The problem is in the probabilities. People usually unconsciously take this into account when discussing what appear to be very basic judgements.
There is only one man (Chadwick) that this could be that would lead to the conclusion "this is Chadwick". On the other hand, there could have been ~15 million white men in the US at the time that could lead to the conclusion "this is not Chadwick". In making the visual comparison and trying to decide between same guy and different guy, it is almost impossible to not have the effect of probability cloud the comparison.
Suppose you were given the picture of Chadwick and the tintype, told the Chadwick photo was Chadwick, and then told that the tintype was either Chadwick or one - only one - other unknown white man. You are also told that the other man was randomly drawn and not intentionally picked for his similar appearance.
Knowing nothing else, you would probably conclude that the tintype is surely Chadwick despite any small differences from the photo, because what are the chances that a randomly drawn second man would look more similar to the tintype than the Chadwick photo?
So it is very hard to purely objectively try to evaluate the similarity of the images without having that sense of probability be part of the mental equation. It's fine to consider the probability in your final determination of whether it is or is not Chadwick, but that's different than using it in your evaluation of pure photo similarity.
Personally, I think it is dead on regardless of the little photographic bobbles that may skew the similarity. But I also think it is not Chadwick because when I factor in the probabilities, now all those little bobbles loom large.
J
|