Academic Overgeneralisation

The second problem (the first problem is here) that I have with English-language studies of Shinto is that they tend to overgeneralise from a limited range of evidence. This problem is certainly not limited to Shinto studies — I read a very interesting book on the “immigrant experience in Japan” that was based on a handful of group sessions with a few dozen people, recruited through about three different groups. My personal experience of the immigrant experience in Japan is broader and deeper than that. However, it is particularly relevant to Shinto. To steal the phrase used in lots of different contexts, if you know one jinja, you know one jinja. Or, to be even more careful, if you know one jinja in one fifty-year period, you know one jinja in one fifty-year period.

In this case, the research itself does not annoy me at all. Quite the reverse: I always find it very interesting, and, so far, I have always learned something new. Unless someone decides to publish about a jinja I know well, that is going to continue.

The problem is that the books and articles almost invariably claim to be shedding light on the general situation of Shinto, and I’m almost always unconvinced that they do so. One study tells us a lot about one situation, and that’s very interesting, but there is no guarantee that things will be similar in the jinja down the road, never mind in jinja at the other end of the country fifty (or five hundred) years earlier or later.

I mentioned last time that this problem may not be practically soluble, and indeed may be best left unsolved. First, why might it be impossible to solve it in practice? Academia is very reluctant to fund people to spend thirty years researching Shinto to write one slim book, but also does not much like books that tell us a lot about one jinja in one city. It wants works that produce sweeping theories of the field, and which can be produced in a couple of years in and between giving 30 hours of lectures a week. (I exaggerate. Slightly.) This means that, in order to get published and, as a result, not get fired, researchers have to make exaggerated claims for the broad significance of their work. This is true in every field of academia I have any familiarity with. I don’t think this is a good thing, but it can only be changed by changing the whole culture of academia — and I doubt that will happen soon.

Further, it might be better not to solve it. One way, and possibly the main way, to advance understanding of a field is to make claims that go way beyond your evidence, and provoke other people into testing those claims against a wider range of cases than you could possibly investigate by yourself. That is what academic studies should be doing. Ideally, in this case you should write things like “If we assume, without any good grounds, that the findings from this one jinja apply to all 80,000 jinja in Japan, we can say that…”, but academic culture does not let you write that. Readers have to figure it out for themselves.

(I was not fond of a number of academic practices back when I was an academic. Distance has not further endeared them to me.)

So, this problem is my problem with the culture of English-language academia, not with the individual researchers or the individual research. (It should be said that the Japanese academic texts on Shinto that I have read also have this problem. It may be endemic to global academia.) If I just treat the general claims as conjectures based on the specific evidence recounted, I will, I think, be able to enjoy them as thought-provoking ideas.