On the Generalization of Relative Clause Instruction in the Acquisition of English as a Second Language1

Abstract

This paper reports on an experimental study intended to test the generalization of instruction in second language learning. A group of students in an English as a second language program served as subjects for special instruction in relative clause formation. The subjects were given a pre-test on combining two sentences into one sentence containing a relative clause where either the subject, object, or object of a preposition was the relativized noun phrase. Based on the pre-test results, four equal groups were formed, three of which served as experimental groups and one as the control group. Each experimental group was given instruction on the formation of only one type of relative clause. The subjects were then given a post-test. From the results of the experiment, it is argued that maximal generalizatioń of learning takes place from structures which are typologically more marked to those structures which are typologically less marked, and not the reverse. Some implications of this interpretation are discussed.