As in the discussion on explanation, there were two sides in the debate of the early narrativists. A number of scholars, such as Arthur Danto, W. B. Gallie, Louis Mink, and Morton White suggested that narrativity is characteristic for historiography and distinguished it from the sciences. On the non-narrativist side, opposing the view that takes narrative as an essential feature of historiography, we find Mandelbaum and Behan McCullagh and a few others supporting their line of argumentation.
One of the earliest noteworthy suggestions is Danto's idea of 'narrative sentences' that amount to 'a differentiating feature of historical knowledge' (Danto 1962, 146). These are sentences that refer to at least two time-separated events although they describe or 'are about' only the earlier event. A simple example offered by Danto is: 'The Thirty Years War began in 1618'.
Danto shows that, because of the narrative form of historical knowledge, historical presentations cannot be made to correspond to the past. Danto defines the 'full description' of an event E as a set of sentences that 'state absolutely everything that happened in E' and says that there is an isomorphism between the full description and the event of which it is true (Danto 1962, 151). Then Danto imagines an Ideal Chronicler, who knows everything that happens, including people's minds, in the moment it happens, who is also able to write down exactly how it happened (Danto 1962, 152). Danto notes that there is a class of descriptions of any event under which the event cannot be witnessed, and which are therefore necessarily excluded from the Ideal Chronicle produced. They are the kinds of descriptions that contain a future-reference or other associated post-dated significance that cannot be observed at the time of occurrence. No one could have observed that The Thirty Years War began in 1618, because no one could have known at the time that it would last 30 years. Or to take another example of Danto's, suppose that two scientists A and B at two different points of time t1 and t2 discover a certain theory T, but that the first scientist A never published his results and that his discovery is found much later, after B has already been given the credit for the discovery. A historian of science might describe the situation saying that 'A anticipated at t1 the discovery by B of T at t2'. 'Anticipation' is something the Ideal Chronicler cannot even in principle witness at the time of an original event (Danto 1962, 154-5, 158-159).
Danto's claim is that historians use narrative type of descriptions, which attach retrospectively such valuations and significations to the events that are not part of the events themselves. Any event can be placed under different descriptions and seen from the angle of a later one. The condition for the Ideal Chronicle is then that the Ideal Chronicler should be able to know not only future events but also the minds of future historians - that is, what future historians are interested in and what later events they will relate to the earlier ones. Danto thus demonstrates that historical accounts cannot be reduced to their object, the past, in any simple way, but that historical interpretation implies unavoidable subject-sidedness; it involves 'an inexpungible subjective factor' (Danto 1968, 142).
Gallie suggested that story or narrative is essential to all historywriting and peculiar to historical understanding. What is more, story is not merely an essential, defining characteristic of historiography, but 'every genuine work of history ... is a species or special application of the genus story’ (Gallie 1969, 169; my emphasis)5. Louis Mink drafted an interesting response and a development of Gallie's contribution. Mink does not criticize Gallie's emphasis on narrative, but writes that it is a mistake to think that the essential feature of narrative is following a story, whose conclusion the reader does not know. The 'configurational mode' as typical for historical comprehension treats a number of separate things as 'elements in a single and concrete complex of relationships' so that they are 'just in balance' (Mink 1970, 551). So a letter may be related to a story or narrative that relieved a misunderstanding at a crucial moment. Further, Mink is convinced that 'stories are not lived but told' (Mink 1970, 557), more specifically that only stories - and not life - have beginnings, middles and endings that weave together separate images of recollection. His idea is thus that the narrative qualities are transferred from art to life. This is something that came to be endorsed by Hayden White in the years that followed, but was intensely disputed by the phenomenological narrativists such as David Carr (1986).
What is particularly interesting in Mink's account is the suggestion, in contrast to almost all other scholars at the time (with the possible exception of Danto), that although narratives are conceived of as stories, time is not of the essence. It is rather that the techniques of narratives are 'instruments for facilitating the comprehension of the story as a whole' (Mink 1970, 555). The actions and events of a story are 'connected by a network of overlapping descriptions' (Mink 1970, 556) that goes beyond a story and temporal sequences. Danto had claimed earlier that narrative imposes a structure upon events, grouping some of them together with others and ruling some out as lacking relevance (Danto 1968, 132; 140). And he saw narrative itself as a form of explanation (Danto 1968, 141; 237). In this way also Danto differs from most other discussants, who doubted whether narration can entail a structure and be explanatory.6
Another major figure who wrote favorably on the narrative aspects of history is Morton White.7 M. White8 writes that narration is 'the typical form of discourse employed by the historian' (M. White 1965, 4) and that narration must have a central subject of which the narrator gives 'a connected account of the development' (M. White 1965, 221). The fundamental problem for M. White is the question of how historians evaluate each other's works, provided that these evaluations need to go beyond the truth and falsity of the statements that compose the works. In other words, it is possible that all competing histories are true in the sense of containing only true statements, which means that on this basis one cannot make a difference between them (M. White 1965, 225). According to M. White, this issue comes down to the question of what reasons the historian has for including some statements rather than others in his narration. Or, 'what do historians mean when they say that one history of a given subject is better than any other?' (M. White 1963, 8). M. White argues for 'a pluralistic view of admissibility', which says that there can be various reasons for inclusion. Sometimes statements are included because they are assumed to present the typical features of a given period or because of their 'colligatory power' (M. White 1965, 257, 263-264). But on other occasions the choice reflects the historian's interests and value judgments about what is historically important or worth remembering. The message is that the historian is allowed to choose 'his facts with eyes on all kinds of considerations, so long as he writes true and connected narrative' (M. White 1965, 259).
Mandelbaum (1967) was the first to react to the line of argumentation that places narrativity in a central role in historiography. Symptomatically with regard to the early discussion on narrativity, Mandelbaum treats stories, narratives and 'connected chronicles' largely as synonymous. His main point is that the historian is not 'engaging in an activity which is best represented by the model of telling a story', not even in a case in which constructing a 'sequence of occurrences' is the main aim (Mandelbaum 1967, 414). According to Mandelbaum, the problem is that 'the sequential story' does not amount to a proper analysis of the complex contextual factors that resulted in the outcome and neither does it amount to a satisfactory explanation as to why the events occurred as they did. Mandelbaum accuses the early narrativists of relativism because they allegedly hold that some facts and their relationships are regulated at least in part by the historian's story and are not given independently.
Also McCullagh (1969) joins the critical chorus and casts doubt on whether a narrative presentation could amount to an adequate historical explanation. The problem is, as he puts it, that: 'To explain how a situation changed is not the same as explaining why the change occurred' (McCullagh 1969, 258). The narrative explanation for McCullagh is describing the steps of change in a temporal chain, but the proper historical explanation has to be predictive. This makes him conclude that the narrative style has only dramatic value in that it may help the reader to experience the same kind of surprise as the historical figures themselves did when the events unfolded.
What kind of conclusions can we draw from the early narrativists' discussion? The early narrativists highlighted the narrativity of historiography and narrativity entails two important ideas: (1) there is a lower and higher level of cognition in historiography; and (2) the truth-values of the lower level statements cannot be translated into a justification (or falsification) of higher-level cognition. This is the central premise in the subsequent narrativist philosophy of historiography. They also shed light on the constructivist and colligatory character of historiography, seen in Danto's subject-sided 'narrative sentences', Mink's 'configurational mode' and M. White's 'central subject'. Further, with the exceptions of Danto and Mink, 'narrative' was fairly straightforwardly equated with 'story'. It was seen to designate temporality, the before-after structure, and sequential order. The third conclusion is that the analysis of narra- tivity focused quite narrowly on chronological or temporal structure. In other words, no early narrative analyst paid attention to actual historical texts or narratives, but they all focused on short abstracted sentences at best. It is just, as Richard Vann writes, that the analytical philosophy of history tended to dissect historical discourse into its smallest intelligible units, such as two-sentence narratives, and as a consequence ignored the questions of genre, plotting and the fundamental organizing principles of history (Vann 1995, 61).
'Narrativity' in this early stage specifically did not address any kind of literary features embedded in historical texts. The predominant question was not what kind of narrative structure is characteristic of historiography - as a historical text was not the given object of analysis - but whether historical presentation requires narrative form and whether narrativity is essential for historiography. The question that preoccupied the minds of the early narrativists was whether historical explanations rely on a general explanatory scheme and whether narrative explanation lowers the degree of generality and scientificity of historiography. This concern related to the debate on the subjectivity and objectivity of historiography. The early narrativists, such as Danto, Gallie, Louch, and M. White, argued that the narrativity of historiography means that historiography inevitably contains subjective elements, while the critics tried to counter the claim by questioning the view that historiography is fundamentally a story or narrative.
The whole discussion took place under the shadow of logical positivism and empiricism, that is, a debate focused on causality and explanation, with specific reference to the covering-law model. As Vann points out, philosophical problems of historiography were often discussed by writing about cars with burst radiators or the indigestion that afflicted Jones after his ingestion of parsnips (Vann 1995, 41). Those who objected to viewing historiography as essentially narrativist typically contrasted explanation to narrativity, implying that a (mere) sequential ordering does not amount to or cannot be translated into a proper explanatory account of historical events. It was often suggested that a narrative account cannot involve contextual explanations, facts external to a sequence of events, or other explanatory principles. And even the early narrativists, who took a positive view of narrativity in historiography, referred to historical narration as a peculiar kind of explanatory account. Narrativity was seen as a 'distinct kind of explanation' (Louch 1969, 58), as 'self-explanatory' (Gallie 1964, 108) or itself as 'a form of explanation' (Danto 1968, 141, 237, 251).
Despite all the often very enlightening analyses of historiography by the early narrativists, disciplinary externalism characterizes their point of view. In other words, concerns and questions emanated more from general philosophy of science than from the field-specific problems of historiography. This is indicated most obviously by the early narrativists' preoccupation with the notion of explanation and by their sensitivity to the unity-of-science theorizing on historiography. The emergence of narrativism in the beginning of the 1970s reset the focus of the theoretical discussion in the philosophy of historiography.