Word meaning in minds and machines

Machines have achieved a broad and growing set of linguistic competencies, thanks to recent progress in Natural Language Processing (NLP). Psychologists have shown increasing interest in such models, comparing their output to psychological judgments such as similarity, association, priming, and comprehension, raising the question of whether the models could serve as psychological theories. In this article, we compare how humans and machines represent the meaning of words. We argue that contemporary NLP systems are fairly successful models of human word similarity, but they fall short in many other respects. Current models are too strongly linked to the text-based patterns in large corpora, and too weakly linked to the desires, goals, and beliefs that people express through words. Word meanings must also be grounded in perception and action and be capable of flexible combinations in ways that current systems are not. We discuss promising approaches to grounding NLP systems and argue that they will be more successful, with a more human-like, conceptual basis for word meaning. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

Contrasting Semantic versus Inhibitory Processing in the Angular Gyrus: An fMRI Study

Recent studies of semantic memory have focused on dissociating the neural bases of two foundational components of human thought: taxonomic categories, which group similar objects like dogs and seals based on features, and thematic categories, which group dissimilar objects like dogs and leashes based on events. While there is emerging consensus that taxonomic concepts are represented in the anterior temporal lobe, there is disagreement over whether thematic concepts are represented in the angular gyrus (AG). We previously found AG sensitivity to both kinds of concepts; however, some accounts suggest that such activity reflects inhibition of irrelevant information rather than thematic activation. To test these possibilities, an fMRI experiment investigated both types of conceptual relations in the AG during two semantic judgment tasks. Each task trained participants to give negative responses (inhibition) or positive responses (activation) to word pairs based on taxonomic and thematic criteria of relatedness. Results showed AG engagement during both negative judgments and thematic judgments, but not during positive judgments about taxonomic pairs. Together, the results suggest that activity in the AG reflects functions that include both thematic (but not taxonomic) processing and inhibiting irrelevant semantic information.

Do salient features overshadow learning of other features in category learning?

Hundreds of associative learning experiments have examined how animals learn to predict an aversive outcome, such as a shock, loud sound, or puff of air in the eye. In this study, we reversed this pattern and examined the role of an aversive stimulus, shock, as a feature of a complex stimulus composed of several features, rather than as an outcome. In particular, we used a category learning paradigm in which multiple features predicted category membership and asked whether a salient, aversive feature would reduce learning of other category features through cue competition. Three experiments compared a condition in which 1 category had among its 6 features a painful "sting" (shock) and the other category a distinctive sound (the critical features) to a control condition in which the sting and sound were represented by much less salient (and not aversive) visual depictions. Subjects learned the categories and then were tested on their knowledge of all 6 features as predictors of the category label. Surprisingly, the experiments consistently found that the salient, aversive critical features did not reduce learning of other features relative to the control. Bayesian statistics gave positive evidence for this null result. Equally surprisingly, in a fourth experiment, a nonaversive salient feature (brightly colored patterns) increased learning of other features compared to the control. We explain the results in terms of attentional strategies that may apply in a category learning context. (PsycINFO Database Record

Feature Availability in Conceptual Combination

Three experiments investigated the interpretation of conceptual combinations such as peeled apples. These experiments focused on verification of combination properties. Some properties (e.g., “round” for peeled apples) were verifiable by virtue of the noun alone, whereas others (e.g., “white” for peeled apples) required the combination of adjective and noun and generation of a new property not associated with either. Surprisingly, combination properties were verified more easily than noun properties, even under conditions of extremely rapid presentation. This finding contradicts a simple compositional model of combination in which components are analyzed prior to interpretation of the overall combination meaning. The implications for models of conceptual combination are discussed.

Learning of role-governed and thematic categories

Natural categories are often based on intrinsic characteristics, such as shared features, but they can also be based on extrinsic relationships to items outside the categories. Examples of relational categories include items that share a thematic relation or items that share a common role. Five experiments used an artificial category learning paradigm to investigate whether people can learn role-governed and thematic categories without explicit instruction or linguistic support. Participants viewed film clips in which objects were engaged in similar actions and then were asked to group together objects that they believed were in the same category. Experiments 1 and 2 demonstrated that while people spontaneously grouped items using both role-governed and thematic relations, when forced to choose between the two, most preferred role-governed categories. In Experiment 3, category labels increased this preference. Experiment 4 found that people failed to group items based on more abstract role relations when the specific relations differed (e.g., objects that prevented different actions). However, Experiment 5 showed that people could identify them with the aid of comparison. We concluded that people can form role-governed categories even with minimal perceptual and linguistic cues.

Categories, concepts, and conditioning: how humans generalize fear

During the past century, Pavlovian conditioning has served as the predominant experimental paradigm and theoretical framework to understand how humans learn to fear and avoid real or perceived dangers. Animal models for translational research offer insight into basic behavioral and neurophysiological factors mediating the acquisition, expression, inhibition, and generalization of fear. However, it is important to consider the limits of traditional animal models when applied to humans. Here, we focus on the question of how humans generalize fear. We propose that to understand fear generalization in humans requires taking into account research on higher-level cognition such as category-based induction, inferential reasoning, and representation of conceptual knowledge. Doing so will open the door for productive avenues of new research.

The neural bases of taxonomic and thematic conceptual relations: an MEG study

Converging evidence from behavioral and neuroimaging studies of human concepts indicate distinct neural systems for taxonomic and thematic knowledge. A recent study of naming in aphasia found involvement of the anterior temporal lobe (ATL) during taxonomic (feature-based) processing, and involvement of the temporoparietal junction (TPJ) during thematic (function-based) processing. We conducted an online magnetoencephalography (MEG) study to examine the spatio-temporal nature of taxonomic and thematic relations. We measured participants' brain responses to words preceded by either a taxonomically or thematically related item (e.g., cottage→castle, king→castle). In a separate experiment we collected relatedness ratings of the word pairs from participants. We examined effects of relatedness and relation type on activation in ATL and TPJ regions of interest (ROIs) using permutation t-tests to identify differences in ROI activation between conditions as well as single-trial correlational analyses to examine the millisecond-by-millisecond influence of the stimulus variables on the ROIs. Taxonomic relations strongly predicted ATL activation, and both kinds of relations influenced the TPJ. Our results further strengthen the view of the ATL's importance to taxonomic knowledge. Moreover, they provide a nuanced view of thematic relations as involving taxonomic knowledge.

Decision making under uncertain categorization

Two experiments investigated how category information is used in decision making under uncertainty and whether the framing of category information influences how it is used. Subjects were presented with vignettes in which the categorization of a critical item was ambiguous and were asked to choose among a set of actions with the goal of attaining the desired outcome for the main character in the story. The normative decision making strategy was to base the decision on all possible categories; however, research on a related topic, category-based induction, has found that people often only consider a single category when making predictions when categorization is uncertain. These experiments found that subjects tend to consider multiple categories when making decisions, but do so both when it is and is not appropriate, suggesting that use of multiple categories is not driven by an understanding of whether categories are relevant to the decision. Similarly, although a framing manipulation increased the rate of multiple-category use, it did so in situations in which multiple-category use both was and was not appropriate.

Stimulus Typicality Determines How Broadly Fear Is Generalized

The ability to represent knowledge at the category level promotes the transfer of learning. How this ability integrates with basic forms of conditioned learning is unknown but could explain why conditioned fear is overgeneralized after aversive experiences. We examined the impact of stimulus typicality—an important determinant of category-based induction—on fear learning and generalization. Typicality is known to affect the strength of categorical arguments; a premise involving typical exemplars (e.g., sparrow) is believed to apply to other members, whereas a premise about atypical exemplars (e.g., penguin) generalizes more narrowly to similar items. We adopted this framework to human fear conditioning and found that fear conditioned to typical exemplars generalized more readily to atypical members than vice versa, despite equal feature overlap across conditions. These findings have implications for understanding why some fearful events lead to broad overgeneralization of fear whereas others are regarded as isolated episodes.

Influence of emotionally charged information on category-based induction

Categories help us make predictions, or inductions, about new objects. However, we cannot always be certain that a novel object belongs to the category we are using to make predictions. In such cases, people should use multiple categories to make inductions. Past research finds that people often use only the most likely category to make inductions, even if it is not certain. In two experiments, subjects read stories and answered questions about items whose categorization was uncertain. In Experiment 1, the less likely category was either emotionally neutral or dangerous (emotionally charged or likely to pose a threat). Subjects used multiple categories in induction when one of the categories was dangerous but not when they were all neutral. In Experiment 2, the most likely category was dangerous. Here, people used multiple categories, but there was also an effect of avoidance, in which people denied that dangerous categories were the most likely. The attention-grabbing power of dangerous categories may be balanced by a higher-level strategy to reject them.

Polysemy in Sentence Comprehension: Effects of Meaning Dominance

Words like church are polysemous, having two related senses (a building and an organization). Three experiments investigated how polysemous senses are represented and processed during sentence comprehension. On one view, readers retrieve an underspecified, core meaning, which is later specified more fully with contextual information. On another view, readers retrieve one or more specific senses. In a reading task, context that was neutral or biased towards a particular sense preceded a polysemous word. Disambiguating material consistent with only one sense followed, in a second sentence (Experiment 1) or the same sentence (Experiments 2 & 3). Reading the disambiguating material was faster when it was consistent with that context, and dominant senses were committed to more strongly than subordinate senses. Critically, following neutral context, the continuation was read more quickly when it selected the dominant sense, and the degree of sense dominance partially explained the reading time advantage. Similarity of the senses also affected reading times. Across experiments, we found that sense selection may not be completed immediately following a polysemous word but is completed at a sentence boundary. Overall, the results suggest that readers select an individual sense when reading a polysemous word, rather than a core meaning.

Semantic Memory Redux: An Experimental Test of Hierarchical Category Representation

Four experiments investigated the classic issue in semantic memory of whether people organize categorical information in hierarchies and use inference to retrieve information from them, as proposed by Collins & Quillian (1969). Past evidence has focused on RT to confirm sentences such as "All birds are animals" or "Canaries breathe." However, confounding variables such as familiarity and associations between the terms have led to contradictory results. Our experiments avoided such problems by teaching subjects novel materials. Experiment 1 tested an implicit hierarchical structure in the features of a set of studied objects (e.g., all brown objects were large). Experiment 2 taught subjects nested categories of artificial bugs. In Experiment 3, subjects learned a tree structure of novel category hierarchies. In all three, the results differed from the predictions of the hierarchical inference model. In Experiment 4, subjects learned a hierarchy by means of paired associates of novel category names. Here we finally found the RT signature of hierarchical inference. We conclude that it is possible to store information in a hierarchy and retrieve it via inference, but it is difficult and avoided whenever possible. The results are more consistent with feature comparison models than hierarchical models of semantic memory.

Ideals and category typicality

Barsalou (1985) argued that exemplars that serve category goals become more typical category members. Although this claim has received support, we investigated (a) whether categories have a single ideal, as negatively valenced categories (e.g., cigarette) often have conflicting goals, and (b) whether ideal items are in fact typical, as they often have unusual attributes. Because past studies on ideals were largely correlational and often used categories not strongly associated to goals (e.g., tree, bird, fish), we took an experimental approach, using categories with obvious goals. Our results indicated that exemplars having goal-fulfilling characteristics are generally judged as less typical than exemplars with average features. Also, although subjects had a general consensus on the ideals of neutral and positive categories, they held opposing opinions on the ideals of the negatively valenced categories. We found that this bimodality in idealness perception was due to differing perspectives taken on the categories; however, perspectives that changed idealness of category exemplars did not influence their typicality. In short, ideal exemplars that best serve category goals are not necessarily perceived as typical. We contrast the goal-fulfilling aspect of ideals with the structural notion of extreme values (e.g., very tall trees), which may influence typicality through other mechanisms.

Category vs. Object Knowledge in Category-based Induction

In one form of category-based induction, people make predictions about unknown properties of objects. There is a tension between predictions made based on the object's specific features (e.g., objects above a certain size tend not to fly) and those made by reference to category-level knowledge (e.g., birds fly). Seven experiments with artificial categories investigated these two sources of induction by looking at whether people used information about correlated features within categories, suggesting that they focused on feature-feature relations rather than summary categorical information. The results showed that people relied heavily on such correlations, even when there was no reason to think that the correlations exist in the population. The results suggested that people's use of this strategy is largely unreflective, rather than strategically chosen. These findings have important implications for models of category-based induction, which generally ignore feature-feature relations.

Uncertainty in category-based induction: when do people integrate across categories?

Two experiments investigated how people perform category-based induction for items that have uncertain categorization. Whereas normative considerations suggest that people should consider multiple relevant categories, much past research has argued that people focus on only the most likely category. A new method is introduced in which responses on individual trials can be classified as using single or multiple categories, an improvement on past methods that relied on null effects as evidence for single-category use. Experiment 1 found that people did use multiple categories when the most likely category gave an ambiguous induction but that few people did so when it gave an unambiguous induction. Experiment 2 suggested that the reluctance to use multiple categories arose from a cognitive shortcut, in which only one source of information is consulted. The experiments revealed significant individual differences, suggesting that use of multiple categories is one of a number of strategies that can be used rather than being the basis for most category-based induction.

Prior knowledge and exemplar frequency

New concepts can be learned by statistical associations, as well as by relevant existing knowledge. We examined the interaction of these two processes by manipulating exemplar frequency and thematic knowledge and considering their interaction through computational modeling. Exemplar frequency affects category learning, with high-frequency items learned more quickly than low-frequency items, and prior knowledge usually speeds category learning. In two experiments in which both of these factors were manipulated, we found that the effects of frequency are greatly reduced when stimulus features are linked by thematic prior knowledge and that frequency effects on single stimulus features can actually be reversed by knowledge. We account for these results with the knowledge resonance model of category learning (Rehder & Murphy, 2003) and conclude that prior knowledge may change representations so that empirical effects, such as those caused by frequency manipulations, are modulated.

Interpreting conjunctions

The interpretation generated from a sentence of the form P and Q can often be different to that generated by Q and P, despite the fact that and has a symmetric truth-conditional meaning. We experimentally investigated to what extent this difference in meaning is due to the connective and and to what extent it is due to order of mention of the events in the sentence. In three experiments, we collected interpretations of sentences in which we varied the presence of the conjunction, the order of mention of the events, and the type of relation holding between the events (temporally vs. causally related events). The results indicated that the effect of using a conjunction was dependent on the discourse relation between the events. Our findings contradict a narrative marker theory of and, but provide partial support for a single-unit theory derived from Carston (2002). The results are discussed in terms of conjunction processing and implicatures of temporal order.

Prior knowledge enhances the category dimensionality effect

A study of the combined influence of prior knowledge and stimulus dimensionality on category learning was conducted. Subjects learned category structures with the same number of necessary dimensions but with more or fewer additional, redundant dimensions and with either knowledge-related or knowledge-unrelated features. Minimal-learning models predict that all subjects, regardless of condition, either should learn the same number of dimensions or should respond more slowly to each dimension. Despite similar learning rates and response times, subjects learned more features in the high-dimensional than in the low-dimensional condition. Furthermore, prior knowledge interacted with dimensionality, increasing what was learned, especially in the high-dimensional case. A second experiment confirmed that the participants did, in fact, learn more features during the training phase, rather than simply inferring them at test. These effects can be explained by direct associations among features (representing prior knowledge), combined with feedback between features and the category label, as was shown by simulations of the knowledge resonance, or KRES, model of category learning.

Blocking in category learning

Many theories of category learning assume that learning is driven by a need to minimize classification error. When there is no classification error, therefore, learning of individual features should be negligible. The authors tested this hypothesis by conducting three category-learning experiments adapted from an associative learning blocking paradigm. Contrary to an error-driven account of learning, participants learned a wide range of information when they learned about categories, and blocking effects were difficult to obtain. Conversely, when participants learned to predict an outcome in a task with the same formal structure and materials, blocking effects were robust and followed the predictions of error-driven learning. The authors discuss their findings in relation to models of category learning and the usefulness of category knowledge in the environment.

Prototypicality in sentence production

Three cued-recall experiments examined the effect of category typicality on the ordering of words in sentence production. Past research has found that typical items tend to be mentioned before atypical items in a phrase--a pattern usually associated with lexical variables (like word frequency), and yet typicality is a conceptual variable. Experiment 1 revealed that an appropriate conceptual framework was necessary to yield the typicality effect. Experiment 2 tested ad hoc categories that do not have prior representations in long-term memory and yielded no typicality effect. Experiment 3 used carefully matched sentences in which two category members appeared in the same or in different phrases. Typicality affected word order only when the two words appeared in the same phrase. These results are consistent with an account in which typicality has its origin in conceptual structure, which leads to differences in lexical accessibility in appropriate contexts.

Time course of retrieving conceptual information: a speed-accuracy trade-off study

Words carry considerable information, but much of that information is not relevant in context. Research has shown that readers selectively activate and remember relevant information associated with words in different contexts, but it is not known when in processing this selection occurs. This experiment investigated whether context can change which properties are initially retrieved, using a speed-accuracy trade-off paradigm. Readers had to verify a property of a modifier-noun phrase (e.g., in the sentence Boiled celery is soft) within a specified interval, from 300-3,000 msec after presentation. Results revealed that properties associated with the noun alone were activated sooner than were properties that required integration of the modifier with the noun. Thus, context did not serve to influence the initial retrieval of properties but only to activate or suppress properties later in processing.

The role of meaning in past-tense inflection: evidence from polysemy and denominal derivation

Although English verbs can be either regular (walk-walked) or irregular (sing-sang), "denominal verbs" that are derived from nouns, such as the use of the verb ring derived from the noun a ring, take the regular form even if they are homophonous with an existing irregular verb: The soldiers ringed the city rather than *The soldiers rang the city. Is this regularization due to a semantic difference from the usual verb, or is it due to the application of the default rule, namely VERB+ -ed suffix? In Experiment 1, participants rated the semantic similarity of the extended senses of polysemous verbs and denominal verbs to their central senses. Experiment 2 examined the acceptability of the regular and irregular past tenses of the different verbs. The results showed that all the denominal verbs were rated as more acceptable for the regular inflection than the same verbs used polysemously, even though the two were semantically equally similar to the central meaning. Thus, the derivation of the verb (nominal or verbal) determined the past-tense preference more than semantic variables, consistent with dual-route models of verb inflection.

Lung Lesions in SCID-bo and SCID-bg Mice after Intratracheal Inoculation with Wild-type or Leucotoxin-deficient Mutant Strains of Mannheimia haemolytica Serotype 1

The purpose of this study was to investigate SCID-bg mice engrafted with bovine haematolymphoid tissues (SCID-bo) as a model for studying bovine Mannheimia haemolytica serotype 1- induced pneumonia, in which leucotoxin (LKT) plays a major role. In experiment A, SCID-bo and SCID-bg mice were inoculated intratracheally with either (1) phosphate-buffered saline (PBS), (2) M. haemolytica wild-type strain 89010807N ("LKT(+)WT"), (3) a M. haemolytica leucotoxin-deficient mutant of strain 89010807N ("LKT(-)mutant"), or (4) the M. haemolytica wild-type Oklahoma strain. Mice were killed for examination at intervals between 20 and 44h after inoculation. Lung lesions consisted of thickened alveolar septa and neutrophil and macrophage infiltrates in the bronchioles and alveoli. Lung lesion scores in the SCID-bo mice inoculated with LKT(+)WT or LKT(-) mutant were significantly (P<0.05) greater than those of the PBS control group, but the two bacterial strains produced results that did not differ significantly. M. haemolytica was isolated from lung, liver and spleen after inoculation but less frequently as time progressed. In experiment B, SCID-bg mice were inoculated intratracheally with live LKT(+)WT or formalin-killed LKT(+)WT and killed 24, 48 or 96 h later. Lung lesions were histologically similar to those observed in experiment A; however, there were no significant differences in the lung lesion scores between groups. It was concluded that the lesions seen in this study were probably not due to LKT, and that the SCID-bo mouse does not provide a good rodent model for bovine pneumonia.

The representation of polysemy: MEG evidence

Most words in natural language are polysemous, that is, they can be used in more than one way. For example, paper can be used to refer to a substance made out of wood pulp or to a daily publication printed on that substance. Although virtually every sentence contains polysemy, there is little agreement as to how polysemy is represented in the mental lexicon. Do different uses of polysemous words involve access to a single representation or do our minds store distinct representations for each different sense? Here we investigated priming between senses with a combination of behavioral and magnetoencephalographic measures in order to test whether different senses of the same word involve identity or mere formal and semantic similarity. Our results show that polysemy effects are clearly distinct from similarity effects bilaterally. In the left hemisphere, sense-relatedness elicited shorter latencies of the M350 source, which has been hypothesized to index lexical activation. Concurrent activity in the right hemisphere, on the other hand, peaked later for sense-related than for unrelated target stimuli, suggesting competition between related senses. The obtained pattern of results supports models in which the representation of polysemy involves both representational identity and difference: Related senses connect to same abstract lexical representation, but are distinctly listed within that representation.

Category dimensionality and feature knowledge: when more features are learned as easily as fewer

Three experiments compared the learning of lower-dimensional family resemblance categories (4 dimensions) with the learning of higher-dimensional ones (8 dimensions). Category-learning models incorporating error-driven learning, hypothesis testing, or limited capacity attention predict that additional dimensions should either increase learning difficulty or decrease learning of individual features. Contrary to these predictions, the experiments showed no slower learning of high-dimensional categories; instead, subjects learned more features from high-dimensional categories than from low-dimensional categories. This result obtained both in standard learning with feedback and in noncontingent, observational learning. These results show that rather than interfering with learning, categories with more dimensions cause individuals to learn more. The authors contrast the learning of family resemblance categories with learning in classical conditioning and probability learning paradigms, in which competition among features is well documented.

Influence of multiple categories on the prediction of unknown properties

Knowing an item's category helps us predict its unknown properties. Previous studies suggest that when asked to evaluate the probability of an unknown property, people tend to consider only an item's most likely category, ignoring alternative categories. In the present study, property prediction took the form of either a probability rating or a speeded binary-choice judgment. In keeping with past findings, the subjects ignored alternative categories in their probability ratings. However, their binary-choice judgments were influenced by alternative categories. This novel finding suggests that the way in which category knowledge is used in prediction depends critically on the form of the prediction.

Frequency of relation type as a determinant of conceptual combination: a reanalysis

C. L. Gagne and E. J. Shoben (1997) proposed that concepts are combined via external relations and that lexical entries include information about which relations are frequent for every modifying noun. As evidence for this view, they showed that relations associated with the modifier affected the interpretation of combinations in several studies in which subjects had to decide whether the combinations were sensible. The authors evaluated the methods and stimuli used in Gagne and Shoben's experiments and present findings suggesting that the effect of relation frequency is likely due to differences between the familiarity and plausibility of different combinations. Although relation frequency could be involved in conceptual combination, the authors concluded that better evidence is needed for this variable, controlling for other more general differences between the combinations.

The two faces of typicality in category-based induction

Studies of category-based induction using different methods have found somewhat contradictory results for whether typical items are a stronger basis for induction. Typical category items are generally more similar to other category items than are atypical ones, and they are also more likely to be categorized into the category in question. We propose that the first aspect (representativeness) influences induction, but the second (uncertainty about the correct category) does not. Two experiments using artificial categories found support for this prediction. Two further experiments manipulated pictures of objects and also found that representativeness in the category influenced the strength of induction, but uncertainty of classification did not. Thus, the two aspects of typicality have different effects on category-based induction.

An apple is more than just a fruit: cross-classification in children's concepts

This research explored children's use of multiple forms of conceptual organization. Experiments 1 and 2 examined script (e.g., breakfast foods), taxonomic (e.g., fruits), and evaluative (e.g., junk foods) categories. The results showed that 4- and 7-year-olds categorized foods into all 3 categories, and 3-year-olds used both taxonomic and script categories. Experiment 3 found that 4- and 7-year-olds can cross-classify items, that is, classify a single food into both taxonomic and script categories. Experiments 4 and 5 showed that 7-year-olds and to some degree 4-year-olds can selectively use categories to make inductive inferences about foods. The results reveal that children do not rely solely on one form of categorization but are flexible in the types of categories they form and use.

The downside of categories

One of the primary uses of categories is to draw inferences about novel objects based on their category membership. In a recent study, Lagnado and Shanks show that people make different inferences about an object depending on whether they first categorize the object at a general or specific level. Indeed, their inference changes even though they have been given no information about the object. This finding reveals limitations of category-based induction.

A knowledge-resonance (KRES) model of category learning

This article introduces a connectionist model of category learning that takes into account the prior knowledge that people bring to new learning situations. In contrast to connectionist learning models that assume a feedforward network and learn by the delta rule or backpropagation, this model, the knowledge-resonance model, or KRES, employs a recurrent network with bidirectional symmetric connection whose weights are updated according to a contrastive Hebbian learning rule. We demonstrate that when prior knowledge is represented in the network, KRES accounts for a considerable range of empirical results regarding the effects of prior knowledge on category learning, including (1) the accelerated learning that occurs in the presence of knowledge, (2) the better learning in the presence of knowledge of category features that are not related to prior knowledge, (3) the reinterpretation of features with ambiguous interpretations in light of error-corrective feedback, and (4) the unlearning of prior knowledge when that knowledge is inappropriate in the context of a particular category.

Causes of taxonomic sorting by adults: a test of the thematic-to-taxonomic shift

The tendency among adults to sort items into taxonomic and thematic categories was examined in two experiments. Past demonstrations of adults' preference for taxonomic categories have usually not used stimuli with a salient thematic organization. The stimuli in Experiment 1 could be divided into three equal-size categories either thematically or taxonomically. Under two sets of instructions, the majority of the college-student subjects sorted thematically. In Experiment 2, a subset of the stimuli was changed so that those within it were strongly taxonomically organized. Subjects then preferred to sort the remaining items taxonomically as well. The two experiments explain why many past sorting studies have yielded a taxonomic preference in adults and provide further evidence against a global change from thematic to taxonomic preference with development.

A genotypically unique Babesia gibsoni-like parasite recovered from a dog in Oklahoma

A small Babesia gibsoni-like parasite was identified and isolated as the cause of clinical babesiosis in a dog from Oklahoma. Because this was potentially the first documented case of B. gibsoni infection in Oklahoma, further characterization was warranted, and the 18S nuclear small subunit ribosomal RNA gene was sequenced. Sequence comparison with other piroplasms from dogs showed significant nucleotide sequence differences between this isolate and both B. canis and B. gibsoni. These findings demonstrate that in domestic dogs in North America there are at least 2 "small" B. gibsoni-like organisms with distinct nucleotide sequences and that the geographic distribution of the "small" canine Babesia species may be wider than previously recognized.

Thematic relations in adults' concepts

Concepts can be organized by their members' similarities, forming a kind (e.g., animal), or by their external relations within scenes or events (e.g., cake and candles). This latter type of relation, known as the thematic relation, is frequently found to be the basis of children's but not adults' classification. However, 10 experiments found that when thematic relations are meaningful and salient, they have significant influence on adults' category construction (sorting), inductive reasoning, and verification of category membership. The authors conclude that concepts function closely with knowledge of scenes and events and that this knowledge has a role in adults' conceptual representations.

Category learning with minimal prior knowledge

In 6 experiments, the authors examined the use of prior knowledge in category learning. Previous studies of the effects of knowledge on category learning have used categories in which knowledge was related to all of the category's features. However, people's knowledge of real-world categories often consists of many "rote" features that are not related to their prior knowledge. Five experiments found that even minimal prior knowledge (1 knowledge-relevant feature and 5 rote features per exemplar) can facilitate category learning. Posttests revealed that although the knowledge aided learning, subjects also acquired the rote features that were not related to knowledge, contradicting predictions of an attentional explanation of the knowledge effect. The results of Experiment 6 suggested that subjects attempt to link even rote features to their knowledge.

Feature distribution and background knowledge in category learning

Three experiments examined the interactions of category structure and prior knowledge in category learning. Experiment 1 examined the distribution of atypical or "crossover" features in category learning. In real categories, crossover features may be unevenly distributed--found primarily in very unusual examples of a category (like whales or ostriches). In contrast, in many psychology experiments, each item has exactly one crossover feature. Even versus uneven distribution of crossover features did not affect category learning when the features were neutral. However, when the features were connected by prior knowledge, it was much harder for subjects to learn the structure with the uneven distribution of crossover features. Experiments 2 and 3 found similar results with a slightly less uneven condition. We conclude that learning is a function of the interaction of category structure and prior knowledge rather than either one alone. Furthermore, knowledge benefits learning even when the category contains contradictions of the knowledge, so long as the contradictions are not very salient.

Category learning with minimal prior knowledge

In 6 experiments, the authors examined the use of prior knowledge in category learning. Previous studies of the effects of knowledge on category learning have used categories in which knowledge was related to all of the category's features. However, people's knowledge of real-world categories often consists of many "rote" features that are not related to their prior knowledge. Five experiments found that even minimal prior knowledge (1 knowledge-relevant feature and 5 rote features per exemplar) can facilitate category learning. Posttests revealed that although the knowledge aided learning, subjects also acquired the rote features that were not related to knowledge, contradicting predictions of an attentional explanation of the knowledge effect. The results of Experiment 6 suggested that subjects attempt to link even rote features to their knowledge.