We have talked about meaning in the world, categories in the world,
semantics, morphology, phonology, even syntax. We have tried to examine from
the child's output how the mechanisms of the brain function in order to
learn language. My impression is that we cannot isolate the child from its
environment, so my big unanswered question would be: How does the input the
child receives from the environment affect the child's vocabulary learning?
At the risk of sounding a bit cliche, I feel that the "biggest" issue
has to do with the reality of an innate language acquisition device and
its nature. This question seems to keep coming up over and over, and
surely influences all approaches to language development. It can, at
times, also be very devisive. Linda touched on this last week when she
mentioned that the nature of such a device is probably not one "black
box" which exists somewhere in the brain, but that it if it exists at
all, it is in parts all over the brain and each of these parts works
together to make up the innate mechanism. I think that more work directed
specifically at answering this question needs to be done.
The next issue, one that I find extremely interesting and important in
word acquisition research, is the role of the influence of culture on the
kinds of words that are learned. This was touched on by Fernald &
Morikawa. This seems to have a great effect on what children acquire
first, and until more data from varied cultures are brought in, I think
we are leaving out an important part of the puzzle.
Children, chimps and connectionist nets all acquire words. But our
models always assume that input is discrete, not a continuous stream. Yet
children and chimps not only acquire words, they acquire the idea of words:
the principle of segmenting that stream into repeatable pieces. So the Big
Question is one of segmentation and structure: how do sentences become
segmented into words, and words into phonemes, in such a way as to be
restorable into sentences again from sounds?
How can we posit a reality that can constrain our conceptual
organization,
and yet deny this reality a mind-independent ontology and structure?
I would like to know the exact neurological description of how words are
learned (both early and late words). Implicit in the answer would be why
there appears to be a sensitive period for language learning.
How do we reliably test what is in the language input to children to
determine what language development is? How can we collect such massive
amounts of data and how do we know what a child does with this language
input?
Why is everybody so enthusiastic about "meaning"?
I feel like meaning is quite abstract and unobservable;
questions about meaning are untractable; theories about
meaning are too sketchy. For example, the hyperspace for
nouns and adjectives posited in Gasser & Smith is just
arbitrary and there is no guarantee whether we are on the
right track or not.
What's "there" before language and how does it affect (and is in turn
affected by) language acquisition?
Just about everything we've seen in class have been interpretations and
explanations of observed phenomena. As long as the explanation is
consistent with what we've seen (and it agrees with our world view), we
accept it until new evidence contradicts its findings. Will we ever
actually "know" anything about the structure and function of language
development mechanisms? Are the workings of language and its development
things that can be "known"? Or are we faced with the reality of "merely"
speculating forever?
What is it when someone "knows" a word (what behavior, knowledge, or
representation best characterize "knowing" a word)? How does a child's
learning of a word approach that "knowledge" (gradually, the same or
differently for each person?)? I guess what I want is a theory of meaning.
What is the necessary and sufficient hardware (wetware)
required to do language? Do the neuroscientists and anatomists
have anything to offer? Should we take the brain seriously or
is it just one of many ways to do this task?
Is language special? If so, why do most people learn it so
effortlessly (or so it seems, anyway)? If not, why is translation so
difficult? (And why have linguists and language development people)?
If it were "proven" that word acquisition was innate/not innate, what
would that mean? In what way(s) would that help us? Do we in fact care?
Last updated: 29 April 1997
URL: http://www.indiana.edu/~gasser/L700/questions.html
Comments: gasser@cs.indiana.edu