Project Details
Projekt
The Prosody of Derived Words in English
Applicant
Professorin Dr. Sabine Arndt-Lappe
Subject Area
General and Comparative Linguistics, Experimental Linguistics, Typology, Non-European Languages
Term
from 2015 to 2020
Project identifier
Deutsche Forschungsgemeinschaft (DFG) - Project number 273681275
Phenomena concerning the prosodic structure of derived words play a central role in modern theories of phonology-morphology interaction. The most wellknown effects are, on the one hand, effects of stress preservation (as e.g. in orìginálity, where the primary stress of the base original survives as a secondary stress in the derivative) and, on the other hand, morphologically determined stress placement, where the position of main stress in derived words is determined by the morphological category. As an example consider again orìginálity, where the suffix -ity determines the location of main stress on the antepenultimate syllable. All -ity nouns uniformly have antepenultimate stress (compare, e.g., pröductivity, monstrósity, etc.).The existence of both stress-shifting and stress-preserving effects is commonly taken to support the idea of interleaving phonological and morphological modules in a stratified lexicon. The existence of stress preservation effects between bases and derivatives has been used as evidence for lexical marking of stress and grammatically relevant paradigmatic correspondence relations. The crucial underlying assumption of most pertinent theoretical work has been that stress rules are categorical, and that variation can only arise as random or lexicalised exceptions. This view, however, has been challenged in recent work, which has brought to light evidence suggesting that both the extent and the systematicity of the variation has hitherto been considerably underestimated (cf. esp. Collie 2008 for stress preservation, Bauer et al. 2013 for stress shift).Furthermore, it is suggested that the systematicity of the variation is codetermined by lexical factors such as frequency and neighbourhood.This project will investigate the extent and systematicity of that variation, on the basis of large quantities of adequate data obtained from production experiments. The variation facts will then be modeled in terms of stratified theories of the lexicon and in terms of competing frameworks to test the empirical adequacy of these models, and if necessary develop a more adequate model. ln particular, we will compare the performance of strata! models with computational models in which phonological generalisation and phonological variation are emergent from properties of lexical organisation, i.e. Analogical Modeling (Skousen & Stanford 2007) and Naïve Discriminative Learning (Baayen et al. 2011).
DFG Programme
Research Units
