The Blue Car Is the Red Car

Once upon a time we had a red car--a red two-door 1987 Acura Integra without air conditioning that we bought in the spring of 1988. We called it "the red car." After thirteen years we replaced it with a blue four-door 2000 Acura Integra with air conditioning. And for the first couple of months we had it, we would occasionally and accidentally (whatever that means) refer to it as "the red car"--even though we knew full well that it was the blue car.

You see, I would argue, the neural circuits were well-engraved: an Acura Integra, the smaller of our cars, the more responsive of our cars, the non-station wagon--the features of the blue car were nearly identical to the features of the red car, so when our brains grasped for a verbal referent they had a good chance of picking the standard phrase we used for the red car. And, of course, neither of us had any trouble understanding what the other meant by the phrase "the red car."

Last week, for the first time in a decade, we drove past the dealership where we bought the red car. And last week--for the first time in five years--we both used the phrase "the red car" when we meant "the blue car."

Thus it seems only natural to me to think that Jerome Feldman is right when he claims, in Jerome Feldman (2006), From Molecule to Metaphor: A Neural Theory of Language (Cambridge: MIT Press: 0262062534), that it is time to try to study human language seriously by starting with the observation that language is produced and controlled by human brains active in the world, each of which has perhaps 100 billion neurons with a thousand connections and millisecond response times--a massively-parallel array of processors in a form almost totally unlike the traditional architecture of all of our electronic computers (except possibly for Google itself).

This may be a bad bet. 100 billion neurons will surely have powerful and remarkable emergent properties that cannot be easily predicted from the behavior of any one of the neurons. Perhaps language is that sort of emergent property that is effectively wetware-independent. But I don't think so. I think it is a very good bet. And Jerome Feldman takes us all on a very interesting and wild ride through the subject.

Highly recommended.

And when Jerome Feldman tries to explain how so much of our high-level metaphorical thinking is grounded in primary metaphors of our immediate physical experience, he can only do so... metaphorically. His discussion of the event structure metaphorical complex contains the sentence: "Difficulties are impediments to motion." Impediment. You can see the Latin roots im- and ped-: an impediment is something that keeps you from footing it. Your impedimenta is your luggage that you have to carry on your journey, but that makes the journey difficult. The more general meaning of impediments as things-that-prevent-completion-of-tasks is rooted in the idea of things-that-make-it-hard-to-cover-ground-fast already.

p.87: Let's look at how long it takes, on average, to do one of these perception-reaction tasks.... How fast can you engage the brakes after seeing an obstacle in your path? It takes about half a second.... This kind ....of calculation can help us understand some basic facts about how the brain computes.... [T]he time for each of the basic processes of neural signalling and firing is about one-thousandth of a second.... Let's compare this with the average human reaqction time of half a second.... The reaction time includes the time for the image information to get from the eyes back to the brain and the time for the motor signals to reach the muscles and contract them. This doesn't leave very much time for the brain to do whatever processing is needed to decide which button to push or whether to slam on the brakes.

If we think of each neural action as one computing step, then our brain is able to compute the reaction in around 100 steps. By way of contrast, computer programs... take many millions of time steps to recognize an image. This discrepancy is one of the main factors leading computer scientists to conclude that neural computation is radically different from ordinary computation....

Earlier work had shown that you could both improve the speed and reduce the probability of error in such a visual reaction task by having the subject hear the target word around the time the image was flashed... the general phenomenon of priming.... When you are planning to buy a care, you are much more likely to notice similar cars....

The psychological literature is filled with discussions of priming, spreading activation, and related ideas. There is usually no specification of how mental connections and spreading mental activation map to neural connections and neural firing. Bridging this gap is an important goal of this book...

p. 141: For embodied cognitive science, any computational-level formalism must be effectively reducible to the connectionist level and thus to brain mechanisms. Computational-levbel descriptions may fail to capture several key neural properties, including massive parallelism, robustness, spreading activation, context sensitivity, and adaptation and learnings. As in all science, the trick is to have levels of description that are mutually consistent, with each facilitating different kinds of reasoning...

p. 200: A general theory elaborated by Joseph Grady in 1996 suggests that the metaphor system is grounded in the body in terms of "primary metaphors." In each primary metaphor, such as affetion is warmth, an experience brings together a subjective judgment... and a sensory-motor occurrence.... For this metaphor, such an experience might be cuddling up to a parent. Such correlations often show up in language.... Affection is warmth.... Intimacy is closeness.... Important is big..... Happy is up.... Bad is stinky.... More is up.... Help is support.... These primary metaphors allow one to express a private internal (subjective) experience in terms of a publily available event; this is one crucial feature of metaphorical language.... Largely universal, primary metaphors provide the grounding for much of the metaphor system.... From our neural perspetive, primary metaphors can be seen as a normal consequence of associative learning... neurons that fire together, wire together....

Causes are forces. States are locations (bounded regions in space). Changes are movements (into or out of bounded regions). Actions are self-propelled movements. Purposes are destinations. Means are paths (to destinations). Difficulties are impediments to motion. Expected progress is a travel schedule; a schedule is a virtual traveler, who reaches a prearranged destination at a prearranged time. External events are large, moving objects. Long-term, purposeful activities are journeys.

p. 273: The core questions in dispute can thus be expressed succinctly: (a) Are formal grammar rules expressed in the brain? (b) Is grammar independent of other brain structures? (c) Is there some special genetic encoding specifically for grammar? The language wars are fought between people.. portraying grammar as a special ability and... [those] suggesting that it is part of our general intelligence.... Neither side in this battle... worries explictly about the details of how language and thought are processed in the brain. The linguists do analysis of language as such, and the PDP connectionists focus on learning rules. By keeping the issues narrowly focused, both sides are able to pursue their arguments without dealing with questions that would be compelling from any broader perspective.... [C]onsider walking, or even better, dancing. Dancing is clearly learned and can be described by rules. Dancing appears to exist in all cultures and can be learned without formal instruction.... There may well be a human proclivity to dance--a dancing instinct. Suppose we recast issues a-c for dancing: (a) Are formal dancing rules expressed in the brain? (b) Is dancing independent of other brain structures? (c) Is there some special genetic encoding specifically for dancing? These questions don't seem to make a whole lot of sense, do they?... [N]one tells us much about how cdancing is actually carried out and learned....

For different reasons, both sides in the language wars reject detailed operational theories. From the PDP general learning position, the only interesting issue is learning from a blank slate.... The fact that the brain has a great deal of elaborate structure before learning begins is ignored....

The extreme believers in innate, autonomous, rule-based grammar can ignore any conflicting biological evidence because of their conviction that neuroscience is not nearly developed enough to be taken seriously.... Chomsky... in... 2003.... "When people say that the mental is the neurophysiological at a higher level, they're being radically unscientific. We know a lot about the mental from a scientific point of view. We have explanatory theories that account for a lot of things. The belief that neurophysiology is implicated in these things could be true, but we have very little evidence for it. So, it's just a kind of hope; look around and you see neurons; maybe they're implicated."... The scientific path to truth [according to Chomsky]... is formal linguistic analysis. If neuroscience is incompatible with this formal analysis, neuroscience must be wrong. The same belief system provides a rational for ignoring inconvenient results from psychological experiments.... These are said to reflect only linguistic performance.... The deep questions concern linguistic competence and can only be addressed by the orthodox methodology of formal grammar...

p. 317: They key to understanding grammar acquisition is not the famous poverty of the stimulus... but rather the opulence of the context. The child comes to language learning with a rich base of conceptual and embodied experience as well as a supportive social environment. Words and rules that describe this experience can be learned without formal training, although not without years of focused effort...

p. 330: My personal favorite among the recent books is Looking for Spinoza by Antonio Damasio (2003). Damasio tries to relate the latest biological and clinical findings to subjective experience with impressive results. As with all current explanations, even if every detail in the book were exactly right, it wouldn't resolve the big question. We simply don't yet have a way to pose the question of subjective experience in a way that could yield a scientific answer...

Comments

[dale]

"language is produced and controlled by human brains active in the world..."

That rings odd to my ears. Seems overly objectified or reified or some sort of fied.
Languaged is produced and controlled by human beings active in their social world. That seems a more realistic and practical statement. And one less likely to lead to mechanistic type errors. But then- I read way to much Habermas.

[Atanu Dey]

I recommend rather highly Jeff Hawkins' book On Intelligence.

http://www.onintelligence.org/

Two word summary: "Pattern Matching."

[shah8]

Chimeing in as I have some background in neuroscience...

The excerpts reprinted here rings bullshit to my ears. I'd need to read the whole thing, but I'm really not impressed as at a minimum, he loses significant accuracy in the metaphors he uses.

Also, the brain uses dedicated reprogramable subsystems (central pattern generators, for one simple example I thought I should have seen mentioned), as well, as chemical soupbowls, as well, as electrical holonets, and I think that's the *big* stuff. Any real discussion has to has some discussion about the interactions and interfaces between all of that mess, because it's pretty much certain (to me) that the emergent properties of both language and consciousness arises from contact points within all the systems in the brain, and *forgetting* (sort of like the reverse of Turing's computer). Hmmm, this would require too much of an explanation...Try one of G. Edelman's book for a major discussion of this tangent...

in any event, he may be talking about neuropsychology, but the excepts doesn't show that much of a clue of neurology...

[Jon Fernquest]

The rule learning aspect of connectionism and neural nets is also important for a very practical economic reason.

Flexible linguistic utilities, to do sophisticated searches through large collections of texts, like the internet, have to be trainable.

Reprogramming the utility everytime you want to do a new kind of search is just not time feasible.

Take Opinmind for instance, that supposedly can classify blog comments as agreeing or not agreeing,

obviously it can't pick up every nuance of language, it's going to miss some, probably a lot, in university blogs...

the only way you're going to get classification accuracy in a reasonable amount of time is through training, and then when it makes mistakes, retraining it, til you converge on accuracy

language is just too complicated and adaptable a thing to list out all the rules

[Ginger Yellow]

Does anyone know how Feldman's neurological approach meshes (or doesn't) with Jackendoff's mental approach, as expounded in The Foundations of language? Jackendoff certainly tries to fit linguistic processing into the results of cognitive psychology, but he does so at the level of what he calls the f-mind, ie the mind instantiated in the brain, rather than at a neuronal level.

[Scott Martens]

"You see, I would argue, the neural circuits were well-engraved: an Acura Integra, the smaller of our cars, the more responsive of our cars, the non-station wagon--the features of the blue car were nearly identical to the features of the red car, so when our brains grasped for a verbal referent they had a good chance of picking the standard phrase we used for the red car. And, of course, neither of us had any trouble understanding what the other meant by the phrase "the red car.""

Bad economist. No cookie for you. Go get Marx' "Theses on Feuerbach" off the shelf. Open to page one:

"The main defect of all hitherto-existing materialism - that of Feuerbach included - is that the Object, actuality, sensuousness, are conceived only in the form of the object, or of contemplation, but not as human sensuous activity, practice, not subjectively."

Do not ask what properties your old red car possessed. Ask: How did the red car fit into your system of praxis? Did the new car fit into the same place?

Then, imagine momentarily that you possessed a transcript of all speech within the Delong household from 1988 (when you bought the red car) to its replacement in 2001. Measure the rate of co-occurance of "red" and "car" in this transcript using any of a number of basic, non-connectionist, techniques. Mutual information is the simplest, but I'm doing my thesis on minimum description length, which is more complicated but has other advantages. Using mutual information, you could compare the entropy of "car" with "red car", and will find that they are at least comparable. "Car" paired with other adjectives will most likely not have comparable entropy.

So, on the one hand we have an object of praxis that has a significant role in your life, and on the other we have a conjunction of words occurring in a pattern comparable to a single word. This is no coincidence, comrade. You could have continued to use the phrase "red car" to indicate the new car. You did not do so because you took some effort not to, either for the sake of an imagined linguistic norm or to ease communication with non-Delongs with whom you might need to discuss your Acura. Most likely the first.

There is no need to invoke features here. The object as practice covers this ground far more easily.

The occurrence of priming in this case - the return to saying "the red car" - is a sign that this type of semiotic creation is a behavioral norm which you have consciously chosen to violate. In fact, you've acted in a manner contrary to the linguistic norm. A "Tudor house" is an exact analogy to the "red car" - you're unlikely to see one in Berkeley that was built when the Tudors were on the throne.

Now, there was no need to invoke connectionism in explaining this outcome. Neurons were, to be sure, involved. But so were hands, feet and carburators. As for modeling, quite non-connectionist information theory was up to the task.

Shah here has it partly right:

"Any real discussion has to has some discussion about the interactions and interfaces between all of that mess, because it's pretty much certain (to me) that the emergent properties of both language and consciousness arises from contact points within all the systems in the brain..."

At least as important to symbol use and production are the contact points between physical cognitive devices within our bodies and our practices and (in Marx' sense of the word) human sensuous activities.

But I have issues with Feldmann here:

"For embodied cognitive science, any computational-level formalism must be effectively reducible to the connectionist level and thus to brain mechanisms."

No, no, no, no! Nothing reduces to brain mechanisms because (*thumping the table*) BRAINS DON'T SIT IN GLASS JARS!!!!! This used to be one of the key tenets of embodied cognition as a research program.

Have we forgotten Bateson's blind man?

"Suppose I am a blind man, and I use a stick. I go tap, tap, tap. Where do I start? Is my mental system bounded at the handle of the stick? Is it bounded by my skin? Does it start halfway up the stick? Does it start at the tip of the stick? But these are nonsense questions. The stick is a pathway along which transforms of difference are being transmitted. The way to delineate the system is to draw the limiting line in such a way that you do not cut any of these pathways in ways which leave things inexplicable."

This basic failure of "embodied cognitive science" to actually embody its cognition in actually existing bodies is the thing that I most have against it. It may be necessary to have a computational level of description for human cognitive activity - it seems reasonable enough - but there can never be a NEURAL level which is complete, descriptively adequate and consistent with macro observation because (*more thumping*) COGNITION IS NOT CONTAINED IN BRAINS!

I'm all in favor of studying brains. I support research into connectionist algorithms and alternative computing principles. But I do not imagine that the solution to the problems of linguistics is primarily to be found in those things.

And I think that is one area where Feldmann and I are really on different pages. He slags the Chomskyans for ignoring brain research when it goes against them, which is right on because their whole research program is based on the idea that the brain must contain certain things. But I don't think connectionism or brain research are really going to bring clarity to much in linguistics either.

It's possible that in this respect we only differ in emphasis and vocabulary. I haven't read his book yet.

As for grounding things in metaphor... I was once an advocate of this approach. I'm not anymore. The information theoric explanation I offered for the "red car" also offers an alternative perspective on metaphor. Rather than seeing metaphor as a consequence of neural association, it makes at least as much sense to see neural association as a consequence of metaphoric usage.

Instead of making metaphors central to cognition, we might reread some Saussure. There is a meaningful, practical, opposition between "This situation is good" and "This situation is bad". We not only think different things when we hear the one and not the other, we do different things, and the things we do have different consequences depending on the objective conditions underlying those statements.

For Saussure, the existence of this opposition is central to linguistics and semiotics. Oppositions of the same sort are central to information theory and computing.

However, there is no opposition between "This situation is bad" and "This situation stinks". That fact alone would, for a lexicographer, be evidence that "to be bad" is the definition of "to stink" in some contexts. Now, situations cannot literally stink. If they could, there would be an opposition between "this situation is bad" and "this situation stinks" because they would entail different actions and different consequences.

This lack of opposition makes it possible for "this situation stinks" to take some type of metaphorical meaning. The semantic overlap between "to stink" and "to be bad" - that things that stink are not generally positively regarded - in combination with the practical fact that what we often wish to communicate about situations is whether they are positive or negative from some perspective, makes this metaphor comprehensible. Regular use, in turn, makes it conventional.

This explanation seems more likely to me because the kinds of metaphors he is invoking are not actually used all that systematically. Even though "bad" can be systematically expressed as a bad smell, the English language does not systematically express good as a good smell. You can't say *"this situation smells like roses".

Again, the mechanisms involved here do include some brain activity, although it is possible to model them using quite non-connectionist explanations, but more importantly they involve practical knowledge of the conditions under which language is used. They turn on activities that go on outside the brain.

Here, I think the cognitivist program in the US and I have to go our separate ways. They tend to see neurobiological modeling and metaphor as being at the root of cognition and semiotics. I tend to see praxis and opposition as central instead. There is some compatibility here, but I think in the end will all be for naught.

It bothers me to see linguistics so easily set adrift from its fundamental Saussurian roots, and it bothers me to see so much of what academic Marxism did right - placing a huge huge emphasis on practice - forgotten as if it never existed.

[a different chris]

I was going to comment but now I'm scared.

[Mike S.]

I'm curious about a much more plebian origin and potential implication -- how old are your kids, Professor? Did they play a role in christening the blue car the new red car?

I ask as a resident of a household where the two children under the age of 4 refer to the cars solely by color. As a parent, I find child-generated malapropisms endearing and look forward to keeping them in general use long after they have ceased to be directly relevent -- indeed, I hope to keep them in use at least until a point at which they cause real embarrassment to the children who originated them.

[bellumregio]

“Suppose we recast issues a-c for dancing: (a) Are formal dancing rules expressed in the brain? (b) Is dancing independent of other brain structures? (c) Is there some special genetic encoding specifically for dancing? These questions don't seem to make a whole lot of sense, do they?... [N]one tells us much about how dancing is actually carried out and learned....”

We many soon know the answer to this in other species such as honey bees. The answer to “a” in bees is certainly “yes”. (here we think more about “instinct” and not necessary “learning” but we could add it into our thinking. The answer to “b” is “no”. The answer to “c” is “not really”. Genes are interconnected in a web and the interactions are context dependent with constraints. There may be some genes for dance but they may also have different roles in other phenotypes.

You don’t need to try to divine human language to get at these kinds of basic questions. This all looks like slippery business to me. We know that similar phenotypes have very different genetic components. Who is say that human metaphors that seem to be similar also have some common neural basis? I would not bet money on it. The real question is not how the complex neural networks interact to create behaviors and perceptions of those behaviors but how complex neural networks interact with equally complex genetic networks in a given environmental context to produce a defined behavior or perception. These two networks are constantly interacting and they are constantly remodeling each other. The are also developmentally plastic but they are, of course, constrained.

[RobLL]

Some ten years ago Philosophy in the Flesh came out. A popular (not quite) treatment of what the authors called the second generation of cognitive studies. I was greatly impressed, and have been wanting to read a follow up (in the sense of being newer) for some time. I shall try Feldman's book. Thanks Rob

[Alex Tolley]

Scott: What are you rambling on about? Of course cognition resides in the brain - that is not in any way falsified by the connection of our brains to our bodies and through our senses, to the external world.

Suggesting that language is not some embodiment of the neural structures is dangerously close to suggesting the mind and brain are separate - which we can agree is wrong?

Whether the actual fine grained details of neural connections is necessary to understand cognition, or grosser structures based on the neural substrate, is not known at present. But I am willing to bet any philosophy - Marxian or otherwise is not going to help us understand it.

[cw]

Having worked with lots of small children, I don't believe that dacing is learned. I've seen 1 1/2 year olds who I don't believe have ever seen anyone dance before, dance when they heared music.

[JR]

If you were reading your story you would say, “Once upon a time we had a “RED CAR … we called it the RED car.” In American English, when an adjective-noun formation becomes the name of a specific item or kind of item, the accent pattern changes to indicate that it is now a compound noun, no longer a modified noun. Compare “a BLACK BIRD” with “a BLACK bird,” a “WHITE HOUSE” with “the WHITE House,” a “HIGH CHAIR” with a “HIGH chair,” a “ROUND steak” with a “ROUND STEAK.”

Once this transformation occurs, the meaning of the adjective is less important than the meaning of the phrase. Sometimes the phrase will persist even though the adjective’s meaning is partially or entirely absent. A legal immigrant has “a GREEN card” although no one has carried “a GREEN CARD” in years. And although all “BLACK BEARS” are black, some “BLACK bears” are brown.

So your first car was no longer just a “RED CAR.” It had become “the RED car.” For you, the color of this object was not its primary characteristic and so there was no difficulty in transferring its name over to its replacement. Perhaps you changed it because it seemed weird when talking to people outside the family.

[Scott Martens]

Alex, is using a calculator a cognitive process? If the answer is no, I think you mean something different by cognition than people usually do, because then designing an airplane is not a cognitive process. If yes, then cognition is not entirely located in the brain.

There's not a lot of agreement on the definition of cognition, but all the serious efforts to define it invoke computation in some sense, perception, and the capacity for action. Brains are clearly capable of some form of calculation, but perception and action take place elsewhere. The whole point of Bateson's blind man is that even he calculation can take place in part elsewhere.

"Suggesting that language is not some embodiment of the neural structures is dangerously close to suggesting the mind and brain are separate - which we can agree is wrong?"

No, we can't agree on that. The mind and the brain are not the same thing. Dead people have brains, but they don't have minds. Imagine taking a brain out of someone's head and using some kind of mad scientist technology to keep it alive in a glass jar. You'd still have a living brain, in principle one capable of all the same internal activity as when it was still in a body. You wouldn't have mind.

Nor is language the embodiment of neural structures. Is playing tennis the embodiment of neural structures? Neurons play a role in any tennis match, true enough. But is there really any demand for a neural theory of tennis? Can you really make sense of the behavior of tennis players by looking at neural structures? Do the physical structures of the tennis court, the material forces at work in balls, rackets and nets, the macro-scale physiology of the players and the social relations of sporting play no role whatsoever in it?

[Ginger Yellow]

"But is there really any demand for a neural theory of tennis?"

In a broad sense, yes. There's certainly a demand for a neural theory of visual perception, motor control, "muscle memory" and so on. Language is, or appears to be, both more specialised (in that it involves some mental processes not used for any other cognition) and more universal than tennis (in that every "normal" child develops language, but not every normal child can play tennis). Therefore a neural theory of language could be very useful. Perhaps not for understanding language in the abstract sense that Chomsky would like to, but certainly for understanding how our minds work and how our brains process stimuli and feedback.

[Peter]

Scott, you wrote:

"You can't say "this situation smells like roses". "

In fact, we can and do say this in English. More conmonly, we say it of people, eg, that he or she "will come out of some situation smelling like roses." But I take your point that we typically don't use our sense of smell as a metaphor for other things. If we were dogs (whose sense of smell is as well-developed as is humans' sense of sight), we could well have more metaphors based on smell.

On your substantive points, as a computer scientist designing languages and protocols for multi-agent systems, I agree wholeheartedly. Great stuff!

[Jerry Feldman]

Here are two more brief quotes from my book that might help.

p3. Language and culture are, of course, carried by the family and by the community. But each child has to rebuild it all in his or her own mind.

p.48 There is nothing unusual or exotic in maintaining several perspectives at the same time. Think of some of the many perspectives you have about your bicycle or your car. It is a mode of transportation, but also an investment, a source of pleasure, and possibly a status symbol. If it isn’t functioning properly, you might start thinking about the details of how it works or who could fix it for you.

A neural theory of language is not intended to replace historical, cultural, and other perspectives on language. On the contrary, we suggest that a better understanding of how people learn and use language is essential for such studies. This is best seen in the work of my NTL partner, George Lakoff.

[Alex Tolley]

Scott:

So we can roughly agree about what cognition means:
http://en.wikipedia.org/wiki/Cognition

"is using a calculator a cognitive process?". *Using* the calculator is, but the inderstanding still resides in the brain. Separate the calculator from the person and you have no residual understanding in the calculator. I would agree that this gets a bit fuzzy if we posit that the calculator is really a cochlear or retinal implant, or ultimately an integrated part of the brain. But I would argue that instruments and cars do not constitute some non-brain centered cognition even though they offer new ways of interacting with the environment. Similarly, separate a piece of the brain from the whole and you won't find much left that understands the world.

Mind-Brain duality. I think you misunderstood my meaning. Of course the mind and brain are different things, but I was referring to the idea that the mind is not an emergent property of the brain, but some other "stuff". Separating language from the underlying neural substrate will lead you into all sorts of difficulties when you consider the evolution of language. There is no binary no_language -> language transition.

Your point about glass jars is in extremis. Every night your mind is in a state like being in a glass jar - sensors and effectors
are largely shut down. Sensory deprivation tanks are even worse. The glass jar analogy is used as a model to suggest that we cannot *grow* minds in isolation.

"But is there really any demand for a neural theory of tennis?". Sure you can - tennis players cannot even return serves if they thought about their actions - the behavior is a trained response. It is far closer to a stimulus-response neural-effector explanation than to some other explanation.

Maybe scientists & engineers do Marx an injustice, but none of the books in my personal library on AI, cognitive computation, neurophilosophy, neurobiology etc. contain any references to Marx or praxis.

[RSA]

Alex:

You may be interested in work on distributed cognition, by Hutchins, Norman, Zhang, and others, if you're not aware of it already. One of the basic ideas is representation and cognition can be off-loaded into the environment, which makes for the not-entirely-implausible idea that some of what we think of as cognition is going on beyond the bounds of our brains.

[Peter]

Why do mathematicians and engineers working alone still use blackboards, or even pen-and-paper, if not to distribute cognitive activities beyond their own brains? Anyone who has built a reasonably-complex spreadsheet has done the same. Likewise most people in the humanities do not typically first think and then write, but think-by-means-of-writing: they often arrive at their view of a topic by writing about it.

[Ginger Yellow]

For the less technically inclined, there are some intriguing philosophical and fictional insights into distributed cognition/consciousness/selfhood in The Mind's 'I', edited by Hofstadter and Dennett. No substitute for empirical research, but well worth a read.

[Mandos]

So I haven't read the book, and maybe I'm a bad boy for not having done so, but I'm confused about something. When I read this:

"The extreme believers in innate, autonomous, rule-based grammar can ignore any conflicting biological evidence because of their conviction that neuroscience is not nearly developed enough to be taken seriously.... Chomsky... in... 2003.... "When people say that the mental is the neurophysiological at a higher level, they're being radically unscientific. We know a lot about the mental from a scientific point of view. We have explanatory theories that account for a lot of things. The belief that neurophysiology is implicated in these things could be true, but we have very little evidence for it. So, it's just a kind of hope; look around and you see neurons; maybe they're implicated."... The scientific path to truth [according to Chomsky]... is formal linguistic analysis. If neuroscience is incompatible with this formal analysis, neuroscience must be wrong. The same belief system provides a rational for ignoring inconvenient results from psychological experiments.... These are said to reflect only linguistic performance.... The deep questions concern linguistic competence and can only be addressed by the orthodox methodology of formal grammar..."

I find myself nodding in agreement. But I get the feeling that this is being quoted critically. But then I know that some readers of this blog believe that much of "orthodox" Chomskyan linguistics can and should be thrown out.

[Maynard Handley]

To paraphrase Scott Martens' point; one of the internal battles currently ongoing in physics is between supporters and detractors of Robert Laughlin (Nobel prize in physics in 1998). Laughlin believes that it is a fundamental characteristic of the world that physics studies that when you get large amounts of stuff, it behaves in ways that are universal, that is, that are more related to the fact that there is lots of stuff, and that it is constrained by some general principles, than by the specific details of how the underlying bits (be they atoms or quarks) behave.
(His point is partly the trite observation that you won't learn much about, say, magnetism in solids, by discovering the exact properties of the Higgs boson, but it is also the rather more depressing notion that you may not get to learn much about the Higgs boson if you are human sized and only able to interact with magnetism in solids, because many many different possible models of the micro-world all result in the same sort of predictions at the human scale by the time they're aggregated.)

Ignoring the second point and focussing on the trite point, there is a lot of evidence in the world of statistical mechanics that this is the way things work out; that rather different underlying systems give rise to the same sorts of behavior of phases and critical points and suchlike.

Of course this example from physics is proof of nothing in the world of linguistics; but IMHO it certainly does not seem ludicrous, like Chomsky, to claim that it is numbers and constraints that give rise to language, not the details of how neurons work and, pace Kandel, the specific proteins involved in modifying the nuclei of cells that encode long-term memory.

Of course there are always situations where the underlying constituents matter --- we can treat hard drives as abstract data stores for most purpose, but we have to remember that they are based on magnetism if we're planning to give them to field engineers who work around transformers ---but the question is whether useful understanding flows from fine-grained details of the constituents, as opposed to higher level abstractions.

[Alex Tolley]

I've changed my mind about the localization of cognition. I can see that calculators, blackboards etc are not intrinsically different from parts of the brain in aiding cognition. Maybe it is another example of Dawkin's "Extended Phenotype".

If one accepts the premise of cognition being part of the environment, does this extend to people interactions and does this lead us to a supermind?

[Auros]

Even ignoring "extended cognition" like calculators, given our ability to process stimuli and react to them in less time than it is known to take for a neural signal to actually reach the brain, be "sensed" by the conscious mind, and send out a planned reaction (the classic example is yanking your hand back from a hot stove), it seems clear that there are some kinds of "cognitive" processing going on in parts of the body other than the brain. There would appear to be some amount of direct activation from neural channels transmitting pain to those creating muscular response, in the spinal cord, well before you reach the brain.

[thomas]

It seems to me that when we use "external, aids to cognition" like blackboards and calculators, we are doing the same thing as when we employ a known or learned strategy, we are talking about the same general phenomenon -- articulation.

I'd draw you a picture if you think it would help. :-)

PETER asks: "Why do mathematicians and engineers working alone still use blackboards, or even pen-and-paper, if not to distribute cognitive activities beyond their own brains?"

They do it in order to form an articulation. Articulation confers an advantage. The advantage is mental distance. Once an idea is fleshed out on the blackboard (or even in working memory), the mind no longer needs to use as many processes to "hold" the particular integration (the articulation) in mind.

"Anyone who has built a reasonably-complex spreadsheet has done the same."

I would agree with this, kind of -- a spreadsheet involves a good number of articulations arranged systematically from which a mind can generate new articulations based on exploiting certain built-in relationships or categories.

"Likewise most people in the humanities do not typically first think and then write, but think-by-means-of-writing: they often arrive at their view of a topic by writing about it."

This really highlights the aid to cognition that a language confers. Once a thought (a real-time integration) is formed in working memory, or drawn or written out (made articulate) on paper, the integrating processes of the functioning mind are freed up to do additional work (on that same integration) if need be. This is thinking. Parts are integrated into an integrative whole, the formed articulation can then itself become part in a further integration, and the idea continues to change.

This is the advantage that articulation conveys -- its role can be understood metaphorically as a step in a flight of stairs. When we gain a step up, our view is literally broadened. And while it may have taken some effort to win the higher step, once there, one enjoys the broader view without expending anymore effort than it took to hold the view at the lower step.

"scott SAYS: No, we can't agree on that. The mind and the brain are not the same thing. Dead people have brains, but they don't have minds."

The mind and brain ARE the same thing when the words mind and brain are used to refer to the same process. Dead people have dead tissue. While we can identify the brain of dead person, as a brain, it does not include the functioning properties of the organ, and thus such a brain is not a brain in the functioning sense, and thus, dead people do not have brains.

It is also common practice to use the term 'brain' to mean the 'brain-body system.' There are many systems-within-systems at work in the human organism. A lot of you writers seem to be objecting to your own unfamiliarity with this branch of study. Not that it isn't of some value, but it really bogs the blog.

"Imagine taking a brain out of someone's head and using some kind of mad scientist technology to keep it alive in a glass jar. You'd still have a living brain, in principle one capable of all the same internal activity as when it was still in a body. You wouldn't have mind."

If you had "the same" internal activity, you'd, in principle, have a mind. However, you wouldn't have the same internal activities with a disembodied brain because the brain and body form an inseparable system.

"Neurons play a role in any tennis match, true enough. But is there really any demand for a neural theory of tennis?"

Tennis is a sport played by human creatures, so cognitive science would be working to identify the various systems involved in realizing physical achievements. Studying how these cognitive processes are integrated and maintained in the form of a tennis player (or a piano player) and how they constitute its mental life , is what a cognitive scientist is trying to do. It's not about a theory of tennis, it's about a theory of human action, which is to say, human thought, feeling and action.

"Can you really make sense of the behavior of tennis players by looking at neural structures?"

Can you make sense of the movements on the face of a clock if you know its inner workings? Sure you can. But neural science doesn't attempt an explanation at a level that makes no sense. If we wanted to know why (player A) hit a backhand return, we would not ask a neuro-scientist, we would ask the player or her coach. That is, we would want the answer to be appropriate to the level of the question. Still, this does not rule out that there might be several relevant neural-level matters that bear upon the "why did the player hit a backhand" level.

I could be wrong but your underlying assumption seems to be that skillful behaviors can't be formalized, therefore, a dance, a game of tennis (any skillful preformance), must in principle be unformalizable. If something is formalizable, we can (by definition) articulate the rules for its operation. We can't seem to generate all the sufficient rules for skillful behaviors like a dance, or playing a song on the keyboard artistically. But if one were describing the event fully from top to bottom, then there would be many levels of neural events that correspond with certain experiences, but none may answer the question as to why this move or why that move. To get those answers we need to know about motivations and personalities and game strategies, etc..

"Do the physical structures of the tennis court, the material forces at work in balls, rackets and nets, the macro-scale physiology of the players and the social relations of sporting play no role whatsoever in it?"

Of course they do; how a person integrates all those elements will bear a great deal on how he or she plays tennis.

"Suppose I am a blind man, and I use a stick. I go tap, tap, tap. Where do I start? Is my mental system bounded at the handle of the stick? Is it bounded by my skin? Does it start halfway up the stick? Does it start at the tip of the stick? But these are nonsense questions."

If you are a skilled blind man, the boundary point between you and the world is extended to the tip of the probe (walking stick). All sensations upon the hand holding the stick are projected meaningfully to the tip of the stick as sensations that bear directly onto (are experienced as delineating) the world.

On the other hand, with novice stick users, the boundary point is at the hand-end of the stick. While the skilled blindman attends to the tip of the stick through the hand sensations, the novice attends to the sensations of the butt end of the stick upon his hands. Both stick-users are integrating information, but only one is doing it skillfully. When integrated skillfully, the brute sensations of force upon ones hand can articulate the sublety of the surface of an orange peel, while the same sensations in a lesser integrator (less skilled) will do little more than form the impression that hard ground is causing the stick to press against ones palm.

"This basic failure of "embodied cognitive science" to actually embody its cognition in actually existing bodies is the thing that I most have against it."

I'm not sure what your comment even means.

Cognition is a label given to an event or a series of neural events of the brain/body system.

Are you saying you are not convinced that the firing of systems of neurons constitutes the computational side of mental operations and that the resulting response (following in time), in the form of systems of neurons firing, is not the awareness of that process?

Finally, you assert that cognition is not contained in the brain. And yet we know there is cognition in the brain. Computation in other non-brain-localized systems is not ruled out, but the brain is where inbound systems converge and where the outgoing circuits originate, and it's where all or most of the computational units reside. Cognitive science understands that the brain and body are systematically related which means you can't account for the functional properties of either one without including the other. But that is what embodiment means.

"As for grounding things in metaphor... I was once an advocate of this approach. I'm not anymore."

It is not that "things" are grounded in metaphor; it is rather that meaning itself relies upon basic mataphors that are a direct result of interacting in a physical world with this particular type of physical body. These physical interactions share similarities with other acts and form the basis for extending meanings beyond the primary, experiential ones. A disembodied conceptual system means that there is no means of grounding the meanings of the concepts in anything other than other, ungrounded symbols.

[thomas]

QUOTE"language is produced and controlled by human brains active in the world..."

RESPONSE:That rings odd to my ears. Seems overly objectified or reified or some sort of fied.
Languaged is produced and controlled by human beings active in their social world. That seems a more realistic and practical statement.

Response to response: The notion of "world" in a philosophical context, already contains the social aspect. Just as "language" here references "human" language, it is understood as a given that the human world is a social one -- as are many animal dynamics.