Posted by Alan Nagel on February 17, 1999 at 10:05:38:
In Reply to: How We Know What We Know? posted by R John Howe on February 16, 1999 at 18:29:22:
: Dear folks -
....
: Could we be involved in an inference that goes something like this? Rugs of a certain color or shade are found more frequently in rugs found in geographic area X. Minor premise: colors found frequently in such rugs are likely produced with local dyes. Therefore it is permissible to infer that rugs with these colors or shades have likely been woven in area X.
: Now all of this may be perfectly sound but as stated it is vaguely unsatisfying. It seems to me possibly to qualify as what Pinner describes as a "convention." We feel better about such conventions if they at least occasionally put a foot down into the factual world.
: R. John Howe
John Howe nicely focuses a logic point it can't hurt us to get at ease with.
1) the color argument is probabilistic, and perhaps sometimes we can determine that one particular dye has a remarkably high frequency association with geographical area x.
2) but there's no certainty to be attained there
3) My question: is there then any other kind of information that can supplement the probability of dye/area associations than connoisseurship? We can challenge the experts with careful research on the probabilities, using certain or gen'ly agreed upon examples. But we have to turn to them when our data have been managed as best we can.
Right on target, then, that John asks about Bohmer, and Wendell cites correspondance from H. Keshishian?
Alan