In order to examine the decoding strategies employed by three contrasting groups in terms of nationality and proficiency (one Venezuelan and two Japanese), their proficiency was measured by two types of tests, tapping linguistic vs. functional competence. The test results obtained, as can be seen in Fig. 1 above, show that while the Venezuelan group achieved a balanced development between linguistic and functional competence, indicated by the developmental line at 45 , the profile of Japanese students is distorted and develops in a biased fashion leaning toward linguistic competence. These contrastive results led me to conclude that the Japanese students were characterized as possessing only linguistic competence, and especially low-achievers were found to suffer from fragmental knowledge of linguistic elements, which did not function as an integrated skill of listening comprehension.

The Japanese students' noticeably poorer performance in the subtest of discourse understanding suggests that they lack skills of higher-level operations, i.e. their functional skills are not developed enough to be commensurate with their knowledge of grammar and vocabulary. This finding thus disclaims the assertion made by Oller and others. In short, listening comprehension is NOT a good measure for overall proficiency in a foreign language, at least as far as Japanese students are concerned. Although LC may share the same repertoire of language elements, it is after all a separate skill and will not develop without training specifically how to integrate those elements into the spoken format.

Unfortunately, such a unique profile of English proficiency is the outcome of the Japanese teaching practice, where grammar-translation is still dominant. However trendy the communicative approach may be, we cannot disregard the prevailing practice in ordinary classrooms in Japan.

(3) This brings us to the third area of conflict with the "universal" standard. What is taken for granted as common sense in the West may not be common in Asia. If one observes an English class-an ordinary class at an ordinary school, not a showy model class where native-speaker teachers are usually invited-what is most obvious is that English is hardly used, except when reading the textbook.Clearly, Japanese teachers tend to repeat the way they themselves were taught, while native-speaker teachers have no alternative but to rely on English. Yet this question about the medium of instruction should not be a categorical decision as to which is good or bad, but the role of L1 should be examined according to the learners' age, the content of the material, etc. There is no arguing that the teacher should provide as much input as possible in the FL classroom. But it is also imperative to make teaching efficient where time and exposure are limited. For these reasons it may be beneficial to take advantage of the cognitive knowledge and skills the learners have in their L1. L1 is more efficient, for example, when it comes to understanding complex grammar rules and abstract terms, or for making a quick summary of the text that has just been read.

The recent emphasis on the communicative approach is confronted with a resurgence of grammar because it has failed to instill grammatical competence. Consequently, the mainstream of research now is conducted under the name of "awareness-raising." However, this is not compatible with the grammar-translation-based tradition in Japan. On the contrary, what is called for is "awareness-lowering" instead. Our students should be given more practice to forget about grammar, in other words, to depart from the conventional form-focused instruction.

Krashen's distinction between learning and acquisition is a useful one when we try to analyze the problems of EFL in most parts of Asia. In our learning-oriented EFL context, nobody denies the importance of input, but Krashen's argument that "learning cannot become acquisition" (1982;22) is rather naive at best. There are a great number of Japanese who have become proficient, thanks in part to English learnt at school, combined with a later exposure through studying or working in an English-speaking country. They did successfully turn learning into acquisition. Thus, what Asians generally should strive for in an input-impoverished EFL context is how best to combine learning with acquisition.

III. Changing Trends in Research
As I have discussed so far, we should pay more attention to the local characteristics of EFL in Asia, which cannot be explained by universal generalizations. When looking at EFL research, a similar trend can be observed: in this case, a widening gap between theory and practice.
As Lightbown points out in her Anniversary Article in Language Learning, 2000, although SLA studies have come to have more impact on teaching, the greatest problem lies in the difference between the teaching environment and the research context. In order to be "scientific," on the one hand, fragmental studies are conducted in the form of mini-experiments, which are too narrowly defined and are hardly applicable. On the other hand, the educational experiments in the classroom are difficult to manage because of too many variables beyond control, which are therefore not scientifically exact.

With such basic conflicts in mind, I embarked on analyzing the articles that appeared in the JACET (Japan Association of College English Teachers) bulletin for the last 29 years since 1975 (Nos. 6-36). 263 articles were classified according to the framework by Nunan (1992), with some modifications to fit into the Japanese context. The objective was to discover any change in the relevant research during the last 3 decades-in particular, with regard to the area of research and the research method.

Figure 2: JACET Articles 1975-2003

As these graphs show, there has been a remarkable change both in the area of research and in the research method. In the area of research, the number of articles that deal with SLA has dramatically increased from 22.6% to 52.3%, whereas articles in English philology and teaching methods/techniques have declined. SLA here includes, for example, learning strategies, order of acquisition, and error analysis, while English philology covers grammatical usage studies, sentence stress, etc.

As for the research method, so-called "literature studies" have declined and elicitation shows a sharp increase from 21.0% to 63.6%. Literature studies refer to comparing and discussing theories and researches, leading to a certain conclusion, while elicitation includes questionnaires, listening tests, production tasks, etc. From these two graphs, we could summarize the changes in our field of study with respect to the content and method, as follows: namely that there has been a shift of emphasis from argumentative studies based on literature studies in EFL to data-driven experimental studies in SLA. In short, data-driven statistics have replaced philosophical discussions. Is this shift a natural evolution of the discipline or is it due to the influence of the trends in the applied linguistic studies in English-speaking countries? Such changes seem to conceal other aspects of practical importance. For example, although English educational reforms have been much talked about in Japan, no articles on language policy can be found in the recent journals.

This discrepancy may be rooted in the historical background. There is an interesting difference in the derivational history between applied linguistics in the West and ELT in Asia. That is to say, the discipline classified as applied linguistics, sometimes pejoratively called "linguistics applied," derives from theoretical linguistics and SLA research tends to show a strong inclination in that direction, whereas the Asian approach to applied linguistics has evolved from the pedagogical tradition of ELT-with more emphasis on teaching and learning in a formal educational setting.

IV. The Role of Non-native Speakers
In the century-long history of language teaching, especially since Chomsky put forth the idea of "ideal speaker-hearer," the native speaker has been idolized. However, looking back at ourselves, we soon realize that none of us are ideal native speakers-to illustrate, I as a native speaker of Japanese may not be able to read some of the difficult Chinese characters used in Japanese, or I may not be able to understand certain economic news in the newspaper. I am a competent but not a perfect native speaker. Furthermore, I am a naive native speaker who can speak but cannot explain. The other day I learned for the first time about the usage of the two particles, 'yo' vs. 'ne,' from a Korean student. I have no problem using them, but I had no conscious knowledge about the distinction, which would disqualify me as an efficient teacher of Japanese as a second language.

This is where the role of the non-native speaker (NNS) teacher fits in-particularly, as a role model and a sympathetic teacher. As Cook (1999) proposes, the NNS as a teacher can provide a valuable example of skilled L2 use and therefore be the role model of a successful L2 user for the students. The NNS teacher is someone who has arrived where the students want to be. Such a teacher not only presents a more achievable model, but shares the linguistic/cultural background. The NN teacher is especially valuable because of the learning experience that the teacher can share with learners. They have gone through the problems and difficulties that language-specific learners are likely to encounter.

L2 users then should be viewed as multicompetent language users, rather than deficient native speakers. Language teaching can tap into a mind that already contains an L1; that is to say, the new language is learnt on the basis of a previous language. Multicompetent minds that know two languages are qualitatively different from those of the monolingual native speaker. As the studies of bilingual brains indicate, multicompetent language users, like bilinguals, have to be looked at in their own right, not as possessing two deficient monolingual competences (Oka, 2002).

V. Intercultural Communicative Competence
These days English is often used as a lingua franca even among non-native speakers from a variety of language backgrounds. However, for cross-cultural communication to successfully take place we need more than a common means of linguistic codes. We need to have cultural and functional literacy, not simply NS-oriented communicative competence. We need a broader framework of "intercultural communicative competence" (ICC).

ICC can be defined as "a person's ability to relate and communicate with people who speak a different language and live in a different cultural context" (Byram,1997:1). This notion thus emphasizes the importance of sharing cultural and functional literacy, so that the intercultural communicator can function in the global world.

Developing ICC may not be within the scope of the language classroom alone. Cultural literacy, for example, requires a basic knowledge of the world, such as who Lincoln is, where Japan is, etc. Functional literacy is the knowledge and skill required to function in modern civilization, such as how to use a telephone, how to ride a train, etc. To achieve cross-cultural understanding, encoding/decoding of information should be appropriate in quantity and quality on the basis of shared knowledge. "Appropriateness" was first introduced into the concept of communicative competence with reference to being socially appropriate, but in ICC it refers to being appropriate in terms of quantity and quality of information. Information should not be just necessary but also sufficient. An appropriate amount of information should be provided in an effective manner according to the knowledge shared between speaker and hearer. For example, depending on how familiar the addressee is with Japanese culture, one might have to explain in great detail how to take a bath in a Japanese home-or no explanation may be necessary. This concept even applies to the language classroom. When asked "Did you have a nice weekend? " Japanese students tend to just respond, "Yes, I did." But they should be made aware that this is not enough. They must be taught how to interact-e.g. by saying, "Yes, I went to Tokyo Disneyland with friends. It was great fun. Have you been there?"

ICC emphasizes the significance of the procedural aspect of communicative interaction. Due to the dynamic nature of interaction, information must be adjusted relative to the gap as well as the common ground between interlocutors. Such procedural competence is as essential to ICC as knowledge and skill. It will put everyone on an equal basis as intercultural communicators. In this way, ICC can become the key concept to liberate the Asian TEFL in the decades to come.

**Special acknowledgement goes to Mr. T. Nakata, who helped me with the JACET articles.