Key points are not available for this paper at this time.
Vocabulary can be learned by tracking cross-situational statistics between words and aspects of the environment. Nouns and verbs – with observable referents in the environment – can be acquired rapidly, but whether more abstract morphological features can also be acquired cross-situationally, and whether such learning is affected by prior language experience, remain open questions. One hundred and seventeen adults with native languages varying in morphological richness (Chinese, English, German) were exposed to an artificial language comprising intransitive sentences with morphology indicating number, tense, and subject-verb agreement. For each sentence, two scenes appeared on a screen, and participants selected which matched the sentence. Participants learned both vocabulary and morphology from cross-situational statistics, but we found no evidence for an effect of language background. Cross-situational learning is a sufficiently powerful mechanism to drive acquisition of morphological features as well as vocabulary, and is robust to differences in morphological expression from speakers’ native languages.
Zhu et al. (Sat,) studied this question.