Where Does History Come From?

History Today Volume: 52 Issue 3 March 2002 pp18-20

Alun Munslow argues that the centrality of narrative to history undermines empirical views of the subject.

Where does history come from? This may seem like an odd question. Surely history comes from the traces of the past that historians find in their sources? However, we might get a different answer if we put the question in another way. What happens if we choose to view history as what, from one perspective at least, it plainly is: a narrative written about the past constructed by the historian in the present? This is clearly not the way history is conventionally defined. To be technical for a moment, it is more usually described as an empirical and analytical undertaking – a source-based and inferential activity concerned with the study of change over time. I am posing this question – where does history come from? – because I think historians still tend to ignore the role of narrative in doing history.

What is a historical narrative? I define it as that written composition of historians that encompasses their source-based data founded on certain principles of selection and organisation. So far so good? But, in addition, the historical narrative also encompasses the arguments used by the historian to establish cause-and-effect relationships between past events. What is more, the historical narrative is also the site of the historian’s emplotment (what the historian thinks the order of the events described lead up to and mean). Additionally, the narrative is where the ideology and the social theory preferences of the historian exist and do their work. The historical narrative is far more than a chronology of events. A study of the historical narrative and how it works highlights another thing. It should make it clear how ‘the past’ and ‘history’ are quite different objects. The former is what actually happened but which is now gone, while the latter, although it is a source-based and inferential inquiry, is only ever its narrative representation. History is, therefore, a substitution for the now absent past.

So, what are the consequences of history as viewed primarily as a narrative act? Inevitably, it raises questions about the objectivity and truth-acquiring character of history. According to established professional practice, historical knowledge is acquired by the operation of reason and rationality as applied to the historical sources. Through this empirical (i.e., source-based) and analytical (i.e., inferential) process we extract what we think is most likely to be the past’s true meaning. Such a perspective (and process) is a reflection of the Enlightenment or modernist theory of knowledge (in philosophical terms its epistemology). It works through the belief that scientific method is the approach to be followed in doing history as much as is possible given the peculiar nature of history’s subject matter – people, human actions, and social cultural, political and economic processes and events. The necessary adjunct to this is that it is possible to represent that knowledge more or less precisely and objectively in our histories.

It is this everyday common-sense belief that prompts the idea of the historical narrative as primarily a chronology of events (what happened) and that its prose should be referential, sober, unembroidered, and fundamentally reflective of the nature of historical change. In other words, the history we write is controlled by the hours of painstaking work undertaken in both the primary and secondary sources rather than any subjective input by the historian. This suggests, as Arthur Marwick has argued in The New Nature of History (Palgrave, 2001) – writing against the grain of my definition of history as primarily an act of narrative making – that history has ‘nothing to do with the nonsense about the need for “emplotment” or “narrative” (in the imagined sense of the postmodernists)’.

Okay, so I am, in Marwick’s terms at least, suggesting a kind of postmodern take on history. Anyway, the point Marwick is making is that the historian does not create or invent the structure found in the history text. Rather, as he insists, there must be

a logical order, a series of connections and interrelationships (in short a ‘structure’), which will be as true to the actual aspects of the past … as it is possible for a historian producing knowledge about these aspects of the past to make it.

In other words, the historian is not devising the narrative of the past but re-telling it and therefore getting the story straight according to the sources.

While this (empirical-analytical) undertaking is a complex and difficult thing to do, it nevertheless fails to adequately address the fundamental role of the historian. What an increasing number of the profession and philosophers of history have tried to do of late is revisit the nineteenth-century critique of this source-based approach. This is not to reject it out of hand, but to look afresh at its claims. They have re-read their Kant, Hegel and Nietzsche as well as those critics of the conventional approaches of the last century – Benedetto Croce, R.G. Collingwood, E.H. Carr, and Michel Foucault. What contemporary critics such as Jacques Derrida, Paul Ricoeur, Keith Jenkins, Hayden White, Richard Rorty, Louis Mink, Frank Ankersmit (and myself for that matter) have done is to more fully explore the cognitive role played by language, specifically the act of writing history in doing history. We have, in our various ways, moved beyond (but not necessarily relegated) the technical procedures of how empirical knowledge is derived, but we have concentrated more on the nature of its representation and what such a study does for the status of history.

The most important early and continuing contribution to this so-called ‘linguistic turn’ away from seeing history as solely an empirical-analytical activity was made by the American historian Hayden White (in Metahistory: The Historical Imagination in Nineteenth Century Europe, published in 1973, and in his work subsequently). While never denying factuality and the pursuit of truth, White confronted the notion that the truth of the events of the past can be read off from the sources and reconstructed in a mirror-like historical text. As the result of this insight White has been said by some to be the voice of ‘anything goes’ and a denier of past reality. This is quite unfair. White’s ‘offence’, it seems to me, is to want the truth but to recognise you can’t have it without it being filtered through the concepts, cultural predispositions and, above all, narrative constructions of the author-historian. This is, it strikes me, a far more realistic approach to knowing what the meaning of the past might be.

Since the 1970s we have experienced the emergence of a new feminist and cultural history encouraged in part by Hayden White’s challenge. By the early 1990s this new history not only broadened and made more inclusive the scope of historical study but it also made us more self-reflexive about our technical and operational procedures, increasingly conscious of the relativism that exists in the humanities generally and in conceptualising and writing history in particular. Such developments impelled the ‘new cultural historians’, while experimenting with the shapes and form they gave to their histories, to explore the possible narratives that existed in the past.

Among those pursuing the new ‘real truth’ have been historians and writers on history such as Natalie Zemon Davis, Mary Poovey, Judith Walkowitz, and Elizabeth Deeds Ermarth. But, there were also some historians who pushed their a priori assumptions further and followed the logic of White’s challenge and became the real epistemological radicals, historians who experimented by placing form (the historical representation or ‘history’) as prior to content (the reality of ‘the past’). Among this group are Simon Schama, Robert Rosenstone, and those increasing numbers of contributors to journals like Clio, Common Knowledge and Rethinking History: The Journal of Theory and Practice.

The fundamental contribution of such radical and experimental historians is the recognition that, as history is a discourse about past objects, then it is the historian who must first imagine and conceptualise the past in narrative terms before she or he can bring to bear upon it the methods and procedures they wish to deploy to explain it. Historians are, of course, self-conscious enough to understand that the process of organising the data (derived from the sources) means history can never reconstruct the past as it actually was. The reasons usually given, however, are problems with data and/or poor inference (missing sources and bad historians). It is not because of the frailties or fictive (not fictional) nature of representation. By fictive I mean that history is subject to the same narrative and imaginative constraints as any other form of realist writing. History is not, as the empirical-analytical approach has it, a discourse outside literature. Arguably, the primary mechanism the historian has for giving a meaning to her sources are the techniques of figurative language. All histories contain such figurative elements and they serve as models of reality and give a form to the past. Marx, for example, used the base-superstructure metaphor. E.P. Thompson described the working class variously in romantic or tragic terms as he believed best fitted the data he had chosen to deploy. Herbert Butterfield once described history’s nature by feminising it: ‘she is a harlot and a hireling … she best serves those who suspect her most’. E.H. Carr referred to facts as fish on a slab, and we still often think of history as an act of discovery with the historian cast as an explorer.

Being conscious of the power of narrative to invest the past with meaning has worried especially the rump of hardcore empiricist historians into highlighting what they perceive to be the most dangerous aspect of the narrativist-linguistic turn: relativism. Their reasoning is that once the historian places form (the historical narrative) on a par with content (the events in the past) then truth is lost. Once truth is lost, next comes moral relativism. As the historian Colin Richmond is quoted as saying: while objectivity is not at all necessary to historical study, knowing the truth is and thus our ability to know the difference between right and wrong. As he is reported as saying, the test of truth is to punch a postmodernist in the face and see if he can explain why it hurts. I guess the postmodernist would counter with: punch an empiricist and then ask if he could tell you what the pain means? Fortunately, examples of historian rage are rare.

The historian of the recent German past, Richard Evans (In Defence of History, 1997), has warned that unless we are very scrupulous in our methods, then language, culture, and ideology can introduce a dangerous relativism into our conclusions. In noting this, the postmodernist historian Keith Jenkins has reminded us that scrupulous methods cannot, in fact, insulate our history from either the relativism introduced by our present experience or the historian’s narrative imposition. While history may be defined by empiricists as the pursuit of the truth of the past (demonstrated by punching postmodernists in the face and asking them why it hurts), it nevertheless remains a pursuit undertaken by people in the present. These people know they have to organise the archive in order to make sense of it and they fictively derive its meaning. History is, after all, something we do; it is not an object we observe.

Nevertheless, reconstructionist historians like Evans, Marwick, Richmond and, most famously Geoffrey Elton in his book Return to Essentials (1991), remain convinced that while they may not be absolutely achievable, objectivity and truthfulness can be approached. This can be done by sticking to technical procedures and maintaining a due respect for the facts (although in his most recent book Marwick has said we should abandon the idea of facts!). In addition we must keep the priority of content over form, and have a faith in the correspondence of the word and the world. What this means is that the honest historian, in getting the story straight, can avoid relativism and its consequence of moral decline.

But, surely, even after the most detailed forensic study of the sources we can only communicate and justify our historical descriptions through a narrative-linguistic and culturally determined form? It may be a dirty job, but someone has to narrativise the past. This is the fact of the matter and it raises doubts in my mind about any mode of thinking that promotes history as a methodology that effectively insulates its knowledge from culture, discourse, and the historian’s figurative imagination – and, what is more, assumes that the sources produce a God’s-eye view. The idea that the past can be recreated in the historian’s mind, and that its shape is not substantially the product of that mind, will not do.

A more sophisticated approach recognises how the historian constructs the nature of history as well as the history eventually written. Quite simply, there can be no historical facts without the application of theory and language-use somewhere along the line. The problem is, of course, one of degree. Once the historian admits that her historical knowledge is always and fully relative to the social and the linguistic, the fear is that it must eventually destroy any possibility of knowing things honestly, accurately and/or ethically about the past.

The issue, then, is will the relativism in knowing and representing which the narrative-linguistic mode of thinking recognises as inevitable, end up denying the existence (and the suffering) of people in the past? It is important to refute this argument. Being an epistemic relativist, that is, believing that knowledge in the humanities is produced in ways more complex than sceptical empiricism allows, does not mean I cannot recognise reality or be a moral person. Acknowledging the boundaries of truth does not make epistemic relativists intolerant, irrational or Holocaust deniers. In fact, those who deny the reality of the past tend not to be postmodernists but empiricists who tell lies. In accepting both the existence of data and the constraints of narrative, epistemic relativists stop well short of denying there is no truth in history. But they acknowledge it to be as much a narrative as an empirical truth.

Because we can only know the past through history (its representation), we really should take the nature of representation into account when we claim to know what is the most likely truthful meaning of the past. I recommend that all history students take a module or two in literary criticism and composition. History does not offer a resemblance to things past; words do not resemble what they represent (class, race, gender, imperialism, nationalism). History is not the past’s verbal or written analogue. Instead, history is a narrative substitution for it. Even when we insert what seem to be well-attested facts into our narrative we are not larding it with atoms of past reality. I would go further and propose that (from my epistemically relativist standpoint) historians are part of history just as we are a part of the past and present.

Although I am a relativist in terms of how I construct the past as history, I do not doubt the existence of the past as located in its source-based data-stream of events and actions. But this data is the raw material out of which ‘facts’, interpretations, and meanings are narratively (fictively) created. In phrases like ‘the Renaissance’ or ‘the Tragedy of Vietnam’ there are no objects to which these descriptions refer as there are buildings, furniture and people making decisions. It is only through the historical representation of narrative that they exist. There is no such ‘thing’ as the Tragedy of Vietnam that we have discovered in the sources.  It is the product of the historians’ narrative construction.

In my history thinking, the data from the sources do, of course, directly influence the history I write, but they do not utterly decide the representational form (the kind of narrative – tragedy, romance, farce) my history must take. The historian can only offer a meaning, one that is as much a product of the imagination that organises the factualised data as the data itself. But what is of most significance is what flows from this. Integral to our historical explanations and the moral lessons we choose to draw from them are our deliberate authorial and prefigurative acts and decisions. It is the historian’s narrative acts – emplotment process, arguments, ideological and moral positions and all the other epistemic choices and preferences – that ultimately invest the past with meaning.