Printer Friendly

Inference based on transitive relation in tree shrews (Tupaia belangeri) and rats (Rattus norvegicus) on a spatial discrimination task.

To use transitive inference (TI) is to deduce a conclusion from premises that have a nonequivalent relationship; for example, a conclusion "Jim is taller than Mike" is deduced from premises "Jim is taller than Bob" and "Bob is taller than Mike." This reasoning typically has been examined with verbal tasks. Bryant and Trabasso (1971) developed a nonverbal version of this task and showed that 4-year-old children can perform TI. In that task, children were trained to choose a longer (or shorter) one of the four pairs of colored rods of five different lengths, namely, A and B, B and C, C and D, and D and E (e.g., A is red, B is blue, C is yellow, D is green, and E is black), where the option on the left in each pair was longer (or shorter). After learning this task, the children tended to choose B rather than D when tested with the novel pair of B and D. Such transitive responding in similar nonverbal tests has been demonstrated in a wide range of mammalian and avian species, including adult humans (Siemann & Delius, 1996), chimpanzees (Boysen, Berntson, Shreyer, & Quigley, 1993; Gillan, 1981), monkeys (McGonigle & Chalmers, 1977, 1992; Rapp, Kansky, & Eichenbaum, 1996; Treichler, Raghanti, & Van Tilburg, 2003; Treichler & Van Tilburg, 1996, 1999), rats (Davis, 1992; Dusek & Eichenbaum, 1997; Roberts & Phelps, 1994; Takahashi & Fujita, 2003), pigeons (Higa & Staddon, 1993; Siemann, Delius, & Wright, 1996; Steirn, Weaver, & Zentall, 1995; Von Fersen, Wynne, Delius, & Staddon, 1991; Weaver, Steirn, & Zentall, 1997; Wynne, 1997), hooded crows (Lazareva, Smirnova, Rayevsky, & Zorina, 2000), scrub jays, and pinyon jays (Bond, Kamil, & Balada, 2003).

However, successful transitive responding in such typical TI tests may not suggest that animals actually perform TI. First, this type of testing may not necessarily provide a transitive situation. Markovits & Dumas (1992) argued that in reward-nonreward discrimination tasks, animals might not interpret A+B- as "A is better than B" but simply as "A goes to reward, B does not." In the latter case there is no room for TI to occur.

Second, some researchers argue that transitive responding could be generated from a difference in the strength of association of each item with rewards (See Couvillon & Bitterman, 1992). Couvillon and Bitterman suggest that probability of choice of one item over the other is determined by the relative strength of each paired item with rewards. Therefore, for instance, when the animal is performing at 80% accuracy, the ratio of the associative strength of the two items should be 4 to 1. Where there are five items, A, B, C, D, and E, and the animal's accuracy is hypothetically 80% for all the adjacent pairs, the associative strength of each item should be 0.8, 0.2, 0.05, 0.0125, and 0.003, respectively. That is, an animal might solve the TI tests simply by choosing items of larger association strength, not by inference.

To demonstrate that animals actually perform the TI in such tests, the problems above must be teased apart. In the present study, we used a spatial discrimination task with a semicircular eight-arm radial maze. The arms of the radial maze were labeled A, B, C, D, E, F, G, and H from right to left, or vice versa. We trained animals on four pairwise discrimination problems for the first five arms, just like ordinary TI tasks, then tested the animals with unused arms. In theory, this procedure provides two important merits in analyzing TI. First, animals can learn not only simple association of each arm with rewards but also relative strength of each arm by abstracting the rule of "right is better than left," a transitive relation. Second, by testing animals in novel arms, we can exclude the possibility that differential previous experience with rewards determines the animal's transitive choice. Thus our first purpose was to test whether animals had inferential ability similar to that of humans.

The second purpose of this study was to explore a factor driving evolution for TI. Previous researchers have suggested two types of adaptive significance in TI. One is a merit in social cognition (Bond et al., 2003; Hogue, Beaugrand, & Lague, 1996); TI helps animals to recognize dominance ranking of the members of the group without observing all the contests among the members. This hypothesis predicts more advanced ability for TI in species that are more social. In fact, Bond et al. found that highly social pinyon jays do better in the typical TI task than closely related but less social scrub jays. Furthermore, domestic hens and scrub jays may actually use TI to recognize dominance among group members (Hogue et al., 1996; Paz-y-Mino, Bond, Kamil, & Balada, 2004). The other is a merit in spatial cognition (Roberts & Phelps, 1994). In fact, Dusek & Eichenbaum (1997) found that after creation of a lesion in the hippocampus, supposed to be related to spatial memory (Morris, Garrud, Rawlins, & O'Keefe, 1982), rats failed to show transitive responding. Roberts & Phelps (1994) found that the spatial layout of olfactory items influenced the TI performance of rats; when the items were arranged linearly (e.g., ABCDE), the animals showed transitive responding, but when the layout of the items was random (e.g., BACED), they did not. Despite these findings, which factor was more contributive was in dispute, since many researchers examined nonsolitary species. To know the adaptive value of TI, it would be important to compare solitary and nonsolitary species.

Tree shrews have been classified as prosimian or insectivorous on the basis of their similarity or dissimilarity to primates. Now they are classified in a separate order, Scandentia (Martin, 1990), placed between insectivores and primates. They are mostly diurnal, arboreal, and omnivorous. These characteristics are similar to those of many primates (Vandenbergh, 1963). However, tree shrews are notably different in sociality from most primates. They live in solitary or in pairs. In contrast to tree shrews, rats are highly social. Comparison with these two species should help to shed light on the evolutionary origin of TI.

Experiment 1

For experiment 1, we asked whether tree shrews and rats would abstract transitive relation in the training task. We examined this question by using the modified TI task described above to test the animals' generalization performance in reaction to novel item pairs.

Method

Subjects

Tree shrews. Six hand-reared tree shrews (Tupaia belangeri; T01, T03, T04, T05, T06, male; T02, female) participated. The ages of T01, T03, T04, T05, T06, and T02 were 1 year 11 months, 3 years 1 month, 2 years 9 months, 1 year 11 months, 1 year 5 months, and 1 year 11 months old, respectively, at the beginning of the experiment. They were donated by Aburahi Laboratories (Shionogi) and were kept individually at the Graduate School of Letters, Kyoto University. T01, T02, and T05 had previously undergone figure discrimination training with Wisconsin General Test Apparatus but had no experience with spatial discrimination. The other subjects were experimentally naive. They were kept individually in a home cage (27 cm wide x 37 cm long x 56 cm high). There were two perches and a potlike nest for small birds as bedding. The temperature of the room was kept at 25[degrees]C to 30[degrees]C. The humidity was maintained at more than 40%. The ratio of bright and dark cycle in the room was 12:12. Once a day, animals were fed (pellets for new world monkeys, a piece of banana, and a piece of baked egg) after the experiment. Body weights of the subjects were kept at about 90% of their free-feeding weights. Water was freely available in their home cages. After this study was ended, the animals were given food freely and were kept in our laboratory.

Rats. The participants were 8 male rats (Rattus norvegicus; numbered 1-8), purchased from a private trader (Shimizu Laboratory Supplies). They were kept individually in a home cage (24 cm wide x 40 cm long x 21 cm high). The temperature of the room was kept at 25[degrees]C to 30[degrees]C. The humidity was kept at more than 40%. The ratio of brightness to darkness in the room was 10:14. All rats were experimentally naive and were approximately 100 days old at the beginning of the experiment. Food (pellets for rats) was given after the experiment, once a day. Body weights of the subjects were kept at about 80% of their free-feeding weights. Water was freely available in their home cages. They were handled for 5 min a day before pretraining. After this study, they were given food freely and were kept in our laboratory.

Apparatus

The apparatus was a semicircular radial maze (Figure 1). The maze consisted of eight arms (10 cm wide x 50 cm long x 20 cm high), a central chamber (45 cm in diameter), and a start box (10 cm wide x 20 cm long x 20 cm high). At the end of each arm was a hole (2 cm in diameter), in which was placed a 45 mg milk pellet (Bio-Serve, Frenchton, NJ). All of the holes had two levels, and a milk pellet was placed on the lower one to exclude possible odor cues. The entrance to each arm had a sliding door. The doors were open when the arms were used as alternatives. A hurdle (10 cm wide x 5 cm long x 3 cm high) was placed 1 cm behind the sliding door. The ceiling of the apparatus was made of transparent Plexiglas. The interior of the apparatus was painted brown with randomly marked black spots so that animals could discriminate individual arms. In front of the start box was a guillotine door that was operated manually. Each arm was assigned a letter in the alphabetic series A through H from left to right for T01, T03, and T05 and from right to left for T02, T06, and T04. The apparatus was wiped with a rag between trials. The positions of the apparatus, the experimenter, and the furniture (desk, chair, monitor, etc.) were fixed throughout experiments. The same apparatus was used for both species.

[FIGURE 1 OMITTED]

Procedure

Pretraining. Tree shrews were trained to visit all of the eight arms so they would become habituated to the apparatus and the experimental setting. All of the entrance doors were open during pretraining. The trial ended when the subject visited all arms or 15 minutes had elapsed. During the first five sessions, only one trial was administered per session. During the sixth session, T01, T05, and T02 completed two trials and T02 completed one trial. After the seventh session, sessions consisted of three trials. The pretraining was continued for 30 sessions. The average number of visits before entering all of the eight arms in the last five sessions was significantly less than chance (21.76), as calculated by means of Monte Carlo simulation repeated for 10,000 trials (T01: 10.8, t14 = 18.16, p < 0.01; T02 : 10.93, t14 = 23.93, p < 0.01; T03 : 13.47, t14 = 9.17, p < 0.01; T04 : 11.73, t14 = 26.11, p < 0.01; T05 : 10.73, t14 = 20.81, p < 0.01; T06 : 13.33, t14 = 10.27, p < 0.01).

Rats received the same pretraining as tree shrews for 15 sessions. The average number of visits before entering all of the eight arms in the last five sessions was significantly less than chance (21.76), as calculated by means of the same analysis of tree shrews (minimum t14 = 11.18, p < 0.01).

Training. Tree shrews were trained on a series of two-choice spatial discriminations immediately after the pretraining phase. During training, only two adjacent arms were open. The task was to go into one of the two arms. The arm on the right was correct for half of the subjects (T02, T06, T04), while that on the left was correct for the other half (T01, T05, T03). Trials started when the guillotine door was opened and the subject went out of the start box. If the subject chose the correct arm, the animal was able to obtain a 45 mg milk pellet as a reward placed at the end of the arm. No food was placed in the incorrect arm. After the choice, the door of the unvisited arm was closed. During trials, the sliding door of the start box was kept open until subjects reentered the start box, which ended the trial. Intertrial intervals lasted 60 sec, during which the subjects waited in the start box for the next trial. No correction trial was administered after incorrect choices.

Tree shrews were trained on discrimination problems of A-B+, B-C+, C-D+, and D-E+, or A+B-, B+C-, C+D-, and D+E- in phases 1 through 7 (see Table 1). Different subjects were trained in different orders; for instance, in phase 1, T01 and T02 were trained A-B+, T03 and T04 were trained D+E-, and T05 and T06 were trained D-E+.

The rats received the same training as tree shrews. The arm on the right was correct for half of the subjects (No. 1 to No. 4), and that on the left was correct for the other half (No.5 to No. 8). The experimental design, numbers of trials, and the learning criterion in each phase for both tree shrews and rats are shown in Table 1.

Test. After reaching the criterion in the last phase (phase 7), the subjects were tested with novel arm pairs. Test sessions consisted of 12 trials of training problems and 8 trials of test problems. To examine whether the subjects learned the rule "right (left) is better than left (right)" in training, two trials of the FG pair and two trials of the GH pair were presented (rule test). Whether the subjects responded transitively was examined in four trials of the FH pair (TI test). The first and the last four trials of each session were baseline trials. The three types of trials, namely baseline, rule test, and TI test trials, were presented in a quasi-random order among the middle 12 trials. The subjects were rewarded no matter which arm they chose in test trials; that is, food was placed in both arms. Test sessions were repeated eight times.

Results and Discussion

All tree shrews and rats were trained to learn spatial discrimination among four pairs of arms of the apparatus. The sessions to criterion in each phase are shown in Table 2. The results of the test in tree shrews are shown in Figure 2. Binominal tests showed that all tree shrews chose correct arms significantly more often than chance in baseline trials (p < 0.01). In the rule tests, all subjects selected the arm that was consistent with the rule incorporated in training (p < 0.01). In the TI tests, all tree shrews showed transitive responding significantly more often than chance (p < 0.01). Some might argue that the responding in the tests could have been a simple preference of the arms. However, there was no such preference in terms of the order of entry into arms F and H, at least in pretraining (T01: the average order of entry H: 9.08, F: 7.40, t43 < 1; T02: H: 7.18, F: 5.87, t43 = 1.20, p > 0.05; T03: H: 7.12, F: 5.63, t43 = 1.32, p > 0.05; T04: H: 8.8, F = 8.82, t43 < 1; T05: H: 8.37, F: 8.04, t43 < 1; T06: H: 9.08, F: 740, t43 = 1.48, p > 0.05). Thus the possibility of preference seems unlikely.

The results of testing in rats are shown in Figure 3. Binominal tests showed that all rats chose correct arms significantly more often than chance (50%) in baseline trials (p < 0.05). In rule tests (FG and GH pairs), 6 rats selected arms that accorded with the original rule significantly more often than chance (50%; p < 0.05), whereas 2 rats selected arms that disagreed with the rule (p < 0.05). In the transitive tests (FH pair), 4 rats showed transitive responding significantly more often than chance (p < 0.05). In contrast, the other rats selected the arm that was consistent with transitive responding significantly less often than chance (p < 0.05). Some might argue that the responding in the tests could have been a simple preference of the arms. However, three rats that showed transitive responding had no such preference in terms of the order of entry into arms F and H, at least in pretraining (No. 1: H : 4.84, F : 4.80, t 24 < 1; No. 3: H : 1.24, F : 1.24, t 24 < 1; No. 5: H : 1.80, F : 1.80, t 24 < 1), though the fourth rat showed a marginally significant preference for H in pretraining (No. 6: H : 1.68, F : 1.32, t 24 = 1.98, p = 0.059).

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

These results suggest that all tree shrews and some rats responded transitively on the basis of the spatial relationships among arms, not on associative strength, whereas the others did not. The results obtained in rats were less persuasive for TI than those in tree shrews. Thus a conservative conclusion seems to be that the rats had learned each of the five conditional discriminations rather than the abstracted rule to choose the right (or left) arm of the pair globally.

Experiment 2

In Experiment 1, all tree shrews and some rats showed transitive responding when a nonadjacent novel pair of arms, FH, was presented. Although the responding was similar to the inference that "if F must be visited earlier than G, and G must be visited earlier than H, then F must be visited earlier than H", tree shrews might have adopted a noninferential strategy, "select the right (left) one among the two opened arms without regard whether they are adjacent or not" (two-arm strategy). In Experiment 2, we presented the same tree shrews three alternatives, F, G, and H (triadic test). If they had adopted the two-arm strategy in Experiment 1, they were not expected to show transitive responding in this triadic test.

Method

Subjects and Apparatus

The subjects and the apparatus were the same as in Experiment 1.

Procedure

Retraining. Subjects were retrained on the same problems in phase 7 of Experiment 1 until they attained the same learning criterion.

Test. After reaching the criterion in the retraining, the subjects were tested with three novel arms, FGH (triadic test). Test sessions consisted of 12 trials of baseline and 4 trials of the triadic test. As Experiment 1, the first and the last 4 trials of each session were baseline trials. The two types of trials were presented in a quasi-random order among the middle 8 trials. The subjects were rewarded no matter which arm they chose in test trials. Test sessions were repeated eight times.

Results and Discussion

The tree shrews reached the learning criterion of retraining after two to six sessions. The results of the test of tree shrews in Experiment 2 are shown in Figure 4. Binominal tests showed that all tree shrews chose correct arms significantly more often than chance (50%) in baseline trials (all ps < 0.01). As in Experiment 1, all tree shrews showed transitive responding significantly more often than chance (33%) in triadic tests (all ps < 0.01). These results suggested that no tree shrews adopted the two-arm strategy to solve the TI tests in Experiments 1 and 2.

[FIGURE 4 OMITTED]

Rats reached the learning criterion after 4 to 12 sessions of retraining. Results of the test for rats are shown in Figure 5. Binomial tests showed that all rats chose correct arms significantly more often than chance (50%) in baseline trials (p < 0.01). In the triadic tests, 2 rats selected the arm that was consistent with transitive responding significantly less often than chance (chance level = 33%, p < 0.05), whereas 5 rats showed transitive responding significantly more often than chance (p < 0.01). Particularly, 3 rats (No.1, No.4, and No.6) that showed transitive responding in Experiment 1 showed the same tendency in Experiment 2. These results suggest that the 3 successful rats did not apply the two-arm rule in Experiments 1 and 2.

[FIGURE 5 OMITTED]

General Discussion

In this study, we examined whether tree shrews and rats would show transitive responding in a modified TI task incorporating a series of pairwise spatial discriminations. In Experiment 1, tree shrews and rats were trained A-B+, B-C+, C-D+, C-D+, and D-E+ or vice versa, then tested with a novel item pair, FH. All tree shrews and some rats showed transitive responding. In Experiment 2, tree shrews and rats were tested with three alternatives, FGH. Again, all tree shrews and the three successful rats in Experiment 1 showed transitive responding in Experiment 2.

As was described briefly in the introduction, previous literature has proposed two noncognitive hypotheses that could account for apparent transitive responding in T1 observed in many species. One assumes that the difference in the strength of association of each test item with reward resulting from reward-nonreward history of each item could support the subjects' transitive choice (Couvillon & Bitterman, 1992; Wynne, 1995). The other was the value transfer theory (von Fersen et al, 1991), which posits that a value of each item transfers to the paired item whenever the items are presented together as alternatives during pairwise training. Neither of these hypotheses accounts for the transitive choices shown by the tree shrews and the 3 rats in this study because they were tested with arms never used in pairwise training.

The results of this experiment suggest that tree shrews and some rats are capable of TI. An alternative explanation is that the animals applied a "two-arm rule," that is, "choose F rather than H because the former is on the right of the pair." This "two-arm rule" does not explain transitive responding of our tree shrews and rats in Experiment 2, in which the animals were tested with three alternatives. A second alternative explanation may be that the animals applied a strategy to start searching from one extreme of apparatus, then to choose the first open arm they encounter. This strategy is consistent with the choice pattern of our subjects. But we believe that this explanation is unlikely, because casual observation identified no such stereotyped searching behavior in either species. Thus the choice patterns shown by our subjects support the view that these animals utilized TI to solve the test problems with untrained arms.

On the one hand, we used new methodology to examine TI; on the other hand, some may argue that the method examined a transposition or generalization of the learned rule. It was certain that transposition or abstract rule learning was included for solving the task, because it was a basis for the understanding of transitive relation. However, it would be difficult to explain the results of the triadic test only by the rule learning. If animals learned the rule and applied it to a triadic situation, they would be confused by the novel situation and the performance would decline. But the performance did not decline; it even increased for some rats. These results would suggest that the animals infer the test relation, based on transitive relation.

In this study, both rats and tree shrews showed TI. These results may give us a hint as to how this ability has evolved. Previous researchers have suggested two types of adaptive significance in TI. One is a merit in social cognition (Bond et al., 2003; Hogue et al., 1996); TI helps animals to recognize dominance ranking of the members of the group without observing all the contests among the members.

In the present study, tree shrews outperformed rats in all tests for transitive responding. The better performance of this less social species implies that the social pressure to recognize dominance ranking among others in daily encounters may not be as important as one might think in the evolution of TI. Bartolomucci, de Biurrun, & Fuchs (2001) found that searching strategy of tree shrews was more similar to that of capuchin monkeys than that of house mice. This finding and our results suggest that cognitive requirement in the spatial domain might have been a more important factor in the evolution of TI than cognitive requirement in the social domain.

It is certain that direct comparison in this study was not fair because of the differences in methodology. However, we believe that the findings in this study showed some aspects of adaptive value of TI. More systematic comparison among closely related species is required to solve the riddle of evolution of this cognitive ability.

References

BARTOLOMUCCI, A., DE BIURRUN, G., & FUCHS, E. (2001) How tree shrews (Tupaia belangeri) perform in a searching task: Evidence for strategy use. Journal of Comparative Psychology, 115, 344-350.

BOND, A. B., KAMIL, A. C., & BALADA, R. P. (2003). Social complexity and transitive inference in corvids. Animal Behaviour, 65, 479-487.

BOYSEN, S. T., BERNTSON, G. G., SHREYER, T. A., & QUIGLEY, K. S. (1993). Processing of ordinality and transitivity by chimpanzees (Pan troglodytes). Journal of Comparative Psychology, 107, 208-215.

BRYANT, P. E., & TRABASSO, T. (1971). Transitive inferences and memory in young children. Nature, 232, 456-458.

COUVILLON, P. A., & BITTERMAN, M. E. (1992). A conventional conditioning analysis of "transitive inference." Journal of Experimental Psychology: Animal Behavior Processes, 18, 308-310.

DAVIS, H. (1992). Transitive inference in rats (Rattus norvegicus). Journal of Comparative Psychology, 106, 342-349.

DUSEK, J. A., & EICHENBAUM, H. (1997). The hippocampus and memory for orderly stimulus relations. Proceeding of the National Academy of Science of the USA, 94, 7109-7114.

GILLAN, D. J. (1981). Reasoning in the chimpanzee: I. Transitive inference. Journal of Experimental Psychology: Animal Behavior Processes, 7, 150-164.

HIGA, J. J., & STADDON, J. E. R. (1993). "Transitive inference" in multiple conditional discriminations. Journal of the Experimental Analysis of Behavior, 59, 265-291.

HOGUE, M. E., BEAUGRAND, J. P., & LAGUE, P. C. (1996). Coherent use of information by hens observing their former dominant defeating or being defeated by a stranger. Behavioural Processes, 38, 241-252.

LAZAREVA, O. F., SMIRNOVA, A. A., RAYEVSKY, V. V., & ZORINA, Z. A. (2000). Transitive inference in hooded crows: preliminary data. Doklady Biological Science: Proceedings of the National Academy of Science of the USSR, 370, 30-32.

MARKOVITS, H., & DUMAS, C. (1992). Can pigeons really make transitive inferences? Journal of Experimental Psychology: Animal Behavior Processes, 18, 311-312.

MARTIN, R. D. (1990). Primate origin and evolution. London: Chapman and Hall.

MCGONIGLE, B., & CHALMERS, M. (1977). Are monkeys logical? Nature, 267, 694-696.

MCGONIGLE, B., & CHALMERS, M. (1992) Monkeys are rational! Quarterly Journal of Experimental Psychology, 45B, 189-228.

MORRIS, R. G. M., GARRUD, P., RAWLINS, J. N. P., & O'KEEFE, J. (1982). Place navigation impaired in rats with hippocampal lesions. Nature, 297, 681-683.

PAZ-Y-MINO, C. F., BOND, A. B., KAMIL, A. C., & BALDA, R. P. (2004). Pinyon jays use transitive inference to predict social dominance. Nature, 430, 778-781.

RAPP, P. R., KANSKY, M. T., & EICHENBAUM, H. (1996). Learning and memory for hierarchical relations in the monkeys: Effect of aging. Behavioral Neuroscience, 110, 887-897.

ROBERTS, W. A., & PHELPS, M. T. (1994). Transitive inference in rats: A test of the spatial coding hypothesis. Psychological Science, 5, 368-374.

SIEMANN, M., & DELIUS, J. D. (1996). Influences of task concreteness upon transitive responding in humans. Psychological Research, 59, 81-93.

SIEMANN, M., DELIUS, J. D., & WRIGHT, A. A. (1996). Transitive responding in pigeons: Influences of stimulus frequency and reinforcement history. Behavioural Processes, 37, 185-195.

STEIRN, J. N., WEAVER, J. E., & ZENTALL, T. R. (1995). Transitive inference in pigeons: Simplified procedures and a test of value transfer theory. Animal Learning and Behavior, 23, 76-82.

TAKAHASHI, M., & FUJITA, K. (2003). Transitive responding with visual stimuli in rats (Rattus norvegicus). The Japanese Journal of Animal Psychology, 53, 17-22.

TREICHLER, F. R., RAGHANTI, M. A., & VAN TILBURG, D. N. (2003). Linking of serially ordered lists by macaque monkeys (Macaca mulatta): List position influences. Journal of Comparative Psychology, 29, 211-221.

TREICHLER, F. R., & VAN TILBURG, D. (1996). Concurrent conditional discrimination tests of transitive inference by macaque monkeys: List linking. Journal of Experimental Psychology: Animal Behavior Processes, 22, 105-117.

TREICHLER, F. R., & VAN TILBURG, D. (1999). Training requirements and retention characteristics of serial list organization by macaque monkeys. Animal Cognition, 2, 235-244.

VANDENBERGH, J. G. (1963). Feeding, activity, and social behavior of the tree shrews, Tupaia glis, in a large outdoor enclosure. Folia Primatologica, 1, 199-207.

VON FERSEN, L., WYNNE, C. D. L., DELIUS, J. D., & STADDON, J. E. R. (1991). Transitive inference formation in pigeons. Journal of Experimental Psychology: Animal Behavior Processes, 17, 334-341.

WEAVER, J. E., STEIRN, J. N., & ZENTALL, T. R. (1997). Transitive inference in pigeons: Control for differential value transfer. Psychonomic Bulletin & Review, 4, 113-117.

WYNNE, C. D. L. (1995). Reinforcement accounts for transitive inference performance. Animal Learning and Behavior, 23, 207-217.

WYNNE, C. D. L. (1997). Pigeon transitive inference: Tests of simple accounts of a complex performance. Behavioural Processes, 39, 95-112.

Makoto Takahashi and Tomokazu Ushitani

Graduate School of Letters, Kyoto University, and the Japan Society for the Promotion of Science

Kazuo Fujita

Graduate School of Letters, Kyoto University

T. Ushitani is now at the Faculty of Letters, Chiba University.

This study was supported by Research Fellowships of the Japan Society of the Promotion of Science (JSPS) for Young Scientists to M. Takahashi and T. Ushitani, by Grants-in-Aid for Scientific Research Nos. 13410026, 14651020, and 17300085 from JSPS to K. Fujita, and by the 21st Century COE Program, D-10, from the Ministry of Education, Culture, Sports, Science, and Technology to Kyoto University.

Address correspondence to Makoto Takahashi, Department of Psychology, Graduate School of Letters, Kyoto University, Yoshida-honmachi, Sakyo, Kyoto, 606-8501, Japan. E-mail: takahashi@psy.bun.kyoto-u.ac.ip
Table 1 Design of Experiment 1

 Training
 Task
Phase T01, T02, Rats T05, T06 T03, T04

1 A-B+ D-E+ A+B-
2 B-C+ C-D+ B+C-
3 A-B+, B-C+ C-D+, D-E+ A+B-, B+C-
4 C-D+ B-C+ C+D-
5 A-B+, B-C+, C-D+ B-C+, C-D+, D-E+ A+B-, B+C-, C+D-
6 D-E+ A-B+ D+E-
7 A-B+, B-C+, C-D+, D-E+ A-B+, B-C+, C-D+, D-E+ A+B-, B+C-, C+D-,
 D+E-

 Training
Phase No. of Trials Criterion

1 16 14/16 x 2 sessions
2 16 14/16 x 2 sessions
3 8 7/8 x 2 sessions
4 16 14/16 x 2 sessions
5 6 5/6 x 3 sessions or 6/6 x 2 sessions
6 16 14/16 x 2 sessions
7 5 4/5 x 4 sessions

 Test Phase
 Task
Test Type T01, T02, Rats T05, T06 T03, T04

Base A-B+, B-C+, C-D+, A-B+, B-C+, C-D+, A+B-, B+C-, C+D-,
 D-E+ D-E+ D+E-
Rule F+G+, G+H+ F+G+, G+H+ F+G+, G+H+
Transitive F+H+ F+H+ F+H+

 Test Phase
Test Type No. of Trials Sessions

Base 3 8
Rule 2 8
Transitive 4 8

Note. "+" denotes correct alternative, and "-" denotes incorrect
alternative.

Table 2 Number of Training Sessions Before Reaching the Criterion for
Each Subject in Experiment 1

 Phase
 Subject 1 2 3 4 5 6 7

Tree shrews T01 8 5 2 3 2 2 4
 T02 18 8 7 3 9 3 4
 T03 5 6 16 5 27 6 8
 T04 12 3 2 2 5 3 5
 T05 4 3 52 4 7 4 6
 T06 13 2 21 3 27 7 18
Rats 1 5 3 2 3 8 3 9
 2 3 2 14 4 2 3 5
 3 2 2 21 3 2 2 6
 4 2 4 2 3 3 3 4
 5 4 3 2 3 2 3 14
 6 4 2 2 2 2 3 5
 7 4 2 15 4 19 7 4
 8 5 3 2 3 5 3 10
COPYRIGHT 2008 The Psychological Record
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2008 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Takahashi, Makoto; Ushitani, Tomokazu; Fujita, Kazuo
Publication:The Psychological Record
Article Type:Report
Geographic Code:9JAPA
Date:Mar 22, 2008
Words:5296
Previous Article:Choice and the initial delay to a reinforcer.
Next Article:A novel method for teaching the first instances of simple discrimination to nonverbal children with autism in a laboratory environment.
Topics:

Terms of use | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters