ABSTRACT This dissertation research project is a cross-linguistic and cross-model study comparing the natural gestural-visual languages of the deaf and the oral languages within which these sign languages are found for the purpose of investigating questions about linguistic universals and human cognition in general, as well as language contact phenomena, particularly with respect to rhetorical practices and communicative conventions. The research focuses on American Sign Language (ASL) and American English, Brazilian Cities Sign Language (BCSL), a language of the urban Brazilian deaf, and Brazilian Portuguese, and Urubu-Kaapor Sign Language (UKSL) and Kaapor, these latter two being languages of an American Indian group native to the Amazonian region of Brazil. Comparative work of this kind can provide crucial evidence for distinguishing constraints imposed by the communication medium, both vocal-auditory and gestural-visual, from linguistic universals. Scholars have tended to approach the analysis of empirical linguistic data as if there were a binary division between universality and relativity, without taking into account the impact of communicative modality. The inclusion of sign language data is crucial not only to the universality versus relativity debate but also to the understanding of human cognition to the extent that such understanding relies on linguistic evidence.
