Brain Hammer

I have a crush on the post “Qualia: The Real Thing” by Keith Frankish, guest-blogging on the Splintered Mind.

Drawing a distinction between classic qualia and diet qualia (though not under those terms) is a common move in the literature, but I’m suspicious of it. I’m just not convinced that there is any distinctive content to the notion of diet qualia. To make the point, let me introduce a third concept, which shall I call zero qualia. Zero qualia are those properties of an experience that lead its possessor to judge that the experience has classic qualia and to make certain judgements about the character of those qualia. Now I assume that diet qualia are supposed to be different from zero qualia: an experience could have properties that dispose one to judge that it has classic qualia without it actually being like anything to undergo it. But what exactly would be missing? Well, a subjective feel. But what is that supposed to be, if not something intrinsic, ineffable, and private? I can see how the properties that dispose us to judge that our experiences have subjective feels might not be intrinsic, ineffable, and private, but I find it much harder to understand how subjective feels themselves might not be.

Now, if we moved in a direction even further from classic qualia, what would the resulting position be (and what would be the appropriate soft-drink inspired name for it)? The position might involve, instead of identifying qualia with properties of experiences that constitute the dispositional basis toward such-and-such judgments, an identification of qualia with some non-dispositional feature of the judgments themselves or, alternately, the contents of the judgments themselves. But, what to call it?

Seltzer Qualia

Tosca, cannoli and coffee

Originally uploaded by renmeleon

Through the miracle of thought experiment, Sellars’ mythical Jones and Clark and Chalmers’s notebook-toting Otto had a baby combining essential features of both daddies. Unfortunately, she got named “Jotto” but let’s go with “Jo” for short.

Jo, all grown up now, carries and utilizes a notebook similar to her daddy Otto’s. If Clark and Chalmers’s remarks about Otto and his notebook are correct, then the story of Otto illustrates an implementation of vehicle externalism for thoughts while being consistent with the falsity of (1) content externalism for thoughts (a la Putnam and Burge), (2) content externalism for experiences (a la Dretske and Tye), and (3) vehicle externalism for experiences (a la Noe and Hurley). Jo, like Otto, likewise implements vehicle externalism about thoughts. But because of contributions from her other daddy, Jones, she will also implement vehicle externalism for experiences. Further, she will do so in ways independent of appeal to the enactive approach heralded by, e.g. Noe and Hurley.

Jones does to Jo what he did to his buddy Dick in “Empiricism and the Philosophy of Mind”:

“And it now turns …that Dick can be trained to give reasonably reliable self-descriptions, using the language of the theory, without having to observe his overt behavior. Jones brings this about, roughly by applauding utterances by Dick of “I am thinking that p” when the behavioral evidence strongly supports the theoretical statement “Dick is thinking that p”; and by frowning on utterances of “I am thinking that p”, when the evidence does not support this theoretical statement. Our ancestors begin to speak of the privileged access each of us has to his own thoughts. What began as a language with a purely theoretical use has gained a reporting role.” (XV, 59)

Unlike Dick, however, Jo will sometimes automatically say “I am thinking that p” after having read it in her notebook. If this doesn’t seem like the sort of privilege we’d expect of mental states, we need only alter the thought experiment to have Jo write in a code only she understands. Thus do Jo’s thought vehicles achieve an at least Sellarsian privacy.

Let’s turn now to see how Jo’s experience vehicles might bleed out of her head and into her notebook. Modeling sensory impressions in a Sellarsian manner, we need notebook states that (1) can be causal products of perceptible objects such as red and triangular objects, (2) achieve a kind of privacy analogous to those achieved for thoughts, and (3) “stand to one another in a system of ways of resembling and differing which is structurally similar to the ways in which the colors and shapes of visible objects resemble and differ” (XVI, 61).

(We likely need much else besides and thus would 1, 2, and 3 be unlikely to jointly suffice for conscious notebook-states.)

Now, if the impressions of red triangles recorded in the notebook are the sorts of things that others would recognize as drawings of red triangular objects, then the privacy condition has not been satisfied. However, given the way that Sellars specifies the notion of similarity utilized in the third condition, the similarity relations between marks in the notebook and perceptible objects need not be apparent to people other than Jo, and thus would they effect a kind of privacy. (An interesting question is whether the similarities in question even need to be apparent to Jo, but I need not take a stand on that issue here.)

I’ll stop this initial sketch for now. The main question that remains to be addressed concerning whether Jo’s notebook states can be partially constitutive of states of phenomenal experiences is the following.

Why not?

If you know that you are not a zombie, then phenomenal character is conceptual and inverted spectra (color qualia being inverted relative to your conceptualizations) are impossible.

Regarding the supposition that you know that you are not a zombie, I interpret this as meaning that you know that you now have states with phenomenal character or qualia. It is useful to compare this kind of knowledge to more ordinary cases of knowing that something is the case. Take for example, my knowing that there is a dog in the room. In order for me to know this, there must be some set of properties that the dog has and that I am able to conceptualize. I can be relatively neutral on exactly which conceptualizations will get the job done. Maybe my conceptualization is that there’s a four-legged furry barker in the room. Maybe my conceptualization is that there’s a domesticated wolf-descendant in the room. Maybe my conceptualization is simply that there’s a dog in the room. But however it goes, there must be some set of properties of the dog (e.g. being domesticated, being wolfish) and I must have some set of concepts adequate for the accurate representation of those properties (e.g. the concept of domestication, the concept of wolves).

Now, my knowledge that I now have states with phenomenal character is seldom if ever analogous to the case in which I simply conceive of the dog as a dog. I am not now simply conceiving of myself as having phenomenal states. There are specific phenomenal states that I conceive myself as having. As I type this note and take breaks to sip coffee there’s a whole slew of qualia that I conceive my states as having. In particular, I conceive myself as seeing my coffee mug as being blue. I have a blue quale and am able to conceptualize it as such. I reject, then, the statement that there is no absolute correct orientation of the color spectrum. I think there is. It involves conceptualizing a blue quale as blue and a yellow quale as yellow and so forth.

Now, if qualia are distinct from my conceptualizations, as they would need to be if inverted spectra are possible, then it would be theoretically possible for my qualia to become inverted without my noticing. My quale that I currently conceptualize as blue would actually be yellow and vice versa. My current conceptualization as having a blue quale would be false, then. And it would be false without my noticing. Further, if qualia are distinct from my conceptualizations, I could have all the same conceptualizations without having any qualia at all, and my belief that I’m not a zombie would be false. If it’s possible for my belief that I’m not a zombie to be false, then I can’t know that I’m not a zombie. Thus does self-knowledge of non-zombie-hood lead to the impossibility of inverted spectra.

Chapter 6 of The Subjective Brain, “The Allocentric-Egocentric Interface Theory of Consciousness” is here.

Excerpt:

In brief, the theory posits that mental processes form a hierarchy of mental representations with maximally egocentric (self-centered) representations at the bottom and maximally allocentric (other-centered) representations at the top. Part of what it means to be higher or lower in the hierarchy is to be further from or closer to the sensory and motor periphery of the nervous system. Focusing on the processing of sensory information, we can trace the path of information from relatively egocentric representations of the stimulus in sensation through stages of processing that increasingly abstract away from egocentric information and represent things in memory in an allocentric way. Further, we can note top-down effects from relatively allocentric representations high up in the hierarchy to egocentric representations lower in the hierarchy. I hypothesize that phenomenally conscious mental states are to be identified with states that are relatively intermediate in this hierarchy. More specifically, conscious states are hybrid states that involve the reciprocal interaction between relatively allocentric and relatively egocentric states: a conscious state is composed of a pair of representations interacting at the Allocentric-Egocentric Interface.

Eric Schwitzgebel, as usual, illuminatingly plumbs the depths of introspective error at his Mind Splinters blog. Of particular recent interest are reports of some exchanges he’s been having with Dan Dennet over how to reconcile the possibility of introspective error with the sorts of authorial authority granted to the introspector by Dennettian heterophenomenology and first-person operationalism. (Schwitzgebel’s most recent post is here, which is a follow-up to this earlier post here.)

I think Schwitzgebel is on to something when he suggests we try to draw the distinction between what errors can and cannot be made in terms of a distinction between phenomenal judgments and what’s “behind those judgments”. I think, however, he missteps in his description of such a distinction in terms of two senses of “seems”. Dennett is no friend of the phenomenal/epistemic distinction between senses of “seems” that many philosophers follow Chisholm and Jackson in drawing. Also worth keeping in mind is Dennett’s negative reaction to so-called “real seemings” expressed in Consciousness Explained and elsewhere.

So how best to flesh out Schwtzgebel’s insight regarding Dennett interpretation? I think one can do this with a single (epistemic) sense of “seems” and a distinction regarding the way’s things are with regard to our seemings.

To give a very clear illustration of this distinction, consider a substance dualist who judges a piece of wax to be melting in the heat of their fireplace. One way things are with respect to the dualist’s seemings is that it seems like he can tell by sight that the wax has changed shape. And about this they are correct: it does seem he can tell by sight that the wax’s shape has changed. But another way things are with respect to the dualist’s seemings is that they, the seemings, are identical to brain states. And about this the dualist is quite wrong (or, more humbly, clearly might be wrong).

To relate this to an example discussed by both Schwitzgebel and Dennett, consider the case of peripherally presented playing cards which, to the surprise of many subjects, cannot be identified by suit or even color (though their motion may be readily apparent). I urge that we avoid cleaving senses of “seems”. We should not describe the case as it epistemically seeming to subjects that the periphery is clear and phenomenally seeming blurry. Instead we should say the following:

It seems to the untutored observer that he or she has detailed information about the color and suit of peripherally presented cards. However, in reality the subject does not have detailed information about the color and suit of peripherally presented cards, even though it may really seem that way to the subject.

[See also, some of my earlier posts on senses of “seems”: Bursting Apart at the Seems; Bursting at the Seems 2: Electric Boogaloo; Transcending Zombies.]

The following figure is from “REDDISH GREEN: A CHALLENGE FOR MODAL CLAIMS ABOUT PHENOMENAL STRUCTURE” by Martine Nida-Rumelin & Juan Suarez (link to pdf) and reproduces stimuli utilized in experiments in which paradoxical visual experiences were induced, such as experiences of a single color patch being colored reddish green.

Some subjects report “seeing a homogeneous color phenomenally composed of red and green whose components are as clear and as compelling as the red and blue components of a purple.”

The phenomenon is induced by presenting equilluminant colored stripes in images stabilized via use of an eye-tracker. The experiments reported are from Billock et al. which reproduce experiments from Crane and Piantanida.

See:

Billock, V. A., Gleason, G. A., & Tsou, B. H. (2001). “Perception of forbidden colors in retinally stabilized equiluminant images: an indication of softwired cortical color opponency?” J Opt Soc Am A Opt Image Sci Vis, 18(10), 2398-403.

Crane, H. D., & Piantanida, T. P. (1983). “On Seeing Reddish Green and Yellowish Blue.” Science, 221(4615), 1078-1080.

See also my “Hyperbolic Mary“

Blast Radius

Originally uploaded by Pete Mandik

Dan Cavedon-Taylor raised some pretty interesting questions here regarding whether the neuro-introspection I defend in Ch. 2 of The Subjective Brain is, as Paul Churchland puts a similar thesis, “direct”. Instead of leaving my response buried in an old-ish comment thread, I reproduce it here:

I think most if not all that needs to be said about the relevant issues concerning perception can be done in terms of a three-way distinction (a tristinction!) between sensation, perception, and inferences based on what’s perceived. A similar tristinction may be drawn between sensation, introspection, and inferences based on what’s introspected. Sensations are non-conceptual and carry information about themselves and their causes. Perceptions are conceptual and involve the automatic conceptual exploitation of information that sensations carry about their causes. Introspections are conceptual and involve the conceptual exploitation of information that sensations carry about themselves. Inferences involve the non-automatic application of concepts.

I try to illustrate all of this stuff in terms of the story of George, John, and the man in the gorilla suit in section 3 of chapter 2. George, the special effects expert, and John, the novice, both have the same sensations, I suppose. George is able to see the person as a man in a gorilla suit. I might just as well say that George is able to see that this is a man in a gorilla suit. John is able only to infer, based on what he perceives (plus what George tells him), that this is a man in a gorilla suit. This is not to deny, however, that John is incapable of seeing this as something or seeing that something is the case. John may being seeing this as a living organism.

I’m not particularly thrilled about the vocabulary of “direct” and “indirect”. A lot of what some people want to call direct perception I would call “sensation” and a lot of what some people want to call indirect perception I wouldn’t call perception at all, but conclusions of inferences. I worry about certain baggage associated with “direct”, in particular, the view that there can be unmediated epistemic access to anything. That strikes me as a nutty view and I don’t want to be read as assuming its truth.

Ok, now to explicitly address your points and questions:

You ask: “[D]o we *see* heat? We can *see that* something is hot (again by seeing some feature of it). Do you think seeing heat and seeing that something is hot are the same? The first sounds direct, the second–indirect.”

Let me start by saying that I assume seeing to be a kind of perceiving. So, visual sensation alone would not count as seeing. Further, I acknowledge a distinction between seeing heat and seeing that something is hot. I also take this to be the same distinction as that between seeing heat and seeing heat as heat. I don’t think, however, that there’s such a thing as seeing heat without seeing it as something or without seeing that something is the case. A visual sensory response to heat without concept application would be sensation, not seeing. I really don’t know how best to apply the “direct” and “indirect” vocabulary to these kinds of cases. Is the direct thing the sensation? Or is it the seeing of the heat as something but not seeing it as heat? I propose to just do without that vocabulary.

You propose:
“[P]erceiving the heat of the coffee by perceiving its steam looks like a case of *indirect* perception (i.e. the claim seems to be I perceive x by perceiving y)”

Another way of describing what’s going on in the coffee case is that I have a visual sensation which is caused by hot steam and thus carries information both about the presence of heat and of steam and I perceive (visually!) both heat and steam though none either more or less directly than the other. (I think I need to be much clearer about this in a revised version of the chapter.)

Regarding introspection, you ask:
“[I]n what sense can the introspection of one’s brain states be direct if it is mediated by mental states? I take it the neuro-introspectionist doesn’t claim that one could introspect their brain states without introspecting mental states. Rather, they claim we introspect brain states as such by first introspecting mental states (again, this sounds like such introspection of brain states is thereby indirect). Or am I wrong??”

Here’s how I’d describe what’s going on. Sensation is one mental state. When I introspect the sensation, that involves a second mental state, which itself is a conceptual representation of the first state. Also, the occurrence of the second state must be an automatic response to the first state. Now, some people who don’t hate the word “direct” would say that the introspection is direct insofar as it is automatic. Others would say it is not direct because it involves a representation. I prefer to say what’s going on without using the word “direct”.

Raising some interesting concerns about modality individuation, you write:
“I would have thought that if heat perception is to be direct, then the sense modality through such an experience is afforded is going to have to be tactile, rather than visual (otherwise it just sounds a bit like a category error).”

I think the distinction between dermal thermoreceptors and retinal photoreceptors will be important for distinguishing between seeing heat and feeling it. But I don’t see that much work can be done about directness in terms of receptors. Patterns of activity in the retina carry information about all sorts of stuff, like what color it is, how hot it is, whether it was manufactured in China, etc. Anything I can figure out by looking at something must involve information that passes through my retina, so I’m not optimistic that one can settle questions concerning directness in terms of what can and cannot be transduced at the site of reception. But then again, I don’t feel particularly motivated to settle questions concerning directness, at least, not in the vocabulary of “directness”.

Below are the slides for the talk based on the chapter available here: link.

In furious response to the alleged irrefutability of Berkeley’s idealism, Samuel Johnson kicked a rock and declared “I refute it thus�. As long as rocks can be Johnsoned, idealism about them is not forthcoming. What else admits of Johnsoning? In particular, are qualia easily Johnsoned?

One might think that they clearly are, for whatever surety we gain from rocks by kicking them will be matched by a painful quale if the rocks are kicked with improper footwear.

However, one thing to be concerned about here is that the Johnson maneuver is too swift and violent to avoid conflating issues of existence and concept-independence. An idealist unmoved by the kick may grant that rocks exist while continuing to deny their existing independently of our rock conceptualizations. And one may grant realism about rocks while maintaining the concept-dependence of various non-rocks. The question of what non-rocks may admit of Johnsoning is best discussed by keeping distinct existence and independence.

Postponing qualia for a moment, let’s adopt or adapt some examples of Dennett’s: being famous and being a suspect. While having star-power is one thing, fame is another and no one is actually famous without the actual conceptualizations of others. Similarly, while being guilty of a murder is one thing, being a murder suspect is another and no one is actually a suspect without the actual conceptualizations of others. Perhaps in the case of being a suspect, others are not required insofar as an amnesiac may suspect themselves of having committed a murder. Nonetheless, being a suspect, like being famous, exhibits concept-dependence.

A sophisticated contemporary idealist about fame is denying the concept-independence, not but not the existence of, famous people quafamous. And so, what’s a sophisticated contemporary Johnsonian supposed to do? It’s not clear that going around kicking famous people is going to prove much of interest to the current question.

A similar lack of clarity attaches to the postponed question of qualia. Whatever painful qualia set in as a result, either directly or indirectly, from a bout of star kicking, the qualia aren’t obviously independent of one’s conceptualizations. Indeed, insofar as it will be obvious to one that one has painful qualia, one will have concepts adequate to forming the thought that one has painful qualia. (This latter point is further developed in the discussion of the first premise of my Transcending Zombies argument [link].)

Fig. 1. …and Johnson’s all, like, WTF!?!?

The Eye in the Pyramid

Originally uploaded by jovike

According to the working definition of conspiracy theories in “Shit Happens”, it is a necessary condition that the hypothesized conspirators “keep their intentions and actions secret.” Since a central point of “Shit Happens” is that conspiracy theories are universally unwarranted, prima facie warranted conspiracy theories (mainstream explanations involving the individuals involved in Watergate, al Queda, Nazi Germany) need to be addressed.

We can focus the concern that needs to be addressed in terms of a pair of questions. Aren’t we warranted in the common belief that, e.g. al Queda blew up WTC? And isn’t the common belief (e.g. that al Queda blew up WTC) a conspiracy theory?

The strategy I currently find most appealing is to answer the first question positively and the second negatively. The next question that immediately arises is why aren’t these prima facie warranted conspiracy theories really conspiracy theories. My answer is that they fail the necessary condition of keeping secret.

There are several ways in which one can fail to keep secrets. One way is by getting caught and being compelled to testify in a criminal investigation. In this case one may have tried then failed to keep the secret. A related way is when direct evidence (video tape of someone building and planting a bomb) renders the secret no longer kept. Another way of failing to keep secret is illustrated by terrorists broadcasting their involvement in a plot in order to take credit for its success. In this case the sense in which they fail to keep a secret is by no longer even trying to keep it secret.

A true conspiracy theory attempts to leap over a wall of posited secrecy via attempts at inference to the best explanation. The main problems arise in establishing that the proffered explanation is indeed the best instead of swamped by multiple equally plausible explanations. In cases that we are warranted in believing, e.g. that al Queda planned the 9/11 bombings, we aren’t stuck making such a leap.