The Fiction of Phenomenality
Keith Frankish comments on Are Feels Real?
Keith Frankish kindly provided some feedback (comments in the margins) on Are Feels Real? I quote them verbatim in somewhat overlapping categories, and then respond to each category. His comments are indented after my summary of the category they (roughly) fall into.
1. Qualities are misrepresentations, not realities. The basic illusionist claim is that experience has no qualities – no phenomenal, qualitative character. The misrepresentation itself has no qualitative content, only propositional. That is, we believe or judge or conceptualize that something in experience has qualities, or have the sense that it does, and thus are disposed to talk about qualities:
…there's an aspect to experience that leads us to talk of it having a phenomenal aspect; w-i-l [what it’s like] talk tracks something (quasi-phenomenal properties; aka 'zero qualia') but we misrepresent (or misclassify) that thing as phenomenal. In my view, this thing is the psychological impact worldly qualities make on us.
I acknowledge that we have a sense (see quote) that experiences have [indefinable but distinctive] characters – meaning that we are disposed to believe they do, etc. I do not acknowledge that such characters actually exist.
…our sense of our experiences having a simple qualitative character is misleading. Introspection leads us to mistake their real character (viz complex patterns of reactions) for simple qualities.
I don’t say that experiences have an indefinable character, just that we have the sense that they do- -- i.e. we are disposed to think and say that they do.
…the primary feature of consciousness is [not qualia but] qualia judgements, reports, and other reactions – what Dennett calls ‘heterophenomenology’
You [Clark] say ‘it’s the illusion of there being qualities that centrally picks out experiences on illusionism’. I don’t think I’m committed to this. I think we recognize experiences straight off, courtesy of introspective mechanisms of some kind. We tend to conceptualize these states as having qualitative feels, but with practice we might be able to overcome that tendency while still being able to recognize the states when they occur.
2. On illusionism, we can still talk about what experience is like, but not suppose this involves qualities, which is a philosophically freighted notion:
Illusionists don’t suggest *giving up* talk of colours (etc) or of what experiences are like. They propose that we adopt better views of what colours are (viz. reflectance or luminance properties) and of what w-i-l talk does (viz. indicates the psychological impact perceived features are making on us). All we need give up are the philosophers’ notions of qualia, phenomenal consciousness, etc.
I'd distinguish here between (1) everyday talk about what experience is like, and (2) philosophers’ notions such as qualia, phenomenal feels, etc. I think the former can be cashed out without appealing to the latter.
Illusionism doesn’t entail that it should be hard to talk about qualities; in fact, it predicts that it will be very natural to talk about them. Compare a theory that offers a psychological explanation for belief in ghosts.
[What it’s like talk is] tracking something (complex reactive patterns), just not what we think it is (pure feels). Hence illusion.
When I say we are not unfeeling, I mean we have feelings in the everyday sense – experiences, emotions, etc. What I deny is that these states involve phenomenal feels in the philosopher’s sense. I think this is clear from the context.
3. What we misconstrue as qualities in experience are properties of mind-independent objects:
Colours, sounds, etc are objective properties of mind-independent objects, though they are identified by their effects on us -- the psychological reactions they evoke.
I think commonsense usage is with me here: colors, tastes, textures, etc. are properties of objects, not of our mind/brains.
I do want to say that colours, etc are properties of objects. I want to say that they are physical properties of surfaces and light sources (reflective and luminance properties). (Parallel claims for other qualities.) That’s what colours *really are*. What I deny (and what I was getting at in the quoted passage) is that there is some further, purely qualitative form of colour *as well*. If we think that there is, that is because we are misrepresenting physical colour in some way. Compare an optical illusion – say, a static image of a spiral which to move when you look at it. This is a misrepresentation. There is no spiral movement anywhere, either in the image itself or in your mind/brain.
4. Frankish’s view is representationalist, but has it that qualities, since they don’t exist, can’t operate as representational contents in terms of which the world appears (my view). Rather qualities are misrepresented as existing in the world (in particular, in consciousness), just as we might misrepresent ghosts as existing:
I too take them [qualities] to be real as representational contents/ intentional objects (see my 2016 paper). Ghosts are real in this sense too (people have beliefs about them). In fact, this is a central claim of illusionism: phenomenal properties – like other illusory properties -- exist only as intentional objects of our mental attitudes.
If they [qualities] are not instantiated anywhere, then shouldn’t we eliminate them from our account of the world? (I suppose we might say that uninstantiated properties exist as abstract property types, but we won’t need to mention them in our scientific theories).
Qualia, like other illusory features, exist in the represented world, not the real world. Ghosts too are representational realities for people who believe in them.
What’s prior [to inquiry about the nature of qualitative character] is our belief in such a character. The belief may be false.
5. Illusionism objectifies sensory experience as identical to observable psychological reactions:
My pain is subjective in the sense of belonging to my body of course, but it's an objective feature of it. Colours, sounds, etc. are objective properties of mind-independent objects, though they are identified by their effects on us -- the psychological reactions they evoke.
It’s like thinking you can observe the university as something separate from all the university buildings. Similarly, you can’t observe the feel separately from all the psychological reactions since feel talk is a way of gesturing at the reaction pattern.
6. Frankish’s illusionism is motivated by pessimism about how physicalist theories, e.g., higher order representationalist theories, can explain phenomenal consciousness:
I’m an illusionist because I can’t see how such theories can explain the existence of phenomenal properties themselves. (Whereas I can see how they might explain our belief in phenomenal properties – i.e. because they misrepresent complex patterns of brain activity as simple qualitative states).
The big problem is giving any account of what these qualities [are] and how we know them. A physicalist is going to end up talking about something like internal monitoring of brain processes (because that’s all there is to talk about) and then the view will be pretty much indistinguishable from illusionism.
The Reality of Phenomenal Consciousness
Tom Clark reply to Frankish comments on Are Feels Real?
Introduction. In his comments above on Are Feels Real? Frankish points out that I was wrong to suppose that indefinable but distinctive characters of sensory experience might survive under illusionism. There are no such characters, he says; we only have the “sense” that there are (see part 1 of his comments above). This seems right since the idea of experiential character is pretty much that of a quality, and there are no qualities in experience according to illusionism. And in fact, at the end of Are Feels Real? I suggest there can’t really be experiential character on illusionism: “And perhaps that’s the logical end point for illusionism: not only are there are no private mental qualities, there are no indefinable but distinctive characters cooked up by introspective consciousness, only propensities to talk about such things.” So feels, on illusionism, are simply psychological reactions, resulting in non-qualitative, mistaken beliefs and judgments to the effect that there are qualities in experience. Even though introspection represents psychological reactions as being qualitative – it “bundles it [the complex physical features of sensory states] all together, representing it as a simple, intrinsic phenomenal feel” (Frankish 2016) – there is nothing qualitative (phenomenal) about the conscious illusion of there being qualities, which is the content of that (mis}representation. That something appears or seems qualitative seems necessarily to involve a quality, but it doesn’t, according to illusionism. There is nothing qualitative anywhere.
In what follows I respond to Frankish’s comments as set forth in the six somewhat overlapping categories above, trying to motivate realism about qualities. My somewhat overlapping responses are that 1) qualities are pretheoretically real elements of sensory experience, 2) the very natural talk of qualities in describing what experience is like doesn’t involve a philosophical posit that makes such talk fictional, 3) qualities are properties not of the world, but of experience, 4) qualities are possibly representational contents, not uninstantiated intentional objects, 5) illusionism wrongly objectifies consciousness, and 6) Frankish’s illusionism about phenomenal consciousness is motivated by an unwarranted pessimism about ever accounting for qualities.
1. The pretheoretical reality of experiential qualities. Frankish says that there’s nothing qualitative about consciousness, but on the face of it – prior to theorizing about consciousness – sensory experience is composed of a host of differentiable qualities (aka “qualitative feels”) combined into a conscious perceptual gestalt. Some of these, such as red, or sweet, are basic in that they aren’t subjectively decomposable into further qualitative elements. Here are some dictionary definitions of “quality”: “the attribute of an elementary sensation that makes it fundamentally unlike any other sensation,” a “peculiar and essential character,” and “an inherent feature” (Miriam Webster); “a characteristic or feature of someone or something” (Cambridge English Dictionary); “an essential or distinctive characteristic, property, or attribute: the chemical qualities of alcohol,” and the “character or nature, as belonging to or distinguishing a thing: the quality of a sound” (Dictionary.com).
These definitions show that the notion of a quality and qualitative character, in particular the various qualitative feels of sensory consciousness, is an uncontroversial, everyday concept we apply when describing our experiences. Illusionism says such descriptions are radically mistaken. But were we to disallow reference to experiential qualities or character, we’d have to find another term to refer to the distinguishable and subjectively irreducible feels of such things as seeing red, feeling pain, tasting sweet, etc. The existence of such discriminable characteristics in experience, whether simple or complex, cannot be dislodged by illusionism, which is a hypothesis about the nature of consciousness. What is in question is how we account for qualities – what their fundamental nature is. We might hypothetically conceptualize them as being illusory in some sense, or physical, or non-physical, or the intrinsic nature of what physics describes (panpsychism), or as representational contents (my current rather standard view), but their existence in experience prior to claims concerning their nature is not an illusion. The illusionist hypothesis might work if it granted the prima facie pretheoretical existence of qualities as experiential elements and then showed that they don’t have some characteristics we suppose they have. Instead, it denies the existence of qualities as possible targets of explanation or revision by saying there is nothing that even appears qualitative about sensations – they only seem to appear qualitative; or, if we grant such an appearance, illusionism says that appearance doesn’t really involve anything qualitative. But it’s difficult to explain on what basis we pretheoretically and uncontroversially describe experiences as involving qualities if nothing like a quality ever appeared to us.
2. Talk of qualities in describing what experience is like is ordinary, not philosophical. Frankish says illusionism predicts that it would be natural for us (philosophers and non-philosophers) to talk about qualities in saying what experience is like, but what we’re really referring to (tracking) in such talk is psychological reactions that we misrepresent as being qualitative. As noted in the first section above, the concept of a quality, including those present in experience, is commonsensical and ineliminable as an everyday descriptor of experience. Frankish asks us to abjure the “philosopher’s sense” of phenomenal feel, of phenomenality, of qualia, since he says that sense is metaphysically loaded in favor of dualism and likely doesn’t refer. We can do this, but the very natural talk of qualities in describing what experience is like needn’t be, and isn’t, pegged to that sense, which might make such talk fictional. Rather, we can be realists about qualities in the ordinary, metaphysically neutral sense in which sensory experience has qualitative character (see the definitions in section 1 above). Qualities are not in the first instance theoretical posits that seem anomalous from a scientific worldview (Frankish 2016), but a pretheoretical reality in need of explanation. Put otherwise, “feelings in the everyday sense – experiences, emotions, etc.” (my emphasis) uncontroversially possess qualitative character, so referring to qualities in saying what experience is like doesn’t involve anything contentiously philosophical or possibly fictional.
3. Qualities are likely representational properties of experience. Frankish grants the existence of colors, sounds, smells, tastes, pains, and other sensory feels but says they are not qualitative characteristics of experiences, but non-qualitative properties of the physical world: “properties of mind-independent objects, not of our mind/brains.” There is nothing qualitative about physical color and other sensory feels since qualities are illusory: there is no “purely qualitative form of colour” which distinguishes red from blue in our experience of them. This means there is nothing qualitative anywhere, either in experience or in the world.
Against these claims, we can see that in dreams, hallucinations, and afterimages there are no mind-independent objects present which could have non-qualitative properties, in which case the colors, sounds, etc. involved in such conscious episodes have to be properties of our mind/brains: something identical to neural processes or perhaps something associated with such processes, e.g., representational content, whether veridical or not. And as argued in section 1, these experiences, as well as our waking experiences of mind-independent objects, really do involve qualities: red is qualitatively distinct from blue; the apple really appears qualitatively red in our experience, whether in a dream or in waking life. That quality no doubt has to do with properties of the apple and its context as reacted to by us as perceivers in normal conditions, but it’s perfectly real as a property of our sensory experience, not an illusion, even if it turns out we’re hallucinating the apple. The way to explain the fact that experiential qualities aren’t discoverable (observable) in the brain or the world is, I suggest, to hypothesize them to be representational contents which, as a general rule, we don’t find in the world they participate in representing (Clark 2019). If indeed there’s nothing qualitative about consciousness, then it’s a mystery why we have an ordinary, pretheoretical conception of quality that we routinely deploy in descriptions of sensory experiences and objects as represented by such experiences.
4. Qualities aren’t like ghosts, but real representational contents. On the one hand, Frankish holds that we misrepresent qualities as actually existing, as we might ghosts, goblins, or unicorns, but on the other agrees that qualities are real as representational contents. The way I see it, representational contents are real (they exist as occurrent contents of representations) whether or not what they pick out in the world exists or not. The experiential content of my hallucinating a red apple is real, but the apple is not. My false belief in ghosts is a real, contentful belief, but ghosts don’t exist. On Frankish’s view, qualities don’t exist as experiential contents that could misrepresent the world, rather we misrepresent them as existing, period. On illusionism there’s nothing qualitative about representational content, nothing phenomenal that might veridically correspond to an object (e.g., the redness of the apple) when we consciously represent reality. So when Frankish says qualities are real as representational contents, he means only that they are the intentional object of our (false) belief that they exist, just as ghosts are the intentional object of belief in ghosts. Content for him is only that which might exist in the world as the referent of belief, what it’s about. On illusionism, since sensory consciousness lacks qualities it is strictly propositional – it only involves conceptual content as in beliefs and judgments about the world. Qualities for Frankish are what we mistakenly believe or judge or conceptualize to exist as features of consciousness, whereas on my view they are real representational contents, prior to belief and judgment and presented in a qualitative format, which may or may not correspond to external objects. There’s nothing ghostly about them since they are the untranscendable qualitative terms (our phenomenal representational reality) in which reality (represented reality) appears so concretely to us. If sensory experience weren’t qualitative, it would indeed be ghostly.
5. Illusionism objectifies consciousness. Since on illusionism feelings are nothing over and above physically instantiated psychological reactions, they are in principle publicly observable goings-on. In looking at the various neural processes that on illusionism are the psychological reactions that constitute my pain, you’d be literally seeing my pain since there is nothing separate from them that is pain. True, in observing those reactions it isn’t your body that’s reacting, so in that sense only I embody the feel of pain as it’s defined on illusionism – a physical criterion of subjectivity. But there’s nothing private about physical embodiment, and on illusionism there’s no qualitative feel of pain that only I undergo, that only I experience. The painfulness of my pain on illusionism is nothing that I have privileged access to since it’s identical to physical happenings in principle accessible to anyone. Experiential feels are thus objectified on illusionism, as they are on any physicalist account of consciousness, but physicalism about feels needs to be established before we can confidently say there’s nothing categorically subjective (private) about consciousness. In the meantime, pretheoretical realism about qualities – the mainstream starting point – takes such subjectivity as a datum to be explained: you can in principle see my brain in action, but you can’t see the red involved in my experience of the apple, so are left wondering whether it’s like your red or not. You can’t and will never feel my pain (and the feel is its essential nature) because the qualities of sensory experience are categorically private phenomena.
6. The explanatory gap motivates illusionism. Frankish says that he’s an illusionist because he can’t see how physicalist theories could explain phenomenality, and indeed it isn’t yet clear how they could. In schematic form his illusionism might look like this:
P1: Physicalism is likely true.
P2: Physicalism can’t account for qualities, which are scientifically anomalous.
C: Therefore, qualities don’t exist but are illusions.
Since both P1 and P2 can be questioned, the current explanatory gap need not and should not lead us to skepticism about the defining feature of sensory experience: that it has qualitative character. Some physicalists, such as Hakwan Lau (see his book In Consciousness We Trust), David Rosenthal, and Richard Brown, take a higher order representational approach similar to Frankish’s; but although they admit their explanations are incomplete, they don’t opt for illusionism. This seems to me the more conservative, patient stance to take, one that respects the manifest reality of the essential character of consciousness prior to theorizing about it. And simply because we can’t locate phenomenal properties in the brain or in the world doesn’t seem to me a reason to eliminate them. Understanding qualities to be representational contents might help explain why they are not locatable, not objectifiable, and thus categorically subjective (Clark 2019). This hypothesis may not pan out, but, alongside other hypotheses, physicalist or otherwise, it merits investigation before we declare the primary explanandum in consciousness studies non-existent.
Conclusion. Frankish might reply to all of the above by saying that of course I and most folks, including perhaps the majority of philosophers, believe we entertain qualities in experience, but this belief is a misrepresentation courtesy of faulty introspection. It might seem that something in experience has a qualitative appearance, but there’s nothing qualitative about how it appears. We can still talk of what experiences are like – how they feel – but we should abandon realism about qualities: there is nothing simple, basic, irreducible, or private about red, or pain, or any sensory experience when we undergo it. Note, however, that this claim follows from a defeasible hypothesis about the nature of consciousness, not a settled, proven theory sufficient to dislodge commonsense pretheoretical realism about qualities. Illusionism solves the hard problem by suggesting how phenomenal consciousness could be illusory, given certain sorts of representational mechanisms. It claims to explain why, if it is true, we only believe, judge, or conceptualize that qualitative feels exist, and thus continue to talk about them. But realism about qualities also can explain why we believe, judge, or conceptualize them to exist. Such realism comports with pretheoretical commonsense and ordinary language but of course raises the difficult question of accounting for phenomenal properties, what Frankish thinks can’t be done within a physicalist framework. Only time will tell if he’s right about that, and of course there are other explanatory options besides a physicalism that objectifies consciousness. Explanatory pessimism should not lead us to endorse the illusionist claim: that experiential qualities don’t exist to be explained, that there’s nothing qualitative about how the world appears as presented in conscious experience.
November 1, 2022
 “Of course, most people do not regard their phenomenology as illusory…” (Frankish 2016)
 “Illusionists can say that one’s experiences are like something if one is aware of them in a functional sense, courtesy of introspective representational mechanisms.” (Frankish 2016)
 “In general, apparent anomalousness is evidence for illusion. If a property resists explanation in physical terms or is detectable only from a certain perspective, then the simplest explanation is that it is illusory.” (Frankish 2016)