Dissertations, Theses, and Capstone Projects

Date of Degree


Document Type


Degree Name



Speech-Language-Hearing Sciences


Loraine Obler

Committee Members

Mira Goral

Martin Gitterman

Subject Categories

Discourse and Text Linguistics | Other Rehabilitation and Therapy | Psycholinguistics and Neurolinguistics | Semantics and Pragmatics | Speech and Hearing Science | Speech Pathology and Audiology


aphasia, discourse analysis, language mixing, communication, bilingual, communication success


Language-mixing (LM) as defined by Chengappa (2009, p. 417) is an “intra-sentential phenomenon referred to as the mixing of various linguistic units (morphemes, words, modifiers, phrases, etc.), primarily from two participating grammatical systems”. LM is influenced by grammatical, environmental, and social constraints (e.g., Milroy & Wei, 1995; Bhat & Chengappa, 2005). Researchers have suggested that LM in patients with aphasia is a communicative strategy used to achieve successful exchanges between speakers; the effectiveness of this mixing, however, had yet to be demonstrated quantitatively.

In the current study we investigated whether LM is present in bilingual speakers with aphasia, and if so, at which linguistic level(s) (morphological, lexical, pragmatic, and phrase) LM is found. Once these questions were addressed, we asked whether the LM patterns were typical or atypical in nature in such individuals. Finally, we investigated the differences in pertinent discourse measures (productivity, dysfluencies, coherence, and communicative success) in bilingual speakers with and without aphasia in order to assess if LM truly helps them to produce a more successful form of communication.

A total of 64 individuals – one group of 32 bilingual individuals with non-fluent aphasia and another group of 32 bilingual healthy control participants were recruited from local hospitals in Mysore, India. The study made use of two types of discourse elicitation tasks: personal narratives and picture description. Healthy control and aphasia participant groups were encouraged to mix languages in one condition. Their performance in this condition was compared to when they were constrained from mixing in Kannada-only and English-only conditions.

Investigating brain damaged and non-brain damaged bilingual speakers from the same speech community allowed for the interpretation of typical and atypical patterns of language usage. The LM patterns that were similar in both groups, hence typical in nature were direction of LM, LM at various levels, LM frequency across tasks, and LM in different word classes. We observed four atypical patterns of language-mixing in individuals with aphasia: 1) they produced a higher percentage of mixing compared to the healthy control participants; 2) they produced a higher percentage of mixing in the Kannada-only condition than the English-only condition, i.e., they did not follow the instructions provided by the examiner during the Kannada-only condition; 3) they produced Kannada matrix language utterances in the English-only condition, which is atypical in the local Kannada-English speaking community; and 4) they produced more word-level mixing during Kannada-only and language-mixed condition than the English-only condition.

It is common practice for clinicians and researchers to promote the usage of two or more languages, as they believe it enhances communication (e.g., Muñoz et al., 1999; Chengappa, et al., 2004). However, in the current study, we found that the ability to use more than one language did not lead to a more successful form of communication. We found that individuals with aphasia produced more morphemes, words, phrases, and utterances when they mixed languages, but they did not score higher on the communicative success and coherence scales compared to when they were constrained from mixing.