maandag 4 februari 2013
The Context Dogma
Context is everything. That neatly sums up the problem. You cannot explain anything by an appeal to everything. To be sure, everything has a context; but to make context an all-encompassing notion robs it of all explanatory power. The principle known as "Ockham's razor" teaches us not make any more assumptions than are needed to prove your thesis, and in particular not to "multiply unnecessary fictional entities". But contextualism, in its current shape, massively overgenerates entities to be taken into account - namely, all that is part of the "context" - and in particular creates one immense fictional object - "The Context". Contextualism, in short, has become The Context Dogma.
This is no plea to chuck the context overboard. [For one thing, I agree with most of the criticism against Jonathan Israel.] But it is crucial to realize where the notion of a context ceases to be a tool for thinking and instead becomes an impediment to reasoning. "The context" is generally used as an argument against theoretical abstractions and schematic simplifications. This argument invokes a set of easily overlooked counter-examples or the occurence of many hybrids, oddities and exceptions to the rule at every point in history, showing that the past as it is contained in direct source evidence is infinitely much more complicated than one would like to believe from reading histories. That argument makes sense. Every generalization can be falsified, and with such complicated topics as human choice and action, of which we never have more than partial record and evidence, the lack of counter-examples would be the greatest anomaly of all.
But that very same argument in turn shuns falsification. There is simply no end to the set of possible anomalies and subcontexts. Thus there is nothing that could prove the contextualist wrong when he invokes "the essential contingency of historical facts". And that, in turn, turns contextualism trivial.
Contextualism becomes a particularly pedantic device when it is used to restrict historical enquiry to the antiquarian investigation of ever more detailed case studies. As research practice currently is, individual researchers end up knowing more and more about less and less. This creates a growing gap between "textbook history" [what is taught] and "real history" [what is investigated], to say nothing of the supposed trickle-down effect. The Context becomes a scary big lump of knowns and unknowns to knock down all the uninitiated who venture upon a certain specialism with.
Anthony Grafton, by no means a pedant, once infavourably compared "parachute-throwing" historical generalizations to the more modest "truffle-seeking" for the telling historical rarity, the luminous detail. But while Grafton's brilliant truffle searches changed the perception of what early modern scholarship was like, and provide one of the most examplary models for how to do intellectual history, they do so by leaving a great deal of generalization implicit. Grafton's work is in fact strewed with salty anachronisms and historical parallelisms that make his descriptions all the more vivid. By showing how the past differed from the present in unexpected ways, he gives all the more rise to cross-temporal analysis that cannot escape generalizations when made explicit.
"Ideas have agents", Quentin Skinner wrote in Meaning & Context, summing up his own philosophy in the most succint possible way. That is very true. But ideas also have content. The very hack of language is that it can be decontextualized - if we couldn't use words in all kinds of possible situations, they would be useless and incomprehensible. Of course, decontextualization doesn't always come cheap. The second wonderful thing about language is that we can say more than we actually say - the actual meaning is revealed only in implications and presuppositions that fan out in all directions, involving our knowledge of the world at large, that we can't possibly oversee at the moment of speaking. That predicament is the same for us as it was for John Locke or Julius Caesar. And it is the same for our deeds.
The nasty thing about the past is that you can't go back and check. I have more than once described my own (novice) historical practice as "journalism for the past" - describing the dead and what they were up to in terms that they would not necessarily agree with, just as I would do discussing contemporary art or politics. But in describing the past, that is a violation of the right of reply.
In the end, as historians, we are all both parachute-throwers and truffle-seekers. Few of us are nerdy enough to be content with just that rare antiquarian find; most of us are still driven by a more comprehensive, somewhat foolhardy and outdated pursuit of general erudition. But we all know how to appreciate the devil in the detail.
What we call the context is often a handful of examples. If it gets beyond the handful, it becomes a matter for statistics rather than deduction. It is precisely because you never know what you're gonna get and what more truffles may be hidden in the mud that good historians should stay hungry - that they should, like good philosophers, want to know everything. That sounds like a plea for contextualism, but in its pedantic version, contextualism hinders itself in restricting this desire. Of course, good historians should also know how to pick and choose and when to stop. That, in effect, is Ockham's razor at work: investigate everything, but keep your argument brief.
The Context Dogma is a dogma because it entails an idealization: call it, in Ranke's words, jede Epoche gleich zu Gott or call it, in more modern terms, actor's categories. The idealization is in constructing a "context" more coherent than past actors can have experienced it. This is not an excess of contextualism; it is inherent in using "context" as an explanatory notion. There is, as Ernest Gellner argued, no "as the Romans do". Agency, then as now, meant acting upon insufficient information, educated guesswork, and occasional creativity.
Donald Davidson argues that we should apply a "principle of charity" in making sense of what others do and say: without attributing to them a logically coherent set of largely true beliefs, it is impossible to construct a coherent interpretation. That, however, concerns the actor. Even so, the principle of charity is an idealization [Davidson calls it a "constitutive ideal of rationality"]; but we can easily picture a historical actor quite as smart as we not knowing what he's up to. The very complexity of the historical situation that is a basic tenet of contextualism contradicts the explanatory use of "context".
Is there a counterposition? Noncontextualism might be a plausible position in semantics, but in historiography, it is not an option. That is like saying: "let's take the past for granted as we know it". Still, it is important to realize that historiography is argumentative and that one consequently has to pick one's arguments judiciously. This means, ultimately, that historians should work towards developing a historically informed position rather than just "doing history". It goes without saying that such a position should not instrumentalize the past; that kind of noncontextualism would make the historical argument rather lame, as there is no surplus value in an appeal to an instrumentalized past. The past is indeed a mess, but one can still pick truffles from it and present them with the mud still on them. One could call this contextualism without the Context Dogma. But without the dogma, it is not an -ism.