<OT> New Posting: ROA-904

[roa at ruccs.rutgers.edu]

ROA 904-0307

A Landmark Underspecification Account of the Patterning of Glottal Stop

Marianne Borroff <mlborroff at gmail.com>

Direct link: http://roa.rutgers.edu/view.php3?roa=904


Abstract:

This dissertation addresses the asymmetry in patterning
between laryngeal and supralaryngeal consonants. In this
dissertation, I consider four patterns: (1) required identity-acr
oss-glottals (in V1?V2, V1 = V2); (2) hiatus resolution-like
processes in V?V(V?V patterns with VV); (3) prohibition
of glottal stop from syllable onset or coda; and (4) temporal
instability of sequences with glottal stop (e.g. vowel intrusion:
Vx?C -> Vx?VxC,coalescence of C? to C'). I present a unified
analysis of these patterns within the frameworks of Articulatory
Phonology (Browman and Goldstein 1986, et seq.) and Optimality
Theory (Prince and Smolensky 1993), in which utterances
are comprised of abstract articulatory gestures (rather
than segments or features).

I argue that the exceptional behavior of glottal stop is
a function of its acoustic properties: in contrast to oral
stop consonants, glottal stop does not condition formant
transitions, and therefore lacks the landmarks of ONSET
(marking the beginning of the gesture) and OFFSET (marking
the end of the gesture. Based on data on temporal relations
within syllables and sequences (e.g. in Browman and Goldstein
2000), I propose that the ONSET and OFFSET landmarks are
points of alignment for phasing relations that underlie
syllabification and sequentiality. Because it lacks these
crucial landmarks (the Landmark Underspecification proposal),
glottal stop cannot participate unambiguously in syllabic
or sequential phasing relations.

This approach provides an account of each of the patterns
described above. Hiatus resolution-across-glottals occurs
because the glottal stop cannot satisfy the constraint that
requires syllable onsets to be precisely phased with respect
to the following vowel; glottal stop is therefore not a
satisfactory syllable onset. Languages in which the vowels
flanking laryngeal consonants are required to be identical
exhibit a subset case of the hiatus resolution pattern,
differing only in the strategies employed to repair hiatus.
Similarly, glottal stop is disallowed pre- or postvocalically
in some languages because it cannot obey the constraints
on phasing of onset or coda consonants with respect to syllable
nuclei. Finally, the lack of clear cues to the temporal
position of glottal stop lead underlying sequences with
glottal stop to surface non-sequentially, manifested as
vowel intrusion or as coalescence of the glottal stop with
a vowel or consonant.


Among the languages discussed are Yatzachi Zapotec (Otomanguean),
Yucatec Maya (Mayan), Kekchi (Mayan), Arbore (Cushitic),
Tukang Besi (Malayo-Polynesian), Kashaya (Pomoan) and Yurok
(Algic).

Comments: 
Keywords: laryngeal consonants, glottal stop, transparency, hiatus resolution, Articulatory Phonology in OT
Areas: Phonology, Phonetics
Type: PhD Dissertation

Direct link: http://roa.rutgers.edu/view.php3?roa=904