English 384: Cultural Theory

Reserve Text: From Raymond Williams, Marxism and Literature. Oxford: Oxford U P, 1977

2. Language

A definition of language is always, implicitly or explicitly, a definition of human beings in the world. The received major categories--'world', 'reality', 'nature,' 'human'--may be counterposed or related to the category `language', but it is now a commonplace to observe that all categories, including the category 'language', are themselves constructions in language, and can thus only with an effort, and within a particular system of thought, be separated from language for relational inquiry. Such efforts and such systems, nevertheless, constitute a major part of the history of thought. Many of the problems which have emerged from this history are relevant to Marxism, and in certain areas Marxism itself has contributed to them, by extension from its basic revaluation, in historical materialism, of the received major categories. Yet it is significant that, by comparison, Marxism has contributed very little to thinking about language itself. The result has been either that limited and undeveloped versions of language as a 'reflection' of 'reality' have been taken for granted, or that propositions about language, developed within or in the forms of other and often antagonistic systems of thought, have been synthesized with Marxist propositions about other kinds of activity, in ways which are not only ultimately untenable but, in our own time, radically limiting to the strength of the social propositions. The effects on cultural theory, and in particular on thinking about literature, have been especially marked.

The key moments which should be of interest to Marxism, in the development of thinking about language, are, first, the emphasis on language as activity and, second, the emphasis on the history of language. Neither of these positions, on its own, is enough to restate the whole problem. It is the conjunction and consequent revaluation of each position that remains necessary. But in different ways, and with significant practical results, each position transformed those habitual conceptions of language which depended on and supported relatively static ways of thinking about human beings in the world.

The major emphasis on language as activity began in the eighteenth century, in close relation to the idea of men having

[p 22]

made their own society, which we have seen as a central element in the new concept of 'culture'. In the previously dominant tradition, through all its variations, 'language' and 'reality' have been decisively separated, so that philosophical inquiry w from the beginning an inquiry into the connections betwee these apparently separate orders. The pre-Socratic unity of logos, in which language was seen as at one with the order of world and of nature, with divine and human law, and with reason, had been decisively broken and in effect forgotten. The radical distinction between 'language' and 'reality', as between 'consciousness' and 'the material world', corresponding to actual and practical divisions between 'mental' and 'physical' activity, had become so habitual that serious attention seeme naturally concentrated on the exceptionally complicated consequent relations and connections. Plato's major inquiry into language (in the Cratylus) was centred on the problem of the correctness of naming, in which the interrelation of 'word' an 'thing' can be seen to originate either in `nature' or in 'convention'. Plato's solution was in effect the foundation of idealist thought: there is an intermediate but constitutive realm, which is neither 'word' nor 'thing' but `form', 'essence', or 'idea'. The investigation of either 'language' or 'reality' was then always, at root, an investigation of these constitutive (metaphysical) forms.

Yet, given this basic assumption, far-reaching inquiries into the uses of language could be undertaken in particular and specialized ways. Language as a way of indicating reality could be studied as logic. Language as an accessible segment of reality especially in its fixed forms in writing, could be studied as grammar, in the sense of its formal and 'external' shape. Finally within the distinction between language and reality, language could be conceived as an instrument used by men for specific and distinguishable purposes, and these could be studied in rhetoric and in the associated poetics. Through prolonged academic and scholastic development, these three great branches of language study-logic, grammar, and rhetoric-though formally associated in the medieval trivium, became specific and eventually separated disciplines. Thus though they made major practical advances, they either foreclosed examination of the form of the basic distinction between 'language' and 'reality', or determined the grounds, and especially the terms, in which such an examination might be made.

[p 23]

This is notably the case with the important medieval concept of sign, which has been so remarkably readopted in modern linguistic thought. 'Sign', from Latin signum, a mark or token, is intrinsically a concept based on a distinction between 'language' and 'reality'. It is an interposition between 'word' and 'thing' which repeats the Platonic interposition of 'form', 'essence', or 'idea', but now in accessible linguistic terms. Thus in Buridan 'natural signs' are the universal mental counterparts of reality and these are matched, by convention, with the 'artificial signs' which are physical sounds or letters. Given this starting-point, important investigations of the activity of language (but not of language as an activity) could be undertaken: for example, the remarkable speculative grammars of medieval thought, in which the power of sentences and of the modes of construction which underlay and complicated simple empirical notions of 'naming' was described and investigated. Meanwhile, however, the trivium itself, and especially grammar and rhetoric, moved into relatively formal, though immensely learned, demonstrations of the properties of a given body of 'classical' written material. What was later to be known as 'literary study', and from the early seventeenth century as 'criticism', developed from this powerful, prestigious, and limited mode.

Yet the whole question of the distinction between 'language' and 'reality' was eventually forced into consciousness, initially in a surprising way. Descartes, in reinforcing the distinction and making it more precise, and in demanding that the criterion of connection should be not metaphysical or conventional but grounded in scientific knowledge, provoked new questions by the very force of his scepticism about the old answers. It was in response to Descartes that Vico proposed his criterion that we can have full knowledge only of what we can ourselves make or do. In one decisive respect this response was reactionary. Since men have not in any obvious sense made the physical world, a powerful new conception of scientific knowledge was ruled out a priori and was, as before, reserved to God. Yet on the other hand, by insisting that we can understand society because we have made it, indeed that we understand it not abstractly but in the very process of making it, and that the activity of language is central in this process, Vico opened a whole new dimension.

It was and is difficult to grasp this dimension, initially

[p 24]

because Vico embedded it in what can be read as a schematic account of the stages of language development: the notorious three stages of divine, heroic, and human. Rousseau, repeating these three stages as 'historical' and interpreting them as stages of declining vigour, gave a form of argument to the Romantic Movement-the revival of literature as a revival of the 'original', 'primal' power of language. But this at once obscured the newly active sense of history (specializing it to regeneration and ultimately, as this failed, to reaction) and the newly active sense of language, which in being specialized to literature could be marked off as a special case, a special entity, a special function, leaving the 'non-literary' relations of language to reality as conventional and as alienated as before. To take Vico's three stages literally, or indeed as 'stages' at all, is to lose sight, as he did, of the dimension he had opened. For what was crucial, in his account of language, was that it emerged only at the human stage, the divine being that of mute ceremonies and rituals and the heroic that of gestures and signs. Verbal language is then distinctively human; indeed, constitutively human. This was the point taken up by Herder, who opposed any notion of language being 'given' to man (as by God) and, in effect, the apparently alternative notion of language being 'added' to man, as a special kind of acquisition or tool. Language is then, positively, a distinctively human opening of and opening to the world: not a distinguishable or instrumental but a constitutive faculty

Historically this emphasis on language as constitutive, like the closely related emphasis on human development as culture, must be seen as an attempt both to preserve some idea of the generally human in face of the analytical and empirical procedures of a powerfully developing natural science, and to assert an idea of human creativity, in face of the increased understanding of the properties of the physical world, and of consequently causal explanations from them. As such this whole tendency was in constant danger of becoming simply a new kind of idealism-'humanity' and 'creativity' being projected as essences-while the tendencies it opposed moved towards a new kind of objective materialism. This specific fission, so fateful in all subsequent thought, was in effect masked and ratified by a newly conventional distinction between 'art' (literature)-the sphere of 'humanity' and 'creativity'-and 'science' ('positive knowledge')-the knowable dimension of the physi-

[p 25]

cal world and of physical human beings within it. Each of the terms--'art', 'literature' and 'science', together with the associated 'culture' and with such a newly necessary specialization 'aesthetic' and the radical distinction between 'experience and 'experiment'-changed in meaning between the early eighteenth and early nineteenth centuries. The resulting conflicts and confusions were severe, but it is significant that in the new situation of the nineteenth century the issues were never really joined on the ground of language, at any radical level, though was precisely in relation to language that the newly conventional distinctions most needed to be challenged.

What happened instead was an extraordinary advance in empirical knowledge of languages, and a wholly remarkable analysis and classification of this knowledge in terms which set some of the basic questions aside. It is impossible to separate this movement from its political history, within the dynamic development of Western societies in a period of extending colonialism. Older studies of language had been largely contained within the model of the dead 'classical' languages (which still effectively determined 'grammar' in both its syntactic and literary senses) and of the 'derived' modern vernaculars. European exploration and colonization, meanwhile, had been dramatically expanding the available range of linguistic material. The critical encounter was between the European and Indian civilizations: not only in available languages but in European contact with the highly developed methods of Indic grammatical scholars; with their alternative body of 'classical' texts. It was as an Englishman in India that William Jones learned Sanskrit and from an observation of its resemblances to Latin and Greek began the work which led to classification of the Indo-European (Aryan ) and other `families' of languages.

This work, based on comparative analysis and classification, was procedurally very close to the evolutionary biology with which it is contemporary. It is one of the major periods of all scholarly investigation, empirically founding not only the major classifications of language families, including schemes of their evolutionary development and relationships, but also, within these schemes, discovering certain 'laws' of change, notably of sound-change. In one area this movement was 'evolutionary' in a particular sense: in its postulate of a proto-language (proto-Indo-European) from which the major 'family' had developed.

[p 26]

But in its later stages it was `evolutionary' also in another sense, Increasing rigour in the study of sound-changes associated one branch of language study with natural science, so that a system of linguistic phonetics marched with physical studies of the language faculty and the evolutionary origins of speech. This tendency culminated in major work in the physiology of speech and in the field significantly designated within this area as experimental psychology.
This identification of language-use as a problem in psychology was to have major effects on concepts of language. But meanwhile within general language-studies there was a new phase which reinforced inherent tendencies to objectivism. What was characteristically studied in comparative philology was a body of records of language: in effect, centrally, the alien written word. This assumption of the defining material of study was already present, of course, in the earlier phase of `classical' language studies: Greek, Latin, Hebrew. But then the modes of access to a wider range of languages repeated this earlier stance: that of the privileged (scientific) observer of a body of alien written material. Methodological decisions, substantially similar to those being developed in the closely related new science of anthropology, followed from this effective situation. On the one hand there was the highly productive application of modes of systematic observation, classification, and analysis. . On the other hand there was the largely unnoticed consequence of the privileged situation of the observer: that he was observing (of course scientifically) within a differential mode of contact with alien material: in texts, the records of a past history; in speech, the activity of an alien people in subordinate (colonialist) relations to the whole activity of the dominant people within which the observer gained his privilege. This defining situation inevitably reduced any sense of language as actively and presently constitutive. The consequent objectivism of fundamental procedure was intensely productive at the level of description, but necessarily any consequent definition of language had to be a definition of a (specialized) philological system. In a later phase of this contact between privileged observer and alien language material, in the special circumstances of North America where hundreds of native American (Amerindian) languages were in danger of dying out after the completion of European conquest and domination, the earlier philological procedures were

[p 27]

indeed, characteristically, found to be not objective enough. imitation of these even more alien languages to the categories of Indo-European philology-the natural reflex of cultural imperialism-was scientifically resisted and checked jar necessary procedures which, assuming only the presence of an alien system, found ways of studying it in its own (intrinsic and structural) terms. This approach was a further gain in scientific description, with its own remarkable results, but at the level of theory it was the final reinforcement of a concept of language as an (alien) objective system.
n

[p 27]

Paradoxically, this approach had even deeper effect through one of the necessary corrections of procedure which followed from the new phase of contact with languages without texts. Earlier procedures had been determined by the fact that a language almost invariably presented itself in specific past texts: finished monologic utterances. Actual speech, even when it was available, was seen as derived, either historically into vernaculars, or practically into speech acts which were instances of the fundamental (textual) forms of the language. Language-use could then hardly ever be seen as itself active and constitutive. And this was reinforced by the political relations of the observer-observed, where the `language-habits' studied, over a range from the speech of conquered and dominated peoples to the `dialects' of outlying or socially inferior groups, theoretically matched against the observer's `standard', were regarded as at most `behaviour', rather than independent, creative, selfdirecting life. North American empirical linguistics reversed one part of this tendency, restoring the primacy of speech in the literal absence of `standard' or `classical' texts. Yet the objectivist character of the underlying general theory came to limit even this, by converting speech itself to --the characteristically persistent word in orthodox structural linguistics. Language came to be seen as a fixed, objective, and in these senses `given' system, which had theoretical and practical priority over what were described as `utterances' (later as `performance'). Thus the living speech of human beings in their specific social relationships in the world was theoretically reduced to instances and examples of a system which lay beyond them.

The major theoretical expression of this reified understanding of language came in the twentieth century, in the work of Saussure, which has close affinities to the objectivist sociology of

[p 28]

Durkheim. In Saussure the social mature of language is expressed as a system (langue), which is at once stable and autonom- ous and founded in normatively identical forms; its `utterances", (paroles) are then seen as `individual' (in abstract distinctions, ,, from `social') uses of `a particular language code' throw h an g enabling `psycho-physical mechanism'. The practical results of n this profound theoretical development, in all its phases, have been exceptionally productive and striking. The great body of philological scholarship has been complemented by a remarkable body of linguistic studies, in which the controlling concept of language as a formal system has opened the way to penetrating descriptions of actual language operations and many of their underlying `laws'.

This achievement has an ironic relation with Marxism. On the one hand it repeats an important and often dominant tendency within Marxism itself, over a range from the comparative analysis and classification of stages of a society, through the discovery of certain fundamental laws of change within these systematic stages, to the assertion of a controlling `social' system which is a priori inaccessible to `individual' acts of will and intelligence. This apparent affinity explains the attempted synthesis of Marxism and structural linguistics which has been so influential a phenomenon of the mid-twentieth century.

But Marxists have then to notice, first, that history, in its most specific, active, and connecting senses, has disappeared (in one tendency has been theoretically excluded) from this account of so central a social activity as language; and second, that the categories in which this version of system has been developed are the familiar bourgeois categories in which an abstract separation and distinction between the `individual' and the `social' have become so habitual that they are taken as `natural' starting-points.

In fact there was little specifically Marxist work on language before the twentieth century. In their chapter on Feuerbach in The German Ideology Marx and Engels touched on the subject, as part of their influential argument against pure, directive consciousness. Recapitulating the `moments' or `aspects' of a materialist conception of history, they wrote:
Only now, after having considered four moments, four aspects of the fundamental historical relationships, do we find that man also posseses

[p 29]

consciousness'; but, even so, not inherent, not `pure' consciousness`. From the start the `spirit' is afflicted with the curse of being 'burdened' with matter, which here makes its appearance in the form of assorted layers of air, sounds, in short of language. Language is as old as consciousness, language is practical consciousness, as it exists for other and for that reason is really beginning to exist for me personally as for language, like consciousness, only arises from the need, the necessity of intercourse with other men. (G), 19)

So far as it goes, this account is wholly compatible with the emphasis on language as practical, constitutive activity. The difficulty arises &e&s it had also arisen in a different form in accounts when the idea of the constitutive is broken down hick. are then temporally ordered. Thus there is an obvious danger, in the thinking of Vico and Herder, of making language 'primary' and 'original', not in the acceptable sense that it is a necessary part of the very act of human self--creation, but in the rated and available - - - sense of language as the founding element of humanity "in the beginning was the Word." It is precisely the sense of language as an indissoluble element of human self-creation that gives any acceptable meaning .to its description as `constitutive'. To make it precede all other connected activities is to claim something quite different.

The idea of language as constitutive is always in danger of this kind of reduction. Not only, however, in the direction of the isolated creative word, which becomes idealism, but also, as actually happened, in objectivist materialism and positivism, where `the world' or `reality' or `social reality' is categorically projected as the pre-existent formation to which language is simply a response.

What Marx and Engels actually say, in this passage, points to simultaneity and totality. The `fundamental historical relationships' are seen as `moments' or `aspects', and man then `also possesses' consciousness'. Moreover, this language is material: the "agitated layers of air, sounds", which are produced by the physical body. It is then not a question of any temporal priority of the `production of material life' considered as a separable act. The distinctively human mode of this primary material production has been characterized in three aspects: needs, new needs, and human reproduction-"not of course to be taken as three different stages . . . but . . . which have existed simultaneously since the dawn of history and the first men, and still assert

[p 30]

themselves in history today". The distinctive humanity of the development is then expressed by the fourth `aspect', that such production is from the beginning also a social relationship. It then involves from the beginning as a necessary element, that practical consciousness which is language.

Thus far the emphasis is primarily constitutive' in the sense of an indissoluble totality of development. But it is easy to see how, in this direction also, what begins as a mode of analysis of aspects of a total process develops towards philosophical or `natural' categories-simple materialist statements which retain the idealist separation of `language' from `reality' but simply reverse their priority-and towards historical categories, in which there is, first, material social production and then (rather than also) language.

In its predominantly positivist development, from the late nineteenth to the mid-twentieth century, a dominant kind of Marxism made this practical reduction: not so much directly in language theory, which on the whole was neglected, but habitually in its accounts of consciousness and in its analyses of the practical language activities which were grouped under the categories of `ideology' and `the superstructure'. Moreover this (tendency was reinforced by the wrong kind of association with `important scientific work on the physical means of language. This association was wholly compatible with an emphasis on language as material, but, given the practical separation of `the world' and `the language in which we speak about it', or in another form, of `reality' and `consciousness', the materiality of language could be grasped only as physical-a set of physical properties-and not as material activity: in fact the ordinary scientistic dissociation of the abstracted physical faculty from its actual human use. The resulting situation had been well described, in another context, by Marx, in the first `thesis' on Feuerbach:

The chief defect of all materialism up to now (including Feuerbach's) is, that the object, reality, what we apprehend through our senses, is understood only in the form of the object of contemplation (anschauung); but not as sensuous human activity, as practice; not subjectively. Hence in opposition to materialism the active side was developed abstractly by idealism-which of course does not know real sensuous activity as such. (GI, 197)

[p 31]

This was indeed the situation in thinking about language. For e active emphases of Vico and Herder had meanwhile been remarkably developed, notably by Wilhelm von Humboldt. Here e inherited problem of the origin of language had been temarkably restated. Language of course developed at some point in evolutionary history, but it is not only that we have virtually no information about this; it is mainly that any human investigation of so constitutive an activity finds language already there in itself and in its presumed object of study. Language has then to be seen as a persistent kind of creation and re-creation: a dynamic presence and a constant regenerative process. But this emphasis, again, can move in different directions. It could reasonably have been associated with the emphasis of whole, indissoluble practice, in which the `dynamic presence' and the `constant regenerative process' would be necessary forms of the `production and reproduction of real life' similarly conceived. What happened instead, in Humboldt and especially after him, was a projection of this idea of activity into essentially idealist and quasi-social forms: either the `nation', based on an abstract version of the `folk-mind' or the (ahistorical) `collective consciousness'; or the `collective spirit', the abstract creative capacity-self-creative but prior to and separate from material social practice, as in Hegel;, or, persuasively, the `individual', abstracted and defined as `creative subjectivity', the starting-point of meaning.

The influence of these various projections has been deep and prolonged. The abstract idea of the `nation' could be readily connected with major philological work on the `families' of languages and on the distinctive inherited properties of particular languages. The abstract idea of the `individual' could be readily connected with the emphasis on a primary subjective reality and a consequent `source' of meaning and creativity which emerged in the Romantic concepts of `art' and `literature' and which defined a major part of the development of `psychology'.

Thus the stress on language as activity, which was the crucial contribution of this line of thinking, and which was a crucial correction of the inherent passivity, usually formalized in the metaphor of `reflection', of positivism and objectivist materialism, was in turn reduced from specific activities (then necessarily social and material, or, in the full sense, historical) to

[p 32]

ideas of such activity, categorized as `nation' or `spirit' or the `creative individual'. It is significant that one of these categories, the `individual' (not the specific, unique human being, who cannot of course be in doubt, but the generalization of the common property of all these beings as `individuals' or `subjects', which are already social categories, with immediate social
implications), was prominent also within the dominant tendency of objectivist materialism. The exclusion of activity, of making, from the category of `objective reality' left it contemp-lated only by `subjects', who might in one version be ignored in the observation of objective reality--the active `subject' replaced' by the neutral `observer'-or in another version, when it became' necessary to speak about language a or about other forms of practice appeared in `inter-subjective' relations-- kin to or at each other passing information or message' between each other, as separate ate or distinguishable identities, rather than ever with each other, the act of language constituting and confirm- ing their relationship. Language here decisively lost it definition as constitutive activity. It became a tool or an instrument or a medium taken up by individuals when the had something to communicate, as distinct from the faculty which made them, from the beginning, not only able to relate and communicate, but in real terms to be practically conscious and so to possess the active practice of language.

Against this reduction of language to instrumentality, the idea of language as expression, which was the main outcome of the idealist version of language as activity, was evidently attractive. It appeared literally to speak to an experience of language which the rival theory, confined to passing information, exchanging messages, naming objects, in effect suppressed. It could include the experience of speaking with others, of participating in language, of making and responding to rhythm or intonation which had no simple `information' or `message' or `object' content: the experience, indeed, which was most evident in `literature' and which was even, by specialization, made identical with it. Yet what actually happened was a deep split, which produced its own powerful categories of separation, some of them old terms in new forms: categorical divisions between the `referential' and the `emotive', between the `denotative' and the `connotative', between `ordinary language' and `literary language'. Certainly the uses towards which these categories point can be distin-

[p 33]

guished as the elements of specific practices, defined by specific situations But their projection as categories, and then their further projection as separate entities , separate `bodies' of age-use, permitted a dissolution and specialization which for a long time prevented the basic issues of the unfinished argument about language from becoming focused within a
single area of discourse.

Marxism might have become this area of discourse, but it had deloped its own forms of limitation and specialization. The most evident of these was a specialization of the whole material, social process to `labour', which was then more and more narrowly conceived. This had its effect in the important argument about the origins and development of language, which could'.
have been reopened in the context of the new science of evolutionary physical anthropology. What happened instead, was an application of the abstract concept of `labour' as the single effective origin. Thus, in a modern authoritative account:

First labour, then articulate speech, were the two chief stimuli under
''the influence of which the brain of the ape gradually changed into
the human brain. (Fundamentals of Dialectical Materialism, ed. Schneier-
Moscow, 1967, 105)

This not only establishes an abstract, two-stage temporal development. It also converts both labour and language to `stimuli', when the real emphasis should be on connected practice. This leads to an abstraction of evolutionary stages:

The development of labour brought the members of the community more closely together, for it enabled them to extend their joint activity and to support each other. Labour relations gave rise to the need for primitive men to speak and communicate with each other.
(Ibid. 105)

This is in effect an idealism of abstracted stimuli and needs. It must be contrasted with a properly materialist theory, in which labour and language, as practices, can be seen as evolutionarily and historically constitutive:
The argument that there could be no language without all the structure of modern man is precisely the same as the old theory that human hands made implement-making and using possible. But the implements are
thousands of years older than hands of the modern human form. Mod-
ern speech-producing structures are the result of the evolutionary suc-

p 34

cess of language, just as the uniquely human hand is the result of the evolutionary success of implements. (J. S. Washburn and J. B. Lancaster, Current Anthropology, vol. 12, No. 3, 1971)

Any constitutive theory of practice, and especially a materialis theory, has important effects beyond the question of origins, in restating the problem of the active process of language at any time: a restatement which goes beyond the separated categorie of `language' and `reality'. Yet orthodox Marxism remaine stuck in reflection theory, because this was the only plausibl materialist connection between the received abstract categories. Reflection theory, in its first period, was itself specialized t crude stimulus-and-response models, adapted from positivis physiology. In its second period, in the later work of Pavlov, it added, as a way of dealing with the special properties of Ian guage, the concept of the `second signal system', the first bein the simple physical system of sensations and responses. This was better than nothing, but it assimilated language to the characteristics of a `signal system', in relatively mechanisti ways, and was in practice unequal to problems of meanin beyond simple models of the associative. Setting out from this point, L.S. Vygotsky (Thought and Language, Moscow, 1934 proposed a new social theory, still named the `second signa system', in which language and consciousness are freed fro simple analogies with physical perception. His work on th development of language in children, and on the crucial prob lem of `inner speech', provided a new starting-point, within historical-materialist perspective. But for a generation, in orthodox Marxism, this was neglected. Meanwhile the work o N. S. Marr, based on older models, tied language to the `super structure' and even to simple class bases. Dogmatic positions taken from other areas of Marxist thinking, limited the necessar theoretical developments. It is ironic that the influence of Ma was in effect ended by Stalin in 1950 with declarations that language was not `part of the superstructure' and that language did not have any essential `class character' but rather a `nations character'. Ironic because though the declarations were necessary, in that context, they simply threw the argument back to a much earlier stage, in which the status of `reflection' and, very specifically, the status of `the superstructure', had, in Marxist terms, needed question. By this time, moreover, linguistics had come to be dominated by a specific and distinctive form of

[p 35]

objectivism which had produced the powerful systems of structuralism and semiotics. It was at this point that generally Marxist positions in other fields, especially in the popular form of objectively determined systems, were practically synthesized with theories of language which, from a fully Marxist position, needed to be profoundly opposed.
A tragic element in this history is that such theories had been rofoundly opposed in the 19206 in Leningrad, where the eginnings of a school of Marxist linguistics, of a significant kind had in fact emerged. It is best represented by the work of V. N. Volosinov, whose Marxism and the Philosophy o f Language appeared in two editions, in 1929 and 1930; the second edition has been translated into English (Matejka and Titunik, New York and London, 1973). Volosinov had been associated with M. M. Bakhtin, author of a study of Dostoevsky (Problemy tvor °estva Dostoevskogo, 1929; new version, with new title, Prob
emy poetiki Dostoevskogo, 1963); see also `P.N. Medvedev' (author of Formal'ny metod v literaturovedenii-kriticeskoe
edenie v sociologiceskuju poetiku-The Formal Method in
literary Scholarship: a critical introduction to sociological poetics-1928). Sometime during the 1930s Volosinov disappeared Nearly half a century was then lost, in real terms, in the development of his exceptionally important realignment of the argument.

Volosinov's decisive contribution was to find a way beyond the powerful but partial theories of expression and objective system. He found it in fundamentally Marxist terms, though he had to begin by saying that Marxist thinking about language was virtually non-existent. His originality lay in the fact that he did not seek to apply other Marxist ideas to language. On the contrary he reconsidered the whole problem of language within a general Marxist orientation. This enabled him to see `activity' (the strength of the idealist emphasis after Humboldt as social activity and to see 'system' ' (the strength of the new objectivist linguistics) in relation to this social activity and not, as had hitherto been the case, formally separated from it. Thus in drawing on the strengths of the alternative traditions, and in setting them side by side showing their connected radical weaknesses, he opened the way to a new kind of theory which had been necessary for more than a century.
Much of his effort went to recovering the full emphasis on

[p 36]

language as activity, as practical consciousness, which had bee weakened and in effect denied by its specialization to a close `individual consciousness' or `inner psyche'. The strength this tradition was still its insistence on the active creation meanings, as distinct from the alternative assumption of closed formal system. Volosinov argued that meaning was necessarily a social action, dependent on a social relationship But to understand .' a depended on recovering a full sense of `social', as distinct both from the idealist reduction of the social to an inherited, ready-made product, an inert crust', beyond': which all creativity was individual, and from the objectivist projection of the social into a formal system, now autonomous and governed only by its internal laws, within which, and solely according to which, meanings were produced. Each sense, at root, depends on the same error: of separating the social from individual meaningful activity (though the rival positions then' valued the separated elements differently). Against the psychologism of the idealist emphasis, Volosinov argued that' "consciousness takes shape and being in the material of signs created by an organized group in the process of its social inter. course. The individual consciousness is nurtured on signs; it de. rives its growth from them; it reflects their logic and laws" (13).
Normally, it is at just this point (and the danger is always increased by retaining the concept of `sign', which Volosinov revalued but continued to use) that objectivism finds its entry, `The material of signs' can be translated as `system of signs'. This system can then be projected (by some notion of a theoretical `social contract', as in Saussure, protected from examination by the assumption of the priority of `synchronic' over `diachronic' analysis) both beyond history and beyond any active conception of contemporary social life, in which socially related individuals meaningfully participate, as distinct from acting out the laws and codes of an inaccessible linguistic system. Each side of Volosinov's argument has a continuing relevance, but it is in his (incomplete) revaluation of the concept of `sign' that his contemporary significance is most evident.
Volosinov accepted that a `sign' in language has indeed a `binary' character. (In fact, as we shall see, his retention of these terms made it easier for the radical challenge of his work to be missed.) That is to say, he agreed that the verbal sign is not equivalent to, nor simply .a reflection of, the object or quality;

[p 37]

it indicates or expresses. The relation within the sign 3n the formal element and the meaning which this ele- carries is thus inevitably conventional (thus far agreeing orthodox semiotic theory), but it is not arbitrary* and, cru- w; it is not fixed. On the contrary the fusion of formal ele- ;and meaning (and it is this fact of dynamic fusion which ~ retention of the `binary' description misleading) is the t of a real process of social development, in the actual (ties of speech and in the continuing development of a rage. Indeed signs can exist only when this active social relationship is posited. The usable sign-the fusion of formal element and meaning-is a product of this continuing speech ~ity between real individuals who are in some continuing dal relationship. The `sign' is in this sense their product, but not simply their past product, as in the reified accounts of an always-given' language system. The real communicative `pro- s' which are usable signs are, on the contrary, living evi- of a continuing social process, into which individuals are and within which they are shaped, but to which they then also actively contribute, in a continuing process. This is at once their socialization and their individuation: the connected sects of a single process which the alternative theories of item' and `expression' had divided and dissociated. We then d not a reified `language' and `society' but an active social language Nor (to glance back at positivist and orthodox aerialist theory) is this language a simple `reflection' or expression of `material reality'. What we have, rather, is a grasping of this reality through language, which as practical consciousness is saturated by and saturates all social activity, eluding productive activity. And, since this grasping is social and continuous (as distinct from the abstract encounters of 'man' and `his world', or `consciousness' and `reality', or language and `material existence'), it occurs within an active and changing society. It is of and to this experience-the lost middle term between the abstract entities, `subject' and `object', on which the propositions of idealism and orthodox materialism

*The question of whether a sign is `arbitrary' is subject to some local confusion. .The term was developed in distinction from the `iconic', to indicate, correctly, that most verbal signs are not `images' of things. But other senses of `arbitrary', in the direction of `random' or 'casual', had developed, and it was these that Volosinov opposed.

[p 38]

are erected-that language speaks. Or to put it more direct language is the articulation of this active and changing exper-enc.; a dynamic and articulated social presence in the world
Yet it remains true that the mode of articulation is specific This is the part of the truth which formalism had grasped. T articulation can be seen, and in some respects has to be seen, both formal and systematic. A physical sound '
natural elements, may be ma e m o a sig-n but its distinction Volosinov argued, is always evident: "a sign does not simply
exist area it -it re it
What distinguishes it as a sign, indeed what made it a sign, is
this sense a Forma process a specific meaning Formalist linguistics s a emphasized amazed this point, but it ha n
discerned that the process of articulation is necessarily also material process, and that the sign itself becomes part of
(socially created) physical and material world: "whether '
sound, physical mass, colour, movement of the body or t like". Signification, the social creation of meanings through t
use of formal signs is then a practical ma material ac m ; it
indeed, literally, a means of production. It is specific form
that practical consciousness w is is inseparable from all social
material activity. It is not, as formalism would make it, and as t
idealist theory of expression had from the beginning assume
an operation of and within `consciousness', which then become
a state or a process separated, a priori, from social material
activity. It is, on the contrary, at once a distinctive material
process-the making of signs-and, in the central quality of its
distinctiveness as practical consciousness, is involved from t
beginning in all other human social and material activity.

Formalist systems can appear to meet this point by referring to the `already-given', the `last-instance determination of the economic structure', as in some current versions of structuralst Marxism. It is to avoid this kind of reduction that we mu consider Volosinov crucial distinction between a `sign' and `signal'. In reflexive theories of language, whether positive kinds of materialism, or such theories as psychological behaviourism, all `signs' are in effect reduced to `signals', within the simple models of `object' and `consciousness' or `stimulus and `response'. Meanings are created by (repeated) recognition of what are then in effect `signals': of the properties of an objet or the character of a stimulus. `Consciousness' and `response

[p 39]

then `contain' (for this is what meaning now is) those properties r that character. The assigned passivity and mechanism of such
accounts have often been recognized. Indeed it was against such passivity and mechanism that formalism had most to contribute,
in its insistence on the specific (formal) articulation of meanings rough signs.
1But it has been less often noticed that quite different theories, based on the determinate character of systems of signs, depend,
ultimately, on a comparable idea of the fixed character of the
sign, which is then in effect a displacement of fixed content to
fixed form. Intense argument between these rival schools has flowed us to overlook the fact that the conversion of the `sign'
(as the term itself always made possible and even likely) into either fixed content or fixed form is a radical denial of active
practical consciousness. The sign, in either case, is moved in the direction of a signal, which Volosinov distinguishes from a sign y the fact that it is intrinsically limited and invariant. The true
quality of a sign (one would have preferred him to say, of a signifying element of a language) is that it is effective in communication, a genuine fusion of a formal element and a meaning (a quality that it indeed shares with signals); but also that as a function of continuing social activity it is capable of modifica
tion and development: the real processes that may be observed i in the history of a language, but which the privileged priority of
`synchronic' analysis had ignored or reduced to a secondary or accidental character.
Indeed since it exists, as a sign, by its quality of signifying relationship-both the relation between formal element and
meaning (its internal structure) and the relations between the people who in actually using it, in practical language, make it a
sign-it has, like the social experience which is the principle of its formation, both dialectical and generative properties.
Characteristically it does not, like a signal, have fixed, determinate, invariant meaning. It must have an effective nucleus of meaning but in practice it has a variable range, corresponding to the endless variety of situations within which it is actively used. These situations include new and changing as well as recurrent relationships, and this is the reality of the sign as dynamic fusion of `formal element' and `meaning'-`form' and `content'-rather than as fixed, `already-given' internal significance. This variable quality, which Volosinov calls mufti-accentual, is

[p 40]

of course the necessary challenge to the idea of `correct' or proper meaning and relationship. Great strength has been given, `proper' meanings, which had been powerfully developed by d continues to be given to theories of language as individual orthodox philology from its studies of dead languages, and depression, by the rich practical experience of `inner which had been taken over both into social-class distinctions of signs'-inner language-in repeated individual awareness of a `standard' language flanked either by `dialects' or by `errors', and into literary theories of a correct' or 'objective' reading. But the quality of varioation--not random variation but variation as a necessary element of practical consciousness--bears heavily also against objectivist accounts of the sign-system. It is one of the decisive arguments against reduction of the key fact of socal determination to the idea of determination by a system. But while it thus beaers heavily against all forms of abstract objectivism, it offers a basis also for a vital reconsideration of the problem of 'subjectivity'.

The signal, in its fixed invariance, is indeed a collective fact. It may be received and repeated, or a new signal may be invented, but in either case the level at which it operates is of a collective kind: that is to say. it has to be recognized but it need not be internalized, at that level of sociality which has excluded (as reductive versions of the `social' commonly exclude) active participation by conscious individuals. The signal, in this sense, is fixed, exchangeable, collective property; characteristically it is easily both imported and exported. The true signifying element of language must from the beginning have a different capacity: to become an inner sign, part of an active practical consciousness. Thus in addition to its social and material existence between actual individuals, the sign is also part of a verbally constituted consciousness which allows individuals to use signs of specifically historical exploration, for their own initiative, whether in acts of social communication or in practices which, not being manifestly social, can be interpreted as personal or private.

This view is then radically opposed to the construction of all acts of communication from pre-determined objective relationships and properties, within which no individual initiative, of a creative or self-generating kind, would be possible. It is thus a decisive theoretical rejection of mechanical, behaviourist, or Saussurean versions of an objective system which is beyond individual initiative or creative use. But it is also a theoretical rejection of subjectivist theories of language as individual expression, since what is internally constituted is the social of the sign, bearing a definite though never fixed or invariant

[p 41]

social meaning and relationship. Great strength has been given, and continues to be given, to theories of language as individual expression, by the rich practical experience of `inner signs'-inner language-in repeated individual awareness of 'inner language activities', whether we call them `thought' or `consciousness' or actual verbal composition. These `inner' activities involve the use of words which are not, at least at that stage, spoken or written to any other person. Any theory of 11 language which excludes this experience, or which seeks to limit it to some residue or by-product or rehearsal (though it may often be these) of manifest social language activity, is again reductive of social language as practical consciousness. What has really to be said is that the sign is social but that in its very <. quality as sign it is capable both of being internalized-indeed . has to be internalized, if it is to be a sign for communicative relation between actual persons, initially using only their own physical powers to express it-and of being continually available, in social and material ways, in manifest communication. This fundamental relationship between the `inner' and the `material' sign-a relationship often experienced as a tension but always lived as an activity, a practice-needs further radical exploration. In individual developmental psychology Vygotsky began this exploration, and at once discerned certain crucially distinguishing characteristics of `inner speech', themselves constitutive rather than, as in Volo"sinov, merely transferred. This is still within the perspective of a historical materialist theory. The complex relationship, from another direction, needs specifically historical exploration, for it is in the movement from the production of language by human physical resources alone, through the material history of the production of other resources and of the problems of both technology and notation then involved in them, to the active social history of the complex of communicative systems which are now so important a part of the material productive process itself, that the dynamics of social language-its development of new means of production within a basic means of production-must be found.

Meanwhile, following Volosinov, we can see that just as all social process is activity between real individuals, so individuality, by the fully social fact of language (whether as `outer' or `inner' speech), is the active constitution, within distinct physical beings, of the social capacity which is the means of realiza-

[p 42]

' tion of any individual life. Consciousness, in this precise sense, is social being. It is the possession, through active and specific social development and relationships, of a precise social capacity, which is the `sign-system'. Vologinov, even after these fundamental restatements, continues to speak of the `sign-system': the formulation that had been decisively made in Saussurean linguistics. But if we follow his arguments we find how difficult and misleading this formulation can be. `Sign' itself-the mark or token; the formal element-has to be revalued to emphasize its variability and internally active elements, indicating not only an internal structure but an internal dynamic. Similarly, `system' has to be revalued to emphasize social process rather than fixed `sociality': a revaluation that was in part made by Jakobson and Tynjanov (1928), within formalist argument, with the recognition that `every system necessarily exists as an evolution while, on the other hand, evolution is inescapably of a systemic nature'. Although this was a necessary recognition, it was limited by its perspective of determinate systems within an `evolutionary' category-the familiar reification of objective idealism-and still requires amendment by the full emphasis of social process. Here, as a matter of absolute priority, men relate and continue to relate before any system which is their product can as a matter of practical rather than abstract consciousness be grasped or exercise its determination.
These changes will have to be made, in the continuing inquiry into language. But the last point indicates a final difficulty. Much of the social process of the creation of meanings was projected within objectivist linguistics to the formal relations-thus the systematic nature-of signs. What at the level of the sign had been abstractly and statically conceived was set into a kind of motion-albeit a frozen, determinate motion, a movement of ice-fields-in the relational `laws' or `structures' of the system as a whole. This extension to a relational system, including its formal aspect as grammar, is in any case inevitable. Isolation of `the sign', whether in Saussure or Vologinov, is at best an analytical procedure, at worst an evasion. Much of the important work on relations within a whole system is therefore an evident advance, and the problem of the variability of the sign can appear to be contained within the variability of its formal relations. But while this kind of emphasis on the relational system is obviously necessary, it is limited by the consequence

[p 43]

of the initial abstract definition of the sign. The highly complex relations of (theoretically) invariable units can never be substantive relationships; they must remain as formal relationships. The internal dynamics of the sign, including its social and material relationships as well as its formal structure, must be seen as necessarily connected with the social and material as well as the formal dynamics of the system as a whole. There have been some advances in this direction in recent work (RossiLandi, 1975).
But there has also been a move which seems to reopen the whole problem. In Chomskyan linguistics there has been a decisive step towards a conception of system which emphasizes the possibility and the fact of individual initiative and creative practice which earlier objectivist systems had excluded. But at the same time this conception stresses deep structures of language formation which are certainly incompatible with ordi
nary social and historical accounts of the origin and development of language. An emphasis on deep constitutive structures, at an evolutionary rather than a historical level, can of course be
reconciled with the view of language as a constitutive human faculty: exerting pressures and setting limits, in determinate ways, to human development itself. But while it is retained as an exclusively evolutionary process, it moves, necessarily, towards
reified accounts of `systemic evolution': development by constituted systems and structures (the constitution now at once permitting and limiting variations) rather than by actual human beings in a continuing social practice. Here Vygotsky's work on inner speech and consciousness is theoretically crucial:

If we compare the early development of speech and of intellect-which, as we have seen, develop along separate lines both in animals and in very young children-with the development of inner speech and of verbal thought, we must conclude that the later stage is not a simple continuation of the earlier. The nature of the development itself
changes, from biological to socio-historical. Verbal thought is not an innate, natural form of behaviour but is determined by a historical
cultural process and has specific properties and laws that cannot be found in the natural forms of thought and speech. (Thought and Language, 51)
Thus we can add to the necessary definition of the biological faculty of language as constitutive an equally necessary definition of language development-at once individual and

[p 44]

social-as historically and socially constituting. What we can then define is a dialectical process: the changing practical consciousness of human beings, in which both the evolutionary and the historical processes can be given full weight, but also within which they can be distinguished, in the complex variations of actual language use. It is from this theoretical foundation thatwe can go on to distinguish `literature', in a specific sociohistorical development of writing, from the abstract retrospective concept, so common in orthodox Marxism, which reduces it, like language itself, to a function and then a (superstructural) by-product of collective labour. But before we can go on to this, we must examine the concepts of literature which, based on earlier theories of language and consciousness, stand in the way.

nnnnnnnn

42 Marxism and Literature
' tion of any individual life. Consciousness, in this precise sense, is social being. It is the possession, through active and specific social development and relationships, of a precise social capacity, which is the `sign-system'. Vologinov, even after these fundamental restatements, continues to speak of the `sign-system': the formulation that had been decisively made in Saussurean linguistics. But if we follow his arguments we find how difficult and misleading this formulation can be. `Sign' itself-the mark or token; the formal element-has to be revalued to emphasize its variability and internally active elements, indicating not only an internal structure but an internal dynamic. Similarly, `system' has to be revalued to emphasize social process rather than fixed `sociality': a revaluation that was in part made by Jakobson and Tynjanov (1928), within formalist argument, with the recognition that `every system necessarily exists as an evolution while, on the other hand, evolution is inescapably of a systemic nature'. Although this was a necessary recognition, it was limited by its perspective of determinate systems within an `evolutionary' category-the familiar reification of objective idealism-and still requires amendment by the full emphasis of social process. Here, as a matter of absolute priority, men relate and continue to relate before any system which is their product can as a matter of practical rather than abstract consciousness be grasped or exercise its determination.
These changes will have to be made, in the continuing inquiry into language. But the last point indicates a final difficulty. Much of the social process of the creation of meanings was projected within objectivist linguistics to the formal relations-thus the systematic nature-of signs. What at the level of the sign had been abstractly and statically conceived was set into a kind of motion-albeit a frozen, determinate motion, a movement of ice-fields-in the relational `laws' or `structures' of the system as a whole. This extension to a relational system, including its formal aspect as grammar, is in any case inevitable. Isolation of `the sign', whether in Saussure or Vologinov, is at best an analytical procedure, at worst an evasion. Much of the important work on relations within a whole system is therefore an evident advance, and the problem of the variability of the sign can appear to be contained within the variability of its formal relations. But while this kind of emphasis on the relational system is obviously necessary, it is limited by the consequence

[p 43]