Mind-body problem-Chomsky/Nagel

[apeiron]

From his most recent talk this is his argument (see video starting at ~27:00 minutes):

Hah, he still knows how to make himself the centre of attention, doesn't he. Fling around the outrageous comments, then retreat into opaqueness so he can't be pinned down by the stung critics. It's a tactic that has longed worked for his political views as well.

Geoff Pullum responds here...http://biolinguistica.wordpress.com/2011/11/20/1516/

These recent talks and papers share a steadfast refusal to engage with anything that might make the debate about the poverty of the stimulus (POS) an empirical one. They issue blanket dismissals of nearly all modern cognitive/linguistic science as worthless, and sweep aside whole genres of work on the basis of what seems to be extremely shallow acquaintance. Claims about parallels in the natural sciences feature prominently, as does a preference for authority over evidence...

...So he portrays current skepticism among cognitive scientists about linguistic nativism as not just obtuse, but actively harmful, a threat to our whole discipline. This is an interesting (if rather risky) new way of stoking enthusiasm for linguistic nativism: appeal to linguists’ self-interest and desire for security (you don’t want to be shut down, do you?). But it’s hard to take seriously. Linguistics is not going to die just because a fair number of its practitioners now have at least some interest in machine learning, evolutionary considerations, computational models of acquisition, and properties of the child’s input, and are becoming acquainted with probability theory, corpus use, computer simulation, and psychological experimentation — as opposed to waving all such techniques contemptuously aside...

...The argument from absence of stimulus is pretty much demolished by this Bayesian insight: the argument form simply is not valid. And for people who use the phrase “the logical problem of language acquisition” (as linguistic nativists have been doing since 1981), that ought to mean something. It certainly seems to me sufficient to justify including at least a brief introduction to Bayesian statistical reasoning in the education of every theoretical linguist...

...Lieberman notes that dramatic evolutionary developments like disappearance of lactose intolerance or radical alteration in the ability to survive in high-altitude low-oxygen environments can take place in under 3000 years; yet (as Chomsky stresses) the evidence that any human being can learn any human languages is strong, suggesting that UG shows no genetic variation at all. Why would UG remain so astonishingly resistant to minor mutations for so many tens of thousands of years? There is no selection pressure that would make it disadvantageous for Australian aborigines to have different innate constraints on movement or thematic role assignment from European or African populations; yet not a hint of any such genetic diversity in innate linguistic capacities has ever been identified, at least in grammar. Why not? Chomsky’s response is basically that it just happened...

 

[bohm2]

Hah, he still knows how to make himself the centre of attention, doesn't he. Fling around the outrageous comments, then retreat into opaqueness so he can't be pinned down by the stung critics.
Geoff Pullum responds here...http://biolinguistica.wordpress.com/2011/11/20/1516/

Do you find Geoff Pullum's argument against the "poverty of stimulus" argument convincing?

We conclude that linguists have some additional work to do if they wish to sustain their claims about having provided support for linguistic nativism, and we offer some reasons for thinking that the relevant kind of future work on this issue is likely to further undermine the linguistic nativist position.

Empirical assessment of stimulus poverty arguments
http://www.ucd.ie/artspgs/research/pullum.pdf

Personally, I agree completely with Chomsky on this issue for the relatively simple and seemingly trivial (for me) reasons he offers here:

Poverty of stimulus problems are ubiquitous. Every aspect of growth and development poses huge poverty of stimulus problems. Now the term isn't used in biology and the reason is it's taken to be so obvious that there is no need for a term, so it's obvious that there is a poverty of stimulus problem when humans develop arms instead of wings or a mammalian visual system but not an insect visual system. There is a stimulus. There's external data like nutrition but there's – no one even bothers to argue about it - there is no way for nutrition to determine that you have a mammalian visual system so that's got to be accounted for by something internal, some genetic property. And then you go on to try to find out what it is and ask why it's that way and not some other way. In the case of language, there is a term, poverty of stimulus, and it's considered highly controversial, but just about everything about language is considered highly controversial, even if it is perfectly obvious, a total truism.

Poverty of Stimulus: Unfinished Business
http://www.stiftung-jgsp.uni-mainz.de/Bilder_allgemein/Mainz_transcript_edited.pdf
‘Poverty of the Stimulus’ Revisited: Recent Challenges Reconsidered
http://csjarchive.cogsci.rpi.edu/proceedings/2008/pdfs/p383.pdf
Poverty of the Stimulus Revisited
http://www.ucl.ac.uk/psychlangsci/news-events/seminars-talks/berwicketal2011

 

[apeiron]

Personally, I agree completely with Chomsky on this issue for the relatively simple and seemingly trivial (for me) reasons he offers here:

It's just a crazy strawman argument. Kids are soaked in language from birth.

You cited Pullum's paper. What in it do you dispute?

...it is calculated that a child in a working-class family will have heard 20 million word tokens by the age of 3, and a child being raised in a family on welfare will have heard only 10 million (p. 132). Nonetheless, even in a welfare household, the cumulative exposure to language use of a 3-year-old amounts to ten times the entire extant corpus of Old English literature...

...By these numbers, even a welfare child would be likely to hear about 7,500 questions that crucially falsify the structure-independent auxiliary-fronting generalization, before reaching the age of 3...

And how do you make sense of Chomsky's statements like this?...

it's obvious that there is a poverty of stimulus problem when humans develop arms instead of wings or a mammalian visual system but not an insect visual system.

You don't learn an arm or a stomach. But you do learn to see, hear, move, and all the other stuff a brain does. Someone needs to take a course on infant brain development. Then quit comparing apples with oranges.

http://www.jneurosci.org/content/13/5/1916.full.pdf

 

[bohm2]

It's just a crazy strawman argument. Kids are soaked in language from birth. You cited Pullum's paper. What in it do you dispute?[/url]

That's why I also posted the other links by Chomsky et al. that go into more detail. It seems the debate between him and Geoffrey Pullum go farther back than this:

On Formalization and Formal Linguistics
http://linguistics.berkeley.edu/~syntax-circle/syntax-group/spr06/chomsky1990.pdf

And I think I'm starting to understand why you don't favour the nativist approach. You favour the empiricist competitor model championed by Friston's 'Bayesian Brain' theory, according to which brains are hierarchically organized statistical inference machines operating via recurrent cascades of predictive coding (the paper you posted before):

The free-energy principle: a unified brain theory?
http://www.fil.ion.ucl.ac.uk/~karl/The free-energy principle A unified brain theory.pdf

I think Chomsky and some of the authors do offer arguments against these type of Bayesian models, as per links I provided. I will look at the stuff more carefully when I finish my stupid exam. They made us have an exam on January 2/2012. I hate my college.

 

[apeiron]

And I think I'm starting to understand why you don't favour the nativist approach. You favour the empiricist competitor model championed by Friston's 'Bayesian Brain' theory, according to which brains are hierarchically organized statistical inference machines operating via recurrent cascades of predictive coding (the paper you posted before):

Exactly. As a reductionist enterprise, science is always seeking to divide itself into either/or scenarios. Something is fundamental, everything else is emergent.

But every definite thesis must harbour its equally definite antithesis (the definition of a dichotomy) and so if someone like Skinner arises to say it is "all just associative learning", someone else gets the job of speaking for the other side, and becoming celebrated for saying "no, in fact it is all just innate knowledge".

So what I actually always seek is the middle ground that describes instead the interaction between what others have divided. And neural network approaches are an example of a fruitful balance between dynamics and computation, nature and nurture, structure and process.

The social dynamics of science really demand that Chomsky set himself up as "the clear other" in major debates. It is the way to make a career. And this is what he is very successful at. It is the trick that makes him even more famous in political discourse as well.

This is why, in the speech you cited, he creates a cartoon version of his enemy - scientists who are only interested in "communication", not real language and its "deep" structure. Or scientists who are like Skinnerians in believing the genome is sculpted gradually, gene by gene - the exact biological equivalent of his bogeyman of blind associative learning, statistics without informational structure.

So I don't favour a nativist approach, nor whatever is its dichotomistic opposite. Instead, if they are a valid dichotomy, then I would favour whatever arises in between as a scientific description of their systematic interaction.

To be against Chomsky, doesn't mean I am for some modern version of Skinner. And what really annoys me is Chomsky continuing to play a stale old game when confronted with modern science that tries to describe material reality from a systems perspective. It may serve his reputation, but it does not serve human knowledge.

Now you are trying to do the same thing with me. If I am not a nativist, then I must be some kind of other - an empiricist. And that is after I don't know how many of my posts highlighting the "deep structure" to be found in Peirce, Salthe, Pattee, Rosen, etc.

Analysis must be matched by synthesis, differentiation by integration. Once ignorance has been divided, the truth is to be found in the space cleared in-between.

 

[bohm2]

The social dynamics of science really demand that Chomsky set himself up as "the clear other" in major debates. It is the way to make a career. And this is what he is very successful at. It is the trick that makes him even more famous in political discourse as well.

I'm incapable of understanding such stuff even if that was the case. And I think it's the same with political discourse. I'll just stick to stuff that is at least within my capabilities. I seem to be right in the middle of this Venn diagram:

 

[bohm2]

This is the major author with Pinker who debated Chomsky/Hauser/Fitch on the evolution of language. I thought this quote on consciousness in a recent preliminary paper is interesting. For whatver reason, I think he is mistaken but it is possible that he might be right:

First he notes Chomsky's position with respect to language and thought:

From this perspective on the language faculty and its place in the mind/brain, let us consider a speculation of Chomsky’s concerning the evolution of the language faculty. For decades he has asserted (e.g. Chomsky 1975) that the basic function of the language faculty is not communication, and that language is in fact poorly designed for communication. More recently (e.g. Chomsky 2000, Berwick and Chomsky 2011) he has added an argument to the effect that we use language mostly to talk to ourselves, in an inner monologue or Joycean stream of consciousness, and he suggests that if language is “designed” for anything, it is for thought. The “externalization” of language as sound, he suggests, was a later stage in the evolution of the language faculty. As we saw in section 6, he further suggests that “externalization” may not even have required a cognitive innovation specific to language.

And contrasts it with his view:

My analysis of this situation (Jackendoff 1987a, 1996c, 2007a, 2012) is that Conceptual Structure, i.e. the formal structure of the thought conveyed by a sentence, is almost completely unconscious, and that what we experience as our inner monologue is actually the phonological structure linked to the thought. We are aware of our thinking because we hear the associated sounds in our head...

Consciousness is linked not with thought, but with phonology and the other forms of perception and perceptual imagery shown in Figure 7. That is, one has the experience of thought through phonological, visual, haptic, proprioceptive, and perhaps auditory imagery. As far as I know, this observation has not been made by any of the numerous philosophers and neuroscientists investigating consciousness. This is in part because they treat “language” as a single cognitive function, failing to recognize the crucial distinction between phonology, syntax, and semantics. As a result, they think of phonological imagery as simply “linguistic imagery” – a characterization that is too coarse to distinguish between a “mere” sound pattern and the thought it expresses. If this account is on the right track, our inner monologue is possible only by virtue of having words – and we have to learn words in the context of a communicative speech community. We can only talk to ourselves by virtue of having learned to talk to others. To be sure, speech and inner speech can enhance thought, by making it possible for us to attend to the combinatorial structure of thought, through its phonological proxies (Jackendoff 1996c, 2007a, 2012). But inner speech is not the thought itself (or the thought alone), as Chomsky appears to be claiming. Going back to evolution, I conclude that hominids could not have had an inner monologue until they had a language with phonology in it. I further conclude that the advantages of language for thought are a byproduct – albeit a substantial byproduct – of its advantages for communication, not the other way round, as Chomsky claims.

Jackendoff does consider deaf children on a footnote on p. 33 and attempts to explain it but I'm not sure he succeeds.

http://ase.tufts.edu/cogstud/incbios/RayJackendoff/humanlanguage.pdf

 

[apeiron]

For whatver reason, I think he is mistaken but it is possible that he might be right:

So, is there a reason why he is mistaken?

First he notes Chomsky's position with respect to language and thought: The “externalization” of language as sound, he suggests, was a later stage in the evolution of the language faculty.

This must count as Chomsky's single most egregious error.

And contrasts it with his view: We can only talk to ourselves by virtue of having learned to talk to others.

Jackendoff is obviously correct, but his error was to think that this is in anyway an original thought.

 

[Pythagorean]

Toddler's have no obvious vocal feedback. Possibly may rely on social cues from parents to regulate proper speech patterning.

Children and adults reacted to this manipulation by changing their vowels in a direction opposite to the perturbation. Surprisingly, toddlers' speech didn't change in response to altered feedback, suggesting that long-held assumptions regarding the role of self-perception in articulatory development need to be reconsidered.

Current Biology 22, 1–5, January 24, 2012, Elsevier Ltd All rights reserved DOI 10.1016/j.cub.2011.11.052

 

[bohm2]

So, is there a reason why he is mistaken?...This must count as Chomsky's single most egregious error...Jackendoff is obviously correct, but his error was to think that this is in anyway an original thought.

These authors suggest the following evidence:

The development of a regularized, structured homesign system in the face of little to no linguistic input, including parental input, gives credence to those arguing for innate language via Poverty of the Stimulus. The fact that the deaf children were found to form recursive, creative, and meaningful signs and systems similar to the natural and more developed languages of the world by their own invention would indicate an innate drive to linguistic communication (Goldin-Meadow et al., 1984). As Feldman, Goldin-Meadow, and Gleitman 91978, p. 408) stated,

We have studied the communicative system developed by young children who are as radically deprived of language input as can be imagined. We have found that these linguistically isolated individuals display communicative skills that are language-like, despite their deprivations. We conclude that there are significant internal dispositions in humans that guide the language acquisition process. Such expressive output from such impoverished, and indeed almost nonexistent, input supports the Poverty of the Stimulus argument and indicates an internal system for language...

A more specific situation can be seen in the case of Simon, a profoundly deaf boy who is the son of deaf parents who learned ASL as teenagers...Additionally, Simon’s performance on the use of movement morphemes was on par with that of the compared deaf children from native-input backgrounds (Newport, 1999). As there is no outside influence to support this compensation, such an improvement points towards an internal correction mechanism and drive towards convergence upon the standard form.

The author concludes:

These last two observations imply that when a feature is simply an element of one specific language some initial input is needed, but when it is universal it can be internally realized. The evidence for an ability to create universals without input suggests an internal, innate language faculty and supports Poverty of the Stimulus-type arguments.

I haven't looked at other similar deaf children research, etc. and I didn't look at her arguments closely.


Incomplete Input as a Poverty of the Stimulus Argument
http://www.dartmouth.edu/~mgullick/gullickpovertystim07.pdf

 

[apeiron]

I haven't looked at other similar deaf children research, etc. and I didn't look at her arguments closely.

You seem to be entangling several separate questions.

1) the evolution of language
2) the neurology of language
3) the relationship between language and thought

If you are citing the evidence of the congenitally deaf in relation to (3), for example, I could point to the studies that show the deaf with fluent signing use signing in their interior dialogue. And those without were once regarded as mentally retarded, on a par with animals, in their cognition.

This then perhaps goes to the other two questions as it illustrates the plasticity and lack of cognitive modularity/hardwiring in the human "faculty" for speech and higher order thinking. It can be related also to the findings that Broca's area is a premotor area for hand as well as oral motor planning.

Then arguments can begin whether this means a gestural language predated oral language, or as seems more plausible, the brain was pre-adapted for syntactical language because of about a million years of evolution for tool-making and tool-use.

But it would really help if you could spell out the exact nature of your hypothesis here.

The bare basics of Chomsky's current views are maybe not controversial. Langauge must be some combination of a genetic endowment, developmental environment and structural principles - the last being the least well defined.

But Chomsky still has big blindspots in his thinking.

- Like the belief that language IS thought, rather than scaffolds thought.

- Like that there is not abundant evidence that sociality has been a prime driver of anthropoid neural evolution and so deserves to be the default hypothesis when it comes to communicative capacity.

- Like there is a problem in the "poverty of stimulus", when there is neither any great lack of stimulus, nor a lack of models, such as generative or Bayesian neural nets, that can manage fast learning.

- Like the fact that the evolution of a vocal tract is precisely the kind of architectural constraint on the unfettered recursive abilities of the brain that would count as one of his "deep structure" principles.

- Like that there is some kind of grammatical module in the brain when current research is revealing just how distributed the syntactical machinery actually is.

For example, three key elements of grammar: three different bits of the brain - only one in the old putative "grammar module".

...three seemingly distinct syntactic operations are supported by mechanisms in distinct brain areas – Movement in Broca’s (perhaps to an extent in Wernicke’s) are; Dative Shift in posterior portions of the right frontal lobe (vPCS, aINS); reflexive-antecedent binding in an anterior part of the right frontal lobe (SFG).

http://freud.tau.ac.il/~yosef1/papers/Grodzinsky_BR.pdf

And as I said, Broca's is now understood to be a premotor area for hand control as much as vocalisation. So there is nothing functionally unique in this organisation either. Broca's just contributes certain computational activities to the mix.

 

[apeiron]

Toddler's have no obvious vocal feedback. Possibly may rely on social cues from parents to regulate proper speech patterning.

Yes, this is another thing they need to learn. And can't even begin to learn until the higher cortex has matured enough to form connections to the parts of the brain that have to be regulated.

 

[bohm2]

- Like there is a problem in the "poverty of stimulus", when there is neither any great lack of stimulus, nor a lack of models, such as generative or Bayesian neural nets, that can manage fast learning.

With respect to the skeptical position on "poverty of stimulus" is that something that you believe to be true or is it something that most linguists/psychologists/cognitive scientists believe? I was under the impression that most in these fields do think it's valid. I'm pretty sure even Pinker et al. who question Chomsky on his view on the evolution of language would agree with Chomsky on this point. And last time I sat in courses in this area, I got the impression that most researchers in this area, take this notion of poverty of stimulus as not very controversial (despite Pullum's and Bayesian learning arguments to the contrary). I honestly can't see how anyone can believe otherwise. Regardless, if I took a poll on this issue from scientists in this area, do you think most would favour the "poverty of stimulus" argument or your position that there is no great "lack of stimulus"? Not that it makes it true, either way.

- Yes, this is another thing they need to learn. And can't even begin to learn until the higher cortex has matured enough to form connections to the parts of the brain that have to be regulated.

Are you sure that is what that study Pythagorean posted really means?

 

[apeiron]

With respect to the skeptical position on "poverty of stimulus" is that something that you believe to be true or is it something that most linguists/psychologists/cognitive scientists believe?

Again, Chomsky's POS was a criticism against the atomistic reductionism of Skinnerian Behaviourism. And I am sure that these days - despite the cult-like grip that Behaviourism can have on people - no one would believe it to offer an adequate account of language learning.

It is also obvious just from having kids that they learn in a generative, world-predicting, fashion.

So then the next step is to develop theories of prediction-based learning.

Instead, Chomsky and his disciples (yes, another cult it looks like) exaggerate POS-like arguments to rule out any sort of learning machines. The role of experience must be limited to something like flicking the developmental switch on some sort of genetic hardware.

And even this hardware mustn't be contaminated by learning/experience of the evolutionary kind.

Chomsky would like the genetics to be the product of some kind of Platonic efficiency principles, some UG deep mathematical structure, rather than something that might actually have an evolutionary history.

So again, he trots out a POS argument against the gradualism and atomism of modern synthesis Darwinism. Langauge and symbolic thought erupted too suddenly in human history. Therefore ordinary evolutionary learning must have a minimal explanatory role.

I didn't quite realize it before, but there is a consistent pattern in Chomsky's thought. At every turn, reject the vagaries of the material world and turn inward towards the certainty of logical structure. Empiricism bad, rationalism good.

As antithesis to Skinner's equally rigid thesis, it catapulted Chomsky to fame/notoriety.

Great for him. Meanwhile serious science can get on with the business of synthesis - a balanced approach to evolution like evo-devo, or to brain organisation, like generative neural network theory, or to metaphysics generally, such as Peircean semiotics.

(Yes, I know Chomsky claims these days to be all evo-devo, but again, read what he actually writes and you can see he is still humming the same rationalist tune.)

 

[bohm2]

Instead, Chomsky and his disciples (yes, another cult it looks like) exaggerate POS-like arguments to rule out any sort of learning machines. The role of experience must be limited to something like flicking the developmental switch on some sort of genetic hardware.

I think you are being a bit unfair/biased/judgemental here (e.g. cult/disciples?). He does offer evidence against the Bayesian models if you check the links I provided (PTR refers to the Perfors/Tenenbaum/Regier paper):

Crucially, however, it does not follow that such learners will acquire grammars in which rules are structure dependent. On the contrary, as we show below, the acquired grammars may still operate structure-independently. In short, inferring that language is hierarchical (in PTR’s sense)leaves the original POS (poverty of stimulus) question untouched, and their Bayesian model does not explain the constrained ambiguity facts. Let us examine why...

http://www.ucl.ac.uk/psychlangsci/news-events/seminars-talks/berwicketal2011

And from reading of the literature, most researchers also recognize the limitations of neural networks/connectionist models you favour. For instance consider this recent review:

The neural basis of structure in language
http://www.illc.uva.nl/Research/Dissertations/DS-2011-11.text.pdf

I realize Chomsky seems very dismissive of these models (and maybe comes off as a bit arrogant?) and probably pisses off a lot of researchers. Then again, it could also be his radical politics that many dislike? I'm not sure. I don't really understand his arguments against these models but he seems pretty aware of all the research in this area, if you read his writings, despite some of your old posts arguing otherwise. Anyway, my excursion into this area is indirectly because of my obsession with the "hard" problem of consciousness and so far, I haven't made any progress which kinda sucks.

 

[apeiron]

The lack of recursion in the Piraha language is a well-known challenge to grammar innatism. This is a really fascinating article by Daniel Everett on the controversy - interesting because of the hints at what having a very simple language might have been like when it comes archaic H.sapiens.

http://edge.org/3rd_culture/everett07/everett07_index.html

But also, he is pretty blunt about Chomsky and the lack of testability of UG.

I think that the way that Chomskyan theories developed over the last 50 years has made it completely untestable now. It's not clear what usefulness there is in the notion of universal grammar. It appeals to the public at large, and it used to appeal to linguists, but as you work more and more with it, there's no way to test it—I can't think of a single experiment—in fact I asked Noam this in an e-mail, what is a single prediction that universal grammar makes that I could falsify? How could I test it? What prediction does it make? And he said, It doesn't make any predictions; it's a field of study, like biology.

 

[bohm2]

The lack of recursion in the Piraha language is a well-known challenge to grammar innatism. This is a really fascinating article by Daniel Everett on the controversy - interesting because of the hints at what having a very simple language might have been like when it comes archaic H.sapiens.

http://edge.org/3rd_culture/everett07/everett07_index.html

It's not clear that is a challenge to grammar innatism. There are papers that even question Everett's claim. For example:

Piraha exceptionality: A reassessment
http://web.mit.edu/linguistics/peop...gues_Piraha_Exceptionality_a_Reassessment.pdf

Moreover, even if accurate, Everett's exceptions (and other known exceptions) don't really impact on Chomsky's thesis or his goals for reasons that T. Fitch (who went to study the Piraha to test Everett's claims) writes:

These regularities will certainly incorporate more general aspects of cognition, including aspects of perception, motor control or conceptual structure that predated language in human evolutionary history. From this abstract perspective, UG is not reducible to a list of properties universally found in every language, nor does its existence imply such a list. As Jackendoff puts it, UG is a characterization of the toolkit the child uses in language acquisition, not a list of universal features of adult languages....It is quite unfortunate, then, that many critics have conflated UG and surface language universals, and proffered the discovery of exceptions to some broad regularity as a refutation of UG. As Roman Jakobson, a tireless defender of the search for universals, pointed out, ‘a rule requiring amendment is more useful than the absence of any rule’. The notion of UG is perfectly compatible with a very broad range of linguistic diversity, evolving via cultural processes, and indeed has developed over many decades with precisely this diversity in mind.

An analogy to the diversity and unity of languages is provided by features of our own vast phylum, the vertebrates. Universal vertebrate features are encompassed in the notion of a Bauplan: a ‘body plan’ that includes (or included during development) a notochord running down the spine, and bony vertebrae built around it. To this are attached ribs and generally appendages. A mouth at the front of the animal serves for both food and respiration, and is followed by branchial arches forming jaws, gills or other diverse structures. Many other shared traits also characterize most vertebrates, but these few suffice to make the point: each of these traits is absent or modified in one or a few species, but this does not render the notion of the body plan vacuous....Thus, when scholars cite unusual languages as a refutation of the entire concept of UG, they both overlook the nature of biological systems, which typically allow exceptions, and ignore many explicit hypotheses about UG that have been offered over the years.

I suggest that the general notion of abstract constraints, operating ubiquitously during the development of a system in time and space, provides one such framework (figure 1). Such systems are familiar: a rich body of mathematics exploring such constraints is the theory of differential equations...A differential equation like x^" =ax expresses a constraint on the movement of an object: its acceleration x^" must be proportional to its location x. In general, there are an infinite number of specific paths that could satisfy this constraint...Because there are an infinite number of solutions, we can think of this differential equation as defining a vast family of solutions, some of which may be superficially very different, but all of which have in common that they satisfy the constraint defined by the original equation. In some cases, we can discover a broader ‘general solution’ (e.g. periodic oscillation) that encompasses an entire set of specific, particular functions...The search for universals is akin to the search for a general solution that encompasses all of these particular solutions, and the goal of biolinguistics is to understand, and make explicit, the specific biological constraints that underlie this general solution...These interacting systems entail dauntingly complex systems of partial differential equations involving genes and the epigenetic control of their expression, brains and their self-wiring depending on the organism and its environment, and individuals as part of cultural systems.

See attached thumbnail for Fitche's analogy/framework.

Both top-down approaches (invoking cultural and historical factors) and bottom-up or ‘reductionist’ approaches (e.g. gene or brain-focused research) will be important for a full characterization of this complex system. No one expects such a task to be easy. Equally, no one can deny the fundamental significance of the search...Rejections of the search for universals, based on a few exceptions to some otherwise universal rule, miss the point of this endeavour.

Unity and diversity in human language
http://rstb.royalsocietypublishing.org/content/366/1563/376

 

[apeiron]

Moreover, even if accurate, Everett's exceptions (and other known exceptions) don't really impact on Chomsky's thesis or his goals for reasons that T. Fitch (who went to study the Piraha to test Everett's claims) writes:

Yes exactly. One minute, UG is a genetic template with experience setting a hardwired switch in of two positions on a large number of grammatical rule settings. The next, it is a very loose claim about recursion and vague noises about genetic endowments and "third factors".

Where UG is definite, it is wrong, where it is vague, it fits just about anyone's theory of language.

I have identified what seem to me the principle blindspots in Chomsky's approach to language evolution. Have you got any response on these specific points?

- Like the belief that language IS thought, rather than scaffolds thought.

- Like that there is not abundant evidence that sociality has been a prime driver of anthropoid neural evolution and so deserves to be the default hypothesis when it comes to communicative capacity.

- Like there is a problem in the "poverty of stimulus", when there is neither any great lack of stimulus, nor a lack of models, such as generative or Bayesian neural nets, that can manage fast learning.

- Like the fact that the evolution of a vocal tract is precisely the kind of architectural constraint on the unfettered recursive abilities of the brain that would count as one of his "deep structure" principles.

- Like that there is some kind of grammatical module in the brain when current research is revealing just how distributed the syntactical machinery actually is.

 

[bohm2]

Like the fact that the evolution of a vocal tract is precisely the kind of architectural constraint on the unfettered recursive abilities of the brain that would count as one of his "deep structure" principles.

I'll start with this point. I'm doing this in between studying for my stupid, retarted exam but I''ll try...This is actually Fitch's specialty but here is a popular article of his on this point:

Let’s start with anatomy. Humans have an unusual vocal tract: the larynx (or voicebox) rests low in the throat. In most other mammals, including chimpanzees, the larynx lies at a higher point, and is often inserted into the nasal passage, creating a sealed nasal airway. In fact, humans begin life this way: a newborn infant can breathe through its nose while swallowing milk through its mouth. But as the infant grows, the larynx descends, and by the age of 3 or 4 this feat is no longer possible. The reconfigured human vocal tract allows the free movement of the tongue that is crucial to make the many distinct sounds heard in human languages.

For a long time, the descended larynx was considered unique to our species, and the key to our possession of speech. Researchers had even tried to place a date on the emergence of language by studying the position of the larynx in ancient fossils. Evidence from two different sources of comparative data casts doubt on this hypothesis. The first was the discovery of animal species with permanently descended larynges like our own. We now know that lions, tigers, koalas and Mongolian gazelles all have a descended larynx – making it a convergent trait. Since none of these species produce anything vaguely speech-like, such changes in anatomy cannot be enough for speech to have emerged.

The second line of evidence is even more damning. X-ray observations of vocalising mammals show that dogs, monkeys, goats and pigs all lower the larynx during vocalisation. This ability to reconfigure the vocal tract appears to be a widespread, and probably homologous, feature of mammals. With its larynx retracted, a dog or a monkey has all the freedom of movement needed to produce many different vocalisations (see diagram, right).The key changes must therefore have occurred in the brain instead.

The evolution of language-Fitch
http://www.newscientist.com/data/do...tant_expert_6_-_the_evolution_of_language.pdf

Like there is a problem in the "poverty of stimulus", when there is neither any great lack of stimulus, nor a lack of models, such as generative or Bayesian neural nets, that can manage fast learning.

I tried to answer this point in post #330. You might look at the link because I'm not a linguist to fully understand their arguments but personally I think the point is obvious to me. I think it's the same in science, mathematics and musical knowledge. The environment is way too poor to allow for this type of knowledge.

 

[apeiron]

Anyway, my excursion into this area is indirectly because of my obsession with the "hard" problem of consciousness and so far, I haven't made any progress which kinda sucks.

Indeed, getting back to the OP , the initial comments you cite from Chomsky seem perfectly reasonable in themselves. You simply can't have a mind~body problem if there is no definite story on what the material realm is all about. If physics is incomplete, you can't even say with any surety whether there is a hard problem or not.

Well, first thing to say is that a lot of people feel there is a mind~body causal issue. So they must feel they know enough about the truth of material reality to believe that an explanatory gap is a major difficulty.

And the justification they give is that they know the "style" of material causal explanations, and even if current physical theory is an unfinished story, further expansions of physical theory will have the same style and so will still not be able to bridge the explanatory gap.

At which point, you can reply either that the style is indeed going to be correct, and one day we may actually find the further ingredients that right now we can't even imagine. This is the panpsychist or quantum consciousness line of thought. One day, the mental will be shown to be a physical property of matter, an atomistic aspect of nature. So our view of causal style is already correct, and we simply need to keep digging to find a material basis to mental experience.

The alternative view is that causality is more complicated.

What we have been talking about here is the reductionist view of reality where everything reduces to a substance ontology, with its embedded principle of locality and its web of additive effective causality. We can already see from quantum theory that "reality is not like that". And we can also tell from biology that more complex causal concepts are needed. And there has been a larger model of causality since ancient times - Aristotle's model of the four causes.

So then comes the choice. Do you stick with a model of causality which its own proponents believe to have intractable problems (mind dualism, quantum nonlocality, epiphenomenal emergence, first causes), or switch to a larger model where all these things are features rather than bugs?

I note that you are now approvingly citing papers that talk about "bauplans" and "abstract constraints".

Both top-down approaches (invoking cultural and historical factors) and bottom-up or ‘reductionist’ approaches (e.g. gene or brain-focused research) will be important for a full characterization of this complex system.

Well yes, of course. Even if people still feel they have to use an apologetic tone when talking about downward causality, global constraints, formal principles, and such-like. It is still such a "novelty" in science. See for example this whole issue devoted to downward causality:

http://rsfs.royalsocietypublishing.org/content/current

There are differing views in science, as well as in philosophy, about the reality of top-down causation. The aim of these articles is to provide answers convincing to both scientists and philosophers. The need is for an interdisciplinary dialogue, which we hope this Theme Issue provides. The issue arises in physics (Bishop [1]; Loewer [2]), chemistry (Scerri [3]), microbiology (Jaeger & Calkins [4]), epigenetics (Davies [5]), evolutionary biology (Okasha [6]), physiology (Noble [7]), neuroscience/psychology/cognitive science (Berntson et al.[8]; Atmanspacher [9]), social science (Elder-Vass [10]) and computer science (Booch [11]). Many examples provide good evidence for top-down causation. Thus, in physiology, it is taken for granted, and it is crucial in cognitive science.

So again, the hard problem boils down to an inadequacy of an overly reduced model of causality. A material conception of nature - as just the outcome of material and effective causes - is not enough to model complexity. It doesn't even look enough to model simplicity. And once you accept the concrete reality of downward causation - of formal and final cause - is there still a hard problem of consciousness, an explanatory gap between matter and mind, information and meaning? Or just an explanatory dichotomy?

If we apply this to Chomsky's thinking in the OP, I think he is wrong to suggest that our material explanations are incomplete, as if more of the same might complete them. Instead, even our fundamental theories would need to include the complementary aspects of nature to complete them. Reductionism needs to be fixed by holism, not by even more reductionism.

And then if we apply this larger model of causality to the issue of human grammar, then I think it can lead to some pretty concrete theories about how it works. The basic story is about the construction of constraints. Syntax is the top-down construction of constrained semantics. Just as genetics is the top-down construction of constrained metabolic dynamics, or dissipative process.

Taking this view, we can even measure language production in entropy terms I think. The more constrained the semantics, the greater the number of alternative meanings that have been disposed. So grammar is a way of making our ideas precise - by the wasting of semantic degrees of freedom.

As I say about Friston's work, mind science is moving to place itself on thermodynamic foundations, much as theoretical biology has done. Thermodynamics is naturally a "four causes" or systems approach to modelling reality, so makes a better conceptual basis than the Galilean/Newtonian atomism that was the original scientific revolution, and which continues to exercise such a grip on the popular imagination.

If I wanted to get to the bottom of human language or consciousness generally, or QM for that matter, I would start from thermodynamics too. Or infodynamics, to distinguish dissipative structure theory from the kind of closed, dead, equilibrium models that are the thermo-world of classical reductionism.

 

[apeiron]

I'll start with this point. I'm doing this in between studying for my stupid, retarted exam but I''ll try...This is actually Fitch's specialty but here is a popular article of his on this point:

This is one of those he said/she said controversies in science where you have to weigh the balance of evidence. So there are a series of standard counterpoints to what Fitch just said.

For instance, what other species have the L-shaped kink which allows a fat tongue to actually separate the initial vocal cord sound production from its later fine-tuned modulation? The descended larynx is just one of a constellation of adaptations as I highlighted in post #283.

It has been pointed out that the radical modification of the hominid vocal tract involves a whole hierarchy of dichotomies in itself. So just to control the equipment demands a recursive motor capacity.

http://zimmer.csufresno.edu/~chrisg/...EvolHierar.pdf

Brown and Golston note how there is a first divide between the larynx and the supralaryngeal filter (or voice box and basically the rest), then the filter divides in turn into its oral vs nasal paths, the oral into its lips vs tongue, the tongue into its front vs back, the front of the tongue into its pointy tip vs broad blade configuration. There is a whole tree of sub-divisions.

You can always explain away some single fact, but it is the weight of facts that carries the case in these kinds of complex arguments.

Perhaps I didn't make it plain. You can see in the design of the vocal tract a physically-embodied recursive hierarchy - "a whole tree of [dichotomous] sub-divisions".

OK, if you are a Chomskyite and apparently very concerned about deep structural principles and the evolution of recursion, then right there in front of your nose is an example of something that must have evolved.

Was this likely to be the result of some blast of cosmic rays, a hopeful monster mutation, because no other standard graduationalist evolutionary hypothesis could explain such an extraordinary development?

No, the evolution of a recursive phonology looks pretty simple, doesn't it?

The radical reorganisation of the brain, or the sudden creation of a brand new functioning module, do seem a tall order for Darwinism - even with the aid of evo-devo fast-tracking. So the idea that it could be syntax first, phonology second, is a real stretch for evolutionary theories.

By contrast, "phonology first" looks an evolutionary doddle. And it then provides the concrete foundation for the subsequent fast-development of syntactical ability. You want symbolic recursion? Well, you've already got vocal recursion to piggyback on. You are way more than halfway there.

 

[bohm2]

And the justification they give is that they know the "style" of material causal explanations, and even if current physical theory is an unfinished story, further expansions of physical theory will have the same style and so will still not be able to bridge the explanatory gap.

Yes, that's Nagel's argument which Chomsky questions. I'm not exactly sure why because in other passages Chomsky quotes Russell approvingly that physics may only reveal the causal/relational properties of physical objects never allowing us to know anything about their intrinsic nature. But, we do seem to know something about the intrinsic structure of one object as Lockwood points out:

Do we therefore have no genuine knowledge of the intrinsic character of the physical world? So it might seem. But, according to the line of thought I am now pursuing, we do, in a very limited way, have access to content in the material world as opposed merely to abstract casual structure, since there is a corner of the physical world that we know, not merely by inference from the deliverances of our five sense, but because we are that corner. It is the bit within our skulls, which we know by introspection. In being aware, for example, of the qualia that seemed so troublesome for the materialist, we glimpse the intrinsic nature of what, concretely, realizes the formal structure that a correct physics would attribute to the matter of our brains. In awareness, we are, so to speak, getting an insider's look at our own brain activity.

If we apply this to Chomsky's thinking in the OP, I think he is wrong to suggest that our material explanations are incomplete, as if more of the same might complete them. Instead, even our fundamental theories would need to include the complementary aspects of nature to complete them. Reductionism needs to be fixed by holism, not by even more reductionism.

To be fair, Chomsky doesn't make a commitment on this issue. He does leave the option open. He just writes that we can't rule it out. And he never claims that it is "more of the same", I think. Still, I'm still not convinced that even the contextuality/holism implied by Bell's/QM can provide a mechanism for spitting out mental stuff, although it does seem to offer a good model for downward causation, in my opinion. I just don't think this is enough for reasons I gave before. But I'm not sure? I posted this question before in another thread as I find it very interesting and is at the heart of this thread:

Given the unification of chemistry with QM and the unification of molecular biology with chemistry (more recently), does anyone see any hope/hint of similar unification of mental phenomena with present-day physics or is this unlikely even with major changes in a future physics as Nagel claims? If such unification is conceivable, is it possible that by noting "some sort of theoretical inference from the character of phenomenal properties" where we do actually have access to its intrinsic properties, it might help/guide us of what is actually required by a future theory of physics for unification to occur? Kind of how chemistry guided the new physics (QM) in the past. Consider Eddington's remarks:

But in one case—namely, for the pointer readings of my own brain—I have an insight which is not limited to the evidence of the pointer readings. That insight shows that they are attached to a background of consciousness in which case I may expect that the background of other pointer readings in physics is of a nature continuous with that revealed to me in this way, even while I do not suppose that it always has the more specialized attributes of consciousness. What is certain is that in regard to my one piece of insight into the background no problem of irreconcilability arises; I have no other knowledge of the background with which to reconcile it...There is nothing to prevent the assemblage of atoms constituting a brain from being of itself a thinking (conscious, experiencing) object in virtue of that nature which physics leaves undetermined and undeterminable. If we must embed our schedule of indicator readings in some kind of background, at least let us accept the only hint we have received as to the significance of the background—namely, that it has a nature capable of manifesting itself as mental activity.

I think McGinn also hints at this where he writes:

I am now in a position to state the main thesis of this paper: in order to solve the mind-body problem we need, at a minimum, a new conception of space. We need a conceptual breakthrough in the way we think about the medium in which material objects exist, and hence in our conception of material objects themselves. That is the region in which our ignorance is focused: not in the details of neurophysiological activity but, more fundamentally, in how space is structured or constituted. That which we refer to when we use the word 'space' has a nature that is quite different from how we standardly conceive it to be; so different, indeed, that it is capable of 'containing' the non-spatial (as we now conceive it) phenomenon of consciousness. Things in space can generate consciousness only because those things are not, at some level, just how we conceive them to be; they harbour some hidden aspect or principle.

I'm not sure if configuration space where the wave function evolves or the non-locality implied in Bell's/QM might meet McGinn's criteria? There are even some physicists who do feel that these non-local correlations/communications do seem to happen outside space-time or at least defy spatio-temporality. Maybe we shouldn't find such stuff in physics surprising since in some ways qualia/consciousness does seem to suggest that if unification is ever possible, physics will have to involve some properties that need to go beyond locality/spatiality. I'm not sure if that, in itself, is enough. Maybe unification would require much more. Maybe Bohm is right and there really are many more sub-quantum levels that we haven't even scratched and these levels are necessary for unification to be completed? Some authors actually see this russian dolls that Bohm suspects as a positive:

What would a metaphysic of infinite descent look like? The most striking feature of an infinite descent is that no level is special. Infinite descent yields an egalitarian ontological attitude which is at home in the macroworld precisely because everything is macro. Mesons, molecules, minds, and mountains are in every sense ontologically equal. Because there can be no privileged locus for the causal powers, and because they must be somewhere, they are everywhere. So infinite descent yields an egalitarian metaphysic which dignifies and empowers the whole of nature. Treat infinite descent as a working hypothesis, and since all entities turn out to be composite, supervenient, realized, and governed, it emerges that these attributes cannot be barriers to full citizenship in the republic of being. The macroworld, once regained, is not easily lost, even should real evidence for fundamentality arrive. Here I am, a human organism, a macroentity, but in no sense unreal for that. I believe that I am both composed of and dependent on certain cells, which are in turn both composed of and dependent on certain molecules, which are in turn both composed of and dependent on certain atoms, which are in turn both composed of and dependent on certain subatomic particles, which are in turn both composed of and dependent on certain quarks and leptons. We just don’t know whether this chain stops. But from this perspective it seems obvious that my realness does not in any sense turn on whether there are preons and so on below, or not. To see that there is no evidence for fundamentality is already to regain the macroworld.

Is there a fundamental level?
http://www.jonathanschaffer.org/fundamental.pdf

 

[apeiron]

Chomsky quotes Russell approvingly that physics may only reveal the causal/relational properties of physical objects never allowing us to know anything about their intrinsic nature.

Yes, I see you are going to stick to the reductionist view and its paradoxes come what may.

But that Royal Society publication I mentioned reminds why "more is different" when it comes to causation.

The Bishop article is worth reading: http://rsfs.royalsocietypublishing.org/content/early/2011/09/02/rsfs.2011.0065.full

While most metaphysicians focus on the ‘upward’ flow of efficient causation from system components to system behaviour as a whole, complex systems such as convecting fluids present plausible examples of a ‘downward’ flow of influence and constraint on the behaviour of system components. Such behaviours clearly raise questions for a programme of discovering the factors influencing complex systems’ behaviour in the fundamental laws alone, an approach to causation championed by those who favour physical accounts of causation.

There are some interesting points, like that sensitivity to initial conditions proves that local effective cause becomes in principle unmeasurable in realistic dynamical situations. You can't measure reality with infinite precision, and coarse graining does not rescue you.

A reductionist will again protest that there are "hidden variables". Reality is actually in some definite state, infinitesimally specified, even if it is beyond our capacity to measure it. The deep answer is still "intrinsic", even if it is over the event horizon of what can be known.

This is bad ontology, and even bad epistemology, when the world can instead be measured in terms of its constraints. In chaos theory, for example, we can model the global attractors of a system. Why fuss about unmeasurable variables which are "hidden/intrinsic" when science actually can spend its time measuring something concrete?

People are recommending Deacon's new book as a systems/semiotic approach to mind/body issues, though I haven't read it yet myself...

https://www.amazon.com/dp/0393049914/?tag=pfamazon01-20

 

[bohm2]

Why fuss about unmeasurable variables which are "hidden/intrinsic" when science actually can spend its time measuring something concrete?

Primarily because as Strawson notes, consciousness/experience is "the most certainly known natural fact." And the fact that physics/neuroscience has no terms specifically for experiential/the mental (leaving aside some interpretations of QM) seems odd since we know it exists. And here, I'm not deying your arguments above, but I honestly don't think there's a hint of anything that may help us solve this gap in the stuff you cite. I also haven't read the book you linked either although I've read some of Ilya Prigogine's stuff and even went to one of his lectures just before he died. Interestingly, one reviewer of that book seems to be echoing Chalmer's arguments:

The focus then shifts to constraints. Constraints prevent things. They cause things to not happen, they cause them to remain absent and to only be what (otherwise) could have been. Incidentally they cause/allow for other, alternative things to happen. (Naturally, they play a role in organization/morphodynamics.) I have a feeling that this doesn't sound like much of a great insight. It wasn't to me. I don't see what can-in respect to the emergence of mind/consciousness-be gained through that, allegedly new, perspective. For one thing, constraints are physically there. They aren't absent/absential features. For another thing, defining things negatively (a banana is a fruit that is not any fruit other than a banana) is not a new invention. I do not see anything resembling the paradigm shift and revolution Mr. Deacon postulates (and the publisher advertises).

Chalmers makes the same point:

A low-level microphysical description can entail all sorts of surprising and interesting macroscopic properties, as with the emergence of chemistry from physics, of biology from chemistry, or more generally of complex emergent behaviors in complex systems theory. But in all these cases, the complex properties that are entailed are nevertheless structural and dynamic: they describe complex spatiotemporal structures and complex dynamic patterns of behavior over those structures. So these cases support the general principle that from structure and dynamics, one can infer only structure and dynamics.

There is one exception, I think. I do find it odd that the same type of "dualism" pervades the meaning of the wave function in QM. There seems to be this "gap" between the picture of the world provided by the wave function and the world provided by our experience and bridging the two appears just as difficult.

 

[apeiron]

And here, I'm not deying your arguments above, but I honestly don't think there's a hint of anything that may help us solve this gap in the stuff you cite.

OK, you offer no rational arguments against Bishop, you are content simply to cite your beliefs here.

Faith-based positions are indeed impregnable to reason.

 

[bohm2]

OK, you offer no rational arguments against Bishop, you are content simply to cite your beliefs here.

Faith-based positions are indeed impregnable to reason.

Sorry and you're right. I will look over that Bishop paper more closely after I write my exam Monday. I'm having trouble studying and paying attention to that crap on my exam because these discussions on the forum are so much more interesting. So you don't think I'm that close-minded, I did read the P. W. Anderson paper (More is different). I read it before also, but I forgot I read it Not a good omen because I've been forgetting a lot of stuff recently. I think those 8 years of clonazepam may have done some damage to my memory.

 

[bohm2]

I thought this was an interesting criticism of Chomsky's position on the possibility of knowledge:

The epistemological implications of Chomksy’s view of knowledge are worth pursuing in a little more detail. On this view, 'knowledge' is not really knowledge as such since it is not about anything; it expresses our nature, not the world's, since it is a fact about how our brain allows us to apprehend the world and implies nothing about what the world is really like (or even if it even exists). If the structure of the brain itself determines what we can know then there must necessarily be 'sharp limits on attainable knowledge', some problems forever remaining 'mysteries' that we are innately unequipped to solve. And yet Chomsky argues that true knowledge would be possible if biologically accessible concepts and theories happened to converge or intersect with properties of reality, although of course 'there is no particular biological reason why such an intersection should exist'. Nevertheless, Chomsky firmly believes that such incredible intersections do happen (and we can know when they do): 'The successful natural sciences, then, fall within the intersection of the scope of SFF (science-forming faculty) and the nature of the world'. This intersection 'is a chance product of human nature’, 'a remarkable historical accident resulting from chance convergence of biological properties of the human mind with some aspect of the real world' . True knowledge depends on a 'kind of biological miracle'...But where Descartes invoked God to explain the correspondence between thinking and being Chomksy claims that it is 'just blind luck if the human science forming capacity, a particular component of the human biological endowment, happens to yield a result that conforms more or less to the truth about the world'.

This author then tries to show that his position is self-defeating:

The terms he uses, such as 'real world', 'discover', 'knowledge', 'objects' etc, really need to be re-translated into the solipsistic biological idiom. ‘Discovering the real properties of objects’ should be understood to mean something like 'interpreting empirical data as instances of biologically determined categories' or, more grandly, ‘projecting a priori categories of mind onto unknowable things-in-themselves'. By his own arguments, the concepts and technical terms of his linguistic theory must be considered the product of 'SFF'. Whether such things really exist he is not in a position to say. The very theory of mind which Chomsky advocates precludes the claim that the mind as he sees it is real. The simple fact is that if you believe in innate ideas (in original or modernized variants) then you cannot be a philosophical realist. In sum, then, Chomskyan nativism, which 'wavers between an antediluvian spiritualism and a genuinely "vulgar" materialism', is a structure built on Humean and Kantian premisses. Empirical hypotheses or alleged empirical discoveries in linguistics, to the extent that they follow from such premises, are bogus. Chomsky's innate grammar is not a scientific discovery but is falsely inferred from indefensible philosophical premises.

Critical realism and scientific method in Chomsky's linguistics
http://shu.academia.edu/PeterJones/...and_scientific_method_in_Chomskys_linguistics

I'm not sure if his conclusions follow?

 

[Mr.Samsa]

As antithesis to Skinner's equally rigid thesis, it catapulted Chomsky to fame/notoriety.

How is Skinner's account of language "rigid"? Are you confusing Skinner's actual position with Chomsky's strawman which made it look like Skinner believed that language was entirely learnt?

Skinner argued that language was a result of the combination of biological hardware in the brain, and learning processes which allows us to learn specific languages and grammars (as of course Skinner was strongly opposed to blank slate accounts of behavior). How is this "rigid", given that it seems to be consistent with the current dominant theories on language?

 

[apeiron]

Skinner argued that language was a result of the combination of biological hardware in the brain, and learning processes...

What was Skinner's theory about the organisational principles of that biological hardware? Can you provide a precis?

The reason I call Skinner/Behaviourism rigid is because it goes to a methodological extreme and sticks to it. Science likes its polarities, and Chomksy does a good job of standing up for strong rationalism, while Skinner stood for its antithesis of strong empiricism.

For a scientist, strong empiricism seems the more defensible. But I believe that Behaviourism leads mind science up an a-theoretic cul de sac. It is just too simplistic to deal with a complex subject.

So what can Behaviourism actually say about the brain if it is modeling its activities in terms of chained responses? All the emphasis gets put on a link between a single act and a perception of the consequences. To get greater complexity, these individual atoms of behaviour must be chained together.

Now this is nothing like a brain with its hierarchical organisation, its networks and feedback, its anticipation-based processing. Gluing together a bunch of learned responses is not going to get you there.

This does not make Behaviourism flat out wrong. Put a rat in a very simple environment and you will reduce it to very simple behaviour. But as I say, what is wrong is believing that a full and rich theory of the brain/mind can be built from the bottom-up.

You say Skinner was not so rigid in fact. He realized there was more complexity going on. Perhaps like Chomsky, he too would have taken the evo-devo position when pressed, agreeing that biological complexity is shaped by some deep principles of self-organising development. It is not all just a reductionist bunch of lego but a much bigger story of how organisation emerges via the top-down causality of constraints.

Well, if Skinner was arguing for anything more than that the brain was implementing response chaining, I have not come across it. How far did he go in investigating the neural and structural underpinnings of verbal behaviour? Did he ever put forward some "proof of concept" argument that you would actually be able to go from operant conditioning to an account of the cognitive architecture of the brain?

As to which side I take in all this, I think the two great Russian - the psychologist, Vygotsky and the neurologist, Luria - were on the money.

Luria, talking about the history of neurolinguistics (long before Chomsky or Skinner), says the divide was between the associationists and idealists. One extreme thought everything was atomistic chains. The other extreme thought the brain had only a "mental" structure - it was a generalised organ of thought where ideas played freely according to their own rules.

But Luria said everything starts with the concept of the hierarchy. Which is a place in between all these other conceptual extremes. The hierarchy couples the bottom-up and the top-down.

In some sense, Skinner's own approach was hierarchical. Operant conditioning was built on a constrast with lower-order Pavlovian conditioning. But that wasn't really hierarchy theory because the two were not theoretically coupled. Whereas in the Russian tradition - I'm thinking here of Sokolov's brilliant investigation of the orienting response in particular - there was an interactive coupling of attentional and habitual mental processes.

You can see a historic pattern here I guess - Continental rationalism/idealism vs Anglo-saxon empiricism/associationism. With the Russians taking the more pragmatic middle ground.

But anyway, the acid test for me is who is talking hierarchical organisation here?

Chomsky is, of course, but in a confused and surprised fashion. Look, syntax is hierarchical - isn't that weird and impossible to explain in terms of neurology and evolution? Err, no. (The thing that needs explaining is the development of an epistemic cut between the hierarchical organisation of phonology and the hierarchical organisation of the semantics it supports.)

And I don't see any evidence that Skinner ever "got" hierarchies either. But maybe you can cite something to that effect?

 

[Mr.Samsa]

What was Skinner's theory about the organisational principles of that biological hardware? Can you provide a precis?

The reason I call Skinner/Behaviourism rigid is because it goes to a methodological extreme and sticks to it. Science likes its polarities, and Chomksy does a good job of standing up for strong rationalism, while Skinner stood for its antithesis of strong empiricism.

For a scientist, strong empiricism seems the more defensible. But I believe that Behaviourism leads mind science up an a-theoretic cul de sac. It is just too simplistic to deal with a complex subject.

I'm still not quite sure why you think behaviorism represents a 'methodological extreme', or why you think it was a position of strong empiricism? Why would behaviorism lead science up an 'a-theoretic cul de sac'?

So what can Behaviourism actually say about the brain if it is modeling its activities in terms of chained responses? All the emphasis gets put on a link between a single act and a perception of the consequences. To get greater complexity, these individual atoms of behaviour must be chained together.

Now this is nothing like a brain with its hierarchical organisation, its networks and feedback, its anticipation-based processing. Gluing together a bunch of learned responses is not going to get you there.

You seem to be suggesting that behaviorism believes in some kind of simplistic stimulus-response approach to psychology, where each aspect of the 'chain' needs to be linked to a particular stimulus and tied together. Skinner was strongly opposed to such a concept, hence why he dedicated much of his early career to debunking the stimulus-response psychology of the time. He'd agree with you that simplistic chains of responses cannot explain the complexity of behavior.

This does not make Behaviourism flat out wrong. Put a rat in a very simple environment and you will reduce it to very simple behaviour. But as I say, what is wrong is believing that a full and rich theory of the brain/mind can be built from the bottom-up.

But behaviorism can also put a human in a complex environment, and explain/predict their complex behavior. The vast array of complex behavior that behaviorist philosophy has explained over the years (e.g. altruism, self-control, choice, signal detection, etc) certainly adds to the strength of its validity.

You say Skinner was not so rigid in fact. He realized there was more complexity going on. Perhaps like Chomsky, he too would have taken the evo-devo position when pressed, agreeing that biological complexity is shaped by some deep principles of self-organising development. It is not all just a reductionist bunch of lego but a much bigger story of how organisation emerges via the top-down causality of constraints.

Skinner did not need to be "pressed" to accept such a position. His entire philosophy was based on the idea of evolution shaping our bodies and behaviors (not to mention the fact that behaviorism as a whole was created by an ethologist, John Watson, who spent much of his life studying innate behaviors), and he always pointed out that his ideas were an extension of evolutionary theory, not meant as a replacement. This is most evident in his book "Selection by Consequences", where he explicitly describes the comparison between the effects of evolutionary selection on phylogenetic behaviors, and the effects of environmental selection on ontogenic behaviors.

Well, if Skinner was arguing for anything more than that the brain was implementing response chaining, I have not come across it.

I'm not sure why you haven't come across it, he ardently and consistently pointed out that behaviors cannot be understood by looking at the environment alone. This was why he rejected the approach taken by the methodological behaviorists, who looked at behavior in isolation of the biological context that it occurred in. He outlines some of the misconceptions about his behaviorism, and why they're wrong, in his book "About Behaviorism". There is an entire section dedicated to the myth that behaviorists reject the influence on biology on behavior.

How far did he go in investigating the neural and structural underpinnings of verbal behaviour? Did he ever put forward some "proof of concept" argument that you would actually be able to go from operant conditioning to an account of the cognitive architecture of the brain?

He didn't go very far at all because he wasn't a neurologist. He didn't know anything about the brain really, he just knew that environment alone wasn't enough to explain the behaviors he was studying, and he knew that the brain was probably where evolved behaviors and predispositions were stored.

He was interested in learning and the environment though, so he focused on that aspect and let the neurologists figure out the rest.

As to which side I take in all this, I think the two great Russian - the psychologist, Vygotsky and the neurologist, Luria - were on the money.

Luria, talking about the history of neurolinguistics (long before Chomsky or Skinner), says the divide was between the associationists and idealists. One extreme thought everything was atomistic chains. The other extreme thought the brain had only a "mental" structure - it was a generalised organ of thought where ideas played freely according to their own rules.

But Luria said everything starts with the concept of the hierarchy. Which is a place in between all these other conceptual extremes. The hierarchy couples the bottom-up and the top-down.

In some sense, Skinner's own approach was hierarchical. Operant conditioning was built on a constrast with lower-order Pavlovian conditioning. But that wasn't really hierarchy theory because the two were not theoretically coupled. Whereas in the Russian tradition - I'm thinking here of Sokolov's brilliant investigation of the orienting response in particular - there was an interactive coupling of attentional and habitual mental processes.

I'm not sure it's accurate to describe Skinner's hierarchical ideas in terms of operant conditioning being placed on top of classical conditioning, since it's not really accurate to consider classical conditioning "lower order". The two processes run more in parallel, rather than linearly up and down. The hierarchy comes from the complex relationships that these two processes form, with interacting stimulus and response classes, generalisation effects, emergent behaviors being generated from equivalence relations, etc.

If Skinner really held to a simplistic response chain theory, with no hierarchy, then it would seem impossible for him to be able to explain novel behaviors and surely pointing that single fact out should have destroyed the entire behaviorist philosophy?

You can see a historic pattern here I guess - Continental rationalism/idealism vs Anglo-saxon empiricism/associationism. With the Russians taking the more pragmatic middle ground.

Where would Skinner fall in that dichotomy though? His theories rejected associationism, and his position held that empiricism is not the be all and end all of investigation - hence why he emphasised the idea that we needed to come up with theories to explain the unobservable entities inside our heads. If he were a strict empiricist, then why would he support theories of unobservable thoughts and feelings which cannot have any direct empirical support?

 

[apeiron]

He didn't go very far at all because he wasn't a neurologist. He didn't know anything about the brain really, he just knew that environment alone wasn't enough to explain the behaviors he was studying, and he knew that the brain was probably where evolved behaviors and predispositions were stored.

OK, my contention here is that Skinner gave no useful account of either the neuro or social aspects of cognition. And his methods never could. As a programme of science, it gave exaggerated importance to some pretty inconsequential results and its main claim to fame was its empirical rigour. It looked like science at a time when Anglo-saxon psychology was very uncertain of itself, but turned out to be just scientism - the triumph of form over substance.

I agree this is harsh criticism. But perhaps it was because for about a term back in the 1970s, I thought operant psychology was really "it". Then I woke up to the actual paucity of results and the rather cult-like approach taken to teaching the subject.

Maybe my view would be much different if my interests were applied psychology rather than mind science. But right here we are discussing the mind~body problem. And Behaviourism was a way to avoid taking either of those things seriously as the object of modelling.

You say Skinner was not just an extension of the associationist, Darwinian, tabula rasa, tradition. But then where are his theories about the structure of the mind, the architecture of the brain? He may have waved his hand in that direction - even waved it vigorously - but so what?

If Behaviourism does not believe in simplistic stimulus-response chains as the material basis of mind, then can you actually articulate what the big theory is?

I can see it might involve evolved instinctual drives and all that. But again, that is the kind of simplistic notion that evaporates as soon as you pick up a neurology textbook. Yes, you might point to the hypothalamus or reticular activating system. Yet then what? Where is the actual construct to guide your descriptions?

 

[Mr.Samsa]

OK, my contention here is that Skinner gave no useful account of either the neuro or social aspects of cognition. And his methods never could. As a programme of science, it gave exaggerated importance to some pretty inconsequential results and its main claim to fame was its empirical rigour. It looked like science at a time when Anglo-saxon psychology was very uncertain of itself, but turned out to be just scientism - the triumph of form over substance.

I can't help but feel that you're criticising behaviorism based on inconsequential grounds - why would behaviorism need to give an account of the neuro aspects of cognition? With that said, there are cross-overs of behaviorists working with neuroscientists to figure out how the two connect, for example the work of Palmer and O'Donohue (e.g. "http://www.sciencedirect.com/science/article/pii/S0376635702000177").

But as for behaviorism not giving any useful account of the social aspects of cognition, I'm not sure how you can even attempt to claim this. Behavioral theories are the dominant explanations in psychology for things like (as I mentioned above) altruism, self-control, choice, signal detection, etc. These are massive areas of cognition. Whether behaviorism has currently explained the phenomenon of cognition as a whole or not is obviously a completely different, and more complicated question.

And this is actually the second time in two days that I've seen Skinner accused of scientism, which is such a weird claim to me. His philosophy was entirely pragmatic, he did not try to extend science to discuss what is real or true, he was just doing what was useful and productive. He attempted to extend science to areas that were previously considered to be beyond science, yes, but he did not misapply science and claim it was the only source of knowledge.

I agree this is harsh criticism. But perhaps it was because for about a term back in the 1970s, I thought operant psychology was really "it". Then I woke up to the actual paucity of results and the rather cult-like approach taken to teaching the subject.

If you gave up in the 1970s, then that might be why you think there is a paucity of results. It wasn't until 1974 that Baum introduced a quantification of choice, which allowed us to predict nearly every behavior (as all behavior is essentially choice behavior, as suggested by Herrnstein). After the mid-1970s is when the field exploded with invaluable results.

Maybe my view would be much different if my interests were applied psychology rather than mind science. But right here we are discussing the mind~body problem. And Behaviourism was a way to avoid taking either of those things seriously as the object of modelling.

The mind-body problem is not something that can be addressed by science, so it seems a little unfair to criticize Skinner's science on that basis. I only jumped into this discussion to question the mischaracterisation of Skinner's position on language.

You say Skinner was not just an extension of the associationist, Darwinian, tabula rasa, tradition. But then where are his theories about the structure of the mind, the architecture of the brain? He may have waved his hand in that direction - even waved it vigorously - but so what?

I don't understand why such a theory would be necessary? He explicitly rejected stimulus-response psychology and associationist theories, and explained why they could not account for complex behaviors. He wanted to scrap everything we thought we knew about the field and begin slowly as a descriptive science, so he wasn't going to come up with theories of the mind or brain without the evidence to support his ideas.

In other words, we can reject blank slate theories without having to create an entire philosophy of mind in its place, in the same way that we can reject creationist theories without having a working theory of abiogenesis.

If Behaviourism does not believe in simplistic stimulus-response chains as the material basis of mind, then can you actually articulate what the big theory is?

Well different behaviorists have different opinions on the matter (especially when we take behaviorism as the general philosophy of science that underpins behavioral psychology, neuroscience, cognitive psychology, ethology, etc), so I'm not sure which area you're particularly interested in. There's Gilbert Ryle's theory in the "Concept of Mind", and there are various functionalist approaches (since functionalism and behaviorism are interchangeable), and Hayes' "Relational Frame Theory" attempts to explain a number of cognitive issues like Theory of Mind, language, etc.

I think the problem is that you're viewing behaviorism as a discrete entity, rather than a broad label for a philosophy of science that spans many disciplines and researchers. Behaviorism simply states that a science of behavior is possible, and makes no real specific claims beyond that.

I can see it might involve evolved instinctual drives and all that. But again, that is the kind of simplistic notion that evaporates as soon as you pick up a neurology textbook. Yes, you might point to the hypothalamus or reticular activating system. Yet then what? Where is the actual construct to guide your descriptions?

Instinctual drives are of course too simplistic and a mistaken notion, but I'm not sure why you're trying to reduce behavioral science to a level that it doesn't attempt to explain.

 

[apeiron]

With that said, there are cross-overs of behaviorists working with neuroscientists to figure out how the two connect, for example the work of Palmer and O'Donohue (e.g. "http://www.sciencedirect.com/science/article/pii/S0376635702000177").

Thanks for that. I've checked a few more related papers too.

http://psych.stanford.edu/~jlm/pdfs/Tryon%20Connectionism%20Selectionism.pdf
http://www.lcb-online.org/html/5.html
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1284592/pdf/9132463.pdf

It is interesting that all of them stress how out-of-favour Behaviourism still is. They say the field does indeed need to find some biological foundation to be taken seriously. And they all agree that neural network connectionism is that foundation.

From Tryon...

Skinner's functional explanation of behavior has been marginalized within psychology for the same general reason that Darwin's theory was marginalized within biology. No proximal causal mechanisms are available to explain behavioral variation and how contingent consequences can selectively reinforce or strengthen target behaviors...

...this article notes that explanation based on selection outside of PDP connectionism trend, if left unchecked, is that fewer and fewer proponents of applied behavior analysis will have less and less impact on science, clinical practice, and education. Representation and influence in professional societies will continue to wane. It is time to act in new more effective ways before extinction fully occurs.

So in fact Behaviourism is seeking salvation in associationist architectures, and even "a little bit hierarchical" architectures with the references to PDP multi-layer networks in particular.

I'm not sure why you insist that Behaviourism has no truck with associationism. It was there from the start with Thorndike and is back there again with neural nets.

 

[Mr.Samsa]

Thanks for that. I've checked a few more related papers too.

http://psych.stanford.edu/~jlm/pdfs/Tryon%20Connectionism%20Selectionism.pdf
http://www.lcb-online.org/html/5.html
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1284592/pdf/9132463.pdf

It is interesting that all of them stress how out-of-favour Behaviourism still is. They say the field does indeed need to find some biological foundation to be taken seriously. And they all agree that neural network connectionism is that foundation.

Indeed. Firstly, behaviorism still of course attracts bad press due to lingering misconceptions and misrepresentations (mostly due to the painfully bad Chomsky review, where by the end of it you're surprised that he could even spell "behaviorism" given how little he clearly knew about the subject), but this doesn't stop psychologists and ethologists using behaviorist methodology without realising it. There's a good paper on the topic ("http://top.sagepub.com/content/19/2/68.abstract") which interviewed a number of psychology students and even lecturers, and found that many of them held a number of false beliefs about Skinner and behaviorism. For example, a large number of them believed that behaviorism was a blank slate ideology. Secondly, the authors emphasise the need for a biological foundation "to be taken seriously" because that's the product they're trying to sell. You find the same thing with authors trying to promote quantitative theories in psychology, where they point out the supposed "resistance" to quantitative theories and then present their "solution" to the problem they just made up.

The results from behavioral science don't need a biological basis to be validated - the results are true, regardless of whether they can point to a structure in the brain or not. A biological basis obviously strengthens the arguments made, as science works by finding general laws which are applicable across various fields, but it's not "necessary" in that behavioral science doesn't cease to be true until it can be found.

From Tryon...

So in fact Behaviourism is seeking salvation in associationist architectures, and even "a little bit hierarchical" architectures with the references to PDP multi-layer networks in particular.

I'm not sure why you insist that Behaviourism has no truck with associationism. It was there from the start with Thorndike and is back there again with neural nets.

Maybe you're using "associationism" in a way that differs from the traditional definition. Associationism, as used in psychology at least, refers to the idea that complex behaviors and entities like the mind, can be understood as being composed of a series of simple stimulus-response pairings. The idea is that the stimulus and response become "paired" or associated as a result of occurring at the same time - that is, it emphasised the importance of contiguity in forming these relations. This is the idea that Locke, Hume and Pavlov (and even John Watson, to an extent) supported.

The (radical) behaviorists rejected the associationist beliefs of the empiricists, and suggested that there was more to it than simple contiguity. It's difficult to summarise the entire field in simple terms, but essentially what they did was that they added two important factors: 1) the intent of the organism itself, and 2) the context. This is why behaviorists spend a lot of time highlighting the fact that saying that something is a "conditioned" stimulus is incorrect. It is now referred to as a "conditional" stimulus (as Pavlov originally intended), because the implication that a stimulus and response are "paired" due to occurring at the same time is blatantly false. Instead the supposed "conditioned" stimulus simply acts as a signal or a 'sign post' that informs the organism of what it about to occur. There is no implication that the organism is compelled or forced to behave in any particular way. (There's an interesting paper on that topic here, if you were interested).

So as for describing neural nets as an associationist concept, I can only assume that you mean that various networks are formed through 'associations' (in the laymen sense), rather than the process behind the generation of neural nets is itself being associationist.

 

[bohm2]

Here's a review piece from Stanford on behaviourism that is relevant to this thread I think:

It should also be noted that Skinner's derisive attitude towards explanatory references to mental innerness stems, in part, not just from fears of explanatory regression but from his conviction that if the language of psychology is permitted to refer to internal processing, this goes some way towards permitting talk of immaterial mental substances, agents endowed with contra-causal free will, and little persons (homunculi) within bodies. Each of these Skinner takes to be incompatible with a scientific worldview (see Skinner 1971; see also Day 1976)...Finally, it must be noted that Skinner's aversion to explanatory references to innerness is not an aversion to inner mental states or processes per se. He readily admits that they exist. Skinner countenances talk of inner events provided that they are treated in the same manner as public or overt responses. An adequate science of behavior, he claims, must describe events taking place within the skin of the organism as part of behavior itself (see Skinner 1976). “So far as I am concerned,” he wrote in 1984 in a special issue of Behavioral and Brain Sciences devoted to his work, “whatever happens when we inspect a public stimulus is in every respect similar to what happens when we introspect a private one” (Skinner 1984b, p. 575; compare Graham 1984, pp. 558–9).

http://www.seop.leeds.ac.uk/archives/fall2011/entries/behaviorism/

This is quite different from Chomsky who wants to treat these mental objects as "real" as any other aspects of the world studied by science. In fact, to not pay focus on these (as Skinner's more behaviourist/instrumental approach) leads to little insight and is not what science is all about (or so he would argue, I think):

I will be using the terms "mind" and "mental" here with no metaphysical import. Thus I understand "mental" to be on a par with "chemical", "optical", or "electrical". Certain phenomena, events, processes and states are informally called "chemical" etc., but no metaphysical divide is suggested thereby. The terms are used to select certain aspects of the world as a focus of inquiry. We do not seek to determine the true criterion of the chemical, or the mark of the electrical, or the boundaries of the optical. I will use "mental" the same way, with something like ordinary coverage, but no deeper implications. By "mind" I just mean the mental aspects of the world, with no more interest in sharpening the boundaries or finding a criterion than in other cases.

There is one final issue that deserves a word of comment. I have been using mentalistic terminology quite freely, but entirely without prejudice as to the question of what may be the physical realisation of the abstract mechanisms postulated to account for the phenomena of behaviour or the acquisition of knowledge. We are not constrained, as was Descartes, to postulate a second substance when we deal with phenomena that are not expressible in terms of matter in motion, in his sense. Nor is there much point in pursuing the question of psychophysical parallelism, in this connection. It is an interesting question whether the functioning and evolution of human mentality can be accommodated within the framework of physical explanation, as presently conceived, or whether there are new principles, now unknown, that must be invoked, perhaps principles that emerge only at higher levels of organisation than can now be submitted to physical investigation. We can, however, be fairly sure that there will be a physical explanation for the phenomena in question, if they can be explained at all, for an uninteresting terminological reason, namely that the concept of “physical explanation” will no doubt be extended to incorporate whatever is discovered in this domain, exactly as it was extended to accommodate gravitational and electromagnetic force, massless particles, and numerous other entities and processes that would have offended the common sense of earlier generations. But it seems clear that this issue need not delay the study of the topics that are now open to investigation, and it seems futile to speculate about matters so remote from present understanding. (Langauge and mind, 1968)

Chomsky, has in fact, argued, that "behavioral sciences" suggests a fundamental confusion between evidence and subject matter. Psychology, for example, he claims is the science of mind; to call psychology a behavioral science is like calling physics a science of meter readings. One uses human behavior as evidence for the laws of the operation of the mind, but to suppose that the laws must be laws of behavior is to suppose that the evidence must be the subject matter. (Searle, 1972)