It is not hard for children to learn a second language, but we are all
familiar with instances of second language learning that fall short of the
language being learned. Researchers have examined cases of language learning
that may not have gone beyond the interlanguage stages documented in
their studies (e.g., Schumann, 1974; Selinker, Swain, & Dumas, 1975; Wong
Fillmore, in press). In Selinker's (1972) terms, they ended up with "fossilized
versions of interlanguages" rather than with fully realized versions of the
target languages. What leads to this kind of outcome rather than the more
complete mastery expected in second language learning? Social and psychological
factors have been implicated in the case of adult language learners,
whereas situational factors have been critical in the case of children. Fossilized
interlanguages are very likely to develop in the language-learning situations
that we find in many schools with big enrollments of immigrant and
refugee students. In the classrooms of such schools, non-English speakers
frequently outnumber English speakers. In fact, except for their teachers,
the learners may have little contact with people who know the language well
enough to help them learn it. In any event, the language learners spend a lot
more time talking with one another than they do with their teachers, and the
English they hear most often is the imperfect varieties spoken by classmates
rather than the more standard varieties spoken by their teachers. That being
the case, the input they base their language learning on being the speech of
learners like themselves, is not altogether representative of the targel! language.
Not surprisingly, language learning based on such input is neither
perfect nor complete.
What happens when students do not learn well a second language after
they have already decided to give up their first language? Can children who
develop neither their first nor their second language fully take full advantage
of the educational opportunities their parents and their teachers have to
offer? These are questions that need to be examined closely in the light of
what we have learned in this study.
So where does all this leave us? Does this suggest that we should abandon
English in programs for language-minority children? Not at all. The problem
is timing, not English. The children have to learn English, but they
should not be required to do so until their native languages are stable enough
to handle the inevitable encounter with English and all it means. Even then,
teachers and parents must work together to try to mitigate the harm that can
be done to children when they discover that differences are not welcome in
the social world represented by the school. Parents need to be warned of the consequences of not insisting that their children speak to them in the language
of the home. Teachers should be aware of the harm they can do when