ They are also aware that marriage with the matrilateral

cross cousin (mother’s brother’s daughter) provides the simplest illustration

of the rule, the form most likely to guarantee its survival On the other hand, marriage with the patrilateral cross cousin (father’s sister’s daughter)would violate it irrevocably&dquo; 15 One must mention, here, a passage in which

11 C Lévi-Strauss, The elementary structures of kinship, London, Social science backs, 1969, p 33 (my italics)

Wittgenstein enumerates all the questions evaded by structural anthropology

and doubtless, more generally, by all intellectualism, which transfers the tive truth established by science into a practice which excludes the disposition

objec-which would make it possible to establish this truth 16 : &dquo;What do I call therule by which he proceeds? The hypothesis that satisfactorily describes his use of

words; or the rule which he looks up when he uses signs; or the one which hegives us in reply if we ask him what his rule is? But if observation does not

enable us to see any clear rule, and the question brings none to light ? For hedid indeed give me a definition when I asked him what he understood by ’N’;

but he was prepared to withdraw and alter it So, how am I to determine therule according to which he is playing? He does not know it himself Or, to

ask a better question: What meaning is the expression ’the rule according to

which he acts’ supposed to have left in it here?&dquo; 11 To consider regularity,

i e what recurs with a certain statistically measurable freguency, as the product

of a consciously laid-down and consciously respected regulation (so having

to explain both their genesis and their effectiveness), or else as the product

of the ullconsciolls regulation of some mysterious cerebral and social

mecha-nism, is to slip from the model of reality to the reality of the model: &dquo;Takethe example of the difference between ’the train is regularly two minutes late’and ’as a rule the train is two minutes late’: [ ] in the latter case it is suggest-

ed that the fact that the train is two minutes late is the result of a policy or

plan [ ] I Rules relate to plans and policies, while regularities do not [ ] To

claim that there ought to be rules in natural language amounts to claiming

that roads ought to be red because they correspond to the red lines on a map&dquo; 18.All sociological statements should be preceded by a sign announcing &dquo;it is

as if&dquo; and should function in the same way as quantifiers in logic, which wouldcontinually remind us of the epistemological status of the constructed concepts

of objective science Everything conspires to encourage the reification of

concepts, beginning with the logic of ordinary language, which is inclined to

infer the substance from the substantive or to award to concepts the power

to act in history in the same way as the words designating them act in the tences of historical discourse, that is as historical subjects As Wittgenstein

sen-remarked, one has only to slip from the adverb &dquo;unconsciously&dquo;

(&dquo;uncons-ciously I have a toothache&dquo;) to the substantive &dquo;unconscious&dquo;, or to a

cer-tain usage of the adjective &dquo;unconscious&dquo; (as in &dquo;I have an unconscious ache&dquo;) in order to produce prodigies of metaphysical profundity 19 Simi-

tooth-16 This is an unwarranted transfer of the same type as that which, according to Ponty, generates the intellectualist and the empiricist errors in psychology ( M Merleau-

Merleau-Ponty, La structure du comportement, Paris, Presses Universitaires de France, 1949, esp p.

124, 135)

17 L Wittgenstein, Philosophical investigations, Oxford, Basil Blackwell, 1963,

pp 38-39.

18 P Ziff, Semantic analysis, New York, Cornell University Press, 1960, p 38.

19 L Wittgenstein, Le cahier bleu et le cahier brun, études préliminaires aux

investi-gations philosophiques, Paris, Gallimard, 1965, pp 57-58.

larly, one can observe the theoretical (and political) effects capable of being

engendered by the personification of collectives (in such sentences as &dquo;thebourgeoisie thinks that &dquo; or &dquo;the working class rejects &dquo;) which amounts

to an assertion of the existence of a group or class &dquo;collective consciousness&dquo; :

by crediting groups or institutions with dispositions that can only arise inindividual consciousness, even if they are the product of collective conditions,such as the awakening of consciousness of class interests, one gets out of ana-

lyzing these conditions and those, in particular, which determine the degree

of objective and subjective homogeneity of the group under consideration andthe degree of consciousness among its members

The paralogism underlying legalism consists in implicitly placing in theconsciousness of individual agents the theoretical knowledge which can only

be constructed and conquered against practical experience; in other terms,

it consists in conferring the value of an anthropological description upon a

theoretical model constructed in order to account for practices The theory

of action as simple execution of a model (in the dual sense of norm and ofscientific construction) is only one example among many of the imaginary anthropology engendered by objectivism when, taking, as Marx puts it, &dquo;the

things of logic for the logic of things&dquo;, it turns the objective meaning of tices or works into the subjective purpose of the activity of the producers ofthese practices or works, with its impossible homo economicus subjecting his

prac-decision-making to rational calculation, its actors carrying out roles or acting

in conformity with models, or its speakers selecting from among phonemes.

It is necessary to go beyond methodical objectivism, which constitutes a

necessary phase in all research, as a tool facilitating the break with primary

experience and as an instrument for the construction of objective relations

To escape from the realism of the structure, which treats systems of objectiverelations as substances by converting them into wholes already constitutedoutside of the history of the individual and the history of the group, it is bothnecessary and sufficient to pass from the opus operatum to the modus operandi,

from statistical regularity or from algebraic structure to the principle of the

production of this observed order: the construction of the theory of practice

or, more precisely, of the mode of generation of practices, is the condition ofthe construction of an experimental science of the dialectic of internality andexternality, that is, of the ititet-iializatioti of externality and of the extertiali.:a-

tiof2 of internality The structures of a particular type of environment (e.g.

the material conditions of existence characteristic of a class condition), which

may be grasped empirically in the form of the regularities associated with

a socially structured environment, produce habitus, systems of durable

dis-positions 2°, structured structures predisposed to function as structuring ures, i.e as the principle of the generation and structuration of practices andrepresentations Consequently, these can be objectively &dquo;regulated&dquo; and

struct-&dquo;regular&dquo; without in any way being the product of obedience to rules,

objec-tively adapted to their purposes without presupposing any conscious aiming

of ends and an express mastery of those operations leading to these ends and,

being all this, collectively orchestrated without being the product of a or’s orchestrating action

conduct-Even when they appear to be determined by the future, that is, by the

expli-cit - and explicitly stated - purpose of a project or plan, the practices produced

by the habitus, as the generating principle of strategies enabling one to copewith unforeseen and ever-changing situations, are determined by the implicit

anticipation of their consequences; being determined by the past conditions

of the production of their principle of production, they always tend to

repro-duce the objective structures of which they are, in the last analysis, the product.

Thus, for example, in the interaction between two agents or groups of agents

possessing the same habitus (say A and B), it is as if the actions of each of them(say, al for A) were organized in relation to the reactions they would call forth

in any agent possessing the same habitus (say bl, B’s reaction to al) in such a

way that they objectively anticipate the reaction which these reactions callforth in turn (say a2, the reaction to bl) Nothing could be more naive,

however, than to accept the teleological description according to which eachaction (say, al) was designed to make possible the reaction to the reaction

it provoked (say a2 as reaction to bl) The habitus generates a sequence of

&dquo;moves&dquo; which are objectively organized as strategies without in any way being

the product of a true strategic intention (which would suppose, for example,

that they be perceived as one strategy among several possible strategies).

We cannot exclude the possibility that the habitus’ responses may be

accom-panied by a strategic calculation tending to carry out, quasi-consciously,

what the habitus carries out in another manner, namely an estimate of thechances based on the transformation of the past effect into anticipated futureeffect These responses are nonetheless primarily related to a field of object-

ive potentialities, immediately contained within the present, things to bedone or not to be done, to be said or not said, which, as opposed to the future

as &dquo;absolute possibility&dquo; (absolute Möglichkeit), in Hegel’s sense, projected

by the pure project of a &dquo;negative liberty&dquo;, has an urgency and a claim toexistence excluding all deliberation Symbolic, that is, conrentional and condi-tional stimuli, which only act upon agents conditioned to perceive them, tend

to impose themselves unconditionally and necessarily when inculcation of

20 The word "disposition" seems particularly appropriate for expressing what is

cover-ed by the concept of habitus (defined as a system of dispositions): firstly, it expresses theresult of an organizing action, having a meaning very close to such words as structure; further-

more, it designates a manner of being, an habitual state (in particular, concerning the body)and, especially, a predisposition, a tendency, a propensity or an inclination.

the arbitrary abolishes the arbitrariness of the inculcation and of the

signi-fications inculcated: the world of emergencies, of goals already achieved, of

objects possessing a &dquo;permanently teleological character&dquo;, as Husserl puts it,

such as tools, of paths already marked out, of values transformed into things,

which is that of practice can allow only a conditional freedom - liberet siliceret - rather similar to that of the magnetic needle which, as Leibniz ima-

gined it, actually enjoyed pointing northwards One regularly observes a very

close relationship between scientifically constructed objective probabilities (e g.opportunities of access to higher education or to museums, etc.) and subject-

ive aspirations (&dquo;motivations&dquo;) : this is not so because agents consciously adjust their aspirations to a precise evaluation of their chances of success - theway a gambler might modify his bets as a function of perfect information

regarding his chances of winning -

as we assume implicitly when, forgettingthe &dquo;it is as if&dquo;, we act as if the game theory or the calculus of probabilities,both of them constructed against spontaneous dispositions, amounted to

anthropological descriptions of practice Completely reversing the tendency

of objectivism, we can, on the contrary, seek in the rules of the scientific

cons-truction of probabilities or strategies, not an anthropological model of

prac-tice, but rather a negative description of the implicit tendencies of the neous strategy or statistics, which they necessarily imply, since they are expli- citly constructed against these implicit tendencies (e.g the propensity to

sponta-ascribe an exaggerated importance to primary experiences) Unlike thescientific calculus of probabilities that is based on controlled experiments

and on data established according to precise rules, the subjective evaluation

of a specific action’s chances of success in a specific situation brings into play

a whole body of semi-formalized wisdom, dicta, commonplaces, ethical

pre-cepts (&dquo;that’s not for us&dquo;) and, more profoundly, the unconscious principles

of the ethos, a general and transposable disposition which, being the product

of a learning dominated by a specific type of objective regularity, determines

&dquo;reasonable&dquo; or &dquo;unreasonable&dquo; behaviour for any agent subject to these

regularities 21 &dquo;We are no sooner acquainted with the impossibility of

satisfying any desire&dquo;, said approximately Hume, in his Treatise on human

nature, &dquo;than desire itself vanishes&dquo; And Marx in the Gl1mdrisse:

&dquo;What-ever I am, if I have no money to travel, then I have no need - in the sense

of a real need to travel -

capable of being satisfied Whatever I am, if Ifeel an urge to study but I have no money to pay for my studies, then I have no

urge to study, that is no effective, true urge.&dquo; Practices may be objectively

21 "We call this subjective, variable probability — which sometimes excludes doubt and engenders certainty sui generis and which, at other times appears as no more than a vague

glimmer — philosophical probability because it refers to the exercise of the higher facultywhereby we comprehend the order and the rationality of things All reasonable men have

a confused notion of similar probabilities; this then determines, or at least justifies, thoseunshakable beliefs we call common sense" (A Cournot, Essai sur les fondements de la connais-

les caractères de la philosophique, Paris, Hachette, 1922, 70)

adjusted to objective chances without the agents having to carry out the slightest

calculation, nor even a more or less conscious estimate of the chances of success:

so, it is as if the a posteriori or e.~ post probability of an event, which is known as

a result of past experience, would determine the a priori or ex ante probability subjectively ascribed to it Because the dispositions durably inculcated by the

objective conditions (which science perceives through statistical regularities asprobabilities objectively attached to a group or a class) gives rise to aspirationsand practices that are objectively compatible with these objective conditions and,

to some extent, preadapted to their objective requirements, the most

impro-bable events are excluded, either without even being examined, as able, or at the cost of a double negation tending to make a virtue out of necessity

unthink-by refusing what is anyway refused and loving the inevitable The veryconditions of the production of the ethos, a virtue fumed into necessity, are

such that the anticipations arising out of it tend to ignore the restriction to

which the validity of any calculus of probabilities is subject, namely thatthe conditions of the experiment should not have been modified Unlikescientific estimates, which are corrected, following each experiment, according

to rigourous rules, practical estimates ascribe a disproportionate weight to

primary experiments: the characteristic structures of a determinate type ofconditions of existence, through the mediation of the economic and socialnecessity which they bring to bear on the relatively autonomous universe of

family relationship, or better, through the mediation of specifically familialmanifestations of this external necessity (e g taboos, worries, lessons in moral-

ity, conflicts, tastes, etc.), produce the habitus structures which, in turn, rate the perception and appreciation of all further experience Finally, as

gene-a result of the effect of hysteresis necessarily entailed in the logic of the genesis

of habitus, practices are always exposed to negative sanctions, hence to a

&dquo;secondary negative reinforcement&dquo;, when the environment with which

they are in fact confronted differs too widely from the environment to which

they are objectively adjusted It is understandable, in the same logic, thatgeneration conflicts oppose, not age classes separated by natural properties,

but classes of habitus produced according to different modes of generation:

by instilling different definitions of what is impossible, possible, probable

and certain, the conditions of existence cause one group to experience as

natu-ral or reasonable the same practices or aspirations which the other group findsunthinkable or scandalous, and vice versa.

In other words, one must abandon all those theories which, explicitly or

implicitly, treat practice as a mechanical reaction, directly determined by

antecedent conditions and entirely reducible to the functioning of

pre-estab-lished mechanisms, &dquo;models&dquo;, &dquo;norms&dquo; or &dquo;roles&dquo;; if not, one is supposed

to assume that these mechanisms exist in infinite number, as the fortuitous

configurations of stimuli capable of releasing them from the outside, thus being

condemned to the kind of grandiose and desperate enterprise undertaken by

the anthropologist who, armed with fine positivist courage, recorded 480