Mediated generalization and the interpretation of verbal behavior: II. Experimental study of certain homophone and synonym gradients.

Abstract

In a previous paper (i), the writers presented a theoretical account of the conditioned response mechanism of mediated generalization, and suggested that many phenomena of verbal behavior might be understood in terms of mediated as well as non-mediated generalization. The present paper reports an investigation designed to determine whether some of the suggested theoretical relationships may be demonstrated experimentally.In the above-mentioned report, it was pointed out that if generalization effects should operate as described, they might occur along any of several different dimensions. Thus, if a given word be reinforced, we should expect generalization effects to appear in its synonyms, homophones (i, Figs. 2 and 3), antonyms, and other classes of words which the subject has learned to associate with it in some way. 2The present investigation is limited to certain synonym and homophone relationships. The study was designed specifically to discover whether generalization may be demonstrated to occur to (1) more than one homophone of a reinforced word, (2) a synonym of the reinforced word, and (3) a synonym of a synonym of the reinforced word which is not, however, itself a synonym (or other semantic relative) of the reinforced word. The first two problems involve generalization to stimuli directly related semantically or homophonically to the reinforced word, whereas the third problem involves semantic generalization to a stimulus only indirectly related to the reinforced wordtwo transformations removed from the original reinforced word, cf. (1).The writers have previously (1) summarized the experiments directly concerned with generalization along synonym and homophone gradients. To our knowledge, all of these experiments have been limited to the demonstration of such generalization along a single synonym or homophone gradient from a given word, the synonym 1 The present paper represents part of a research project being conducted by the writers under a grant-in-aid from the American Association for the Advancement of Science.* For a list of certain of these possible gradients along which mediated generalization might be expected to occur, cf. (i).168

Genetic changes in semantic conditioning.

“…The use of mediated generalization studies at different cultural levels and under different environmental stresses may adduce more substantial evidence than^ etymology for the linguistic similarities apparent in the form and structure of the words. "• SUMMARY 4. For the youngest group, mean age 7-9, the homophone received most transferred EDR, the antonym and synonym following in the order stated.…”

“…After, discussing the general principles of semantic conditioning and mediated generalization in relation to verbal behavior (i), these authors performed experiments involving the presentation of lists of words for memorization followed by study of the gains in retention from these lists to lists related to the original in various semantic ways. .Among the relationships studied were antonyms (2), homophones, first and second order synonyms (4). In other experiments, the : authors varied the language.background of the Ss (6) and the professional and occu-pational fields in which the Ss were trained (5).…”

Mediated generalization and the interpretation of verbal behavior: V. 'Free association' as related to differences in professional training.

“…In previous papers of this series, direct and indirect, or mediated, types of generalization were theoretically discussed (3) and experimentally demonstrated along various gradients (8,4). The development of mediated generalization in an everyday-life situation was also experimentally investigated (9).…”

Mediated generalization and the interpretation of verbal behavior: III. Experimental study of antonym gradients.

“…2. Procedure and Scoring.-The procedure and scoring used in this experiment were the same as in the previous study (2). Briefly, the procedure was as follows: Every S first learned the Buffer List and then the E. L. list On the basis of E. L. score (used to equate the reinforcement sub-groups for 'learning ability'), the S was assigned to a reinforcement procedure.…”