Now showing 1 - 3 of 3
ItemInvestigating universals of sound change: the effect of vowel height and duration on the development of distinctive nasalizationHAJEK, JOHN ; Maeda, Shinji (Cambridge University Press, 2000)It is widely assumed that the development of vowel nasalization is conditioned by vowel height. Most commonly it is thought that low vowels are preferentially nasalized. However, there is conflicting cross-linguistic evidence of low vowels in some languages and high vowels in other languages being subject to preferential nasalization. Experimental evidence is also found to provide similarly conflicting results. These differences can be accounted for by different vowel duration effects: longer vowels are more likely to be perceived as nasal. Where low vowels are longer, they will be preferentially nasalized, where they are not longer than higher vowels, the latter will be preferentially nasalized.
ItemTaba and Roma: clusters and geminates in two Austronesian languagesHAJEK, J. ; Bowden, J. ( 1999)Some lesser-known Austronesian languages, such as Taba and Roma, with relatively simple phonological inventories, nevertheless have very complex phonotactic structures. The presence of a wide range of typologically unusual cluster combinations in word-initial position has important implications for generally accepted notions about Austronesian languages and about segment sequencing and the sonority hierarchy.
ItemUniversals of nasal attritionConnell, B. ; HAJEK, J. ( 1991)The claim that there is a hierarchy governing the attrition of nasals according to place of articulation is put to test in this paper by examination of cross linguistic data from two language groups which are unrelated genetically and geographically: the Romance dialects of Northern Italy and the Lower Cross group of South-Eastern Nigeria. Results of this new survey provide interesting food for thought: developments in the Northern Italian dialects support, to a large extent, predictions that follow from phonetic considerations. However, the Lower Cross languages at first appear to contradict expectations. This suggests that other factors may need to be taken account of, before a true universal tendency, if one exists, can be established.