Summary: language "locations" in the bilingual brain

[George Elgin, Suzette Haden Elgin]

        Recently I posted a request on this list (and several others) for
reactions to the 7/10/97 article by Kim and Hirsch in *Nature* reporting
functional MRI research showing a difference between bilinguals acquiring
both languages as infants or young children, and bilinguals acquiring their
second language as adults. A Memorial Sloan-Kettering news release had the
following quotation attributed to Dr. Joy Hirsch: "A second language
acquired during the teenage years, which is late in developmental life, is
represented in the brain in a separate location from the native language.
But when both languages are learned at the same time early in life, they
are represented in areas that have a considerable amount of overlap."
("Bilinguals Devote Distinct Areas of Brain to Native and Second
Languages," MSKCC: Press Releases, online.) The NY Times for 7/15/97 ran a
story titled "When an Adult Adds a Language, It's One Brain, Two Systems";
the Wall Street Journal's story on 7/10/97 had two headlines -- "How
Language Is Stored in Brain Depends on Age" and "Where Languages Are Stored
in Brain Depends on Your Age." I immediately began receiving phone calls
and mail with questions about this research. I am not a neuroscientist and
did not feel competent to answer the questions with even minimal accuracy;
I am very unwilling to add to the usual confusion created by such media
reports.  I wrote the authors requesting clarification, and got no
response; I did an online search and found nothing that I felt I could
trust; I then asked for help from the lists. (Depending on the precise
wording, a search on "language locations in the brain in
bilinguals/multilinguals" on Neuroscience Web Search gets roughly 3000
hits, through which I have been doggedly working my way.)  My sincere
thanks to all who responded.

The majority of responses were from individuals expressing interest and
asking that I share whatever information came my way. A number of responses
can be summarized as "This research is nothing new, nor is it especially
significant." (Medical professionals responding, who rely on information of
this kind in order to do neurosurgery without catastrophic effects on
language capacity in patients, disagreed with that judgment.)  A number of
bilingual or monolingual responders wrote with accounts of personal
experiences that were extremely useful and interesting.  Excerpts from
responses that strike me as of general interest follow; they are taken from
very lengthy postings and should be understood *only* as excerpts.

        From Joe Hilferty: "Kim et al. ...just investigated neural
activation in Wernicke's and Broca's areas, because the two sites are well
known from the long tradition of aphasia studies. Their findings do not
mean, however, that language is only processed in these areas of the
brain." References suggested: ** Bates 1994, "Modularity, Domain
Specificity and the Development of Language," Discussions in Neuroscience
10:1-2;136-149. **Maratsos and Matheny 1994, "Language Specificity and
Elasticity," Annual Review of Psychology 45: 487-516.

        From Brian MacWhinney: "I have argued that early bilinguals project
the input linguistic data to a single space, because that space is not yet
saturated by weights on synaptic connections and the two systems can be
learned together in a computationally reasonable sense without worrying
about catastrophic interference. In adult L2 learning, the optimal area for
an ability has already been occupied and new learning must either use the
old territory in a new way or else coopt adjacent 'new' territory. ... A
lot of what is at issue here is exactly what the role of Broca's area in
language processing might be. ... I would like to think of Broca's as
controlling high level sequential planning for language and related
abilities."


 From Liz Bates: "I believe that the article in Nature on differential
localization
for Language 1 and Language 2 represents the kind of gross
over-interpretation of neural imaging data that has become all
too common in the last few years.  Let's be clear about exactly
what this paper (like many others before it -- this is not the
first paper on neural imaging in bilinguals) has really shown:
(1) patterns of activation associated with covert speech in L1 and
L2 are largely very similar, but (2) there are some reliable differences
in the center and extent of activation in the Broca region.  None of
this has anything necessarily to do with STORAGE of L1 vs. L2! ...

Aside from this serious problem of confusing patterns of slightly
non-overlapping activity with separate modules, separate mechanisms,
separate storage of knowledge, there are other problems here as well.
For example, the assumption that Broca's area mediates grammar while
Wernicke's area mediates semantics is HIGHLY controversial, and
in my view, probably dead wrong.  Both Broca's and Wernicke's aphasics
have severe grammatical problems, but they take a somewhat different
form (e.g. omission in the former case, substitution in the latter).
And those studies that have tried to find a 'grammar area' through
neural imaging of normals have generally found EITHER that grammar
and lexical semantics activate the same areas, OR they have found
differences that vary markedly from study to study, and even from
one individual patient to another -- consistent with a "task demand"
interpretation of the data.  Another problem lies in the assumption
that any area which mediates language is a "language area."  In fact,
a number of recent neural imaging studies have shown that EVERY SINGLE
PIECE of the Broca's area complex (and, by the way, even the boundaries
of Broca's area of controversial, varying from study to study) is
activated by one or more covert motor tasks involving non-linguistic
motor activities, of the hands or tongue or both.  So it is possible
that ALL we are seeing in this study in Nature is an effect in which
bilinguals set their mouths a little differently (covertly, of course)
while speaking their second language -- possibly reflecting greater
difficulty in L2 in this case."

>>From Lise Menn: "It's not an unreasonable result given earlier indications
that late bilinguals had a more bilateral representation of language than
early bilinguals (using one-hand motor tasks as indicators of hemispheric
involvement), and given Damasio's claims that meaning of words with
concrete referents is represented in a way that is linked to our sensory
expereince of those referents (hence, language independent).

Additional references that were suggested to me by a number of list members:

Hull, Philip, 1990.  UC Berkeley Psych. Dept. dissertation . "Bilingualism:
Two languages, two personalities."
Bain, Bruce, 1996. Pathways to the peak of Mount Piaget and Vygotsky:
Speaking and cognizing monolingually and bilingually. Rome: Bulzoni
Editore.
Mollica, Anthony and Marcel Danesi, 1995. "The Foray into the
Neurosciences: Have We Learned Anything Useful Yet?" Mosaic 2:4;12-20.
Danesi, Marcel, 1994. "The Neuroscientific Perspective in Second Language
Acquisition Research: A Critical Synopsis. International Review of Applied
Linguistics 32:3;201-228.

In addition to my desire not to answer questions about this matter stupidly
and/or ignorantly, I was interested in this research because I have been
investigating the perception in many bilinguals/multilinguals that they are
"a different person" when they use different languages. (I won't bore you
with an account of this; I mention it just to provide context.) I became
interested in this when Diana Cook sent me "The Bilingual Self: Duet in Two
Voices," by RoseMarie Perez Foster, "Cultural and Conceptual Dissonance in
Theoretical Practice," by Carla Massey (a response to the previous item),
and "The Bilingual Self--Thoughts from a Scientific Positivist or Pragmatic
Psychoanalyst, by Perez Foster (a reply to Massey). All are from
Psychoanalysitc Dialogues 6:1, 1996, pp. 99 ff. (Sample from Perez Foster's
article, page 101: "Thus the bilingual person presents a packaging puzzle,
as it were, in which two language-bounded experiential systems are housed
in the confines of a single mind." And from page 100, "Bilinguals may
possess different experiences of the self, which are organized around their
respective languages.") It would have been extremely interesting if the
individuals in my database who report this "different person" perception
and those who do not correlated in some fashion with the "different storage
locations in the brain" research; they do not.

I found especially helpful, and thorough, and informative, the article
"Brain evolution and neurolinguistic preconditions," by Wendy K. Wilkins
and Jennie Wakefield,  in Behavioral and Brain Sciences (1995) 18:161-226,
together with the many pages of commentary on their target article in the
same and subsequent issues and their responses to those comments.   Wilkins
and Wakefield (page 170) propose that Broca's area is "a processing module
whose inherent specialization is the hierarchical structuring of
information in a format consistent with a temporally ordered linear
sequence reflective of that structure."

Suzette Haden Elgin
ocls at ipa.net