Does knowledge of language consist of mentally-represented rules? Rumelhart and McClelland have described a connectionist (parallel distributed processing) model of the acquisition of the past tense in English which successfully maps many stems onto their past tense forms, both regular (walk/walked) and irregular (go/went), and which mimics some of the errors and sequences of development of children. Yet the model contains no explicit rules, only a set of neuron-style units which stand for trigrams of phonetic features of the stem, a set of units which stand for trigrams of phonetic features of the past form, and an array of connections between the two sets of units whose strengths are modified during learning. Rumelhart and McClelland conclude that linguistic rules may be merely convenient approximate fictions and that the real causal processes in language use and acquisition must be characterized as the transfer of activation levels among units and the modification of the weights of their connections. We analyze both the linguistic and the developmental assumptions of the model in detail and discover that (1) it cannot represent certain words, (2) it cannot learn many rules, (3) it can learn rules found in no human language, (4) it cannot explain morphological and phonological regularities, (5) it cannot explain the differences between irregular and regular forms, (6) it fails at its assigned task of mastering the past tense of English, (7) it gives an incorrect explanation for two developmental phenomena: stages of overregularization of irregular forms such as bringed, and the appearance of doubly-marked forms such as ated, and (8) it gives accounts of two others (infrequent overregularization of verbs ending in t/d, and the order of acquisition of different irregula...

Selectional constraints are limitations on the applicability of predicates to arguments. For example, the statement “The number two is blue” may be syntactically well formed, but at some level it is anomalous — BLUE is not a predicate that can be applied to numbers. According to the influential theory of (Katz and Fodor, 1964), a predicate associates a set of defining features with each argument, expressed within a restricted semantic vocabulary. Despite the persistence of this theory, however, there is widespread agreement about its empirical shortcomings (McCawley, 1968; Fodor, 1977). As an alternative, some critics of the Katz-Fodor theory (e.g. (Johnson-Laird, 1983)) have abandoned the treatment of selectional constraints as semantic, instead treating them as indistinguishable from inferences made on the basis of factual knowledge. This provides a better match for the empirical phenomena, but it opens up a different problem: if selectional constraints are the same as inferences in general, then accounting for them will require a much more complete understanding of knowledge representation and inference than we have at present. The problem, then, is this: how can a theory of selectional constraints be elaborated without first having either an empirically adequate theory of defining features or a comprehensive theory of inference? In this dissertation, I suggest that an answer to this question lies in the representation of conceptual

Abstract not found

Abstract not found

In this paper, we present a conceptual framework for indexing different aspects of visual information. Our framework unifies concepts from the literature in diverse fields such as cognitive psychology, library sciences, art, and the more recent contentbased retrieval. We present multiple level structures for visual and non-visual information. The ten-level visual structure presented provides a systematic way of indexing images based on syntax (e.g., color, texture, etc.) and semantics (e.g., objects, events, etc.), and includes distinctions between general concept and visual concept. We define different types of relations (e.g., syntactic, semantic) at different levels of the visual structure, and also use a semantic information table to summarize important aspects related to an image. While the focus is on the development of a conceptual indexing structure, our aim is also to bring together the knowledge from various fields, unifying the issues that should be considered when building ...

This study explores the commonalities between linguistic and visual representations of space. In particular, because common types of spatial relations, specifically closed-class spatial forms in language and qualitative spatial relations in perception, have been proposed in both representational systems, we investigate whether they share underlying structural similarities. Moreover, while visual spatial relations are a basic element of several theories of object representation, they have been characterized mainly in terms of their linguistic counterparts and without direct evidence about their organization. In order to illuminate the nature of these structures, as well as demonstrate possible correspondences between the two systems, we compare how the spatial relationship between pairs of objects in a scene is encoded linguistically and visually. Spatial language was investigated by having subjects either generate (Experiment 1) or rate the applicability of (Experiment 2) spatial terms for describing the spatial relationship between object pairs. Both the frequency of use and the applicability of spatial terms were highest when the two objects were in vertical or in horizontal alignment. Spatial representation was investigated by paradigms in which subjects either recalled the position of one object relative to the other (Experiment 3) or judged whether one object presented sequentially was in the same or a different position relative to the other (Experiment 4). The accuracy of position estimates and the sensitivity to shifts in position were both highest when the rated object was in a spatial location where spatial terms had been judged to have high applicability in Experiments 1 and 2. These results indicate that the structure of space as encoded by language may b...

Address correspondence to

Theories of categorization have typically focused on the internal structure of categories. This paper is concerned with the external structure of categories. In particular , it is suggested that many categories specify the relational role that is played by category members. To support this claim, the paper distinguishes between traditional feature-based categories, relational categories (which specify a relational structure) and role-governed categories (which specify that an item plays a particular role within a relational structure). After discussing the relationship among these types of categories, the implications of this view for the study of category learning and category use are discussed.

A rule-instantiation model and a similarity-to-exemplars model were contrasted in terms of their predictions of typicality judgments and speeded classificalions-for-members-oflogically defined categories. In Experiment 1, subjects learned a unidimensional rule based on the size of objects. It was assumed that items that maximally instantiated the rule were those farthest from the category boundary that separated small and large stimuli. In Experiment 2, subjects learned a disjunctive rule ofthe form “x or y or both. ” It was assumed that itemsthat maximally instantiated the rule were those withbothpositive values (x and y). In both experiments, the frequency with which different exemplars were presented during classification learning was manipulated across conditions. Thesefrequency manipulations exerted a major impact on subjects ’ postacquisition goodness-of-example judgments, and they also influenced reaction times in a speeded classification task. The results could not be predicted solely on the basis of the degree to which the rules were instantiated. The goodnessjudgments were predicted fairly well by a mixed-exemplar model involving both relative-similarity and absolute-similarity components. It was concluded that even for logically defined concepts, stored exemplars may form a major component of the category representation.

Human concept learning presents a version of the classic problem of induction, which is made particularly difficult by the combination of two requirements: the need to learn from a rich (i.e. nested and overlapping) vocabulary of possible concepts and the need to be able to generalize concepts reasonably from only a few positive examples. I begin this thesis by considering a simple number concept game as a concrete illustration of this ability. On this task, human learners can with reasonable confidence lock in on one out of a billion billion billion logically possible concepts, after seeing only four positive examples of the concept, and can generalize informatively after seeing just a single example. Neither of the two classic approaches to inductive inference -- hypothesis testing in a constrained space of possible rules and computing similarity to the observed examples -- can provide a complete picture of how people generalize concepts in even this simple setting. This thesis prop...