In _The Boundaries of Babel_, Andrea Moro tells the story of an encounter between two cultures: contemporary theoretical linguistics and the cognitive neurosciences. The study of language within a biological context has been ongoing for more than fifty years. The development of neuroimaging technology offers new opportunities to enrich the "biolinguistic perspective" and extend it beyond an abstract framework for inquiry. As a leading theoretical linguist in the generative tradition and also a cognitive scientist schooled in the new imaging technology, (...) Moro is uniquely equipped to explore this. Moro examines what he calls the "hidden" revolution in contemporary science: the discovery that the number of possible grammars is not infinite and that their number is biologically limited. This radical but little-discussed change in the way we look at language, he claims, will require us to rethink not just the fundamentals of linguistics and neurosciences but also our view of the human mind. Moro searches for neurobiological correlates of "the boundaries of Babel" -- the constraints on the apparent chaotic variation in human languages -- by using an original experimental design based on artificial languages. He offers a critical overview of some of the fundamental results from linguistics over the last fifty years, in particular regarding syntax, then uses these essential aspects of language to examine two neuroimaging experiments in which he took part. He describes the two neuroimaging techniques used, but makes it clear that techniques and machines do not provide interesting data without a sound theoretical framework. Finally, he discusses some speculative aspects of modern research in biolinguistics regarding the impact of the linear structure of linguistics expression on grammar, and more generally, some core aspects of language acquisition, genetics, and evolution. (shrink)

In _Reliable Reasoning_, Gilbert Harman and Sanjeev Kulkarni -- a philosopher and an engineer -- argue that philosophy and cognitive science can benefit from statistical learning theory, the theory that lies behind recent advances in machine learning. The philosophical problem of induction, for example, is in part about the reliability of inductive reasoning, where the reliability of a method is measured by its statistically expected percentage of errors -- a central topic in SLT. After discussing philosophical attempts to evade the (...) problem of induction, Harman and Kulkarni provide an admirably clear account of the basic framework of SLT and its implications for inductive reasoning. They explain the Vapnik-Chervonenkis dimension of a set of hypotheses and distinguish two kinds of inductive reasoning. The authors discuss various topics in machine learning, including nearest-neighbor methods, neural networks, and support vector machines. Finally, they describe transductive reasoning and suggest possible new models of human reasoning suggested by developments in SLT. (shrink)

Covert behavior may also be strong behavior which cannot be overtly emitted because the proper circumstances are lacking. When we are strongly inclined to go skiing, although there is no snow, we say I would like to go skiing. It is not very  ...

1 Cognitive Architectures 2 Multilayer Perceptrons 3 Relations between Variables 4 Structured Representations 5 Individuals 6 Where does the Machinery of Symbol Manipulation Come From? 7 Conclusions.

The idea that the language we speak influences the way we think has evoked perennial fascination and intense controversy. According to the strong version of this hypothesis, called the Sapir-Whorf hypothesis after the American linguists who propounded it, languages vary in their semantic partitioning of the world, and the structure of one’s language influences how one understands the world. Thus speakers of different languages perceive the world differently. Although the last two decades have been marked by extreme skepticism concerning the (...) possible effects of language on thought, recent theoretical and methodological advances in cognitive science have given the question new life. Research in linguistics and linguistic anthropology has revealed striking differences in cross-linguistic semantic patterns, and cognitive psychology has developed subtle techniques for studying how people represent and remember experience. It is now possible to test predictions about how a given language influences the thinking of its speakers. Language in Mind includes contributions from both skeptics and believers and from a range of fields. It contains work in cognitive psychology, cognitive development, linguistics, anthropology, and animal cognition. The topics discussed include space, number, motion, gender, theory of mind, thematic roles, and the ontological distinction between objects and substances. I Introduction 1 Whither Whorf 3 Dedre Gentner and Susan Goldin-Meadow II Position Statements 15 1 Languages and Representations 17 Eve V. Clark 2 Language and Mind: Let’s Get the Issues Straight! 25 Stephen C. Levinson 3 The Key Is Social Cognition 47 Michael Tomasello III Language as Lens: Does the Language We Acquire Influence How We See the World? 59 4 Sex, Syntax, and Semantics 61 Lera Boroditsky, Lauren A. Schmidt, and Webb Phillips 5 Speaking versus Thinking about Objects and Actions 81 Barbara C. Malt, Steven A. Sloman, and Silvia P. Gennari 6 The Effects of Spatial Language on Spatial Representation: Setting Some Boundaries 113 Edward Munnich and Barbara Landau 7 Language and Thought Online: Cognitive Consequences of Linguistic Relativity 157 Dan I. Slobin IV Language as Tool Kit: Does the Language We Acquire Augment Our Capacity for Higher-Order Representation and Reasoning? 193 8 Why We’re So Smart 195 Dedre Gentner 9 Does Language Help Animals Think? 237 Stan A. Kuczaj, II, and Jennifer L. Hendry 10 What Makes Us Smart? Core Knowledge and Natural Language 277 Elizabeth S. Spelke 11 Conceptual and Linguistic Factors in Inductive Projection: How Do Young Children Recognize Commonalities between Animals and Plants? 313 Kayoko Inagaki and Giyoo Hatano 12 Language for Thought: Coming to Understand False Beliefs 335 Jill G. de Villiers and Peter A. de Villiers V Language as Category Maker: Does the Language We Acquire Influence Where We Make Our Category Distinctions? 385 13 Space under Construction: Language-Specific Spatial Categorization in First Language Acquisition 387 Melissa Bowerman and Soonja Choi 14 Reevaluating Linguistic Relativity: Language-Specific Categories and the Role of Universal Ontological Knowledge in the Construal of Individuation 429 Mutsumi Imai and Reiko Mazuka 15 Interaction of Language Type and Referent Type in the Development of Nonverbal Classification Preferences 465 John A. Lucy and Suzanne Gaskins 16 Thought before Language: Do We Think Ergative? 493 Susan Goldin-Meadow. (shrink)

We compare our model of unsupervised learning of linguistic structures, ADIOS [1, 2, 3], to some recent work in computational linguistics and in grammar theory. Our approach resembles the Construction Grammar in its general philosophy (e.g., in its reliance on structural generalizations rather than on syntax projected by the lexicon, as in the current generative theories), and the Tree Adjoining Grammar in its computational characteristics (e.g., in its apparent affinity with Mildly Context Sensitive Languages). The representations learned by our algorithm (...) are truly emergent from the (unannotated) corpus data, whereas those found in published works on cognitive and construction grammars and on TAGs are hand-tailored. Thus, our results complement and extend both the computational and the more linguistically oriented research into language acquisition. We conclude by suggesting how empirical and formal study of language can be best integrated. (shrink)

We report a quantitative analysis of the cross-utterance coordination observed in child-directed language, where successive utterances often overlap in a manner that makes their constituent structure more prominent, and describe the application of a recently published unsupervised algorithm for grammar induction to the largest available corpus of such language, producing a grammar capable of accepting and generating novel wellformed sentences. We also introduce a new corpus-based method for assessing the precision and recall of an automatically acquired generative grammar without recourse (...) to human judgment. The present work sets the stage for the eventual development of more powerful unsupervised algorithms for language acquisition, which would make use of the coordination structures present in natural child-directed speech. (shrink)

In this paper, we explore the possibility that machine learning approaches to naturallanguage processing being developed in engineering-oriented computational linguistics may be able to provide specific scientific insights into the nature of human language. We argue that, in principle, machine learning results could inform basic debates about language, in one area at least, and that in practice, existing results may offer initial tentative support for this prospect. Further, results from computational learning theory can inform arguments carried on within linguistic theory (...) as well. (shrink)

The investigations reported here are the result of three lucky events. The first occurred in 1986. I had recently done the work reported in Pesetsky (1987), and received in the mail a copy of Kiss (1986). Since I had argued at length that D-linked wh-phrases do not display Superiority effects. I was astonished by a paradigm reported by Kiss, which appears here as example (98). These facts remained stubbornly in my mind for the next decade as an unsolved puzzle. Kiss (...) did not publish this paper in the form that I received — and, in fact, did not even recall discovering the crucial facts when she heard this work presented as a talk in 1998. But the facts are hers nonetheless. (shrink)

The Pirahã language challenges simplistic application of Hockett’s nearly universally accepted design features of human language by showing that some of these features may be culturally constrained. In particular, Pirahã culture constrains communication to nonabstract subjects which fall within the immediate experience of interlocutors. This constraint explains a number of very surprising features of Pirahã grammar and culture: the absence of numbers of any kind or a concept of counting and of any terms for quantification, the absence of color terms, (...) the absence of embedding, the simplest pronoun inventory known, the absence of "relative tenses," the simplest kinship system yet documented, the absence of creation myths and fiction, the absence of any individual or collective memory of more than two generations past, the absence of drawing or other art and one of the simplest material cultures documented, and the fact that the Pirahã are monolingual after more than 200 years of regular contact with Brazilians and the Tupi-Guarani-speaking Kawahiv. (shrink)

In this paper, we explore the possibility that machine learning approaches to naturallanguage processing being developed in engineering-oriented computational linguistics may be able to provide specific scientific insights into the nature of human language. We argue that, in principle, machine learning results could inform basic debates about language, in one area at least, and that in practice, existing results may offer initial tentative support for this prospect. Further, results from computational learning theory can inform arguments carried on within linguistic theory (...) as well. (shrink)

In this article, we review some of the most provocative experimental results to have emerged from comparative labs in the past few years, starting with research focusing on contingency learning and finishing with experiments exploring nonhuman animals' understanding of causal-logical relations. Although the theoretical explanation for these results is often inchoate, a clear pattern nevertheless emerges. The comparative evidence does not fit comfortably into either the traditional associationist or inferential alternatives that have dominated comparative debate for many decades now. Indeed, (...) the similarities and differences between human and nonhuman causal cognition seem to be much more multifarious than these dichotomous alternatives allow. (shrink)

We describe a linguistic pattern acquisition algorithm that learns, in an unsupervised fashion, a streamlined representation of corpus data. This is achieved by compactly coding recursively structured constituent patterns, and by placing strings that have an identical backbone and similar context structure into the same equivalence class. The resulting representations constitute an efficient encoding of linguistic knowledge and support systematic generalization to unseen sentences.