Using their previously demonstrated gaze-following abilities, juvenile chimpanzees (and 3-year-old human children) were tested to determine if they interpreted seeing as the mental state of attention. The studies tested predictions generated by a low-level model of chimpanzee gaze-following which assumes that chimpanzees do not understand attention as an unobservable, internal mental state, and a high-level model which assumes that they do. In Expts 1 and 2, chimpanzees were rst trained to respond to a cup to which an experimenter pointed, and then tested on probe trials to determine if they could respond correctly when the experimenter either oriented his or her whole head toward the correct cup, or just looked with the eyes. In Expt 1 these cues were static, whereas in Expt 2 the experimenter actively moved his or her head and/or eyes back-and-forth form the subjects' faces to the correct cup as they were attempting to make their choice. Expt 3 validated the logic of Expts 1 and 2 by demonstrating that 3-year-old human children responded in a manner predicted by the high-level model. The results of the experiments converged on supporting the predictions of the low-level model of juvenile chimpanzees' understanding of seeing.

No abstract

Three studies examined the ubiquity of the bowed serial position effect in comparative judgments: the tendency for pairs of extreme magnitude to be discriminated more readily than pairs of intermediate magnitude. Although prior research has demonstrated that this effect occurs quite regularly in finite set experiments that repeatedly present a small number of items, there has been some ambiguity about the robustness of the bowed serial position effect in infinite set experiments, in which items are never repeated. Based on the extensive norms collected in Experiment 1, Experiment 2 demonstrated that a bowed serial position effect does in fact occur in a large infinite set experiment. The results of Experiment 3 indicate that this bowed serial position effect is not an artifact of our norms. The results are consistent with models that emphasize categorization of magnitudes and inconsistent with models that emphasize positional discriminability.

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.

Four experiments investigate a novel finding in the area of symbolic magnitude comparisons: Congruity effects may occur with subsets of objects. Such multiple congruity effects appear to signal the creation of size-ordered categories. Experiment 1 observed separate congruity effects for large and small pairs despite the intermingling of pairs within a session. Experiment 2 determined whether this result was an artifact of the items used. Experiments 3 and 4 examined whether linear separability on a dimension of size or on some other correlated dimension was a prerequisite for multiple size-ordered categorization. The results of these experiments suggest that congruity effects are properly regarded as indicating the presence of an organized structure or category. Thus, to the extent that congruity effects typify magnitude comparisons, the processing of relational information appears to implicate categorization.Research on comparative judgments in which subjects are asked to determine the greater or lesser of two objects has focused on whether the underlying magnitude representation is propositional or analogical. At the heart of this debate are three primary findings: the symbolic distance effect (Moyer & Landauer, 1967), whereby objects that differ greatly in magnitude are discriminated more easily than objects that differ little; the congruity effect (Banks, Clark, & Lucy, 1975), whereby the larger of two large objects is more readily discriminated than the smaller, but in small objects, the reverse is found; and the bowed serial position effect (Shoben, Cech, Schwanenflugel, & Sailor, 1989), whereby items of extreme magnitude are discriminated more readily than items of central magnitude.Rather than address the nature of the underlying representation, however, we examined the role of categorization in comparative judgments. Previous studies have shown apparent categorization effects in comparative judgments. We argue that the relevance of such effects has not been appreciated sufficiently. Before arguing this, however, we consider some of the findings relevant to categorization. One notable finding concerns the distance effect. Both Maki (1981) and Pliske and Smith (1979) found a moderation of the distance effect with comparisons of objects from noticeably different categories. Maki obtained distance effects for determining the relative easterliness or westerliness of cities located within a state, but not for cities located in neighboring states. Pliske and Smith similarly found no distance effect for determining relative intelligence when the members of a pair differed on a feature This research was supported in part by National Science Foundation Grants BNS82-17674 and BNS86-08215 to Edward J. Shoben.We are grateful to Larry Barsalou, Ruth Maki, and two anonymous reviewers for valuable comments on an earlier draft of this article and Peter Piazza for his help with the figures.