Influence of second language immersion in Cherokee on children's development of past tense in their first language, English

Tracy E. Hirata-Edds
Child Language Program, University of Kansas
August, 2006


Previous researchers have suggested that metalinguistic skills may develop differently for multilingual children compared to monolingual peers, although there is no clear evidence of overall advantage or disadvantage. I investigated the effect of learning Cherokee as a second language (L2) on attention to properties of English (L1) for children aged 4;5--6;1 in a Cherokee immersion program in Tahlequah, Oklahoma. To determine if learning Cherokee influenced metalinguistic skills related to noticing morphological forms and patterns of language, I compared English past tense marking skills for 10 children in the Cherokee immersion program with 13 peers in English-medium classrooms. I compared children's English skills on two non-word and two real word tasks. On the non-word production task, both groups productively applied regular -ed suffixes to indicate past tense. For the non-word forced choice task, the groups were not significantly different in their choice of irregular verb forms versus overregularizations. On the elicited imitation of sentences task, regular and irregular verbs scores were not different between groups. However, the finiteness measure (irregular forms plus over-regularizations) was significantly different, with children in the Cherokee immersion group performing better than those in the English-medium group, indicating the former used their own representational systems to override the input (by using -ed ). Scores on the Test of Early Grammatical Impairment for regular, irregular, and finiteness were not significantly different.

My research indicated that development of English past tense marking by the children learning Cherokee was progressing as well as that of their peers and that, notably in one area, they had developed increased attention to productive morphological patterns and focus on language properties not as apparent to their monolingual peers. Because children in the Cherokee group were only in their second year in the immersion program, they may not yet have reached a threshold level of fluency necessary to consistently reflect an influence of an L2 on metalinguistic abilities of an L1. However, even in this early developmental stage, their English past tense skills seemed to show some signs of developing abilities to extract structures, rules, and concepts.