Palmer was puzzled by the nature of Linear B. If, as Evans had believed, the samples were from 1400 bce and the samples from Pylos were from 1200 bce, and the language was the same, he wondered how was this possible, because the nature of all linguistic model involves change. Those linguists who were pro-Evans talked about it being a scribal language and that's it, but Palmer, unconvinced, consulted the daybooks of Duncan MacKenzie, Evan's right-hand man on the digs, and Palmer sussed out that Evans had consciously or unconsciously, fudged data.

Whichever way you slice it, it's not an attractive picture. While bias rules us all, to some degree or another, a scientist should be vigilant against it. I prefer to think that Evans did it unconsciously, so propelled was he by his passion for his Minoan vision, but that makes it hardly less wrong.

I also found it interesting that he refused to publish his Linear B samples, so that the language wasn't deciphered until Carl Blegen published his findings from Pylos, and Michael Ventris got hold of it.