A THEORY OF SENTIENCE. By Austen Clark. New York: Oxford University Press, 2000. Pp. v, 288.

Three central theses of A Theory of Sentience are these:

First, what it is for a sentient being to sense is for it to employ two distinct capacities: one for representing places-at-times; the other for representing "features" (60, cf. 70). Exercised together, the result is akin to feature-placing, which brings us to the second thesis: what sensory systems represent is that features are instantiated at place-times. Accordingly, sensory systems do not, for instance, attribute properties to objects, such as trees, tables, bodies, or persons (163).

Finally, an empirically correct account of the mechanisms of spatial representation for an organism, combined with an account of the neural correlates, in that organism, of its sensations of features, would suffice to explain "how…a merely material system" can "produce the…tangy phenomenology of sense" (vii). The phenomenology of sense is distinguished from that of conscious states more generally (166).

Drawing on the notion of quality space developed in Sensory Qualities (1993), Clark outlines the form that an account of an organism's neural correlates of sensation of features could take. The book is divided into a helpful recap of this strategy (chapter 1 and an appendix entitled "How to Close the Explanatory Gap"), three chapters arguing that spatial representation and representation of "features" are distinct capacities (chapters 2-4), a discussion of empirical models that would accord with Clark's theory of sentience (chapter 5), and a discussion of the correctness conditions of sensations of colors (chapter 6). There is much of interest in these chapters, which simultaneously engage philosophical problems of perception and empirical results. I will focus on Clark's characterization of his theory's explanandum--sensory phenomenology.

It goes with Clark's first thesis above that the phenomenology of sense varies along two dimensions: one qualitative, the other spatiotemporal. The two dimensions are needed, Clark says, because "there seems to be nothing analysable in the quality of a tone that makes it a tone from here rather than there" (61). Pairs of cheeps may yield qualitatively identical sensations, even though the cheeps differ in apparent location (59). Two experiences vary along the qualitative dimension if they vary with respect to their qualitative properties. But what are qualitative properties?

In the terminology that Clark adopts from Galen Strawson (Strawson 1989), qualitative properties are distinguished from phenomenal ones. Phenomenal properties characterize how external things appear: "If the apple one spots, picks and bites looks red, feels cold, and tastes sweet, then red, cold and sweet are phenomenal properties" (2). Qualitative properties, in contrast, are properties of mental states--namely, experiences--themselves: "qualitative properties are properties of such internal states, in virtue of which things out there, dangling in front of one's grasping fingers, appear as they do" (2).

Clark's view is that variations in qualitative properties just are variations in which phenomenal properties the qualitative ones represent. ("[Qualitative property] Qr = the property of sensations in virtue of which they are sensations of [phenomenal property] Pr" (257).) A quibble: the definition of 'qualitative property' is supposed to be suitable for stating debates in which it is controversial whether there any qualia. (Cf. 3: "The neutral term "qualitative properties'…should not be assumed to have any of the defining characteristics of the now infamous 'qualia'.") To the extent that Clark means to engage debates about the relation of experience to representation, his definition seems unfortunate, since some philosophers think experience has entirely non-representational features, and sometimes they call those features "qualia". For the purpose of engaging that view, it would be useful to have a notion of qualitative properties according to which non-representational features count as such. The quibble isn't as serious a worry as it would be, if Clark had failed to address any such views. As it is, he responds (in chapter 3) to Boghossian and Velleman's claim (in Boghossian and Velleman 1989) that the phenomenology of sense is not limited to representation of external things. So the view that phenomenology is exclusively a matter of representing external things is not simply taken for granted.

We're never given a list of which properties are the phenomenal ones, but throughout the book various examples are mentioned: hue, saturation, brightness, temperature, roughness, pain, cheeps, beeps, and clicks (the last three are noises). These examples conform to Clark's claim that sensory systems are akin to users of feature-placing languages, a paradigm sentence of which Clark takes to be "cheep cheep here". The mechanisms accounting for representations of places-at-times, Clark says, "function like singular terms"; whereas those accounting for representations of phenomenal properties have a "proto-predicative function" (74). When both kinds of mechanisms operate, the result is, Clark speculates, a sensory ancestor of linguistic reference and predication (73).

Feature-placing languages as P. F. Strawson describes them (Strawson 1963, chapter 7, section 1) steer clear of predication altogether: "It is raining here" does not mean that a place is raining; it is silent on what, if anything, is doing the raining. Although Clark says that Strawson's account inspires his theory of sentience (75), he repeatedly says that sensory systems attribute qualities to places. For example "The sensation of a red triangle…picks out places and attributes features to them" (147; cf. 69, 70, 77, 165, 167, 185). Taken literally, these claims seem questionable. If audition told us that it was a place, rather than something at that place, that was cheeping, we would have all sorts of errors to correct in the move from audition to thought. We would be similarly misled by vision if it told us that a certain region of space was red, while remaining neutral on whether anything occupying that place was red.

A better interpretation of these passages thus seems to be that sensation tells us that phenomenal properties are instantiated at a certain place, without telling us anything about which particulars, if any, are doing the instantiating. To maintain such neutrality, sensation would not have to stop short of predication; after all, it is hard to understand the difference between the claim Something is red at location L and the claim Redness is instantiated at L. How could redness be instantiated at L, if it were not instantiated by something or other? Perhaps the best interpretation of the "proto-predicative" status of sensation's contents as Clark understands them is not that they are free of predication as such, but rather that they stop of short predicating anything to particulars.

As a piece of phenomenology, however, this seems implausible. What happens in sensory phenomenology when a subject sees a basketball make its way from the player's hands to the basket? The information that it's one and the same basketball traversing a single path is not given by sentience if sentience is limited to feature-placing. On Clark's view, the information that it's one and the same basketball traversing a single path has to be given non-sensorily. The subject's visual experience stops short at delivering a series of momentary presentations of orange-and-roundness at a series of places.

Given his commitment to the thesis that sentience is limited to feature-placing, Clark could grant that reidentifying moving basketballs is phenomenologically impressive only by distinguishing sensory phenomenology from that of other kinds, and by holding that the reidentification phenomenology is non-sensory. Although it is not clear how the distinction between sensory and non-sensory phenomenology should be drawn, by invoking it Clark could at least respect the point there is some phenomenology to seeing moving objects. But to the extent that the phenomenology of sense includes representing moving objects, Clark's theory seems to underestimate its complexity.

Susanna Siegel

Harvard University


Boghossian, P. and Velleman, D. 1989. "Color as a Secondary Quality." Mind 98: 81-103.

Clark, A. Sensory Qualities. 1993. Oxford: Clarendon.

Strawson, Galen. 1989. "Red and 'red'". Synthese 78: 193-232.

Strawson, P.F. 1963. Individuals. New York: Anchor Books.