~ ROA Version, 8/2002. Essentially identical to the Tech Report, with new pagination (but the same footnote and example numbering); correction of typos, oversights & outright errors; improved typography; and occasional small-scale clarificatory rewordings. Citation should include reference to this version.

T he distinction between marked and unmarked structures has played a role throughout this century in the development of phonology and of linguistics generally. Optimality Theory (Prince and Smolensky 1993) offers an approach to linguistic theory that aims to combine an empirically adequate theory of

A data-oriented (empiricist) alternative to the currently pervasive (nativist) Principles and Pa-rameters approach to the acquisition of stress assignment is investigated. A similarity-based algorithm, viz. an augmented version of Instance-Based Learning is used to learn the system of main stress assignment in Dutch. In this nontrivial task a comprehensive lexicon of Dutch monomorphemes is used instead of the idealized and highly simplified description of the empirical data used in previous approaches. It is demonstrated that a similarity-based learning method is effective in learning the complex stress system of Dutch. The task is accomplished without the a priori knowledge assumed to pre-exist in the learner in a Principles and Parameters framework. A comparison of the system's behavior with a consensus linguistic analysis (in the framework of Metrical Phonology) shows that ease of learning correlates with decreasing degrees of marked-ness of metrical phenomena. It is also shown that the learning algorithm captures subregularities within the stress system of Dutch that cannot be described without going beyond some of the theoretical assumptions of metrical phonology.

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.

Abstract not found

It is argued that a satisfactory global theory of intonation will require four levels of analysis : (i) physical (acoustic, physiological) (ii) phonetic (iii) surface phonological and (iv) deep phonological. The theoretical and cognitive status of each level is discussed and specific proposals are made for a model respecting such an overall architecture as well as a condition of interpretability which requires that each level of representation be interpretable in terms of adjacent levels. The level of phonetic representation is conceived of as providing an interface between abstract cognitive representations and their physical manifestations. This level is also assumed to provide an interface between constraints on production and perception. For fundamental frequency an algorithm, MOMEL, for the automatic derivation of a representation as a sequence of target-points is presented. The level of surface phonological representation is seen as the prosodic equivalent of the International Ph...

This paper discusses a perceptron model of the learning and assignment of linguistic stress, using data from nineteen human languages. First, we point out some interesting parallels between aspects of the model and the constructs and predictions of metrical phonology, the linguistic theory of stress. Second, we develop a novel analysis of linguistic stress in terms of ease of perceptron-learnability. These two sets of results suggest that simple statistical learning techniques have the potential to complement, and provide computational validation for, abstract theoretical investigations of language. We then examine why such methodologies should be of interest for linguistic theorizing. Our analysis began at a high level by observing inherent characteristics of various stress systems, much as theoretical linguistics does. However, our explanations changed substantially whenwe included a detailed account of the model's processing mechanisms. Our higher-level, theoretical accou...

this paper * , from the viewpoint of Optimality Theory, is how to handle exceptional patterns in Turkish stress: in the lexicon, with underlying templatic metrical structure, or in the grammar, through morpheme-specific constraints? The question is of particular interest because of recent work by McCarthy and Prince 1993b which challenges one of the strongest arguments in favor of underlying metrical structure in phonological theory. 1.1 Doing without templates in Optimality Theory Since McCarthy and Prince 1986, reduplication has been a major source of evidence for the existence of underlying metrical structure, or templates, in lexical entries. In "classical" reduplication, the phonological representation of a reduplicative morpheme consists virtually entirely of metrical structure (usually a syllable or a foot). However, McCarthy and Prince 1993b, 1994 have argued that Optimality Theory makes templates 1 unnecessary even in the analyses of the very phenomena that originally motivated their existence. The essential idea is that in Optimality Theory, the work of a template can be taken over by a grammatical constraint determining the type of metrical structure to which a morpheme corresponds in the surface form of the word. The particular implementation adopted in McCarthy and Prince 1993b is stated in (1): a morpheme-specific grammatical constraint specifies the exact metrical shape of each reduplicative affix. (1) "...a classical template is really nothing more than an assertion about how some morphological category... is to be aligned with some prosodic category, such as a heavy syllable or a trochaic foot." [M&P 1993b:139] Consider, for example, a simple case of prefixing reduplication of a light syllable. An analysis along the lines proposed by McCarthy and Prince...

This paper presents an Optimality-theoretic (Prince and Smolensky 1993) analysis of quantity-insensitive stress. A set of grid-based constraints is shown by means of a computer-generated factorial typology to provide a relatively tight fit to the full range of stress systems attested in an extensive survey of quantity-insensitive stress patterns, many of which have not been previously discussed in the theoretical literature.

The position in which an argument is projected is usually assumed to be determined strictly by conditions of the computational system (X-bar theory) and the LF interface (Θ-theory). We will argue that these systems, although they obviously restrict the distribution of arguments, still allow in many cases a range of positions in which a given argument can be generated (merged). In the structures we will consider here, the choice between these is made,