Word class effects on representation and processing in non-brain damaged speakers and people with aphasia

This thesis focuses on processing of proper nouns (e.g., Hugh Grant) and common nouns (e.g., actor). Empirical studies have shown poorer retrieval and slower learning for proper nouns in comparison to common nouns in non-brain-damaged individuals, and poorer retrieval for proper nouns relative to common nouns in people with aphasia. However, no censensus has been reached regarding the underlying cause(s) of these discrepancies. Differences in production of proper and common nouns have generally been attributed to the existence of separate processing mechanisms for these two noun classes. However, different statistical properties (such as frequency) and the effects they may have on the retrieval of proper relative to common nouns have also been discussed. A stumbling block for making any firm conclusions regarding differences in proper versus common noun processing based on previous experimental designs is that matching proper and common nouns on statistical properties is very hard, if not impossible. This thesis attempted to address this methodological problem with five experiments to inform our understanding of word class effects (proper versus common nouns) on representation and processing. Chapters Two, Three and Four explore learning mechanisms for novel proper and common nouns in younger and older non-brain-damaged speakers as well as in people with aphasia. By using novel word forms paired with novel meanings, we held statistical properties of words equal. Therefore, the design allowed us to more directly address the possibility that proper and common nouns are processed by two separate mechanisms. Younger and older individuals showed the same pattern in retrieval: Proper nouns were not more vulnerable than common nouns. In learning, proper nouns were, however, more demanding (possibly due fine visual discrimination required). People with aphasia did not demonstrate greater impairment on newly learned proper nouns than on common nouns relative to their age-matched controls. Chapter Five focuses on proper versus common noun production in category fluency tasks in non-brain-damaged speakers as well as in people with aphasia. Category fluency tasks were argued to reduce the influence of frequency on word retrieval and thus, represent a more objective tool (in comparison to picture naming, for example) to assess differences in proper versus common noun retrieval. The study found no evidence for proper nouns being harder to retrieve. Finally, in Chapter Six, repetition of a word in the presence of a picture was used to facilitate naming of proper and common nouns in people with aphasia. Proper and common nouns were found to benefit equally from facilitation. Overall, convergent evidence from the studies in this thesis shows that when statistical properties are held equal or affect performance on the task less, proper nouns are no longer more vulnerable in retrieval than common nouns. Consequently, statistical properties of proper and common nouns should be given more consideration in accounting for retrieval patterns and should be incorporated into explanatory theoretical models.