[etnolinguistica] FW: Lenition in Northwest Ge

[Dan Everett]

Andres,

Very useful stuff in the answer. People do seem to be confused by the
claims of Structure Preservation, since there are so many ways around
it, as you mention in the reference to pre-compiled allomorphy. And
Ellen Kaisse has some work on post-lexical processes that suggest the
possibility of SP, though nothing quite like the Suya case.

One question: why do you think that fortition would solve the problem of
/k/?

Of course, there are many potential phonetic explanations for the Suya
facts, but those facts to show that the purely architectonic solution of
Structure Preservation is problematic for traditional Lexical Phonology.

In recent work, Paul Kiparsky has argued for re-ranking Optimality
Theoretic constraints according to morphological status, e.g. different
rankings for root, stem, and word, etc. Let us say that the Suya rule of
Lenition is the Context Sensitive rule and that there is a Context Free
rule of feature markedness for Suya of the type: *[dorsal,continuant].
Then the ranking would be, throughout the Suya phonology: CF >> CS.
Typical Structure Preservation would be CF >> CS in the lexicon but CS
>> CF everywhere else. This proposal of Kiparksy's allows OT to capture
Structure Preservation (and opacity) without appeal to things such as
'Sympathy Theory'. It does not, however, account for the rarity of the
Suya process. On the other hand, rarity is not a very good argument for
anything, as Peter Ladefoged and I argued in our 1996 Language article
'The problem of phonetic rarities'.

If this discussion is a bit too opaque for readers who have not seen the
facts, let me know and I can post an abstract of the Suya research to
this list.

Once again, Andres, thanks very much for your remarks, which quite
useful indeed.

-- Dan

P.S. Do discussions on this list normally take place in English? I could
have responded in Portuguese, but am not sure of what normally takes
place on this list. 


.........................
Dan Everett
Professor of Phonetics and Phonology
Department of Linguistics
Arts Building
University of Manchester
Oxford Road
M13 9PL
Manchester, UK
dan.everett at man.ac.uk 
Phone: 44-161-275-3158
Dept. Fax and Phone: 44-161-275-3187
http://lings.ln.man.ac.uk/info/staff/de

-----Original Message-----
From: Andres P Salanova [mailto:kaitire at MIT.EDU] 
Sent: Wednesday, April 02, 2003 2:27 AM
To: Daniel Everett
Cc: etnolinguistica at yahoogrupos.com.br
Subject: Re: Lenition in Northwest Ge


Hello Dan:

Your message was forwarded to me, and I'm taking the liberty of
forwarding my answer to the etnolinguistica list, since I think this
discussion will be of interest there.

It makes me happy that someone is working on the phonology of Suyá, as
I've always had great curiosity for it. Please make your paper available
whenever it's ready.

You ask whether there's anything special about /k/ in the northern Je
languages. A couple of things come to mind. One thing that singles out
/k/ from other voiceless stops is that in coda position it is not
opposed to a homorganic nasal in at least Mebengokre and Apinaye.

Another point about /k/ is made by Burgess & Ham in their 1968 article
about Apinaye. If I remember well, they say that it's inserted for
emphasis in words like ngryk, tyk, etc., rather than being part of the
lexeme. I don't know how valid an interpretation this is, but it is true
that /k/ in Apinaye has a peculiar behavior: whereas other coda
consonants are deleted before a homorganic stop in the following onset,
/k/ is deleted everywhere except before /r/.

About other processes of lenition, there's voicing of obstruents in
unstressed syllables in Apinaye, but nothing very similar to what you
describe.

One question I would like to ask is what the nature of the evidence is
for believing that the process in question is lenition rather than
fortition. If this were fortition, it would be quite unproblematic,
right?

The facts are in fact quite striking, and I'm also surprised that so
many people report that things of the sort haven't been encountered
before. The consonantal mutations in languages such as Mende (which are
structure-preserving) were, at least for a while, _believed_ to be
post-lexical phonology (i.e., they were thought of as phonological
processes that applied at the edges of certain prosodic domains). I
don't know if this sort of thing is abundant, but in my mind the (ugly)
device of "precompiled allomorphy" was precisely to address that kind of
problem.

Best,
	Andres


There is a fascinating process of phrase-final lenition in Suya in which
t --> r and p --> w, both then followed by what Marymarcia calls, I
think, and 'echo vowel', a copy of the vowel immediately preceding the
lenited consonant. The special interest of the process is that k fails
to undergo it. Occasionally a phrase-final k will be marked by a
retarded voice onset time or even be pronounced as g. However, it
generally just surfaces as k in my data. Because r and w are phonemes of
Suya, according to my analysis, but Suya lacks any velar approximant
phoneme, this is a case of what Lexical Phonology (remember all those
years ago?) called Structure Preservation. However, it seems to be the
first recorded case of non-lexical Structure Preservation, thus
representing a severe problem for Lexical Phonology-type theories [...]

[...] (i) what the
Suya /k/ corresponds to in other Northwest Ge languages and (ii) whether
Lenition of this sort (phrase final on voiceless plosives) is found in
these languages. I *think* that there is nothing unusual about /k/ in
Northwest Ge, but am not sure. I do not think that other Ge languages
have this kind of Lenition either, but I could not assert categorically
that they do not.





Visite a página do grupo Etnolingüística, no endereço seguinte:
http://br.groups.yahoo.com/group/etnolinguistica/.
Submeta textos para discussão, contribua com a "Biblioteca Virtual", sugira contatos que venham a ser do interesse dos membros do grupo.  Enfim, participe!
Para cancelar sua assinatura deste grupo, envie um e-mail para:  
etnolinguistica-unsubscribe at yahoogrupos.com.br

 

Seu uso do Yahoo! Grupos é sujeito às regras descritas em: http://br.yahoo.com/info/utos.html