Prior expectation and the interpretation of natural language quantifiers

Abstract

The interpretation of natural language quantifiers depends, in part, on one's prior expectations about what the proportion being described might be. This hypothesis is tested empirically, by comparing the interpretation of 10 different quantifier expressions in three contexts which contained proportions about which subjects' expectations were reliably different (Experiment 1). It is further argued that quantifiers perform many functions which cannot be observed by considering only the proportion(s) denoted by a quantifier. The experimental evidence presented shows that quantifiers can provide information about what the user of the quantifier had expected before “knowing the facts” (Experiment 2). It is also clear that when writers produce a quantifier, they can reveal their assumptions about the prior expectations held by their readers (Experiment 2).     

Understanding natural language

This paper describes a computer system for understanding English. The system answers questions, executes commands, and accepts information in an interactive English dialog.     

Learning theory and natural language

Formal Learning Theory may be conceived as a means of relating theories of comparative grammar to studies of linguistic development. After a brief review of relevant concepts, the present paper surveys formal results within Learning Theory that suggest correponding constraints on linguistic theory. Particular attention is devoted to the question: How many possible natural languages are there?     

Bilateral brain processes for comprehending natural language

Comprehension of natural language--stories, conversations, text--is very simple for those doing the comprehending and very complex for cognitive neuroscientists. It also presents a paradox: the advantage of the left hemisphere (LH) for most language tasks is one of the best-established facts about the brain; yet, when it comes to comprehending complex, natural language, the right hemisphere (RH) might play an important role. Accumulated evidence from neuropsychology, neuroimaging, and neuroanatomy suggests at least three roughly separable (but highly interactive) components of semantic processing. Each process in turn has bilateral components, with the RH component performing coarser computations for the same general process. Examining asymmetrical brain and cognitive functions provides a unique opportunity for understanding the neural basis of complex cognition.     

The redundancy of recursion and infinity for natural language

An influential line of thought claims that natural language and arithmetic processing require recursion, a putative hallmark of human cognitive processing (Chomsky in Evolution of human language: biolinguistic perspectives. Cambridge University Press, Cambridge, pp 45–61, 2010; Fitch et al. in Cognition 97(2):179–210, 2005; Hauser et al. in Science 298(5598):1569–1579, 2002). First, we question the need for recursion in human cognitive processing by arguing that a generally simpler and less resource demanding process—iteration—is sufficient to account for human natural language and arithmetic performance. We argue that the only motivation for recursion, the infinity in natural language and arithmetic competence, is equally approachable by iteration and recursion. Second, we submit that the infinity in natural language and arithmetic competence reduces to imagining infinite embedding or concatenation, which is completely independent from the ability to implement infinite processing, and thus, independent from both recursion and iteration. Furthermore, we claim that a property of natural language is physically uncountable finity and not discrete infinity.     

Connectionist natural language parsing

The key developments of two decades of connectionist parsing are reviewed. Connectionist parsers are assessed according to their ability to learn to represent syntactic structures from examples automatically, without being presented with symbolic grammar rules. This review also considers the extent to which connectionist parsers offer computational models of human sentence processing and provide plausible accounts of psycholinguistic data. In considering these issues, special attention is paid to the level of realism, the nature of the modularity, and the type of processing that is to be found in a wide range of parsers.     

Concurrent learning of temporal and spatial sequences

In a serial reaction time task, stimulus events simultaneously defined spatial and temporal sequences. Responses were based on the spatial dimension. The temporal sequence was incidental to the task, defined by the response-to-stimulus intervals in Experiment 1 and stimulus onset asynchronies in Experiment 2. The two sequences were either of equal length and correlated or of unequal length. In both experiments, spatial learning occurred regardless of sequence length condition. In contrast, temporal learning occurred only in the correlated condition. These results suggest that timing is an integrated part of action representations and that incidental learning for a temporal pattern does not occur independently from the action. Interestingly, sequence learning was enhanced in the correlated condition, reflecting the integration of spatial-temporal information.     

Dialogue-Games: Metacommunication structures for natural language interaction

Our studies of naturally occurring human dialogue have led to the recognition of a class of regularities which characterize impoltant aspects of communication. People appear to interact according to established patterns which span several turns in a dialogue and which recur frequently. These patterns appear to be organized around the goals which the dialogue serves for each participant. Many things which are said later in a dialogue can only be interpreted as pursuit of these goals, established by earlier dialogue. These patterns have been represented b) a set of knowledge structures called Dialogue-Games, capturing shared, conventional knowledge that people have about communication and how it can be used to achieve goals. A Dialogue-Game has Parameters, which represent those elements that vary across instances of a particular pattern-the particular dialogue participants and the content topic. The states of the world which must be in effect for a particular Dialogue-Game to be employed successfully are represented by Specifications of these Parameters. Finally, the expected sequence of intermediate states that occur during instances of a particular conventional pattern are represented by the Components of the corresponding Dialogue-Game. Representations for several Dialogue-Games are presented here, based on our analyses of different kinds of naturally occurring dialogue. A process model is discussed, showing Dialogue-Game identification, pursuit, and termination as part of the comprehension of dialogue utterances. This Dialogue-Game model captures some of the important functional aspects of language, especially indirect uses to achieve implicit communication.     

Thought beyond language: neural dissociation of algebra and natural language

A central question in cognitive science is whether natural language provides combinatorial operations that are essential to diverse domains of thought. In the study reported here, we addressed this issue by examining the role of linguistic mechanisms in forging the hierarchical structures of algebra. In a 3-T functional MRI experiment, we showed that processing of the syntax-like operations of algebra does not rely on the neural mechanisms of natural language. Our findings indicate that processing the syntax of language elicits the known substrate of linguistic competence, whereas algebraic operations recruit bilateral parietal brain regions previously implicated in the representation of magnitude. This double dissociation argues against the view that language provides the structure of thought across all cognitive domains.     

Generalization from natural language text

Generalization and memory are part of natural language understanding. This paper describes a model of generalization that is part of a system for language understanding. the Integrated Partial Parser (IPP). The generalization process includes the retrieval of relevant examples from long-term memory so that the concepts to be created can be determined when new stories are read. Generalizations are analyzed as to their critical parts and evaluated in light of later evidence. The knowledge structures used and a number of examples of the system in operation are presented.