The study of phonotactics (e.g., the ability of English speakers to distinguish possible words like blick from impossible words like *bnick) is a central topic in phonology. We propose a theory of phonotactic grammars and a learning algorithm that constructs such grammars from positive evidence. Our grammars consist of constraints that are assigned numerical weights according to the principle of maximum entropy. Possible words are assessed by these grammars based on the weighted sum of their constraint violations. The learning algorithm yields grammars that can capture both categorical and gradient phonotactic patterns. The algorithm is not provided with any constraints in advance, but uses its own resources to form constraints and weight them. A baseline model, in which Universal Grammar is reduced to a feature set and an SPE-style constraint format, suffices to learn many phonotactic phenomena. In order to learn nonlocal phenomena such as stress and vowel harmony, it is necessary to augment the model with autosegmental tiers and metrical grids. Our results thus offer novel, learning-theoretic support for such representations. We apply the model to English syllable onsets, Shona vowel harmony, quantity-insensitive stress typology, and the full phonotactics of Wargamay, showing that the learned grammars capture the distributional generalizations of these languages and accurately predict the findings of a phonotactic experiment.

There is an active debate within the field of phonology concerning the cognitive status of substantive phonetic factors such as ease of articulation and perceptual distinctiveness. A new framework is proposed in which substance acts as a bias, or prior, on phonological learning. Two experiments tested this framework with a method in which participants are first provided highly impoverished evidence of a new phonological pattern, and then tested on how they extend this pattern to novel contexts and novel sounds. Participants were found to generalize velar palatalization (e.g., the change from [k]asinkeep to [t�ʃ]asincheap) in a way that accords with linguistic typology, and that is predicted by a cognitive bias in favor of changes that relate perceptually similar sounds. Velar palatalization was extended from the mid front vowel context (i.e., before [e]asincape) to the high front vowel context (i.e., before [i]asin keep), but not vice versa. The key explanatory notion of perceptual similarity is quantified with a psychological model of categorization, and the substantively biased framework is formalized as a conditional random field. Implications of these results for the debate on substance, theories of phonological generalization, and the formalization of similarity are discussed.

Abstract not found

Among the most fascinating data for phonology are those showing how speakers incorporate new words and foreign words into their language system, since these data provide cues to the actual principles underlying language. In this article, we address how speakers deal with neutralized obstruents in new words. We formulate four hypotheses and test them on the basis of Dutch word-final obstruents, which are neutral for [voice]. Our experiments show that speakers predict the characteristics ofneutralized segments on the basis ofphonologically similar morphemes stored in the mental lexicon. This effect of the similar morphemes can be modeled in several ways. We compare five models, among them STOCHASTIC OPTIMALITY THEORY and ANALOGICAL MODELING OF LANGUAGE; all perform approximately equally well, but they differ in their complexity, with analogical modeling oflanguage providing the most economical explanation.* 1. INTRODUCTION. The

This paper describes in detail an algorithm for the unsupervised learning of natural language morphology, with emphasis on challenges that are encountered in languages typologically similar to European languages. It utilizes the Minimum Description Length analysis described in Goldsmith 2001 and has been implemented in software that is available for downloading and testing. 1. Scope of this paper This paper describes in detail an algorithm used for the unsupervised learning of natural language morphology which works well for European languages and other languages in which the average number of morphemes per word is not too high. 1 It has been implemented and tested in Linguistica, and is based on the theoretical principles described in Goldsmith 2001. The present paper describes that framework briefly, but the reader is referred there for a more careful development. The executable for this program, and the source code as well, is available at

inherently graded.

Abstract not found

Preprint version; to appear in Phonology, vol. 23, no. 1 In Hungarian, stems ending in a back vowel plus one or more neutral vowels show unusual behavior: 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 internalized 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. *

Two questions regarding the non-local nature of long-distance agreement in consonantal harmony patterns (Hansson 2001, Rose and Walker 2004) are addressed: (1) How can such patterns be learned from surface forms alone? (2) How can we understand a a major feature of the typology—the absence of blocking effects? It is shown that a learner which generalizes only by making distinctions with respect to the order of sounds (and by not making distinctions with respect to the distance between sounds) is able to learn major classes of long-distance phonotactic patterns, and is unable to learn hypothetical long-distance phonotactic patterns with blocking effects. Thus not only is the learner able to acquire attested patterns, it explains the absence of unattested ones. Furthermore, this result lends support to the idea that long distance phonotactic patterns are phenomonologically distinct from spreading patterns contra the hypothesis of Strict Locality (Gafos 1999, et seq).

Probabilistic methods are providing new explanatory approaches to fundamental cognitive science questions of how humans structure, process and acquire language. This review examines probabilistic models defined over traditional symbolic structures. Language comprehension and production involve probabilistic inference in such models; and acquisition involves choosing the best model, given innate constraints and linguistic and other input. Probabilistic models can account for the learning and processing of language, while maintaining the sophistication of symbolic models. A recent burgeoning of theoretical developments and online corpus creation has enabled large models to be tested, revealing probabilistic constraints in processing, undermining acquisition arguments based on a perceived poverty