It is a remarkable fact that the truth or falsity of a judgment about what color some particular object or surface is is and is generally recognized to be independent of how that color happens to look to the particular person who utters that judgment. Someone looking at an object, who suffers from some as yet unrecognized defect of color vision, or who is looking in visually unfavorable circumstances, may have her or his false judgment about its color corrected by someone blind who has been told what color it is by a reliable informant.

Wittgenstein appears to deny this at one point in the Remarks on Colour, when he not only asserted “That it seems so (so scheint) to human beings is their criterion for its being so” , but added that only in exceptional cases might being and seeming be independent of one another. If Wittgenstein meant by this no more than that it is a necessary condition of our color judgments being as they are that, for example, “we call brown the table which under certain circumstances appears brown to the normal-sighted”, then it would be difficult to disagree. But the use of the word “criterion,” as I shall suggest later, is misleading. For the recognition of a color is not generally the application of a test.