M.A. (Applied Linguistics) / The South African English Academy requested that an investigation be undertaken on the feasibility of producing a battery of English proficiency tests in the four modes (listening, speaking, reading and writing) which are free of ethnic labels. For the purpose of this study the aural and oral modes will be developed first as these are the two modes from which language arises. These tests are intended to be universal language tests which will be appropriate for both English first and second language learners. They will be undertaken with mother tongue standards as the first point of departure. It is proposed that mother tongue speakers represent the "universal" standard to which second language speakers aspire, although we are fully aware that native speakers show considerable variation in ability. First language norms will be taken as the norm of performance for this study because it assumed that this is what most second language learners would aspire to. This study will attempt to provide a valid and reliable measuring instrument to facilitate the correct placement of pupils in schools where English is the medium of instruction. Of particular importance is the fact that these are not tests of absolute competence, but are tests of progressive competence. No pupil is a failure, the pupils performance is measured on a continuum of six levels in the senior primary phase. This will enable the teacher to assist each pupil to progress at his own pace according to his own ability. There is a very definite need for performance tests in English since tests of communicative competence in English that are "universal in nature" are not available in South Africa. A standardised test battery aimed at measuring the language proficiency (of the heterogeneous population at the senior primary level) across all four language modes, does not exist in South Africa. The problem is to .develop tests with tasks at the appropriate level for a heterogeneous population and produce evidence regarding their validity. Although subjective and objective assessment methods are commonly used, the merits of the one method as opposed to the other are still debated. The importance of investigating the relationship between the multiple-choice aural test (by means of objective and indirect assessment) and the productive oral test (by means of subjective and direct assessment) cannot be underestimated.

We as teacher of English are faced with the task of preparing students for the world they will meet when they leave the classroom—a world that will judge them in part by their ability to manipulate the language to their advantage. And yet to measure the use of language, we rely almost exclusively on written measurements. It seems likely that students may easily learn to give the teacher the answer he wants on written tests of English usage. Oral tests are commercially available, but many are prohibitively expensive; others require hours of instruction before the teacher can consider himself qualified to administer the test; and still others include the measurement of so many varied objectives that they do not allow the teacher to pinpoint a particular language problem on which he may want to concentrate. Some oral tests have all of these drawback. It is hopes that a test of spoken English usage—one which directs itself to a specific language problem—can be constructed that will overcome the drawbacks mentioned above. And it is hoped that such a test will prove to be adaptable to classroom uses in much the same way that a written test would be; that is, it can be constructed by any classroom teacher. It is hoped that such a test can be proven valid. What form should a test take that is designed to measure spoken English usage and at the same time, is designed to be specific, usable, and readily adaptable to classroom use?