【正文】 scribed one mon problem amongst EFL students: not being able to access a word in speech and lacking the ability to circumvent that word by providing a definition in the target language. He attributed this problem primarily to students39。 use of bilingual dictionaries and strongly advocated the use of monolingual dictionaries that would encourage conversational definition (p. 335). In general, Baxter reiterated the basic concerns of most language teachers, that bilingual dictionaries 1) encourage translation。 2) foster onetoone precise correspondence at word level between two languages。 and 3) fail to describe adequately the syntactic behaviour of words. By contrast, Thompson (1987) argued against monolingual dictionaries and supported the development of a new generation of learners39。 bilingual dictionaries (p. 286). He pointed out that monolingual dictionaries tend to be circular in their definitions, ., laugh, amuse, amusement and humour are normally used in each other39。s definitions. Even if defining vocabulary is restricted, monolingual dictionaries still employ a special register which is not necessarily the most useful or rewarding for learners to be exposed to (p. 284), and are therefore of little value to foreign language learners below the advanced level. Thompson did admit that objections to traditional bilingual dictionaries are valid, and he advocated the pilation of new bilingual dictionaries that, in addition to providing clearer understanding in the learners39。 L1, avoid reinforcing the belief in a onetoone relationship at word level (p. 285), and provide full semantic, grammatical, and stylistic information, examples, and usage notes that are not available in traditional bilingual dictionaries. [7]Since a bination of good features of both types of dictionaries is not impossible, there has been considerable interest in the last twenty years in the new bilingualised promise dictionaries, hybrid dictionaries that essentially provide translations in addition to the good features of monolingual dictionaries (Hartmann, 1991, p. 79). Evaluation of the effectiveness of such dictionaries emerged mainly in the 1990s. Laufer and Hadar (1997), for example, pared monolingual, bilingual, and bilingualised dictionaries among 123 EFL learners in Israel. They found that irrespective of the learners39。 proficiency level, the bilingualised version was either significantly better than, or as good as, the other two types in both prehension and production tasks. Edictionaries Recent developments in puters have triggered a whole line of interest in electronic dictionaries, online dictionaries or vocabulary glosses integrated into language learning software or web pages (., Hulstijn, 1993。 Knight, 1994。 Koren, 1999。 Laufer amp。 Hill, 2000。 Lomicka, 1998。 Roby, 1999). In general, the same questions that have been asked of traditional dictionaries are being asked of their eversions. Not surprisingly, very similar answers are obtained, for example, electronic dictionaries that contain not much more than L1 equivalents are not quite helpful to the language learner (Koren, 1997). Two new developments warrant special mentioning: 1) puters offer researchers a powerful and convenient tool in terms of logs or trackers of learner behavior in dictionary use。 and 2) online vocabulary glosses offer the learner a quick access to the information s/he needs which in turn might encourage more dictionary use. However, clicking on a hyperlink is a lookup strategy totally different from flipping through a bulky dictionary, locating the relevant entry, and finding the contextually meaningful information. While the ease and speed might encourage more dictionary use and reading, the convenience might not always be a good thing for vocabulary learning. By the same token, the online logs we obtain about learners39。 dictionary behaviors might not contain exactly the same behaviors the same learners would demonstrate when they use paper dictionaries. Clearly we need to see more work along this exciting line of research before we can arrive at any fortable conclusions about online dictionaries and glosses. Perhaps we have bee more skeptical about a single most appropriate dictionary format, perhaps we are more wary about dogmatic statements on which dictionary is good for you, and realize that more research is needed on what real dictionary users do in real situations of dictionary lookup (Hartmann, 1991, p. 79). Indeed, the field is beginning to take up this challenge. In addition to the experimental parison of different types of dictionaries, more studies are emerging that aim to discover what exactly learners do and how their dictionary strategies influence their learning results.Dictionary strategiesLearners39。 dictionaries are certainly piled with the language learner in mind. And almost every such dictionary is acpanied by at least one workbook (most notably Underhill, 1980。 Whitcut, 1979) aiming for the training of dictionary strategies to maximize the effect of dictionary use in language learning. With only a handful of exceptions, little has been done empirically to find out what dictionary strategies are used by learners and whether and how these strategies influence their learning outes. [8]In an analysis of the steps learners need to look up a word for prehension, Scholfield (1982) suggested seven steps for the language learner, and analysed each step in great detail (pp. 186193). He argued that making use of a dictionary should not be seen as a straightforward technical and passive activity, it is rather a plex process of hypothesis testing that involves the active participation of the learner. Similar views and strategies are also presented in Nation (2001, ). What Scholfield and Nation described is a synthesized and idealized dictionary strategy a learner should use when looking up a word during reading. We do not know, however, whether L2 dictionary users do make use of these steps. Nor do