-------------------------------------------------------
The current research uses representative linguistic structures to test the
viability of our approach, applied to Southern British English. The three
focal areas of research are intonation, morphological structure, and
systematic segmental variation.
Our control structure takes the form of a linguistically informed,
head-driven prosodic hierarchy, implemented in XML. Database text material
is given a detailed syntactic parse, supplemented by a parse into the
prosodic and sub-syllabic constituents of the hierarchy, with links to the
syntactic tree. The lexicon itself is in the form of a partially parsed
representation.
Timing relations in terms of temporal compression and overlap are treated
as the phonetic interpretation of phonological representations. Necessary
linguistic information includes a grammar of syllable and word joins,
using various types of structure sharing (ambisyllabicity) and an
appropriate feature system.
The temporal extent of systematic spectral variation due to coarticulatory
processes is measured in segmental sequences that differ in whether
intervening segments are likely to block the spread of coarticulatory
effects, and in prosodic domains differing in structure and length.
----------------------------------------------------------------------
I'm not sure whether I'm being overly picky, but the introductory words
(taken from our Summary on the website etc.) say "intonation,
morphological structure, and systematic segmental variation", while the
paragraphs afterwards don't (apart from Sarah's!) mention these things.
Instead they are about linguistic structures and timing relations. Would
you like me to revise the last sentence of the first paragraph quoted here
so that it reflects more accurately what follows, or would you prefer us
to write something that addresses the issues raised in that sentence?
Alex says he'd rather not have his name on the abstract. It seems to me,
he was there at the planning stage and has worked on ProSynth as much as
anyone, so like Jill, I think his name ought to be on it.
PS: We're submitting this (so far) as a poster abstract. Is that what you
want?
Richard