Première publication dans Marxism and Literature, Oxford University Press, 1977, pp. 45-54.


It is relatively difficult to see ‘literature’ as a concept. In ordinary usage it appears to be no more than a specific description, and what is described is then, as a rule, so highly valued that there is a virtually immediate and unnoticed transfer of the specific values of particular works and kinds of work to what operates as a concept but is still firmly believed to be actual and practical. Indeed the special property of ‘literature’ as a concept is that it claims this kind of importance and priority, in the concrete achievements ofmany particular great works, as against the ‘abstraction’ and ‘generality’ of other concepts and of the kinds of practice which they, by contrast, define. Thus it is common to see ‘literature’ defined as ‘full, central, immediate human experience’, usually with an associated reference to ‘minute particulars’. By contrast, ‘society’ is often seen as essentially general and abstract: the summaries and averages, rather than the direct substance, of human living. Other related concepts, such as ‘politics’, ‘sociology’ or ‘ideology’, are similarly placed and downgraded, as mere hardened outer shells compared with the living experience of literature.

The naivety of the concept, in this familiar form, can be shown in two ways: theoretically and historically. It is true that one popular version of the concept has been developed in ways that appear to protect it, and in practice do often protect it, against any such arguments. An essential abstraction of the ‘personal’ and the ‘immediate’ is carried so far that, within this highly developed form of thought, the whole process of abstraction has been dissolved. None of its steps can be retraced, and the abstraction of the ‘concrete’ is a perfect and virtually unbreakable circle. Arguments from theory or from history are simply evidence of the incurable abstraction and generality of those who are putting them forward. They can then be contemptuously rejected, often without specific reply, which would be only to fall to their level.

This is a powerful and often forbidding system of abstraction, in which the concept of ‘literature’ becomes actively ideological. Theory can do something against it, in the necessary recognition (which ought hardly, to those who are really in contact with literature, to need any long preparation) that whatever else ‘it’ may be, literature is the process and the result of formal composition within the social and formal properties of a language. The effective suppression of this process and its circumstances, which is achieved by shifting the concept to an undifferentiated equivalence with ‘immediate living experience’ (indeed, in some cases, to more than this, so that the actual lived experiences of society and history are seen as less particular and immediate than those of literature) is an extraordinary ideological feat. The very process that is specific, that of actual composition, has effectively disappeared or has been displaced to an internal and self-proving procedure in which writing of this kind is genuinely believed to be (however many questions are then begged) ‘immediate living experience’ itself. Appeals to the history of literature, over its immense and extraordinarily various range, from the Mabinogion to Middlemarch, or from Paradise Lost to The Prelude, cause a momentary hesitation until various dependent categories of the concept are moved into place: ‘myth’, ‘romance’, ‘fiction’, ‘realist fiction’, ‘epic’, ‘lyric’, ‘autobiography’. What from another point of view might reasonably be taken as initial definitions of the processes and circumstances of composition are converted, within the ideological concept, to ‘forms’ of what is still triumphantly defined as ‘full, central, immediate human experience’. Indeed when any concept has so profound and complex an internal specializing development, it can hardly be examined or questioned at all from outside. If we are to understand its significance, and the complicated facts it partially reveals and partially obscures, we must turn to examining the development of the concept itself.

In its modern form the concept of ‘literature’ did not emerge earlier than the eighteenth century and was not fully developed until the nineteenth century. Yet the conditions for its emergence had been developing since the Renaissance. The word itself came into English use in the fourteenth century, following French and Latin precedents; its root was Latin littera, a letter ofthe alphabet. Litterature, in the common early spelling, was then in effect a condition of reading: of being able to read and of having read. It was often close to the sense of modern literacy, which was not in the language until the late nineteenth century, its introduction in part made necessary by the movement of literature to a different sense. The normal adjective associated with literature was literate. Literary appeared in the sense of reading ability and experience in the seventeenth century, and did not acquire its specialized modern meaning until the eighteenth century.

Literature as a new category was then a specialization of the area formerly categorized as rhetoric and grammar: a specialization to reading and, in the material context of the development of printing, to the printed word and especially the book. It was eventually to become a more general category than poetry or the earlier poesy, which had been general terms for imaginative composition, but which in relation to the development of literature became predominantly specialized, from the seventeenth century, to metrical composition and especially written and printed metrical composition. But literature was never primarily the active composition ─ the ‘making’ ─ which poetry had described. As reading rather than writing, it was a category of a different kind. The characteristic use can be seen in Bacon ─ “learned in all literature and erudition, divine and humane” ─ and as late as ─ “he had probably more than common literature, as his son addresses him in one of his most elaborate Latin poems”. Literature, that is to say, was a category of use and condition rather than of production. It was a particular specialization of what had hitherto been seen as an activity or practice, and a specialization, in the circumstances, which was inevitably made in terms of social class. In its first extended sense, beyond the bare sense of ‘literacy’, it was a definition of ‘polite’ or ‘humane’ learning, and thus specified a particular social distinction. New political concepts of the ‘nation’ and new valuations of the ‘vernacular’ interacted with a persistent emphasis on ‘literature’ as reading in the ‘classical’ languages. But still, in this first stage, into the eighteenth century, literature was primarily a generalized social concept, expressing a certain (minority) level of educational achievement. This carried with it a potential and eventually realized alternative definition of literature as ‘printed books’: the objects in and through which this achievement was demonstrated.

It is important that, within the terms of this development, literature normally included all printed books. There was not necessary specialization to ‘imaginative’ works. Literature was still primarily reading ability and reading experience, and this included philosophy, history and essays as well as poems. Were the new eighteenth-century novels ‘literature’? That question was first approached, not by definition of their mode or content, but by reference to the standards of ‘polite’ or ‘humane’ learning. Was drama literature? This question was to exercise successive generations, not because of any substantial difficulty but because of the practical limits of the category. If literature was reading, could a mode written for spoken performance be said to be literature, and if not, where was Shakespeare? (But of course he could now be read; this was made possible, and ‘literary’, by texts.)

At one level the definition indicated by this development has persisted. Literature lost its earliest sense of reading ability and reading experience, and became an apparently objective category of printed works of a certain quality. The concerns of a ‘literary editor’ or a ‘literary supplement’ would still be defined in this way. But three complicating tendencies can then be distinguished: first, a shift from ‘learning’ to ‘taste’ or ‘sensibility’ as a criterion defining literary quality; second, an increasing specialization of literature tocreative’ or ‘imaginative’ works; third, a development of the concept of ‘tradition’ within national terms, resulting in the more effective definition of ‘a national literature’. The sources of each of these tendencies can be discerned from the Renaissance, but it was in the eighteenth and nineteenth centuries that they came through most powerfully, until they became, in the twentieth century, in effect received assumptions. We can look more closely at each tendency.

The shift from ‘learning’ to ‘taste’ or ‘sensibility’ was in effect the final stage of a shift from a para-national scholarly profession, with its original social base in the church and then in the universities, and with the classical languages as its shared material, to a profession increasingly defined by its class position, from which essentially general criteria, applicable in fields other than literature, were derived. In England certain specific features of bourgeois development strengthened the shift; the ‘cultivated amateur’ was one of its elements, but ‘taste’ and ‘sensibility’ were essentially unifying concepts, in class terms, and could be applied over a very wide range from public and private behaviour to (as Wordsworth complained) either wine or poetry. As subjective definitions of apparently objective criteria (which acquire their apparent objectivity from an actively consensual class sense), and at the same time apparently objective definitions of subjective qualities, ‘taste’ and ‘sensibility’ are characteristically bourgeois categories.

‘Criticism’ is an essentially associated concept, in the same development. As a new term, from the seventeenth century, it developed (always in difficult relations with its general and persistent sense of fault-finding) from ‘commentaries’ on literature, within the ‘learned’ criterion, to the conscious exercise of ‘taste’, ‘sensibility’ and ‘discrimination’. It became a significant special form of the general tendency in the concept of literature towards an emphasis on the use or (conspicuous) consumption of works, rather than on their production. While the habits of use or consumption were still the criteria of a relatively integrated class, they had their characteristic strengths as well as weaknesses. ‘Taste’ in literature might be confused with ‘taste’ in everything else, but, within class terms, responses to literature were notably integrated, and the relative integration of the ‘reading public’ (a characteristic term of the definition) was a sound base for important literary production. The reliance on ‘sensibility’, as a special form of an attempted emphasis on whole ‘human’ response, had its evident weaknesses in its tendency to separate ‘feeling’ from ‘thought’ (with an associated vocabulary of ‘subjective’ and ‘objective’, ‘unconscious’ and ‘conscious’, ‘private’ and ‘public’). At the same time it served, at its best, to insist on ‘immediate’ and ‘living’ substance (in which its contrast with the ‘learned’ tradition was especially marked). It was really only as this class lost its relative cohesion and dominance that the weakness of the concepts as concepts became evident. And it is evidence of at least its residual hegemony that criticism, taken as a new conscious discipline into the universities, to be practised by what became a new para-national profession, retained these founding class concepts, alongside attempts to establish new abstractly objective criteria. More seriously, criticism was taken to be a natural definition of literary studies, themselves defined by the specializing category (printed works of a certain quality) of literature. Thus these forms of the concepts of literature and criticism are, in the perspective of historical social development, forms of a class specialization and control of a general social practice, and of a class limitation of the questions which it might raise.

The process of the specialization of ‘literature’ to ‘creative’ or ‘imaginative’ works is very much more complicated. It is in part a major affirmative response, in the name of an essentially general human ‘creativity’, to the socially repressive and intellectually mechanical forms of a new social order: that of capitalism and especially industrial capitalism. The practical specialization of work to the wage-labour production of commodities; of ‘being’ to ‘work’ in these terms; of language to the passing of ‘rational’ or ‘informative’ ‘messages’; of social relations to functions within a systematic economic and political order: all these pressures and limits were challenged in the name of a full and liberating ‘imagination’ or ‘creativity’. The central Romantic assertions, which depend on these concepts, have a significantly absolute range, from politics and nature to work and art. ‘Literature’ acquired, in this period, a quite new resonance, but it was not yet a specialized resonance. That came later as, against the full pressures of an industrial capitalist order, the assertion became defensive and reserving where it had once been positive and absolute. In ‘art’ and ‘literature’, the essential and saving human qualities must, in the early phase, be ‘extended’; in the later phase, ‘preserved’.

Several concepts developed together. ‘Art’ was shifted from its sense of a general human skill to a special province, defined by ‘imagination’ and ‘sensibility’. ‘Aesthetic’, in the same period, shifted from its sense of general perception to a specialized category of the ‘artistic’ and the ‘beautiful’. ‘Fiction’ and ‘myth’ (a new term from the early nineteenth century) might be seen from the dominant class position as ‘fancies’ or ‘lies’ but from this alternative position were honoured as the hearers of imaginative truth’. ‘Romance’ and ‘romantic’ were given newly specialized positive emphases. ‘Literature’ moved with all these. The wide general meaning was still available, but a specialized meaning came steadily to predominate, around the distinguishing qualities of the ‘imaginative’ and the ‘aesthetic’. ‘Taste’ and ‘sensibility’ had begun as categories of a social condition. In the new specialization, comparable but more elevated qualities were assigned to ‘the works themselves’, the ‘aesthetic objects’.

But there was still one substantial uncertainty: whether the elevated qualities were to be assigned to the ‘imaginative’ dimension (access to a truth ‘higher’ or ‘deeper’ than ‘scientific’ or ‘objective’ or ‘everyday’ reality; a claim consciously substituting itself for the traditional claims of religion) or to the ‘aesthetic’ dimension (‘beauties’ of language or style). Within the specialization of literature, alternative schools made one or other of these emphases, but there were also repeated attempts to fuse them, making ‘truth’ and ‘beauty’, or ‘truth’ and ‘vitality of language’, identical. Under continuing pressure these arguments became not only positive assertions but increasingly negative and comparative, against all other modes: not only against ‘science’ and ‘society’ ─ the abstract and generalizing modes of other ‘kinds’ of experience ─ and not only against other kinds of writing ─ now in their turn specialized as ‘discursive’ or ‘factual’ ─ but, ironically, against much of ‘literature’ itself ─ ‘bad’ writing, ‘popular’ writing, ‘mass culture’. Thus the category which had appeared objective as ‘all printed books’, and which had been given a social-class foundation as ‘polite learning’ and the domain of ‘taste’ and ‘sensibility’, now became a necessarily selective and self-defining area: not all ‘fiction’ was ‘imaginative’; not all ‘literature’ was ‘Literature’. ‘Criticism’ acquired a quite new and effectively primary importance, since it was now the only way of validating this specialized and selective category. It was at once a discrimination of the authentic ‘great’ or ‘major’ works, with a consequent grading of ‘minor’ works and an effective exclusion of ‘bad’ or ‘negligible’ works, and a practical realization and communication of the ‘major’ values. What had been claimed for ‘art’ and the ‘creative imagination’ in the central Romantic arguments was now claimed for ‘criticism’, as the central ‘humane’ activity and ‘discipline’.

This development depended, in the first place, on an elaboration of the concept of ‘tradition’. The idea of a ‘national literature’ had been growing strongly since the Renaissance. It drew on all the positive forces of cultural nationalism and its real achievements. It brought with it a sense of the ‘greatness’ or ‘glory’ of the native language, for which before the Renaissance there had been conventional apology by comparison with a ‘classical’ range. Each of these rich and strong achievements had been actual; the ‘national literature’ and the ‘major language’ were now indeed ‘there’. But, within the specialization of ‘literature’, each was re-defined so that it could be brought to identity with the selective and self-defining ‘literary values’. The ‘national literature’ soon ceased to be a history and became a tradition It was not, even theoretically, all that had been written or all kinds of writing It was a selection which culminated in, and in a circular way defined, the ‘literary values’ which ‘criticism’ was asserting. There were then always local disputes about who and what should be included, or as commonly excluded, in the definition of this ‘tradition’. To have been an Englishman and to have written was by no means to belong to the ‘English literary tradition’, just as to be an Englishman and to speak was by no means to exemplify the ‘greatness’ of the language ─ indeed the practice of most English speakers was continually cited as ‘ignorance’ or ‘betrayal’ or ‘debasement’ of just this ‘greatness’. Selectivity and self-definition, which were the evident processes of ‘criticism’ of this kind, were, however, projected as ‘literature’ itself, as ‘literary values’ and even finally as ‘essential Englishness’: the absolute ratification of a limited and specializing consensual process. To oppose the terms of this ratification was to be ‘against literature’.

It is one of the signs of the success of this categorization of literature that even Marxism has made so little headway against it. Marx himself, to be sure, hardly tried. His characteristically intelligent and informed incidental discussions of actual literature are now often cited, defensively, as evidence of the humane flexibility of Marxism, when they ought really to be cited (with no particular devaluation) as evidence of how far he remained, in these matters, within the conventions and categories of his time. The radical challenge of the emphasis on ‘practical consciousness’ was thus never carried through to the categories of ‘literature’ and ‘the aesthetic’, and there was always hesitation about the practical application, in this area, of propositions which were held to be central and decisive almost everywhere else.

When such application was eventually made, in the later Marxist tradition, it was of three main kinds: an attempted assimilation of ‘literature’ to ‘ideology’, which was in practice little more than banging one inadequate category against another; an effective and important inclusion of ‘popular literature’ ─ the ‘literature of the people’ ─ as a necessary but neglected part of the ‘literary tradition’; and a sustained but uneven attempt to relate ‘literature’ to the social and economic history within which ‘it’ had been produced. Each of these last two attempts has been significant. In the former a ‘tradition’ has been genuinely extended. In the latter there has been an effective reconstitution, over wide areas, of historical social practice, which makes the abstraction of ‘literary values’ much more problematical, and which, more positively, allows new kinds of reading and new kinds of questions about ‘the works themselves’. This has been known, especially, as ‘Marxist criticism’ (a radical variant of the established bourgeois practice) though other work has been done on quite different bases, from a wider social history and from wider conceptions of ‘the people’, ‘the language’ and ‘the nation’.

It is significant that ‘Marxist criticism’ and ‘Marxist literary studies’ have been most successful, in ordinary terms, when they have worked within the received category of ‘literature’, which they may have extended or even revalued, but never radically questioned or opposed. By contrast, what looked like fundamental theoretical revaluation, in the attempted assimilation to ‘ideology’, was a disastrous failure, and fundamentally compromised, in this whole area, the status of Marxism itself. Yet for half a century now there have been other and more significant tendencies. Lukács contributed a profound revaluation of ‘the aesthetic’. The Frankfurt School, with its special emphasis on art, undertook a sustained re-examination of ‘artistic production’, centred on the concept of ‘mediation’. Goldmann undertook a radical revaluation of the ‘creative subject’. Marxist variants of formalism undertook radical redefinition of the processes of writing, with new uses of the concepts of ‘signs’ and ‘texts’, and with a significantly related refusal of ‘literature’ as a category. The methods and problems indicated by these tendencies will be examined in detail later in this book.

Yet the crucial theoretical break is the recognition of ‘literature’ as a specializing social and historical category. It should be clear that this does not diminish its importance. Just because it is historical, a key concept of a major phase of a culture, it is decisive evidence of a particular form of the social development of language. Within its terms, work of outstanding and permanent importance was done, in specific social and cultural relationships. But what has been happening, in our own century, is a profound transformation of these relationships, directly connected with changes in the basic means of production. These changes are most evident in the new technologies of language, which have moved practice beyond the relatively uniform and specializing technology of print. The principal changes are the electronic transmission and recording of speech and of writing for speech, and the chemical and electronic composition and transmission of images, in complex relations with speech and with writing for speech, and including images which can themselves be ‘written’. None of these means cancels print, or even diminishes its specific importance, but they are not simple additions to it, or mere alternatives. In their complex connections and interrelations they compose a new substantial practice in social language itself, over a range from public address and manifest representation to ‘inner speech’ and verbal thought. For they are always more than new technologies, in the limited sense. They are means of production, developed in direct if complex relations with profoundly changing and extending social and cultural relationships: changes elsewhere recognizable as deep political and economic transformations. It is in no way surprising that the specialized concept of ‘literature’, developed in precise forms of correspondence with a particular social class, a particular organization of learning, and the appropriate particular technology of print, should now be so often invoked in retrospective, nostalgic, or reactionary moods, as a form of opposition to what is correctly seen as a new phase of civilization. The situation is historically comparable to that invocation of the divine and the sacred, and of divine and sacred learning, against the new humanist concept of literature, in the difficult and contested transition from feudal to bourgeois society.

What can then be seen as happening, in each transition, is an historical development of social language itself: finding new means, new forms and then new definitions of a changing practical consciousness. Many of the active values of ‘literature’ have then to be seen, not as tied to the concept, which came to limit as well as to summarize them, but as elements of a continuing and changing practice which already substantially, and now at the level of theoretical redefinition, is moving beyond its old forms.

Pour citer cet article :

Raymond Williams, « Literature », réédition sur le site des ressources Socius, URL :, page consultée le 25 mars 2017.