<OT> New Posting: ROA-1017

[roa at ruccs.rutgers.edu]

ROA 1017-0309

On the Sociolinguistic Application of the Paradigm Uniformity Model

Andrew Kostakis <akostaki at indiana.edu>

Direct link: http://roa.rutgers.edu/view.php3?roa=1017


Abstract:
Traditional Optimality Theory (Prince and Smolensky 1993)
has often been an effective tool for capturing the internal
factors of language (phonology, morphology, etc.).  However,
the field of sociolinguistics has demonstrated that external
factors (gender, age, style, register, and social class)
are almost always involved in linguistic behavior as well.
Because traditional OT is not adept at integrating external
factors into the model, describing variation in OT remains
a challenge.


In this paper I propose Vestige Theory.  What Vestige Theory
claims, is that language change is always the result of
constraint demotion and that the demotion of a constraint
leaves behind a vestige of itself.  I argue that variation
can be defined in OT as the extent to which different demographic
s make use of the vestiges of old constraints.


Vestige constraints are kinds of output-output constraints.
They therefore differ from the “true” constraints that are
demoted as part of language change.  As an output-output
constraint, a vestige constraint has the link to the outside
world (i.e. to external factors) that all previous methods
of describing variation in OT (McCarthy 1993, Antilla and
Cho 1998, Boersma and Hayes 2001, etc.) have failed to incorporat
e: the use of tableaux without some kind of reference to
output in the environment surrounding the individual will
always fail to capture the principled heterogeneity (Weinreich
et al. 1968) that is characteristic of all the world’s languages.


The kind of output-output model that is employed in this
paper is derived from the Base-Identity model used in explaining
phenomena of Paradigm Uniformity (cf. Downing et al. 2005
and sources therein). I argue that an individual is a cell
of a social paradigm and that the norms of that social paradigm
cause an individual to produce contextually appropriate
outputs that are not surface true.  This is analogous to
the influence that a base has on a grammatical paradigm.


Social forces regularly generate resistance to sound changes.
This resistance causes an antiquated form to be preferred
by one demographic and avoided by another.  For example,
Labov (2001: 319) has observed, “Almost all the sound changes
in progress studied here and elsewhere involve gender differentia
tion,” with females leading over men in the use of innovative
forms (Labov 2001: 366-67; Cheshire 2002: 425-26).  I argue
that the details of such facts are not incompatible with
phonological theory, because output-output models offer
the crucial window to the outside world that is missing
from earlier models of variation in OT.

Comments: 
Keywords: Sound Change, Variation, Constraint demotion
Areas: Phonology,Historical Linguistics,Sociolinguistics
Type: Masters Dissertation

Direct link: http://roa.rutgers.edu/view.php3?roa=1017