LINGUIST List 15.3166

Thu Nov 11 2004

Diss: Phonology: Curtis: Geminate weight: Case Studies

Editor for this issue: Naomi Fox <>

To post to LINGUIST, use our convenient web form at


        1.    Emily Curtis, Geminate weight: case studies and formal models

Message 1: Geminate weight: case studies and formal models

Date: 08-Nov-2004
From: Emily Curtis <>
Subject: Geminate weight: case studies and formal models

Institution: University of Washington
Program: Department of Linguistics
Dissertation Status: Completed
Degree Date: 2003

Author: Emily KJ Curtis

Dissertation Title: Geminate weight: case studies and formal models

Linguistic Field(s):

Dissertation Director:
Sharon L. Hargus
Ellen M Kaisse
Richard A. Wright

Dissertation Abstract:

Geminate Weight: case studies and formal models

On the common assumption that segmental length is prosodic, its
representation depends on the syllable model that is adopted, and the
cross-linguistic patterning of geminate consonants in syllable structure
and weight provides evidence for determining which model is the most
cross-linguistically predictive and explanatory model.

In this dissertation I examine the predictions of four models of geminates
and syllable weight corresponding to two basic models, a skeletal model and
a moraic model, and compare them with patterns of geminate distribution and
weight in 16 languages’ weight-sensitive phenomena (including stress,
minimality, metrics, compensatory lengthening, reduplication and language
games). Some of the data has arisen in the literature but is reanalyzed
here in the context of additional data from each language and the attempt
to understand each weight system as a whole and to clarify the implications
of the data and of the competing syllable models.

All four syllable models compared are ultimately inadequate to account for
the attested patterns. While a moraic model of syllable weight predicts
and models attested prosodic patterns most successfully, the moraic model
of geminates as inherently weight-bearing segments is an insufficient
definition of geminates because it fails to distinguish geminates from
weight-bearing singleton consonants in some cases. Because the skeletal
models fail to account for the majority of the prosodic patterns and make
failed predictions, however, the X-slot is untenable as a prosodic unit. I
argue that the representational distinction between geminates and
singletons is not one of X-slots in a skeletal or composite syllable model
(Hume et al 1997), but one of root nodes, similar to that proposed by
Selkirk (1990). This model of geminates entails a revised syllable model,
resembling the moraic and composite models, but with important
distinctions. The root-and-mora model of syllables makes predictions that
must be further examined, but some are supported by recent findings such as
the interactions of segmental phonetics and/or features and prosody
(Kavitskaya 2002, Gordon 1999, Zec 1988). The prosodic evidence supports a
root-based model of segments and segmental length which also must be
further examined with respect to segment types such as affricates and
prenasalized stops.

This investigation into the representation of geminates concerns many
subfields of phonology that also individually merit cross-linguistic
studies (minimal word constraints, stress patterns, reduplication and
weight, etc.). The representations examined are fundamental to defining
the prosodic units to be utilized in a constraint-based approach and in
defining input and output forms that are evaluated in that approach. This
study also clarifies patterns that are (and are not) crucial in comparing
weight models and provides data and data-driven analysis of the weight
systems of the 16 languages.

Key Terms: phonology, prosody, syllable weight, segment length, geminate,
mora, X-slot, skeletal model; Arabic, Cuna, Estonian, Hausa, Hindi,
Hungarian, Italian, Japanese, Korean, Koya, Latin, Leti, Malayalam, Sinhala
(Sinhalese), Yakima Sahaptin, Truk (Chuuk)

Respond to list|Read more issues|LINGUIST home page|Top of issue