13.07.2015 Views

Untitled

Untitled

Untitled

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

122 Bogus clusters, syllabic consonants and vowel syncope. . .it is only the leftmost nuclear position that behaves oddly and that thenumber of such initial clusters is relatively small (25 items). Despite thefact that in the Government Phonology literature such nuclei have acquireda peculiar status in that they resist being properly governed (Charette1991, Yoshida 1990), 13 Brockhaus (1995) resorts to this verymechanism in order to explain the forms in (14). One immediate problemwith this solution, however, is the fact that such sequences invariably consistof ‘TR’ clusters. Thus, even if we separate the initial clusters in (14)with a properly governed nucleus, the question remains why they are alwaysof the ‘TR’ type. 14 To put it differently, there is nothing in the natureof Proper Government that would impose the restriction on the segmenttypes surrounding the properly governed empty nucleus, hence we shouldexpect various combinatorial possibilities and not only ‘TR’ clusters. Finally,it is worth mentioning that Löhken (1995), tracing the developmentof consonant clusters such as those in (11) above from OHG to NHG, arrivesat similar conclusions concerning the key role played by sonorants inwhat we call here bogus clusters. She claims that the latter clusters appearas a result of vowel syncope which takes place only before [l] or [n]. Specifically,the syncope process depends on the context in that it did not takeplace where the vowel was sandwiched between two obstruents or twosonorants.It must be clarified here that in the end Brockhaus (1995) adoptsan alternative solution, namely, she suggests that bogus clusters contractthe Interonset Government relation (Gussmann and Kaye1993). Despite the fact that this mechanism is controversial, InteronsetGovernment, according to Brockhaus (1995:209) holds a promise ofshedding new light on the rigid order of consonants in bogus clusters. 15Finally, while discussing Brockhaus’s (1995) analysis, it is worthmentioning that NSG speakers do not apply FOD to obstruents in certaincontext, e.g. Rudrer [ ® ru:drŒ] instead of expected *[ ® ru:trŒ]. Recall thatbefore other sonorants the obstruents in this dialect do undergo devoic-13Recently Cyran (2003) has made a similar observation. He notes that Englishmysteriously lacks the word-initial consonant clusters separated by the empty nucleus.14The key role of sonorants in such clusters was indicated by Rubach’s (1990)analysis of FOD.15Interestingly enough, if we accepted the idea that bogus clusters are not separatedby an empty governed nuclear position, Proper Government would become superfluousin the GP analysis of German and English. The idea that Proper Government doesnot exist in Germanic languages has been proposed by Kaye (cited by Brockhaus(1995:210) as a personal contact). This idea has been recently pushed to its extremepoint by Cyran (2003) who proposes to get rid of Proper Government altogether in thetheoretical model he develops.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!