Re: Instincts and Bioprograms
Ronald Kephart (rkephart@OSPREY.UNF.EDU)
Tue, 20 Aug 1996 09:51:44 -0400
In message <19960820003222.AAA10966@LOCALNAME> Robert Snower writes:
> What are the criteria for deciding what is genetic and what is acquired--in
> language, that is?
First, all would agree that there are clearly aspects of language that are
cultural. For example, the word "dog" is not innate; the lexicon of any given
language is a set of arbitrary symbols (dog, inu, sobaka, perro, chien, etc.)
which must be acquired. (Even these words, however, illustrate a language
universal, which is that if a language has a word for a plant or animal it will
tend to correspond to the generic level in taxonomy; see George Lakoff's Women,
Fire, and Dangerous Things.)
A major piece of evidence for innateness is things which seem to be "known"
prior to experience. For example, children at about six months can perceive and
categorize features such as vowel height and voicing of stop consonants (refs
forthcoming if you want). This is before they can talk of course; indeed, at
six months they can't produce the acoustic equivalent of adult vowels.
Chomsky's classic example involves sentences like "The cat that is black is on
the couch". English speaking children correctly form the yes-no question
"Is the cat that is black t on the couch?"
by moving the "is" which is in the predicate (The "t" marks the place which the
main verb was moved from). They never produce something like
*"Is the cat that t black is on the couch?"
which involves incorrectly moving a verb out of a clause embedded in the
subject. They do not require instruction in this, or even prior experience of
such question-formation. The conclusion is that phrase structure is built-in, a
part of universal grammar.
Less obvious, maybe, are things like the distinction between "null-subject" and
"non null-subject" languages. Languages either insist on an overt subject
(English, French), or allow for both overt and empty subjects (Spanish,
Italian). Chomsky calls this a "principle" with two "parameters." The
principle is innate, a built-in feature of all human languages. During
acquisition, the parameter is "set" for one of the two possibilities.
Bickerton's thesis is a bit different, and probably even more controversial
(Roots of Language, Language and Species). He makes a distinction between what
is innate, as a part of his language bioprogram, and what is "cultural" i.e.
learned in a social context. He argues further that shared features of creole
languages represent the bioprogram kicking in in the absence of cultural input.
This is because these languages were first nativized by children of people who
were generally able to communicate with each other only thru pidgin languages,
such as slaves on Caribbean sugar plantations. Pidgin languages are by
definition stripped of almost all functional categories; the children of these
slaves took the pidgins (which had European lexicon) and added to them the
grammatical forms and functions necessary for fully human language.
Features common to creole language around the world include:
(1) marking of definite vs. indefinite on nouns
(2) marking of punctual vs. nonpunctual (progressive) aspect on verbs
(3) marking of tense, aspect, etc. by preverbal particles
Innate features of the bioprogram such as these can be over-ridden by "culture".
Russian and Aymara, for example, do not require marking for definite vs.
indefinite on nouns; English does not require aspectual marking on verbs.
Bickeron (especially in Language and Species) cites language acquisition studies
which suggest that children appear to assume that these bioprogrammed features
exist in their language until they get evidence to the contrary.
Bickerton's thesis inspires heated, but very interesting, debate among
linguists, especially those who focus on creole languages. I don't have time
right now to go into that in detail, however.
Let me just end, for now, by asking whether this at least points in the
direction of your inquiry.