27
at a policy level" (p. 50). Cummins points out a pattern of results
in hundreds of evaluations of bilingual programs as evidence of the
common underlying proficiency of language acquisition. He further
points to his correlational study (1979) of first and second languages
as evidence in support of the CUP theory in terms of academic skill.
BICS are not so easily transferred across languages. Age of entrance
into the second language learning environment also affects speed and
amount of transfer, with older children mastering second language
morphology and syntax skills, the skills taught at school, more quickly
than younger children. Younger children have an advantage in phonology,
essentially a BICS area of language proficiency. Of importance here is
the fact that students had mastered the second language sufficiently
to score equally, or nearly equally as well, on both first and second
language tests. Cummins (1979) points out the need for empirical
investigation of the relationship of first language proficiency and
second language acquisition.
Cummins et al. (in press) conclude after researching dissimilar
populations of Japanese and Vietnamese students that first language
". . cognitive/academic proficiency accounted for a highly significant
proportion of variance in L2 (second language) cognitive/academic pro
ficiency, as predicted by the interdependence hypothesis" (p. 32).
Older Japanese students continued to develop first language cognitive/
academic skills to a greater degree than younger Japanese children in
the study. According to the researchers (Cummins et al., in press),
their recent work has been exploratory but has allowed them to conceptu
alize the interdependence theory (of the relationship of first language