Human language is an incredibly rich yet incredibly constrained system. Learning and generalizing these systematic constraints from small, sparse, and underspecified data presents a fundamental inference problem. Therapidity and ease by which humans learn these constraints has made this a foundational study in cognitive science, linguistics, and…

The past thirty years have shown a rise in models of language acquisition in which the state of the learner is characterized as a probability distribution over a set of non-stochastic grammars. In recent years, increasingly powerful models have been constructed as earlier models have failed to generalize well to increasingly complex and realistic…

Syntactic structures are unobserved theoretical constructs which are useful in explaining a wide range of linguistic and psychological phenomena. Language acquisition studies how such latent structures are acquired by human learners through many hypothesized learning mechanisms and apparatuses, which can be genetically endowed or of general…

Shows how to characterize language learning in a finite parameter space, such as in the "principles-and-parameters" approach, as a Markov structure. Explains how sample complexity varies with input distributions and learning regimes. Finds that a simple random-step algorithm always converges to the right target language and works faster than a…

Competence is defined as the knowledge of a mature speaker of a language, irrespective of questions about acquisition. It is questioned whether there can be a procedure for generating sentences as a model of this competence; it is concluded that there cannot be a precise, non-random model for competence. (CHK)

This paper describes an attempt to combine the advantages of both example-based translation and stochastic translation methods in an attempt to develop a method for inferring symbolic transfer functions from a bilingual corpus. By formalizing the translation process and applying standard optimization techniques, a system can be developed that will…

PRESENTED IN THIS PAPER IS A SIMPLIFIED DESCRIPTION OF AN ALGORITHM FOR TRANSFORMATIONAL ANALYSIS OF ENGLISH SENTENCES. IT IS NEITHER ASSUMED NOR IMPLIED IN THIS ALGORITHM THAT ANY KIND OF PRIOR ANALYSIS (EITHER STRING OR CONSTITUENT ANALYSIS) IS REQUIRED AS A PREREQUISITE. IN ORDER TO DEFINE THE SET OF ALL TRANSFORMS, THE AUTHORS FIRST DEFINE A…

The augmented transition network described in this report was developed in the course of work in semantic interpretation in the context of a computer system which answers English questions. In order to provide mechanical input for the semantic interpreter, a parsing program based on the notion of a "recursive transition network grammar" was…

The computerized morphological parser, AMPLE, grew out of work in computer assisted dialect adaptation. AMPLE contains no language-specific code, but is controlled entirely through external, user-written files, the notations of which were designed for linguists. AMPLE's constructs are linguistic: e.g., allomorph, morpheme, conditioning…

THE PAPERS IN THIS VOLUME--CATEGORIZED AS EITHER (1) AUTOMATIC ANALYSES OF NATURAL LANGUAGES, (2) STATISTIC AND SEMANTIC ANALYSES OF LINGUISTIC DATA, OR (3) ALGEBRAIC THEORIES OF LANGUAGES--ARE THE COLLECTIVE PRODUCT OF THE INTERNATIONAL CONFERENCE ON COMPUTATIONAL LINGUISTICS HELD AT GRENOBLE, FRANCE, AUGUST 23-25, 1967. TOPICS RANGE FROM…

This paper describes the notation and underlying organization of an augmented, recursive-transition network grammar and illustrates how such a grammar is a natural medium for expressing and explaining a wide variety of facts about the psychological processes of sentence comprehension. A general discussion of transformational grammar and…

There is a need for a new kind of linguistic theory which, while being concerned with both generation and analysis, must include the roles of memory, non-linguistic knowledge, and inference. The role of logic is diminished according to such a theory because inference has no real logical content. Meaning must be studied with respect to the actual…

The final part of a four-part report of research on the development of a computerized, phrase-structure grammar of modern Hebrew describes the computerized algorithm for analyzing the sentences generated based on a complex-constituent-phrase structure grammar. The first section here discusses a structural model for modern Hebrew; the second…

This first part of a four-part report of research on the development of a computerized, phrase-structure grammar of modern Hebrew presents evidence to demonstrate the need for material to train teachers of Semitic languagues in the theory of grammar. It then provides a discussion of the research already done on the application of computational…

This document contains the reports summarizing the main discussion held during the March 1972 Computational Linguistics Conference. The first report, "Computational Linguistics and Linguistics," helps to establish definitions and an understanding of the scope of computational linguistics. "Integrated Computer Systems for Language" and…