Evaluating stress “deafness” in Amharic vs. English speakers

Abstract

Speakers of languages that lack contrastive word stress, such as French, are observed to display stress “deafness”, a relative difficulty with learning nonce prosodic contrasts for the purpose of memory-based perception tests. We assessed stress “deafness” in Amharic, a South Semitic language of Ethiopia, to clarify the status of word stress in that language. Various descriptions have been provided for Amharic prosody, including that the language may lack word stress. We found that 22 Amharic and 22 English speakers did not significantly differ in accuracy in a sequence recall test for a nonce segmental contrast (múku vs. múnu), but Amharic speakers performed relatively worse on a nonce prosodic contrast (númi vs. numí); this behavior supports the claim that Amharic lacks word stress. In addition, Amharic speakers who reported higher proficiency in English were more accurate in the prosodic test and more likely to pass a prosodic training, suggesting gradient degrees of stress “deafness” based on experience with a language with contrastive stress.