ITL - International Journal of Applied Linguistics - Volume 87, Issue 1, 1990

The aim of the present paper was to find out which learning strategies secondary school pupils of different ages employ to acquire complex words in English as their second language: either by applying rules and analogies or by learning by heart. As a working hypothesis, it was postulated that younger pupils probably preferred the latter approach and older pupils the former. In order to test this hypothesis, a 122-item complex word derivation test was devised, containing three categories of words: (1) words of which both the base-form and the derived form had been studied, (2) words of which just the base-form had been studied and (3) words of which neither form had been studied. The test was administered to pupils in three grades of secondary school and a group of 1st year university students of English. Statistical treatment of the data neither confirmed nor falsified the original hypothesis, but it showed many correlations and gave rise to a large number of additional conclusions. Amongst other things, it was concluded that the presence of the proposed tentative change-over in learning approach, from learning words as whole entities to applying word-formation rules, may or may not have been present, but if it were, it had been completely obscured. It was evident from several different indications that a dominant influence on the pupils' scores was exerted by exposure. In addition, the data collected revealed numerous correlations concerning the influence of education level, word category, regularity, frequency, etc. Finally, suggestions are given for application of the test results in second language education in secondary school in general.

Various parameters pertaining to selection and sequencing patterns are defined on phonological chains broadly transcribed. We apply some multivariate schemes with these parameters to produce quantitative frames for the phonological study of part of the Sankoff-Cedergren corpus of Montreal French, containing more than half a million phonological units. The frames bring a quantitative description for the notion of phonological plane with its paradigmatic and syntagmatic axes. We generalize the procedure to describe more precisely sequencing patterns of the phonological chain with various planes and spaces. We also show that there is a quantitative interpretation, in terms of our variables, for the usual but imprecise distinction between the registers of Montreal French. The same methodology can be used for the similar planes defined in the upper language hierarchy, and for a new quantitative description of perturbed languages based on the direct observation of spontaneous language chains.