3 Motivating pessimism

Two distinctions will prove helpful in what follows. One distinction is between forms of pessimism that concern only our capacity to identify our phenomenal states and forms of pessimism that call into question our capacity to both discriminate and categorize our phenomenal states. A second distinction concerns the scope of pessimism. At one end of the spectrum are local forms of pessimism that concern only a relatively circumscribed range of phenomenal states (say, imagery experiences), while at the other end of the spectrum are forms of pessimism that are unrestricted in scope. Perhaps no theorist has ever embraced a truly global form of pessimism—even Schwitzgebel grants that introspection is trustworthy with respect to certain aspects of consciousness—but some forms of pessimism are clearly wider in scope than others. These two distinctions are, of course, orthogonal to each other. One could be a moderate but global pessimist; alternatively, one could endorse a radical but highly local form of pessimism.

So much for the varieties of pessimism—how might one argue for the view? One influential line of argument for pessimism—or at least something very much like it—appeals to the alleged privacy of introspection. Because an individual’s introspective judgments cannot be checked by anyone else, it follows—so the argument runs—that it would be inappropriate to trust them. This argument is often used to motivate the view that introspection is scientifically illegitimate, but it could also be used to motivate the view that one should adopt a sceptical attitude towards one’s own introspective capacities.[6] Although it has been influential, I will leave this argument to one side in order to focus on a trio of arguments that aim to establish not merely that there is no positive reason to trust introspection (as the argument just outlined attempts to do), but that there is positive reason not to trust it. My presentation of these arguments will draw heavily on Schwitzgebel’s work, for he has done more than any other author to develop and defend them.[7]

But before I examine those arguments, I want to consider the overall structure of Schwitzgebel’s case for global scepticism. As I read him, Schwitzgebel employs a two-step strategy (2008, p. 259). The first step involves attempting to establish a form of local pessimism via one (or more) of the three argumentative strategies to be explored below. The second step involves generalizing from the kinds of phenomenal states that are the targets of local pessimism to phenomenal states in general. The second step is clearly required, for without it we would have no reason to regard introspection in general as “faulty, untrustworthy, and misleading—“not just possibly mistaken, but massively and pervasively” (Schwitzgebel 2008, p. 259).[8]

I will consider both steps in due course, but the crucial point to note for now is that, considered in the abstract, the second step of the argument looks somewhat suspect (Bayne & Spener 2010). Even if there are hard cases for introspection—that is, cases in which introspective access to phenomenology is insecure—there also easy cases—that is, cases in which introspective access to phenomenology is clearly secure. Indeed, Schwitzgebel himself grants that introspection “may admit obvious cases” and that some aspects of visual experience “are so obvious it would be difficult to go wrong about them” (Schwitzgebel 2008, p. 253). But if that’s the case, then one might well ask why we shouldn’t generalize from those cases rather than from the hard cases on which he focuses. Schwitzgebel complains that to generalize about introspection only on the basis of the easy cases “rigs the game”. That’s true. But it’s equally true that to generalize only on the basis of the hard cases—as Schwitzgebel seems to do—would also rig the game. In fact, it would seem pretty clear that any comprehensive account of the epistemic landscape of introspection must take both the hard and easy cases into consideration. Arguably, generalizing beyond the obviously easy and hard cases requires an account of what makes the hard cases hard and the easy cases easy. Only once we’ve made some progress with that question will we be in a position to make warranted claims about introspective access to consciousness in general. What this suggests is that although there is a formal distinction between the two steps of Schwitzgebel’s argument, the steps are not entirely independent of each other, for the fortunes of the second step rest in part on the case that can be made for the first step. With that thought in mind, let us now turn to the arguments for pessimism.

3.1 The argument from dumbfounding

One line of argument that features prominently in Schwitzgebel’s work is what I call the argument from dumbfounding.[9] Arguments of this form involve posing introspective questions that allegedly stump us—questions that we find ourselves unable to answer with any significant degree of confidence. Here’s an example of such an argument:

Reflect on, introspect, your own ongoing emotional experience at this instant. Do you even have any? If you’re in doubt, vividly recall some event that still riles you until you’re sure enough that you’re suffering from renewed emotion. Or maybe your boredom, anxiety, irritation, or whatever in reading this essay is enough. Now let me ask: Is it completely obvious to you what the character of that experience is? Does introspection reveal it to you as clearly as visual observation reveals the presence of the text before your eyes? Can you discern its gross and fine features through introspection as easily and as confidently as you can, through vision, discern the gross and fine features of nearby external objects? Can you trace its spatiality (or non-spatiality), its viscerality or cognitiveness, its involvement with conscious imagery, thought, proprioception, or whatever, as sharply and infallibly as you can discern the shape, texture and color of your desk? (Or the difference between 3 and 27?) I cannot, of course, force a particular answer to these questions. I can only invite you to share my intuitive sense of uncertainty. (Schwitzgebel 2008, p. 251)

This argument does not appeal to independent evidence in order to motivate pessimism. Rather, it appeals to first-person considerations: introspection itself seems to suggest that there are aspects of our own conscious experience that elude our grasp. As Schwitzgebel puts it, “it’s not just language that fails us—most of us?—when we confront such questions […] but introspection itself. […] in the case of emotion, the very phenomenology itself—the qualitative character of our consciousness—is not entirely evident” (Schwitzgebel 2008, pp. 249–250).

Before examining the force of this argument, let us first consider what kind of pessimism it aims to establish. Does the above passage call into question our capacity to accurately categorize our emotional phenomenology, or is the claim rather that we lack even the capacity to discriminate our emotional experiences from one another and from the rest of our phenomenal states? Although Schwitzgebel’s concern seems to include questions of discriminative access—after all, the passage begins by asking if we can even tell whether or not we have any emotional phenomenology—I take his worries to centre on our capacity to accurately categorize our emotional phenomenology. As I read him, Schwitzgebel’s questions focus on our ability to determine how our emotional experience is structured, both internally and in terms of its relations to phenomenal states of other kinds.

I think that the questions Schwitzgebel raises are difficult to answer. However, it is not clear to me that this fact provides quite as much support for introspective pessimism as Schwitzgebel thinks it does. Lying behind the dumbfounding strategy is the assumption that the questions being posed have determinate answers—that they are appropriate questions to ask. However, I suspect that in an important range of cases this assumption may be unjustified. With respect to the phenomenology of emotion it is natural to assume that the boundaries between the phenomenal states associated with emotion are as clean and sharp as the boundaries between our standard ways of categorizing emotional states. We regard boredom, anxiety, and irritation as distinct emotional states, and we also regard each of these states as associated with distinctive forms of phenomenology. On the basis of these two thoughts we assume that the phenomenal states associated with these categories can themselves be cleanly distinguished from one another. Thus, when one finds oneself at a loss to know whether one is in the phenomenal state associated with boredom, anxiety, or irritation one naturally assumes that the fault lies with one’s introspective capacities. But perhaps the mistake was to assume that the phenomenology of emotion can be cleanly demarcated into states that are uniquely associated with either boredom, anxiety, or irritation. Perhaps the phenomenal states associated with these emotional states overlap and interpenetrate each other. If this were the case, then although there might be certain contexts in which one’s emotional phenomenology is purely that of (say) boredom, there may also be other contexts in which one’s emotional phenomenology involves a complex mix of the phenomenal states associated with boredom, anxiety and irritation. And if one were in a context like this, one might be at something of a loss to know just how to categorize one’s emotional state. The only categories that might come to mind would be those associated with the folk psychology of emotion—<boredom>, <anger> and <irritation>—but these categories might fail to cut the phenomenology of emotion at its joints. In other words, emotional phenomenology may pose a particular introspective challenge not because introspection does a poor job of acquainting us with emotional phenomenology, but because the structure of the phenomenology of emotion fails to map onto the structure of our folk categories of emotions in a straightforward manner.

Other versions of the argument from dumbfounding raise a different set of challenges for introspective optimism. Consider the question of introspective access to visual imagery. Schwitzgebel asks his readers to form a visual image of the front of his or her house, and to then consider the following questions:

How much of the scene are you able vividly to visualize at once? Can you keep the image of your chimney vividly in mind at the same time you vividly imagine (or “image”) your front door? Or does the image of your chimney fade as your attention shifts to the door? If there is a focal part of your image, how much detail does it have? How stable is it? Suppose that you are not able to image the entire front of your house with equal clarity at once, does your image gradually fade away towards the periphery, or does it do so abruptly? Is there any imagery at all outside the immediate region of focus? If the image fades gradually away toward the periphery, does one lose colours before shapes? Do the peripheral elements of the image have color at all before you think to assign color to them? Do any parts of the image? If some parts of the image have indeterminate colour before a colour is assigned, how is that indeterminacy experiencedas grey?or is it not experienced at all? If images fade from the centre and it is not a matter of the color fading, what exactly are the half-faded images like? (Schwitzgebel 2002, pp. 38–39)

I think that this line of questioning poses one of the most significant challenges to optimism. Further, it is doubtful whether this challenge can be resisted in the way that the previous version of the dumbfounding challenge can, for these questions don’t seem to rest on any problematic assumptions. Schwitzgebel isn’t assuming that visual imagery must be pictorial in nature, or that it will always be fully detailed and determinate. Rather, one issue that he explicitly puts on the table is whether the phenomenology of visual imagery can be purely “generic” or “gisty”, or whether it must instead always be specific in some way or another.

But perhaps the dumbfounding challenge can be met in another way. As Jakob Hohwy (2011) has noted, one striking feature of visual imagery is its instability:

In the absence of specific goal parameters for simulations there will be much phenomenal variability because in such conditions subjects must themselves make up the purposes for which they imagine things, or engage in ‘simple’ free-wheeling imagery. For example, there is an indefinite number of purposes for which you can imagine the front of your house (walking up to it, standing close by, assessing its shape, its prettiness, flying around it, how the postman sees is, smelling it, repairing it, buying it, selling it etc), each of these purposes will constrain the imagery, and thus the introspected phenomenology, in different ways. This means that subjects probably do have variable phenomenology, and introspectively report so reliably. (2011, p. 279)

Hohwy’s comments are intended to explain the variability in the introspective reports that individuals give, but they also bear on the dumbfounding argument. Perhaps we are not sure how best to describe the phenomenology of imagery because it is so variable. Imagery experiences cannot be pinned down, but are constantly shifting in response to our own imagistic activity. Precisely how much of the scene we vividly visualize “all at once” depends on the goals that constrain the act of visualization. And, as Hohwy suggests, when we have no such goals our imagery may end up “freewheeling”, such that we move from one state to another. Hohwy grounds his analysis in a predictive-coding account of cognition, but his fundamental point is independent of that theoretical framework and should be fairly uncontentious: imagery surely is more labile than perceptual experience or bodily sensation. No wonder, then, that its phenomenal structure is that much more difficult to articulate.

I have suggested that the optimist has the resources to meet (or at least “problematize”) two of the leading versions of the dumbfounding argument. But suppose that my responses are found wanting, and that the pessimist is able to show that our introspective access to both emotional and imagery phenomenology is insecure and impoverished. Even so, there would be a further question as to how such a finding would motivate global pessimism. It is certainly true that questions about the nature of certain kinds of experiences (e.g., emotional and imagery experiences) strike us as difficult to answer and may leave us flummoxed, but it is equally true that many introspective questions strike us as easily answered. Indeed, as the quotation from Gertler makes vivid, many of our introspective judgments appear to be accompanied by a sense of epistemic certainty. Why should we generalize from the first set of cases rather than the second? Without an account of why certain introspective questions leave us dumbfounded it is difficult to see why pessimism about a particular range of introspective questions should undermine the epistemic credentials of introspection more generally. So even if the threat posed by dumbfounding arguments were able to establish a form of local pessimism, that threat would appear to be easily quarantined.

3.2 Dissociation arguments

A very different case for introspective pessimism is provided by what I call dissociation arguments. Such arguments appeal to a lack of congruence between a subject’s introspective judgments and their capacity to produce reliable first-order judgments—that is, judgments about the objects and properties in their environment. An example of this kind of argument is provided by Schwitzgebel’s treatment of the so-called “grand illusion” (Noë 2002). Most people, Schwitzgebel claims, hold that a broad swathe of their environment—perhaps thirty or more degrees—is clearly presented within visual experience with its “shapes, colours, textures all sharply defined”. Schwitzgebel argues that we have good reason to regard such claims as false. In making the case for this claim, he appeals to an example first popularized by Dennett (1991):

Draw a card from a normal deck without looking at it. Keeping your eyes fixed on some point in front of you, hold the card at arm’s length just beyond your field of view. Without moving your eyes, slowly rotate the card toward the centre of your visual field. How close to the centre must you bring it before you can determine the colour of the card, its suit, and its value? Most people are quite surprised at the result of this little experiment. They substantially overestimate their visual acuity outside the central, foveal region. When they can’t make out whether it’s a Jack or a Queen though the card is nearly (but only nearly) dead centre, they laugh, they’re astounded, dismayed. (Schwitzgebel 2008, pp. 254–255)

How might we explain the dissociation between subjects’ introspective judgments and their first-order judgments? One explanation is that the subjects’ introspective beliefs are false, and that people wrongly take themselves to have detailed visual phenomenology outside of the focus of attention. This is the explanation that Schwitzgebel endorses. But as Schwitzgebel (2008, p. 255) himself notes, it is possible to explain this dissociation by supposing that individuals are wrong not about which phenomenal states they are in but only about the origin of that state. With respect to the card trick example, the proposal is that subjects do indeed have detailed visual phenomenology outside of the origin of attention, but that this phenomenology derives from background expectation rather than environmental input—that is, it is “illusory”.

Schwitzgebel’s account of the dissociation may have more intuitive appeal than the account I have just outlined, but it is not clear how the data furnished by the dissociation argument allows us to choose between them. However, reasons to favour Schwitzgebel’s account can be gleaned noting that the judgment on which we have focused—“thirty or more degrees of my visual field presents itself to me clearly in experience with its shapes, colours, textures all sharply defined”—is available to introspection only indirectly. This judgment is not the direct reflection of any one introspective act, but is a belief about the nature of one’s visual experience that one forms by tracking one’s introspective capacities over time. Call such judgments indirectly introspective. Indirectly introspective judgments can be contrasted with directly introspective judgments—that is, judgments of the kind that one makes in the very context of the card trick experiment, such as “I am now experiencing the shape, colour, and texture of this card (which is presented to me slightly off centre) in sharp detail”. We can now see that although there is a dissociation between the first-order judgments that subjects make and their indirect introspective judgments, there is no such dissociation between their first-order judgments and their direct introspective judgments. Subjects in the card-trick experiment don’t report experiencing the shape, colour, and texture of cards that are presented slightly off centre to them “in sharp detail”—rather, they claim to lack sharp and detailed experiences of such objects. Direct introspective judgments clearly have more warrant than indirect judgments, and thus there is good reason to prefer Schwitzgebel’s explanation of the dissociation over the alternative account.

But although we have found reasons to support Schwitzgebel’s analysis of the dissociation, we have seen that these very reasons undermine his pessimistic attitude to introspection in general, for the evidence in favour of Schwitzgebel’s account involves an appeal to introspection. In other words, the pressure that the dissociation argument puts on indirect introspective judgments assumes that direct introspective judgments are trustworthy. The card trick case does indeed cast doubt on the epistemic security of our background beliefs about our own visual experience, but there is no reason to extend such doubts to include our direct introspective judgments; and it is surely direct introspective judgments that are at the heart of debates about the trustworthiness of introspection. (Indeed, indirect introspection judgments are not really a genuine form of introspection at all.)

Let us turn now to the second step of the dissociation argument: the inference from local pessimism to general pessimism. Suppose that we were to find a dissociation between a certain range of introspective judgments and the subject’s capacity to make the corresponding first-order judgments. Suppose, furthermore, that one could show that this dissociation is best explained by assuming that the introspective judgments in question were false. Would one have any reason to think that introspection in general ought to be regarded with suspicion? Not as far as I can see. It seems to me that our faith in the robustness of introspective access to domains in which such dissociations are not to be found ought to remain completely untroubled by such a finding. In fact, one might even argue that coherence between first-order judgments and (direct) introspective judgments would provide evidence in favour of introspective optimism. If dissociations between a person’s introspective capacities and their first-order capacities can disconfirm their introspective judgments (as the dissociation argument assumes), then associations between a person’s introspective judgments and their first-order capacities ought to confirm them (Bayne & Spener 2010). In other words, the fact that a person’s introspective judgments cohere with their capacity to produce reliable reports of their environment ought to provide us with positive reason to trust those judgments.[10] And a great number of our introspective reports clearly do cohere with our first-order capacities. Although there are cases in which such coherence fails to obtain—for example, Schwitzgebel (2011, Ch. 3) provides a plausible case for the claim that introspective reports of visual imagery are only weakly correlated with the kinds of first-order cognitive capacities that one would expect visual imagery to subserve—such cases are striking precisely because they stand out against the backdrop of coherence that characterizes the relationship between our normal introspective reports and our first-order perceptual capacities.

3.3 Arguments from introspective variation

Perhaps the strongest case for introspective pessimism derives from the phenomenon of introspective variation. Such arguments have as their starting point a disagreement about how best to describe some aspect of phenomenology. Pessimists then argue that the best explanation for the introspective dispute is that at least one of the two groups is mistaken about its own phenomenology, and thus that introspective access to the relevant phenomenal domain is insecure: despite their best efforts, at least one of the two parties to the dispute is wrong about its own phenomenology.

Schwitzgebel (2008) examines a number of arguments from introspective variation, but his central case study concerns a debate about the nature of conscious thought—the so-called “cognitive phenomenology” debate (Bayne & Montague 2011; Smithies 2013).[11] On one side of this dispute are those who deny that thought has a distinctive phenomenal character. Those who hold this view typically allow that conscious thought has a phenomenology of some kind, but they regard that phenomenology as purely sensory—as limited to the phenomenology of inner speech, visual imagery, and so on. We might call this the conservative account of conscious thought, for it treats phenomenal consciousness as limited to sensory aspects of the mind. On the other side of this dispute are those who adopt a liberal conception of conscious thought, according to which conscious thought is characterized by a range of non-sensory phenomenal states—states of “cognitive phenomenology”. It is tempting to conclude that at least one of these two sides is guilty of a fairly radical introspective error: introspection either fails to inform conservatives of a wide range of phenomenal states that they enjoy on a regular basis, or it misleads liberals into thinking that they enjoy a wide range of phenomenal states that they don’t enjoy. Either way, introspection would seem to be untrustworthy with respect to what is clearly a central feature of phenomenology.[12]

But before we follow Schwitzgebel (and many others) in embracing this conclusion, we need to consider alternative explanations of the cognitive phenomenology dispute. One possible explanation appeals to group differences in phenomenology. Perhaps the descriptions of conscious thought that both liberals and conservatives give are right when applied to themselves but wrong when taken to describe conscious thought in general. In other words, perhaps both parties to the dispute are guilty of over over-hasty generalization rather than introspective error.

Although an appeal to group differences might explain (away) some instances of introspective disagreement, it is unlikely to provide the best explanation of the cognitive phenomenology dispute. First, this account requires a degree of variation in phenomenology for which there are few (if any) parallels. This is not to say that phenomenal differences between individuals might not run much deeper than common-sense tends to assume—consider, for example, the phenomenal differences that characterize synaesthesia (Robertson & Sagiv 2005)—but the kinds of phenomenal differences that we already recognize are nowhere near as fundamental as the kinds of differences required by this explanation of the cognitive phenomenology debate, for liberals claim that conscious thought is characterized by a sui generis kind of phenomenology—a kind that is non-sensory in nature. Second, the group difference proposal predicts that there are cognitive and behavioural differences between the advocates of cognitive phenomenology and their detractors that simply don’t appear to obtain. In sum, it seems highly unlikely that the debate about the existence of cognitive phenomenology can be explained by supposing that what it is like to be a liberal is different from what it is like to be a conservative.

But there is another deflationary explanation of the debate about cognitive phenomenology that cannot be so easily dismissed. Perhaps the parties to the debate are operating with very different conceptions of what it would take for thought to possess distinctive phenomenal character, and are thus talking passed each other (Bayne unpublished). On this proposal, liberals are willing to extend the notion of phenomenal consciousness beyond its sensory paradigms in a way that conservatives are not. If this account is right, then the dispute surrounding the existence of cognitive phenomenology is largely verbal. Rather than disagreeing about what introspection reveals, the two sides instead disagree about how the term “phenomenal consciousness” and its cognates ought to be employed.

Why take this proposal seriously? Well, one argument for it is that it would provide a good explanation of why there is such widespread disagreement about the nature of conscious thought—the very terms in which the debate are couched are contested. It is also widely acknowledged that there are different notions of “what it’s likeness” (see e.g., Tye 1996; Flanagan 1992; Georgalis 2005). Although this proposal clearly needs much more defence and development than I can give it here, I think it is not unreasonable to suppose that the disagreement surrounding the existence of cognitive phenomenology might turn out to be largely verbal. At any rate, it seems to me that this account provides at least as good an explanation of the dispute as that which is required by the argument from variation.[13]

There are, of course, other introspective disagreements besides that concerning the phenomenology of thought, and nothing that I have said here goes any way towards showing that they too succumb to a deflationary analysis. Indeed, I suspect that certain introspective disputes—for example, those relating to the richness of visual imagery—may well be best explained by appeal to introspective error. But even if the argument from variation succeeds in establishing a local form of pessimism, it seems to me there is little reason to think that this pessimism generalizes. Indeed, domains that feature disagreement in introspective reports stand out against a general backdrop of introspective agreement. Arguably many domains of consciousness exhibit a great deal of uniformity with respect to introspective reports once individual differences and verbal disputes are taken into account. Now, although inter-subjective agreement doesn’t entail that the individuals in question are right, it does need to be explained, and it seems plausible to suppose that leading explanations of inter-subjective agreement will appeal to the trustworthiness of introspection.