1 Introduction
Attention increases acuity, allowing the perceiver to see details that would otherwise be missed. In addition, for items that the perceiver does actually see, attention changes their appearance, increasing, for example, the appearances of contrast, (differences between light and dark), speed of a moving object, spatial frequency (a measure of how closely spaced light and dark areas are) and the size of a gap—as in Figure 4. But when attention makes something appear bigger or faster, does it work like a magnifying glass, trading off a gain in information at the cost of making something appear bigger or faster than it is? Or does attentive perception portray the item more as it really is? Or are both percepts veridical—or are both non-veridical? Similar issues arise with regard to inhomogeneities in the visual field. Vision in the lower visual field is about 65% more sensitive to contrast (and orientation discrimination, texture segmentation, gap size, speed, spatial frequency) than vision equidistant from fixation in the upper visual field. (See Figure 1 for examples of low and high contrast.) In addition, there is a great deal of noise in perceptual systems. Percepts involving the same area of the visual field and the same degree of attention will typically differ in visual response from occasion to occasion. So on different occasions, one can see the same object or event in the same conditions, with the same degree of attention, and from the same vantage point and it will look different in size or speed or contrast because of random factors.
What is the consequence of these facts for the veridicality of perception? One viewpoint says that perception is mostly slightly mistaken. We usually see length, speed and contrast non-veridically but the extent of error is small enough not to be problematic. However, this viewpoint cannot be right since it is only in virtue of a history of veridical representation both in our own lives and in the past of our species that our perceptual representations even have representational contents (Burge 2010). Without such a history of veridical representation it is not clear that perceptual representation really makes sense.
An alternative way of thinking about the issue is that perception is sufficiently imprecise in its representational content for all these varying percepts to be veridical. If a person is said to be 5 feet to 6 feet tall on one occasion and 6 feet to 7 feet tall on another, both are veridical if the person is 6 feet tall. One could put this by saying that perceptual representation is “intervalic”. The intervals however would have to be pretty large given the size of these effects—notably the 65% difference between lower and upper visual field just mentioned. And it is hard to square such large differences with the phenomenology of foveal vision. Hold a piece of lined paper in front of you. You seem to see the difference between the white space and the lines fairly precisely. “Irrelevant!,” you may retort, “Those differences in the visual field affect only peripheral perception; attentive foveal perception is much more precise than inattentive peripheral vision.” And this resolution seems to be reflected in our phenomenological judgments: move the piece of lined paper out to 30o away from the line of sight. Doesn’t your visual impression of the contrast between the lines and spaces seem, well, less precise? Surprisingly there is evidence that unattended and peripheral perception of some properties (notably contrast) are about as precisely represented in attentive foveal vision as in inattentive vision and vision in the near periphery (up to a 30o angle from the line of sight). The upshot is that the phenomenology of perception may mislead us with regard to the precision of the representational content of perception.
One might suppose that help will come from bodily action. Goodale & Murphy presented 5 rectangular blocks to subjects at various positions in the visual field ranging from 5o to 70o away from the line of sight (1997). They compared accuracy of perceptual discrimination of one block from another with accuracy of grip via a device that measured the aperture between thumb and forefinger as subjects reached out to pick up one of the blocks. Grip accuracy is roughly the same at 5o as at 70o. The fine details of action are controlled by a largely distinct system from the system that underlies conscious vision. So what this result dramatically illustrates is that the precision of bodily action is unlikely to cast any light on the precision of perceptual phenomenology.
This is the puzzle of the title. I argue that the disconnect may be real and that perceptual phenomenology may mislead about perceptual representation. Perceptual phenomenology may not be grounded in the representational content of perception. Further, there may be no “phenomenal content”, that is no representational content that emerges from the phenomenology of perception.[1]
This is a very long paper so it might be useful to know what parts to focus on. You can see the basic lines of the dialectic from reading sections 1- 3. Sections 4- 7 concern the experimental data concerning attention and can be skimmed without losing the thread. The argument resumes with 8- 10. 11 can be skipped without loss of continuity. 12 covers some of the results that the argument is based on. 13 can be skipped. 14 is the conclusion.
Figure 1: Six levels of contrast. The Wikipedia caption reads “Different levels of contrast - original image top left - less contrast to the left (50%, 75%), more to the right (25%, 50%, 75%)”. I take this to mean that the mid-left photo has 50% less contrast than the upper left, the lower left photo as 75% less contrast than the upper left, etc. These percentages are differences from photoshop, not absolute measures of contrast of the sort to be discussed later in the paper. Percent contrast in the sense to be discussed is the difference between the luminance of the lightest and darkest parts divided by the sum of these luminances. These images come from the Wikipedia entry on contrast. According to Wikipedia, “Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation; with no Invariant Sections, no Front-Cover Texts, and no Back-Cover Texts. A copy of the license is included in the section entitled GNU Free Documentation License.”