|Title:||Morphological haplology in a constraint-based morpho-phonology|
|Abstract:||Morphological haplology in a constraint-based morpho-phonology
The paper is published in
Wolfgang Kehrein and Richard Wiese (eds.) 1998. Phonology and Morphology
of the Germanic Languages, 199-215. Tübingen: Niemeyer.
Morphological haplology has been defined as the absence of "an affix or
clitic [...] when the adjacent part of the stem is homophonous to it"
(Stemberger 1981:791), or as the avoidance of adjacent identical morphs
(e.g. Menn/McWhinney 1984). Although haplology is dealt with in a large
number of publications, a satisfactory general account of this
phenomenon is still lacking. Thus, it is often difficult to arrive at
firm generalizations because identity avoidance, though wide-spread in
the world's languages, seems extremely variable not only across
languages, but also within one language.
Part of the problem lies in the conceptualization of the phenomenon
itself. The above-mentioned definitions of haplology indicate that
identity avoidance may affect repeated morphs, morphemes that are only
partially identical, and even non-morphemic material. It is unclear
whether these phenomena should indeed all be subsumed under the label
of haplology. Furthermore, the jury is still out on at least three
other questions, the first being whether haplology can be defined in
universal terms, or only as peculiar to individual morphological
processes. The second question is whether haplology should be accounted
for by rules or by output-oriented restrictions. Finally, it remains to
be shown whether haplology can be regarded as a purely phonological
phenomenon or not.
Starting from earlier constraint-based approaches to haplology (e.g.
Yip 1998) this paper proposes that morphological haplology results from
a family of universal, violable output-oriented constraints on the
repetition of identical phonological elements (OCP (feature), OCP
(segment), OCP (onset), OCP (nucleus), OCP (onset, nucleus)). This
proposal differs from earlier constraint-based models (e.g. Yip 1998)
in important respects. It eliminates constraints that make reference to
morphological instead of phonological structure, i.e. Yip's OCP (affix)
and OCP (stem), and adds subsyllabic constituents to the inventory of
elements to which OCP constraints can refer. This has two main
consequences. First, in this approach, morphological haplology is
exclusively triggered by phonological constraints (in interaction with
other prosodic and morphological constrtaints), which is both
empirically and theoretically preferable. Second, by eliminating OCP
(affix) and OCP (stem) as constraints, I argue against the Repeated
Morph Constraint. It is shown that both the Repeated Morph Constraint
and Stem-End-Haplology are different consequences of the same
In the account I propose, the similarities of haplology phenomena
across and within languages are explained by the universality of the
phonological constraints at work. At the same time, the observable
variability of haplology among different morphological categories in
one language and across languages turns out to be the consequence of
language-specific rankings of the OCP constraints. The interaction of
OCP constraints with other prosodic and morphological constraints can
also explain the apparent variability of haplology among derivatives of
the same morphological category (cf., for example strychnin-ize vs.
femin-ize vs. *feminin-ize).
These points are illustrated with data from English (-ize derivatives),
German (female person nouns in -in) and Dutch (person nouns and
comparatives in -er), for which a unitary account has not yet been