Stochastic phonological knowledge: the case of Hungarian vowel harmony

Abstract

In Hungarian, stems ending in a back vowel plus one or more neutral vowels show unusual behaviour: for such stems, the otherwise general process of vowel harmony is lexically idiosyncratic. Particular stems can take front suffixes, take back suffixes or vacillate. Yet at a statistical level, the patterning among these stems is lawful: in the aggregate, they obey principles that relate the propensity to take back or front harmony to the height of the rightmost vowel and to the number of neutral vowels. We argue that this patterned statistical variation in the Hungarian lexicon is internalised by native speakers. Our evidence is that they replicate the pattern when they are asked to apply harmony to novel stems in a ‘wug’ test (Berko 1958). Our test results match quantitative data about the Hungarian lexicon, gathered with an automated Web search. We model the speakers' knowledge and intuitions with a grammar based on the dual listing/generation model of Zuraw (2000), then show how the constraint rankings of this grammar can be learned by algorithm.

Japanese loanword devoicing revisited: a rating study

“…The dispreference against devoicing of word-initial singletons may have its roots in the psycholinguistic prominence of initial positions (Hawkins & Cutler, 1988;Horowitz et al, 1969Horowitz et al, , 1968Nooteboom, 1981); since word-initial position plays an important role in lexical access, speakers disprefer changing segments in this position (Beckman, 1997;Kawahara & Shinohara, 2010). 9 More generally, the results show that, in line with other recent studies, grammatical intuitions are gradient (e.g., Chomsky 1965;Cohn 2006;Coetzee 2008Coetzee , 2009Coleman & Pierrehumbert 1997;Fanselow et al 2006;Frisch et al 2000Frisch et al , 2004Greenberg & Jenkins 1964;Hay et al 2003;Hayes 2000Hayes , 2009Hayes & Londe 2006;Myers 2009;Pierrehumbert 2001;Schütze 1996;Zuraw 2000) in the following two senses. First, it is not the case that some devoicing patterns are grammatical and some other devoicing patterns are ungrammatical.…”

“…This study is inspired and motivated by an increasing interest in testing the quality of linguistic data using experimental methodology (Berko 1958;Cowart 1997;Dabrowska 2010;Hayes & Londe 2006;Kawahara 2011;Myers 2009;Nolan 1992;Schütze 1996, among others). Here I briefly summarize why experimentation is necessary beyond intuition-based data collection (see the work cited and references cited therein for more general and elaborate discussion of the following points).…”

“…The effects of lexical frequencies on phonological patterns have been receiving an increasing interest in recent phonological and psycholinguistic studies (see e.g. Bybee 1999Bybee , 2001Coleman & Pierrehumbert 1997;Ernestus & Baayen 2003;Frisch et al 2000;Hay et al 2003;Hayes 2009;Hayes & Londe 2006;Zuraw 2009). For example, in Usage-based Phonology (Bybee, 1999(Bybee, , 2001), frequency is one major factor that governs and shapes phonological regularity (see also Coleman & Pierrehumbert 1997;Frisch et al 2000;Hay et al 2003).…”

Aspects of Japanese loanword devoicing

“…This question is in part motivated by a growing body of interests in to what extent lexical frequency affects phonological regularity (for different proposals on this issue, see e.g. Boersma and Hayes 2001;Bybee 1999Bybee , 2001Coetzee 2009b;Coetzee and Kawahara 2010;Coleman and Pierrehumbert 1997;Frisch et al 2000;Hay et al 2003;Hayes and Londe 2006).…”

Surface Correspondence and Discrete Harmony Triggers

“…Hayes & Londe 2006, Wayment 2009, Zymet 2014. A question is whether IDENT-XX[αG](F) constraints with evaluation over all segments belonging to the [αG] surface correspondence set could provide an account of such effects if refinements were developed to bias for local interactions.…”