There has been a controversy in the perception literature about whether the kind of effect I have been describing is at least in part genuinely perceptual as opposed to an effect on the decision process involved in generating a report (Schneider & Komlos 2008; Valsecchi et al. 2010).
There probably are effects of attention on aspects of decision, including on conceptualization of a stimulus (Botta et al. 2014). However, I think the case is overwhelming that the attentional effect is at least in part genuinely perceptual. One reason involves “perceptual adaptation” a phenomenon known to Aristotle in the form of the “waterfall illusion”. As Aristotle noted, “...when persons turn away from looking at objects in motion, e.g., rivers, and especially those which flow very rapidly, things really at rest are seen as moving” (1955). Looking at something moving in a direction raises the threshold for seeing motion in that direction, biasing the percept towards motion in the opposite direction.
Perceptual adaptation is involved in the “tilt aftereffect”. If one looks at a left-tilting patch, the neural circuits for the left direction raise their thresholds. This is sometimes described (evocatively but inaccurately—see Anton-Erxleben et al. 2013) as neural fatigue. Then when one looks at a vertical patch, it initially looks tilted to the right. (See Figure 6 of Block 2010). The reason is that the neural circuits for rightward tilt dominate the percept because of the “fatigue” of the leftward tilt neurons. Ling & Carrasco (2006) showed that attending to the adaptor increased the size and duration of a variant of the tilt-aftereffect as if the contrast of the adaptor had itself been raised. Attending to a 70% contrast grating ramped up the tilt after-effect as if the contrast had been raised from 11 to 14% (different magnitudes in different subjects). Ling and Carrasco directed subjects to attend to gratings for 16 seconds. They found a benefit of attention at first in allowing subjects to distinguish tilts, since attention increases acuity, but then as adaptation increased, discrimination of the adapted tilt was impaired. This kind of adaptation is ubiquitous in perception but does not appear to occur in cognition or decision (Block 2014b). In case anyone thought that the attentional effect was entirely an effect on decision or cognition, this experiment suggests otherwise.
But even apart from the adaptation results, there is strong evidence going back at least to the 1990s from single cell recording in monkeys and in brain imaging for the conclusion that attention increases activity in the neural circuits responsible for the perception of contrast in a manner roughly consonant with an increase in the perception of contrast. Much of this evidence is summarized in sections 4.6 and 4.7 of a review article (Carrasco 2011). My hedge “roughly” stems from debates about the exact effect of attention. There are two kinds of “multiplicative” effects. In “contrast gain” the effect is just as if the contrast of the stimulus has been multiplied by a constant factor. In “response gain” the response is multiplied by a constant factor. The balance of these effects depends on the difference between the size of the target and the size of the “attentional field” (Herrmann et al. 2010). (These ideas are very clearly explained in Chapter 2 of Wu 2014.) A further kind of amplification effect is additive rather than multiplicative: the baseline or “floor” level of activation in the circuit is increased. There is some evidence (Cutrone et al. in press) for increased input baseline as a major part of the attentional effect.
Further, there is plenty of evidence for the conclusion that attention modulates specific cortical circuits depending on what feature is attended. A recent experiment (Emmanouil & Magen 2014; Schoenfeld et al. 2014) compared brain activation when subjects attended to a surface on the basis of its motion and when subjects attended to a surface on the basis of its color. Many of the stimuli involved both color and motion but which feature was task relevant was varied. The result was that motion sensitive areas of visual cortex were activated first when motion was task relevant and color sensitive areas of visual cortex were activated when color was task relevant. In Carrasco’s experiments, subjects’ attention is drawn to the specific features that the experiment concerns. In the experiment diagrammed in Figure 6, subjects are directed to report the location of the bigger gap, thereby directing attention to gap size. In the analogous experiment connected with Figure 7, subjects are asked to report the tilt of the patch that is higher in contrast, thereby directing attention to contrast. In experiments concerned with color saturation, subjects are shown stimuli that vary in saturation and asked to report the tilt of the patch that is higher in saturation. Similarly for many other features—speed, spatial frequency, flicker rate, motion coherence, shape, brightness, etc. These instructions can be expected to direct attention to the indicated features with amplification in the circuits that register those features.
Schneider (2011) seems to think that when subjects are asked to report on the side of the larger gap and the gap on the attended side is .20o while the gap on the unattended side is .23o, the subject finds that there is no difference in apparent gap size so the subject just chooses the more salient side. I will discuss salience in the next section, but there is one thing about this charge that raises a distinct issue: that subjects register the increase in apparent size only unconsciously. I now turn to that issue.