100

Non-selective Attention 1 There are many convincing examples of attention changing appearance in a way that does not involve selecting some properties and de-selecting others. The effect I will be appealing to requires a small amount of practice in moving one’s attention without changing fixation, but once one manages this it is a strong effect. To get the effect, fixate (that is: aim your eyes at) any one of the 4 squarish dots in Figure 2A. Then move your attention around from one disk to another without moving your eyes. This appears to be much easier for some people than others. Still, every one of sixteen observers in Peter Tse’s (2005) experiment said that the attended disk darkened and that they could darken a circle by shifting their attention to it. Tse argues convincingly that the visual system constructs the interpretation which minimizes the number of transparent (or translucent) layers seen. I won’t go through the reasoning, but the reader can see one piece of evidence that seeing the disks as transparent is necessary for the effect by noting that the effect does not32 NedBlock Figure 2. The Tse Illusion. In A, you can alter the brightness of a disk by attending to it while keeping your eyes fixated on one of the dots. This does not work for B. work for the stimulus in 2B. (Zero of sixteen subjects reported darkening disks in that experiment.) Because the background of B is seen as black, the disks do not look like transparent or translucent disks. The effect involves grouping and is selective in that sense, but it is not selective in the sense described in the last section, the sense that lends itself so easily to treatment by the direct realist.Attention and Mental Paint 33 I mention the Tse Illusion here for three reasons. First, in case any holdouts denied that change of voluntary attention can change phenomenology, this example should put that view to rest for once and for all. Second, it is useful by way of contrast to the next phenomenon I will be presenting, to see why this phenomenon does not pose an immediately obvious problem for direct realism or representationism. The reason is that it is an illusion. The three disks are really equally bright and what the moving of attention does is make one of them, illusorily, darker. Direct realism says that in veridical perception one is directly aware of actually instantiated properties, and that view is not challenged by an illusion. (Later, I will be discussing one direct realist attempt to provide a theory of illusion.) The representationist’s representational contents are grounded in veridical perception and those contents can misrepresent in illusion so they will have no problem with this case. Finally, as I mentioned, moving attention independently of fixation takes practice, and I hope the reader’s practice with Figure 2 will allow a better appreciation of Figure 4 to come. 5. Non-selective Attention 2 I will now present the evidence that supports my challenge to direct realism and representationism. Let me go over the empirical facts that I will be appealing to. 1. As I mentioned in connection with Figure 2, when one’s eyes are pointed straight at a thing or a place, one is said to be “fixating” it. However, and this is fact number one, one can consciously see something without fixating it, as when in Figure 2 one fixates one of the dots but nonetheless sees disks that one is not fixating. (I add “consciously” here to make it clear that I am not talking about subliminal perception or other sorts of unconscious perception. I will omit the word ‘consciously’ in what follows, but talk of seeing in this paper should be understood as conscious seeing, unless otherwise indicated.) Sometimes we speak in this connection of looking at something out of the corner of one’s eye. 2. One can see something without focally attending to it as well as without fixating it. 3. Most importantly: Focal attention changes perceptual qualities such as perceived contrast, perceived color saturation, perceived object size, perceived spatial frequency (a measure of stripe density), perceived gap size, perceived speed and perceived flicker rate. As with most results concerning consciousness, the experimental evidence for this conclusion has some loose ends that I will describe. Terminological note: contrast12, saturation, size, spatial frequency, gap size and flicker34 NedBlock rate are objective properties of objects or events. When I talk of perceptual qualities such as perceived contrast, perceived saturation, etc., I mean the relevant phenomenological qualities of the perceptual states that ground the perceiver’s knowledge of objective contrast, saturation, etc. 4. Attentional effects, both excitatory and suppressive, pervade the visual field (Datta & DeYoe, 2009; Downing & Pinker, 1985; Hopf et al., 2006). The first three of these points were appreciated by William James (1890, p. 425): “…to some extent the relative intensity of two sensations may be changed when one of them is attended to and the other not.” Although he emphasizes attention to sensations in this passage, the examples James uses to illustrate it include distributions of attention to colors and sounds in the environment. Hermann Ebbinghaus (1908a, 1908b) appears to have had a similar view. James also recognizes that “The subject is one which would well repay exact experiment, if methods could be devised,” something that has now been done in experimental psychology, notably by Marisa Carrasco and her colleagues. The experiments to be described provide strong evidence for the claim that the phenomenal appearance of a thing depends on how much attention is allocated to it. Figure 3 diagrams one of Carrasco’s paradigms (Carrasco, 2009; Carrasco et al., 2004). The subject is instructed to look at the “fixation point” throughout each trial, never moving the eyes. (Obedience is often checked with an infra-red camera, but in any case in most versions of Carrasco’s paradigm for involuntary attention there is not enough time for the subject’s eyes to move.) The fixation point is presented for half a second, then a dot can appear (randomly) on the left, right or center. The subject is instructed to ignore the dot and informed that the dot does not indicate or predict the orientation or side of the higher contrast stimulus. However, low level attention mechanisms ensure that the subject cannot ignore the dot. The results of the experiment reveal that a dot on one side attracts the subject’s attention to that side. The subject then is shown another brief fixation point, then two “Gabor patches”, small grids (made up of sinusoidal luminance stripes) each of which can face in one of two directions. The subject’s task is to report the orientation of the Gabor patch that is higher in contrast (see footnote 12 for a definition of ‘contrast’) by pressing one of the keys pictured in the lower right of Figure 3. The subject is supposed to press one of the keys on the left if the left Gabor patch is higher in contrast and one of the keys on the right if the right Gabor patch is higher in contrast. The subject is prepared for this judgment by being shown examples of comparisons of contrast. (If the subject is to judge e.g. color saturation, the subject is showed examples of differential color saturation.) The idea of the procedure is to avoid bias by placing the contrast judgment at one level of remove from what the subjectAttention and Mental Paint 35 Figure 3. This diagram of an experiment reported in Carrasco, Ling & Read (2004) is to be read from left to right, starting with the presentation of a fixation point for half a second that the subject is supposed to look at and ending with the fixation point, at which time the subject is supposed to respond, either by pressing one of the pair of keys on the left or one of the pair of keys on the right (pictured in the lower right hand corner). ISI = inter-stimulus interval. The subject’s task is to report the orientation of the grid-like patch (called a Gabor patch). Further details are presented in the text. is officially judging, namely angle. Of course the experimenter is interested in the apparent contrast, not the angle. The result is that if an attended Gabor patch is slightly lower in actual contrast, attention can boost its apparent contrast to the point of apparent equality of contrast between the two patches. (This effect no doubt involves decreased apparent contrast of the less attended patch.) If the two Gabor patches are the same in actual contrast, an attended patch looks higher in contrast. Carrasco was able to map out the effect quantitatively for a wide range of contrasts. In Figure 4, the square dot between the two patches is a fixation point. If one is attending to the fixation point, one can still see both Gabor patches. (Try it!) The patch on the left looks lower in contrast than the one on the right, a veridical perception. However, if one is attending to the patch on the left, that patch looks equal in contrast to the patch on the right. The combination of increased attention to the left patch and decreased attention to the right patch makes for a 6 point boost in apparent contrast to the left figure relative to the right in Figure 4. (The boost effect increases with increasing average contrast.)36 NedBlock Figure 4. The Gabor patch on the left has 22% contrast. The patch on the right has 28% contrast. A 6% contrast difference is easy to detect as the viewer can confirm. The effect described in the text is one in which involuntary attention attracted to the left increases the apparent contrast of the 22% patch (and decreases apparent contrast of the 28% patch) to the point of subjective equality between the two patches. With a bit of practice, one can voluntarily shift attention to the left or to the right and experience the increase from this diagram. Figure 5 reports four comparisons of the sort just discussed. For example: If the subject fixates the upper right square dot, and attends to the dot, the Gabor patch on the left of the fixation point (6% contrast) looks, veridically, to be lower in contrast than the one on the right; but if the subject attends to the 6% contrast patch it looks equal to the 8.5% contrast patch. So at this level of contrast, attention is worth 2.5 points of contrast. Anespecially interesting comparison is that on the upper left because a 3.5% patch is invisible if one is not attending to it. So, fixating the square dot on the upper left, the difference between attending to the left and attending elsewhere makes for a difference between seeing the patch and not seeing it. The kind of attention involved here is involuntary, attracted by the appearance of a dot, but Carrasco and her colleagues have recently shown (Liu, Abrams, & Carrasco, 2009) that the same effect occurs with voluntary attention. Subjects were asked to fixate on a cross and to attend to the right if the right side of the cross thickened. Likewise for the left. Although voluntary attention is deployed much more slowly than involuntary attention, the boost to perceived contrast was about the same as for involuntary attention. Interestingly, some perceptible features showed the effect and others did not. For example, color saturation shows the effect but hue does not. The effect works for gap size (Gobell & Carrasco, 2005), size of moving objects, flicker rate and spatial frequency (more stripes in a Gabor patch of a given size constitutes a higher spatial frequency). Further, building on earlier work (Golla, Ignashchenkova, Haarmeier, & Thier, 2004; Shiu & Pashler, 1995),Attention and Mental Paint 37 Figure 5. The previous figure comes from the lower right of this diagram. Carrasco and her colleagues showed (Montagna, Pestilli, & Carrasco, 2009) that attention increases acuity at the attended location. One of the underlying neural mechanisms of this fundamental feature of vision appears to be that attention shrinks the relevant receptive fields of neurons in the visual system, the receptive field of a neuron being the area of space that the neuron responds to (see references in Montagna et al., 2009).