We develop a connectionist approach to processing in quasi-regular domains, as exemplified by English word reading. A consideration of the shortcomings of a previous implementation (Seidenberg & McClelland, 1989, Psych. Rev.) in reading nonwords leads to the development of orthographic and phonological representations that capture better the relevant structure among the written and spoken forms of words. In a number of simulation experiments, networks using the new representations learn to read both regular and exception words, including low-frequency exception words, and yet are still able to read pronounceable nonwords as well as skilled readers. A mathematical analysis of the effects of word frequency and spelling-sound consistency in a related but simpler system serves to clarify the close relationship of these factors in influencing naming latencies. These insights are verified in subsequent simulations, including an attractor network that reproduces the naming latency data directly in its time to settle on a response. Further analyses of the network's ability to reproduce data on impaired reading in surface dyslexia support a view of the reading system that incorporates a graded division-of-labor between semantic and phonological processes. Such a view is consistent with the more general Seidenberg and McClelland framework and has some similarities with---but also important differences from---the standard dual-route account.

Abstract not found

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

Introduction 2 0.1. Licensing: by cue or by prosody 2 0.2. Phonetics in phonology: the downward arrow and alternatives 3 0.3. An example of cue licensing: retroflexion 4 0.4. Cues 6 0.5. Cue weighting 9 0.6. Cue duration 10 0.7. The descriptive system 10 0.8. Excessive variability 13 0.9. Extensions 14 0.9.1 Direct reference to cues? 14 0.9.2 Intersegmental timing 15 0.9.3 Intrasegmental timing 15 0.9.4. Variable timing 16 Part I: Against syllable-based accounts of neutralization 1.1. Lithuanian 17 1.2. The representation of neutralized voicing 21 1.3. Word domain effects in voicing neutralization 24 2. Generalizing from Lithuanian 25 2.1. Greek and Sanskrit 25 2.1.1. Sanskrit 27 2.1.2. Greek 28 2.1.3 Aspiration neutralized 29 2.2. Voicing neutralization in Polish and Russian 30 2.3. German syllabification and devoicing 38 2.3.1 The facts 38 2.3.2. Correlations between neutralization and the syllable 40 3. A second voicing neutralization pattern: before obstr

The speech code is a vehicle of language: it defines a set of forms used by a community to carry information. Such a code is necessary to support the linguistic interactions that allow humans to communicate. How then may a speech code be formed prior to the existence of linguistic interactions? Moreover, the human speech code is discrete and compositional, shared by all the individuals of a community but different across communities, and phoneme inventories are characterized by statistical regularities. How can a speech code with these properties form? We try to approach these questions in the paper, using the ‘‘methodology of the artificial’’. We build a society of artificial agents, and detail a mechanism that shows the formation of a discrete speech code without presupposing the existence of linguistic capacities or of coordinated interactions. The mechanism is based on a low-level model of sensory–motor interactions. We show that the integration of certain very simple and non-language-specific neural devices leads to the formation of a speech code that has properties similar to the human speech code. This result relies on the self-organizing properties of a generic coupling between perception and production within agents, and on the interactions between agents. The artificial system helps us to develop better intuitions on how speech might have appeared, by showing how self-organization might have helped natural selection to find speech.

Abstract not found

Co-occurrence restrictions on word-initial consonant clusters are traditionally viewed as a consequence of the relative sonority of both members of the CC. In the first part of this paper, I aim to show that the reasoning underlying this approach is circular. The observation that sonority does increase in word-initial clusters is relabelled explanation in saying that sonority must increase. Since the crucial part of this circular argumentation is expressed by a constraint (“sonority must increase within word-initial clusters”), I address the more general issue of constraints in linguistic theory. In the second part of the paper, I propose a constraint-free theory where restrictions on word-initial clusters follow from the interaction of more general principles. The main principles I draw on are Government-Licensing (Charette, 1990), segmental complexity (Harris, 1990) and a strict CVCV syllable-structure (Lowenstamm, 1996). None of these devices makes special reference to wordinitial clusters. Since word-initial restrictions crucially depend on idiosyncratic properties of the consonants involved, I also investigate the internal structure of consonants. In the representations I introduce, the set of observations commonly subsumed under the label sonority is assigned no phonological status. Rather, it is shown to be a function of known phonological primitives. Finally, a

Functionalist phonetic literature has shown how the phonologies of human languages are arranged to facilitate ease of articulation and perception. The explanatory force of phonological theory is greatly increased if it can directly access these research results. There are two formal mechanisms that together can facilitate the link-up of formal to functional work. As others have noted, Optimality Theory, with its emphasis on directly incorporating principles of markedness, can serve as part of the bridge. Another mechanism is proposed here: an algorithm for inductive grounding permits the language learner to access the knowledge gained from experience in articulation and perception, and form from it the appropriate set of formal phonological constraints.

Abstract not found

1 Synchronic and historical explanation Evolutionary Phonology. Evolutionary Phonology seeks to derive typological generalizations from recurrent patterns of language change, themselves assumed to be rooted in perception, production, and acquisition. The goal is to eliminate UG by providing diachronic explanations for the cross-linguistic evidence that has been used to motivate it. (2) shows a schema of this program, where the arrows can be read as “explains ” and/or “constrains”. 1