Abstract: ions and general knowledge more extensive than that of individual categories are required in any account of categorization. For example, Pittenger & Shaw (1975, Pittenger et al 1979) argue that higher order knowledge about transformations serves as the perceptual invariant under­ lying certain types of categorization. In the tradition of constructive mem­ ory research (e.g. Bartlett 1932, Bransford & McCarrell 1974), categories are treated as part of very general schemas. Finally, large scale computer models (Collins & Loftus 1975, Schank & Abelson 1977, Winograd 1 972) treat categories and categorization processes as inseparable from world knowledge and the inference processes used in such knowledge. DECOMPOSABILITY OF CATEGORIES INTO ELEMENTS Virtually all accounts of the representation and processing of categories assume that categories are decomposable into more elementary qualities. This is not surprising: as Dreyfus (1979) has pointed out, since the time of Plato one of the major aspects of what has been meant by an explanation has been the decomposition of the thing to be explained into its elements. In psychology, however, arguments against the indiscriminate use of expla­ nation-by-decomposition have been with the field since its inception (see e.g. James 1 890a on the unitary nature of a single complex thought). The issue of decomposability was the focus of the debate in the early part of this century between the structuralists, who saw all experience as built out of primitive meaningless sensations, and the GestaItists, who emphasized irre­ ducible emergent properties of wholes (see Boring 1 950). At present, while decomposition is the unmarked assumption in model building, the possible need for less analytic factors is periodically acknowl­ edged. For example, in the pattern recognition literature, analysis into features and holistic matching to a template are generally presented as the two major types of alternative models (Reed 1 973), although for good reasons templates are usually treated as straw men (see Palmer 1978). CATEGORIZATION OF OBJECTS 105 Various empirical developments have brought the nature and role of decomposition into current debate. First, categorization has been investi­ gated for types of stimuli that do not have obvious elements at a cognitive level. The most notable of these are color (Rosch 1 973b, 1974) and overall configuration (e.g. Attneave 1 957, Posner 1969, Lockhead 1 972-but see also Barresi et a1 1 975, Homa & Vosburgh 1976). It is not surprising that such stimuli do not strike us as obviously decomposible, since they are themselves normally treated as attributes, i.e. the qualities into which more complex objects are decomposed. There has also been considerable dis­ agreement as to whether faces should be considered special holistically perceived objects (Hochberg & Galper 1 967, Yin 1 969, Rock 1973, Brad­ shaw 1 976). A second development which has offered the opportunity to view categories as wholes is the possibility for spatial representation of withinand between-category structures through techniques such as mul­ tidimensional scaling. For example, Hutchinson & Lockhead (1977) have argued that categories can best be conceived as unanalyzed points in metric multidimensional space. A third trend has been use of the concept of a prototype and the facts of gradients of representativeness to suggest holistic processing (e.g. Rosch 1973b, Dreyfus 1 979). The great majority of arguments over decomposition concern specifying the level of abstraction at which a particular kind of decomposition can or cannot be said to occur. While most categorization models include decom­ position, it is never to the point of infinite regress. Some elements are included as the primitives, although usually by default, rather than by explicit labeling as primitives. Because some elements are not decomposed, many accounts of catego­ rization include an explicit holistic component. For example, this can be introduced by means of a (relatively) holistic processing stage (E. E. Smith et al 1 974). Another possibility is that a given level of abstraction may be a basic and (potentially) holistically perceived level, even if other levels require more analytic mechanisms (Rosch et al 1 976a). Perceptual process­ ing of figures (such as a large letter constructed of smaller letters) has been shown to proceed from global to local analysis under some stimulus condi­ tions (Navon 1 977) but to proceed from local to global under others (e.g. Kinchla & Wolf 1979). In the perception models of Palmer (1975) and Winston ( 1975), the decomposition of a visual scene is viewed as a hierar­ chical network of subscenes, and it is claimed that higher-order properties are processed first, followed by lower-order properties. Depending on the circumstances, however, a given aspect of a scene might be either the higheror the lower-order property. The controversy over lexical decompo­ sition in linguistics and artificial intelligence (does "kill" really mean "cause to die") can be seen as largely an issue of which linguistic level to consider a whole and which to consider the elements (Fodor 1970; McCawley 1971,

Abstract: Performance appraisal is construed as the outcome of a dual-process system of evaluation and decision making. Attention, categorization, recall, and information integration are carried out via either an automatic or a controlled process. In the automatic process, aspects of an employee and his/her behavior are noted and the employee is categorized without conscious monitoring. The automatic process is dominant except when decisions are problematic, in which case a consciously monitored categorization process takes place. Subsequent recall of the employee is biased by the attributes of prototypes (abstract images) representing categories to which the employee has been assigned. Dispositional and contextual factors influence the availability of categories during both assignment and recall. Categorization also biases any subsequent search for information about the employee, and interacts with task type to produce halo, leniency/ stringency biases, and racial, sexual, ethnic, and personalistic bias as well. The same automatic and controlled processes can, however, account for accuracy of evaluations. Suggested research includes the study of behavior taxonomies, individual differences in cognitive structure, the validation of behavior-sampling techniques, and laboratory studies of appraisal processes.

Abstract: A comparison was made between two procedures for testing whether an alpha-numeric character, which was pattern masked to prevent awareness, could access higher-order information concerning its category. In the indirect test, subjects made categorization responses to a letter or digit target which was immediately preceded by a masked letter or digit stimulus. In the direct test, the target stimulus was not presented; nevertheless, categorization responses were required, and it was assumed that any higher-order information available from the masked stimulus would be evident in these responses. Using the indirect procedure, category information was shown to be accessed by the masked stimulus, since target discrimination was affected by the category relationship between the stimuli. In contrast, categorization responses when the target was absent (in the direct test) were uninfluenced by the category of the masked stimulus. It was suggested that pattern masking disrupts the information necessary for (direct) report. However, it does not prevent the processing of stimuli to semantic levels, or the (indirect) effect the information accessed on the analysis of subsequent stimuli.

Abstract: KOSSAN, NANCY E. Developmental Differences in Concept Acquisition Strategies. CHILD DEVELOPMENT, 1981, 52, 290-298. The studies examined developmental differences in concept acquisition strategies as a function of the complexity of the internal structure of the concept to be learned. 3 types of concepts were examined: concepts defined by sufficient features, concepts which possessed necessary and sufficient features for concept membership, and concepts composed of exemplars with distinctive features. Secondand fifth-grade subjects learned the concepts in either a concept reception procedure which encouraged abstraction of common features or a paired associate procedure which fostered exemplar learning. The strategy most effective for concept learning was found to be a function of the age of the subject and the complexity of the concept. Subjects at both ages learned concepts defined by necessary and sufficient features better using a common feature abstraction strategy. For concepts defined by sufficient features, second graders were better at categorization following exemplar learning, while fifth graders were more successful using a common feature abstraction strategy. Concepts composed of exemplars with distinctive features were best learned using an exemplar strategy by subjects at both age levels. Convergent evidence was presented for the use of the common feature abstraction strategy and the exemplar strategy following the 2 training conditions.

Abstract: Studied the dimensionality of judgments made by groups of persons evaluating described face-to-face social interactions. In these interactions, three situational elements are varied systematically, to wit: topic of conversation, partner status, and purpose of interaction. Twenty female and 20 male British social studies students compared the similarity of 12 such descriptions of dyadic interaction. A Principal Component Factor Analyses indicates that two dimensions (‘constraint’, ‘involvement’) are significant, by providing a two-subset partition of the set of adjective pairs used by the judging groups. Furthermore, a Multiple Regression Analysis identifies the situational elements which account most for the two dimensions (partner's status for ‘constraint’, topic of conversation for ‘involvement’). The results are discussed in terms of social categorization and social order rheories. Conclusions are drawn as 10 further research on the relationship between perception of social situations and patterns of behaviour.

Abstract: A study was conducted to identify what job incumbent's interpret as dimensions of job difficulty. In this study job difficulty is defined as a job activity that requires significant physical and/or mental effort to complete. First, interviews, card sorting, and categorization procedures were used with randomly selected staff nurses in three hospitals to develop an occupation specific job difficulty composite scale. The scale was then administered to 130 medical-surgical staff nurses. Factor analysis revealed three job difficulty dimensions for nurses—overload, conflict, head nurse practices, Second, ANOVA and stepwise multiple regression were used to examine the relationships between the factor analytically derived dimensions and satisfaction, commitment, job tension, and performance ratings. It was determined that the various dimensions were differentially related to the various outcome measures.

Abstract: Husaim and Cohen's focus (Merrill-Palmer Quarterly, 1981, 27, 443-456) on the learning of ill-defined categories by infants is securely motivated. Still, some of the particular questions they pursue—namely, how many dimensions are used to form the categories and what is the salience hierarchy of the di mensions—are tricky and perhaps misleading. Underlying their design and analysis is the basic assumption that the dimensions or attributes of the stimulus as defined by the experimenter have psychological reality for the infants. This assumption is questioned. Infants may perceive different attributes in the stimulus or they may not articulate the stimulus into attributes at all.

Abstract: Six groups of subjects rated word pairs for the degree to which they exemplified one of six semantic relationships. The relationships that subjects were instructed to rate were antonymy, synonymity, subordination, superordination, coordination, and similarity. Stimulus pairs represented antonyms, synonyms, subordinates, superordinates, and coordinates. The pairs representing each stimulus relationship varied across four levels of typicality, ranging from good examples of the relationship to unrelated pairs. The highest rating in each group was given to the stimulus relationship corresponding to the relationship being judged (e.g., antonyms received the highest rating under antonym judgment instructions). This interaction was strongest for high-typicality pairs and decreased across the levels of typicality. Semantic decision models cannot explain these results unless the models are modified so that decisions are based on relationship similarity, the degree to which a stimulus pair exemplifies the relationship subjects are instructed to judge.

Abstract: This chapter reports the results of experiments carried out with children who suffer from either specific perceptual or general cognitive handicaps. The studies represent an attempt to compare the effects of such specific sensory handicaps with general cognitive deficit, but they were also intended as a breaking away from the traditional role of psychology in relation to psychiatry, a role in which psychologists have tended to accept uncritically the classificatory framework of clinically defined groups or, alternatively, to reject such diagnostically defined groupings out of hand. A further source for these studies lay in the problems which characterize the psychological investigation of subnormality, which had for long been dominated by the important concept of intelligence. Although the experimental investigation of general cognitive handicap has proceeded in the last two decades by specific investigations concerned with learning processes, this particular departure from the traditional approach via the intelligence test could nonetheless be criticized. Explicitly or implicitly, most experimental workers in the field of subnormality have assumed the existence of a linear information-handling process, beginning with the focusing of attention and followed by perception and short-term retention of input. The categorization of this retained input in long-term memory, the selection of a verbal equivalent, and the subsequent verbal or motor output associated with the stimulus have been the other stages which have been presumed.

Abstract: Summary Ss performed a categorization task in which size of target category and semantic relatedness of negative items to the target category were varied. One group of 16 Ss saw high typicality positive items—e.g., bird sparrow (high typicality group); the second group of 16 Ss saw low typicality positive items—e.g., bird peacock (low typicality group). Both groups saw the same negative items. Responses were slower in the low typicality group than in the high typicality group for both positive and negative items. For both groups responses were faster to large than to small target categories for related negative items. These findings support models that attribute categorization effects to decision processes rather than to structural properties of semantic memory.

Abstract: Social scientists often measure variables that are best thought of as having continuous underlying distributions by using crude rank-category techniques (i.e., measures which have few rank categories). The result is a large amount of grouping error. This workshop uses computer simulations to investigate the effects of four factors which contribute to grouping error: (1) the underlying distribution of scores, (2) the marginal proportion of cases in each category, (3) the scoring system used to score categories, and (4) the number of categories used. A distinction is made between two types of grouping error-transformation error and categorization error. This distinction is used to help clarify two distinct strategies which can be used to reduce grouping error.

Abstract: Under the general term ‘mathematical models’ is now subsumed a sufficiently diversified collection of distinct types of models to merit categorization and an attempt at a unified definition. In this study a broad ‘working definition’ for the concept ‘model’ is proposed; basic types of models are identified and illustrated, and the impact of conscious instruction of models on mathematics education is analysed.

Abstract: The hypothesis that the judgemental categorizing behaviour measured by Pettigrew's Category-Width scale is linked to semantic categorization is investigated. Individual differences in broad vs. narrow categorizing behaviour measured by the C-W scale are reconceptualized within the framework of Rosch's (1977) theory of human categorization. Two extreme groups on the C-W scale, broad categorizers and narrow categorizers, were selected as subjects. One experiment employed a short-term release-from-proactive-inhibition paradigm. The amount of discrepancy in recall between the preshift and the shift trials was used as an index to perceived semantic distance. The result showed that narrow categorizers perceived the same fuzzy items in a semantic category as less similar to the good exemplars of that category than broad categorizers did. In a further experiment using a semantic-categorization technique, narrow categorizers judged the same fuzzy items as category members less frequently than broad categorizers did, indicating that narrow categorizers possess more differentiated semantic categories than do broad categorizers. These findings are seen as an extension of the differential categorizing behaviour on artificial categories in the C-W scale to semantic categorization for which there is a pre-existing system.

Abstract: A Language Arts Test of Cognitive Functioning was administered to seventh grade students to estimate levels of cognitive functioning in language arts tasks. The six tasks developed are based on Shayer's curriculum taxonomy: (1) interest and investigation style, (2) reasons for events, (3) relationships, (4) use of a model as a theory, (5) type of categorization, and (6) depth of interpretation. Student responses were analyzed and students were classified as concrete, transitional, or formal thinkers. When compared with results from Bumey's Logical Reasoning Test, students were generally rated lower in cognitive functioning on the LATCF. Of greater importance was the predictability of the LATCF for language arts thinking. Student answers in response to the six tasks elicited specific characteristics of concrete, transitional, and formal thinkers.

Abstract: Our thanks to Kemler (1981) for her thought-provoking com ments regarding the Husaim and Cohen (1981) experiment on in fants' learning of ill-defined categories. She raised a number of important issues that need to be resolved before an understanding of infant categorization is complete. Needless to say, we agreed with several of these points and shall indicate those areas of agreement in this paper. On the other hand, some of her comments suggested a misunderstanding both of the goals of the experiment and of how well those goals were met. We shall also try to correct those mis understandings. Three basic questions that were addressed by Husaim and Cohen will be re-examined here, (a) Can 10-month-old infants learn to respond differentially to two separate categories? (b) Can these infants learn to distinguish between categories that have no clearly defined, criterial attributes differentiating between them? (i.e., Can they learn ill-defined categories?). And (c) if the answer to either the first or second question is yes, can the approach used by Husaim and Cohen be useful in discovering the bases on which those categories are learned?

Abstract: The aim of this book as elaborated in preceding chapters is to classify speech acts by means of a classification of speech act verbs1 More precisely this means that we want to start from the set of all the verbs of a given language which designate (aspects of) speech activities, list them, and categorize them according to a well-specified system which we shall describe immediately The resulting categorization will reflect the ontological and conceptual structuring of the linguistic behavior of the speakers of the language in its various facets

Abstract: Four tasks measuring breadth of categorization in different ways were administered to 128 4-year-old children. Independent personality data were available on these children at ages 3,4, and 7 in the form of behavior ratings by independent sets of teachers using the California Child Q-Set and in the form of behavior ratings by independent sets of examiners. The four category breadth measures were standard-scored and composited to generate a domain-representative index of category breadth. This index was then correlated with the independent personality ratings available at each of the four ages. A core set of relationships, common across both age and sex, indicated that broad categorizers, relative to narrow categorizers, had difficulty in stimulus exclusion, lacked autonomous structuring, and generally behaved with few internal constraints. Boys were significantly broader in their categorization than girls, although there was appreciable overlap between the sexes. Personality correlates of breadth of categorization observed as early as 3 years of age were also observed at age 7. Results were placed in a larger theoretical context by introducing the concept of underand overperceptualization. The act of categorizing objects and events is a way of structuring information about the world. Categorization provides an efficient strategy for managing inputs because it permits dealing with classes of objects rather than single entities, allows the specification of salient criterial features of categories according to which newly encountered events or objects can be classified, and facilitates information retrieval. Individuals differ in their categorizing proclivities: Category definitions may not be incisively formulated; criteria for category membership may be applied inconsistently or shift as a function of context; categories may be overly narrow and restrictive, resulting in the overexclusion of reasonably includable exemplars; and categories may be excessively broad, resulting in overinclusion of dissimilar exemplars.

Abstract: How do strategies affect the learning of categories that lack necessary and sufficient attributes? The usual answer is that different strategies correspond to different models. In this article we provide evidence for an alternative view— Strategy variations induced by instructions affect only the amount of information represented about attributes, not the process operating on these representations. The experiment required subjects to classify schematic faces into two categories. Three groups of subjects worked with different sets of instructions: roughly, form a prototype of each category, learn each category as a rule-plus-exception, or standard neutral instructions. In addition to learning the faces (Phase 1), subjects were given transfer tests on learned and novel faces (Phase 2) and speeded categorization tests on learned faces (Phase 3). There were performance differences in all three phases due to instructions, but these results were readily accounted for by specific changes in the representations posited by the context model of Medin and Schaffer; that is, strategies seemed to affect only the amount of information stored about each exemplar's attributes.