Native language acquisition

[Rudy Troike]

Bill Poser is right -- it is particularly sad that so much effort is
expended on the study of the acquisition of well-known and accessible
languages, while the potential for our understanding of the acquisition
of typologically distinct and even unique languages is allowed to slip
away as they cease to be learned by children. Bill's reference to the
one study of Navajo L1 acquisition is probably that of Muriel Saville-
Troike, who found, in agreement with native-speaker intuition that
Navajo is a verb-centered language, that children somehow extracted
verb stems from the prefixed forms they commonly heard and used these
before nouns. Older children in preschool who were asked to describe
pictures (which English and Spanish speaking children had described in
terms of names for the objects represented) by constructing action/event
scenarios using verbs (e.g. a picture of a boy and a wagon: the boy is
pulling the wagon).

   Contra the studies that Bill references on the convergence of babbling
with the phonological/phonotactic structure of the ambient language(s),
our twins, now 21 months old, have consistently produced very distinct
"babbling" (as we adults call it, because we don't understand it -- it
is presumably meaningful to them), neither sounding like either English
or Spanish, both of which they are regularly exposed to. One of their
earliest common "pre-words" was [nga] (velar nasal + vowel), and this
week they began pronouncing their word for triangle, "ga", as [gah],
sounding very much like the Navajo word for "rabbit", although neither
English nor Spanish (at least not the variety they are exposed to) uses
syllable-final [h].

   It would be particularly valuable to see how children might acquire
Cheyenne, since computationally the possible combinations of sets of
interrelated suffixes is a huge number, which was successfully mastered
by many generations of children in the past. This would be one of the
most interesting potential studies for child language acquisition, given
the particular morphological structure of the language, and could resolve
some major theoretical issues, if it could be engineered somehow.

   Some years ago one of my students discovered a girl from a Spanish-
speaking home in a first-grade classroom who had been labeled a "learning
problem", since for six weeks she had wandered around the classroom
paying no attention to the teacher, who spoke entirely in English and
knew no Spanish. When my student (Caroline Willard) spoke to the girl
in Spanish, she was perfectly fluent in her responses. Further inquiry
revealed that when her parents went to work, not having money for a
baby-sitter, they left the TV on tuned to a local channel which was
entirely in English. The girl had apparently learned to ignore the English
she heard on the TV as "background ambient noise", and so continued to
ignore it when she heard the teacher producing English with other students.
Only after Caroline showed her the equivalence between Spanish and English
terms did she begin to pay attention to the teacher's language, and within
six weeks was ahead of all the children in the class. So it would appear
that introducing a language as "background ambient noise" would be the
best way to insure that it would NOT be learned.

    Rudy Troike