one-parent one-language/Grammont's principle

[Fred Genesee]

We have been studying children who are learning English and French
simultaneously in the home in the Montreal area for many years.
Montreal is rather unique in that both English and French enjoy
majority-like socio-cultural status -- both languages are widely
valued and widely used and children have opportunities of hearing both
languages from many other children, adults, and in the media. As a result,
language usage
patterns in other communities that might reflect the differential status of
the languages under investigation is probably less of an issue here than
elsewhere, although, arguably, it is always an issue to some extent.

We have not observed any child who was confused as a result of parents
mixing their languages; in fact, one of the most sophisticated bilingual
children
we studied came from a family in which the parents tended to mix their
langauges quite a lot -- hardly evidence of confusion. Overall, we have
found that
French-English bilingual children in Montreal use their languages
differentially and appropriately with parents and with strangers from very
early in development --
from the two-word stage onward, and possibly earlier.

However, we have not systematically looked a bilingual children's language
development as a function of how much mixing there is in the input -- this
is what would be needed to establish if  confusion results form extensive
mixing in the input. This research is not straightforward to do since
identifying such families would be difficult because many parents do not
admit readily to mixing their langauges a lot with their children because
they believe that it is not recommended practice. In some communities, in
fact, such as can be found in New Brunswick, English and French are mixed
quite extensively -- on the one hand, this would be fertile ground for
studying this issue. However, one would need to document how extensive
adult mixing is and what form it takes and then use this as a basis for
studying children's mxiing. Children who grow up in such communities might
mix a lot -- not because they are confused, but because this is a
communicative norm in their community.

In a quasi-experimental study with Liane Comeau (it was her PhD
dissertation research), we examined how much 3-year old French-English
bilingual children mixed with an unfamiliar interlocutor who changed her
rates of mixing from one session to another, on different days. We found
that even these young bilinguals and even the least proficient ones,
matched the adults rates of mixing -- suggesting to us that bilingual
children who mix are often being responsive to the input they are hearing.

Thsu, extensive child bilingual code-mixing may reflect language
socialization (see Liz Lanza's work in Norway). It might also reflect lack
of full proficiency in one or both languages -- the children we have
studied often code-mix from their weaker into their stronger language
because there are lexical gaps in the weaker language. Thus, a child who
does not get sufficient input in one or both of his/her languages might
code-mix a lot in order to fill these gaps. IThere are probably multiple,
non-mutually exclusive reasons for child code-mixing.

Fred





At 10:07 PM 18/04/2005 -0700, Betty Yu wrote:
>Hello all,
>
>  I am curious if there is convergence in current research on the
>usefulness (or not) of separating languages as a strategy for teaching
>children more than one language (e.g., one-parent/one-language,
>one-situation/one-language). Is there evidence that children really
>become confused by mixed linguistic input given that there's evidence
>that code-switching and other language mixing behaviors are quite
>normal in bilingual communities? I'm especially interested in this
>topic as it relates to children with language impairments.
>
>Thank you for your attention.
>
>Betty
>Doctoral Student at UC Berkeley/SFSU
>
>

Psychology Department			Phone: 1-514-398-6022
McGill University 				Fax:     1-514-398-4896
1205 Docteur Penfield Ave.
Montreal QC
Canada H3A 1B1