Connoisseurship still plays a role. It's much denigrated and criticized, but ultimately, without connoisseurship, we really wouldn't know Leonardo's work at all. It's still a fundamental tool in establishing what was done by him and when it was done, since none of it is signed, none of it is dated, and, apart from "The Last Supper," nothing has a continuous provenance. So you still have a lot of that rather old-fashioned judgment by eye to do.

So, in the flesh, you look at it. It's on vellum, and you can see the extent to which the surface is deteriorated, which you can't see, really, in a digital file, which smooths out the surface. You can begin to see where it's been restored — as you look at it in different light and from different angles, the physicality of it becomes apparent. But that's only your starting point. Then, all the heavy-duty research comes in, and we now have, of course, an enormous body of extra things we can look at. So the initial connoisseur's reaction merely tells you that something is worth looking at, but at any point one wrong thing can throw that all away — a later pigment, a bit of something that might come up about its history to indicate it was forged at some point, and so on. I was trained as a scientist, and if you have a scientific theory, you only need have one bit of the experiment that says, "this is not right," and the whole thing collapses. You always have to be looking for that one thing that is going to demolish the whole expectation that's being set up.

I would not now probably say much about it at all, because on reflection I think we don't have an adequate reference bank of Leonardo fingerprints. I've talked to fingerprint specialists, and they typically require a full set of reference prints. We don't have that for Leonardo. My sense is — and this is Pascal's sense, too — that it's probably premature, given what we know about Leonardo's fingerprints, to come up with matches at all.