14th International Congress of Phonetic Sciences (ICPhS-14)
San Francisco, CA, USA
Frequently, an earlier historical stage of a language where a single phoneme has two major allophones yields to a later stage where two phonemes have one major allophone each. Here, schematically, a phonological system where /X/ = (has the realization(s)) [x] and [y] is replaced by a system where /X/ = [x] and /Y/ = [y]. Treatments of such "phonemicization/phonologization of (former) allophones" tend to exaggerate the importance of distributional factors and underestimate the role of phonetic (dis)similarity, psycholinguistic aspects of categorization, and social considerations. Hence accounts of phonologization require a greater focus on two different sorts of exaggeration: the trend for each successive generation of younger speakers to set itself off from older ones by using phonetically more extreme values of existing phonological variables, and the tendency for such variants to originate, in the first place, via the hyper- and hypo-correct processing and (re)production of perceived speech.
Bibliographic reference. Janda, Richard D. (1999): "Accounts of phonemic split have been greatly exaggerated - but not enough", In ICPhS-14, 329-332.