Conscious Experience and Public Report

Reports by subjects of what something looks like are even more problematic since what people report in an experimental setting is known to be affected by many factors, including what subjects think the experimenter wants (such compliance effects have been called experimenter demands), what they believe the task to be (which have been called task demands) as well as subjects' general beliefs and utilities. Every response requires making a decision that may involve weighing the costs (including embarrassment) of different sorts of errors, particularly errors of commission versus errors of omission. For example, work on subliminal perception or ''perceptual defense'' shows that accuracy in reporting whether one has ''seen'' a briefly presented word is different for taboo words than for neutral words equated for frequency of occurrence (Freeman 1955). These are typically not cases of subjects being disingenuous, but of making rational choices— choices that can be traced to processes described by models of decision making, such as utility theory and signal-detection theory. Sometimes signal-detection theory can separate contents from reporting biases in a fairly direct way by providing different measures for response criteria (the parameter usually written as p) and for sensitivity (the parameter d') in experiments involving thresholds. This is done by taking into account not only the correct responses, but also the relative rate of errors of omission and errors of commission. If the subject has a bias to report seeing something independently of whether there was a signal, then both the hit rate and the rate of errors of commission will increase. Such a bias would be useful if the signal were present on most of the trials or if the utility of detecting the signal was high (e.g., if it signaled danger). If, on the other hand, the subject has a conservative bias, then the hit rate will be lowered but the rate of errors of omission will also increase. This sort of bias would be useful if the signal were present only rarely. These tendencies can be used to separate response bias from the availability of conscious contents (Snod-grass 2002). Experimental psychology has learned that sincere reports of conscious contents have to be evaluated in relation to other sources of evidence and in the light of developing theories.

Consider, for example, the problem of interpreting such findings as those reported by Wittreich (1959). A well-known illusion is that when people walk across the floor of a specially designed room called the Ames room (shown in figure 4.2) they appear to change in size.5 Wittreich confirmed

5. It does not reveal a magician's proprietary trick to tell you that the room is actually distorted, having been constructed with one side much lower and shallower than the other. The design specifications are such that rays drawn from a peephole to every visual feature—i.e., corner and vertex (the room has windows)—bear the same visual angles to one another as they would have in a regular rectangular room. Thus when viewed through the peephole all the visual cues in the Ames room are identical to those that would have been available in its corresponding phantom rectangular room. Of course it is not possible to build a distorted room such that the illusion persists as the viewer moves inside the room, though this could (almost) be

Figure 4.2

From a fixed peephole vantage point the Ames room seems like a normal rectangular room, but people in the room look to be different sizes depending on where they stand. (© The Exploratorium, www.exploratorium.edu, used with permission.)

Figure 4.2

From a fixed peephole vantage point the Ames room seems like a normal rectangular room, but people in the room look to be different sizes depending on where they stand. (© The Exploratorium, www.exploratorium.edu, used with permission.)

this observation, but he also found that this did not happen when the people were well known to the observer—for example, the observer's spouse— even if these people were accompanied by a stranger, whose size did appear to change! Notwithstanding the presumably sincere reports made by the subjects, there remains the question whether to interpret this finding as showing the malleability of judgments of the content of conscious experience, or of the operation of the visual system itself. The problem is not that subjects are disingenuous, but simply that the lines between what we report and what we believe with great conviction, as well as between what we report to others and what we report to ourselves, are not so clear. If, as many have supposed (e.g., Block 1995;Dennett 1991), part of conscious content is what vision (or imagination) reports to the rest of the mind, done in an electronic virtual reality room (the reason for the qualifier is that VR displays cannot reproduce all cues exactly; in particular, since objects are not actually located at different depths but on the same 2-D surface, the eyes do not focus at different depths, which results in some conflicting cues).

then what it reports may be different from the information that it actually possesses. In other words, there may well be a partial dissociation between the content of our conscious experience and the information that is passed on to other stages in mental processing. Sometimes we can show this fairly directly by comparing measures from which the reporting bias has been mathematically factored out, as we do when we use the signal-detection measure d' rather than percent correct. Such measures not only separate what information observers have from what they report to an experimenter, but also from what observers report to themselves—that is, what they are aware of. More often than not, general questions such as this are ultimately adjudicated according to whether a theory that takes certain observations at face value simply misses underlying (causal, functional) principles.

Focusing on the conscious contents of perception has also encouraged direct perception theories (such as those of Gibson 1979), which claim that perception allows us to directly access (''pick up'') information about properties of the world that are prominent in our experience, such as the property that things have of being suitable for certain purposes—from being eaten to being sat upon (suitability is referred to as having certain ''affordances''). James J. Gibson has argued, quite reasonably, that we see not patterns of light and shadow and patches of color, but familiar things such as tables, chairs, and people. Moreover, we never see just the front surface of objects;we see entire objects and we see them as particular things, such as our car or our spouse, or as having certain affordances, such as being graspable or edible. Although the urge to shun visual representations led Gibson and his followers to embrace what is essentially a behaviorist position, they were right to claim that perception eventuates in the extraction of abstract properties rather than low-level sensory patterns (''sensations''). The moral of this observation should have been that what we see is a reconstruction of the properties of distal objects: we never experience the preconstancy proximal stimulus. But in direct realism theories (for various modern versions, see Smith 2003) this is not the moral that is drawn. Rather, these observations are taken to be an indictment of the view that perception begins with properties as described by physical science and constructs a representation of a scene ( perhaps in some cases with the aid of inference from general knowledge). Instead, they are taken as support for the radical view that the world should be redescribed according to the categories of experience, which are assumed to be the starting point of perception; these are the categories to which perception is inherently attuned and which are ''picked up'' the way a tuning fork picks up the notes in its immediate environment. To make this picture work Gibson also had to deal with the problem of misperception, which, in turn, led him the view that theories should be applied to perception in an ''ecologically valid'' environment (for a critical discussion of these ideas, see Fodor and Pylyshyn 1981;Pylyshyn 1984, chap. 6). Although it is not usually put in this way, it is the temptation to see the categories of conscious experience as the primitive bases for (or inputs to) perception that has been one of the siren calls of direct realism.

Notice that the position I have been describing in this book bears some similarity to Gibson's. I too do not believe that we should take the starting point of vision (the nonconceptual first steps) to be sensations, if by sensations we mean consciously experienced colors, shapes, textures, and so on (or whatever the primitive sensations turn out to be). Rather, the starting point should be nonconceptual, in particular it should be nonconceptual demonstrative references to proto-objects or FINGS.

Was this article helpful?

0 0
Self Help Affirmations

Self Help Affirmations

The big book of affirmations from personal development authors. An affirmation is something you say to yourself. Everybody uses them on purpose or accidentally. You get up in the morning, leap out of bed and proclaim,

Get My Free Ebook


Post a comment