typological universal

[Steve Parker]

It is often observed that complex syllable onsets follow the Sonority
Sequencing Principle, rising in sonority toward the nucleus, such
as /plan/, /kro/, etc. This can lead to claims about universal typological
implications. For example, if a language allows liquid + obstruent
clusters, then it must also allow the unmarked sequence obstruent + liquid
as well (Greenberg). I'm looking for a language which falsifies this claim.
Specifically, a language which has well-formed onsets of the type liquid +
obstruent, but crucially does NOT exhibit any obstruent + liquid. I would
even be happy to see nasal + obstruent but not obstruent + nasal, or glide
+ obstruent but not obstruent + glide. Please note, however, that I am specifically not interested in pre-nasalized stops since these are technically just single segments, not true clusters. If it can be shown that in a word-initial /nd/ sequence the /n/ is a separate segment, then that would be helpful to me.

Thanks,

Steve Parker
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/lingtyp/attachments/20091202/036298c0/attachment.htm>