chusetts, Amherst for help and discussion. I would like to particularly acknowledge Bruce Hayes and Michael Kenstowicz for very detailed and useful comments on the manuscript, as well as Glyne Piggott for his encouragement and discussion of many previous versions of this analysis. This research was supported by SSHRCC fellowship 752-93-2773, and by SSHRCC research grant 410-98-1595, for which I am grateful. Non-uniformity in English secondary stress 239 non-uniformity, that is, explained why stress is consistently preserved in one context (1a), never preserved in another (1b) and variably preserved in yet another (2). As Prince (1993) and many others have shown, phonologically conditioned non-uniformity is an expected consequence of the ranked and minimally violable constraints of Optimality Theory, and is thus captured more straightforwardly than in theories that do not allow for constraint violability (see especially McCarthy & Prince 1993a, 1994a, Prince & Smolensky 1993 and McCarthy 2000). In Optimality Theory, a constraint is violated in a particular environment because its satisfaction would conflict with the requirements of a higher-ranked constraint. If in another context the higher-ranked constraint makes no conflicting demands, the lower-ranked one is obeyed. In frameworks that do not allow for violability, non-uniformity is handled in a variety of ways, but accounting for it usually requires a loss of generality in the formulation of constraints or rules. 3 Given the ease with which Optimality Theory generally deals with nonuniform constraint application, it is perhaps not too surprising that constraint ranking would help to provide insight into the complexities of the English secondary stress system. 4 §1 of this paper employs ranked constraints to account for the regular patterns of pretonic stress, in which syllable weight non-uniformly determines the distribution of secondary stress. The analysis of stem stress preservation and monomorphemic exceptions in §2 incorporates prosodic faithfulness constraints into the hierarchy established in §1. The result is an account that relies largely on independently motivated constraints and rankings to explain the nonuniform distribution of stem-based stress and of underlyingly specified exceptional stress. Lexically conditioned non-uniformity requires elaboration of the basic theory of Prince & Smolensky (1993), which provides no means of relativising the activity of a constraint to a particular set of lexical items. To handle the lexical idiosyncrasy of stem stress preservation, I propose in §2.4 that prosodic faithfulness constraints can be multiply instantiated in the constraint hierarchy : in a general version and in a lexically indexed version. This allows stress preservation to optionally overcome some, but not all, constraints. The resulting analysis captures the non-uniformity of stress preservation outlined above. This approach to lexical non-unisyllables, words of this type vary as to whether the rhotic of the stem is preserved or not. Assuming that ...

Harmonic Grammar (HG) and Optimality Theory (OT) are closely related formal frameworks for the study of language. In both, the structure of a given language is determined by the relative strengths of a set of constraints. They differ in how these strengths are represented: as numerical weights (HG) or as ranks (OT). Weighted constraints have advantages for the construction of accounts of language learning and other cognitive processes, partly because they allow for the adaptation of connectionist and statistical models. HG has been little studied in generative linguistics, however, largely due to influential claims that weighted constraints make incorrect predictions about the typology of natural languages, predictions that are not shared by the more popular OT. This paper makes the case that HG is in fact a promising framework for typological research, and reviews and extends the existing arguments for weighted over ranked constraints.

This articl-, examines some consequences of optimality theoretic constraint ranking and violability for the study of phonological deyelopment. It is well known that the phonetic shape of child utterances is subject to strict restrictions. When these restrictions are captured in terms of minimally violable, rather than inviolable, constraints, the effects of a constraint that is overcome in development can continue to persist through successive developmental stages and into the mature system. The empirical base for this study is provided by previously unpublished data from a longitudinal corpus of phonetically transcribed speech from four English-learning children (Compton and Streeter (1977)). Evidence for this view of development as constraint reranking is found in a comparison of the prosodic structure of child and adult English and in a developmental change in the application of consonant harmony: minimal violation in the child system itself occurs in the patterns of content preservation displayed by truncations.

Stager and Werker (1997) show that fourteen-month-olds engaged in a word-learning task fail to respond to a switch between the minimal pair [bI] and [dI], though they do respond to a switch between [lIf] and [nim] in the same task. In this article we show that the [bI]/[dI] results extend to stimuli that respect English phonotactics ([bIn] vs. [dIn]), to a voicing contrast ([pIn] vs. [bIn]), and to voicing and place combined ([pIn] vs. [dIn]). Our interpretation of these results is that when a phonological contrast like place or voicing is first acquired, it remains only partially integrated and can be lost under the processing demands of word learning. We formalize partial integration in terms of unranked optimality-theoretic constraints and discuss the predictions of this account for further research.

Walker, and the participants in LING 751, UMass, and the Rutgers/UMass Joint Class Meeting, Spring 1995, for useful suggestions, and to Heather Goad and Glyne Piggott for their indispensible guidance. I would also like to acknowledge Choirul Djamhari for his help with the Indonesian data, and Lisa Travis for making it possible for me to work with him, as well as Jan Voskuil for his hospitality and assistance in securing many of the Austronesian materials during an all too short visit to Leiden.

No abstract

The publisher of this book does not permit the archiving in RUcore of this or any other version of the chapter. See the full description for a link to the book.

Artificial analogues of natural-language phonological patterns can often be learned in the lab from small amounts of training or exposure. The difficulty of a featurallydefined pattern has been hypothesized to be affected by two main factors, its formal structure (the abstract logical relationships between the defining features) and its phonetic substance (the concrete phonetic interpretation of the pattern). This paper, the second of a two-part series, reviews the experimental literature on phonetic substance, which is hypothesized to facilitate the acquisition of phonological patterns that resemble naturally-occurring phonetic patterns. The effects of phonetic substance on pattern learning turn out to be elusive and unreliable in comparison with the robust effects of formal complexity (reviewed in Part I). If natural-language acquisition is guided by the same inductive biases as are found in the lab, these results support a theory in which inductive bias shapes only the form, and channel bias shapes the content, of the sound patterns of the worlds languages.