Abstract Brown & Gullberg
Changes in L1 Encoding of Motion after Exposure to an L2
In second language acquisition research on the expression of motion, like in many areas of L2 research, the relationship between the L1 and the L2 is generally portrayed as being unidirectional. Thus, L1 features have been shown to transfer to the L2 in speech (Cadierno, 2004) and even in gesture (Negueruela et al., 2004). Such findings imply that linguistic patterns of encoding motion, while often simply preferences, are robust features of an L1. The current study, however, examines the extent to which linguistic preferences for encoding motion in the L1 are indeed robust and impervious to change by investigating whether emerging L2 patterns can influence established L1 patterns within individual speakers.
Narrative descriptions of motion events from the cartoon Canary Row were elicited from monolingual Japanese speakers (n=16), monolingual English speakers (n=13), and native Japanese speakers with intermediate L2 English (n=28). The latter group was divided into two sub-groups: participants living in Japan with little or no exposure to U.S. culture (n=15), and participants living in Boston with full cultural immersion (n=13). These two sub-groups were matched on formal proficiency in English.
Analyses of expression of Path confirmed previously found differences in monolingual lexicalization patterns. In speech, monolingual Japanese speakers lexicalize Path primarily in the verb as in (1), whereas monolingual English speakers lexicalize Path primarily outside the verb as in (2). In their L1, however, native Japanese speakers with intermediate L2 English packaged significantly more Path elements of both linguistic types into each verb clause than either monolingual group, as in (3).
Neko-ga kondo-wa amadoi-no naka-wo tsutat-te
Cat-NOM this.time-TOP downpipe-GEN inside-ACC go.along-CON
“This time the cat goes along the inside of a gutter”
Sylvester crawls up the drainpipe
Chijyou-kara tokoro-made nobot-te ik-outo shi-tan-desu-kedo
Ground-from place-to climb-CON go-VOL do-PAST-be-but
“(He) went climbing up as hard as he could from the ground to the place”
In gestures depicting translocational motion, Japanese monolinguals used significantly more character perspective (sagittal and bi-manual, employing an enactment hand-shape) than English monolinguals, who preferred observer perspective (lateral and single-handed, employing no enactment hand-shape). Again, native Japanese speakers with intermediate L2 English patterned significantly differently in their L1, showing equal tendencies to adopt either perspective. Moreover, similarities in L1 gesture patterns between learners resident in Japan and the U.S. imply that variation from the monolingual Japanese baseline does not result from cultural exposure, but may arise from linguistic convergence between the L1 and L2.
Results suggest that encoding of motion in an L1 may be affected by the presence of an L2 system, even at intermediate levels of L2 proficiency. Thus, the domain of motion, an area exhibiting considerable crosslinguistic differences, is also a domain where we can observe the interaction of languages within the multilingual mind.