Redefining the Class of Qualitative States—A Reply to Shoemaker

The so-called qualia-type objections to functionalism seem to imply that some qualitative states are not functionally definable (or identical to some functional state type). In “Functionalism and Qualia,” Sydney Shoemaker concedes that functionalists can allow for some types of mental states to be functionally undefinable without committing themselves to a view that cannot account for the class qualitative states. If qualitative states are construed as a relation of qualitative similarity, Shoemaker argues that qualia are functionally definable, and thus do not pose a serious problem for functionalists. In this paper I argue that (i) Shoemaker’s argument against the possibility of absent qualia is untenable, (ii) if cases of absentqualia are possible, then Shoemaker’s reconciliation fails, and (iii) even if his reconciliation succeeds in functionally defining the class of qualitative states, it still fails to account for qualitative states being capable of existing independently from functional characterization. The so-called qualia-type objections to functionalism seem to imply that some qualitative states are not functionally definable or identical to some functional state type. So, on the assumption that an attractive theory of mind must account for the qualitative character of mental states, Functionalism fails because it cannot functionally account for the role of qualitative states. In “Functionalism and Qualia,” Sydney Shoemaker concedes that functionalists can allow for some types of mental states to be functionally undefinable without committing themselves to a view that cannot account for the class qualitative states. If qualitative states are construed as a relation of qualitative similarity, Shoemaker argues that qualia are functionally definable, and thus do not pose a serious problem for functionalists. In Section I of this paper I will explain the version of functionalism endorsed by Shoemaker, two prominent versions of qualia-type objections, and how Shoemaker’s argument supposedly rebuts these objections. In Section II, I argue that (i) Shoemaker’s argument against the possibility of absent qualia is untenable, (ii) if cases of absentqualia are possible, then Shoemaker’s reconciliation fails, and (iii) even if his Res Cogitans (2013) 4 Pratt | 53 2155-4838 | commons.pacificu.edu/rescogitans reconciliation succeeds in functionally defining the class of qualitative states tailored to visual experiences, it still fails to account for qualitative states being capable of existing independently from functional characterization.

The so-called qualia-type objections to functionalism seem to imply that some qualitative states are not functionally definable or identical to some functional state type.So, on the assumption that an attractive theory of mind must account for the qualitative character of mental states, Functionalism fails because it cannot functionally account for the role of qualitative states.In "Functionalism and Qualia," Sydney Shoemaker concedes that functionalists can allow for some types of mental states to be functionally undefinable without committing themselves to a view that cannot account for the class qualitative states.If qualitative states are construed as a relation of qualitative similarity, Shoemaker argues that qualia are functionally definable, and thus do not pose a serious problem for functionalists.
In Section I of this paper I will explain the version of functionalism endorsed by Shoemaker, two prominent versions of qualia-type objections, and how Shoemaker's argument supposedly rebuts these objections.In Section II, I argue that (i) Shoemaker's argument against the possibility of absent qualia is untenable, (ii) if cases of absentqualia are possible, then Shoemaker's reconciliation fails, and (iii) even if his 2155-4838 | commons.pacificu.edu/rescogitansreconciliation succeeds in functionally defining the class of qualitative states tailored to visual experiences, it still fails to account for qualitative states being capable of existing independently from functional characterization.

I
Shoemaker advocates what he calls "functionalism in the broad sense," which characterizes mental states in terms of their casual relations to inputs, outputs, and other mental sates.For example, the mental state of 'being in pain' is defined by its causal role or relation to other variables, such as physiological damage and subsequent mental states (e.g., the belief that one is in pain, the desire to assuage the pain, beliefs about what action will stop the pain, and so on). 1 More specifically, as Shoemaker stipulates, for some mental state 'pain' there is a functional state identical to it if and only if it shares (1) its tendency to influence behavior in observable ways, (2) its tendency to produce the belief that one is physiologically damaged, and (3) its tendency to give rise to qualitative beliefs about that pain.For example, the qualitative belief 'I have a headache' is given rise by a functional state that has the qualitative character constituting 'a headache' (namely, the pain identified as being in the head, its tendency to throb, to be exacerbated by light, sound, etc.).
Before explaining how qualia-type objections pose a problem for functionalists, I will will briefly cash out what qualia are.Thomas Nagel famously coined the notion of qualia as the 'what-it-is-like' component of psychological states, insofar as some organism or being has 'phenomenally conscious states' if and only if there is something that 'it is like' to be that organism. 2More specifically, qualia refers to those properties of mental states-in particular of sensations and perceptual states-which determine or how it feels phenomenologically to be in a particular mental state.E.g., the mental state of pain has some qualitative character insofar as a 'stub your toe pain' is qualitatively different from, for example, a 'burn your face pain', in such a way that you could recognize and discern the differences between them (namely, that one burns in the 'searing' sense, whereas the other throbs, aches, and so forth).Though the point has been argued, qualia are considered to be non-intentional or non-representational, and moreover, fundamentally intrinsic. 3th these terms briefly explained, I will now lay out two prominent versions of the socalled qualia objections-namely, the Absent Qualia and Inverted Qualia Objections.The Absent Qualia Objection (often argued for via the 'zombie hypothesis' or the 'China-brain argument') rests on the assumption that the functional organization characteristic of mental or conscious states can be instantiated in a variety of physical systems-presumably, even systems that are radically dissimilar from humans 4 .In recent discussion, the zombie hypothesis has garnered the most attention in the absent qualia debate.A "philosophical zombie" refers to a being that is functionally and 2155-4838 | commons.pacificu.edu/rescogitansbehaviorally indistinguishable from us, yet lack qualia (or qualitative experiences).Since zombies would presumably instantiate the cognitive system requisite for a functional "mind", zombies would thus be the subject of functional states.But, the objection concludes, a zombie-while having a functional organization equivalent to the human mind-would by definition not have the qualitative character that ours do, and thus functionalism fails since it cannot account for the qualitative aspect of psychological states. 5e Inverted Qualia Objection posits cases in which subjects suffer from spectrum inversion.Spectrum inversion generally comes of two varieties: (1) intersubjective inversion and (2) intrasubjective inversion.Let's start with (1).Coined by John Locke in his Essay Concerning Human Understanding, the idea of intersubjective spectrum inversion suggests the possibility that "the same object should produce in several men's minds different ideas at the same time; for example, the idea, that a violet produces in one man's mind by his eyes, were the same that a marigold produced in another man's, and vice versa" (Locke, 1960).I.e., though two individuals-say Smith and Blackshare the same exact functional organization, the sensation that Smith has when visually perceiving the color red is phenomenally the same as the sensation Black has when perceiving green, and vice versa.Thus, because qualia are non-intentional, the visual experiences of Smith and Black when viewing, say, a ripe tomato, would be similar in their intentional features (namely, of a red, round, bulgy surface), but would have dissimilar qualitative character.
Wittgenstein, more than 200 hundred years later, modified the spectrum inversion problem by considering a case of intrasubjective inversion. 6Intrasubjective spectrum inversion considers the possibility of an individual waking up to find that everything that they typically identified as green now, inexplicably, appears red.To clarify, what essentially distinguishes intrasubjective inversion from intersubjective inversion is their detectability by the subject(s).That is, it's possible that two individuals suffering from intersubjective inversion would perceive a color's intentional features similarly (while having fundamentally dissimilar qualitative character) such that the qualitative difference would not be detectable through behavior or verbal identification. 7By contrast, a case of intrasubjective inversion would presumably be detectable by the subject due to memory (i.e., they remembered that ripe tomatoes usually appeared to them as red, but now appear to them as green).Nonetheless, it seems that both versions of the Inverted Qualia Objection pose problems for functionalism.Namely, that a particular mental state could be functionally isomorphic (identical) to another mental state, even though both states are phenomenally or qualitatively different, and thus functionalism fails because it cannot account for such cases.Some have suggested that to successfully meet this objection the functionalist need only respond that an organism's mental state having qualitative character is irrelevant 2155-4838 | commons.pacificu.edu/rescogitansto whether that being actually is in some psychological state. 8However, as noted by Block and Fodor, once the functionalist concedes that qualitative character is essentially irrelevant to some functional state, it is difficult to resist the implication that such a state could exist without any qualitative content at all. 9 Hence, it may be nomologically possible for two mental states to be functionally identical, even if one of the two states has no qualitative character at all.This thus implies that an organism could be in a pain state even if it literally feels nothing at all.As Shoemaker construes it, opponents of functionalism seem to be committed to the view that qualitative states cannot be functionally defined.So, the objection runs, a psychological state that is not functionally definable cannot enter into definitions or analyses of other functional mental states-and thus Functionalism is either false or incomplete.In sum, it is to this problem that Shoemaker tailors his response.For, if the objection hinges on the fact that qualitative states cannot be functionally defined, Shoemaker need only demonstrate that, in some respect, qualitative states can be functionally defined to meet the objection.
To reconcile functionalism with spectrum inversion, Shoemaker concedes that (assuming the possibility of cases like spectrum inversion) even if particular qualitative states cannot be functionally defined, the 'class of qualitative states' can be.To clarify, a particular qualitative state would refer to, for example, the individual instance of a 'green quale' in a single experience-whereas the class of qualitative states corresponding to 'green qualia' would contain all instances of qualia identified as a 'green quale'.Hence, to reconcile functionalism and spectrum inversion, Shoemaker functionally defines qualia classes in such a way that the possibility that they might invert doesn't matter.To illustrate, there are 'sinus headache qualia' and 'migraine headache qualia'.For normal cases of mind (non-inverted subjects), the first of these classes are associations with certain functionally describable comparative judgments (e.g., "this feeling or qualia is worse than that one").The second of these classes corresponds to associations with particular functionally describable judgments comparative to other subjects (e.g., Smith's headache qualitatively feels like a 'dull ache between the eyes and around the nose', whereas Black's feels like a 'throbbing, one-sided head pain').Thus, these two qualia could be inverted, but Shoemaker thinks that this doesn't matter; the two classes will have been defined functionally in such a way that even if a subject suffers from inversion, the result will be functionally indistinguishable behavior when compared with a "normal" non-inverted subject.In Shoemaker's terms, qualitative similarity and difference is to be defined in terms of (1) a subject's experiences being qualitatively similar or different in particular ways, (2) the subject's believing in the existence of certain types of objective similarities or differences in the world, and (3) the subject behaving in a certain way, such as forming qualitative beliefs. 10This relationship between experience and judgments of qualitative similarity is thus the kind of "functional" role that Shoemaker has in mind for qualitative states.In doing so, Shoemaker claims to have demonstrated that 2155-4838 | commons.pacificu.edu/rescogitanspsychological states having qualitative character does not, in fact, pose a problem for functionalism.

II
With all the preliminary moves laid out, I will now consider some problems facing Shoemaker's reconciliation, namely: (i) If 'the causal account' of knowledge is false or problematic, then his argument against the possibility of absent qualia is untenable.(ii) If cases of absent qualia are possible, then Shoemaker's solution to the Inverted Qualia Objection fails.(iii) Even if his reconciliation succeeds in functionally defining the class of qualitative states, it still fails to account for qualitative states being capable of existing independently of functional characterization.I.e., qualia can be described in non-functional terms, and so there is a language for describing qualia that is not functional, and this fact is not accounted for by Shoemaker's view.
Let's start with worry (i).The possibility of absent qualia cases has been and continues to be widely controversial; and an independent argument for their possibility would require more time than I have here. 11Rather than developing an independent argument for the possibility of absent qualia, I will instead show that Shoemaker's argument for their impossibility is untenable.Shoemaker-while allowing for the possibility of inverted qualia-thinks that cases of absent qualia are logically impossible.Shoemaker's argument against the possibility of absent qualia maintains that if cases of absent qualia are possible, then a mental state's having qualitative character would be irrelevant to whatever causal consequences followed from it.So, assuming a causal theory of knowledge, one could not have knowledge of a mental state's qualitative character.But Shoemaker thinks this is mistaken, for it seems obvious that we do have knowledge of a mental state's qualitative character.Thus, he concludes that absent qualia are impossible.Explicitly constructed, Shoemaker's argument takes the following form: 1. Given some subject S, where p = 'S is the subject of qualitative states'; S knows p if and only if the fact that p is causally connected in an appropriate way with S's believing p. 12 2.If cases of absent qualia are possible, then S knowing p is irrelevant to p being causally connected in an appropriate way with S's believing p. 3. So, given that premises 1 and 2 are true, S could not have knowledge of p. 4. But, because S does have knowledge of p, absent qualia are therefore impossible.
2155-4838 | commons.pacificu.edu/rescogitansSince Shoemaker's argument assumes a casual theory of knowledge (premise 1), it seems that his argument cannot go through if the causal account of knowledge is false or at least problematic.In his book Epistemology, Richard Feldman (2003) spells out the problems facing the causal theory of knowledge by offering various counterexamples, the most pertinent of which are cases involving perception and evidence.The following case is the counterexample considered by Feldman: "Trudy and Judy are identical twins.Smith sees one and, for no good reason, forms the belief that he sees Judy.It is true, and it is a case of perception.He reconstructs the causal chain between Judy's presence and the belief properly.He knows about Trudy, but rashly discounts the possibility that she is the one he sees" (Feldman, 2003).This type of case is problematic for the causal theory of knowledge because there is a causal connection between a fact (the fact that Smith sees Judy) and a belief in that fact (Smith's belief that he sees Judy), but Smith nevertheless lacks knowledge.This is because Smith lacks justification for the belief he has formed-for Smith does not have good reasons for believing that he sees Judy rather than Trudy, and thus he arrives at the (true) fact merely by accident.To clarify, consider this case in more explicit terms: Where p = 'I see Judy' and p* = 'I see Trudy'; Smith knows p if and only if the fact that p is causally connected in an appropriate way with Smith's believing p. Smith sees one of the twins and rashly forms the belief p.As it turns out, p is true-the twin Smith sees is in fact Judy, and thus there is a causal connection between p and his believing p.However, given that Smith knew that Judy and Trudy were identical twins, Smith had no good reason for believing p rather than p*-even if there is a causal connection between p and Smith's belief in p-and so Smith does not actually have knowledge of p because his belief lacked justification.In sum, the existence of a causal connection is not good enough to count as knowledge if knowledge requires truth, belief, and proper justification.So premise 1 of Shoemaker's argument is either false or at least problematic, and thus his argument against the possibility of absent qualia is untenable. 13ven that knowledge of qualia or qualitative states would fall within the realm of perception (perhaps an inward perception), this type of problem motivates rejecting Shoemaker's use of a causal account of knowledge if the aforementioned case is considered applicable in the relevant respects. 14If problems with a causal theory of knowledge are convincing, Shoemaker's argument against absent qualia loses its force.Moreover, if his argument cannot go through, it seems reasonable to give cases of absent qualia the benefit of theoretical doubt; and with cases of absent qualia back on the table, I argue that Shoemaker's reconciliation fails.This is the essence of worry (ii).For, it seems like any external input for a philosophical zombie would functionally define that input's corresponding qualitative character in terms of its similarity to other 2155-4838 | commons.pacificu.edu/rescogitans(causal-historical) equivalence classes.However, if a Zombie's qualitative states are defined in terms of their similarity to past qualitative states, then all of its qualitative states would be (in terms of equivalence classes) qualitatively indiscernible, for any experiential input would instantiate the same phenomenal property-that being, the absence of qualitative character.
To illustrate, consider a mind that is functionally identical to a zombie's but does have qualitative content.This "normal" mind could functionally account for the qualitative character of the class of qualitative states identified as 'headaches'-insofar as it could discern between, for instance, the phenomenological properties of a 'sinus headache' and a 'migraine'.However, a zombie mind would not even be capable of distinguishing between distinct equivalence classes-such as the phenomenal properties of visual experience and the qualitative properties of pain-let alone between two qualitative states that fall within the same equivalence class, such as migraines or sinus headaches (within the equivalence class 'headache').
Cashing it out in Shoemaker's terms, it seems like a zombie could evaluate visual sensations or inputs in terms of (1) its experiences being similar or different in particular ways, (2) its believing in the existence of certain types of objective similarities or differences in the world, and (3) its behaving in a certain way.But, since by definition a zombie's mental states would not have intrinsic qualitative character like a "normal" mind, it seems that Shoemaker's solution to the Inverted Qualia Objection fails if cases of absent qualia are possible.
With worries (i) and (ii) laid out, I will now sketch out worry (iii).The essence of worry (iii) is this: even if functionalism can plausibly account for the class of qualitative states, it still doesn't account for qualitative states being capable of existing independently from functional definition.As mentioned earlier, it has been argued that qualia are non-intentional or non-representational, which seems to suggest that qualitative states (or phenomenal properties) can not only exist independently from, but can come apart from and completely outrun their representational properties and hence functional characterization.To clarify, the claim that qualitative states or phenomenal properties can "come apart from their representational properties" or "exist independently of functional characterization" is a feature of the antirepresentationalist thesis.If this thesis is true, two things follow: (1) there could be pairs of experiences with identical representational properties but with different qualitative properties, and (2) there could be pairs of experiences with the same phenomenal properties but with different representational properties.Hence, on this view, the phenomenal or qualitative properties of an experience can "come apart from" or "exist independently of" representational properties. 15 4 his assumption is couched in the principle of multiple realizability, which maintains that mental states can be instantiated or realized in a different kinds of beings (e.g., humans, computers, martians, etc.).For more on how this principle relates to Functionalism, see Churchland (1988). 5Before advancing to the Inverted Qualia Objection, I think it's important to note here that Shoemaker-while allowing for the possibility of inverted qualia-thinks that cases of absent qualia are logically impossible.Shoemaker's argument against the possibility of absent qualia maintains that if cases of absent are possible, then a mental state's having qualitative character would be irrelevant to whatever causal consequences that followed from it.So, assuming a causal theory of knowledge, one could not have knowledge of a mental state's qualitative character-but since we do have knowledge of qualitative states (knowing, for example, that it feels like one is in pain), absent qualia are impossible.While, for the sake of time, I will not offer an independent argument for the possibility of absent-qualia, it is worth mentioning that Ned Block (1980) and David Chalmers (1996) have argued for their possibility.