The Batman fallacy

This is a historians’ problem which has bugged me for some time, but I’ve never had a name for it. Now I do, courtesy of my ten-year-old son.

‘You loved Batman as a boy,’ he declared confidently to me over the weekend. His evidence: a photo of me which hangs in our hallway, aged about eight, wearing a Batman T-shirt.

Now I happen to know that that’s not the case. I vividly remember my enthusiasms at that age: Star Wars was peaking, the now-forgotten Micronauts were putting in a respectable showing and Lego was just beginning to register. Superheroes of any kind: meh. I was willing to wear the T-shirt. (Or do I only believe I remember these things? - We all know contemporary documents, like the photo, are more reliable than later recollections …)

But it was a perfectly sensible hypothesis for him to make based on the very limited and fragmentary evidence he had to draw on. And historians do this all the time. All we have are a few fragments, chance survivals. The temptation to assume that they are keys to understanding everything is very strong. We can formulate hypotheses from them which are both plausible and legitimate. At least they seem legitimate.

Since Karl Popper we’ve measured the legitimacy of a hypothesis by whether it is falsifiable. If not – if there is no test which could be devised which could in principle prove it wrong – then it is not a scientific hypothesis, but something else. That works fine in the experimental sciences. But in observational sciences like history (or, say, palaeontology), there is a grey area. Some hypotheses – lots of hypotheses in fact – are falsifiable in principle, but not in practice. That is, you can imagine the evidence which might allow us to test it. But we don’t have it, we know we don’t have it, and we are pretty sure we’re never going to have it. In which case, what you have is not really a hypothesis. It is a speculation, or a generalisation. It is, in fact, a sand-sculpture: perhaps very attractive, but not something you'd be wise to build on, or indeed something that's likely still to be there once the tide has come in.

Now historians need to speculate, we all do it, and there’s nothing wrong with that. That’s not the Batman fallacy. The Batman fallacy is when we imagine that because a speculation is compatible with all the evidence we have it is therefore likely or even proven. In short, we forget how much we do not know.

It is hard to emphasise this enough. For most of the human past, we know almost nothing. Even for my own period, the 16^th and 17^th centuries, huge swathes of ordinary life are simply mysterious to us. For earlier or less well-documented periods the problem increases exponentially. We tend to skate over or conceal this ignorance: books stating baldly that we know nothing and that there is no evidence are short and do not sell well. Textbooks particularly, which are required to give an overview, do so by giving an illusion of knowledge. Students tend only slowly to realise that their own persistent ignorance about the past is not a mark of their own stupidity, but the historian’s condition.

No scholarly field that I know is more vulnerable to this problem than Biblical study. There a body of evidence which is both tiny and enormously unbalanced meets a huge enterprise of focused scholarly attention. Genuinely new evidence does appear, but it is pretty rare. So the danger of overinterpretation is everpresent. Naturally scholars make plausible and often ingenious guesses about the authorship, redaction history, contexts, social meanings, cultural assumptions and so forth surrounding the Biblical texts. Which is wonderful. The danger is that they start to believe that these hypotheses are established and proven. There are, in fact, very few hypotheses for which we have enough evidence for us to be able to establish them. Most of the Bible's historical context is simply lost to us, and if there is one thing we can be sure of, it is that any substantial new evidence about it would contain surprises.

Genuinely substantial evidence is of course unlikely. So in the meantime we read the texts and make the best of what we have. But we need to remember that we are doing the equivalent of deducing a boy’s life and enthusiasms from a single snapshot.