It is surprising that Hume offers a counter example almost immediately after saying that this is all that is required to refute his claim. Some scholars have said that this is precisely the kind of example that should have led Hume to reject his copy principle. A more generous interpretation is that Hume is knowingly modifying his general claim and it should now be understood as saying, ‘In the normal course of events...’
A counter example isn’t necessarily more damaging to Hume’s epistemology than a counter example is to Newton’s laws of motion—his epistemology might be perfectly adequate within certain limits.
Throughout this section Hume assumes that colours are simple impressions and simple ideas. This can be challenged and if a colour can be complex due to it being possible to analyse it down into hue, tint, tone and shade, or even being able to construct them from a mixture of red, green and blue, then the ability to imagine new colours becomes less problematic.
Hume assumes everyone will agree with this but it isn’t obviously true. If I see two similar colours at the same time I may well form two ideas that are distinguishable but if I saw those two colours separately and then later tried to recall them I may have indistinguishable ideas and, as a result, mistakenly think the two colours I had seen were the same.
The problem here is the same as with different colours.
If our perception of a colour depends not just on the colour itself but also on the context within which that colour is seen then it is no longer possible to conclude that just because two adjacent colours are indistinguishable the ends of the gradient will be indistinguishable.
For the reductio argument to work you would have to assume that indistinguishable impressions and the corresponding indistinguishable ideas were, nevertheless, different. It isn’t clear that there can be ideas that are both indistinguishable and different.
Hume is also assuming that our idea of a gradient is the same as perceiving a series of graduated steps and this isn’t necessarily so.
It is clear that for Hume there is nothing special about blue. Since he also mentioned sounds 'that are conveyed by the ear' it is reasonable to suppose that he might have also chosen a graduated series of sounds, and presumably other graduated series as well.
Since such cases are either very rare or nonexistent it is difficult to be sure that we can accurately determine how they would perceive the world or, more precisely, whether they would form the relevant ideas in the same way that we do.
Hume is assuming that a gradient is a series of discrete steps and not a continuum. However, this doesn’t seem to be a serious problem with the counter example.
This isn’t obviously true if the steps in the gradient are very small. The difference between steps may be so small that not only are adjoining steps indistinguishable but when one is removed the newly adjoining steps remain indistinguishable. Again, this may not be fatal to the thought experiment. It would be possible to specify a sequence of steps such that normally experienced people can perceive a gap and make the claim that the experience deficient person would also recognise the gap.
This seems less credible. The normally experienced person finds it difficult to generate an idea of an intermediate colour so this doesn’t give strong grounds for assuming that the experience deficient person will be able to generate such an idea. To verify this claim requires an experiment and one for which it will be difficult to find an appropriate subject.
Clearly Hume is using blue as just an example and it is important to avoid the naive criticism that Hume is unaware that there might be a missing shade of green or a missing note on a scale, etc. The singularity presumably refers to the type of example, i.e. the tightly ordered sequences. Perhaps the challenge that Hume should have offered is, ‘If anyone thinks that this kind of example does undermine the claim that in the normal course of events...then all they need to do is explain why that is so.'