A formal ideal-based account of typicality

Inspired by Barsalou’s (Journal of Experimental Psychology: Learning, Memory, and Cognition, 11, 629–654, 1985) proposal that categories can be represented by ideals, we develop and test a computational model, the ideal dimension model (IDM). The IDM is tested in its account of the typicality gradient for 11 superordinate natural language concepts and, using Bayesian model evaluation, contrasted with a standard exemplar model and a central prototype model. The IDM is found to capture typicality better than do the exemplar model and the central tendency prototype model, in terms of both goodness of fit and generalizability. The present findings challenge the dominant view that exemplar representations are most successful and present compelling evidence that superordinate natural language categories can be represented using an abstract summary, in the form of ideal representations. Supplemental appendices for this article can be downloaded from .     

Exemplar by feature applicability matrices and other Dutch normative data for semantic concepts

Features are at the core of many empirical and modeling endeavors in the study of semantic concepts. This article is concerned with the delineation of features that are important in natural language concepts and the use of these features in the study of semantic concept representation. The results of a feature generation task in which the exemplars and labels of 15 semantic categories served as cues are described. The importance of the generated features was assessed by tallying the frequency with which they were generated and by obtaining judgments of their relevance. The generated attributes also featured in extensive exemplar by feature applicability matrices covering the 15 different categories, as well as two large semantic domains (that of animals and artifacts). For all exemplars of the 15 semantic categories, typicality ratings, goodness ratings, goodness rank order, generation frequency, exemplar associative strength, category associative strength, estimated age of acquisition, word frequency, familiarity ratings, imageability ratings, and pairwise similarity ratings are described as well. By making these data easily available to other researchers in the field, we hope to provide ample opportunities for continued investigations into the nature of semantic concept representation. These data may be downloaded from the Psychonomic Society’s Archive of Norms, Stimuli, and Data, www.psychonomic.org/archive.     

Distinguishing prototype-based and exemplar-based processes in dot-pattern category learning

The authors contrast exemplar-based and prototype-based processes in dot-pattern categorization. In Experiments 1A and 1B, participants provided similarity ratings of dot-distortion pairs that were distortions of the same originating prototype. The results show that comparisons to training exemplars surrounding the prototype create flat typicality gradients within a category and small prototype-enhancement effects, whereas comparisons to a prototype center create steep typicality gradients within a category and large prototype-enhancement effects. Thus, prototype and exemplar theories make different predictions regarding common versions of the dot-distortion task. Experiment 2 tested these different predictions by having participants learn dot-pattern categories. The steep typicality gradients, the large prototype effects, and the superior fit of prototype models suggest that participants refer to-be-categorized items to a representation near the category's center (the prototype), and not to the training exemplars that surround the prototype.     

Representation at different levels in a conceptual hierarchy

The present study examines the influence of hierarchical level on category representation. Three computational models of representation – an exemplar model, a prototype model and an ideal representation model – were evaluated in their ability to account for the typicality gradient of categories at two hierarchical levels in the conceptual domain of clothes. The domain contains 20 subordinate categories (e.g., trousers, stockings and underwear) and an encompassing superordinate category (CLOTHES). The models were evaluated both in terms of their ability to fit the empirical data and their generalizability through marginal likelihood. The hierarchical level was found to clearly influence the type of representation: For concepts at the subordinate level, exemplar representations were supported. At the superordinate level, however, an ideal representation was overwhelmingly preferred over exemplar and prototype representations. This finding contributes to the increasingly dominant view that the human conceptual apparatus adopts both exemplar representations and more abstract representations, contradicting unitary approaches to categorization.     

Ideal is typical.

A well-established finding in research on concepts and categories is that some members are rated as better or more typical examples than others. It is generally thought that typicality reflects centrality, that is, that typical examples are those that are similar to many other members of the category. This interpretation of typicality is based on studies in which participants had little knowledge about the relevant categories. In the present study, experienced fishermen were asked to give goodness-of-example ratings to familiar freshwater fish. These fishermen were of two cultural groups with somewhat different goals and ideals. Typicality was well predicted by fishes' desirability and poorly predicted by their centrality. Further, the two cultural groups differed in their typicality ratings in ways that corresponded to their different goals and ideals. For knowledgeable reasoners typicality in natural taxonomic categories appears based on ideals rather than on centrality.     

Prototype abstraction by monkeys (Macaca mulatta)

The authors analyze the shape categorization of rhesus monkeys (Macaca mulatta) and the role of prototype- and exemplar-based comparison processes in monkeys' category learning. Prototype and exemplar theories make contrasting predictions regarding performance on the Posner-Homa dot-distortion categorization task. Prototype theory--which presumes that participants refer to-be-categorized items to a representation near the category's center (the prototype)--predicts steep typicality gradients and large prototype-enhancement effects. Exemplar theory--which presumes that participants refer to-be-categorized items to memorized training exemplars-predicts flat typicality gradients and small prototype-enhancement effects. Across many categorization tasks that, for the first time, assayed monkeys' dot-distortion categorization, monkeys showed steep typicality gradients and large prototype-enhancement effects. These results suggest that monkeys--like humans--refer to-be-categorized items to a prototype-like representation near the category's center rather than to a set of memorized training exemplars.     

Developmental trends in children’s typicality judgments

Category typicality norms from 12 natural language categories are presented for kindergarten, third-grade, sixth-grade, and college students. Subjects first selected examples of familiar word concepts and rated them on a 3-point scale in terms of category typicality. Age differences in the percentage of items included as category members were found primarily for the less typical items, with inclusion rates varying as a function of both age and typicality level. The absolute level of typicality judgments increased with age, although correlations between the children’s and college students’ ratings were generally significant for all three children’s groups, with average correlations increasing somewhat with age. It was suggested that the rating data would be useful to developmental investigators interested in children’s processing of category information.     

Dutch norm data for 13 semantic categories and 338 exemplars.

A data set is described that includes eight variables gathered for 13 common superordinate natural language categories and a representative set of 338 exemplars in Dutch. The category set contains 6 animal categories (reptiles, amphibians, mammals, birds, fish, and insects), 3 artifact categories (musical instruments, tools, and vehicles), 2 borderline artifact-natural-kind categories (vegetables and fruit), and 2 activity categories (sports and professions). In an exemplar and a feature generation task for the category nouns, frequency data were collected. For each of the 13 categories, a representative sample of 5-30 exemplars was selected. For all exemplars, feature generation frequencies, typicality ratings, pairwise similarity ratings, age-of-acquisition ratings, word frequencies, and word associations were gathered. Reliability estimates and some additional measures are presented. The full set of these norms is available in Excel format at the Psychonomic Society Web archive, www.psychonomic.org/archive/.     

Linguistic Distributional Knowledge and Sensorimotor Grounding both Contribute to Semantic Category Production

The human conceptual system comprises simulated information of sensorimotor experience and linguistic distributional information of how words are used in language. Moreover, the linguistic shortcut hypothesis predicts that people will use computationally cheaper linguistic distributional information where it is sufficient to inform a task response. In a pre-registered category production study, we asked participants to verbally name members of concrete and abstract categories and tested whether performance could be predicted by a novel measure of sensorimotor similarity (based on an 11-dimensional representation of sensorimotor strength) and linguistic proximity (based on word co-occurrence derived from a large corpus). As predicted, both measures predicted the order and frequency of category production but, critically, linguistic proximity had an effect above and beyond sensorimotor similarity. A follow-up study using typicality ratings as an additional predictor found that typicality was often the strongest predictor of category production variables, but it did not subsume sensorimotor and linguistic effects. Finally, we created a novel, fully grounded computational model of conceptual activation during category production, which best approximated typical human performance when conceptual activation was allowed to spread indirectly between concepts, and when candidate category members came from both sensorimotor and linguistic distributional representations. Critically, model performance was indistinguishable from typical human performance. Results support the linguistic shortcut hypothesis in semantic processing and provide strong evidence that both linguistic and grounded representations are inherent to the functioning of the conceptual system. All materials, data, and code are available at https://osf.io/vaq56/ .     

Idealness and similarity in goal-derived categories: A computational examination

The finding that the typicality gradient in goal-derived categories is mainly driven by ideals rather than by exemplar similarity has stood uncontested for nearly three decades. Due to the rather rigid earlier implementations of similarity, a key question has remained—that is, whether a more flexible approach to similarity would alter the conclusions. In the present study, we evaluated whether a similarity-based approach that allows for dimensional weighting could account for findings in goal-derived categories. To this end, we compared a computational model of exemplar similarity (the generalized context model; Nosofsky, Journal of Experimental Psychology. General 115:39–57, 1986) and a computational model of ideal representation (the ideal-dimension model; Voorspoels, Vanpaemel, & Storms, Psychonomic Bulletin & Review 18:1006-114, 2011) in their accounts of exemplar typicality in ten goal-derived categories. In terms of both goodness-of-fit and generalizability, we found strong evidence for an ideal approach in nearly all categories. We conclude that focusing on a limited set of features is necessary but not sufficient to account for the observed typicality gradient. A second aspect of ideal representations—that is, that extreme rather than common, central-tendency values drive typicality—seems to be crucial.     

Conceptual Hierarchies in a Flat Attractor Network: Dynamics of Learning and Computations

The structure of people's conceptual knowledge of concrete nouns has traditionally been viewed as hierarchical ( Collins & Quillian, 1969 ). For example, superordinate concepts ( vegetable ) are assumed to reside at a higher level than basic-level concepts ( carrot ). A feature-based attractor network with a single layer of semantic features developed representations of both basic-level and superordinate concepts. No hierarchical structure was built into the network. In Experiment and Simulation 1, the graded structure of categories (typicality ratings) is accounted for by the flat attractor network. Experiment and Simulation 2 show that, as with basic-level concepts, such a network predicts feature verification latencies for superordinate concepts ( vegetable ). In Experiment and Simulation 3, counterintuitive results regarding the temporal dynamics of similarity in semantic priming are explained by the model. By treating both types of concepts the same in terms of representation, learning, and computations, the model provides new insights into semantic memory.     

What some effects might not be: The time to verify membership in 'well-defined' categories

Armstrong, Gleitman, and Gleitman (1983) reported shorter categorization times for members of well-defined categories judged more typical. They concluded that these effects could not originate in a graded, similarity-based category representation and consequently that the typicality effects obtained with natural categories might not be indicative of such a structure either. In this article, we re-examine this conclusion, focusing first on the performance obtained with well-defined categories of different sizes. Only the larger categories used showed variations in typicality ratings and produced typicality effects on categorization times. However,multiple regression analyses showed the effects on categorization times to be better explained by a measure of associative strength, called category dominance. The range of various predictor variables was equated in a follow-up experiment involving large, natural, and well-defined categories. Results obtained with well-defined categories showed pronounced dominance effects when typicality was controlled, but no reliable typicality effect when category dominance and instance familiarity were controlled. Results were opposite for natural categories. By showing that well-defined categories fail to produce unbiased typicality effects, our results bring added support to the hypothesis that the effects obtained with natural categories originate in a graded, similarity-based category structure.     

Structural aspects of face recognition and the other-race effect

The other-race effect was examined in a series of experiments and simulations that looked at the relationships among observer ratings of typicality, familiarity, attractiveness, memorability, and the performance variables ofd’ and criterion. Experiment 1 replicated the other-race effect with our Caucasian and Japanese stimuli for both Caucasian and Asian observers. In Experiment 2, we collected ratings from Caucasian observers on the faces used in the recognition task. A Varimax-rotated principal components analysis on the rating and performance data for the Caucasian faces replicated Vokey and Read’s (1992) finding that typicality is composed of two orthogonal components, dissociable via their independent relationships to: (1) attractiveness and familiarity ratings and (2) memorahility ratings. For Japanese faces, however, we fond that typicality was related only to memorahility. Where performance measures were concerned, two additional principal components dominated by criterion and byd’ emerged for Caucasian faces. For the Japanese faces, however, the performance measures ofd’ and criterion merged into a single component that represented a second component of typicality, one orthogonal to thememorability-dominated component. A measure offace representation quality extracted from an autoassociative neural network trained with a majority of Caucasian faces and a minority of Japanese faces was incorporated into the principal components analysis. For both Caucasian and Japanese faces, the neural network measure related both to memorability ratings and to human accuracy measures. Combined, the human data and simulation results indicate that the memorahility component of typicality may be related to small, local, distinctive features, whereas the attractiveness/familiarity component may be more related to the global, shape-based properties of the face.     

Exemplar theory's predicted typicality gradient can be tested and disconfirmed

One of exemplar theory's central predictions concerns the shape of typicality gradients. The typicality gradient it predicts is a consequence of its exemplar–based comparisons and appears no matter how the theory is evaluated. However, this predicted typicality gradient does not fit the empirical typicality gradients obtained in an influential version of the dot–distortion category task, and this is true even when the exemplar model is made more flexible and mathematically powerful. Thus, exemplar theory is disconfirmed in this domain of categorization. In contrast, prototype theories are consistent with the empirically obtained typicality gradients.     

Category typicality effects in children's memory performance: Qualitative and quantitative differences in the processing of category information

Kindergarten, third, and sixth graders (6, 9, and 12 years of age, respectively) received a cue-at-input/cue-at-output recall task, using category typical and atypical items that were based on either (1) children's conceptions of item typicality, or (2) adults' conceptions of item typicality. At each grade level, recall was greater with the child-defined lists than with the adult-defined lists, and typical items were recalled to a greater extent than atypical items. Further analyses items were recalled to a greater extent than atypical items. Further analyses revealed that the recall of typical items varied as a function of children's typicality ratings of items, and that the “typicality effect” in the adult-norm condition was due primarily to the childrennot realizing that many of the atypical items were appropriate category exemplars. In contrast, typicality effects in the child-norm condition were attributed to qualitative differences in the judged “goodness of example” of the typical and atypical items. The results were discussed in terms of the appropriateness of typicality as a dimension of children's natural language concepts, the role of age differences in knowledge base in affecting performance on a cognitive task, and of the importance of using child-generated norms in studies of children's processing of category information.     

How many exemplars are used? Explorations with the Rex Leopold I model

The goal of this research is to test the hypothesis that a category is not necessarily represented by all observed exemplars, but by a reduced subset of these exemplars. To test this hypothesis, we made use of a study reported by Nosofsky, Clark, and Shin (1989), and replicated their Experiment 1 in order to gather individual-participant data. Both a full exemplar model and a reduced exemplar model were fit to the data. In general, the fits of the reduced exemplar model were superior to those of the full exemplar model. The results suggest that only a subset of exemplars may be sufficient for category representation.     

Refining the visual-cortical hypothesis in category learning

Participants produce steep typicality gradients and large prototype-enhancement effects in dot-distortion category tasks, showing that in these tasks to-be-categorized items are compared to a prototypical representation that is the central tendency of the participant’s exemplar experience. These prototype-abstraction processes have been ascribed to low-level mechanisms in primary visual cortex. Here we asked whether higher-level mechanisms in visual cortex can also sometimes support prototype abstraction. To do so, we compared dot-distortion performance when the stimuli were size constant (allowing some low-level repetition-familiarity to develop for similar shapes) or size variable (defeating repetition-familiarity effects). If prototype formation is only mediated by low-level mechanisms, stimulus-size variability should lessen prototype effects and flatten typicality gradients. Yet prototype effects and typicality gradients were the same under both conditions, whether participants learned the categories explicitly or implicitly and whether they received trial-by-trial reinforcement during transfer tests. These results broaden out the visual-cortical hypothesis because low-level visual areas, featuring retinotopic perceptual representations, would not support robust category learning or prototype-enhancement effects in an environment of pronounced variability in stimulus size. Therefore, higher-level cortical mechanisms evidently can also support prototype formation during categorization.     

The opposite of republican: polarization and political categorization

Two experiments examined the typicality structure of contrasting political categories. In Experiment 1, two separate groups of participants rated the typicality of 15 individuals, including political figures and media personalities, with respect to the categories Democrat or Republican. The relation between the two sets of ratings was negative, linear, and extremely strong, r = ?.9957. Essentially, one category was treated as a mirror image of the other. Experiment 2 replicated this result, showing some boundary conditions, and extending the result to liberal and conservative categories. The same method was applied to two other pairs of contrasting categories, healthy and junk foods, and male and female jobs. For those categories, the relation between contrasting pairs was weaker and there was less of a direct trade‐off between typicality in one category versus typicality in its opposite. The results are discussed in terms of implications for political decision making and reasoning, and conceptual representation.     

Categorization of novel stimuli in well-known natural concepts: a case study

In this study, we investigated to what extent exemplar-based and prototype predictors can be applied to predicting categorization in natural language concepts. Participants categorized novel tropical foods into two well-known natural language concepts:fruits and vegetables. The results indicate that both the prototype predictors and the exemplar predictors contribute significantly in accounting for the categorization choices but that the contribution of the prototype predictor comes from just a limited number of features.