Online research in philosophy

I examine one of the conceptual cornerstones of the field known as computational neuroscience, especially as articulated in Churchland et al. (1990), an article that is arguably the locus classicus of this term and its meaning. The authors of that article try, but I claim ultimately fail, to mark off the enterprise of computational neuroscience as an interdisciplinary approach to understanding the cognitive, information-processing functions of the brain. The failure is a result of the fact that the authors provide no principled means to distinguish the study of neural systems as genuinely computational/information-processing from the study of any complex causal process. I then argue for two things. First, that in order to appropriately mark off computational neuroscience, one must be able to assign a semantics to the states over which an attempt to provide a computational explanation is made. Second, I show that neither of the two most popular ways of trying to effect such content assignation -- informational semantics and 'biosemantics' -- can make the required distinction, at least not in a way that a computational neuroscientist should be happy about. The moral of the story as I take it is not a negative one to the effect that computational neuroscience is in principle incapable of doing what it wants to do. Rather, it is to point out some work that remains to be done.

If a computational account of visual perception were correct, then perception would involve at least two sorts of rule-guided inference processes: inference from primitive input to complex perceptual output (constructional inference) and inference from perceptual content to the organism's environment (representational inference). Psychologist J.J. Gibson argues that such accounts are circular. Fodor and Pylyshyn argue that Gibson's alternative account, though intended to be non-inferential, actually requires the above two sorts of inference. But their arguments for the necessity of inference work only if (1) complex properties cannot be transduced, and (2) we assume a signal transmission model of perception. The force of their arguments is weakened once we see that (1) their criterion for non-transducibility is itself problematic, and (2) an interactive model of visual perception does not require signal transmission.

Computational semantics is the study of how to represent meaning in a way that computers can use. For the authors of this textbook, this study includes the representation of the meaning of natural language in logic formalisms, the recognition of certain relations that hold within this formalization (such as synonymy, consistency, and implication), and the computational implementation of all this. I think that, while there probably are not many courses devoted to computational semantics, this book could profitably be incorporated into more traditional computational linguistics courses, especially when two courses are offered serially. The material here could be spread out and integrated into parts of a more standard pair of these courses, and it would result in a substantial widening of the knowledge that students come away with from these courses.

There is a prevalent notion among cognitive scientists and philosophers of mind that computers are merely formal symbol manipulators, performing the actions they do solely on the basis of the syntactic properties of the symbols they manipulate. This view of computers has allowed some philosophers to divorce semantics from computational explanations. Semantic content, then, becomes something one adds to computational explanations to get psychological explanations. Other philosophers, such as Stephen Stich, have taken a stronger view, advocating doing away with semantics entirely. This paper argues that a correct account of computation requires us to attribute content to computational processes in order to explain which functions are being computed. This entails that computational psychology must countenance mental representations. Since anti-semantic positions are incompatible with computational psychology thus construed, they ought to be rejected. Lastly, I argue that in an important sense, computers are not formal symbol manipulators.

1. Formal semantics in linguistics -- 2. Generalized quantifier theory -- 3. The interface between syntax and semantics -- 4. Anaphora, discourse, and modality -- 5. Focus, presupposition, and negation -- 6. Tense -- 7. Questions -- 8. Plurals -- 9. Computational semantics -- 10. Lexical semantics -- 11. Semantics and related domains.

Lexical semantics has become a major research area within computational linguistics, drawing from psycholinguistics, knowledge representation, computer algorithms and architecture. Research programmes whose goal is the definition of large lexicons are asking what the appropriate representation structure is for different facets of lexical information. Among these facets, semantic information is probably the most complex and the least explored.Computational Lexical Semantics is one of the first volumes to provide models for the creation of various kinds of computerised lexicons for the automatic treatment of natural language, with applications to machine translation, automatic indexing, and database front-ends, knowledge extraction, among other things. It focuses on semantic issues, as seen by linguists, psychologists, and computer scientists. Besides describing academic research, it also covers ongoing industrial projects.

Semantics is concerned with meaning: what meanings are, how meanings are assigned to words, phrases and sentences of natural and formal languages, and how meanings can be combined and used for inference and reasoning. The goal of this chapter is to introduce computational linguists and computer scientists to the tools, methods, and concepts required to work on natural language semantics. Semantics, while often paired with pragmatics, is nominally distinct. On a traditional view, semantics concerns itself with the compositional buildup of meaning from the lexicon to the sentence level whereas pragmatics concerns the way in which contextual factors and speaker intentions affect meaning and inference (see, e.g., Potts to appear in this volume). Although the semantics-pragmatics distinction is historically important, and continues to be widely adopted, in practice it is not clearcut. Work in semantics inevitably involves pragmatics and vice versa. Furthermore, it is not a distinction which is of much relevance for applications in computational linguistics. This chapter is organized as follows. In sections 2 and 3 we introduce foundational concepts and discuss ways of representing the meaning of sentences, and of combining the meaning of smaller expressions to produce those sentential meanings. In section 4 we discuss the representation of meaning for larger units, especially with respect to anaphora, and introduce two formal theories that go beyond sentence meaning: Discourse Representation Theory and Dynamic Semantics. Then, in section 5 we discuss temporality, introducing event semantics, and describing standard approaches to the semantics of tense and aspect. Section 6 concerns the tension between the surface-oriented statistical methods characteristic of much of computational linguistics and the more abstract methods typical of formal semantics and includes discussion of a range of phenomena for which it seems particularly important to utilize insights from formal semantics..

In this article we discuss what constitutes a good choice of semantic representation, compare different approaches of constructing semantic representations for fragments of natural language, and give an overview of recent methods for employing inference engines for natural language understanding tasks.

Just as war can be viewed as continuation of diplomacy using other means, computational semantics is continuation of logical analysis of natural language by other means. For a long time, the tool of choice for this used to be Prolog. In our recent textbook we argue (and try to demonstrate by example) that lazy functional programming is a more appropriate tool. In the talk we will lay out a program for computational semantics, by linking computational semantics to the general analysis of procedures for social interaction. The talk will give examples of how Haskell can be used to begin carrying out this program.

How can computers distinguish the coherent from the unintelligible, recognize new information in a sentence, or draw inferences from a natural language passage? Computational semantics is an exciting new field that seeks answers to these questions, and this volume is the first textbook wholly devoted to this growing subdiscipline. The book explains the underlying theoretical issues and fundamental techniques for computing semantic representations for fragments of natural language. This volume will be an essential text for computer scientists, linguists, and anyone interested in the development of computational semantics.