[D. Bergin]

Originally Posted by sphere and ash

I have spent over 25 years collecting photographs, learning in the darkroom, and training in nineteenth century photographic processes. None of that, as the original poster correctly points out, gives me any advantage in facial recognition. But—it does give me some me some perspective. Finding a photograph that resembles another person is a commonplace. Take, for example, the two albumen photographs shown below. Couldn’t almost all of us agree on their identification?

I’m not trying to poke fun in any way—I’m just trying to impart some very hard-won learning. Within the last two weeks, the subject of a very valuable photograph that I own was shown to be someone else entirely—and this was a baseball photograph of the team founded by the subject in question.

Before I forget: for those wishing a good reference book on dating stereoviews, I recommend the work of William C. Darrah. I would consult my own copy now, but everything I own is in storage pending a house sale.

Also: someone above made the point that nearly all discussion participants agreed that the ‘Knickerbocker stereoview’ attribution was questionable. Although my own skepticism was correctly included on that list, I do not believe popularity is a good measure of the truth of an idea.

And: I was wrong to say in my first post that probabilities have margins of error. I should have said ‘Estimates have margins of error.’