ksagnostic

As requested, I am starting this as a new thread. Rufus responded to this statement by me:

Before I go further, let me elaborate on the above a little bit. I am quite critical of the set of positions I refer to as strong innatism, as articulated by people such as Noam Chomsky, Steven Pinker, and from my own field, John Muma. Strong innatism assumes that the ability of humans to create/acquire language requires the explanation that humans, and only humans, have specific neurological structures hard wired for learning grammar. Specifically, strong innatism implies that humans have a language acquisition device (or language module) in their heads that allows them to learn language without it being explicitly taught. As evidence, strong innatists point out the complexities in languages at the phonological, semantic, and syntactic levels, and the ability of the speaker of any language to potentially generate an infinite variety of messages through manipulations of finite vocabularies and grammatical rules. Such a complex system, the strong innatist will argue, requires a unique and detailed biological predisposition to learn. That position, frankly, reminds me of the irreducible complexity argument used by the ID movement.

I will not be arguing against the proposition that humans have some unique physiological and even specific (but usually emergent) neurological features which assist them in learning language (see discussion of the FOXP2 gene below). For example, humans have a vocal tract that can be altered to rapidly produce a wide variety of differentiated sounds. Other mammals, including our closest great ape relatives, do not have such a tract. The sheer size of human brains, plus the extended length of neurological development, I am sure are not incidental to the ability of humans to rapidly coordinate the movements required to produce speech, from respiration to vocalization to the alteration of the relatively long supra-laryngeal tube that is the human vocal tract. Nor are these features incidental to the ability of humans who use manual sign languages to rapidly shape their hands into different shapes in different locations, and to pair these movements with facial expressions and mannerisms to enrich the message. However, I do not think that a specific hard wired language acquisition device is needed for a child to learn language. Rather that the typically developing human brain, a typical primate brain except for its increased volume and its greater length of maturation time, will learn language when the people around it use language to communicate with it and the people around it, and to describe the environment around it.

1. Not really. While it is true that language functions do, assuming unremarkable development, end up in the left hemisphere of the brain, that doesn’t mean that they are hard wired there. Children who have had left hemispherectomies still develop full language skills commensurate with their overall ability. Similarly, focal damage to the “language areas” of the brain does not have the devastating or long lasting effect on the language skills of children that it does on the language skills of adults. Also, the areas of the brain associated with language, such as Broca’s and Wernicke’s area, are found in chimpanzees and bonobos as well as humans, indicating that the development of language was probably not instrumental in the development of those structures. All told, I believe that the evidence is that language areas develop into the brain, with general anatomical commonalities in the brains of humans resulting in the language areas ending up in same areas of the same hemisphere, assuming unremarkable development. Language areas are associated with the temporal lobe in the vicinity of the auditory cortex and associative areas, and the frontal and prefrontal motor areas associated with fine oral motor and hand movements. Thus, the neurons of these areas of the brain have a selective advantage to become involved in language. In sum, the language organs appear to be structures that emerge with development, rather than pre-existing, structures.
   
   
   Quote:

   Children will naturally develop a language in a social setting even if they have no language of their own. (See Nicaraguan sign language.)[

   In the case of Nicaraguan sign language, the children who were deaf were first placed together into a single setting in the early 1980’s. Prior to that time, there were no educational programs for children who were deaf in Nicaragua (indeed, it is my impression that there was a dearth of educational programs for hearing children under the Samoza government). The early attempts were to educate these children in an oral program. According to accounts, these methods were by and large failures, at least in terms of teaching children oral language and through oral language. However, there are still several things to keep in mind:
   
   1) Teachers/caregivers and children still needed to communicate with one another, and shared gestural communication would almost certainly have figured in such interactions.
   2) This was a group of children who had the same set of strengths and limitations when it came to communicating, and therefore strategies that worked for one worked for other situations.
   3) The impression given by some strong innatist is that this was a situation where a group of children who were isolated communicatively developed their own language. However, even if the approach used was initially a failed oral approach, this was still an educational environment, where a lot of attempted teaching and enrichment would have been going on. This would have been an environment conducive to language development, and even if the instructional language used was not accessible to the children, there was still an exposure to language use, and an environment that was designed (as best as the instructors there could design it) to elicit verbal behavior. In this sense, the children were still in a relatively language enriched environment.
   4) Despite descriptions that the development of Nicaraguan sign language was “almost instantaneous”, if you read accounts carefully you will see that it was not. It took about a decade, decade and a half, with subsequent groups of children building on what had been done before. While rapid, it was still a language that developed. This is fully compatible with the developmental/emergent theoretical viewpoint that I tend to think that the evidence favors.
   
   
   Quote:

   FOXP2, a gene for language that when "broken" produces both grammar and motor dysfunction.

As you yourself say, both grammar and motor dysfunction. Specifically, the FOXP2 gene appears to support the neuronal architecture for the development of the fine motor skills necessary for the development of articulation, a key component skill in the development of language, and therefore a “broken gene” would have an effect on the development of language. Difficulties in sequencing the fine motor movements necessary for the clear production of speech, for example, would also be likely to have an effect on other sequential skills, including syntax/grammar. Incidentally, in my professional (as well as personal) experience, there is almost no such thing as a “pure” language disorder. Even those children diagnosed with “specific language disorders”, for example, end up having more going on, such as attention deficit issues, or auditory processing disorders such as deficits in fine discrimination in timing and separation, or short auditory memory spans. In other words, again, issues involving component skills that underlie language development.

I hate to pull an Ed, but see above with regards to my comments on the invention of a sign language.

Come now. I don’t need a language acquisition device to make inferences from an association between pictures and heretofore unfamiliar sequences of letters. Plus, I would fully agree we have a bias to acquire human communication systems, including language. But that is by no means the same thing as saying that we have a hard wired language module designed for language acquisition.

Here is the list of papers on the homepage of Elizabeth Bates, who argues in favor of an emergent, rather than a strong innatist, approach. The emergent approach is closer to my viewpoint on the issue of language acquisition, and the one I think that will eventually be validated. I would particularly recommend the paper, Innateness and Emergentism as an introduction on the subject.

http://crl.ucsd.edu/~bates/papers.html

beyelzu

i found your post to be very interesting and a strong argument. i wish that i could actually offer constructive criticism. but i do not know enough about the argument to offer any. i do think however that strong innatism does smack of id and god. and that your post offers a strong counter argument for cultural evolution of langauge.

Dr Rick

Rick

RufusAtticus

There are at least two problems with that.

1. It doesn’t explain the language window of early childhood. Early childhood acquisition is different than later acquisition of language. Children with little effort do pick up language. However, the ease is lost by the time they reach puberty. Later acquisition uses our thinking abilities to solve communication problems. Early acquisition does not. Furthermore, it doesn’t explain why a baby’s innate ability to imitate is so important in acquiring language.

2. It doesn’t explain for know instances of novel language development with no input. Let’s start with the typical transition of a pidgin to a creole. A pidgin is a simple language that exists in contact zones when people of different languages need to communicate with one another, say for business or trade. The remarkable thing is that children brought up in such contact zones not only learn the pidgin, but through using it with other children transform it into a fully featured language. This is where creoles come from.

An extreme example of this is Idioma de Signos Nicaragüense. Children, who had no language ability except rudimentary “home signs” because they’d been isolated from other deaf children, were placed together in deaf schools. Through interacting with one another they turned their home signs into a crude, pidgin language. The next wave of students who were younger entered this pidgin environment and developed it into a full fledged grammatical and structured language. They didn’t acquire this language from their teachers or anyone else; they developed it on their own as they socialized and communicated. If there was no instinct for language and its existence was learned or cultural, then this ISN wouldn’t exist. If language was just a way to describe the environment then the home signers would have had fully developed languages.

The existence of similar structures in our relatives in no way casts doubt on the innateness of language, nor any of the other remarks about brain morphology. Take the above argument apply it to the visual cortex and you’d be trying to prove that sight wasn’t an innate ability but rather a result of us using our complex brains to decide to see.

It is suspected that gestural communication is a preadaptation to (spoken) language. So it is not surprising to see language organs tied closely to such abilities.

So in your view, where do the communication biases lie if not in a language acquisition device? Here is the issue described by Hauser et al. (2002) Science 298(1569-1579).

Jackalope

I'll bite on this one. Rufus, there's also a fairly small development window in the visual system. Children who do not have corrective surgery for strabismus (cross-eye) before the age of 2 usually do not develop stereoscopic vision. There are some (like myself) who, when they have the corrective surgery at a much later date, develop at least some stereoscopic vision, but do not have full stereopsis. I will posit that in both this case and the case of language aquisition, the real window has to do with the elasticity of the brain at a very early age. In other words the age of the child has a bearing on brain structures that are still developing. Once the window is past, the likelihood of developing a given ability to its fullest is lowered.

The small window for language aquisition seems remarkable when viewed in isolation. When viewed in the context of development going on throughout the whole brain, it starts to look less amazing.

--Lee
(look, it's amazing that the brain develops at all.)

RufusAtticus

Jackalope,

Exactly. Just like sight, languange is an innate ability.

ksagnostic

Sure it does. As Jackalope points out, there are critical periods, or windows, for acquisition of several human abilities, and all of them seem tied to brain development. In other words, language acquisition or development requires a developing brain. In fact, it is well known that species specific communication behaviors across the board are best acquired during a critical period of development. For example, Kim Bard down at the Yerkies Primate Center in Georgia has a project that has laboratory caregivers attempting to imitate as best as possible the communication gestures and mannerisms of adult chimpanzees, because lab raised orphans were turning out to be socially and communicatively disabled in comparison to conspecific raised peers. It is also no coincidence, IMO, that Kanzi the bonobo was able to acquire a considerable receptive understanding of English, as presented in both English and lexigrams, during simple exposure to linguistic input during failed attempts to teach his adoptive mother to use lexigrams. Differences between typically developing humans and Kanzi and other cross fostered or semi-cross fostered great apes I believe is largely explainable by the fact that maturation of the brain takes a lot longer for humans than it does for apes. However, when both human and great ape brains are in their critical periods of development, both are acquiring language if they are growing up in an accessible language rich environment. Look again at the quote you took from my post above, particularly the part I italicized.

In fact, I am suspecting that you really are not understanding the emergent/developmental viewpoint I am arguing for here. If you look at the quote that you reply to above, I specifically tie the emergence of language development into the maturation of the brain. An emergent viewpoint is that language emerges with the development of the brain. The strong innatist position, on the other hand, argues that there is a pre-existing brain structure (or structures) designated for language acquisition. If anything, strong innatism would infer to me that there is a dedicated neurological structure for language acquisition that could not be co-opted for other purposes, and that other parts of a developing brain could not be co-opted for language development. As I explained in my first post, this is not the case. A strong innatist tends to argue that a language acquisition device is tied into development, but then becomes locked in as maturation occurs. A developmentalist/emergentist will argue that language skills emerge with a developing brain, therefore the raw matarials necessary for easy language development include a growing, developing brain. Neurological and genetic research indicates that the difference between humans and great apes is that brain growth and development, and the activity of genes that mediate brain growth, takes place and operate for a much longer relative period in humans. On the other hand, finding differences in neurologial structures between humans and great apes, other than of a quantitative nature, have met with failure. Figuring out which viewpoint the evidence fits, to me, is a (ahem) no brainer.

Again, none of the above contradicts an emergent approach. You had children who had similar communicative strengths and weaknesses placed together in an environment that was, relative to their home environents, relatively enriched. You also had people trying to teach these children in this environent, and even to create situations which favored linguistic communication. However, an oral approach was not effective for these children, and in particularly it was not an effective approach for the children to communicate with one another. However, you still had an environment that favored linguistic communication, and the environment was still relatively enriched. It was not, as I said in my first post, input free.

And again, the transition from a pidgeon to a creole is consistent with the emergent view of language development. The first generation of children took idiosyncratic gestures and made them common gestures, and also began to combine these gestures with other gestures. There were almost certainly common gestures between instructors and children as well (even those who attempt to insist on an oral approach soon learn the importance of gestures--how do you think the instructors signaled the children to go to lunch, or to class, or to bed?). However, the first generation of kids did not develop a full blown language from this input, they developed a pidgeon. The next generation of children, however, were presented with a pidgeon to work with, and developed a creole/full blown developing sign language, the difference being the raw material these kids had to work with with their developing brains. Nicaraguan sign language, in other words, emerged, and emerged in an environment where every effort was made to make it an enriched one. One thing, incidentally, that I really don't think is clear is whether the instructional conditions were the same throughout this process. At what point, in other words, were instructors starting to encourage rather than ignore the children when they were communicating in their incipiant sign language?



And actually as an aside, there is some rather goulish evidence, from some animal experiments that the visual cortex can be moved if the optic nerve fibers are surgicially rerouted to another part of the brain. However, it is pretty clear you are missing my point here. Remarks about brain morphology are pretty fundamental when it comes to the innatist view articulated by the likes of Chomsky and Pinker. You yourself, incidentally, brought up language organs in the brain in your post on the Evolutionary Psychology thread, so my responses with regards to brain morphology are appropriate. The emergent viewpoint is not contradicted by the fact that humans seems to have pre-adaptations that favor the development of language (see discussion of the vocal tract, larger brains, and extended neurologial development). In fact, if one argues that language is innate to human beings because there are pre-adaptations that favor language development, I would go along with that. However, that is not what Chomsky, Pinker, and their theoretical cohorts are arguing. They claim that humans have a specific language learning device/acquisition device/module/organ/predisposition designed for the acquisition of language. Such a viewpoint does make predictions about specific neurological structures dedicated to language development. Such predictions have not withstood scrutiny.

Pardon me, but what????? In one paragraph you state "The existence of similar structures in our relatives in no way casts doubt on the innateness of language, nor any of the other remarks about brain morphology." (italics added by me). Then you say in the very next paragraph "So it is not surprising to see language organs tied closely to such abilities." (again, italics added by me). Where the frick would language organs be, except in the brain???? You say that remarks about brain morphology casts no doubt on the innateness of language. Well, if by innateness you mean that language acquisition seems to be a more or less unique trait of humans, and that humans have some pre-adaptations that favor the development of language, you get no argument from me or any other person who favors the emergent/developmental view of language acquisition. The only argument you might get is from a strict verbal behaviorist. If, however, by the innateness of language you mean specific language organs dedicated to acquiring language, then remarks about brain morphology cast quite a bit of doubt upon the innateness of language.



I really think I have answered this question. Humans are biased to communicate with other humans, and humans have an extended window of development and a greater amount of raw material in the brain to work with for language to emerge. Non-human animals have, given sufficient enrichment from humans, acquired some language (The differences between their "language" and full blown human language almost certainly comes from a much shorter critical period, and also not incidentally from the nature of their expressive systems. Kanzi, for example, has substantial and even arguably open ended receptive language ability with regards to spoken English. His expressive language is much more limited. That situation, incidentally, is very similar to that of many of the people I work or have worked with over the years.).


I am not going to take the time to respond to this in detail. Just a couple of quick points.

1) "The astronomical variety of sentences any natural language user can produce and understand has an important implication for language acquisition, long a core issue in developmental psychology. A child is exposed to only a small proportion of the possible sentences in its language, thus limiting its database for constructing a more general version of that language in its own mind/brain. This point has logical implications for any system that attempts to acquire a natural language on the basis of limited data." I always tend to chuckle when I read claims like this. The rules of languages, from phonological to semantic to syntactic, are limited. It is what you can do with them that results in the astronomical variety of sentences. The exposure to the rules of languages is arguably, almost certainly, complete.

2) "No known 'general learning mechanism' can acquire a natural language solely on the basis of positive or negative evidence, and the prospects for finding any such domain-independent device seem rather dim. The difficulty of this problem leads to the hypothesis that whatever system is responsible must be biased or constrained in certain ways. Such constraints have historically been termed “innate dispositions,” with those underlying language referred to as “universal grammar.” Although these particular terms have been forcibly rejected by many researchers, and the nature of the particular constraints on human (or animal) learning mechanisms is currently unresolved, the existence of some such constraints cannot be seriously doubted."

Several comments come to mind here. First, a general comment. The argument that "general learning mechanisms" cannot account for language development strikes me as an argument from incredulity. Secondly, what exactly is a "general learning mechanism" when it comes to child development anyway? Children learn lots of rule based systems when brain development is most active, not only regarding language acquisition but also social rules acquisition and knowledge acquisition. Psycholinguist Keith Nelson, for example, has argued that rule and rare event learning devices would account for language acquisition as well as a specific language acquisition device. Finally, when the authors say "Although these particular terms have been forcibly rejected by many researchers, and the nature of the particular constraints on human (or animal) learning mechanisms is currently unresolved, the existence of some such constraints cannot be seriously doubted", that is in fact a reference to the issues I am discussing. The differences between strong innatists, developmentalists/emergentists, and for that matter interactionists, are discussions about the nature of learning mechanisms, including the constraints.

I really encourage you to use the link I provided, and read the article I recommended. I think it would help you to make sense of what I am saying, and not saying.

Regards.

ksagnostic

Posted by Jackalope

Exactly Lee!

Oolon Colluphid

Okay, now I’m really curious. Is, say, human bipedal walking innate? Kids seem to have to learn how to do it.... and there’s a window in their development when this happens, and doubtless the developing brain wires itself up accordingly.

How about, say, hunting in carnivores? Young polar bears have to learn how to thump themselves down so as to break through to a seal pup below the snow surface... young cheetahs have to practice hunting skills. If something takes practice, it must be learned, no? Or what about the sexual behaviour of bonobos? All that social shagging must be purely cultural, nothing innate in it.

So I take it that walking, hunting and bonobo sex are not innate, not part of the ‘standard equipment’ that characterises the species?

DT

Oolon Colluphid

... come to that, what of bower birds? “The construction of high quality bowers appears to require experience, learning and practice,” apparently. And it is related to brain architecture. See Madden, Sex, bowers and brains (Proc Royal Society of London B, 268, 833-838). Therefore, presumably, there’s nothing innate about it. They just do it because, well… they saw their daddies doing it, presumably.

Can behaviours be innate at all? And if so, why might human language -- the tendency to learn one, to use it as our human means of communication -- not be just as much part of the standard human set-up as a bower bird's bower? Is it not just what you do when you've got that sort of brain -- shaped by evolution cos it works?

DT