Perception and the nature of the phoneme

Traditional interpretations of the phoneme have viewed it either in terms of physical properties (Jones, 1944b), psychological reality (Sapir, 1933, cf. Jaeger, 1980), or function, typically to serve in opposition to other phonemes within a phonological system (Saussure, 1915, Trubetzkoy 1939, Penzl, 1971). More recently, some phonologists have questioned the value of phonemes to phonological description in a post-generative world (Goldsmith, 1999). I argue that Sapir and Trubetzkoy’s ideas about the psychological reality of phonemes and the role of contrastive oppositions in sound systems are as relevant as ever, a claim justified by research in phonetics, which has demonstrated the importance of perception in some types of phonological change (Ohala, 1993, cf. Kuhl, 1991 and Sendlmeier, 2000). This chapter considers four examples of phonological contrast and change that may have been rooted in perception. First, fortis/lenis and geminate/singleton contrasts may constitute phonological oppositions when their members are perceived as different (Lisker, 1957, cf. Penzl, 1974). Second, perceptual ambiguity may have played a role in the lack of affricates post-vocalically for old short stops in Old High German texts (Callender, 2017). Third, perception may be relevant to understanding the English Great Vowel Shift. Liberman (1995) argued that the GVS had no beginning, in that there was always some degree of allophonic variation in vowels. To extend his analysis, I argue that it is the perception of new vowels that may have triggered the shift. Finally, I suggest that perceptual salience may be responsible for the maintenance of /ai/ before voiceless consonants in southern US English, where it is often monophthongized in other phonological environments. As each of the changes discussed is rooted in the perception of new sounds, phonological oppositions and psychological reality remain relevant to our understanding of phonemes.