Hume supported the claim by appealing to the fact that people who lack access to a particular realm of sense experience, e.g. sight in the case of the blind or sound in the case of the deaf, are incapable of forming the corresponding ideas—they are unable to form the visual and auditory images familiar to those able to see and hear.
However, as it stands, the claim cannot be accepted. Hume himself pointed out that he could form the idea of a golden mountain without ever having come across one, in which case there are ideas that are not copies of prior impressions. But Hume went on to argue that his ability to form this idea does at least depend on his having had an impression of things that are golden and things that are mountains. This suggests that a weaker thesis might be acceptable: perhaps every idea that we have is formed by
‘compounding, transposing, augmenting, or diminishing’ ideas which are copies of previously experienced impressions. Even this may not be quite right, for as Hume points out, a person may be capable of imagining a specific shade of colour that he has never previously experienced, but Hume considers this exception too minor to challenge the general thesis. An alternative response to the challenge apparently posed by such cases would involve arguing that they are not really counter-examples to the principle, provided that the terms ‘augmenting’ and ‘diminishing’ are understood in a sufficiently broad sense. For presumably the image of an unencountered colour shade can always be derived from a previously encountered one by augmenting or diminishing one or more qualities of the latter (e.g. its degree of reddishness). So understood, the principle would allow a great deal of extrapolation from the material of experience. What it would disallow is an ability to imagine impressions that are of a radically different sort from those actually encountered at any point in the subject’s life. The thesis is really an empirical one—it must itself be judged by experiential evidence. As a general statement applicable to all human beings at all times and places, it could arguably never be wholly certain, since however often it was confirmed in our experience, there would always be the logical possibility of finding a counter-instance. On the other hand, the claim does have enough empirical support to make it highly plausible. In other words, our evidence strongly supports the claim that ordinary human beings are unable to form images of a radically different kind from those that they have encountered within the course of their actual experience.
It has been suggested that the thesis is untenable, since if all my ideas derive from my sense experience and all your ideas derive from your sense experience, it follows that we can never share the same idea, as we cannot have the same sense experience, and this is an absurd conclusion. However, the conclusion does not in fact follow. For consider a situation where I have an impression of