Art or science?
In his paper 'History Brought under the General concept of Art', Benedetto Croce defended the view that historiography is art, in contrast, for example, to his contemporary Johann Droysen, who thought that historiography was definitely a science. The fundamental problem was that historiography had typically been assumed to study particular individuals, singular facts or particulars. Croce writes that, by contrast, science investigates types or 'the general - what exists in all the individual objects' (Croce 2012, 492). As a result, historiography as a science would be a science of the individual, which would seem to entail a contradiction, as Schopenhauer had put it. Indeed, it is the subsumption under the general that often defined science and the representation of the particular that characterized art in nineteenth-century discourse.
Croce writes that while some forms of art represent 'the possible', 'history may be defined as that type of artistic production which has the real event as the object of its representation' (Croce 2012, 500). Perhaps surprisingly for the modern reader, Croce thinks that both artists and historians aim at accurate representation and work in the spirit of observation: 'as the artist cannot lapse into the false, so the historian cannot lapse into the imaginary' (Croce 2012, 500). The only rationale for historiography is 'to tell the facts', and telling the facts means narrating. In passing it may be noted that, for this reason, some authors had suggested that historiography was a special kind of non-explanatory science, that is, a descriptive science (Croce 2012, 492) - but for Croce historiography was a special kind of art that deals with what is real and of historical interest.
In accordance with Croce's view, Daston and Galison (2010) show in their book Objectivity that subjectivity has not always been seen as an obstacle to acquiring truth in science. The norm that guided scientific investigation in the eighteenth-century, which the authors call 'truth-to- nature', required the scientist's (subjective) eye and the artist's (subjective) skill to discern the true essence of natural objects without any of the deficiencies of the actual perceivable things - to correct nature's imperfect specimens. This requirement is similar to the demands on Croce's historian, although such a historian is faced with a yet more challenging situation in the sense that constructing 'a complete narrative' is, by Croce's own admission, only rarely achievable, yet represents an ideal that the historian nevertheless has to try to reach.
While Ranke may have emphasized the subjective, intuitive skills of the historian more than has generally been realized, the philosophical school Rankeanism stressed the need to remove subjectivity entirely from historical narration.1 The historian was asked to 'extinguish oneself' in order to pin down wie es eigentlich gewesen, or 'how it really was'. Rankean historiography may also be dubbed a scientific historiog- raphy2, which relied on the idea that by using critical source methods it is possible to acquire an objective and true history. 'Objectivity' here means neutrality, and thus excludes any literary, poetic and other speculative elements from the historian's narrative. One might say that, in the Rankean paradigm and in its implied historical realism, the rationale of historiography derives directly from its epistemological standing. To acquire knowledge and truths about the world amounts to a self-justified rationale for historiographical practice.
G. R. Elton is an exemplary modern advocate of scientific historiography and historical realism. He writes that 'like all sciences, history, to be worthy of itself and beyond itself, must concentrate on one thing: the search for truth' (Elton 2002, 44). That is, the search for truth, is the whole and self-evident point of doing history. Elton also believed that historical method provides a means of 'extracting from what the past has left the true facts and events of the past' (Elton 2002, 59). Provocatively he states that the historian can only 'discover', and in this way prevent the tendencies of the observer and 'experimenter' being reflected upon the subject matter (Elton 2002, 49). Or even closer to the modern day, we may listen to Arthur Marwick, who says that historians work in the same spirit as natural scientists ('always working from the evidence, always basing their generalizations, interpretations, or theses on the evidence (not on metaphysical speculation)'): 'history should be judged as a scientific activity' (Marwick 2001, 248, 249).
The previous chapter demonstrated how the narrativist philosophy of historiography questioned historical and narrative realism, something present in one form or another in all of the above attitudes. The narrativists specifically dispute that the narratives of complete works of history could, even in principle, be true and that the truth-making qualities of narrative could be 'discovered', because the past does not possess such properties. It follows that the function of history cannot be anything like 'accurate representation' or the 'search for truth' with regard to its central scholarly products. The narrativists also likened historiography to art, arguing initially that historiography resembles literature and subsequently that it should be compared to visual art. This suggests that the worth and the rationale of historiography is to create artistic products with aesthetic value, without any entailment of their reality or truthfulness.
Now, the postulation that historiography is art, or science, or something else, is one thing, and the actual possibility of creating the kinds of entities produced in these disciplines is quite another. For example, one may wish that historiography be a science and that it produce true narratives. But if it so happens that a 'true narrative' is an unattainable goal, as the narrativists have argued, then attaining the status of science on this basis becomes an impossible desire. The narrativists have stated that historiography necessarily creates representations. Commitment to the notion of representation and representationalism is interesting because it cuts across various philosophical orientations. Croce wrote about representations. And similarly to the anti-realist narrativists, historical realists typically think that historiography creates representations, the difference being that they believe those representations can represent the past as it was. In the previous chapter, I began the analysis of what historical representations are representations of, provided that they are not 'copies of the past'. Now this investigation continues. How tenable is the narrativist theory of representation? And how necessary is it for historiography?