A good way into Johnston’s article is to first get clear on the distinctions between three views about the metaphysics of value (or, more generally, about the metaphysics of an arbitrary sensible property). After getting clear on this, and seeing what Johnston’s arguments establish on the question of the metaphysics of value, we can turn to the consequences of this for our view of the content of perception. Roughly:
Projectivism about some property F is the view that nothing is F, and that when we judge that things are F, we are mistakenly projecting some property of our sensations onto the world. Thus projectivism about color is the view that nothing is colored, and that when we judge that things are colored we are ascribing mistakenly properties, which are really properties of our own sensations, to things in the external world.
Dispositionalism about some property F is the view that (in contrast to projectivism) some things are F, but that what it is for something to be F is for it to have the disposition to cause certain kinds of sensations or, more generally, responses, in us. For example, let red* be the property which all visual experiences which are (as we might put it) ‘as of something red’ have in common. Then the dispositionalist about redness might say that what it is for something to be red is for it to be disposed to produce red* sensations in us (in certain circumstances).
Detectivism about some property F is the view that things are F, and that this is a fact about how they are in themselves, not a fact about how they’re disposed to affect us. Thus detectivism about color might (but need not) take the form of a view which says that colors are spectral reflectance properties of surfaces, and that in our visual perception we ‘detect’ these properties.
Sometimes it will be convenient to use ‘realism’ as a shorthand for ‘detectivism.’ Typically dispositionalists will also want to call their view a form of realism, so it’s important to be clear about whether one is using the term in an expansive sense which includes dispositionalism, or a narrow sense which excludes it.
We will be concerned with the question of which of these is the right view about value. After discussing the metaphysical question, we’ll turn to its implications for the perception and epistemology of value.
Johnston considers and rejects three arguments for projectivism about ‘sensuous values.’
A story about the evolution of affective responses which seems to support projectivism. Two objections to the use of this sort of story as an objection to detectivism:
The argument: “the enormous range across different sorts of sentient beings of sensibilities or stable dispositions to fell cannot all be treated as potentially detective off sensuous goods. So, the argument from variation concludes that none can.” (185)
We can give a causal explanation of our affective responses without ‘invoking exemplifications of sensuous goods’; we have reason to believe only in things which are needed in giving causal explanations of the data we observe; therefore we have no reason to be detectivists about sensuous values.
Johnston’s objection to this argument is that it proves too much. Is this right? What sorts of beliefs are wrongly ruled out by this criterion? The problem of other minds.
Johnston’s main challenge to the projectivist is that he cannot explain ‘the authority of affect.’ He puts it as follows:
“By ‘the authority of affect’ I mean not to refer to its sheer effectiveness as a source of desire or action, but rather to the fact that the presence of the affect can make the desire or action especially intelligible to the agent himself.”
The analogy with perceptual experience; experience as making perceptual beliefs intelligible in the same way that affective responses make some desires and actions intelligible. The example of the imageless clairvoyant. Quinn’s example of the urge to turn the knobs of doors counter-clockwise.
One way to put the challenge to the projectivist is as follows: explain why the doorknob-turning urge (or the action which issues from it) is unintelligible in a way that desires and actions are not normally unintelligible. The detectivist explains this in terms of the lack of a perceptual-affective response to the object of the act; this explanation is not open to the projectivist, since the projectivist does not think that there are any sensuous goods in the worlds to be perceived.
Johnston considers three explanations which are consistent with projectivism:
Explanation 1: Lack of evaluative judgements
On explanation is that in the doorknob case, I don’t judge that turning the doorknob is worthwhile. Johnston objects that this is not enough to explain the lack of intelligibility, since acts can be intelligible from the inside in this sense even if I do not judge that they are worthwhile. The example of the Power 5000 Protein Bar.
Explanation 2: Lack of pleasure/pain
A second explanation: in the doorknob case, I there is no pleasure which accompanies my act, and no discomfort in not relieving my urge to turn the doorknob. Why this seems like a nonstarter.
Explanation 3: Lack of identification (second-order desire)
A third explanation: I do not identify with the doorknob turning urge, in the sense that I do not desire to have that urge be effective. The example of the chess addict; such an addict might find his addiction intelligible without this sort of identification. The example of yelling obscenities in department meetings (192).
The moral is supposed to be the same as in the above cases: this sort of identification is neither necessary nor sufficient for ‘intelligibility from the inside’, so its absence can’t be taken to explain the lack of intelligibility in the doorknob case.
Suppose that the projectivist grants the argument to this point. How might he modify his view of sensuous values in a way which is suited to meet Johnston’s challenge? A natural idea is that he might try to find some substitute for the detectivist’s ‘affective disclosure of sensory values.’ Perhaps we can explain the lack of ‘intelligibility from the inside’ in the doorknob example in terms of the absence of some affective state without construing this affective state in the way that the detectivist construes it. If such a state is not thought of as perception-like, then it has to be thought of as some kind of ‘raw feel.’ The idea might be that such a raw feel could not make desire and action intelligible from the inside any more than could pleasure and pain, or any of the other options discussed above. But why not? What is the argument for this?
Another reply on the part of the projectivist: a hybrid view, which combines one of more of the above explanations. E.g., a view which identifies Johnston’s ‘affective desire’ with the conjunction of evaluative judgement and pleasure. (See Wedgwood, ‘Sensing values?’)
So how could the projectivist make room for some sort of perception-like relation to sensuous values, without giving up too much of his metaphysics? The natural move here is a shift from projectivism to dispositionalism: a move from the denial of sensuous values to an identification of those values with the dispositions to produce such-and-such affective responses in us. Johnston puts it like this:
“So far I have just been assuming that the Projectivist cannot consistently have resort to the idea that affectivity is a refinement of sensing ...This may not be obvious. ...it may seem that the Projectivist can ...go on to endorse a Dispositionalist or ‘Response-Dependent’ account of the sensuous goods.
For example, he might say, with considerable initial plausibility, that a scene’s being ethereally beautiful is its being disposed to appear to be so under appropriate conditions. And he might go on to insist that a scene’s appearing so is to be understood as essentially involving an affective response.” (195)
The idea is that dispositionalism keeps in place the basic metaphysics of the projectivist, while leaving room for the kind of sensory state which is supposed to explain the intelligibility of action and desire in much the same way that perception makes intelligible the formation of beliefs.
Johnston argues in §6 that this idea is incoherent. He imagines that the dispositionalist adopts the following view of sensuous values like being ethereally beautiful:
being ethereally beautiful = having the disposition to appear (be presented as) ethereally beautiful in C
We can then consider three versions of this view, corresponding to two views of appearances which treat them as structured contents -- Russellian and Fregean -- and one which treats being appeared to as a kind of sui generis qualitative mental state which doesn’t represent exemplifications of being ethereally beautiful at all. (Johnston uses ‘Millian’ in the way that we have been using ‘Russellian.’)
Against Russellian dispositionalism
If appearances are Russellian, then they have as their constituents objects and properties. So the appearance of being ethereally beautiful contains the property of being ethereally beautiful itself. Against this, Johnston writes:
“...a disposition is a relational property and presentation is another relation. So the Dispositionalist will be identifying a property -- in this case the property of being ethereally beautiful -- with a relation to the holding of another relation that has that very property as a relatum. But a property cannot be contained within itself in this way.”
An analogy: the idea that something can be a proper part of itself. Why this seems incoherent.
Against Fregean dispositionalism
So suppose instead that the dispositionalist thinks of appearances as Fregean modes of presentations of properties. Then the above problem does not seem to arise, since the appearance of being ethereally beautiful will not contain that property as a proper part, but only a mode of presentation (Fregean sense) of it. So the dispositionalist is offering the following analysis of the property of being ethereally beautiful:
being ethereally beautiful = having the disposition to appear (be presented as) M in C
where M is a mode of presentation of the property of being ethereally beautiful.
Johnston thinks that this only obscures essentially the same incoherence which was visible on the surface of Russellian dispositionalism. An argument for this from the hierarchical view of properties.
Against non-representational dispositionalism
The non-representational dispositionalist says that
being ethereally beautiful = having the disposition to cause E-sensations in C
where E-sensations are not representations, whether Russellian or Fregean, of the property of being ethereally beautiful, but just certain ‘raw feels.’
The disanalogy between properties like being ethereally beautiful and dispositions to produce sensations (199).
What sort of expansion of the possible contents of experience is being suggested? Are there any properties of sensible particulars which, on this sort of view, can’t be perceived? If so, what marks out the difference?
Johnston’s view of the relationship between thought and perception. Keeping Kant’s idea that the contents of experience can be sensitive to one’s conceptual resources without lapsing into a kind of Kantian anti-realism; keeping empiricism’s idea that perception should be explanatory of thought without limiting the possible objects of experience to the ‘sparse contents’ which would be available to someone who lacks one’s conceptual capacities.