Cross-linguistic treatment generalisation in Welsh-English bilingual aphasia

Tainturier, M. & Roberts, J. R.

School of Psychology, Bangor University

This study investigates the conditions under which treating one language can lead to improved naming in the other. We examined this issue using a phonological cueing treatment in a bilingual anomic woman. We hypothesized that cross-linguistic transfer would be maximal for cognates of treated words, based on psycholinguistic models positing stronger connections between cognates in the
bilingual lexicon. HBL, 67, a Welsh-English early proficient bilingual, suffered a LH stroke in 1986. An extensive investigation of her naming abilities revealed a deficit to the phonological lexicon in both languages.
Her picture naming was treated in English using phonemic cueing. HBL
received treatment twice a week until reaching ceiling on the treated set (15 sessions). She was asked to name pictures and, when unsuccessful, was given phonemic cues of progressive length until the item could be named. Generalisation to untreated English and Welsh words was assessed by comparing performance pre- and post treatment on treated English words, untreated unrelated English control words, untreated Welsh translations of English treated words (including half cognates) and untreated unrelated Welsh words.
In English, HBL improved on treated words but not on untreated control words. As predicted, untreated Welsh cognates of treated words improved. There was no change in the naming of non-cognate translations, and a decrease for unrelated Welsh controls. In conclusion, this study shows that treating picture naming in one
language can lead to improvement in the untreated language. Generalisation was limited to cognates, supporting models positing stronger lexical connections between cognates. We propose this allowed for an indirect activation of Welsh cognates when corresponding English words were treated, boosting the strength of their lexical representations. This study is the first to demonstrate cross-linguistic generalisation using phonological cueing and with a patient with a selective deficit to the phonological lexicon.