Bilingual Recognition of Concrete and Abstract Sentences

Abstract

Eight French-English bilinguals were presented two lists of sentences constructed from different scenarios. For some of the scenarios, the language of the sentences was constant but for others, varied. On a subsequent recognition test, subjects received three types of sentences: (a) some previously presented (literals), (b) some semantically related but not previously presented (recognitions), (c) some differing only in language of presentation (translations). For both abstract and concrete scenarios in which initial presentation was in one language subjects made significantly fewer false acceptances than correct acceptances. However, for concrete scenarios in which the initial language of presentation had varied subjects accepted as many translations as recognitions and as many recognitions as literals. With the abstract scenarios which varied in the language of initial presentation subjects accepted as many recognitions as literals and more translations than literals. The results were interpreted in terms of the principle of variability of encoding.