Online research in philosophy

Three different ways to understand the representational content of the feature maps employed in early vision are compared. First is Stephen Kosslyn's claim, entered as part of the debate over mental imagery, that such areas support "depictive" representation, and that visual perception uses them as depictive representations. Reasons are given to doubt this view. Second, an improved version of what I call "feature-placing" is described and advanced. Third, feature-placing is contrasted with the notion that the representational content of those feature maps could be conveyed in a list of sentences about visual objects. Some problems with this last alternative are described.

Drawing on work in psychophysics, psychometrics, and sensory neurophysiology, Clark analyzes the character and defends the integrity of psychophysical explanations of qualitative facts, arguing that the structure of such explanations is sound and potentially successful.

One of the biggest challenges in understanding perception is to understand how the nervous system manages to integrate the multiple codes it uses to represent features in multiple sensory modalities. From different cortical areas, which might separately register the sight of something red and the touch of something smooth, one effortlessly generates the perception of one thing that is both red and smooth. This process has been variously called "feature integration", "binding", or "synthesis". Citing some current models and some historical precursors, this paper makes some simple observations about the logic of feature integration. I suggest that "feature conjunction" is not strictly speaking conjunction at all, but rather joint predication; and that the critical task in "binding" is not simply grouping scattered representations together, or providing them a common label, but rather identifying those that have a common subject matter-those that are.

1. Clark’s book is a detailed study of the nature of sensory representation. It is highly informed by empirical results in the psychology of perception, and philosophically rich and significant. I admire the book and learned a great deal from reading it. As it covers a wide range of topics, and as I have no overarching critique to present, in this commentary I will briefly address three issues that come up in the book: Clark’s relational type-identity thesis for sensory qualities, his theory that sensory representations involve proto-singular terms referring to spatio-temporal regions in the subject’s environment, and his interesting proposal concerning color to treat it as “difference coding”. Some of my remarks will be critical, but others will just explore some of the implications of his view. 2. Clark distinguishes “phenomenal properties” from “qualitative properties”, the former being appearance properties of things in the world (their colors, shapes, tastes, odors, etc.) and the latter being the properties of sensations by virtue of which they are sensations of their corresponding phenomenal properties. So when I see a red ball I am “directly” aware of the ball’s redness and roundness - it appears red and round to me. This awareness of the ball’s redness and roundness is accomplished, however, by my having a visual experience with certain qualitative properties; those that are of the sort one has when seeing something as red and round. It is these latter qualitative properties that are the subject of his relational type-identity thesis. Before addressing that thesis, however, I want to quickly note and respond to another point Clark makes concerning the qualitative properties of sensory states. He..

This paper contrasts three different schemes of reference relevant to understanding systems of perceptual representation: a location-based system dubbed "feature-placing", a system of "visual indices" referring to things called "proto-objects", and the full sortal-based individuation allowed by a natural language. The first three sections summarize some of the key arguments (in Clark, 2000) to the effect that the early, parallel, and pre-attentive registration of sensory features itself constitutes a simple system of nonconceptual mental representation. In particular, feature integration--perceiving something as being both F and G, where F and G are sensible properties registered in distinct parallel streams--requires a referential apparatus. Section V. reviews some grounds for thinking that at these earliest levels this apparatus is location-based: that it has a direct and nonconceptual means of picking out places. Feature-placing is contrasted with a somewhat more sophisticated system that can identify and track four or five "perceptual objects" or "proto-objects", independently of their location, for as long as they remain perceptible. Such a system is found in Zenon Pylyshyn's fascinating work on "visual indices", in Dana Ballard's notion of deictic codes, and in Kahneman, Treisman, and Wolfe's accounts of systems of evanescent representations they call "object files". Perceptual representation is a layered affair, and I argue that it probably includes both feature-placing and proto-objects. Finally, both nonconceptual systems are contrasted with the full-blooded individuation allowed in a natural language.

In his book A Theory of Sentience, Austen Clark argues that the content of sensory representations can be expressed as sentences constructed from a language of sentience. Such sentences specify that a determinate feature obtains in a particular space-time region, but the language's limited vocabulary prohibits the sentences from referring or attributing features to objects. In this paper, I show that this view is flawed in at least two ways. First, if sensation has the capacities that Clark and others attribute to it, then the vocabulary of sense extends further than he supposes, and a limited language of sentience cannot justify a prohibition of object representation within sentience. Second, even if the language of sentience is as impoverished as he claims, and thus even if the representational capacity of sentience is correspondingly limited, object individuation can plausibly occur at the level of sensation. In the course of defusing Clark's major argument for an “object-less” theory of sentience, I offer reasons to believe both that sensory representations can be impressively sophisticated in what they say about the world, and that object representations can be surprisingly basic.

In Clark (2000), Austen Clark argues convincingly that a widespread view of perception as a complicated kind of feature-extraction is incomplete. He argues that perception has another crucial representational ingredient: it must also involve the representation of "sensory individuals" that exemplify sensorily extracted features. Moreover, he contends, the best way of understanding sensory individuals takes them to be places in space surrounding the perceiver. In this paper, I'll agree with Clark's case for sensory individuals (.

The paper argues that material objects are the primary referents of visual states -- not places, as Austen Clark would have it in his A Theory of Sentience.

Austen Clark offers a general account of the forms of mental representation that we call "sensory." Drawing on the findings of current neuroscience, Clark defends the hypothesis that the various modalities of sensation share a generic form that he calls "feature-placing." Sensing proceeds by picking out place-times in or around the body of the sentient organism, and characterizing qualities (features) that appear at those place-times. The hypothesis casts light on many other troublesome phenomena, including the varieties of illusion, the problem of projection, the notion of a visual field, and the existence of sense-data.