Click to read: OnDennett
This paper will explain what qualia are and why Dennett thinks that it is a confused concept without a referent. Subsequently, it will spell out how Dennett substitutes qualia for public, extrinsic properties that relate to one’s certain internal states and acts; this sounds curiously like direct realism. This paper will conclude by explaining that Dennett’s argument provides good reasons to take direct realism seriously, but that it does not support the materialism in eliminative materialism and so does not dissuade us from most forms of dualism.
Qualia can be thought of as the immediate properties of phenomenal experience. Nothing is more ordinary than qualia; it is the way that something appears to you (226), the way that the letters look as you read this paper. It is the pain you feel when that first sip of coffee burns the roof of your mouth. According to Dennett, qualia are characterized by four features (229):
- Qualia are directly or immediately apprehensible in consciousness. This means that through introspection I can apprehend that I’m seeing some yellowish over there, some bluish over here – for it is just evident in my conscious experience. I am directly aware of the state of my qualia.
- Qualia are private. This is to say that my experiences and qualia cannot be compared with others, for any comparison must rely on intersubjective report and there is no “machine” that we could check our qualia against for objective comparison (for no machine could detect what we mean by qualia) (228).
- Qualia are intrinsic. This means that they are atomic and unanalyzable (229); you cannot break down and analyze a unit of qualia. There are no further constituent properties that can be used to describe qualia; they are simple.
- Qualia are ineffable, partly in virtue of their intrinsicality. To say they are ineffable is to say that one cannot describe precisely what the state of one’s qualia is, no matter how fanciful and determined a description. For instance, you could never adequately characterize what red is, and the experience of seeing it, to a colorblind person, no matter how much describing you do. (Indeed, you could never adequately describe red to anyone. For even if they agreed to your description, you still couldn’t know whether your red was their red as opposed to their green, for we all came to learn and use color words the same way.)
Dennett will argue that this conception of qualia is empty and confused and that, in fact, there are no qualia at all. Here’s part of Dennett’s motivation. He characterizes his position on qualia as that of “eliminative materialism” (244), an alternative to reductive materialism. Accounting for qualia is the typical stumbling block for materialism (229). If there are no qualia, then this is no longer an issue for the materialist; materialism should enjoy more success and dualism should fall into disrepute. Dennett argues for this through a series of intuition pumps aimed to show that the immediate properties of our experience, our qualia, are not directly apprehensible, private, intrinsic, or ineffable.
Section 1. How Dennett quines qualia.
1.1 The argument that qualia are not directly apprehensible.
Consider the following (231). An evil neurosurgeon performs an operation on your brain in your sleep. When you wake up, you see that the grass is red, the sky is yellow, etc. Initially upon introspection, it seems as though you have undergone qualia inversion. But this is not necessarily so, for there are two ways that the neurosurgeon could have inverted your experience. (1) The inversion process occurs in a “qualia-producing” channel, like your optic nerve. (Presumably, the qualia “comes after” light striking your eye. With the eye inverted, the qualia produced from it are inverted.) Or, (2) the inversion process does not occur in a qualia-producing channel, but rather in your memory access pathways. (Such that qualia are not inverted, but memory-anchored dispositions to react to them are. You remember a different spectrum of qualia; even though there has been no change in your qualia, it seems as though there is when you check against your memory.) Noticing a difference in your conscious experience, therefore, does not entail a difference in qualia (231). So I cannot know the state of my qualia through introspection, because I will never be able to determine whether the shift in my conscious experience is the result of qualia inversion or memory inversion. So the state of my own qualia is as unknowable to me as it is to any other observer. But if I cannot determine the state of my own qualia, then how is it that my qualia can be directly apprehensible to me? It seems that they can’t. Or consider: am I running a fever or just feeling feverish (or some combination thereof )? I can’t tell by introspection. Perhaps I could consult a thermometer to determine the state of my qualia, but then I cannot say that my qualia were directly apprehensible to me as I must conduct some empirical test. (And empirical tests have their own challenges, as will be discussed in 1.2)
1.2 The argument that qualia are not private.
Recall that the state of your own qualia is as unknowable to you as it is to some observer. Presumably, we have privileged access to those things of ours that are private. But in what sense do we have privileged access to our own qualia, if we cannot know the state of it? We have no more privileged access to our qualia than anyone else, so it seems strange that we would characterize that thing as somehow private to us. The properties of qualia, its constancies and changes, are either entirely beyond our knowledge or inferable from third-person examination (236). We cannot tell whether our qualia have changed, have stayed constant while our reactive attitudes toward the qualia have changed, or some combination thereof. Empirical methods may be able to determine between one extreme version or the other. For instance, if I can reliably match my qualia to an object and re-identify (e.g. I perform well in blind tastings), then that indirectly supports the hypothesis that my dispositions to react to the qualia have remained constant and it must be my qualia that have shifted. So in some sense, in some cases, the state of my qualia is amenable to third-person examination. But suppose both my qualia and my dispositions to react to them have shifted. Empirical methods will not be able to determine the proportions by which they shifted, for what it measures is always the resultant of both the qualia change and the change in reactive attitudes. If they are entirely beyond our knowledge, then it does not make sense to call them private, for no one has any access to them. But if they are inferable from third-person examination, then they cannot be private, for others have access to them. So there is no circumstance in which qualia are private.
1.3 The argument that qualia are not intrinsic.
Properties can seem intrinsic, but on closer inspection are actually extrinsic and relational. Dennett asks us to consider the worldwide eugenics experiment (237). Consider substance X, which tastes bitter to three-fourths of humanity and tastes like water to the rest. The way X tastes is a genetically transmitted trait. So we could prevent the water-tasters from breeding and after a few generations X will be unequivocally bitter. Or, alternatively, we could commit mass eugenics and prevent the bitter-tasters from breeding and after a few generations X just will be uncontroversially tasteless. So X may seem to be intrinsically tasteless, but really how it tastes is a relational, extrinsic property, for the property changes depending on the class of tasters. Or, consider that there is a certain intrinsic, felt quality about the value of a ten-dollar bill (237). But the value of a ten-dollar bill must depend on the willingness of others to accept it as payment, which makes it clearly an extrinsic, relational property.
Moreover, the intrinsic/extrinsic distinction itself is poorly defined (240). Here’s as good a candidate as you’ll get for an intrinsic property: think of the redness of Suzie’s ball. Public redness, like the ball, however, must really be an extrinsic, relational property – for the same reasons that the taste of substance X, a secondary quality like color, was an extrinsic, relational property. To be publicly red is just to be the sort of thing that affects people in such a way that they make a red-identification. Public properties are relational, and no intrinsic property could count as a public property, for public properties do not require the supposition of anything but the person’s disposition to react behaviorally (241). We can specify the effect a public property produces just by specifying the relevant reactive behavioral dispositions, e.g. the value of a dollar is established by establishing how people are disposed to react to it. This fact will be useful later.
1.4 The argument that qualia are not ineffable.
Dennett notes that it seems as though qualia are ineffable, but then suggests that they are merely practically ineffable (241). Consider an osprey cry (241). Here’s a true, accurate, poetically evocative description of the cry, “a series of short, sharp, cheeping whistles, cheep cheep or chewk chewk, [and so forth]; sounds annoyed” (241). With this we can rule out many bird calls that may or have been heard, but there are still many potential cries which could satisfy this description, and so we cannot uniquely pick out the osprey cry. So the cry is ineffable, but only practically so, for we can still rule out many other potential cries. But suppose I see an osprey cry and hear it and see that the sound is coming from it. I think, “that’s what an osprey sounds like” (the sound is recognized as S). But from this single experience I cannot know which physical variations and constancies in stimuli would produce an indistinguishable experience, nor can I know whether I would react the same or have the same experience if I were presented with a re-stimulation identical to the first, for I cannot know the effect that variations in either my mind or body have on my experience (241). So when I hear the osprey cry, I am just acquainted with some sort of property detector, and I cannot tell what property the property detector actually detects. This is to say that I hear the osprey cry, but I do not know on the basis of that experience what actually qualifies as an osprey cry (e.g. what other pitch-ranges and patterns count); I have picked up on and tagged a property (in virtue of my property-detector), but I cannot say anything about the property I have just detected. Indeed, the only way that I can refer to the property detected is by referring to that property which was detected by the property detector in some event. That I am able to refer to the property in this way amounts to a new way of thinking about the property. That I have a new way of thinking about the property means that I can mentally access it in a unique way, viz. by thinking of that property that was detected in that event (242).
This way of thinking is practically ineffable for two reasons. (1) It has an untested profile in response to perceptual circumstances. This means that I haven’t detected the property in enough cases to be able to give it an adequate characterization. I sensed S and detected it as a certain property p, but I do not know whether sensations R and Q also count as detecting p. (2) It is a highly informative way of thinking because it uniquely picks out a perception, representing it with a richness and complexity that cannot be captured by description (and only too many other properties could satisfy any description). As such, it seems that any description of it could never be as informative, making it appear practically ineffable. But because we have that unique way of referring to the property, it is not actually ineffable.
Or consider a guitar string (243). Pluck a single string – how does it sound? Is the sound describable, or does it sound ineffably guitarish? It seems more intuitive or natural to suggest the latter. But should you pluck the string and another to create a harmonic, a new sound is heard. And this harmonic overtone one can hear distinctly, the ineffability of the first plucking is gone, replaced by a duality that is more clearly describable. There was a difference between the first and second plucking, viz. there was a complexity to the second plucking which we could discriminate and respond to, whereas the first plucking seemed simple, unanalyzable, and ineffable. But upon reflection, it seems that there was actually the same complexity in the first plucking as the second plucking. For consider, it was by the first complex patterns of harmonic overtones that you could recognize the sound as that of a guitar, as opposed to a sitar or a mandolin. The subjective experience between the first and second plucking may be strikingly different, but in each case you were responding to a highly informative complex property that in many ways cannot be described.
Dennett introduces the horizon of distinguishability which determines the simple or atomic properties that one consciously experiences (244). In this way, it sets the limit to our capacity to describe things, for, as we showed earlier, atomic properties are unanalyzable and indescribable, which means that our descriptions of our experience must bottom out at the atoms. So the horizon of distinguishability limits our capacity to describe. But that there is this limit does not mean that there are absolutely indescribable features in our experience. For the experience was either more or less describable in the first or second plucking, depending on what our horizon of distinguishability was. In the first case, there wasn’t much for us to distinguish between. In the second case, with the overtones, there were more ways we could distinguish between the sounds because there were more sounds, more tones, that were present for comparison and could bring out the relevant features. So qualia are not absolutely ineffable.
1.5 A breather.
What has been shown thus far? Qualia were defined as directly apprehensible, private, intrinsic, and ineffable. But Dennett has shown that this is a confused and empty definition. Qualia cannot be directly apprehensible, for we cannot determine the state of our own qualia through any amount of introspection. Moreover, should we determine the state of our qualia through elaborate empirical means, then it they cannot be directly apprehensible. Nor can qualia be private, for they are either beyond our knowledge and thus inaccessible to anyone (and thus not private to anyone), or they are verifiable through empirical testing, but then they are accessible to everyone (and thus not private to anyone, for no one has privileged access). Furthermore, qualia cannot be intrinsic. There is no adequate characterization of the intrinsic/extrinsic distinction – it is not clear what an intrinsic quality could even be. And in addition, we showed how public properties like taste can seem to be intrinsic, but that actually they are extrinsic and relational. And lastly we showed that qualia cannot be absolutely ineffable, for we do in fact have a way of directly referring to them, and the fact that there are limits to our ability to describe qualia does not make them ineffable, for they can still be more or less describable (depending on the horizon of distinguishability). So the term “qualia”, as defined, cannot actually refer to anything. There is no referent picked out by the description, “directly aprehensible, private, intrinsic, and ineffable”. The term is empty.
Section 2. Dualism’s teeth.
But if there are no qualia that we are acquainted with, how should we make sense of the phenomenal quality of our experience? It seemed as if there were qualia, but if there are not, then what are we actually talking about when try to talk about the things that qualia were meant to characterize? Well, we know that whatever properties they are, they are public, extrinsic, relational properties (as was established in the substance-X example). We know that we can refer to these properties by reference to that property detected by our property-detectors. The reference to our property-detectors is private, but only in the sense that it is idiosyncratic, meaning whatever our personal behavioral response is, so it will not be private per se. If we want to talk about states within us of a certain type, we can only do so by talking about relational, extrinsic properties, e.g. how my internal state a can provoke an act of apparent re-identification. So these are the characteristics of the properties we were trying to capture with qualia.
What’s the takeaway supposed to be? Dennett’s position on qualia is supposed to be an instance of “eliminative materialism” (244). Qualia were supposed to be one of the dualist’s last defenses against growing support for materialism. For as long as there are qualia, the materialist must be able to account for them. And this seems unlikely, as qualia are just the sort of thing that evade materialist explanation, as they were intrinsic, private, and so forth. But Dennett has shown that qualia don’t actually exist, so the materialist need not account for them. Thus we proclaim that a victory has been won for materialism and that dualism has been defanged.
Or has it? I will argue that, though qualia have indeed been quined, this in no way supports the materialism in eliminative materialism, nor does it defang dualism. Instead, Dennett’s conclusions provide good reasons to support the direct realist’s thesis: that objects exist independently of observation and that we have direct epistemic access to these objects.
It is important to clarify part of the motivation for dualism (including nonreductive materialism). A dualist is not a dualist because she has noticed qualia and thinks that there is something irreducible about it. A dualist is a dualist because apart from acknowledging all the physical and neurophysiological facts in the universe, she acknowledges that the world also appears in her conscious experience. My conscious experience has an undeniable phenomenal character. I am not “mentally dark”, so to speak. This is the mystery that motivates dualism. Qualia are invoked only as a way to characterize and talk about this apparent phenomenal character of conscious experience. The dualist’s challenge isn’t “why should there be qualia?”, but rather, “why should there be conscious experience?”
Perhaps there are no qualia, no mystical, internal, phenomenal properties. But nevertheless, I feel a pain when I stub my toe – and so do you. There is something that it is like to stub my toe, to feel that pain, meaning that there is something (maybe not qualia) phenomenal about the experience. It is a fact about the world that we have experience of pains, tastes, smells and the like. If that is not qualia, then what could it be? It clearly is the sort of mysterious thing that calls for explanation. In “Quining Qualia”, Dennett has offered no solutions to this problem. His suggestion is that “insofar as we wish to cling to our subjective authority about the occurrence within us of states of certain types or with certain properties, we can have some authority…only if we restrict ourselves to relational, extrinsic properties like the power of certain internal states of ours to provoke acts of apparent re-identification” (244). But there is more to conscious experience than our internal states provoking acts of apparent re-identification. For consider at the means by which you apparently re-identify; in what does this apparent re-identification consist?
There is something that it is like to apparently re-identify. I tend to identify one past experience with another when I judge that the phenomenal character of the past experience is sufficiently similar to the other. Upon what basis do I make this judgment? It seems as though I remember something of the phenomenal character of the past experience, but I also am acquainted with the phenomenal character of another experience. By way of memory, I make the identification between the phenomenal character of each experience, provided that the phenomenal character of each experience is sufficiently similar. Dennett successfully shows that we cannot trust our memory in such instances, however. For memory-access pathways can be inverted such that I judge that one past experience matches the phenomenal character of another, even when their phenomenal character at the time of my experiencing was actually quite different (such was the cunning of the evil neurosurgeon). This is the sense in which acts of re-identification can only be apparent. We cannot reliably, authoritatively re-identify. But though the phenomenal character of each experience may be different, contrary to our judgment, there is still a phenomenal character to each experience, regardless of whether or not we are able to authoritatively identify them – there is still something that it is like to undergo each experience. This is an important fact to recognize; Dennett is not showing that there is nothing that it is like to undergo a particular episode of conscious experience. So the eliminative materialist must still show how phenomenal, conscious experience is reducible to wholly material constituents, even as he has shown that there are no qualia. Because this hasn’t been accomplished, eliminative materialism will not provide an attractive alternative to reductive materialism, or improve upon it in any significant way. Therefore, dualism remains a viable approach to consciousness and mental life, for the dualist will have the resources to talk about the phenomenal character of our experience while the materialist will not.
Section 3. “Quining Qualia” as an argument for scientific realism.
So if Dennett’s argument does not provide us with good reasons for materialism, what does his argument provide good reasons for? What does the elimination of qualia actually motivate us to think? In this section, I will argue that Dennett’s argument actually amounts to an argument in favor of direct realism. First, we should further specify the features of direct realism. Most broadly, direct realism is the thesis that objects like tables, coffee mugs, and your mother’s favorite frying pan, exist independently of their being perceived and that, moreover, we directly engage with these objects – there is no intermediary (e.g. sense-data). In naive realism, these objects have all the properties we perceive them to have, e.g. warmth, blueness, figure. In contrast, scientific realism maintains that only some of the properties an object is perceived to have exist independently of a perceiver and that an unperceived object will not have certain (secondary) properties, like blueness. This is because secondary qualities exist only in virtue of their relation to the perceiver – an object will not be blue apart from a subject perceiving it as such. Dennett’s argument lends support to scientific realism (but not naive realism).
Recall that sense-data are the objects of perception and that qualia are their properties. The thought is that sense-data serve as an intermediary between us and the external world. We are given sense-data in perception and become acquainted with their qualia. On this basis, we make an inference to the presence of some real, external object. From the way things appear to me, I infer what must be so.
But Dennett has shown that there are no qualia. If there are no qualia, then sense-data have no properties, for qualia are the properties of sense-data. But an object cannot exist without any corresponding properties. So if sense-data have no properties, then sense-data simply do not exist. This means that there is no intermediary between us and the world, entailing that we must be directly acquainted with the world, for nothing mystical comes in between. Note that this is one of the features of scientific realism.
Reconsider some of Dennett’s conclusions. Instead of qualia, we have practically ineffable public properties, to which our internal states respond and provoke acts of re-identification. Bitterness is not an intrinsic property, but a public property, for recall that the taste of substance X was determined by the reference class of normal detectors. The properties of our sensory experience are therefore relational. Substance X was bitter in virtue of a set of perceivers perceiving it as bitter. Dennett asserts, however, that you are epistemically justified in reporting that the relation between your tasting activity and your judging activity has changed. For instance, if you order a tasty americano from Caffe Strada everyday, but then after enough time find that you no longer think the americano tasty, you are justified in saying that the relation between your drinking Strada americanos and your enjoyment of them has changed. What you are not justified in saying is that either the taste of the americano has changed or your reactive-attitudes have changed (or some combination thereof), for this is not knowable by introspection (as was shown). This highlights the fact that secondary properties like taste are relational. The taste of the americano is contingent upon the taster. This echoes scientific realism, that the secondary property of an object is contingent upon there being a perceiver to which the object is related.
Furthermore, Dennett observes that “the properties of the ‘thing experienced’ are not to be confused with the property of the event that realizes the experiencing” (243). For example, the difference between someone’s imagining a green reindeer and as opposed to a purple reindeer might just be a particular physical difference, e.g. the presence or absence of a particular zero or one in a “brain register” (243). This sort of difference is sufficient for generating the dispositional differences between imagining a green reindeer and a purple reindeer in response to some intrinsic fact – that is, qualia a may dispose one to either imagine a green reindeer or imagine a purple reindeer depending on the minutiae of one’s physical brain state. This means that one’s disposition to respond to qualia a (by way of imagining a green or purple reindeer), was actually dependent on some external, physical fact (one’s brain state), as opposed to the capricious flow of some spiritual goo (e.g. qualia). The properties of the event that realizes the experiencing (the brain register state), are different from the properties of the object of experience (whatever was picked out by qualia a). This further underscores the relationship between the extrinsic properties of the object and the state of a perceiver’s perceptual apparatus – a relationship emphasized on a scientific realist picture.
We have shown how Dennett thinks that we have direct epistemic access to objects and how the secondary qualities of objects, those sensory qualities, are relational and dependent upon a perceiver. Thus we have shown that Dennett’s “Quining Qualia” has provided an argument against intermediary sense-data and in support of scientific realism.
Section 4. The compatibility of direct realism and dualism.
But here’s an important observation: direct or scientific realism is not incompatible with dualism. This is why Dennett is not able to support the materialism in eliminative materialism nor is he able to defang dualism. This section briefly notes some conceptions of dualism that are compatible with direct realism.
One example is provided by David Chalmers in his essay “Facing Up to the Problem of Consciousness”. After a thorough discussion of the hard problem of consciousness and why it evades material explanation, Chalmers proposes some speculative solutions. One proposal is the double-aspect theory of information. Information consists in information states embedded in an information space. An information space is a structure of difference relations between elements (without having to specify what those relations or elements are). Information can be physically embodied in a space of distinct physical states – this would constitute an information space. Chalmers observes that there is an isomorphism between physical information spaces and phenomenal information spaces – that is, there is an isomorphism between the physical world and our phenomenal experience of the world. This leads to the idea that information has two aspects, viz. a phenomenal aspect and a physical aspect. It is possible to describe the laws of physics purely in terms of information, by expressing the relations between certain states and the effects they have on other states without ever saturating those states with meaning. So it may be that it is information that is truly fundamental to the world, and that it corresponds to both the physical and the phenomenal. If information has both a physical and a phenomenal aspect, then the objects we are acquainted with by perception just are the phenomenal aspect of physical information. Perceptual objects and objects-in-the-world are the very same thing – just two different aspects of it. Thus, in perception we can still maintain that we are directly acquainted with objects, we can continue our acknowledgement of the phenomenal character of experience without invoking a confusing and empty concept like qualia. In this way, we can keep our dualism without sacrificing direct realism.
Another example is present in the work of Umrao Sethi. In “Hallucination Defanged: Perception and the Dual Nature of Its Objects”, she maintains that we are directly acquainted with the world and that the particular appearances of objects are objective, even though they are qualitative in character and may have mind-dependent instantiations – objects have appearances that are independent of our experience. This is a direct realist line that acknowledges the phenomenal character of our experience and even the phenomenal character of things outside our experience. So it is also something of a dualist view. So we see that direct realism is compatible with the idea of objective phenomenal information. Thus direct realism is compatible with dualism.
I will conclude by briefly summarizing the happenings of this paper. We showed how Dennett quines qualia, revealing it to be a confused concept without a referent. We argued, however, that Dennett’s paper only succeeds in eliminating sense-data and supporting scientific realism. It was then shown that dualism is compatible with direct realism. Therefore, Dennett’s argument does not decide between dualism and eliminative materialism. So the dualist is not compelled to become a materialist.
Chalmers, David J. “Facing Up to the Problem of Consciousness.” N.p., n.d. Web. 17 Apr. 2016.
Dennett, Daniel. “Quining Qualia.” Philosophy of Mind: Classical and Contemporary Readings. By David John Chalmers. New York: Oxford UP, 2002. 226-45. Print.
Sethi, Umrao. “Hallucination Defanged: Perception and the Dual Nature of Its Objects.” Umrao Sethi. N.p., n.d. Web. 17 Apr. 2016.