The purpose of this paper is to briefly describe the functions of a set of computer programs developed locally at The Takefuta Laboratory of Linguistic Engineering to examine and compare English passages objectively. The first program counts the frequencies of the various linguistic units appearing in the passages, and then the other programs accomplish several kinds of data processing, including transformation of the frequency data, extraction of quantitative and qualitative features, application of a factor or a cluster analysis, etc., in, order to analyze, compare, and classify the sets of passages. We hope that such methods as the ones described here will improve the reliability and the validity of the criteria for selecting appropriate teaching materials, or testing materials, for a better teaching of English as a second language in Japan. We want to acknowledge here that the facilities of the Computer Center of Tokyo University and those of the Data Station of Chiba University were used to develop and test the programs.

抄録全体を表示

According to Rivers' theory, it is important to develop the communicative competence of students in the "natural uses of language." Therefore, "skill-getting" exercises are not enough: in our teaching we must lead on to "skill-using" practice. To do this, 1) we have to give a situational context and lexical content to the students; and 2) we must shift from teacher-directed practice to student-directed practice. The French section of the Language Center of Nagoya University made, from that point of view, teaching-materials comprising a series of skits on 8mm films and a textbook containing explanations of grammar and many exercises, which are also recorded on cassettes. To save time in class, students can practice the exercises at home using the cassettes (skill-getting), and in class, use the skills they have obtained in a kind of pseudo-communication (skill-using) based on the situations and vocabulary given in the films.

抄録全体を表示

From a theoretical point of view, listening comprehension(LC) is interpreted as a highly integrative skill which demands an active process of "analysis by synthesis," rather than receptive decoding. This paper investigates LC empirically: first, by correlations between LC and overall language proficiency, second, by correlations between LC and other language skills, and finally, by correlations among different types of test items designed to measure LC. The so-called "integrative approach" to testing, in contrast to the "discrete-point approach," provides a useful means of measuring overall language proficiency through "doze tests,""noise tests," and dictation. Present research is based on students' performance in the entrance examination and on a LC test. The result (correlation 0.51) suggests that common factors are involved in both entrance and LC tests on the level of linguistic elements, whereas differences can be detected on the higher level of integrative skills. Various studies of the relationship between LC and other language skills have shown that although LC involves basic linguistic features common to all skills, it also exhibits inherent qualities found in more specific skills. This paper discusses experiments in three different English courses: reading, composition, and LL. It demonstrates that the generally low correlations were due mainly to reliance on achievement rather than proficiency, a reflection of student attitudes toward learning. It is generally recognized that different tests measuring a particular skill reveal varying results, even from the same subjects. The very low correlation obtained here between a phoneme discrimination test and an LC test implies that the former cannot be used to measure LC; otherwise, its testing validity would be lost. After investigating four sub-tests of the LL term examination, the paper concludes that the four test items are appropriate on two grounds: (1) they manifest a contributory/distinctive power with their similar high correlations to the total score, and (2) they demonstrate rather low intercorrelations among themselves, which facilitate the validity and reliability of the total examination.

抄録全体を表示

Accompanying the popularization of audio-visual aids-the language laboratory in particular-the pronunciation of English individual sounds spoken by Japanese college students who major in English is getting better year by year. But if you take a closer look at it, you will find that the pronunciation of all these sounds is not becoming equally easier for them. It is interesting to note that some sounds cannot be imitated by Japanese students or understood by native English speakers while others can easily be transferred into audible English sounds. In addition, it is generally said that the listening and pronouncing abilities are highly correlated, but this may not always be the case for every individual sound. In the scope of consideration the present paper deals with the problems on (1) which sounds are more difficult to be pronounced or corrected, (2) what kind of replacements occur among rather advanced learners, (3) how native speakers respond to the pronunciation of particular English sounds spoken with complete Japanese "accent," and (4) what is the correlation between listening and pronouncing abilities.

抄録全体を表示