Gary and Peirce on Mind and Functionalism

Posted on August 4, 2010
Filed Under Peirce, Philosophy | 18 Comments

Gary, over at Minds and Brains has a discussion of functionalism and mind. The actual topic is a debate over Metzinger’s theory of consciousness which Gary is having with various OOP critics. Now I’ve never read Metzinger and have barely dipped into OOP so I’ll wisely remain silent on those topics. However Gary wrote something interesting I wish to take as a starting point for a tangental discussion.

Although I do not totally agree with everything Metzinger has to say (he is an internalist whereas I am an externalist), I do agree with him that the self is not an object, but rather, a function. It is an operation rather than a thing or repository.

This got me immediately thinking about C. S. Peirce and his famous statement “man is a symbol”


There is no element whatever of man’s consciousness which has not something corresponding to it in the word…the word or sign which man uses is the man himself…that every thought is a sign, taken in conjunction with the fact that life is a train of thought, proves that man is a sign; so, that every thought is an external sign, proves that man is an external sign…the man and the external sign are identical. Thus my language is the sum total of myself; for the man is the thought. (“Some Consequences of Four Incapacities”, CP 5.314)

While this is from his early period (1868) he says fairly similar things through his life.

Now Peirce is using language fairly broadly here. Effectively he is saying all thinking is in signs, to produce a thought involves this sign-process and the communication must have an external sign. For Peirce to be real is a process of practical verifiable communication.

This is fairly similar to the functionalists although I think Peirce would be very skeptical of the manner the functionalists proceed. More key I think Peirce would criticize functionalism in terms of his categories. Functionalism gets thirdness but misses firstness (the experience in and of itself – similar to qualia) and secondness (the pure action in terms of Other). Firstness is basically what zombie arguments against functionalism attempt to bring out. Secondness is what arguments about the importance of the material of the mind as opposed to the simulation bring out. (Say the difference between a simulation of motor and having a motor and physical gasoline)

Related posts:

  1. Functionalism and AI
  2. The Extended Mind
  3. Peirce & Being
  4. Davidson: Knowing Ones Own Mind 2
  5. Morris vs. Peirce
  6. Peirce on Reference

Comments

18 Responses to “Gary and Peirce on Mind and Functionalism”

Very interesting post. That’s a great quote by Peirce. You raise this issue of “firstness” or qualia. This is the standard argument against functionalism: that it can’t explain the subjective “feel” of consciousness because it’s just a brute function and why should the special features of consciousness be associated with a function. Well, my thinking in regards to functionalism is heavily inspired by Julian Jaynes and I think Jaynes has a clever way of getting “firstness” out of a function.

For Jaynes, the function is that of a lexical metaphor-spinning operation. The metaphor function spins a lexical web of narrative-based metacognition that is grounded by experience in the physical world (here, I think Jaynes is very much compatible with Lakoff and Johnson). The conscious mind is an metaphorical function or analog of physical behaviors and physical space. Moreover, there are associated experiential properties of the metaphor-function which Jaynes calls “paraphrands”. Since the source domain for the metaphor function is our knowledge of the spatial world, “spatialization” or “interiority” becomes the paraphrand for the conscious function. This “interiority” or “mind-space” effect explains the “insideness” or “firstness” of conscious introspection and conscious thought. This is the very interiority which has bedazzled thinkers for ages. Jaynes is able to “explain” interiority by unraveling its structure in terms of the process of analogical reason.

While somewhat abstract, I think it is plausible to assume that if a function is operational within the organism, there would be an experiential component for the function (provided we understand that there can be nonconscious experience). The consciousness function would be a function of metaphorical generation based on analogical reason. Jaynes shows that if we look at the structure of metaphors, they all have paraphrands. The most important paraphrand for consciousness is the spatialized “mind-space”. So the experiential component of the consciousness function is “spatialized” or “interiorized” precisely because the metaphor it is based on is a spatial one (generating the mind-space by modeling physical space). Moreover, I think this mind-space can now be modeled by global work-space theory.

Previous functionalists were unable to explain firstness because they couldn’t understand the metaphorical structure of consciousness, and thus couldn’t connect the features of interiority to the analogical processes of linguistic cognition turn upon itself. Because Jaynes’ theory is able to break firstness down into its experiential components, I think he can answer the Zombie question about functionalism quite well.I hope this all made sense.

I’d always felt that there was something really Kantian about Jaynes’ psychology – I mean, like it could have been written by Kant, if he’d come under the influence of bad cognitive science/philosophy of mind – but I’d never seen its Kantianism put so succinctly. Your summary, Gary, sounds like it is the Transcendental Aesthetic as read and reworded by a cognitive linguistics student (I don’t mean this as a dig on you, but on Jaynes). The problem of course is that, as with the Transcendental Aesthetic, and perhaps even to a greater degree, Jaynes’ explanation of “firstness” or “qualia” or conscious experience, simply pushes the problem back a level. How does the metaphorical or analogical mapping produce qualitative experience? Through another metaphor? Is it metaphors all the way down?

I’m only being partly facetious (it really does sound a lot like Kant, especially with phrases like “consciousness is the spatialized ‘mind-space,’” which is somewhat meaningless without a severely Kantian spin.

What’s strange here, though, is that for functionalists, qualia are functions, which makes explaining how functionalism might account for qualia by making qualia something other than functions (which are, of necessity, multiply realizable, whereas these metaphorical mappings are either so context dependent or entirely bodily dependent as to make them anything but multiply realizable) somewhat problematic. Unless that’s not what you’re trying to do, in which case, there are still much better approaches than Jaynes’ pseudo-scientific claptrap. And I say this as someone who is pretty convinced that analogical reasoning, or analogical mapping, is one of the most ubiquitous and important processes in human cognition and perception.

Chris,

Jaynes’ notion of the conscious operation is no more “pseudo-scientific” than current discussions of “executive function” or Gazzaniga’s notion of “the interpreter”. Jaynes’ conception of consciousness is in fact an early progenitor of executive function theories and linguistically mediated self-regulation models. In fact, he often calls the function of consciousness “narratization”. I take it that this function of narratization is scientifically plausible.

You ask how it is that this conscious function actually generates qualitative experience. I would first question you about what you mean by “qualitative” experience. If by qualitative you mean the “quality” of experience in the sense that a bee or lizard has qualitative experience, then I would ask, “What is there to explain?” Why is behavior a “problem” to be explained? For Jaynes, nonhuman behavior can be explained by behaviorism because asking why there is something it is like to be a bee is a pseudo-question unless you presuppose that there is an epistemic inside-outside distinction operation for bees which isolates them from the immediacy of their actions in the environment. For me, inside-outside talk is only applicable for humans because I see the experience of “insideness” being meditated by linguistic-social concepts and the metaphor-functions that I discussed earlier.

So, in regards to qualia, I would first distinguish behavioristic qualia (that shared by bees or mice) and introspective qualia, which is the experience of insideness given through the conscious operation of introspecting upon a mind-space (available to those who have the requisite powers of introspection). Behavioristic qualia is only a “problem” if you buy into certain assumptions about the nature of animal experience. Introspective qualia however is actually a problem insofar as there seems to be a disconnect between behavior and introspection. I hold that Jaynesian theory can explain this disconnect in terms of analogical processes with associated paraphrands of spatiality. Now, you can keep asking “But why does metaphor generate qualitative experience?” but this question is only sensible if we first distinguish between behaviorial (nonconscious) qualia and introspective (conscious) qualia. Of course, metaphor doesn’t “explain” behaviorial qualia but this is because there is nothing to “explain”.

Introspective qualia on the other hand, does need an explanation insofar as it is not quite clear what the “function” of introspection is. On Jaynes’ account though, the function of conscious introspection is for the guidance and control of behavior through narratization. This can answer Chalmer’s hard problem insofar as we now have an empirically plausible answer to the question of “Well, what does consciousness do?” It narratizes behavior. While this is still abstract, we could flesh this out in terms of contemporary research on working memory, global workspace, the left hemisphere “interpreter”, inner-speech, thought monitoring, etc. All these functions can be modeled in computational terms. There is nothing philosophically mysterious about them other than their complexity. We only think they are mysterious because we are prone to think about them from the perspective of bad metaphors (the inner Cartesian theater).

Your point about Kant is interesting. Jaynes comments in his book that he thinks his concept of the “Analog ‘I’” (that which *does* the introspecting) is comparable to Kant’s notion of the transcendental ego insofar as it is contentless. But Jaynes’ notion of the Analog ‘I’ is far from pseudo-scientific insofar as its ontology can be modeled in terms of the “virtual”, which is a well-established ontological category. In this way, the concept of a “mind-space” is salvageable without resorting to Kantian metaphysics (which are still Cartesian insofar as the ontological ground of the transcendental synthesis is obscure).

To be honest, I haven’t the slightest idea what the “executive function” and “linguistically mediated self-regulation” models of consciousness are. Could you point me to something I could read on them? Also, I don’t think Gazzaniga’s “interpreter” is a very good contrast, because it’s not a very widely respected theory of consciousness, at least not within the scholarly fields that study consciousness.

On the actual issues, the “narrative” aspect is, of course, not unique to consciousness, making it difficult to use it as an explanatory tool (contrary to popular belief, even “linguistic” isn’t limited to conscious experience). In fact, one could argue that much of what our higher-order cognitive processes are doing is “narrativizing” the world. This is, in a very literal sense, what concepts, schemata, analogical reasoning, etc., are, and they are, for the most part, unconscious processes. Hell, in a sense, this is what our visual system does much of the time, and that’s a pretty low level. Granted, much of this doesn’t take place in language, in the sense of inner dialogue, but as we now know quite well, inner dialogue isn’t a cause of consciousness, even if it does enable certain higher-level conscious functions. So the narrative part doesn’t explain much, with respect to consciousness, nor does it explain much with respect to the content of cousciousness (neither do paraphrands, a concept which hasn’t cought on within the study of consciousness or metaphor for a reason, most specifically that it’s just a fancy name for something that we already know a lot about, and something that’s not very explanatory, namely associations). It doesn’t even explain the problem away. All it does, as I said earlier, is push it back a level. For example, when you write, “For me, inside-outside talk is only applicable for humans because I see the experience of ‘insideness’ being meditated by linguistic-social concepts and the metaphor-functions that I discussed earlier,” this is wonderful, except it doesn’t explain, in an even remotely reasonable, much less scientific way, how the experience of insideness (whatever that is, and certainly this, if anything, needs both description and explanation, eh?) arises. How can these concepts create a conscious experience which is, in fact, not actually a conscious experience (since conscious experience, which is just what qualia are, isn’t real, it’s an illusion created by the very thing that must be, in fact, a conscious experience).

Now, I’m all for doing away with inside-outside talk, at least in many philosophical/scientific contexts (in some, it serves as a helpful heuristic, while in others it just creates confusions), along with things like subjective-objective, mind-world, etc., but if we’re going to do that, we need more than a rewording of behaviorism and a fairly abstract theory about what consciousness is “for” (there are much more concrete theories, though no one seems to agree on them; I’m partial to Damasio’s, at least through the lense of direct realism, as I don’t like much else in Damasio), as the former has been tried and failed, and the latter is just a part of the beginning of a theory of conscious experience. At this point, I see no reason to take Jayne’s seriously.

By the way, if it’s not pseudoscientific, I’d definitely want to know a.) what some of the testable hypotheses of Jaynes’ theory are, and b.) some cites of research testing those hypotheses.

Gary, I’m not familiar with the figures you mention but Chris more or less makes the point I would. Interestingly Peirce talks a lot about phenomenological firstness but not consciousness as such. When he does talk about it it’s clear he’s not quite sure how to deal with it. Interestingly in a few places he suggests it’s chance experienced from the inside. Peirce adopts the old epicurean notion of swerve and has the universe basically being a semiotic development from pure chance to pure determination – the development of laws. So in any sign process there will be elements of chance or swerve and in those aspects is consciousness. He doesn’t seem too committed to this view of consciousness though. (IMO) More it’s the best hypothesis he can come up with to explain it.

A functionalist can, of course, merely state that during the execution of a function there is this element of firstness (or qualia – although the two don’t exactly correspond to each other). It’s interesting that so few make that move. (Although I don’t claim to be well enough read on the topic to perhaps justifiably make that claim)

Chris, from what I gather, Gazzaniga’s interpretor theory is widely regarded to be more or less correct. It’s very well respected and trust me, I keep up to date in the field of consciousness studies. Look up some recent research on “anti-correlated” default mode networks, mind-wandering, and “secondary consciousness”. As for executive function, do a google scholar search string. It’s a very well researched notion and fits nicely with Jaynes’ theories. As for the testable hypotheses of the book, I highly recommend you just read his book. For an overview of recent research that corroborates Jaynes’ theory, see the following:

http://julianjaynes.org/pdf/jaynesian_volume3_issue1.pdf

I don’t need to justify why you should take Jaynes seriously. He was a Princeton university professor, not some crackpot theorist. His views are well researched, philosophically sound, interdisciplinary, and well corroborated by evidence from diverse fields. It’s impossible to convey the full scope and weight of his argumentation and theorization within a few short paragraphs. What I have presented here represents but a fraction of what he has to say. In my opinion, Jaynes was the great neglected theorist of the 20th century. You would do yourself a great disservice if you brushed him off as a “pseudo-scientist”.

I know a lot about “executive functions” in cognitive and neuroscientific theories, but I don’t know of any such theory of consciousness. If you merely mean to say theories of executive function, then I’m not sure how it relates to Jaynes’ theory of consciousness, and if you mean executive function theories of consciousness, then Google Scholar isn’t going to help me. It also doesn’t help much as a theory of the function of consciousness, though I suppose the (linguistic or non-linguistic) self-regulation theory, which probably has something to do with the function of consciousness, does get us somewhere, though not far (as it would only account for a part of, likely early or low-level, consciousness).

And while I’m sure you keep up with consciousness studies, it must be said Gazzaniga’s “interpreter” theory is certainly not a part of anyone but Gazzaniga and his students’ standard theory of consciousness, or anything else for that matter. It’s hardly ever mentioned, except by Gazzaniga and his students, in mainstream neuroscience or cognitive science journals. That’s a big clue.

Also, I didn’t mean to imply that you should defend Jaynes to me. I read The Origin of Consciousness several years ago (I think around ’99 or ’00, when I was just beginning grad school), and I wasn’t impressed then or now. It’s very interesting (if dry), but it’s largely untestable, and almost completely untested. The summary of supporting evidence you linked is as vague as it could possibly be, and for good reason. Those findings fit much better with more complete theories, for one, and they’re somewhat ambiguous findings in themselves, based on advances in neuroscience that happened long after Jaynes’ book. I doubt you, or anyone but the data, could change my mind on that.

I do have to say that I find it interesting that you pick a homuncular theory of (one aspect of) consciousness as a defense of Jaynes, though. I wonder if you can point to some literature discussing Gazzaniga and Jaynes together.

Chris,

Theories of executive function are related to consciousness insofar as many theorists think that the executive function is the function of consciousness. This would be the position of people like Bernard Baars and his Global workspace theory. Global workspace, working memory, and executive function are all more or less synonymous in terms of how they are purported to function. However, this connection between executive function and consciousness rests on the commonly made distinction between primary consciousness and secondary consciousness, with executive function being associated with secondary consciousness (advanced metacognition). This distinction goes by many names e.g. prereflective/reflective cognition, primary process/secondary process, unconscious/conscious, automatic/controlled, etc. Primary consciousness is what Jaynes calls cognition and secondary consciousness is what Jaynes called consciousness. If you look at how secondary consciousness is defined in the literature, you will see a close connection between executive function or “ego function”. Baars and others are quite explicit about this. Moreover, many theorists have associated executive skills with linguistic cognition and this is another respect in which Jaynes was ahead of his time.

And insofar as executive function theories postulate similar “inner speech” and “inner rehearsal” mechanisms for self-regulation, Gazzaniga’s theory of the interpreter is deeply related to consciousness, provided we understand consciousness in terms of “secondary consciousness” and not “primary” consciousness.

And if you cannot find references to Gazzaniga, then perhaps you aren’t looking in the right places. His work is commonly taken up under “narrative” theory of mind and discussions of the “narrative self” and “narrative mind”. See:

Gallagher, S. 2000. “Philosophical conceptions of the self: implications for cognitive science,” Trends in Cognitive Science 4, No. 1: 14-21.

Menary, R. (2008). Embodied Narratives. Journal of Consciousness Studies, 15(6), 63-84.

Hutto, D. (2009). Folk Psychology as Narrative Practice. Journal of Consciousness Studies, 16(6), 9-39.

Also, I find the claim that Jaynes’ hypotheses are “untestable” to be an exaggeration. His hypotheses about neurology, schizophrenia, and bicameral vestigiality are testable and have stood up to the test of time. His ideas about auditory hallucinations are testable and have been corroborated. His hypotheses about archeology and anthropology can be corroborated by empirical evidence (and have been). His hypothesis of language and consciousness can be tested, and have been corroborated (and philosophically supported as well). Jaynes’ theory is scientific in the same way that evolution is scientific. We can’t run a test and “falsify” evolution as a whole, but we can gather corroborative evidence from a variety of sources and then piece together the facts into a unified story. Jaynes has done exactly this. To accuse his theory as being somehow unscientific is ludicrous in light of these facts.

Ugh, I don’t mean to sound like a jerk, but no, you’ve got it all wrong. I mean, until the references, nothing you say is right. First of all, working memory and executive functions are not identical, though (some) executive functions may be a component of working memory (e.g., in Baddeley), and working memory may be a component in some executive function, but they are not in the least synonymous. Then, of course, for Baars, executive functions take place primarily on the unconscious side of his Global Workspace theory. And of course, Global Workspace does contain conscious experiences, but Baars himself is quite clear that it is not the same thing as consciousness. And of course, working memory isn’t always conscious (holding something in working memory and holding it within conscious awareness, at least, are not the same thing at all). You might do better to look at directed attention, or something to that effect, as a scientific synonym with consciousness, though even that is hardly uncontroversial, and I’m not sure it would help you with Jaynes at all. And no, nothing I describe as unconscious (including the executive function component of Baars model) are parts of secondary consciousness.

Of course, if we’re going to equate primary and secondary consciousness with things like automatic/controlled, unconscious/conscious, and most of your other dichotomies, then “consciousness” has lost most of its meaning, and certainly loses the connotation of conscious experience. And since Baars always treats the background stuff, including executive function, as unconscious in a way that is distinct from consciousness (be that consciousness basic/core or extended, say, in Damasio’s sense).

As for Gazzaniga, I know that he is referenced a lot (though not nearly as much in others as he is in his own work; he’s like Damasio in that regard), but not generally in the neuroscience of consciousness.

Anyway, this is pretty far afield. I’m perfectly satisfied that, while Jaynes’ theories of hallucination might have some validity, though we’ve advanced well beyond them, he has nothing important to say about consciousness except as a discussion starter.

Here is a quote from Baars (Baars, B. J., & Franklin, S. (2007). An architectural model of conscious and unconscious brain functions: Global Workspace Theory and IDA. Neural Networks, 20(9), 955-961.)

“GWT suggests an active view of the conscious aspects of human cognition. It is the consciously evoked “broadcast” that serves to mobilize and guide the many unconscious knowledge domains that enable Working Memory functions like inner speech, visual problem solving and executive control”.

So, Baars says that global workspace enables executive control and working memory, not that executive control and WM are totally unconscious. While true, executive control, working memory, and GW are not exactly the same, they are wrapped up with each other which is why I said they are “more or less synonymous in purported function”. This much is true.

In this paper (Alvarez, J., & Emory, E. (2006). Executive Function and the Frontal Lobes: A Meta-Analytic Review. Neuropsychology Review, 16(1), 17-42.), theorists define working memory as a component of executive function:

“executive functions, including verbal and design fluency, ability to maintain and shift set (as measured by the WCST), planning, response
inhibition,working memory, organizational skills, reasoning, problem-solving, and abstract thinking”

These are all more or less the functions which Jaynes associated with consciousness. As for working memory, Baars says that “all the classical “boxes” of Alan Baddeley’sWM models have a conscious component—including conscious perception of input, conscious access to verbal rehearsal, and conscious decisions regarding verbal report.”

So while you are right to distinguish the two, GWT and WM are closely related insofar as working memory needs consciousness and GWT is a theory of consciousness. In his 1997 book, Baars says “GWT is consistent with models of working memory by Alan Baddeley, the mind’s eye by Stephen Kosslyn, explicit knowledge after brain damage by Daniel Schacter and others, the thalamocortical searchlight elaboriate by Francis Crick, and society models outlined by Micahel Gazzaniga and Marvin Minksy.”

I’m skeptical of your notion that we have “advanced” in terms of understanding auditory hallucinations. We might understand their neurological substrate, but what about their origin in history? What theory of auditory hallucination other than Jaynes’ manages to connect the origin of religious hallucination (ubiquitous in our society), schizophrenia, and nonpsychotic auditory hallucination into a comprehensive theory? None.

Also, you mention that my breakdown of nonconscious/conscious “certainly loses the connotation of conscious experience”. While it is true that my notion of consciousness is distinct from how contemporary analytic philosophers define consciousness (in terms of qualia), Jaynesian consciousness is actually much more in line with the denotative definition of consciousness that Descartes introduced: that which is introspectable.

Gary, I just don’t know enough about the cognitive science here to really follow Chris and your debate here. I am intrigued if you could maybe flesh out Jaynes’ argument a bit. (i.e. is it as Kantian as Chris suggests and exactly how does it handle the transition from 3rd person to 1st person)

If you want the full treatment, I have recently completed a long paper defending Jaynes’ complete theory of mind against Ned Block’s famous criticisms. It also compiles contemporary research and theory which is in support of the Jaynesian paradigm (Personally, I think it is the most comprehensive philosophical defensive of Jaynes yet written, though this is largely due to the paucity of literature on the subject). It is currently under review (fingers crossed) for the journal Phenomenology and the Cognitive Sciences. You can download it here:

http://lsu.academia.edu/GaryWilliams/Papers/156099/What-Is-It-Like-to-Be-Nonconscious-

Gary, nothing you say about Baars contradict what I said. Some of it even repeats it.

However, if Jaynes is redefining consciousness to include things like automaticy, non-linear category rules, and cognitive processes and representations that we have no “conscious” access of, then consciousness itself becomes a scientifically and likely philosophically meaningless concept.

Chris, Jaynes claims that we have no conscious access to the generation of consciousness, but like a map maker and a map user, how a map is used is different from how its made. So using consciousness is different from how consciousness is generated. Accordingly, the analogical laws that generate the paraphrands of “mind space” are not accessible by the user-construct that is generated by such laws, what Jaynes called the “Analog I”. The Analog ‘I’ is Jaynes’ most Kantian concept. It is what “does” the introspecting when we close our eyes and imagine a mind-space. But there is no reason to suppose that the linguistically mediated analog construct would have access to the mechanics of its own generation.

I would agree with most of that. In fact, most of it is non controversial. But the linguistic part is antiquated and wrong for one, or possibly for reasons. First, consciousness isn’t entirely linguistic, nor is it built on language. Second, not all linguistic cognition is conscious.

What is strange about this is that we haven’t mentioned the body at all. The strongest argument against functionalism is that the purpose of experience relates it to an embodied organism, making multiple realizability problematic. Individual processes may be functions in this sense, but the whole is not. Does the individual function have any meaning or reason outside of the whole, then? If not, functionalism is untenable.

Under this view, it probably makes more sense to treat the quality of experience as arising from its position within the whole. This doesn’t require language or a narrative — such theories are poor attempts at reducing experience — but it does require concepts. I suppose a sort of McDowell meets Merleau-Ponty, or something. But the redness of red doesn’t need language, or a concept of consciousness; only a concept of red within the space (and I use that term metaphorically) of color experience.

Narratives may give us rich concepts of self and history, but they aren’t required for consciousness nor do they require consciousness.

At times one notes a “naive idealist” reading of Pierce (or perhaps postmod reading)–some seem to think Peirce was denying “externalism” so forth, claiming humans are nothing but a play of signifiers, etc., (granted Peirce does at times hint at a nearly mystical conception of nature…then he’s not entirely free of…quackness either). I don’t think that’s what Peirce intended (also recall his “secondness” as the realm of brute fact, more or less). Perhaps he did believe reality was mediated by semiotic, and that mind was irreducible…yet he was too much a scientist to deny externalism of some sort: the pragmaticist theory of truth itself would demand something like measurement, correlation, observation, so forth (ie, secondness, and thirdness). Peirce was ergo arguably a type of functionalist, at least in terms of his ideas of truth of theories–

Recall that Peirce called his philosophy objective idealism following Schelling. I also don’t think this semiotic realism is opposed to externalism in the least. Reality is the force of secondness acting which leads to habits in thirdness. That’s definitely his cosmology. So it’s the interplay between secondness and thirdness which is important.

Leave a Reply