Edward J. Shoben

Differential Context Effects in the Comprehension of Abstract and Concrete Verbal Materials.

Three experiments were performed to test contrasting predictions of a dual-representation theory and a context availability model of concreteness effects in verbal processing. In one experiment, abstract and concrete sentences were presented with and without a paragraph context. Without context, subjects took longer to read abstract sentences than concrete sentences. With context, the reading times did not differ. A similar result was observed in a second experiment in which lexical decision times were measured for abstract and concrete words. In the absence of context, lexical decision times for abstract words were longer than for concrete words. With a sentence context, however, the lexical decision times for these two word types were equivalent. A subsequent rating experiment indicated that rated context availability was a good predictor of reaction time in both experiments. The results were discussed as providing support for the context availability model.

Context and structure in conceptual combination

Three experiments evaluated modifications of conceptual knowledge associated with judgments of adjective-noun conceptual combinations. Existing models, such as the Smith and Osherson modification model, assume that the changes associated with understanding an adjective noun combination are confined to the corresponding adjectival dimension. Our experiments indicate that this assumption is too strong. The first study found that naming one dimension affects correlated dimensions. For example, participants judge small spoons to be more typical spoons than large spoons, but for wooden spoons, large spoons are more typical than small spoons. The second study demonstrated that the similarity of adjectives is not independent of the noun context in which they appear. For example, white and gray are judged to be more similar than gray and black in the context of hair but this judgment reverses in the context of clouds. The third study showed that a property equally true (or false) for two concepts may be more central to one concept than the other (e.g., it is more important that boomerangs be curved than that bananas be curved). These results pose serious problems for current theories of how people combine concepts. We propose instead that we need richer views of both the conceptual structure and the modifications of it required by conceptual combination. We suggest that theoretical knowledge and the construct of centrality of meaning may play useful roles.

The Effect of Context on the Structure of Categories

Abstract Three experiments examined the effect of context on the representativeness ordering of exemplars of a category. Experiments 1 and 2 employed an online reading time paradigm to examine the effect of context on the time it takes to establish an anaphoric reference between an exemplar and a category term. Experiment 1 demonstrated that context can change the relation between a category term and an exemplar at the time of comprehension. Experiment 2 showed that category terms presented in context generate graded goodness-of-example distributions of exemplars that are different from the distributions generated in the absence of explicit context. These distributions cannot be derived by assuming that the exemplar most strongly suggested by the context serves as the category representation. Experiment 3 employed a membership verification paradigm. Response time was found to be a function of degree of relatedness to the contextdependent category representation. Typicality, as determined in the absence of explicit context, had no effect on decision time. Several models, including some extensions of current semantic memory theories, are developed to account for the results of these experiments.

Influence of Thematic Relations on the Comprehension of Modifier–noun Combinations

To comprehend a nonpredicating combination involving a modifier (e.g., mountain) and a head noun (e.g., stream), one must specify a thematic relation (e.g., a stream in the mountains) that links the 2 constituent concepts. The authors investigated the influence of thematic relations on the comprehension of nonpredicating combinations. Experiments 1 and 3 demonstrated that people use information about what relations the constituents typically instantiate during conceptual combination. More specifically, a combination is easier to interpret when it uses a frequent relation of the modifier than when it uses a less frequent relation. The results of Experiment 2 indicated that these results are not an artifact of the individual component words. The authors propose a model of conceptual combination called the competition among relations in nominals (CARIN) model in which ease of comprehension depends both on the frequency of the to-be-selected relation and on the frequency of the alternatives.

The influence of sentence constraint on the scope of facilitation for upcoming words

Abstract Four experiments were performed to examine the influence of sentence constraint and cue validity on the processing of expected and unexpected congruous sentence completions. Experiment 1 showed that high constraint sentences aided lexical decisions only for expected completions whereas low constraint contexts demonstrated a broader, although weaker context effect. Increasing the proportion of expected completions in Experiment 2 caused inhibition of lexical decisions for unexpected words appearing in high constraint sentences. A similar manipulation for low constraint sentences in Experiment 3 did not show such an effect for unexpected completions. The addition of an incongruous completion condition in Experiment 4 had a negligible effect on the relative proportions of facilitation found in every condition. These findings are consistent with the view that more featural restrictions are generated as sentence constraint and cue validity increase.

Scripts as linear orders

Abstract Two models of peoples knowledge of routine activities (e.g., eating in a restaurant) are outlined. The first of these script models assumes that events in a script can only be accessed in sequential (that is, temporal) order. In contrast to this retrieval-dependent model, the second model is derived from a theory of linear orders, or the ordering of objects on a given dimension (e.g., animals on a dimension of size). This retrieval-independent model asserts that all items can be retrieved with equal facility. In the first of two reaction time experiments, subjects judged which of two events in a given script occurred sooner (or later). The results were similar to those observed in linear order experiments, and supported the retrieval-independent model. A second experiment ruled out an artifactual explanation of our results. The relation between script theory and the theory of linear orders is discussed.

Priming relations in ambiguous noun-noun combinations

We conducted two experiments to examine whether the interpretation of an ambiguous noun phrase is influenced by exposure to a similar combination. In Experiment 1, we found that it was easier to verify a definition for a combination (e.g.,adolescent doctor, a doctor for adolescents) when the prime used the same relation as the target (e.g.,adolescent magazine, a magazine for adolescents;animal doctor, a doctor for animals) than when the prime used a different relation (e.g.,country doctor; adolescent experience). In Experiment 2, we found that the interpretation generated for an ambiguous combination was affected by prior exposure to sentences containing a combination with the same modifier or head noun as the target combination. The data are inconsistent with key predictions of schema-based theories of conceptual combination. Although the results do not contradict key assumptions of relationbased theories, modifications to these theories are required to account for these data.

Component processes and the utility of the conceptually-driven/data-driven distinction.

Three experiments tested conceptually-driven/data-driven distinction as an account of memory dissociations

Context Effects in Symbolic Magnitude Comparisons

Five experiments demonstrate that context has an effect on the ease with which people can determine the relative sizes of pairs of large and small animals. In a standard context, people are faster at choosing the larger of two large animals and the smaller of two small animals. However, when only pairs of small animals are presented (Experiment 1), relatively large pairs (RABBIT-BEAVER) are treated as if they were large animals and are discriminated more rapidly under the choose larger instruction. Similarly, when only large animals are presented (Experiment 2), the smaller of these are now treated as if they were small animals. Several models are presented that account for these effects of context, and these models are examined in subsequent experiments that impose yet other variations in magnitude pairings. The results demonstrate the importance of context in comparative judgement and place important constraints on theories of linear orders.

Event memory: the temporal organization of social action sequences.

We used comparative judgment procedures in two experiments to investigate the cognitive processes that mediate peoples reconstruction of the social events they read about. Subjects in Experiment 1 read a passage describing a series of behaviors manifested by a person in three situations. Subsequently, they were given pairs of these actions and were asked to judge either which action occurred sooner or which occurred later. These judgments were (a) faster when the behaviors being compared occurred near the middle of the situation to which they pertained than when they occurred near either the beginning or the end, (b) faster when the three situations were unrelated to one another than when they were thematically related, and (c) faster when the behaviors being compared occurred in different situations than when they occurred in the same situation. Actions were compared more quickly if they were far apart in the overall series presented than if they were close together, replicating the symbolic temporal distance effects obtained when scripted actions are judged on the basis of general knowledge. However, a semantic congruity effect (a tendency for actions near the beginning of the series to be discriminated more quickly when subjects are asked which comes sooner, but for actions near the end to be discriminated more quickly when subjects are asked which comes later) was not evident. In Experiment 2, subjects read a passage about a persons visit to a restaurant in which both generic actions (e.g., ordering the meal) and particularized actions (e.g., salting the fries) were described. Symbolic distance had a greater effect on judgments of particularized actions than on judgments of generic actions. Congruity effects were found only for judgments of generic actions. To account for these effects, a model of temporal order judgments is proposed that considers both the manner in which situation-specific actions are encoded into memory at the time they are learned and the process of comparing the actions at the time they are judged.