Popular memory: theory, politics, method

The Popular Memory Group

In this article we explore an approach to history-writing which involves becoming 'historians of the present too'. It is important to stress 'explore'. We do not have a completed project in 'popular memory' to report. We summarize and develop discussions which were intended as an initial clarification. These discussions had three main starting-points. First, we were interested in the limits and contradictions of academic history where links were attempted with a popular socialist or feminist politics. Our main example here was 'oral history', a practice that seemed nearest to our own preoccupations. Second, we were attracted to projects which moved in the direction indicated by these initial criticisms. These included experiments in popular autobiography and in community-based history, but also some critical developments with a base in cultural studies or academic historiography. Third, we tried, as in the case of all the articles in this book, to relate problems of history-writing to more abstract debates which suggested possible clarifications.

What do we mean, then, by 'popular memory'? We give our own provisional answers in the first part of this essay. We define popular memory first as an object of study, but, second, as a dimension of political practice. We then look, in the second part, at some of the resources for such a project, but also sketch its limits and difficulties. These are discussed in turn and at more length in the third and fourth parts. They range from problems of theory and method to the social organization of research and writing.

On its own this essay is incomplete in another way: though it sketches the field as a whole, it explores one side of the popular memory relation, the side nearest to oral history as a practice. The larger argument is extended, in important ways, in Chapter 7. Although this has a different authorship it grew from the same discussions.

Defining popular memory

Popular memory as an object of study

The first move in defining popular memory is to extend what we mean by historywriting (and therefore what is involved in historiographical comment). The other essays in this book have not drawn too formal a distinction between academic history and more popular or politicized kinds. Such a distinction makes little sense if applied to Marx; or the Hammonds, or the early years of the Communist Party Historians' Group, or to much feminist history. We have tended to use some loose notion of 'the history-writer', who works under the sign of history as art or science, and even so flexible a definition does set limits. This is especially true for the present time. The looser, potentially amateur notion of 'tine history-writer' has hardened into 'the historian' as specialist academic. Left historiographies have not been immune from this process. Socialist and feminist histories have developed with at least one foot in universities or polytechnics. The expansion of a radicalized college-going constituency, characteristic of the 1960s and 1970s, has created a kind of specialized readership. Some of the contradictions, and two different ways of handling them, can be seen in the contrasted strategies of two recently founded historical journals, Social History and History Workshop Joumal: the one pursuing a cautious historical avant-gardism, the other avowedly socialist and committed to the idea of a genuinely 'popular' history.1

For the modern period there is a real problem of the implicitly non-popular effects of focusing on formal history-writing, a practice largely colonized by academic and professional norms. (As we shall see this is also the case with new methodologies, especially 'oral history', which are sometimes seen as intrinsically 'popular' and democratizing.) If we retain this focus, we risk reproducing some very conservative forms: a closed circle of comment between left social historians and what Marx would have called 'critical critics'. Ken Worpole has pointed out the effects of this in justly sceptical terms:

It is obvious to anyone that the last two decades have produced an outstanding growth in the range of work done in the field of Labour studies and the more informal modes of working-class self-organisation and forms of cultural identity. . . . There has been a proliferation of research papers, published essays and fulllength books emerging from this powerful intellectual current. Yet . . . I seriously wonder whether we could with any confidence suggest that we have a more historically conscious labour movement now than we have done at previous periods of crisis in the past. I would think not. 2

He suggests 'two observable trends' that might account for this: the concentration of history in higher education (the move from 'draughty Co-op Halls and Trades Halls to modern Polytechnic lecture rooms') and the expense of commercially published books ('expensive hardbacks for the higher education libraries, rather than pocketbooks for the people'). What is so important about these comments is that they direct attention to the form of historical works and the social conditions

within which they are produced, distributed and (sometimes) read. These questions are completely taken for granted in the ordinary run of critical reviewing, though the problem of the accessibility of language and of jargon' sometimes stands in for the larger problems.

Worpole's comments show the need to expand the idea of historical production well beyond the limits of academic history-writing. We must include all the ways in which a sense of the past is constructed in our society. These do not necessarily take a written or literary form. Still less do they conform to academic standards of scholarship or canons of truthfulness. Academic history has a particular place in a much larger process. We will call this 'the social production of memory'. In this collective production everyone participates, though unequally. Everyone, in this sense, is a historian. As Jean Chesneaux argues, professionalized history has attempted to appropriate a much more general set of relationships and needs: 'the collective and contradictory relationship of our society to its past' and the 'collective need' for guidance in the struggle to make the future.3 We have already noted a similar stress in Christopher Hill's work: the recognition of a larger social process in which 'we ourselves are shaped by the past' but are also continually reworking the past which shapes us.' The first problem in the pursuit of 'popular memory' is to specify the 'wet in Hill's formulation or 'our society' in Chesneaux's. What are the means by which social memory is produced? And what practices are relevant especially outside those of professional history-writing?

It is useful to distinguish the main ways in which a sense of the past is produced: through public representations and through private memory (which, however, may also be collective and shared). The first way involves a public 'theatre' of history, a public stage and a public audience for the enacting of dramas concerning 'our' history, or heritage, the story, traditions and legacy of tthe British People'. This public stage is occupied by many actors who often speak from contradictory scripts, but collectively we shall term the agencies which construct this public historical sphere and control access to the means of public-ation 'the historical apparatus'. We shall call the products of these agencies, in their aggregate relations and combinations at any point of time, 'the field of public representations of history'. In thinking about the ways in which these representations affect individual or group conceptions of the past, we might speak of 'dominant memory'. This term points to the power and pervasiveness of historical representations, their connections with dominant institutions and the part they play in winning consent and building alliances in the processes of formal politics. But we do not mean to imply that conceptions of the past that acquire a dominance in the field of public representations are either monolithically installed or everywhere believed in. Not all the historical representations that win access to the public field are 'dominant'. The field is crossed by competing constructions of the past, often at war with each other. Dominant memory is produced in the course of these struggles and is always open to contestation. We do want to insist, however, that there are real processes of domination in the historical field. Certain representations achieve centrality and luxuriate grandly; others are marginalized or excluded or reworked.

Nor are the criteria of success here those of truth: dominant representations may be those that are most ideological, most obviously conforming to the flattened stereotypes of myth.

Historical constructions are most obviously public when linked to central state institutions. The governmental and parliamentary systems, especially in their 'Englishness', are historical apparatuses in their own right. Aided (sotto voce) by BBC pomposity, they 'breathe' a sense of 'tradition', guaranteeing the inviolability of the broad ground-rules of formal politics, 'our democratic constitution'. Actually (and contradictorily) it is not parliamentary institutions that are the important foci for most pageantry, the main form of historical theatre. The monarchy and the military are much more closely involved here, providing the very stuff of tradition. Both loom large in the regular metropolitan spectacles and in the more occasional shows: jubilees, royal weddings, state visits, state funerals and commemorative events. Nor are the historical interventions of the monarchy and the military only metropolitan, appropriated by visiting tourists (though they certainly benefit the tourist industries). Loudly amplified through the media, they intersect with everyday life in the localities. 'Our Royal Family' may be costly consumed at the fireside. Children may learn about a militaristic past at the war museums, through handbooks of military strategy and technology and through the local airshow or open day, commemorating, perhaps, the Battle of Britain. Historical recreations (popular now in the grounds of the better preserved local castles) may figure military moments (the Civil War) or pugnacious popular myth, robbed, however, of political significance (Robin Hood versus the Sheriff of Nottingham perhaps). Such events produce too their own historiographies of brochures, guides, of ficial (e.g. regimental) histories and massive academic and popular literatures on royal and military personages and themes. Despite their official origins such representations have a real life within the patterns of popular leisure and pleasure.

Other institutions, though linked to the national or local state, have a greater degree of autonomy, operating with high-cultural, educational, presenational or archival purposes. We include here the whole world of museums, art galleries, record of fices, the Department of the Environment's of ficial preservation orders, the 'National' Trust, the 'National' Theatre, and in general the sphere of history as 'cultural policy' - much of what is explored in Chapter 7 as 'National Heritage'. Perhaps the educational system itself belongs here too: the academic producers and all those definitions of historical significance carried in the formal curricula, in O and A level syllabuses and examinations, and in the texts that are widely used in schools. In this 'cultural' field, the relations between scholarly and dominant historiographies are especially intimate; the historian's criteria of truthfulness are more likely to prevail here than in the more overtly politicized versions.

History is also business. It is important for the whole range of publishing activity, especially since historical writing retains much more of an amateur or 'lay' public than other social sciences. Best-seller lists commonly contain items that are marketed as 'historical', especially biographies and autobiographies, historical fictions and military histories. In Britain, more than in Europe, World War II has provided an inexhaustible supply of historical fact and fiction, much of it in heavily militaristic guise and reinforced by the close convergence of war, fighting and a boy culture in men, young and old.S (The historical paradigm here is definitely not academic history but the tradition of masculine romance that runs from Boy's Own to the Super-hero comics of today.) To popular fiction and the modern form of the glossy illustrated documentary book, we have to add the historical movies, somewhat displaced in the block-buster market by the contemporary salience of science fiction. More interesting because less remarked on is the massive contemporary growth of 'historical tourism'. We mean the way in which historically significant places become a resource, physically or ideologically, for the leisure and tourist industries. The way is led there by the owners of palaces, mansions, castles and other 'country houses', and, in its own discreet way, by the Anglican Church. But the last decade or so has seen the commercial colonization of many lesser sites with historical or mythical capital. The guide books, also commercially produced and with very large and expertly promoted circulations, point us to these places, encapsulating their historical meaning.

The public media too - especially radio, television and the press - are a principal source of historical constructions. We include the intersections of history, journalism and documentary, but also the media arts, especially historical drama. The media certainly produce their own historical accounts - they-produce a contemporary history daily, for instance, in the form of 'pews'. But they also select, amplify and transform constructions of the past produced elsewhere. They increasingly draw, for example, from oral history and 'yesterday's witness'. They give a privileged space to conceptions of the past which accompany the party-political battles. Of all parts of the historical apparatus the electronic media are perhaps the most compelling and ubiquitous. Access here may often be decisive in gaining currency for an historical account.

More removed from the patronage of state and of capital are the voluntary associations of the world of history. Most counties and many towns have their own historical and archaeological societies, often with a long nineteenth-century pedigree. Like the Historical Association which links schoolteachers and academics, these societies draw on a fund of amateur historical enthusiasm, bounded by a strong sense of locality. To these we must add newer growths: the preservation societies and community-based groups and WEA classes, including those with socialist and feminist purposes. The growth of 'oral history' and of the History Workshop movement has added whole layers, sometimes of a radically new kind, to these local and participatory forms.

As this last set of examples suggests, the various sites and institutions do not act in concert. To make them sing, if not in harmony at least with only minor dissonances, involves hard labour and active intervention. Sometimes this has been achieved by direct control (censorship for example) and by a violent recasting or obliteration of whole fields of public history. More commonly today, in the capitalist West, the intersections of formal political debates and the public media are probably the crucial site. Certainly political ideologies involve a view of past and present and future. Ranged against powers such as these, what price the lonely scholar, producing (also through commercial channels) the one or two thousand copies of the latest monograph?!

There is a second way of looking at the social production of memory which draws attention to quite other processes. A knowledge of past and present is also produced in the course of everyday life. There is a common sense of the past which, though it may lack consistency and explanatory force, none the less contains elements of good sense. Such knowledge may circulate, usually without amplification, in everyday talk and in personal comparisons and narratives. It may even be recorded in certain intimate cultural forms: letters, diaries, photograph albums and collections of things with past associations. It may be encapsulated in anecdotes that acquire the force and generality of myth. If this is history, it is history under extreme pressures and privations. Usually this history is held to the level of private remembrance. It is not only unrecorded, but actually silenced. It is not offered the occasion to speak. In one domain, the modern Women's Movement well understands the process of silencing and is raising the 'hidden' history of women's feelings, thoughts and actions more clearly to view. Feminist history challenges the very distinction 'public'/'private' that silences or marginalizes women's lived sense of the past. But similar processes of domination operate in relation to specifically working-class experiences, for most working-class people are also robbed of access to the means of publicity and are equally unused to the male, middle-class habit of giving universal or 'historic' significance to an extremely partial experience. But we are only beginning to understand the class dimensions of cultural domination, partly by transferring the feminist insights. Nor is this only a question of class or gender positions. Even the articulate middle-class historian, facing the dominant memory of events through which he has actually lived, can also be silenced (almost) in this way. One telling example is the difficulty of writers of the New Left in speaking coherently about the Second World War:

One is not permitted to speak of one's wartime reminiscences today, nor is one under any impulse to do so. It is an area of general reticence: an unmentionable subject among younger friends, and perhaps of mild ridicule among those of radical opinions. All this is understood. And one understands also why it is so.

It is so, in part, because Chapman Pincher and his like have made an uncontested take-over of all the moral assets of that period; have coined the war into Hollywood blockbusters and spooky paper-backs and television tedia; have attributed all the value of that moment to the mythic virtues of an authoritarian Right which is now, supposedly, the proper inheritor and guardian of the present nation's interests.

I walk in my garden, or stand cooking at the stove, and muse on how this came about. My memories of that war are very different.6

This is followed by a reassuringly confident passage which is a classic text for studying the popular memory of the 1940s, but the struggle is intense, the victory narrow, and the near-silencing of so strong and masculine a voice in the shape of its domestication is very revealing.

It is this kind of recovery that has become the mission of the radical and democratic currents in oral history, popular autobiography and community-based publishing. We will look at these attempts to create a socialist or democratic popular memory later in the argument. But we wish to stress first that the study of popular memory cannot be limited to this level alone. It is a necessarily relational study. It has to take in the dominant historical representations in the public field as well as attempts to amplify or generalize subordinated or private experiences. Like all struggles it must needs have two sides. Private memories cannot, in concrete studies, be readily unscrambled from the effects of dominant historical discourses. It is often these that supply the very terms by which a private history is thought through. Memories of the past are, like all common-sense forms, strangely composite constructions, resembling a kind of geology, the selective sedimentation of past traces. As Gramsci put it, writing about the necessity of historical consciousness for a Communist politics, the problem is' "knowing thyself" as a product of the historical process to date which has deposited in you an infinity of traces, without leaving an inventory'. Similarly the public discourses live off the primary recording of events in the course of everyday transactions and take over the practical knowledges of historical agents. It is for these reasons that the study of 'popular memory' is concerned with two sets of relations. It is concerned with the relation between dominant memory and oppositional forms across the whole public (including academic) field. It is also concerned with the relation between these public discourses in their contemporary state of play and the more privatized sense of the past which is generated within a lived culture.

Popular memory as a political practice

Socialist, feminist and radical historians have always understood that history matters politically. History-writing has sometimes been seen as a way of fighting within one branch of 'science', an attempt to dislodge 'bourgeois' history from its predominant place within the professional intellectual field. Left historians have shared a general, often vague, sense that a history informed by marxist or socialist premises must serve the politics of the present and future. Only rarely, at least in contemporary debates, has this general association been challenged.7 The intrinsically 'historical' character of marxism, as science or as critique, has usually seemed to guarantee the connection.

The political uses of history do seem to us more problematic even from a marxist perspective. This is especially the case when history is defined as 'the study of the past'. We have come to see this as one of the key features of professional history, and indeed, of historical ideologies. Certainly it is deeply problematic from the viewpoint of 'popular memory'. For memory is, by definition, a term which directs our attention not to the past but to the past-present relation. It is because 'the past' has this living active existence in the present that it matters so much politically. As 'the past' - dead, gone or only subsumed in the present - it matters much less. This argument may be clarified if we compare a number of approaches to the political significance of history, returning to some of Bill Schwarz's formulations in Chapter 2 on the Communist Party Historians' Group.8

We may follow Schwarz in distinguishing three main approaches to the political relevance of history. The first approach, while retaining in a strong form the notion that the object of history is 'the past', seeks to link past and present in the form of salutary 'lessons'. These may have a negative force, warning, for instance, against returns to past disasters. The contemporary argument about 'the 1930s', which draws on a conventional left historiography of that decade, is a case in point.9 But this argument may also work more positively, typically by identifying 'traditions' which then become a resource for present struggles. Raymond Williams's Culture and Society tradition, Edward Thompson's tradition of libertarian socialism or communism, formed in the junction of marxism and romanticism, and the socialist feminist succession uncovered in the historical work of Sheila Rowbotham are salient examples here. An even better case is that already discussed by Schwarz: the Communist Historians' construction of a long lineage of popular democratic struggles from the Levellers and Diggers to the socialisms and communism of the twentieth century. More generally still, and the move is typical of Edward Thompson's history, the re-creation of popular struggles shows us that despite retreats and defeats, 'the people', 'the working class' or the female sex do 'make history' even under conditions of oppression or exploitation. In the same way, especially if we are conscious of this lineage, we can make history too.10 The link between past and present, between history-writing and the construction of historical futures today, is in essence an exhortatory one.

A second way of conceiving the past-present relation is to employ historical perspectives and methods as an element in strategic analysis. We start from the need to understand contemporary political problems. We seek to examine the conditions on which contemporary dilemmas rest. In looking at the nature and origins of current oppressions, we trace their genesis as far back as it is necessary to go. Here the relation between past and present is necessarily more organic, more internal. The past is present today in particular social structures with determinate origins and particular histories. This cooler,'scientific' evaluation of the past is best exemplified, as Schwarz suggests, in Perry Anderson's historical project, from 'Origins of the present crisis' to his sequence of major studies in the origins of the modern capitalist world. It is characteristic too of Marx's own historical projects, Capital itself but also the essays on French and English politics, though in Marx it is allied with a more inspirational or agitational mode of history-writing, closer to the first of our own categories and with a similar risk of triumphalism.

A stress on popular memory adds something to both these conceptions, though it does not displace them. The construction of traditions is certainly one way in which historical argument operates as a political force though it risks a certain conservatism; similarly any adequate analysis of the contemporary relations of political force has to be historical in form as well as reaching back to more or less distant historical times. It must also attempt to grasp the broader epochal limits and possibilities in terms of a longer history of capitalist and patriarchal structures.

What we may insist on in addition is that all political activity is intrinsically a process of historical argument and definition, that all political programmes involve some construction of the past as well as the future, and that these processes go on every day, often outrunning, especially in terms of period, the preoccupations of historians.l1 Political domination involves historical definition. History - in particular popular memory - is a stake in the constant struggle for hegemony. The relation between history and politics, like the relation between past and present, is, therefore, an internal one: it is about the politics of history and the historical dimensions of politics.

Some examples may make the implications of this argument clearer. It can be argued that conceptions of the past have played a particularly central role in political life in Britain especially in popular conceptions of nationhood. The intersections between a popular consenative historiography and the dominant definitions of British (but especially English) nationalism have been especially intimate. One problem of left historiographies in Britain that retain 'the nation' or 'the national-popular' as a major affirmative category is precisely the thoroughness of this consenative appropriation of nationhood and the more structural conditions, especially the long history of Empire and of cultural separation from Europe on which it rests.l2 The dominant memory of the Second World War (in which the 'island race' was united under a great leader, Winston Churchill), and its recurrent re-evocation (they/we really can pull together when it's absolutely necessary) is a case in point.l3 Similarly, contemporary racism feeds upon a memory of a nation and of a working class that was white, chauvinistic and dominant on a world-wide scale. Indeed 'the British People' (to whom the legacy of Alfred, Drake, Wellington and Churchill is bequeathed) is, often half-consciously, a racist construction. The dominant nationalist themes, grown cosy and thoroughly naturalized by repetition, disguise or celebrate the actual history of imperial and colonial domination. They define what it means to be British, to 'belong' today. In so doing they marginalize and oppress black people in Britain whose history is precisely the reverse side of the consenative chronicle .14

More particular political questions are fought out on this ground too. Each major political settlement involves its own historiography, academic and popular. The dominant social-democratic and liberal-consenative post-war tendencies, for example, constructed their own history of the 1940s as a period of massive social transformation. For Labour Party 'revisionism' particularly, the 1950s were a postrevolutionary era, to which traditional left analyses of a marxist kind were quite irrelevant. The dominance of this historical account (classically expressed in the political writing of John Strachey and Anthony Crosland) marginalized socialist and marxist explanations in this period.l5 We can see similar processes at work today. Contemporary Thatcherism has constructed its own historical account that centres on the failure of the whole arc of post-war politics and the growth of a bureaucratic statism. Similarly, attempts to create a new liberal and socialdemocratic centre in British politics are pursued, as Dave Sutton has suggested, partly by historical means, including an extensive re-evaluation of the 'new Liberalism' of the late nineteenth and early twentieth centuries.l6 The historio- graphy of the Labour Party, too, has taken on new dimensions in the light of the growth of socialist currents within it. The historical debate about the 1930s has acquired a new urgency with the growth of unemployment and the constant referencing (sometimes in very conservative terms) of this decade by Labour politicians and trade union leaders. Here again one finds a significant 'revisionist' project, now launched from a position right of centre in British politics, that discovers that the 1930s were not really so bad after all!'7 If we remind ourselves of the centrality of history-writing and teaching for a long succession of left intellectuals in Britain, the fact that marxism has taken characteristically historical forms and the strength of a feminist, especially a socialist-feminist historiography, the importance of history as a ground of political struggle seems confirmed. Explanations of this cultural centrality of history are evasive and need a proper comparative context: do all nation-states with long histories develop an extensive historiographical culture? Perhaps the very early formation of the nation-state and the long subsequent continuities, unbroken by revolution except in the seventeenth century, is one part of an explanation. If we are looking to fill the 'absent centre' of British national culture, history and the sense of the past might prove a very strong candidate.l8

The formation of a popular memory that is socialist, feminist and anti-racist I is of peculiar importance today, both for general and for particular reasons. Generally, as Gramsci argued, a sense of history must be one element in a strong popular socialist culture. It is one means by which an organic social group acquires a knowledge of the larger context of its collective struggles, and becomes capable of a wider transformative role in the society. Most important of all, perhaps, it is the means by which we may become self-conscious about the formation of our own common-sense beliefs, those that we appropriate from our immediate social and cultural milieu. These beliefs have a history and are also produced in determinate processes. The point is to recover their 'inventory', not in the manner of the folklorist who wants to preserve quaint ways for modemity, but in order that, their origin and tendency known, they may be consciously adopted, rejected or modified.l9 In this way a popular historiography, especially a history of the commonest forms of consciousness, is a necessary aspect of the struggle for a better world.

More particularly, the formation of a popular socialist memory is an urgent requirement for the 1980s in Britain. Part of the problem is that traces of a politicized memory of this kind chart, on the whole, a post-war history of disillusionment and decline. In particular, there is a sense of loss and alienation so far as the Labour Party is concerned. But the problem is deeper than this difficulty (which, even now, the socialist revival within and outside the Labour Party may be lessening). For what are to be the forms of a new socialist popular memory? A recovery of Labour's past will hardly do; nor is it helpful to chart the struggles only of the male, skilled, white sectors of the working class who have formed the main subjects of 'labour history' to this day. We need forms of socialist popular memory that tell us about the situation and struggles of women and about the convergent and often antagonistic history of black people, including the black Britons of today. Socialist popular memory today has to be a newly constructed enterprise; no mere recovery or re-creation is going to do. Otherwise we shall find that nostalgia merely reproduces conservatism.

Resources and difficulties


The resources for such a project are great but they are also, in important ways, very disorganized, systematically disorganized that is, not merely 'lacking organization'. This has much to do with the diverse social origins of different kinds of resources and the immense difficulties of their combination. For many resources have, in the last two decades, been created through the critical work of academic practitioners - especially, in our field, historians, sociologists, philosophers and so on, dissatisfied with the limits and ideologies of their professional discipline. 'Cultural studies' has developed along these lines, but belongs to a very much wider field of radical and feminist intellectual work where much of the stress has been, till lately, upon theoretical clarification and development. But there have been important breaks outside the academic circles too, or in a tense relation to them. They have been most commonly connected to adult education (especially the WEA) or to schoolteaching or to post-1968 forms of community action. The principal aim of these tendencies has been to democratize the practices of authorship; in the case of 'history' to lessen or remove entirely the distance between 'historian' and what Ken Worpole has called 'the originating constituency'. The characteristic products of this movement have been popular autobiographies, orally based histories, histories of communities and other forms of popular writing. But it has also developed a characteristic critique of academic practice that stresses the inaccessibility even of left social history in terms both of language and price, and the absorption of authors and readers in the product (book or journal) rather than the process by which it is produced and distributed. Partly because of the stress on 'language' and the commitment to 'plain speech', oral-historical or popular-autobiographical activists are often deeply critical of the dominant forms of theory. It is this division that is, in our opinion, a major source of disorganization. The tensions between the 'activist' and 'academic' ends of radical historical tendencies are explosive to a degree that is often quite destructive. They are often qualitatively less productive than directly cross-class encounters in which workingclass people directly interrogate academic radicals. Even so there is a beginning of useful connections between academic 'critics' and community activists (who are not always different persons); where patience holds long enough on either side there are the beginnings of a useful dialogue. Some of this can be traced in the pages of History Workshop Journal, the conference volume to History Workshop 13 and in the writings, especially, of some authors whose experience spans an 'amateur' and 'professional' experience.20 In general History Workshop (as journal and as 'movement') has been distinguished by its attempt to hold together these two unamiable constituencies along with other groups under the banner of 'socialist' or 'people's' history. In this sense History Workshop is the nearest thing we have to an alternative 'historical apparatus', especially if its own recently-formed federation is placed alongside the older Federation of Worker Writers and Community Publishers.21 In what follows we want simply to note some developments, within and outside the History Workshop movement that seem to us already to point towards the study of popular memory.

It is oral history - the evocation and recording of individual memories of the past - which seems, at first sight, nearest to the popular memory perspective, or one aspect of it. In fact the term oral history embraces a very large range of practices only tenuously connected by a 'common' methodology. What interests us most about oral history is that it is often the place where the tension between competing historical and political aims is most apparent: between professional procedures and amateur enthusiasm, between oral history as recreation (in both senses) and as politics, between canons of objectivity and an interest, precisely, in subjectivity and in cultural forms. Later, we want to illustrate these tensions by looking at the early work of the oral and social historian Paul Thompson. There are good reasons for choosing Thompson's work. He is both a socialist and a professional historian. He has done more than anyone to introduce and codify the use of oral methods in this country. With Thea Vigne, he organized the first large-scale SSRC-funded oral history research project.22 He is also author of the first lengthy 'introduction to the use of oral sources for the historian'.23 He is editor of Oral History, the main medium of communication between oral historians, and is closely associated with History Workshops.24

In focusing part of our argument around Thompson's work, we do not mean to imply that there are not alternative models.25 Other adaptations of oral history are, indeed, much nearer to our own concerns. We would cite for example the critique of oral history, in its more empiricist forms, to be found in Luisa Passerini's work.26 Her pursuit of the structuring principles of memory and of forgetfulness, her concern with representation, ideology and subconscious desires, her focus on 'subjectivity' as 'that area of symbolic activity which includes cognitive, cultural and psychological aspect',27 and her understanding of subjectivity as a ground of political struggle, all bring her work very close to British traditions of cultural studies, especially where they have been influenced by feminism. Her critique of oral history seems to us much more radical than its sometimes guarded expression might suggest. And we agree absolutely with her criticisms of English debates for the failure to connect oral history as a method with more general theoretical issues.25 The beginnings of her analysis of popular memories of Italian fascism in Turin mark a large advance on most thinking about the cultural and political (as opposed to merely 'factual') significance of oral history texts.

Although there is a beginning of a more self-reflexive mood in Britain, the strengths here lie more in a developed practice of popular history, often building on the social and labour history traditions. This is the case, for example, with the most stunning single work drawing on evoked memories of participants - Ronald Fraser's Blood of Spain.29 The lessons of this book for future practice lie more in the way it is written than in any self-conscious prescriptions by the author, a long- time practitioner of oral history or 'qualitative sociology'. What we found interesting in Blood of Spain was the use of oral remembered material in something like the form in which it is first evoked: not as abstracted 'facts' about the past, but as story, as remembered feeling and thought, as personal account. The whole book is woven from such stories and retrospective analyses, sometimes quoted, sometimes paraphrased, clustered around the chronology of the Spanish Civil War or the make- or-break issues that were debated and literally fought out in its course. There is a sense in which Fraser's interviewees actually 'write' Blood of Spain by providing the author with the cellular form of the larger work: innumerable tiny personal narratives from which is woven a larger story of heroic proportions and almost infinite complication. Blood of Spain is history through composite autobiography, the re-creation of experience in the form of a thousand partial and warring view- points.30

But is arguable that the most significant development has been the growth of community history, popular autobiography and working-class writing more generally, where the terms of authorship have been more completely changed. In one sense, all these texts and projects are evidence for the forms of popular memory; they are all about the relation of past to present, whether self-consciously 'historical' or not. Some projects, however, have specifically focused on these themes: the chronologically-ordered sequence of accounts of work in Centreprise's Working Lives, part of the People's Autobiography of Hackney, is one example ;3 the work of the Durham Strong Words Collective, especially Hello Are You Working? (about unemployment) and But the World Goes on the Same (about past and present in the pit villages) is another.32 The Durham work is especially organized around contrasts of 'then' and 'now', often viewed through inter- generational comparisons. As the editors put it:

The past exerts a powerful presence upon the lives of people in County Durham. The pit heaps have gone but they are still remembered, as is the severity of life under the old coal owners and the political battles that were fought with them. As they sit, people try to sort things out in their minds - how were things then? How different are they now? And why?33

Different from either of these projects are the politically located, culturally sensitive projects around history and memory that have developed within the contemporary Women's Movement. There is already a strong past-present dialogue at work within contemporary feminism as Chapter 8 shows. Much feminist history also draws on oral materials, sometimes using them in innovative ways.34 The autobiographies evoked by Jean McCrindle and Sheila Rowbotham, and published as Dutiful Daughters, are framed by the editors' feminism and by a distinctive politics of publication. The aim is to render private feminism oppressions more public and more shared, thereby challenging dominant male definitions and the silencing of women.35 Works like this continue a long feminist tradition of writing about past and present through autobiographical form. We might also note in this collection, in the Durham work, in Jeremy Seabrook's What Went Wrong? and elsewhere the beginnings of an interest in a specifically socialist popular memory.

It was interesting that both Dutiful Daughters and What Went Wrong? were the subjects of 'collective reviews' at History Workshop 14 36

Not all relevant practices and debates belong to what would usually be thought of as 'historical' work. Indeed, there is a real danger that 'History', who is often a very tyrannous Muse, will draw the circumference of concerns much too narrowly. That is one reason why the broader categories - black, or women's or working-class 'writing' for example - are sometimes preferable. Even here, though, there are unhelpful limitations: the commitment, for example, to the printed word and the tendency to neglect other practices including the critique of dominant memory in the media. It is here that debates on 'popular memory' which come out of a completely different national and theoretical tradition are so important, especially debates in France around Michel Foucault's coinage of 'popular memory' as a term.37 French debates focus on such issues as the representation of history in film and around the 'historical' policies of the French state - for example the Ministry of Culture's promotion of popular history and archival retrieval during the of ficial Heritage Year of 1979.38 Another important French voice for us has been Jean Chesneaux's Pasts and Presents: What is History' For?, a militant and sometimes wildly iconoclastic attack on French academic history, including academic social history written by marxists.

One importance of the French debates is that they have directed attention to the possibility of radical cultural practice of en 'historical' kind outside the writing of history books.39 It is important to note developments of this kind in film, community theatre, television drama and radical museum work. The film 'Song of the Shirt', the television series 'Days of Hope', the television adaptation of Vera Brittain's Testament of Youth and the strong historical work of radical theatre groups like 7:84, Red Ladder and The Monstrous Regiment are examples of 'history-making' often with a real popular purchase, yet usually neglected by historians. Innovations in this area are intrinsic to popular memory both as a study and a political practice. They should certainly receive as much interest and support from socialist and feminist historians as the latest historical volume, or the newest issue of 'the journal'.

Diffculties and contradictions

What, then, are some of the difficulties in realizing the potential of these resources? Oral history and popular autobiography have, after all, now been around for some time, initially generating a real excitement. Why have the political effects been fairly meagre? What are the remaining blocks and inhibitions here?

There are, perhaps, four main areas of difficulty. Very often these have to do with the tensions that exist between the academic or professional provenance of new practices and their adaptation to a popular politics. We will summarize the four areas of difficulty briefly here, then in the rest of the article consider each at more length.

The first set of difficulties is epistemological in character. They arise from the ways in which 'historical' objects of study are defined. They revolve around the empiricism of orthodox historical practice. They are not purely technical matters for philosophers to adjudicate. The historian's empiricism is a real difficulty. It blocks political progress. That is why it is so important to return to these questions once more, showing the political effects of this persistently empiricist stance.

The second set of difficulties derive initially from the form in which the 'raw material' of oral history or popular autobiography first arises: the individual testimony, narrative or autobiography. This poses, in a very acute form, the problem of the individual subject and his or her broader social context. In what sense is individual witness evidence for larger social changes? How can these changes themselves be understood, not as something that evades human action, but also as the product of human labour, including this individual personality? This difficulty runs through the oral history method and through the autobiographical form. It is also reflected in larger divisions of genres: history, autobiography, fiction (with its particular experiential truth). Such divisions in turn encapsulate hierarchies of significance. The oral-historical witness or the autobiographer, unless held to be a personage of exceptional public power, speaks only for herself; it is the historian who, like the Professor in Lucky Jim, speaks literally for 'History'. Some resolution of this persistent problem, some way of thinking the society of individuals, would be an important additional resource.

We have already touched on a third set of difficulties: the tendency to identify the object of history as 'the past'. This largely unquestioned feature of historical common sense has extremely paradoxical results when applied to oral history or popular autobiography. Indeed it shows us that this definition cannot be held without a radical depoliticization of the practice of research. What is interesting about the forms of oral-historical witness or autobiography are not just the nuggets of 'fact' about the past, but the whole way in which popular memories are constructed and reconstructed as part of a contemporary consciousness. In this section we will look at some of the characteristic ways in which a sense of the past has been constructed in private memories.

The fourth set of difficulties is more fundamental. It concerns not just the manifest intellectual and theoretical blockages, but the social relations which these inhibitions express. In oral history and in similar practices the epistemological problem - how historians are going to use their 'sources' - is also a problem of human relationships. The practice of research actually conforms to (and may in practice deepen) social divisions which are also relations of power and of inequality. It is cultural power that is at stake here, of course, rather than economic power or political coercion. Even so research may certainly construct a kind of economic relation (a balance of economic and cultural benefits) that is 'exploitative' in that the returns are grossly unequal ones. On the one hand there is 'the historian', who specializes in the production of explanations and interpretations and who constitutes himself as the most active, thinking part of the process. On the other hand, there is his 'source' who happens in this case to be a living human being who is positioned in the process in order to yield up information. The interviewee is certainly subject to the professional power of the interviewer who may take the initiative in seeking her out and questioning her. Of course, the problem may be solved rhetorically or at the level of personal relations: the historian may assert that he has 'sat at the feet of working-class witnesses' and has learnt all he knows in that improbable and uncomfortable posture. It is, however, he that produces the final account, he that provides the dominant interpretation, he that judges what is true and not true, reliable or inauthentic. It is his name that appears on the jacket of his monograph and his academic career that is furthered by its publication. It is he who receives a portion of the royalties and almost all the 'cultural capital' involved in authorship. It is his amour propre as 'creator' that is served here. It is his professional standing among his peers that is enhanced in the case of 'success'. In all this, at best, the first constructors of historical accounts - the 'sources' themselves - are left untouched, unchanged by the whole process except in what they have given up - the telling. They do not participate, or only indirectly, in the educational work which produces the fnal account. They may never get to read the book of which they were part authors, nor fully comprehend it if they do.

We have deliberately overdrawn this case, to make the point polemically. But we do not describe an untypical situation for the more professionalized types of oral-historical practice. The question is what are the wider effects of such social divisions? Are they transformable? To what extent, locally, fragilely, have they already been transformed? And what are the difficulties and opportunities involved in further transformations? Much is at stake here. We are discussing a particular form of class relation (that between working-class people and sections of the professional middle class) and how it can be transformed into a more equal alliance. It is an alliance that happens to have been a crucial one in the history of left politics and one which is certainly central to the future of socialism and feminism today.

The problem of empiricism

By empiricism we mean the epistemological doctrine that holds that the test of true knowledge lies in observation, 'experience' or the collection of 'facts'. This may be understood in a classically inductivist way, factual accumulation producing knowledge in a more general form, or it may rest on more strictly positivist procedures: the validation or falsification of particular hypotheses by experiment or observation. When professional historians describe their research procedures they almost always employ empiricist formulations. But the historian's empiricism takes particular forms, influenced by the elementary experiences of archival research.

Research is seen as a dialogue between the historian and 'his evidence'. In this, the historian takes the active part but represents also the subjectivity of inquiry or hypothesis. It is the 'source', the product of a now unchangeable past, that provides the possibility of a knowledge that is objective if it is honestly and critically interrogated. Indeed in some justifications of historical research this procedure is seen as more objective than those of natural science:

As a matter of fact, in a very real sense the study of history is concerned with a subject matter more objective and more independent than that of the natural sciences. He [the historian] cannot escape the first condition of his enterprise, which is that the matter he investigates has a dead reality independent of the inquiry.40

Although Elton's formulations are perhaps extreme and were developed in a polemic against what he saw as E. H. Carr's relativism,41 we also think they still accurately represent the common sense of historical professionalism. It is this sense of the facticity and determinancy of the source which unites a discipline otherwise fragmented into thematic and period subspecialisms. This is even reflected in what, not entirely in jest, one might call the small talk of the historical subculture. How do you start a conversation with a historical researcher? Why by asking, of course, 'What's your period? What are your sources?' Or, 'What are you working on?'

Oral history as it has developed in Britain within a professional historical practice has conformed very much to this historical common sense. The tensions that have resulted are very well illustrated in the early work of Paul Thompson, especially in The Voice of the Past (1978) which is Thompson's authoritative introduction to oral history method and in The Edwardians (1975) which is the first fruit of the Essex Project, a national interview study of the family, work and community life before 1918. In what follows we will use the Voice of the Past in particular to suggest ways in which an historian's empiricism may limit the radical potential of oral history practice. We should stress that we are not concerned to criticize Thompson himself as author or as historian, recognizing the necessary limits of pioneering projects. But the example is a particularly salient one for some typical historian's dilemmas.

Oral history - the wrone defence?

The Voice of the Past is addressed to 'the practitioner' who emerges as a curious composite figure. On the one hand it is addressed to professional historians who are to be persuaded that the use of oral sources is 'perfectly compatible with scholarly standards'. On the other it is addressed to a popular socialist audience with which Thompson himself identifies:

I myself believe that the richest possibilities for oral history lie within the development of a more socially-conscious and democratic history.42

From this perspective oral history is a transformative socialist practice in terms of purpose, content and the social relations of its production:

(History) should provide a challenge, and understanding which helps towards change. And for the historian who wants to work as a socialist the task must be not simply to celebrate the working class as it is but to change its consciousness.... (Oral history) provides the means for a radical transformation of the social meaning of history.43

Like all texts The Voice of the Past is susceptible to different readings. Our own reading, developed in detail elsewhere, is that these two aims are inherently contradictory and are mainly resolved in a conservative direction. It is professional history, not history as a popular socialist practice, that remains in dominance in this text.

There are two main ways in which this happens. In the end both depend upon the adoption of empiricist solutions, hardened, here and there, by a reliance on borrowings from a more narrowly positivist social science, especially a quantitative sociology and an experimental psychology. The first empiricist move is to rest the case for the radicalism of oral history less upon a theoretical and political standpoint, more upon the method itself. The Voice of the Past does not develop political and theoretical criteria for a socialist practice; indeed its political content remains quite unspecified. In a typical empiricist move, it is the method itself which guarantees, or at least encourages, the political effects. Oral history makes available a new and untapped source in the testimony of living people.

The historian's traditional sources are administrative and other records of authority. In Chesneaux's characterization: 'Our memory is that of the power structure which functions as a gigantic recording machine'.44 Thompson provides an interesting critique of such sources and argues that, by contrast, oral history makes possible 'a much fairer trial' - 'witnesses can now also be called from the underclasses, the unprivileged and the defeated'.45 This may provide a challenge to the established account. But this is not all. Oral history also demands of the historian an altogether different mode of working:

Historians as field-workers, while in important ways retaining the advantages of professional knowledge, also find themselves off their desks, sharing experience on a human level.46

Tempted out of the record of fice into the living room, the historian finds that the process of research no longer involves working on inert material. We are in face to face contact with living people, asking them to talk about their experiences and share their understandings. The process of research is necessarily interpersonal and depends on human rapport. The straightforward human contact implicit in oral history works against any tendency to 'objectify' the material:

For the historian comes to the interview to learn: to sit at the feet of others who, because they come from a different social class, or are less educated, or older, know more about something.47

Oral history, by main force of the method, provides the means to deprofessionalize history: 'it gives history back to the people in their own words'.4B According to this view, then, oral history will transform the social relations of research because it is inherently democratic. It will transform the content of history because it necessarily provides an alternative viewpoint, a viewpoint 'from below'.

These formulations are empiricist because they rest on argument about alternative content and transformative practice on the nature of the historical source itself. They neglect the relations of power that enter into the method, unconsciously because not theorized, at every point from the devising of an interview schedule to the presentation of the final explanatory account. Something more than method is needed, in other words, to render oral history into socialist practice. But this conservatism is greatly strengthened by the other coexisting discourse in Thompson's text, the discourse of 'professional standards'. We can best understand this strand by recalling some of the criticisms made of oral sources from a stance more firmly within orthodox history.

These criticisms amount to a two-pronged attack. The first points to the 'fallibility' of memory. Memory is a highly selective process which registers some processes and discards others. Stephen Koss, reviewing The Edwardians, supported this familiar point with an example. He once interviewed a woman about an episode in her life which, during the course of his research, he had read about in a letter. She could remember nothing about it and denied that the episode had taken place as he described it. Such comparisons, Koss argued, testify to the incompleteness of memory, its gaps and absences. But Koss raised another set of problems:

His Edwardians, after all, have lived on to become 'Georgians' end now, 'Elizabethans'. Over the years certain memories have faded or, at very least, may have been influenced by subsequent experience. How many of their childhood recollections were in fact recalled to them by their own elders? What autobiographies or novels might they have since read that would reinforce certain impressions at the expense of others? What films or TV programmes have had an impact on their consciousness? It would be interesting to know whether Lady Violet Brandon (a character quoted in The Edwardians) had read Vita Sackville-West's The Edwardians, or whether Grace Fulford watched Upstairs Downstairs. More generally, to what extent might the rise of the Labour Party in the post war decade have inspired retrospective perceptions of class status and conflict?49

These are tricky questions indeed, but there is more than one way of responding to them. Within an empiricist framework in which sources are used to re-create a given factual past, the continuous transformations of memory under the force of subsequent experience and shifting interpretative frameworks, do indeed pose major difficulties. For Koss these difficulties are adjudged insuperable; oral witness provides at best ungeneralizable, vivid, poignant insights that are usually impossible to validate. Factual reality - what did happen as opposed to what people might believe happened - can only be known if the necessary limits, biases and distortions of the source material are located and taken into account by the tried and tested methods of the craft. The problem then becomes how to apply such methods to a new type of source, to show in fact, that it is as reliable as any other.

It is this route that Thompson takes in his defence of oral history against the professional histotians. He adopts two main lines of argument. He shows first, rather convincingly, that all historical sources are biased in this sense, since they share a specifically human provenance and social purpose.50 But he also develops a whole battery of methods designed to safeguard the accuracy of oral history practice. These involve the careful comparison across sources, the demonstration, from experimental social psychology, of certain regularities in patterns of memory and forgetfulness, an excursion into the biochemistry of the brain in processes of ageing, guidance on the proper forms of questioning to avoid 'retrospective bias', and the adoption of norms of 'representativeness' drawn from the sampling methods of quantitative sociology, and the adequate (sociological) classification of the statistical population.51

The inadequacy of these responses has been fully discussed elsewhere.52 We wish here to stress their costs, especially in relation to the first (more productive) strand in The Voice of the Past. As we will argue in more detail later, these responses serve inexorably to reproduce the very social divisions which oral history is supposed to transform. They systematically privilege the historian and researcher as bearer of the scientific canon. This can be best seen, perhaps, in the way 'scientific'requirements must necessarily impose on the forms of the 'human contact' of the interview itself. The historian must needs approach 'respondents" with a set of standardized questions. A project of the scale of The Edwardians requires an hierarchical division of labour among the researchers themselves. The historian employs a team of inteniewers to perform a particular and relatively programmed research process under his or her direction. Methodology is then largely concerned with 'interview technique': with the means by which people can be persuaded to give you the information which you require of them. The contradictions cannot but be carried through to the moment of interviewing itself:

The problem is to introduce sufficient standardisation without breaking the interview relationship through inhibiting self expression.... Encouraging the informant to free expression but gradually introducing a standard set of questions . . . protects the inteniew relationship but makes the material less strictly comparable.53

These are all too familiar problems within the 'parent' discipline from which the techniques are drawn. In sociology, however, this kind of confidence in a positivist methodology has long been undermined.

A second major cost of the empiricist defence of oral history is to render cultural determinations and effects quite marginal. This has two separate and related aspects. Empiricist canons tend to render invisible the 'culture' of the historians themselves. The belief in a model of neutral factual validation underestimates the practical influence of the researcher s own values, theories and preoccupations. Readings of sources, while perhaps not infinite and certainly not arbitrary, are none the less various and competing. Contrary to Paul Thompson's implicit epistemology, major shifts in interpretation or new directions in historiography are more likely to arise from changes in political or theoretical preoccupations induced by contemporary social events than from the discovery of new historical sources. If confirmation is needed here we have only to cite the origins of the new social history of the 1950s and 1960s whose specific political and theoretical genesis has been described by Bill Schwarz. The Communist Party Historians' Group induced a revolution in historiography (without the benefit of oral history). New historical questions inspired the search for new sources. In fact, Paul Thompson's history does rest upon specifically theoretical premises which inform, for example, the conceptions of 'society' and of 'social change' that organize the use of oral witness in The Edwardians. The difficulty of the empiricist stress on 'method' and on 'source' has always been to disguise the theoretical premises of accounts.

But empiricist procedures also disguise the culturally-constructed character of 'the sources' themselves. If we treat historical sources only as bearers of 'fact', we will tend, like Paul Thompson, to be interested mainly in certain orders of facticity, concerning past action and behaviour. We will tend to be less concerned with the historically and socially constructed values which are the very medium of this 'information'. At its most banal, historical practice simply ignores this cultural framing of 'fact' - all the symbolic and linguistic features through which human meaning is conveyed. Historians do sometimes treat sources as though they were a transparent medium. In the more critical professional mode frameworks of meaning are treated only negatively: they are a problem, a bias, a distortion. Critical procedure attempts to strip them away revealing the verifiable 'fact' inside. Actually, it is very doubtful if empiricism even describes this process accurately, let alone tells us what it should be. Since all historical interrogation operates, however flexibly, within its own frameworks of meaning, the reading of sources is better described as a process of the competition of theories. Our own explanations are judged not against the pre-given facticity of the source (except in the vulgar sense of its material existence) but against the human constructions of meaning that are found there. In the end, the fact-value distinction is itself difficult to sustain since facts only signify, only have human meaning, within explanatory frames.

The problematic of 'bias' or 'distortion' is particularly destructive of any historiography, like the discussion of 'popular memory', that wishes to give due weight to cultural determinations. Following Luisa Passerini, we would add that it is also peculiarly destructive of any open, sensitive and full use of oral history.

We cannot afford to lose sight of the peculiar specificity of oral material, and we have to develop conceptual approaches - and indeed insist upon that type of analysis - that can succeed in drawing out their full implications. Above all, we should not ignore that the raw material of oral history consists not just in factual statements, but is pre-eminently a representation and expression of culture, and therefore includes not only literal narrations but also the dimensions of memory, ideology and sub-conscious desires.54

From the point of view of the study of popular memory or of cultural phenomena generally, empiricist methods (whether the 'administered questionnaire' or the empiricist interrogation of the source) are of very little value. We might say, indeed, that the study of popular memory can begin only where empiricist and positivist norms break down.55 The alternative and stronger responses to Koss-like criticisms of oral history are, then, as follows: yes, indeed, memory and its narratives are cultural constructions in much the way that your histories are. To illuminate both, and especially to help popular memory to a consciousness of itself, requires an understanding of specifically cultural processes and particularly of the making and remaking of memory, on both an individual and a social level. In this way your 'problem' becomes our 'resource', your insuperable difficulty our agendum, your closure our starting-point.

Alternative readings: for structure and culture

The critique of empiricism is familiar enough. There is, however, one persistent problem in breaking with this framework. The fear is that once the cultural character of accounts ('secondary' or 'primary') is admitted, there is no recourse from a relativism in which all accounts are equally socially determined, equally 'valid' and equally subject to political pragmatism. This is to vacate the ground of a serious intellectual project, certainly of a social science.

A concern with 'scientistic' criteria of validity sits well with the monopolization of certain forms of knowledge. None the less, the problem is a serious and general one. It is necessarily posed whenever knowledge is seen as a strategic political resource. In such situations the truth content of any account, and its competitive advantage over rival accounts, matter very much.

There are, however, epistemological alternatives to empiricist codes that recognize the importance of the cultural without falling into a subjectivism. Some of them have been discussed by Greg McLennan in this book under the general rubric of 'realism'.56 These options start from the position that there is indeed an objective social world which has changed, historically, in ways that are potentially knowable, but does not reveal its secrets by simple observation or the testing of hypotheses. Conscious human activity is the main constituent of this social world even where it unconsciously reproduces existing structures and processes. The subjective dimensions of human action are as much 'social facts' as externally obsened behaviour. They possess their own forms, their own dependence on non-cultural processes and their own determinacy on social outcomes. Our theories are both socially determined products and active interventions in the social world.

They are never neutrally 'scientific' or 'objective' in some socially external sense. They are, however, more or less adequate as guides to practice because they conform, more or less, to actual social processes and historical change. If this description could apply to a range of theoretical and epistemological positions, it is especially characteristic of those forms of historical materialism that take cultural and ideological processes seriously while holding to a broadly materialist frame.

Although the argument against empiricism seems abstract, it does have impor- tant implications for the way we understand what is going on in oral-historical and autobiographical accounts. It suggests, in fact, two different but mutually dependent ways of reading which we might call the 'structural' and the 'cultural'. A structural reading (which resembles in important respects the historian's 'factual' reading of a 'source') is interested in the conditions which the author of an account has appropriated subjectively - the conditions, structures, processes that have formed, sometimes consciously, sometimes not, this particular lived experience. Such a reading is based on realist premises in that it assumes that what is signified in such accounts has some real existence outside the text and is not wholly constituted in the writing itself. It departs from an empiricism, however, to the extent that the relation between events or process on the one side and the account on the other is rendered thoroughly problematical and subject to its own transformations. This allows us to see that in the search for validity in the face of professional criticism oral history practitioners have greatly underestimated the complexity of their own 'sources'. Oral or autobiographical accounts are both richer and less strictly 'reliable' than has been suggested.

The process of 'reading through' an account to its 'factual' substratum is an extremely complicated business. It depends on at least two further conditions: the presence of sources of relevant knowledge other than the account itself and, as important, the presence of some explicit and productive theory of social relations and of forms of consciousness. The first allows us to bring to view determinations that lie outside the range of vision of a given account; the second allows us an understanding of the relation between social being and the forms in which authors become conscious of their social condition and history. In this respect we agree very strongly with Jerry White in his recent exchange with Ken Worpole and Stephen Yeo. He argues for a more critical, more explanatory and more socialist local history, suggesting that this may involve confronting the way in which people understand their own oppressions.57 For White popular autobiography is a raw material for the development of socialist understandings (a raw material in Marx's classic sense - already produced under given conditions, not a simple 'given'). It needs, however, a further labour. For Worpole and especially for Yeo the very act of working-class writing and expressivity is a slow, necessarily accumulative step along the road to a 'long revolution'.

The difficulty with this argument is that it uncritically assumes that working-class experience will, in the end, produce socialist or proto-socialist forms of understanding.58 Though this differs from the more formal empiricisms so far discussed, it shares with them a lack of curiosity about specifically cultural determinations, especially those which are of an ideological kind because they hide contradictions, reproduce limitations of understanding, and divert or stultify political energies.

Part of the problem is that White seems to be arguing a deeply unpopular case. This is because, in these debates, secondary analysis, second thoughts, further research, the struggle for deeper and more explanatory accounts are always associated with the intervention of middle-class commentators, if not professional historians or sociologists. In this way two different but related questions are confused: the need for more adequate knowledge, both in political and explanatory terms and, on the other hand, the existing forms of the social division of intellectual labour. To advocate, therefore, a more analytical approach is to side with the theoretical arrogance of the intellectuals. Historical or community 'activists', in reaction to this threat of cultural domination, take a protective or even a possessive stance on behalf of what they see as really indigenous or authentic popular understandings. This circle (which we have seen repeated many times) actually re- produces the existing relations of power and knowledge. It separates theory (with its tendency to idealism) from the realistic grasp of circumstances (with its tendencies to fatalism and parochialism). It evades or forecloses the one encounter that is of vital political significance and inevitable explosiveness: between a theoretical middle-class left politics and the structures, cultures and problems of everyday working-class life. In fact there is no necessary connection between a more analytical or explanatory local history and a middle-class cultural domination. The aim should be to generalize the skills of secondary analysis and ancillary research, not to hold them at bay in deference to a more accessible wisdom.

Cultural readings

A cultural reading would focus on the ways in which the account makes sense of a structured experience or life history. This perspective rests on two main premises. The first is that all accounts, whether in answer to the researcher's questions or more autonomously produced, are highly constructed texts or performances. Certainly the model of (more or less adequate) recall or of the tapping of memories is completely inadequate here. It is plain, reading such accounts, that they are the product of thought, artifice, verbal and literary skills, always involving authorship in this sense, having (like all 'sources') an active presence in the world. These skills are not necessarily literary in form; they are very commonly verbal, bearing the signs of transcription from lively conversations and story-telling.59 So information about the past comes completely with evaluations, explanations and theories which often constitute a principal value of the account and are intrinsic to its representations of reality. Hence the feeling not uncommonly experienced in reading secondary interpretations of first accounts: we wish the bloody historian would go away and let us listen to the account itself! It seems more interesting, more nuanced, more complex and actually more explanatory than its secondary appropriation allows. It is here that we understand the force of the Worpole/Yeo argument and the frequent feeling of revelation to the middle-class listener. Much of this (and the exaggerated inversions which it sometimes produces) comes from the force of dominant ideologies which daily imply that working people have very imperfect understandings - just as it was once argued they had 'no culture'.

The second premise of a cultural reading is, however, that the cultural features of accounts are not simply the product of individual authorship; they draw on general cultural repertoires, features of language and codes of expression which help to determine what may be said, how and to what effect. In charting such repertoires, we might start, for example, from the repeated observation of the centrality of story-telling to working-class accounts of social reality. More or less extended narratives about past events, often of an intimate and always of a personal kind, are certainly one elementary form of popular memory and the commonest way in which past and present are compared and evaluated. These stories are made to carry great significance and may be deliberately told (in the face of probings) in place of more generalized conclusions. Stories come also with intricate variations of form: on the one side tightly sewn up and 'closed', punctuated with a moral at the end, a particular, confident male form; on the other, more open narratives, completed rather by a laugh, often in self-depreciation, dealing with some embarrassing or difficult past happening; then again stories of a largely apochryphal status, more in the form of the proverb, often encapsulating some element of collective memory.60 A developed cultural reading would have to understand such forms and something of their provenance and of their effects (which we would not suppose to be inherently 'biased' or 'ideological').

It would be in the relation of these two readings - the structural and the cultural - that the most important understandings might be reached. Very often this will be a matter of convergence or the confirmation of existing theories. From a marxist-feminist perspective for instance, we would expect to find that life histories accord a special place to those events that are critical for the social existence of a worker or a woman/housewife/mother. And indeed autobiographical accounts do give an important organizing role to salient experiences of this kind: the first entry into labour in the shape of stories about 'my first job', the entry into the employer's household as a critical moment in domestic service, the comparison among women of 'my own' childhood and the way 'I wish to bring my own children up', the moments of courtship and marriage and of major life transitions in general. In these eases 'salient experience' is, in Edward Thompson's phrase, genuinely 'a junction concept':61 it highlights the historial and social position of the author; it is a structural hinge or condition around which, in historical reality, a life has so far been lived. But it also organizes accounts culturally, appearing there with a particular force and emphasis, as something around which many significant stories are told.

This pattern of convergence is, however, by no means invariable and the two readings may actually be productive of surprises that force the modification or extension of understandings. On the one hand, for example, obvious structural or historical features (known from other sources or predicted theoretically) may not show up at all in accounts. On the other hand there may be surprising and initially inexplicable preoccupations. Most interesting here are Passerini's findings on the memories of working-class militants in Turin during the struggles of the fascist period:

Oral sources refuse to answer certain kinds of questions; seemingly loquacious, they finally prove to be reticent and enigmatic, and like the sphinx they force us to reformulate problems and challenge our current habits of thought.... Indeed, I received what to my ears were either irrelevant or inconsistent answers. 'Irrelevant' answers were mainly of two sorts; silences and jokes.62

Silences were two kinds. Sometimes whole life stories were told 'without any reference to fascism, except for casual ones'. Sometimes, especially among the politically conscious, there was simply 'a striking chronological gap' between Mussolini's rise to power and the Second World War. Later on, in second, more probing interviews, similar patterns were repeated. Passerini concludes that in an important sense the period between 1922-5 and 1941-3 has been erased from memory: 'this self-censorship is evidence of a scar, a violent annihilation of many years in human lives, a profound wound in daily experience'.63

Historically or structurally significant events or processes do not necessarily show up, then, in retrospective reconstructions. Our two readings, put together, will often refer us to cases of these kinds: to silences, suppressions, amnesias and taboos. These cases, in turn, force us to 'think again' about structural significance, modifying secondary theories. These surprises may, however, happen in a different way: a particular story or episode may carry more significance for its author than its structural or historical salience may, at first sight, warrant. There are several possibilities here. We may be experiencing a specifically cultural or subjective process by which stories about one domain of social life stand in for other experiences that are too problematic, difficult, embarrassing or traumatic to be spoken of in any other way. Sexual themes seem often to be handled in this way, either as vehicles for especially condensed meanings, or as an emotional charge that electrifies, so to speak, the telling of a tale about something else. What is interesting about this kind of surprise is that it indexes experiences and conditions that are not adequately understood within our own dominant frame of reference.64 Certain-rather limited and mechanical forms of marxism or socialism, for instance, but also many kinds of social-structural sociology, have been notoriously incapable of handling the dimensions of subjectivity in general and of gender and sexuality in particular. These dimensions are often present even in accounts where formal closure has not occurred. In cases such as these, where explanatory theories remain open and flexible, a really dialectical relation of a socialist feminist kind may be established between first accounts-and a more explanatory secondary discussion of them.

In conclusion of this part of the essay it is worth reasserting two key themes. First, the process of working up first thoughts about our social existence and our history and our own ways of understanding the world is part of the central dynamic of 'making socialists'. This is not a process by which, whatever our class and gender position, we simply express or confirm our existing view of the world. This holds both for 'common sense' end for 'theory'. Since everyone possesses a social consciousness and is, in that sense, 'a philosopher', there is no necessary feature of this situation that privileges the 'intellectual' (in its usual narrow meaning). The problem is that intellectual skills and functions are subject to a social division in our society: there are processes of cultural dispossession and monopoly. The characteristic result is to rob theory of the elements of practical realism, especially in relation to the social conditions of the great majority of people, but also to withhold from common sense the skills and perspectives that might enable it to become more wide-ranging and critical. To fight to preserve the authenticity of common sense against both dominant discourses and radical intellectual alternatives is, in this situation, a kind of conservatism. The solution must lie in the establishment of connections between socialist and feminist theory and popular movements which work at and attempt to transform the intellectual divisions of labour in new, more democratic educational forms, forms of learning, that is, for all those involved. We will look later in this article at attempts to achieve such transformations in the broad areas of popular historical production.

The second point recapitulates the argument about empiricism, realism and an understanding of the cultural character of accounts. Commitment to a more culturally sensitive understanding, behind which lies the belief in the political significance of the cultural, is not incompatible with a realist concern for historical and structural conditions. On the contrary these two forms of understanding are complementary. A concern with symbolic and cultural forms is part of historical and contemporary analysis, not just a problem of historical 'bias' in source material. As Roland Barthes, a key figure for the development of the critical analysis of cultural forms, has put it:

Less terrorised by the spectre of formalism, historical criticism might have been less sterile; it would have understood that the specific study of forms does not in any way contradict the necessary principles of totality and History. On the contrary: the more a system is specifically defined in its forms, the more amenable it is to historical criticism. To parody a well-known saying, I shall say that a little formalism turns one away from History, but that a lot brings one back to it.65

Historical process and autobiographical form

Oral history and popular autobiography are forms which systematically individualize; yet an historically-informed political knowledge requires a much broader sense of social context. How can autobiographical materials be the source of, or contribute to, a more structural account of past and-present? Like the problem of the empiricist use of sources, the difficulty can be found in a representative form in Paul Thompson's work, where it appears, however, as a technical or methodological dilemma to do with the way of presenting oral-historical material.

Thompson suggests three different ways of presenting oral-historical accounts: 'the single life-story narrative', especially important for informants with 'a rich memory'; the collection of life histories, grouped, perhaps, around common themes, perhaps the better way of presenting most 'typical life-history material'; 'cross-analysis' in which oral material is used as a 'quarry from which to construct an argument'.66 The first two of these, which may be grouped together as effectively one mode, preserve most of the original material intact. The second mode is governed by the logic of the historian's own argument, and involves a briefer, less contextualized, more promiscuous use of quotation.

On the whole Thompson prefers the second mode of analysis since 'argument and cross-analysis are clearly essential for any systematic development of history'. The notion that oral history material 'speaks for itself' is 'blind empiricism,.67 Yet he is distinctly uneasy about this preference. He has qualms on two main grounds. First, he realizes that the use of materials as a 'quarry' and a form of analysis that gives primacy to the historian's own explanations is incompatible with part of his political aim: giving 'history back to the people in their own words'. Second, he doubts whether 'analysis' uses the oral material to the full advantage. There is something there that excites him - a quality frequently identified as 'richness' - which is often absent from other types of source.

One of the deepest lessons of oral history is the uniqueness, as well as representativeness, of every life-story. There are some so rare and vivid that they demand recording, whatever the plan.68

The oral evidence can be evaluated, counted, compared and cited along with the other material . . . But in some ways it is a different kind of experience. As you write, you are aware of the people with whom you talked; you hesitate to give meanings to their words which they would wish to reject.69

In The Edwardians Thompson uses a combination of the two methods of life history and cross-analysis. More correctly, perhaps, the two methods coexist, rather unhappily, in different parts of the same text. Part I of The Edwardians employs cross-analysis to 'establish the most important dimensions of social change in the early twentieth century'.70 In Part III Thompson analyses 'the main reasons for social change'.71 But in Part II he marshals 'twelve accounts of childhood in real families' 'to give a place, within this evaluation of general social change, to the contribution and experience of ordinary individuals'.72

Those parts of The Edwardians that are based on cross-analysis none the less use direct quotation a great deal. Passages are selected for quotation according to their contribution to the argument and this necessarily involves removing them from their original context and re-presenting them with a new meaning. It is worth taking one example of Thompson's characteristic method here: the announcement of some general theme (derived from his working through of oral and other sources) followed by a quotation from oral sources, in illustration.

Nevertheless there were many teenagers who frequently had to accept being disciplined by informal violence from adults: from the police, from employers, from schoolteachers, from parents, sometimes in bewildering succession. One Essex youth worked before starting school as the kitchen boy at a farm, cleaning boots among his tasks. The farmer's son would give him a cigarette if he told him when the farmer had his best boots cleaned, for 'he knew he'd gone to London, you see, on business, which enabled him to have the day off going on the spree'. The unfortunate kitchen boy one morning 'was smoking in the woodshed, chopping sticks, when the farmer caught me and picked up a twig and gave me - not a thrashing because I just bolted, but he caught me three times.... [I came home, crying, to father, who said] "What did he hit you for?" I said, "Because I was smoking" . . . so he gave me a clip with his hand . . . a backhander right across me cheek. And off I went to school crying. By that time, you see, it had gone ten . . . and that's where the headmaster was strict . . . so we got one on each hand, then, the cane on each hand'.

Farmers were perhaps the employers most likely to strike a boy whom they employed. 73

There are two main points to make about this passage. First, we are given no contextual information about 'the youth', nor do we discover what significance he attributed to his story, or why it remained in his memory. There is no commentary on the way in which the story was told: in cold print, it has a faintly humorous structure and tone, rather like the typical chapter of physical accidents in silent film comedy; but was such a pattern of punishment regarded as 'normal' in that period? Did the three punishments rank as equally legitimate? Attention to the specificity of this anecdote (and further questions about it) could elicit important indications of the interrelation between the three spheres of power - family, school and work - and the three kinds of authority - employer's, father's and school-teacher's - to which the young man was subjected. But Thompson offers no analysis either of the content of this quotation nor of its cultural forms, passing on immediately to his next generalization.

So, second, the quotation adds very little to the overall argument. As evidence it has already contributed to the 'factuality' of the preceding statement, in its form as an imprecise average ('many teenagers frequently had to accept'). It now functions only as an illustration. As such, we might ask, what is its value? Surely it has been included because it is an interesting story in itself: it injects 'life', 'richness' into the otherwise conventional account. In this way, argument and quotation are used to support each other, but at the expense of locking away the meanings of the quotation itself and the further structural relations that lie behind it.

What of the second form of the use of oral history material in The Edwardians? Here the life histories (or rather the succession of answers to the interviewer's questions)74 are presented in 'the untidy reality upon which, though too many scholastics would wish it forgotten, both theoretical sociology and historical myth rest'.75 They are presented not as integral to the analysis of social change or social structure but as a kind of unanalvsable extra. ThomDson sneaks of the life histories as indeed 'an antidote to the simplifications with which I have had to outline the dimensions of social structure'.76 Despite being produced under very definite conditions, the accounts are treated as a rather precious resource. In the end, there is nothing for it but simply to present them 'raw', to allow them to 'speak for themselves'.

These problems of method index many of the weaknesses of orthodox oral history, including some we have already discussed. That sense of richness betrays the failure to deal with cultural forms, with memory as an active restructuring process, with the forms of intersubjective transaction that produced the 'evidence' in the first place, and with the large question of the relation between the historically produced subjectivity of this particular 'witness', then and now. The richness remains elusive because it cannot become a specific object of knowledge within Thompson's own epistemological and theoretical framework. Similarly there is a kind of guilty pull between the desire to present popular memory directly (with its own unacknowledged interpretative framing) and the desire not to abandon the overarching interpretative responsibility of 'the historian'. What especially concerns us here, however, is the failure to secure a really organic and internal relation between large historical categories like 'social change' and 'social structure' and the individual stories and interviews. What is missing in Thompson's method, or more correctly in his theory, is some way of bridging the large-scale social processes which are the usual objects of historical accounts and the small-scale 'private' narratives which are the very stuff of personal memories.

The individual and social relations

We need to ask what it is about autobiographies or life histories that makes them not merely 'rich', but significant and 'representative' in some stronger and more definite sense. Part of the answer has already been given: once read in a non-empiricist way, these forms may make available elements of lived culture and of subjectivity not easily reached otherwise. But we need to go further than this along lines indicated by our insistence on 'structural' readings. Such accounts are repre- sentative and significant for a larger account not because they express a general abstract humanity, or a particular unique subjectivity, but because they are the product of social individuals. Their authors speak out of particular positions in the complex of social relations characteristic of particular societies at particular historical times. These accounts appropriate and make a sense of salient features of social relations within which their authors have been implicated and within which they have acted and struggled. Oral historians, like contemporary ethnographers, require a theory of these social relations - of structure in that sense - to see the representative elements of individual life histories as part of a more general history: a history, for example, of the formation or recomposition of a class, or of a break or reinforcement in the patriarchal relations between men and women. They require, moreover, some parallel understanding of the same or similar relations of force in contemporary society, the society in which a collective memory of these processes is, from particular social standpoints, to be formed. Without some such understanding, we can only judge of the 'representativeness' of accounts according to some passive, descriptive or quantitative criteria (so many working-class witnesses'; so many women) or fall back on unexamined quasi-literary criteria, similar to those used in the evaluation of 'great' literature.

This is one of the reasons why a meeting point between debates in marxist theory and the practice of oral history is so important. Marxism has always contained within it both a stress upon human individuality as a socially produced phenomenon and a recognition that our constitution within social relations is an active process of which we may become conscious and over which we may engage in struggles of a collective kind. Particular marxisms, it is true, have stressed one or other side of Marx's double vision, but marxist debate today is beginning to recover both sides of these insights. This is not the place for a long theoretical detour, but we would point to the debate between 'structuralist' currents in contemporary marxism (Althusser's stress for example on the determinacy of social relations) and the broadly 'humanist' stress on agency and struggle (Edward Thompson's Poverty of Theory or Victor Seidler's critique of Althusser in the collection One-Dimensional i farxism).77 More synthesizing enterprises are to be found in Perry Anderson's discussion of 'agency' in Arguments within English Marxism and, very importantly for our themes, Raymond Williams's successive treatments of the problem of society, the social individual and fictional and autobiographical ways of writing from his earliest work to Politics and Letters.78 In the contemporary debate about 'realism', the work of Bhaskar and Wal Suchting's translation and reading of Marx's Theses on Feuerbach are especially important.79 But much of what we take from these debates is most fully expressed in Gramsci's attempt to redefine the old philosophical question 'What is Man?' in the light of a marxist understanding of social relations.

In his notes on the study of philosophy,80 Gramsci argues against the tendency of Catholic social theory to locate the sources of evil in 'the individual man himself'. This is to treat human beings as essentially 'spiritual'' es uniquely individual, and also as already 'defined and limited', not capable, that is, of further development. As Gramsci says, such a view of human beings is not limited to a religious framework: it has been characteristic, perhaps, of 'all hitherto existing philosophies'. It certainly remains an important mainstream philosophical legacy fifty years after Gramsci was writing. It remains common, for example, in both historical and literary studies: in the portrayal of leading national politicians or great writers as uniquely, individually, creative persons. There is more than a hint of this perspective in the notion of individual oral witness as irreducible and 'unique'.

It is on this point, Gramsci argues, that 'it is necessary to reform the concept of man'. Consciousness of one's individuality is important - it is to become very important in Gramsci's own argument - but it is not the only element to be taken into account. We have also to consider the relations between human beings and 'the natural world', and their relations to each other. But the relationship between human beings and nature is not a simple one: it is above all 'active'. Human beings don't just 'dwell' in the natural world - they are related to it actively by work and technique. Similarly, our relations to each other are not those of adjacent individual persons but involve membership of socially-organized groups, of a more or less complex kind. One might conceive of 'Man' then as 'a series of active relationships'.

Relations of this kind, however, must not be seen as mechanical or deterministic: they involve both human subjectivity and the possibility of change:

Further: these relations are not mechanical. They are active and conscious. They correspond to the greater or lesser degree of understanding that each man has of them. So one could say that each of us changes himself, modifies himself to the extent that he changes and modifies the complex relations of which he is the hub.

Of course, social relations are of different kinds: they have varying degrees of determinacy and are more or less easily changed. Yet to be conscious of them, in the sense of knowing how they might be modified, is already to modify them. This is even the case with relations that are found to be 'necessary' (for example the necessity to earn a living from nature) which take on a new light when viewed in this way. All this suggests, indeed, a change in the way we define 'philosophy'. Knowledge really is power:

In this sense the real philosopher is, and cannot be other than, the politician, the active man who modifies the environment, understanding by environment the ensemble of relations which each of us enters to take part in.

Similarly, the old philosophical project of 'know yourself' takes on a new light, both as a project for all (everyone as a 'philosopher') and as a more than contemplative practice:

If one's individuality is the ensemble of these relations, to create one's personality means to modify the ensemble of these relations.

But this form of understanding, which remains static, is not enough. We have to understand 'Mans not just as a social and natural being, but as an historical one too:

It is not enough to know the ensemble of relations as they exist at any given time as a given system. They must be known genetically, in the movement of their formation. For each individual is the synthesis not only of existing relations, but of the history of these relations. He is a precis of all the Past.

It is because it allows us to grasp 'history' in this sense that the approach to 'Man' through social relations is greatly to be preferred:

That 'human nature' is the 'complex of social relations' is the most satisfactory answer, because it includes the idea of becoming (man 'becomes', he changes continuously with the changing of social relations) and because it denies 'man in general'.

Indeed, if it is useful to talk about 'man in general' at all, we must refer the inquiry to history - ' "human nature" cannot be located in any particular man but in the entire history of the human species'.81

Gramsci's discussion of 'What is Man?' recovers and restates arguments found in the origins of marxism, especially in the tussles between Feuerbachian humanism and the emerging perspectives of historical materialism. It recapitulates and develops, especially, The Theses on Feuerbach. It could be criticized on several grounds: the failure to develop a thesis about the 'natural' character of human beings (the way they do not only work on nature but are also a part of it) ;82 the incompleteness of a theory of individual subjectivity. But the most important criticism, for today, is Gramsci's neglect of social relations other than those of class, and in particular his neglect of the historical construction of gender relations, a neglect indexed by his acceptance of the philosophical figure of 'Man' to indicate the whole human race.

In fact, modern feminism has made peculiarly important contributions here, filling out, developing, and critiquing existing marxist conceptions. The contribution has been especially important in the context of our own argument because it has centrally concerned the relation between individual experience and social relations.

Feminist theory has worked from the practical implications of women's politics, especially from the forms of 'tine consciousness-raising' group.83 The experiences of individual women, often 'publicly' spoken about for the first time, and in the form of individual witness, were seen as evidence for a general social condition. Women's sense of oppression was not the product of individual problems or personal inadequacies: it sprang from systematic causes, a general social condition shared by women as a whole. This general social condition, even where given the same name - 'patriarchy' - was subsequently defined in many different ways, sometimes as existing, rather weakly, in archaic or residual attitudes. In stronger versions of the theory of patriarchy, greatly to be preferred in our view, the stress is upon patriarchal social relations, a particular organization of relationships between men and women in the process of which gender differences are also produced and defined. Relations of gender, in their historically specific forms, have resembled relations of class in that they have conferred radically unequal chances for the development and satisfaction of human needs. They have, therefore, necessarily involved struggle, wherever women have become conscious of the forms of male control and the unequal exchanges involved in most gender-based transactions - from romantic love and the marriage contract to the production and bringing up of fresh human beings.

Feminism insists, then, that 'tine ensemble of social relations' tee thought of as still more complex. Moreover, because of the personal character of women's oppressions in the forms in which they are first experienced, many of the major feminist texts have been quite unlike the main productions of marxist theory. They have taken the form of stories and novels and autobiographical writing, forms of writing that have also managed to capture, however, a sense of representative experience, true for women in a particular historical phase.

Vera Brittain's Testament of Youth seems to us to be a superb example of this kind.84 It centres upon the experience of Vera herself, yet it tells us more than most works of 'history' about the experiences of the First World War more generally, and even about the determinations upon events. If such an assessment were agreed, it would be worth considering what it is that makes this book so successful.

Our own answers relate back to the more general arguments about individuals and social relations. Part of the strength of Testament of Youth lies in its author's location within key sets of social relations, especially, in this case, relations of gender. This is 'lived' in her own personal relationships, especially with her brother, her lover and their mutual male friends in the years before the War. The War impacts on these relationships in a way that is not merely representative of the position of women at this time, but has a kind of super-representative character. Vera's own life is massively, crushingly over-determined by the European holocaust; she loses a much-loved brother, the man whom she hopes to marry, and her other male friends. Her own active response to the threat and reality of this is to become involved in the War herself, as directly as she possibly can. So she becomes a nurse in the battlefields of Europe, and further experiences the contrast between the idealistic hopes of young men going to the War and the physical and mental effects of their mutilation. This in turn removes her both from the certainties of her own middle-class family and from the naively of most of her later fellow-students at Oxford. In a very important sense, Vera personifies the woman's experience of the First World War, something that she herself began to realize as she wrote Testament of Youth over twenty years after the War had ended. She wanted to show what the whole period meant to 'the men and women of my generation'. She wanted 'to write history in terms of personal life'. After various experimentations with form, including a novel and the reproduction of her war diary, she opted for autobiography:

There was only one possible course left - to tell my own fairly typical story as truthfully as I could against the larger background, and take the risk of offending all those who believe that a personal story should be kept private, however great its public significance and however wide its general application. In no other fashion, it seemed, could I carry out my endeavour to put the life of an ordinary individual into its niche in contemporary history, and thus illustrate the influence of world- wide events and movements upon the personal destinies of men and women.85

We cannot explain the force of Testament of Youth, however, simply in terms of 'structure' and a strategic personal location in social relations in their particular historical form. Vera also developed a particularly heightened consciousness of herself, her intimate relationships and of a wider catastrophic context. Moreover this was a developed and critical consciousness, unrelentingly suspicious of easy explanations, comforting resolutions and surface appearances. It expressed itself typically in a kind of measured irony, turned especially against patriotic and religious dogmas, and, in effect, the whole tissue of pre-war nationalistic pieties. It was also a political consciousness: feminist and pacifist in its active orientations. In many ways, then, Testament of Youth is one powerful model for the kind of practice we have been discussing in this essay - a socialist and feminist popular memory.

The mature properties of 'emotion remembered in tranquillity' have not been my object, which, at least in part, is to challenge that too easy, too comfortable relapse into forgetfulness which is responsible for history's most grievous repetitions.86

It is important to add that our own advocacy has nothing in common with some other arguments in favour of autobiographies of 'representative individuals'. It has nothing in common, for example, with elitist versions of this argument that ascribe an importance in principle to a biography when it concerns a personage of great formal power or influence. An interesting case to set beside Vera Brittain's account of the First World War is Sir Winston Churchill's extended autobiographical account of the Second World War.87 Here, a sense of personal significance and 'historic role' is asserted in almost megalomaniac proportions, but the accounts that follow, though certainly personally revealing, are rarely self-consciously so, and remain locked up within an upper-class military and high-political culture and a highly mythical Conservative version of national character and history." 'Representativeness' in our sense, is more likely to be found in popular autobiographical forms where dominant social relations are viewed from the typical subject position: that of daily oppressions and of the struggle against them. Representativeness, moreover, is a feature of social positions that are understood to be shared and collective: the main feature of much autobiographical writing is to distinguish the author from the people and the determinations that surround him. Such accounts belong not to the construction of 'popular memory', but to the reproduction and dissemination of 'dominant memory', of which the Churchillian myth is indeed a salient and persistent modern feature.

Our argument does have implications, however, for a practice of popular memory within which oral history or popular autobiography will have a central place. There are, perhaps, two main implications: that the choice of subject (individual, or community, or historical period) will matter very much; and that our method must give maximum opportunities for self-conscious political reflection upon first accounts.

It does matter what subjects are chosen or what subjects choose themselves. And And once they have emerged, such subjects have to be understood critically in relation to their representativeness. Some subjects will have a massive general salience; all remain important 'in themselves' but may speak less directly to widespread social experiences. They may be severely limited, indeed, by the social milieu from which they arise. Brittain's Testament, for example, is, in important ways, bounded by the middle-class character of her own social position, and her subsequent profession as writer and agitator. But accounts may be community-bounded too, locked up in the specificities of a particular area, where, for example, class relations take a particular, perhaps more archaic, form. In the same way, certain communities may have a peculiar significance, representing new social forms, new transformations, the sudden destruction of a whole way of life and the emergence of a new. It is important not to press this argument too far, too mechanically. In particular we must not assume that we have a perfect knowledge of what is typical before the accounts themselves emerge. But relevant issues are raised: why, for instance, are most 'community-based' studies focused on areas with a traditional industrial infrastructure - mining valleys or villages, or older working-class urban neighbourhoods? Why not the newer housing estates, or the inter-war or post-war settlements, many like the steel-town Corby, now suffering equally dramatic transformations? Again, the assumption that history anyway is about the past encourages complicity with a certain conservative nostalgia. The ideal cases may well be those not with the greatest or longest continuities but those with the most dramatic contemporary transformations.

The implications for method follow from the recognition that social location is no guarantee of politically useful knowledge. If our arguments about 'consciousness' are exact, the method needs to maximize opportunities for second thoughts, for further analysis of primary results and first impressions, for retheorizing and 'making strange' familiar appearances. Some of us rebel against this, because the ugly figure of the 'historian' (or 'sociologist') once more intrudes, telling us what our explanations should be, fitting our 'facts' to his theories, presenting our experience back to us, sometimes in unrecognizable forms. But secondary analysis need not take this form, or be constrained within the existing social divisions of intellectual labour. It could be a more internal process: a working up of first accounts by authors themselves, in the light of further research and thinking. Again we need to detach the general question (the necessity of study for overcoming ideological ways of thinking) from the particular social and educational forms in which this problem currently appears. Our point is that it is not enough that the production of first accounts be respected in the sense of being left untouched. Really to 'respect' them is to take them as the basis for larger understandings, for the progressive deepening of knowledge and for active political involvement.

Presents and pasts

In this section we want to draw out some implications of the view that the proper object of history is not the past but the past-present relationship. So obvious a point would not be worth stressing had not historians striven so strenuously to deny it. Our argument, here, parallels the case against empiricism: just as history-writing is necessarily a theoretical and political activity, so it is also a practice in and for the present. Theory, politics and contemporaneity are basic conditions of the practice, present even when denied. To treat organizing assumptions and political implications as 'bias' or as 'ideology'(in the everyday sense) is to expect professional integrity or technical methodology to transform the very conditions of knowing.

Again oral history is an especially useful ground on which to think through these problems. In accounts of oral-historical methodology, the past-present relation appears mainly as a problem - the unreliability of memory. There is a model of memory implicitly at work here and it is a particularly passive one. Memory is the sedimented form of past events, leaving traces that may be unearthed by appropriate questioning. It is a completed process, representative of the past which is itself dead and gone and therefore stable and objective. Once laid down in this way, memories may certainly cease to be available, but this, too, is a mainly technical problem. In The Voice of the Past, for example, Paul Thompson has recourse to the laboratory experiments of, social psychologists and to knowledge about the biochemistry of the brain in old age. His strategy, again a conservative one, is to cite psychologists' findings as evidence for the relative reliability and constancy of memory over longer periods of time.89

The most conspicuous absence in such accounts is the present conjuncture in which oral witness is actually recorded. The present is absent both as a source of determinations on and meanings of the stories that are told and as the location of current responsibilities and needs. Thompson does have a short discussion of the problem of 'retrospective bias', the perils of which he tends, however, to minimize.90 Yet there remains a point to Koss's objection that the 'Edwardians' are 'Elizabethans' too. Their stories are necessarily influenced by present events and by the restructuring of what it is possible to think and say. Oral history testimonies do not form a simple record, more or less accurate, of past events; they are complex cultural products. They involve interrelations, whose nature is not at all understood, between private memories and public representations, between past experiences and present situations. Figure 3, though not intended as a developed alternative to the notion of memory as record, indicates something of this complexity.

Imagine yourself my oral history witness. The tape recorder has been turned off and since you are co-operative to a fault (and because I am writing this and can make you do anything I want) we start to consider the fuller history that lies behind what you have been saying. We are seated in the right-hand room on the first floor of Figure 3 (which has, you see, particularly strong walls so that you couldn't get out anyway). I insist you start by considering your present social position and experiences, which I am sure have influenced your story. You are, besides, an avid reader of the Daily Mail and a supporter of Mrs Thatcher; you have been sure, for some time, that the country has suffered a moral decline. I insist we take that into account too. 'But', you say, 'all that can't change the reality of what happened to me then.' 'That's true', I reply, 'but it will frame, shape, evoke or even obliterate your memory of it. Anyway, your reality was a complicated thing in itself. It too was part material relation, part cultural form. And since then you have had many experiences, many dramatic changes of heart, even engaged in guilty self-criticism. Some things you would rather forget and have in fact forgotten. Other things now acquire an importance which they didn't have before. (I noted how your moment of adolescent socialism, for example, was rather skated over.)' 'I see what you mean', you say, apparently capitulating, 'but haven't you forgotten the most important thing of all? It was your questions to which I was replying; my answers are also produced by your power.'

Oral history testimony, in other words, is profoundly influenced by discourses and experiences in the present. That is the standpoint from which oral accounts (and formal histories) are constructed. Memory is therefore itself a profoundly complicated construction and a very active process. In memory past events, in their own complexity, are worked and reworked. Of course, there are also continuities and people do relive certain past events imaginarily, often with a peculiar vividness. This may be especially the case for those (e.g. the elderly) who have been forced into a marginal position in the economic, cultural and social life of a society, and, fearful of absolute oblivion, have little to lose but their memories. But, in addition, as a further set of mediations, the intervention of the historian or sociologist is of crucial significance, as a catalyst for a whole process of structured remembrance.

There is, besides, something, absolutely problematic, politically, in treating oral history witnesses as 'sources'. It is to treat them as a form of walking, talking documentation, the person as archive. Is there a flat contradiction here between oral history as professional research and oral history as politics? We think there is, insisting, however, that we should not equate 'professional' with 'intellectual', still less with explanatory or analytic as such. But the approach of a professional historian to a person-source does involve an entirely different relationship than that which ought to characterize political-intellectual encounters. The historian's attitude must needs be preservative, rather like a museum curator's. The impulse is 'to get this on tape'. The attitude of mind is non-interventionist (unavailingly however). The rich mental material must not be disturbed too much: the point is to record it. Political engagements, by contrast, where they are not self-seeking, narrowly partisan or actually manipulative, must needs involve an active mutual incitement to rethink experiences and understandings, to struggle to see the world differently, to go beyond existing social relations, to see that, under other arrangements, personal problems might become collective solutions. Such an active 'educative' process will necessarily affect memory, even in its private forms. Older constructions may be preserved, but they will be overlaid by newer thinking. Sometimes we will wish to disavow whole passages in our own past both in terms of how we have acted and how we have understood our actions. The past is understood, in this practical framework, not as a given 'thin"' which we must preserve, but as a force constantly resonating in the present, producing new layers of sound and meaning.

There are two practical implications of this argument, the one broad, the other more specific. Our argument about past and present points to the unity of history and politics, to historical work as an aspect of politics. If this seems paradoxical, even 'dangerous', it is often because we conceive of 'politics' too narrowly, as dealing only with 'great public issues' in ways familiarized in existing political routines. God forbid that history should be linked to 'politics' only in that sense! Yet in a larger political project - that of debating and struggling over the whole social future - 'memory' is a vital resource, helping us to understand the nature, historicity and changeability of current conditions.

More specifically, socialists and feminists need to develop a politics of later life. It is clear that the contribution of older people to a socialist and feminist popular memory involves such a political strategy. Our existing understandings are insignificant here, not least because 'age' seems yet another kind of relation to be compared with class, gender or race. But such a politics should certainly start from a proper analysis of the experience of older people today - the specific forms of oppression and of cultural marginality - and from the assumption, very different from of ficial or conventional wisdoms, that their memories, sometimes actively rethought, may contribute to a politics of today in which they are, themselves, active participants.

Some forms of popular memory

Ideally, at this point in the argument, it would be useful to take some examples of popular memories and try to see the various processes we have tried to describe actually at work there. Aside from resenations about doing this outside a more situated political project, it would really require a larger separate study. We do, however, want to look more briefly at some of the typical forms in which memories are expressed in published oral-historical accounts and in popular autobiography. We will be trying to generalize over a wide range of material here, at the risk of large simplifications.

We have already noted some of the recurrent features of these accounts: the different forms of artifice that have gone to make them, from literary to verbal skills; the omnipresence of the story-telling form; the necessary constitutive influence of the past-present relation as the basic retrospective condition. In addition it seems important to stress the intensely 'personal' character of most of the accounts we have read and their distance, therefore, from the common forms of public historiography, most academic history, and especially from histories of a 'structural' or non-narrative kind.

By 'personal' in this context we mean two rather different things. First, and most generally, we refer to the 'humanism' of the texts, using this term quite descriptively, and neither pejoratively nor in a simple, morally-affirmative way. What marxism or an explanatory social history will wish to treat as social relations or as social classes, these accounts tend to treat as persons. If this is a fetishism (and the term may not be appropriate) it is a fetishism of people and not of things. These worlds are not only 'dense', 'concrete' and specific but also very heavily 'peopled'. Social relations are understood through the qualities of persons who inhabit them. Structural determinations appear as relationships between people, This 'humanism' runs through the commonest form of story-telling, in which real or imaginary events are re-cre ted in the form of conversations between particular people: 'My father said to him .... He said to my father ....', a form, incidentally, which crosses national and class cultures and is certainly not limited to working. class culture in Britain. It informs, more generally, the 'theatre' of relived memory with its dramatic personae of 'strong characters', vivid feelings and minutely described episodes. One striking instance is the centrality accorded to persons in retrospective accounts of work experience, as in the collection Working Lives. These tend to centre on the personalities of employers (slave-driving, posh and distant, or 'reasonable') and on the qualities of forepersons or workmates. The character of persons is seen as making an immense difference to working experience, the difference between a good job and a bad. As Betty Ferry puts it in Working Lives:

This man was a tonic to anyone that was feeling a bit down in the dumps. He was always the same. If more men like him were employed in factories what a pleasure it would be to go to work! He gave a lot of pleasure to us all. People that worked with him could never forget him.91

Similarly, waged employment, especially the very first job, is almost always remembered, in part, through the personality of the employer or the employer's agent, often through a short, pungent pen-portrait. In Working Lives these range from the Pickwickian:

So again the next day we found ourselves in the office of the head of the firm. Mr Thomas Muddiman was an important but kindly and pleasant looking little man, dressed in the conventional morning dress of that period with a gold watch chain and gold rimmed glasses, topped by a shining bald head.92

To the villainous:

Then there was Charlie Griffiths, which is now a big firm. In those days there was just Charlie Griffiths and his two sons, Bert and Sid. I could have spat in Sid's eye more than once. He was a slave driver.93

The ubiquity of such portrayals in the Hackney accounts may have much to do with the small scale of industry in the area in the period of many of the memories of volume 1, but the stress on character, personality and the capacities and abilities of individuals does seem to be a key feature of the whole genre, and indeed, of the lived culture from which it arises.

The second feature of the personal character of the texts is the way in which the past-present relation is handled through the intimacies of personal relationships and especially through the comparisons of generations. We might take Dutiful Daughters as an example here, because it is especially rich in 'historical' references and because it may give particular insights into a specifically feminine view of the past. On the whole there is very little that is recognizable in these accounts as a 'public history' though we certainly glimpse the personal implications of larger changes. If we worked with a very conventional view of 'the historical' it is doubtful if we would recognize these accounts as 'historical' et all. Only occasionally do 'great' public events intrude on a more personal story. The major exception here is the Second World War, almost always experienced as a major reorganization of everyday life. Even the War, however, most often serves as a sort of historical marker for more intimate events, a reminder of what 'we' were doing when this or that happened. We have a strong sense that the history of monarchy, undoubtedly a popular history in many of its forms, acts more generally in this way, as the bearer of a more general chronology, in relation to which private stories may be placed. The following passage from Peggy Wood's account in Dutiful Daughters condenses much of what we are saying here:

I was telling the children the other night, and the telly was on, they had this thing about 1945 and when the War ended, and David [her husband] was sneering at all the crowds coming round the palace and waving to the King and Queen and little princesses, and I said, 'Don't you sneer, one of those people's me.' Jonathan ther son] looked at me in amazement, and I said, 'Yes, what else could you co?' The pubs were dry, and we just wandered round the streets, and the lights were still out. It took weeks for them to be all fixed up again. But the black-outs were torn down so the lights shone into the streets for the first time. Terrific.94

A lot of the basic conditions and forms of popular memory come out of this short extract: the centrality of generation ('telling the children') as a fundamental impulse to remember; the precipitating role of the public media entering right into domestic relations; the struggle to assert the validity of memories against adult scepticism and youthful incredulity; the use of royal occasions (in an unroyalist family) as a marker for 'experience'; the struggle to connect public events and private experiences which are actually obverse sides of the same processes ('one of those people's me'); and, again, the story form, the narrative of personal experience. In the same way, in other parts of Dutiful Daughters, a sense of past and present is constructed through accounts of personal life transitions - childhood, puberty, school, courtship, marriage, children - or through generational comparisons. One common form of this, related, perhaps, to the feminist context of the project, is the gauging of change in the lot of women through comparisons between 'my' childhood and the childhood wished for my own children. Often, in the most painful parts of accounts, the comparison rests on a really strong hostility to mothers and a determination that such relationships will never be repeated again: 'And I often used to say how I was treated I would never treat mine'.95 But such comparisons may be the basis for much larger historical constructions which embrace an understanding of the intensity of the problems faced by an older generation of women, and some appraisal of real, if limited improvements especially in matters to do with sexuality.96 This very short account does not begin to exhaust the forms and conditions of popular memory. It would be interesting, for example, to examine the capacity of things, especially old things with personal associations, to embody conceptions and emotions about the past.97 It would be interesting to compare this with the role of historical things in a culture of the museum. But we have said enough, perhaps, to pose some further questions. How do the dominant historical constructions acquire a purchase within these more lived senses of the past? How and to what extent are the forms of popular memory rendered ideological? How are they made to conform, for example, to a thoroughgoing political Conservatism?

These questions are complex and can scarcely be resolved without a close account of particular historical ideologies. We can approach some of this complexity, however, if we return to the vexed question of 'humanism'.

If we recall the full weight of Gramsci's 'philosophy' of social relations and our own discussion of the social individual, it is clear that not all humanism can be disposed of as ideological or as socially conservative. Social relations are produced, reproduced or changed in and through the activity of concrete social individuals. To think otherwise implies that social relations lie behind or outside practical activity, 'determining' it externally. To centre on the roles and characters-of individuals is not in itself to hide or disguise social relations or larger structured causes. These must be seen to work precisely through human action and subjectivity. Moreover the notion of humanism as ideology disguises the importance of the attempted humanization of social relationships as a way of resisting their force or rendering them meaningful. You appropriate my labour but I insist on treating your person, shiny pate and all, as an object of amusement or fun. In some of the accounts in WorkingLives this theme of humanization through personal humour is made completely explicit and is, as Paul Willis has argued, a quite persistent and general feature of shop-floor culture. As Betty Ferry puts it, talking about a particularly scrupulous manager:

Mr Howarth was a very stern looking man, rather tall and very businesslike. He came from Liverpool and sounded just like my favourite comedian, Al Read .... I could never take Mr Howarth seriously when he got annoyed or angry with anyone, because he always sounded so funny. He was most entertaining for me. Sometimes I had to run to the toilet out of the way in case he caught me laughing, especially if he was telling someone off, which he did quite often.98

It is only after we have grasped these features - the 'realism' of a humanism and the role of humanization in resistance - that we must, in the end, return to the question of costs and to the elements of truth in the Althusserian critique of a 'theoretical humanism', humanism, that is, in a systematically explanatory form. For what if the personal capacities of individuals (as unique representatives of 'humanity') becomes our main explanatory principle, such that we blame ourselves and others, all the time, for the things that go wrong in our lives? Such frameworks, which undoubtedly have a penasive common-sense existence, are easily assimilated into a thoroughgoing consenative individualism with its characteristic heroes and heroines ('those who work hard') and its characteristic scapegoats and villains. The objective tendency of such ways of thinking is to hold us within existing social arrangements and to increase the depth of inequalities. In these particular respects, then, a thoroughgoing humanism may inhibit deeper and more activist under- standings of everyday life such as produce, for instance, collective struggles against difficulties. But there is no warrant here for thinking of humanist figures of expression and understanding as ideological through and through, in all their manifestations.

We suggest, still more tentatively, that historical ideologies work most power- fully in relation to general conceptions of social improvement or decline. This is a crucial point of intersection between popular memory and public political discourse: it is an aspect of the construction of a national-popular will. It also unites conceptions of the past, analyses of the present crisis, and hopes and possibilities for the future. It is part of everyday historical consciousness: as people reflect upon their own experience, they make evaluations of this kind, by no means limited to a simple nostalgia or a one-sided progressivism. In Britain today, however, such evaluations are almost always framed by a sense of national decline, connected to the question 'what's wrong with this country?'. As Stuart Hall and others have argued, in studies especially of education and 'law and order', it is the New Right that has presented the most powerful and wide-ranging answers to this question, making an historical sense of the experience of crisis,99 connecting with a larger Conservative historiography in which 'we' are again an embattled people, prepared for (very unequal) sacrifices.

There is, so far, little to set against this. The main themes of a liberal and social-democratic historiography - of real popular progress through 'economic growth' end 'social reform' - are revealed as idealist constructions, rebuked by the realist elements in common sense. In particular the notion that late capitalist development is necessarily associated with the progressive extension of welfare provisions of all kinds now looks entirely implausible. On the other hand, the historical rhetoric of the Left seems to be dominated by the negative, static backward-looking image of the 1 930s, a form of popular history with no orientation towards a future different from the Thatcherite or social-democratic utopias. Among men and women who can recall a more synthetic socialism in the past, there is a widespread sense of loss that recognizes how politics have been narrowed and professionalized since the 1940s, that idealism has been lost in an everyday pragmatism, and that there has been a profound liberalization of the Labour Party. As Annie Davidson puts it in Dutiful Daughters, speaking from her own socialist formation in the Glasgow of the 1920s:

What I call myself is an 'idealist socialist', and it's laughed at nowadays because the world has become so complicated. Idealism is not very popular nowadays because there are so many practical things that have to be decided where your ideals have to be pushed aside to some extent in order to get a result now. But I feel that a long-term result must be the real goal of socialism, and in the end the best rather than trying to do something as a short-term measure. If the short-term measure becomes permanent then it's no use, because the short-term measure is the wrong measure to me. That's not very well-expressed but that's what I have a feeling about. 100

Perhaps what is elusive here is the need for a historically-informed view of a socialist future. Annie's sense of loss points to the absence, in contemporary socialism, of a popular historiography that links past, present and future. It is only within such a framework that the long-term and short-term issues can be properly worked out. It is interesting that the socialists (and especially the Communists) of her youth were often possessed of such a vision, informed also by a marxist historical science. It is interesting to find, for example, that Annie herself read the historical novels of Eugene Sue, getting from them a comprehensive sense of proletarian history linked to a 'family saga'.101 Bill Schwarz and Raphael Samuel have noted how a long-durational marxist history, back often to 'primitive communism', allowed socialists and marxists to place their own struggles with a much larger (indeed often grandiose) historical framework.102 The problem, of course, is not to simply recover this older marxist historical anthropology. A popular memory that is socialist, feminist and anti-racist will have to start from what already exists in relation to some popular constituencies, particularly in feminist and black history. It will have to rethink a history of the white, male working class too. But the need for an active, popular and politicized sense of the past has never been clearer.

On the social relations of research

We have stressed throughout that epistemological and theoretical problems rest, in the end, upon certain social conditions, especially the position of authors of different kinds in the social relations of intellectual production. We have been especially concerned with the role of 'the historian' as a professional monopolist, or would-be monopolist of historical knowledge production. How can these forms of monopoly (and the professional ideologies which they tend to reproduce) be broken down?

There is already a range of qualitatively different practices in this area. They seem to us to conform, more or less, to three main types of organization. There are, first, the large, heavily funded 'projects' in oral history, sometimes, as in the USA, with of ficial backing too.103 The key features of such projects tend to accent social divisions and, indeed, multiply them: not only divisions between the subjects and objects of research, but also among the researchers themselves - between those who write, think and theoretically frame projects and those who mainly administer them. These divisions have reached such a stage in the USA that it is now possible to design a research project and then hire a commercial fimm to conduct the fieldwork. It seems to us, even where conducted under politically radical auspices and with the help of marxist or feminist theorizations, such projects are problematic in their very forms, though this may not be sufficient reason for abandoning them. They certainly, in their actual working as practices, tend to reinforce existing relations of knowledge-and power. They may exercise a kind of symbolic violence by which people's memories are taken from them and used to further an academic or of ficial career with no proportionate return to the originating constituency.

In Britain, however, such large-scale projects have been unusual. Much more commonly oral history and similar practices have been linked to the contradictory social position of the researchers themselves. The most obvious instance is the salience of oral history end 'people's history' in adult working-class education, as in the projects based on Ruskin College.l04 It is in adult education (especially in the residential colleges) that the contradictions between working-class origins and loyalties and the academic and middle-class character of knowledge and of knowledge-production are experienced most sharply. One common form of this is the pull between 'going back to the community' and 'going to university', the main form in which an intensive continuation of intellectual activities presents itself. Whether or not the academic choice is eventually made, projects in 'people's history' may serve as a way of affirming a continuing class or local loyalty. At the same time the actual relation to a community of origin is being subtly changed by the acquisition of new languages, knowledges and, sometimes, habits of life. We suggest, but cannot demonstrate, that the main impetus for people's and oral history derives from this or similar situations.105 In this respect social-historical studies conform to a broader social pattern, common in sociology and in cultural studies too: the importance of the experience of class mobility (especially through formal education) in precipitating political questions and intellectual work around class-cultural differences and relations.l06

Despite the intimacy of the subject and object of research, in these cases, the division of labour remains. Academic careers can be made out of personal memories or communities of origin too! As in the previous case, this does not necessarily negate the intellectual results of the research, but it does raise the further question: for whom, then, to guide whose practice, is this knowledge intended? Much depends, here, upon the extent to which such knowledge returns to its originating constituency, and no doubt this sometimes happens. More commonly the reading constituencies are similar to those for this book: not the popular constituencies themselves but those particular social groupings to which the author now belongs, the more academic or intellectual elements in the professional middle class, especially the more radicalized sections. We might add that the conditions of life and, especially, the terms of access to this group, are currently undergoing a major restructuring.107

It is in this context that some individuals and groups within the radical intellectual constituency have sought to develop the forms of community-based writing and publication which we have already discussed. One aim here has been to challenge, even to recompose, the social relations of intellectual production, distribution, and readership. Working-class authors, as individuals or collectives, have been encouraged to develop popular ways of writing history, autobiography, or fiction, for a primarily working-class audience and often a local distribution. The community-based movements very accurately identify the social divisions of intellectual production as a crucial area of struggle. We cannot conceive of an adequate politics of popular memory that did not involve, as one moment, the production and circulation of first accounts with a direct popular authorship. Even so, such projects are surrounded by difficulties, the product of the unreconstructed context in which they necessarily operate. We want to end by stressing two of these, the latter to be further developed in the last essay in this book.

One major difficulty is that divisions of labour and relations of power, challenged and modified in one place, promptly reappear in another. We democratize authorship so that the oral history 'witness' now writes as popular autobiographer. We may organize such productions collectively, as at Centreprise. But social divisions, abolished in the shape of the historian and 'the source', may reappear as, for example, differences of skill and cultural power within an organizing group. In educational or quasi-educational contexts, in the WEA or extramural class, they may be reworked as a problem of pedagogy, as a teacher-taught relation. As anyone with experience of 'collective work' knows, small-scale organization does not abolish powers and inequalities of initiative, direction or definition (though it may make their negative sides easier to identify and struggle over). In all this, however, there is a danger that social divisions (now removed 'elsewhere') are resolved by an imaginary identification: middle-class initiator with working-class group, teacher with class, activist with 'community'. In such cases the danger is that the exercise of cultural power (which remains) becomes quite unconscious or, in its effects, merely recycles what is known in the community already.

Such contradictions aren't going to be solved by arguments. But we think it is productive to restate the problem less in terms of social identity than in terms of standpoint and connection. Organic intellectual work is not, in other words, only or primarily a question of who you are; it is more a question of whose problems you set out to study and resolve. The decisive question is the standpoint of projects and the connections they require. Social identity may help or hinder here, though not always in predictable ways. More important is the source of the agenda in the everyday life of subordinated classes and social groups and the production of knowledge which is useful from that standpoint. This applies as much to 'critical' academies as to political activists. It is the connection between popular experiences and a developed intellectual function that is the essential thing. The key problems then become how to organize such a connection politically, how to generalize skills of secondary analysis and how to connect this popular education to other daily struggles.

The second main difficulty concerns the relation between popular historical writing and the field of public representations of history. From the standpoint of the community activist, who may value authentic popular expression above all else, the pervasiveness of a dominant memory may not be visible. Yet the dominant historical constructions will necessarily be a presence in the first sketches of a people's history. For the ruling ideas are not just those of the dominant social group - those they live their own lives by, those they seek to generalize. They are ideas also with particular tendencies, 'bourgeois' not necessarily in origin, but in their conformity to dominant social relationships. Since we all live in a capitalist, patriarchal and racist social order, such ideas may be spoken by anyone. There are limits, then, to the pursuit of authenticity, of really popular expression. Struggle at this level - the production and working up of alternatives - has to take account of the dominant discourses and their means of production. It is to these public histories and their general forms that Chapter 7 is addressed.

Chapter 6 Popular memory: theory, politics, method

1 Compare the editorials in History Workshop Journal, no. 1 (1976) and Social History, vol. 1, no. 1 (1976).

2 Ken Worpole, 'A ghostly pavement: the political implications of local working-class history', in Raphael Samuel (ed.), People's History and Socialist Theory (Routledge and Kegan Paul, 1981), p. 23.

3 Jean Chesneaux, Pasts and Futures or What is History For? (Themes and Hudson, 1978),especially pp. 1 and 11.

4 Quoted p. 95.

5 We draw here on the work of Rita Pakleppa and Hans Poser who were members of the Popular Memory Group in 1979-80. Their presentations on representations of the Second World War in Britain and West Germany enlivened and informed the work of the group. We hope their study will eventually be available in English.

6 E. P. Thompson, Writing by Candlelight (Merlin, 1980), pp. 130-1.

7 One provocative exception was B. Hindess and P. Q. Hirst, Pre-Capitalist Modes of Production (Routledge and Kegan Paul, 1975).

8 Cf. p: . 94-5.

9 For example, Michael Foot's speech to the Liverpool mass demonstration against unemployment, subsequently used in a Labour Party Political Broad- cast on 5 December 1980.

10 This is one of the ways - by connecting up the instances of struggle - that Edward Thompson uses history politically. See especially the title essay in Writing by Candlelight and Greg McLennan's assessment in Chapter 3.

11 We may expect something of a boom of historical work on the 1940s, for example, partly because of the release of archives. But this comes long after popular memory, the media and the politicians have invested deeply in the period which looms increasingly large in political discourse.

12 This applies especially to attempts to apply theorizations derived from other national cases which 'forget' their specific origin: e.g. applications of Gramsci's 'hegemony'; the discussion of socialism and nationalism in Ernesto Laclau, Politics and Ideology in Marxist Theory (New Left Books, 1977).

13 CCCS Popular Memory Group is currently working on various present-day memories of the 1940s.

14 For an example of popular Conservative historiography of a proto-racist character see the works of Sir Arthur Bryant, e.g. English Sage (Collins and Eyre and Spottiswoode, 1940).

15 See especially, C. A. R. Crosland, The Future of Socialism (Jonathan Cape, 1962).

16 See Chapter 1.

17 See especially Chris Cook and John Stephenson, The Slump: Society and Politics during the repression (Jonathan Cape, 1977), but as can be seen in the citations of this book economic historians have been developing a parallel argument for some time now.

18 Perry Anderson, 'Components of the national culture', in A. Cockburn and R. Blackburn (eds.), Student Power (Penguin, 1969).

19 Quintin Hoare and Geoffrey Nowell-Smith (eds. and trans.), Selections from the Prison Notebooks of Antonio Gramsci (Lawrence and Wishart, 1971), passim but especially pp. 324-5.

20 See especially the debate between Ken Worpole, Jerry White and Stephen Yeo in Samuel, People's History, pp. 22-48.

21 The FWWCP was founded in 1976 and 'links some twenty or more working class writers' workshops and local publishing initiatives around the country'. For a useful account of the history of History Workshop see Raphael Samuel, 'History Workshop, 1966-80' in Samuel,People'sHistory, pp. 410-17.

22 For the first results: Paul Thompson, The Edwardians (Weidenfeld and Nicolson, 1975).

23 Paul Thompson, The Voice of the Past: Oral History (Oxford University Press, 1978).

24 Oral History: The Journal of the Oral History Society (first published 1971).

25 Nor do we want to imply that Paul Thompson's work is peculiarly 'flawed' in any way. We are interested in it as typical of oral history in its more profes- sional historical connection. We are grateful to Paul Thompson himself for correspondence and discussions which have helped to clarify points of agree- ment and disagreement.

26 Luisa Passerini, 'Work ideology and consensus under Italian fascism', History Workshop Journal, no. 8 (1979), pp. 82-108; Luisa Passerini, 'On the use and abuse of oral history' (mimeo translated from L. Passerini (ed.), Storia Orale: Vita Quotidiana e CulturaMateriale delli Classe Subalterne (Rosenberg and Sellier, 1978.) We are grateful to the author for sending us a copy of this paper. See also her position paper given at History Workshop, 13: 'Oral history and people's culture' (mimeo, Nov.-Dec. 1979).

27 Passerini, 'Italian fascism', p. 83.

28 Passerini, 'Use and abuse', pp. 7-8.

29 Ronald Fraser, Blood of Spain: The Experience of Civil War 1936-39 (Allen Lane; 1979). See also Ronald Fraser, Work: Twenty Personal Accounts, 2 volumes (Penguin, 1967).

30 We are grateful to Bill Schwarz for sharing his responses to this book.

31 'A people's autobiography of Hackney', Working Lives, 2 vols. (Hackney WEA and Centreprise, n.d.). For Centreprise more generally see Ken Worpole, Local Publishing and Local Culture: An Account of the Centreprise Publish- ing Project 19 72- 77 (Centreprise, 1977), and Centreprise Report (December 1978).

32 Keith Armstrong and Huw Beynon eds.), Hello, Are You Working? Memories of the Thirties in the North East of England (Strong Words, 1977); Strong Words Collective, But the World Goes on the Same: Changing Times in Durham Pit Villages (Strong Words, 1979). We are grateful to Rebecca O'Rourke for introducing us to the work of this collective.

33 ibid., p. 7.

34 For example, the use of autobiographical material in J. Liddington and J. Norris, One Hand Tied Behind Us (Virago, 1978).

35 Jean McCrindle and Sheila Rowbotham (eds.),DutifulDaughters (Penguin, 1979).

36 Jeremy Seabrook, What Went Wrong? Working People and the Ideals of the LabourMovement (Gollancz, 1978).

37 Michel Foucault, 'Interview', in Edinburgh '77Magazine (originally published in French in Cahiers du Cinema (1974)). See also Radical Philo- sophy, no. 16 (1975).

38 Philippe Hoyau, 'Heritage year or the society of conservation',LesRevoltes Logiques (Paris), no. 12 (1980), pp. 70-7. See also the report on Cahiers du Forum - Histoire in Les Revoltes Logiques, no. 11 (1979-80), p. 104, a group with similar interests and aims to our own. These French debates are further discussed on pp. 255-7.

39 Hence the debate in Britain on radical filmic practices and historical drama. See, for example, Colin MacCabe, 'Memory, phantasy, identity: Days of Hope and the politics of the past', Edinburgh '77Magazine; Keith Tribe, 'History and the production of memories', Screen, vol. xvii, no. 4 (1977-8); Colin McArthur, Television and History (British Film Institute, 1978).

40 G. R. Elton, The Practice of History (Sydney University Press and Methuen), pp. 52-3.

41 The Practice of History is in large part a response to E. H. Carr, What is History? (Penguin, 1961).

42 Thompson, Voice of the Past, p. x.

43 ibid., p. 17.

44 Chesneaux, Pasts and Futures, esp. ch. 2.

45 Thompson, Voice, p. 5.

46 ibid., p. 8.

47 ibid., p. 11.

48 ibid., p. 226.

49 Stephen Koss, 'Review of The Edwardians', The Times Literary Supple- ment, 5 December 1975.

50 Thompson, Voice, pp. 9 1 -8.

51 ibid.,pp.100-37.

52 Graham Dawson, 'Oral history: a critique of Voice of the Past', CCCS Stencilled Paper (forthcoming).

53 Thompson, Voice, pp. 91-8.

54 ibid.,pp.100-37.

55 This holds, perhaps, for cultural studies more generally. It is interesting, for example, that early critiques of the media and especially of news and current affairs programmes were heavily influenced by a dual critique of empiricism - in Roland Barthes's early semiological writing and in the pheno- menological sociologies of Schutz and Cicourel. See, for example, Stuart Hall, Yhe structured communication of events', CCCS Stencilled Paper, no. 5.

56 See Chapter 4 and Gregor McLennan, Marxism and the Methodologies of History (Verve, 1981).

57 Jerry White, 'Beyond autobiography' in Samuel, People's History, especially p. 47. For examples of White's own practice see 'Campbell Bunk: a lumpen community in London between the wars', History Workshop Journal, no. 8 (1979); Rothschild Buildings: Life in an East End Tenement Block 1887- 1920 (Routledge and Kegan Paul, 1980).

58 As Yeo puts it, 'Yes, Jerry, the assumption underpinning nearly all of this work is that for working people to speak for themselves, about their own history, is somehow a political act in itself.' (Stephen Yeo, 'The politics of community publications' in Samuel, People's History, p. 46.) For the idea of a long revolution' by this route see Worpole's 'A ghostly pavement', ibid., pp. 31-2. For a full critique of this view of 'experience' see Perry Anderson, Arguments Within English Marxism (Verve, 1980), pp. 25-39.

59 Compare, for example, the autobiography of Emily Bishop (clearly a writer) and that of John Welsh (clearly a talker) in Working Lives, vol. i, pp. 8-12 and pp. 31-50.

60 WorkingLives is full of stories of the first two kinds. For the confident 'male' form (with a notably class-conscious moral) see John Welsh's story about the death of his brother (p. 38); for examples of the more 'open' kind see Betty Ferry's story about 'Bill' (pp. 11-12) and Lil Smith's memories of bath night (pp. 62-3). For examples of the proverbial form of collective historical memory see Dave Douglas, ' "Worms of the Earth": the miner's own story', in Samuel, People's History, pp. 61-7.

61 'For experience' as a 'junction concept' see E. P. Thompson, 'The politics of theory', ibid., pp. 405-7.

62 Passerini, 'Italian fascism', p. 91.

63 ibid., p. 92.

64 For the idea of 'surprise' more generally in qualitative research see Paul Willis, 'Notes on method', in Hall et al. (eds.), Culture, Media, Language (Hutchinson, 1981), pp. 90-1.

65 Roland Barthes,Mythologies (Paladin, 1972), p. 112.

66 Thompson, Voice, pp. 204-5.

67 ibid., pp. 219-20.

68 ibid., p. 129.

69 ibid., p. 209.

70 Thompson, Edwardians, p. 3.

71 ibid., p. 3.

72 ibid., p. 4.

73 ibid., p. 66.

74 It is important to note that the life histories used in The Edwardians are compilations from answers to questions from interviewers and not, or only in a lirnited sense, autobiographies. The basic materials for all the accounts are the simple 'factual' responses. The life histories arise from stringing together these answers in a narrative form. Often the questions (see Voice of the Past, pp. 243-52) can be reconstructed under the narrative.

75 Thompson, Edwardians, p. 4. Evidently 'tidied up' a little, however.

76 ibid., p. 4.

77 Louis Althusser, Essays in Self-Criticism (New Left Books, 1976), esp. pp. 201 ,1; E. P. Thompson, The Poverty of Theory and Other Essays (Merlin 1978; Victor Seidler, 'Trusting ourselves: marxism, human needs and sexual politics', in Simon Clarke et al., One-Dimensional Marxism: Althusser and the Politics of Culture (Allison and Busby, 1980), pp. 103-56.

78 Anderson, Arguments, especially ch. 2; Raymond Williams, 'Individuals and society' in TheLongRevolution (Penguin, 1965);Williams,Politics and Letters (New Left Books, 1979), pp. 271-302.

79 Wal Suchting, 'Marx's theses on Feuerbach', in John Mepham and D.H. Ruben (eds.), Issues in Marxist Philosophy, vol. ii, pp. 5-34; Roy Bhaskar, 'On the possibility of social scientific knowledge and the limits of naturalism', in Mepham and Ruben, Issues, vol. iii, pp. 107-39.

80 For what follows see 'What is Man?', Prison Notebooks, pp. 351-7.

81 ibid., p. 356.

82 Gramsci is clearly aware of the issues but, in the commonest marxist move, takes 'biological' conceptions of human nature as the main antagonist. On these important issues see Kate Soper, 'Marxism, materialism and biology' in Mepham and Ruben (eds.), Issues, vol. ii, pp. 61-100.

83 We do not wish to imply that feminist thinking is homogeneous and what follows represents our own appropriations. For recent statements of some of the key issues see Michele Barrett, Women's Oppression Today (Verve, 1981), the debates in Feminist Review on the work of Christine Delphy and over the concept of 'patriarchy' and the exchanges between Sally Alexander, Barbara Taylor and Sheila Rowbotham in Samuel, People's History.

84 Vera Brittain, Testament of Youth (Fontana with Virago, 1979). We are grateful to Rebecca O'Rourke for drawing attention to the signuDicance of this example and for discussions on the book and TV programme.

85 ibid., p. 12.

86 ibid.

87 Winston S. Churchill, The Second World War, 6 vols. (Cassell, 1948-54).

88 Something of the character of what follows is caught in Churchill's Preface to vol. I (e.g. 'I am perhaps the only man who has passed through both the supreme cataclysms of recorded history in high executive of fice', ibid., vol. 1,p.vii.)

89 Thompson, Voice, pp. 100-4.

90 ibid., pp. 129-37.

91 Betty Ferry, in WorkingLives, vol. i, p. 125.

92 Emily Bishop, ibid., p. 10.

93 John Welch, ibid., p. 39.

94 Peggy Wood, Dutiful Daughters, p. 174.

95 Janet Daly, ibid., p. 19.

96 Editor's introduction, ibid., pp. 4-5.

97 For an interesting example see Maggie Fuller on 'Nowadays' where a whole historical evaluation is woven around 'things' (ibid., pp. 135-6).

98 Betty Ferry, Working Lives, vol. i, p. 117.

99 Stuart Hall et al., Policing the Crisis (Macmillan, 1978); Stuart Hall, 'Thatcher- ism - a new stage', Marxism Today (February 1980), pp. 26-8; CCCS Educa- tion Group, Unpopular Education: Schooling and Social Democracy since 1944 (Hutchinson, 1981).

100 Annie Davison, DutifulDaughters, pp. 65-6.

101 ibid., pp. 61-2.

102 Raphael Samuel, 'British ma xist historians, 1880-1980, part i', New Left Review, no. 120 (1980), pp. 23,34-7, 85-91; Schwarz, Chapter 2 of this volume.

103 For oral history in the USA see Passerini, 'On the use and abuse of oral history' (cited above, note 26).

104 For example, the contributions from Ruskin ax-students in Raphael Samuel (ed.), Miners, Quarrymen and Saltworkers (Routledge and Kegan Paul, 1977), and Village Life and Lahour (Routledge and Kegan Paul, 1976).

105 These thoughts are based in part on discussions with students at Coleg Harlech and are stimulated by a paper given by students there (at a day school with members of CCCS) on the contradictions of their situation.

106 We have in mind, especially, Richard Hoggart and Raymond Williams but lots more examples could be given.

107 Especially through the decimation of the Social Science Research Council, the cuts in higher and further education, the disappearance of new posts at colleges, polytechnics and universities and the rising fees for postgraduate study. All this contrasts sharply with the expectations and possibilities of the 1960s and early 1970s.

* This essay is based upon the collective work of the Popular Memory Group in the centre for Contemporary Cultural Studies which met between October 1979 and June 1980. The group consisted of: Michael Bommes, Gary Clarke, Graham Dawson, Jacob Eichler, Thomas Eock Richard Johnson, Cim Meyer, Rebecca O,Rourke, Rita Pakleppa, Hans-Erich Poser, Morten Skov-Carlsen, Anne Turley and Patrick Wright. This piece was written by Richard Johnson with Graham Dawson.

"Popular Memory: Theory, Politics, Method" in Making Histories: Studies in History-writing and Politics. Ed. Richard Johnson. London: Hutchinson, 1982.