From the ALWS archives: A selection of papers from the International Wittgenstein Symposia in Kirchberg am Wechsel

1. Ascriptions of Taste

I shall defend the view that the sentences of ordinary languages which ascribe tastes to ordinary physical objects, most notably to edible bits of the world, are truth bearers. This is to say that the statements we may wish to make in order to promote the view that such items have a taste, either general like "Strawberries are sweet", or particular like "This strawberry is sour", and the utterances resulting from the fulfillment of that wish on a particular occasion, may be true. To claim that the sentences and the related statements and utterances of a particular kind, or belonging to a particular class, are truth bearers, is to say that they are in the market for truth, that there is a truth predicate for the language, or fragment of langage, to which they belong, and that they may be assessed by us in terms of truth.

If such sentences may be true, the question arises as to whether they may be so whether or not we are able to tell that they are true, or, on the contrary, whether we must always be able to tell that they are true whenever this is indeed the case. I shall also address this question here, but I shall first restrict myself to an argument to the effect that they may be true simpliciter.

For the sake of convenience and conciseness, let " " be the modal sentential operator "It is possible that", let "x" range over ordinary physical objects of which taste may be genuinely predicated and let "T" range over tastes (let "T" take sweetness, sourness and the like as values) ; finally, until the argument is spelled out, let "Tr" be a putative truth predicate for any ordinary language in which ascriptions of taste may be framed and expressed.

Given these notational specifications, the view I wish to defend may be symbolically noted thus :

• (1) (Tr [Tx]).

There are two complementary ways of arguing for particular instances of this schema : by showing that particular ascriptions of taste have truth makers and by showing that they bear the marks of truth.

With respect to the first point, I shall argue that Lockean dispositions (Locke [1690] 1975) are the truth makers of ascriptions of taste. With respect to the second, I shall argue that these ascriptions bear the marks of truth listed by David Wiggins (Wiggins 1980).

2. Dispositions as Truth Makers

According to the dispositional analysis of secondary qualities inherited from Locke, a judgement to the effect that an object possesses a secondary quality must be justified by means which are based on the experience of an appearance. This doesn't mean or imply that we must, or let alone can, justify in this way the attribution of every perceptible quality to an object. Although primary qualities do indeed have an appearance, our experience of that appearance may not be part of our justification for judging that an object possesses these qualities. Our experience of secondary qualities, on the other hand, must be part of our justification for judging that an object possesses them. This is what distinguishes primary qualities like shape from secondary qualities like taste, with respect to the justifiability of our ascriptions.

Our sense of sight, touch and hearing may help us to determine the size and shape of an object. For instance, we may judge correctly that an orange is round and small by looking at it or by grasping it with our hands, or we may guess correctly that a bell is big by listening to the sound it makes when struck. But in each of these cases, there are experience-independent criteria for whether the primary qualities of shape and size are instantiated in those objects, or, on the contrary, merely seem instantiated, whereas the distinction between the real and the apparent instantiation of a secondary quality, e.g. between the real and the apparent sweetness in an orange, must be drawn from within the realm of experience.

Measurement is, quite typically, the sort of experience-independent criterion which justifies the ascription of primary qualities like shape and size. According to the dispositional analysis, we lack such a criterion, or type of criterion, in the case of taste. For an orange to have a given taste, or to taste a certain way, is for it to have exactly that taste, or to taste just that way to someone who happens to eat it or to put a slice of it in one's mouth without actually consuming it ; but for an orange to be round or solid - i.e. to have one of those qualities identified by Locke as primary - is not to seem to have precisely that shape and solidity to whoever happens to be touching it or looking at it, or experiencing it with the help of the relevant sensory equipment.

This distinction between the ways in which the real and the merely apparent occurrence of qualities is to be drawn is crucial. It isn't merely that, according to the tradition inherited from Locke, the instantiation of a secondary quality in an object consists in a disposition of the object to produce sensory experiences of a certain phenomenological character in us, whereas the instantiation of a primary quality consists in an intrinsic feature of the object. What counts here, specifically, is that the dispositional analysis prevents us from drawing a distinction between sweetness as it appears and sweetness as detected by the application of an experience-independent criterion, i.e. by non-phenomenological means. There is just no such thing, the dispositional analysis goes, such as objective or intrinsic sweetness in that sense.

This does not prevent us, though, from drawing a distinction between genuine sweetness and apparent sweetness from within the realm of appearance. As a matter of fact, this is just how the distinction must be drawn if it is to be drawn at all, and the normal way to do this is by accounting for the distinction by reference to those experiences of sweetness which we may reasonably judge to be standard.

Now standard sweetness, or sweetness as it standardly appears, as opposed to an individual's or a community's illusory sweetness, need not necessarily be true to anyone's phenomenology of sweetness. This should not worry us. After all, it is possible for the analysans of a dispositional analysis not to be true to the phenomenology of a nevertheless correctly analysed secondary quality. For instance, the dispositional analysis tells us that sweetness is a relational property. It claims that an object is sweet just in case it tastes sweet to some perceiver. The phenomenology of our perception tells us otherwise, for when, say, a fruit tastes sweet, it appears to us as having a monadic property, not a relational one.

It is therefore possible for a secondary quality like taste to be analysed dispositionally in terms of a condition which must be fulfilled (e.g. the phenomenological condition of seeming sweet, or of seeming sour) without it following that, if some object falls under the concept of the quality, the condition must necessarily hold. And if that is the case, then the distinction we need between the genuine and the illusory occurrence of the quality may be drawn within the limits of the realm of appearance, e.g., the realm which we take to be constitutive of the quality of sweetness. So we are perfectly allowed, without thereby giving up the dispositional analysis, to say that someone could suffer an illusion with respect to sweetness because, say, an orange which would standardly taste sweet to anyone under standard conditions, does not seem sweet to him.

This undermines, at least indirectly, via the additional explicit consideration that what standardly tastes sweet constitutes a bona fide norm for sweetness - a type of argument which is not unfrequently used to show (allegedly) that ascriptions of taste may not be true at all, that all instances of schema (1) are false. To distinguish it from arguments which do or might take into consideration intersubjective agreements with respect to matters of taste, we may call it the "pure subjectivity argument" . It is the argument to the effect that ascriptions of taste may not be true because they do nothing more than register our individual responses to the causal ground of taste.

This objection to (1) is undermined by a feature of the phenomenology of taste which accounts for the distinction between the genuine secondary quality of taste (or the genuine occurrence of that quality) and the merely illusory secondary quality of taste (or the merely illusory occurrence of that quality). We know that if our experience of a given taste is such and such, then the experienced object tastes such and such. It isn't the causal ground of taste, constituted, among other relevant elements, by the anatomy of our tongue and its taste buds, which make ascriptions of taste true. Once we have accepted the dispositional analysis, we have of course precluded the possibility of reducing secondary qualities to their causal ground. But although no such reduction is to be hoped for, the phenomenology of taste is on a par with the causal ground of taste in that it imposes restrictions.

Just as some features of the causal ground of taste make it impossible for some tastes represented in Hans Henning's taste tetrahedron (with pure primary tastes at the apex of each angle and combinations of any two tastes as points on one of the edges of the figure) to physically blend, the standard experience of taste make it impossible for edible bits of the world to appear to instantiate certain combinations. Standard experiential facts about taste fix key features of the appearance, or of the disposition, and they do so not only in the sense that they are constitutive of the quality, but also in the sense that they fix limits to how things may seem to a taster. It is only this corrected manifest standard appearance which makes ascriptions such as "Strawberries are sweet" or "This strawberry is sour" true. So it isn't merely that the occurrence of a given taste, or of a combination of tastes in an edible object is constituted by how things sensorily seem to the taster. How things standardly seem to tasters is subject to phenomenological laws precluding phenomenological incompatibilities, i.e. incompatibilities with respect to the appearance or disposition which constitutes the quality.

3. Marks of truth

Standard dispositions à la Locke make ascriptions of taste true. Conversely, these ascriptions bear the marks of truth listed by Wiggins (1980, 205-213) because corrected appearences determine the content of particular tastes.

Here is Wiggins's complete list :

• MARK 1. The assertibility (and therefore the truth) of a sentence is the primary dimension of its assessment.
• MARK 2. A sentence is assertible (and is therefore true) if it has a content such that, with respect to that content, there should be a tendency, under favourable conditions of investigation, for disagreement to diminish and for opinion to converge in agreement.
• MARK 3. (i) Every sentence which lacks the property of assertibility (and therefore of truth) lacks it independently of a speaker's means of recognizing it. (ii) Every sentence which possesses the property of assertibility (and therefore of truth) possesses it independently of a speaker's means of recognizing it.
• MARK 4 [CONDENSATION OF MARKS 1-3]. Every assertible sentence is assertible in virtue of something. Consequently, every true sentence is true in virtue of something.
• MARK 5. Every assertible sentence s is co-assertible with every other assertible sentence s'. Consequently, if s is true and s' is also true, then so is (s o s').

Why do ascriptions of taste bear these marks? Because of the restrictions imposed on the form of our perceptual experience with respect to taste. Although there are telling differences from person to person with respect to abilities to distinguish flavours, tastes must appear a certain way. In his Remarks on Colour, Wittgenstein talked of laws of appearence as truths about how things can look. Much of what he says about, e.g., the blending of colours may, I think, be said about the blending of tastes : there are laws which make some blendings impossible. As Wittgenstein remarked, "blending in white removes the colouredness from the colour" (Wittgenstein 1977, II, 3). Similarly, combinations of salty, sweet, bitter and sour float inside Henning's tetrahedron so that such combinations may not be properly located, and from the dispositionalist's point of view, no edible object may appear to instantiate this particular combination. So there are favourable conditions of investigations with respect to ascriptions of taste, namely those determined by how tastes must be experienced. This directly implies that ascriptions of taste bear MARK 1. Although there may be a tendency towards convergence in this respect (MARK 2), there is no reason to believe that we must be able to recognize the truth or falsity of every ascription of this kind (MARK 3). It remains to be investigated under which conditions these ascriptions could bear MARK 5, but this is the occasion of another investigation into the logic of these statements.