Bücher veröffentlicht von Language Science Press

Throughout much of the history of linguistics, grammaticality judgments ¿ intuitions about the well-formedness of sentences ¿ have constituted most of the empirical base against which theoretical hypothesis have been tested. Although such judgments often rest on subtle intuitions, there is no systematic methodology for eliciting them, and their apparent instability and unreliability have led many to conclude that they should be abandoned as a source of data. Carson T. Schütze presents here a detailed critical overview of the vast literature on the nature and utility of grammaticality judgments and other linguistic intuitions, and the ways they have been used in linguistic research. He shows how variation in the judgment process can arise from factors such as biological, cognitive, and social differences among subjects, the particular elicitation method used, and extraneous features of the materials being judged. He then assesses the status of judgments as reliable indicators of a speaker's grammar. Integrating substantive and methodological findings, Schütze proposes a model in which grammaticality judgments result from interaction of linguistic competence with general cognitive processes. He argues that this model provides the underpinning for empirical arguments to show that once extragrammatical variance is factored out, universal grammar succumbs to a simpler, more elegant analysis than judgment data initially lead us to expect. Finally, Schütze offers numerous practical suggestions on how to collect better and more useful data. The result is a work of vital importance that will be required reading for linguists, cognitive psychologists, and philosophers of language alike.'Native speakers' judgments of the acceptability of linguistic examples have always formed a major part of the data of linguistics, but linguists generally either have elicited such data in a haphazard fashion and accepted the results uncritically or have rejected acceptability judgments altogether and equally uncritically. Schütze's book is a welcome relief from the failure of linguists to deal responsibly with what can be either the most fertile or the most misleading source of information about languages.'¿James D. McCawley, University of Chicago'Schütze has written an extraordinarily useful and timely book. In it, he provides a clear and readable review of past studies of the methodology of generative syntax. But this is not merely a survey: it is also a call for more careful and objective scientific methods in syntax, including a set of practical methodological suggestions for working syntacticians. If heeded, they will greatly strengthen the empirical base of linguistic theory.\u2014Tom Wasow, Stanford University

One of the hardest problems in science is the symbol grounding problem, a question that has intrigued philosophers and linguists for more than a century. With the rise of artificial intelligence, the question has become very actual, especially within the field of robotics. The problem is that an agent, be it a robot or a human, perceives the world in analogue signals. Yet humans have the ability to categorise the world in symbols that they, for instance, may use for language.This book presents a series of experiments in which two robots try to solve the symbol grounding problem. The experiments are based on the language game paradigm, and involve real mobile robots that are able to develop a grounded lexicon about the objects that they can detect in their world. Crucially, neither the lexicon nor the ontology of the robots has been preprogrammed, so the experiments demonstrate how a population of embodied language users can develop their own vocabularies from scratch.

This book looks at some phenomena within the grammar of the noun phrase in a group of traditional North Germanic varieties mainly spoken in Sweden and Finland, usually seen as Swedish dialects, although the differences between them and Standard Swedish are often larger than between the latter and the other standard Mainland Scandinavian languages. In addition to being conservative in many respects ¿ e.g. in preserving nominal cases and subject-verb agreement ¿ these varieties also display many innovative features. These include extended uses of definite articles, incorporation of attributive adjectives, and a variety of possessive constructions. Although considerable attention has been given to these phenomena in earlier literature, this book is the first to put them in the perspective of typology and grammaticalization processes. It also looks for a plausible account of the historical origin of the changes involved, arguing that many of them spread from central Sweden, where they were later reverted due to the influence from prestige varieties coming from southern Scandinavia.

The Talking Heads Experiment, conducted in the years 1999-2001, was the first large-scale experiment in which open populations of situated embodied agents created for the first time ever a new shared vocabulary by playing language games about real world scenes in front of them. The agents could teleport to different physical sites in the world through the Internet. Sites, in Antwerp, Brussels, Paris, Tokyo, London, Cambridge and several other locations were linked into the network. Humans could interact with the robotic agents either on site or remotely through the Internet and thus influence the evolving ontologies and languages of the artificial agents.The present book describes in detail the motivation, the cognitive mechanisms used by the agents, the various installations of the Talking Heads, the experimental results that were obtained, and the interaction with humans. It also provides a perspective on what happened in the field after these initial groundbreaking experiments. The book is invaluable reading for anyone interested in the history of agent-based models of language evolution and the future of Artificial Intelligence.