There is no doubt that Hayden White's Metahistory: The Historical Imagination in Nineteenth-Century Europe (1973) transformed the landscape of theory and philosophy of history and historiography. A telling indicator of his influence is the praise that has been heaped upon him and his Metahistory, both at the time of its appearance and retrospectively by many eminent scholars in the field. Soon after the publication of Metahistory, Mink said, quite correctly from the contemporary vantage point, that it was 'the book around which all reflective historians must reorganize their thoughts on history' (1987, 22). Ankersmit has called Metahistory the 'most revolutionary work on philosophy of history' (Ankersmit 1986, 18). Brian Fay, writing at the end of the millennium, in turn suggests that 'an important shift in philosophy and theory of history occurred twenty-five years ago' when Metahistory appeared (Fay 1998, 2).
There are many ways in which one could characterize White's revolutionary impact as well as many different aspects in his thinking that one could highlight. Nevertheless, one of the most fundamental changes that White brought about is arguably the shift of focus from individual statements about the past onto entire texts of history. This perspective on historiography is absolutely crucial if one wishes to understand what the discipline is about. Take any book of history into your hand and open it; you will notice that it is indeed a text. Studying and reading it a bit more closely reveals that it contains lots of sentences, but also chapters. Typically, the book has a beginning, middle sections and a conclusion. If one is studying a monograph on history, by far the most common book type in historiography, we soon realize that all these elements, the text in its entirety, seem to function as a whole to form a view or views of the past. If one is very busy, one could read only the blurb on the back, which typically provides an overview in a few sentences. Perhaps one would read the conclusion as well, but if one really wants to understand what is argued and how, one has to read through all or most of the text in the book. The text encapsulates the most important product of historical scholarship.
In order to illustrate the narrativist perspective, I mention an example. Fritz Fischer's Germany's Aims in the First World War (1967) (Griff nach der Weltmacht) is an eminent study of the origins of the Great War and contains a wealth of information and statements about the past. On page 310 it is claimed that Austro-Hungary's aim in the war was to 'get the greatest possible increase of power and security when things are re-arranged'. To take another random example, on page 510 we are told that Soviet Russia recognized the independence of Finland on January 4 following personal negotiations between Svinhufvud and Lenin. However, Fischer's book is not about the war aims of Austro-Hungary or the independence of Finland or the Soviet policy at the end of the war. It is about the aims of Germany in the Great War and implies a substantive thesis: the First World War was a premeditated German bid for power. That is the central historical view of the book and something that the reader of the book should not miss. In other words, theses like the one in Fischer's book are the most important contributions of historiography to our understanding of the past. One can of course be interested, and often is, in some singular claim included in a book and may wonder how the historian justifies it, but such claims are clearly subsidiary and integrated to (the argument for) the general point put forward.
Frank Ankersmit's Narrative Logic: A Semantic Analysis of the Historian's Language from 1983 amounts to another landmark publication in contemporary narrativism. It is a book that developed narrativism into new and philosophically more explicit directions and can be said to have opened a whole new discursive level in the philosophy of historiography. Similarly to White, Ankermit suggests that we understand historical works as producing holistic literary or linguistic theses, such as the idea that there was an industrial revolution at the end of the 18th century, that the 17th century was an age of crisis or that there was a cold war between the superpowers after the Second World War. These kinds of theses are compared to 'comprehensive, panoramic interpretations of large parts of the past' (1983, 7, 15). Indeed, as already mentioned previously, the absolutely fundamental function of narratives is, according to Ankersmit, 'how historians integrate a great number of historical facts into one synthetical whole' (1983, 15). He points out that the study of narrative is something that the earlier philosophy of history left out (2001, 53), which is largely but not entirely true, as we saw above. In brief, 'Narrativism is that view which requires the historical theorist to focus on the whole of the historian's text and not its constituent parts (for example, its individual constative or causal statements)' (Ankersmit 1995a, 155).
Ankersmit stresses that, if we omit the narrative level of historiography, we ignore the feature that is most characteristic of historiography. This idea is expressed powerfully in the following quotation:
In the first place, it cannot reasonably be doubted that these narrative sentences are paradigmatic of all historical writing: historians state facts for no other purpose than to relate them to other facts. So take away from historical writing these narrative sentences and what they effect, and you have transformed historical writing into a corpse without a heart. (1987, 68)
Historical writing without narrative sentences is thus a body without its heart. Who would like to reduce such a flourishing and dynamic practice to a corpse?
The difference with the analytic philosophy of history is clear as analytic philosophy did not seek to understand and analyze the nature of historical texts and the comprehensive historical theses that they contain. The analytic philosophy of history was too much driven by its normative ambitions to link historiography to the family of sciences. Even the early narrativists in all their well-intentioned attempts at studying the narrative aspects of history-writing ended up examining contrived 'two-sentence narratives'. This neglect makes the contribution of the narrativist philosophy of historiography extremely valuable since it highlighted an overlooked narrative aspect of historians' practice and analyzed the products that result from this practice.
Both the name 'narrativism' and the discussion above suggest that the central theoretical notion of the narrativists is that of 'narrative'. It is something that organizes textual material into one intelligible form. Now, it is of course a contentious issue what, exactly, narrative is. We saw above that most of the early narrativists understood narrative as a story, typically containing a beginning, a middle and an end. But note that Hayden White actually says very little about narratives as such in his Metahistory, despite the fact that he is given the role of 'founding father' of contemporary narrativism. On the other hand, Ankersmit discusses the nature of 'narratives' at length in his Narrative Logic. And it can be said that 'narratives' and equivalent technical terms 'narratios' and 'narrative substances' are certainly not stories to him. He writes that 'whenever in this book the terms "narration" and "narrative substance" are to be used, all associations with the belles-lettres and with a storytelling kind of historiography should be avoided' (Ankersmit 1983, 16; similarly 1986, 2). The function of narrative is to propose points of view on the past. Metaphysically they are 'primary logical entities in historiographical accounts of the past' (1983, 94) and amount to 'the third logical entity' in addition to the subject and predicate already known in propositional logic (1983, 95).
Now, it is not important to nail down here what 'narrative' is. Much more important is to recognize that White and Ankersmit both paid attention to the unifying structures of historiographical studies and attempted to define them in more precise terms. Even if we assumed that a historical study necessarily contains a beginning, a middle, and an end, it is obvious that there is much more to say about how historians present their most important research results. In other words, the unifying structure is much more complicated and much more philosophically involved than the talk about 'story' implies. This is the issue that the two central narrativists have pursued in their own ways. Ankersmit's strength is that he has detailed the philosophical implications, which we discuss in the following chapter. White's take is more elaborate when, in Metahistory, he describes the structural components that result in concrete 'stories' or 'narratives'.