Acoustic-Based and Knowledge-Based Processing of Mandarin Tones by Native and Non-native Speakers

Abstract

A fundamental issue in spoken language comprehension is how listeners process the acoustic signal to retrieve intended linguistic representations. This issue is discussed by reviewing selected studies on acoustic-based and knowledge-based processing of lexical tone in speech perception and spoken word recognition. Research on acoustic-based processing suggests that native listeners are able to use phonetic knowledge to compensate for compromised F0 information, whereas non-native listeners rely primarily on syllable-internal, canonical F0 information for tone identification. Research on knowledge-based processing shows that native listeners effectively track information, such as a syllable-tone’s lexical status, the probability of syllable–tone co-occurrences, morpheme and word frequency, and the density of homophonous syllable–tone neighborhoods. Non-native listeners also show evidence of knowledge-based tone processing, although the difference between native and non-native listeners remains to be explored.