As is well known, prawn cocktail flavour crisps are manufactured without the involvement of the humble prawn. By contrast, a prawn cocktail contains prawns. (We can leave open for present purposes whether a prawn cocktail is a cocktail.) The flavour that you taste when you taste prawn cocktail flavour crisps is, therefore, (in at least one sense) not the flavour of a prawn cocktail. The crisps are prawn cocktail flavour, not flavoured. But are the flavours of crisp and cocktail the same? Could they be? And could one learn what prawn cocktail tastes like, or what it’s like to taste it, through tasting only prawn cocktail flavour crisps?
These questions are akin to the question whether identical twins—individuals that we can assume look the same—have the same look. They look the same. But is there, in addition, a look that both has? Or does each, instead, have its own look? And can one learn what one of them looks like, or what it’s like to look at them, by looking only at the other?
One set of questions here is broadly metaphysical. Does it follow from the fact that the flavours of distinct substances are qualitatively the same that the flavours themselves are the same? And supposing that we have reason to allow that crisps and cocktail have the same flavour, does it follow that there are not, in addition, distinct instances of flavour in the cocktail and the crisps?
Another set of questions is broadly epistemological. Under what conditions would one be in a position to recognize that crisps and cocktail have the same flavour? Is it enough that one cannot tell presented instances apart in a blind tasting? Or must one in addition compare those instances with a wider range of flavoured substances? And must one also consider the effects on flavour of mixing cocktail and crisps with other foodstuffs? Finally, should one consider, in addition, the responses of other tasters, including those with greater experience or delicacy of palate? And can it even be required that one cannot tell presented instances apart, given that the different textures of normal instances of crisps and cocktail are liable to serve as a confound by making it easy to tell instances apart? Finally, just as we agree that the twins are distinct even though we cannot tell them apart by looking, might we be willing to allow that the flavours of crisps and cocktails differ, despite our inability in principle to tell them apart on the basis of how they taste?