Re: Natural Language and Natural Selection

From: Stevan Harnad (harnad@cogsci.soton.ac.uk)
Date: Sun May 26 1996 - 18:38:02 BST


> From: "Alexandra Bilak" <AB495@psy.soton.ac.uk>
> Date: Fri, 19 Apr 1996 11:45:33 GMT
>
> IS LANGUAGE A BIOLOGICAL CAPACITY OR A CULTURAL
> INVENTION ?
>
> The study of the nature and origins of language seems to be the
> object of a lot of speculation and inquiry. Indeed, the origins of
> man's ability to use and comprehend language are slightly ambiguous,
> and people's views on it have often differed. Language can be seen as
> a basis, a universal, biological capacity common to every individual,
> but can also be considered as a cultural element that would have
> evolved in relation to different external, environmental factors.

What is language? Is it one thing, or a bundle of things? Is what's true
of one true of all?

> First of all, it is interesting to examine the aspect of language
> that links it to a biological predisposition of man to use and
> understand it. There are some clear genetic factors essential to the
> understanding of the origins of language. Although a little
> simplistic, a first element to consider would be a human
> morphological predisposition to language. Indeed, the sounds of
> language seem closely related to the morphology of the vocal tract:
> certain characteristics of a human face seem to have a decisive
> influence upon speech sounds. Simple elements such as the shape of
> the mouth and lips can be related to the development of language in
> man, or at least seem to be prerequisite for speech articulation,
> and oral motivity in man. However, anatomical descriptions do not
> seem to provide an accurate explanation for the nature, or the
> ORIGIN of language itself, and does not lead to insight into man's
> capacity for language, ie the underlying complexities of the matter.
> Some much more interesting aspects of study exist and must be
> considered. It seems indeed more interesting to examine the genetic,
> or biological account for the emergence of language. The "nativist"
> view, therefore, strongly insists on a human specific capacity for
> language peculiar to each individual.

Paragraphs are too long!

> One of the most influencial
> nativist theorists is the linguist Noam Chomsky, who insists on the
> fact that man has mental structural capacities including an innate
> concept of human language.

Not "concept of language." UG.

> This concept appears to be genetically
> determined, and present in every human individual. He explains that
> a "normal" human child is thus predisposed to learn any language
> fairly easily, by combining a set of innate rules with " language
> data", that she or he hears. Language would therefore appear in man
> as an abstract system of rules that cannot be acquired by
> traditional learning principles.

But why not? Chomsky doesn't just SAY this; he has evidence and
reasons...

> An interesting example illustrating
> this predisposition to language is the creation of the "Creole
> language" among children of a mixed community of Japanese, Chinese,
> Korean, Portuguese, and Puerto Rican parents at the turn of the
> century. Immigrants of these various communities were "forced" to
> live together and thus communicate by some sort of hybrid (pidgin)
> language. A generation later, their children had invente (or come up
> with) a new language ("Creole"), being a highly developed linguistic
> system whose structure appeared to be similar among all the children
> of the community. This amazing fact seems to prove that, according
> to Bickerton (1983), "first-language acquisition is mediated by an
> innate device (...) the device provides the child with a single and
> fairly specific grammatical model". A iological basis of language
> would indeed seem to be more than just a possibility.

> Another
> important aspect of the nativist view emphasises on the fact that
> environmental (external) stimuli alone, such as learning, cannot
> account for language possession. This idea is summarised in the
> theory of the "Poverty of the Stimulus", that explains the ways in
> which language is born into an individual. Lightfoot (1982) explains
> that "there are no data available to the child that will suffice to
> establish some rule or principle". This implies the idea of an
> "innately-primed learning", that is, of an innate set of rules and
> information, of internal mechanisms suitable for the acquisition of
> language. The "poverty of the stimulus" phrase refers to the
> insufficient information obtained from OUTSIDE the individual, thus
> implying the idea of some sort of "compensating" internal,
> biological mechanism.

Good.

> This is where it becomes interesting to examine
> S. Pinker's view on the origins of language being determined by
> "natural selection". Darwin's theory of natural selection explains
> the existence of human organs in terms of "adaptive complexity";
> this means that every organ of the system has a specific function,
> and is adapted, arranged to perform a certain function, some sort of
> task. For example the eye (ie the visual organ) is adapted
> specifically for visual imagery (Function/task). For Darwin,
> therefore, it is through a process of natural selection that these
> organs have developed this specific function. It is this theory that
> Pinker relates to language. Indeed, he attempts to explain the
> origins of language as a natural process: for him, natural selection
> is a plausible explanation for the complex cognitive mechanisms
> underlying language. Language being a " complex design system",
> Pinker explains, it can only have occured through a process of
> natural selection.

But is "it" one thing, or several?

> An interesting analogy is made in Pinker's article
> Natural language and natural selection, extract of Behavioural and
> Brain Sciences (1990). He uses the image of a spandrel to describe

What is a spandrel? (Remember, kid-sib?)

> the process of language acquisition : the centerpiece is an innate,
> biological capacity for language and is necessary for the holding up
> of what lies around it, ie language acquisition. Finally, the most
> important argument in favour of a biological foundation of language
> is probably the universality of its structures, ie it definitely
> seems to be common to every individual. The linguist Chomsky for
> example insisted on the fact that there are similar linguistic
> traits all over the world.

Insisted? Or provided evidence?

> In order to understand this, it is
> interesting to consider the idea of a "Language Acquisition Device",
> which explains the transformation of a "corpus of speech", ie a set
> of utterances (some grammatical, some not) into a complex
> grammatical system. This is generally studied in relation to the
> acquisition of language in a child. For example, the corpus of
> speech will be utterances overheard by a child in a given
> environment.

Or spoken by the child.

> Upon receipt of this corpus, the Language Acquisition
> Device creates a grammatical system. The device therefore represents
> some kind of internal structure, that must be able to acquire any
> natural language. It is therefore universally applicable to the
> fundamental human capacity for language. The idea of an internal and
> universal capacity for language is closely associated to that of
> Universal Grammar (UG), whose study is clearly an attempt to specify
> the forms and features common to all human language. UG suggests a
> universal set of grammatical rules, and therefore also suggests that
> a child is predisposed to entertain certain hypotheses concerning
> the structure of language. The idea of UG clearly implies the idea
> of a biological foundation of language, and suggests that no
> individual will thus "start from scratch" his/her acquisition of
> language.

Need to relate this to poverty of the stimulus, otherwise sounds
arbitrary.

> A number of elements seem thus to support the claim that language is
> most of all a biological capacity. However, it is easy to notethat
> it is not this biological basis alone that makes up for the study of
> language. Indeed, it seems to have a context as well, some sort of
> external, CULTURAL aspect.
>
> Indeed, language is often considered a cultural invention, precisely
> because it evolves differently among peoples and cultures as a means
> of communication. Language often seems to have been invented by man
> simply in order to communicate with his peers. The capacity to use
> language is a capacity to use and comprehend the use of SIGNS, and
> it is the arbitrariness of the sign that implies that there is
> something cutural about language. The fact that for example the word
> "tree" is used to define the concept of a tree in the English

"Tree" does not define a concept (it doesn't define anything). tree
refers to an object: to a tree. A tree is not the same as a concept of a
tree. And the word "tree" is not the same as either of the other two.
The definition of "tree" is yet another thing. We have a concept of a
tree: A tree is a green wooden thing. A concept MIGHT be like a
definition, or it may be more like an image, or a filter for picking out
trees when we see them. It is important to begin to try sorting these
things out, as vague use fo words will only lead to vague
understanding.

> language, but that in French one will use the word "arbre" to the
> same concept, proves the arbitrariness of the sign, ie the word.
> This arbitrariness shows that language is culturally determined to
> some extent, and relates to the idea of "particular grammar" in
> opposition to UG. "Particular Grammar" studies the features of a
> particular language ( though always in relation to universal
> features of Grammar.)

Unclear what you mean.

> Another important aspect to consider with the
> cultural aspect of language is the external factors having an
> influence on the acquisition of language. Language may well be
> culturally determined, in that some aspects of a culture or a given
> environment may well act as stimuli to the development of it. The
> "nurture" view (as opposed to a "nativist" account of language)
> emphasises on an account of traditional learning principles being
> key-elements in the development of language. The theorist
> B.F.Skinner (1957) proposes a theory of reinforcement, in which he
> explains that it is certain elements in a child's environment (eg
> parents) that are capable of "reinforcing", that is, improving the
> child's early babbling sounds. The concept of imitation is here very
> important, since it is this activity that enables the child to
> "reinforce" its language abilities. It is therefore in this sense
> that language can be considered as culturally determined, because
> certain aspects of a culture tend to either provide a language with
> a particular grammar peculiar to that culture, or simply reinforce
> the process of acquisition.

But the story is different for UG and for vocabulary. You have to sort
these out.

> It would seem that the origins of language are a little complicated
> to undersand. Although it would appear obvious that language
> represents a cultural invention ( because of its communicative
> function and the fact that it actually symbolises man's will to
> communicate), language seems to be at the root biologically
> determined, proving man's predisposition to acquire and use a
> language. However, one must take into account both sides of the
> answer. Indeed, language seems definitely to be genetically
> determined and universal, but finds its evolution and development in
> external factors aswell. An internal structure would therefore be at
> the core of language acquisition, whereas external stimuli lie
> around this key-structure to make it more solid and complex. One can
> thus say that language is both a biological capacity and a cultural
> invention.

Kid-sib would not be too happy with that conclusion!



This archive was generated by hypermail 2b30 : Tue Feb 13 2001 - 16:24:16 GMT