Permalink : http://hdl.handle.net/2115/45665

This paper argues for the importance of collocations, as opposed to single words, as basic units in L2 vocabulary acquisition. Collocations are defined here as commonly occurring combinations of two or more lexical words in a syntactic construction: examples are ame ga yamu, ofuro ni hairu, karada ga ookii, atsui oyu, kawatta namae, nabe no soko, hayaku neru, kirei ni arau, hijooni muzukashii. Collocations are seen as providing models of natural language use which contribute to both understanding and production; they highlight interesting points of difference as well as similarity between languages; and since they involve syntactic constructions, they also reinforce the acquisition of grammar. The importance of collocations in acquisition is discussed with reference to four semantic types of Japanese words: verbs which are used in highly general meanings such as kakeru (as in bureeki o kakeru); mimetic adverbs like tekuteku whose meanings are, by contrast, very specific; nouns like terebi, whose meanings are arguably identical with L1 items; and cases like tsumetai and samui, where two or more L2 items cover the semantic range of a single item in L1 (as with Eng cold). In the first two cases, the need to acquire collocations like enjin o kakeru and tekuteku aruku is self-evident, but even with semantically unremarkable items like terebi the control of collocations such as terebi o tsukeru is necessary if the word is to be used naturally in sentences. With tsumetai and samui, collocations like samui hi and tsumetai mizu serve to reinforce the basic semantic distinction between these two words; even when this distinction is grasped, however, borderline cases for learners still arise, and further collocations such as kaze ga tsumetai, ashi ga tsumetai ('feet feel cold'), ashi ga samui ('legs feel cold') provide models of natural use in these areas. As Lewis and others have stressed, control of vocabulary requires mastery of the behaviour of words in combination as well as of their individual meanings, and increased focus on collocations appears to offer great potential benefits for vocabulary acquisition. This needs to be supported by further studies of collocation patterns and corpus frequency in Japanese linguistics, and by the incorporation of more detailed collocational information in Japanese learner dictionaries.