LINGUIST List 11.631

Tue Mar 21 2000

Qs: Clitic-final Consonants, NLP Devices

Editor for this issue: Karen Milligan <karenlinguistlist.org>

Directory

1. Walcir & Daria, Question: assimilation of clitic-final consonants
2. Philip A. Bralich, Ph.D., Re: Question and Answer NLP Devices

I am a currently working on my dissertation that 
involves several topics in the phonology of Picard, a Gallo-Romance dialect 
spoken in Northern France. I am interested in the shape of proclitics and the 
types of assimilation processes that they undergo across languages. Initial 
observations have led me to conclude that clitic-final consonants behave in an 
unexpected manner in the grammar of many closely related and unrelated 
languages. First, clitics often have a (C)VC shape, while prefixes tend to be 
(C)V in form. Given that stems are typically consonant initial, the result of 
the concatenation of a proclitic and a following word is rather marked from the 
point of view of syllable structure: (C)VC.CV. Second, the coda-onset 
assimilation patterns observed at the clitic + stem boundary are not typical of 
those observed morpheme-internally or between an affix and a root. For example, 
the C.C sequence in procliticization (clitic + stem) is often subject to total 
assimilation, even when the clitic-final consonant is a well-formed coda in the 
language, as is the case in Picard.

Some languages where total assimilation is found 
include Picard (e.g. /Sol fEt/ = [So(f f)Et] 'the party'), Classical Arabic 
(where the assimilation process is restricted to coronal consonants only) (e.g. 
/?al nabi/ = [?a(n n)abi] 'the prophet'); and Aranese (Hualde 1992) (e.g. /et 
bi/ = [e(b b)i] 'the wine')

uestion: do you know other languages besides 
Picard, Aranese and Classical Arabic that display similar assimilation patterns 
or where proclitics and prefixes have different shapes (in terms of their CV 
structure)?

I will post a summary of the responses. Thank you 
very much.

Walcir Cardoso
wcardopo-box.mcgill.ca

I am currently trying to put together a focused and complete list of
Question and Answer dialog NLP (Natural Language Processing) systems
that exist as for sale products on the web or as stand alone
products. The list I am compiling will include products that can
be purchased or downloaded as well as much larger projects that
require a long term commitment of resources for both the
developer and the client. The list should include search engine
projects such as "Ask Jeeves" or those efforts that are
proposing even more targeted question and answer systems than
that company. Though I am pimarily interested in working, ready to go
products, promised products are also of interest. Also early
stage start ups (e.g. ThrowNet.com) will be included as well. Please
send your responses to me privately and I will post a summary to the
list. University research may also be included, but I am primarily
interested in work being done by the private sector.


Phil Bralich

Philip A. Bralich, Ph.D. 
President and CEO 
Ergo Linguistic Technologies 
2800 Woodlawn Drive, Suite 175 
Honolulu, HI 96822
Tel: (808)539-3920 
Fax: (808)539-3924