TESTING ENGLISH AS A SECOND LANGUAGE
(WHAT IS MEANT BY SPEAKING A SECOND LANGUAGE)
If we compare the inventory with that which was presented in the preceding chapter on the testing of writing, one essential difference is apparent. As used in the chapter, the term writing was applied to a rather formal and sophisticated language activity not ever fully developed by many students, even in their first language. In contrast, when we refer to a student’s skill in speaking a second language, our fundamental concern is with his ability to communicate informally on everyday subjects with sufficient ease and fluency to hold the attention of his listener. Thus in our tests of speaking ability, we are primarily, if not solely, interested in the foreign student’s control of the signaling systems of English—his pronunciation, grammar, and vocabulary—and not with the idea content or formal organization of the message he conveys.’
THE MAJOR PROBLEM IN MEASURING SPEAKING ABILITY
In earlier chapters, we observed how three of the speech components—grammatical structure, vocabulary, and auditory comprehension—are now being tested by reliable and relatively simple objective techniques. It is highly probable that performance on these tests is positively related to general ability to converse in a foreign language, although, as will be explained directly, we still lack very reliable criteria for testing out this assumption.’ General fluency, too, is fairly easy to assess, at least in gross terms: it usually takes only a few minutes of listening to determine whether a foreign speaker is able to approximate the speed and ease with which native speakers of the language typically produce their utterances.’ It is only when we come to the crucial matter of pronunciation that we
are confronted with a really serious problem of evaluation. The central reason is the lack of general agreement on what “good” pronunciation of a second language really means: is comprehensibility to be the sole basis of judgment, or must we demand a high degree of both phonemic and allophonic accuracy? And can we be certain that two or more native speakers will find the utterances of a foreign speaker equally comprehensible, or do some listeners decode a foreign accent with greater facility than others? Until we can agree on precisely how speech is to be judged and have determined that the judgments will have stability, we cannot put much confidence in oral ratings. All that we can offer in this chapter, then, is a brief summary of the present state of a very imperfect art. Let us hope that future research may yet transform it into a reasonably exact science.
TYPES OF ORAL PRODUCTION TESTS
Most tests of oral production fall into one of the following categories:
- Relatively unstructured interviews, rated on a carefully constructed scale
- Highly structured speech samples (generally recorded), rated according to very specific criteria
- Paper-and-pencil objective tests of pronunciation, presumably providing indirect evidence of speaking ability
Of the three, the rated interview is undoubtedly the most commonly used technique and the one with the longest history. Paper-and-pencil tests of pronunciation have been used off and on for some years, generally in combination with other types of assessment. Highly structured speech samples, as the term will be used here, appear to be relatively recent and have not as yet won much acceptance in American testing of English as a second language.
- Sentence conversion. The examinee is instructed to convert or transform sentences in specific ways (from positive to negative, from statement to question, from present tense to past, etc.). The voice on the tape gives the sentences one at a time, the examinee supplying the conversion in the pause that follows. Scoring procedure: The rater scores each converted sentence on the basis of whether or not it is grammatically acceptable.
- Sentence construction. The voice on the tape asks the examinee to compose sentences appropriate to specific situations. Scoring procedure: The rater scores each sentence on an acceptable-unacceptable basis
6.Response to pictorial stimuli. The examinee is given time to study each of a series of pictures and then briefly describes what is going on in each scene. Scoring procedure: For each picture, the rater gives a separate rating of the examinee’s pronunciation, grammar, vocabulary, and fluency. Oral-production tests comprising the above, or similar, types of highly structured speech tasks offer considerable promise as replacements for the unstructured interview, for they greatly increase both test and scorer consistency. However, it must not be forgotten that the scoring still requires human judgments, and satisfactory
7.Because of the relatively complicated nature of the directions, it would probably be best for these to be given in both oral and written form. Even then, poor response or lack of response might conceivably indicate a reading/listening difficulty on the part of the subject.
reliability can be achieved only if the raters are carefully selected and are put through rigorous training sessions.
- Moreover, such tests demand a great deal of time and care in preparation and usually re-quire mechanical devices for proper administration. In short, structured speech-sample tests provide no shortcuts to effective oral testing. Paper-and-Pencil Tests of Pronunciation So far in this chapter we have described oral rating techniques which call for the eliciting and evaluating of samples of actual speech. For some years test writers have experimented with objective paper-and-pencil tests of pronunciation in which the subjects have merely to check responses indicating how they pronounce English vowels and consonants and how they stress words and phrases. Such tests assume
(i) that what the foreign learner “hears” himself say silently is, in fact, what he says aloud, and
(ii) that a sufficiently broad range of pronunciation problems can be tested by this indirect method to allow us to generalize about a subject’s overall control of the English sound system. Before discussing these assumptions, let us illustrate characteristic item types appearing in these paper-and-pencil pronunciation tests.
Rhyme words: The examinee is first presented with a test word which he is instructed to read to himself, after which he is to select the one word from among several alternatives which rhymes with the test word.