Historiography was invented in response to a replication crisis. Herodotus and Thucydides recognised that different people offered different accounts of the same events – that even eye-witnesses, for a variety of reasons, remembered things differently. Thucydides in particular observed how they might then seek to apply this unreliable knowledge, and the false understanding derived from it, to other situations, and were then shocked and surprised that this didn’t in fact work. The political principles that the Athenians established on the basis of their story of the overthrow of the Peisistratids were misconceived and hence not replicable in the case of Alcibiades…
There has been some discussion recently on social media of a piece by the always-interesting historian of technological development Anton Howes, asking whether history might have a replication crisis. The idea of developing such an argument, given how much controversy there has been about the actual replication crisis in various areas of the social sciences, is really clever. Unfortunately, in practice this seems like a category error. A researcher asserting the existence of a general law or principle of psychology or social life on the basis of experimental data, but if the experiment is repeated there’s no sign of said law or principle, is not the same thing as two historians producing different interpretations of the same set of fragmentary and ambiguous data relating to a single specific event or series of events. The historical data are neither controlled nor complete, and there is no possibility of re-running events under the same or slightly varied conditions. To be fair, Howes recognises this, and suggests that we might instead describe it as a ‘reproducibility crisis’, but I’m not sure that gets us much further.
Could we talk about a replication crisis or reproducibility crisis in literary studies? It might be rather fun as a way of annoying people, but, a couple of would-be objective approaches to textual analysis from the 20th century aside, the fact that different people have different interpretations of and responses to works of literature is generally seen as a feature rather than a bug. It’s interesting, therefore, that the idea seems superficially more plausible in the case of history; in part, one imagines, because the past appears to be A Thing Of Objective Existence that we ought therefore to be able to analyse objectively and reach the same conclusions about, and in part because of the legacy of the mid-19th-century Geschichte Als Wissenschaft crowd and their drive to make historiography as science-ish as possible.
Fine, Howes might respond; granted that the problems of historical knowledge are not in fact problems of replication in a strict sense, is it not still a problem that two historians can look at the same lot of evidence and reach different conclusions? Well, if you think historiography is, or ought to be, ‘scientific’ in a reasonably robust way, then, yes, the extent to which the whole thing is based on subjective judgements about the selection, weighting and interpretation of different bits of evidence, the rhetorical and literary representation of that evidence and the historian’s understanding of it, and the place of this activity within a set of wider, overlapping discourses makes it look distinctly unscientific.
It’s a problem, if you have such an unrealistic idea of history in the first place. It’s not a new problem; it’s all there in Hayden White’s ‘The Burden of History’ essay from nearly seventy years ago, and most of it can be found in Nietzsche’s ‘Zum Nutzen und Nachteil der Geschichte für das Leben’ and other mid-19th-century discussions (my old friend Wilhelm Roscher’s book on Thucydides shows a powerful awareness of the subjective and discursive nature of historiography even as he’s desperately trying to bring it up to the status of a science).
From that perspective, the issue is not the fact that historians spend their whole time arguing over the interpretation of evidence and the reconstruction of the past – it’s basically the whole point of the enterprise – but rather the gap between this professional understanding and what (parts of) a wider public believes about and expects from history. It’s there in the (perhaps disingenuous, perhaps just stupid) claims of people like Restore Trust or the opponents of the 1619 Project that the development of new interpretations and more rounded, socially-conscious accounts is somehow Destroying The Past (rather than enriching our understanding of it).
There is a desire for a single version, so long as it’s the right version (‘right’, for values of cultural conservatism and invented tradition). Questioning the ‘scientific’ status of historical knowledge is potentially a threat to this project, insofar as it implies the existence of multiple perspectives on the past rather than a single shared narrative – but it is also potentially an opportunity, as a basis for rejecting any new interpretations (since these are self-confessedly subjective and provisional) in favour of the traditional story. In other words, the idea of a ‘replication crisis’ in history might be deployed as a means of condemning its failure to lend wholehearted support to national and cultural myths. How dare Thucydides question the Athenians’ beliefs about their own past? It can only be because he’s a disaffected opponent of the established order…
Quite a lot of Howes’ piece falls firmly within a Thucydidean approach; he’s concerned with the persistence of various ‘myths’ of technological development despite the absence of concrete evidence when the issue is actually investigated properly (so in fact the reproducibility problem is about there being too much reproduction rather than not enough…). Historians are sometimes lazy and cut corners; peer reviewers really don’t have time to check every single point; speaking as an obsessive pedant who is completely paranoid about the possibility of getting anything wrong without realising it, I totally get this, but I don’t think it amounts to an epistemic or methodological crisis. Nobody in history thinks about it in these terms, as Howes notes – perhaps because they are not the right terms in which to think about it?
There are Left critics of the 1619 Project, to be fair; in fact, the main critique I’ve seen is precisely that it represents an attempt to impose a (new) single version of history, overriding cross-cutting histories of America as seen through lenses of class, empire, democracy, republicanism etc. Restore Trust… no.