Omission of definite and indefinite articles in the spontaneous speech of agrammatic speakers with Broca's aphasia

Abstract

Background: Cross-linguistic investigation of agrammatic speech in speakers of different languages allows us to tests theoretical accounts of the nature of agrammatism. A significant feature of the speech of many agrammatic speakers is a problem with article production. Månsson and Ahlsén (2001) report that Swedish agrammatic speakers have more problems with the indefinite than with the definite article. In Swedish, the indefinite article is a free morpheme, whereas the definite article is a suffix to the noun. However, in other languages all articles are free morphemes. We ask whether the indefinite article is more difficult. Aims: The aim of the present study is to examine whether the discrepancy between the production of the definite and indefinite article is because it is indefinite or because it is a free morpheme. Our objective is to examine whether a discrepancy in definite and indefinite article production is also found in Dutch agrammatic speakers, where all articles are freestanding morphemes. Methods & Procedures: The spontaneous speech of eight Dutch agrammatic speakers with Broca's aphasia was analysed with respect to the production of definite and indefinite articles. Outcomes & Results: The Dutch agrammatic speakers had problems with the production of articles. However, no significant difference was found between definite and indefinite articles, although a trend was observed towards definite articles being more impaired than indefinite articles in a language where all articles are free morphemes. Conclusions: The finding of Månsson and Ahlsén (2001) that Swedish agrammatic speakers have particular problems with the indefinite article is probably due to the fact that the definite article is a bound morpheme in Swedish, which is less prone to error than the freestanding indefinite article. Definiteness per se does not appear play a role in the problematic production of articles in agrammatic speakers. Theoretical implications of this finding are discussed.