Emergent relations in the formation of stimulus classes by pigeons.
Reflexivity, also referred to as identity, involves the principle that two stimuli are identical (i.e., Stimulus A is the same as Stimulus A). Reflexivity is demonstrated when following training in which Conditional Stimulus A is associated with Comparison Stimulus A, an emergent relation is found between Conditional Stimulus B and Comparison Stimulus B. Symmetry, also referred to as bidirectionality or backward associative learning, involves the principle that associations between two events function in both directions. Symmetry is demonstrated when, after training in which Stimulus A is associated with Stimulus B, one can show that an emergent relation exists between Stimulus B and Stimulus A. Transitivity, also referred to as mediation, can be demonstrated when, if Stimulus A is associated with Stimulus B, and Stimulus B is associated with Stimulus C, one can show that an emergent relation exists between Stimulus A and Stimulus C.
In a number of experiments, Zentall and Hogan (1974, 1975, 1976, 1978) have shown that when pigeons are trained on either an identity matching or mismatching task and are then transferred to the same task with novel stimuli, significantly better transfer performance is found when the training and transfer tasks are the same than when they are different. If, for example, the training task involves the matching or mismatching of red and green hues and the transfer task involves the matching or mismatching of yellow and blue hues, the differences in transfer performance between pigeons shifted to the other task (i.e., from matching to mismatching or from mismatching to matching), versus those that were nonshifted (i.e., from matching to matching or from mismatching to mismatching), can be quite large.
Even when transfer involves stimuli from a dimension different from that of the training stimuli (e.g., black and white shapes in training and distinctive colors in transfer, Zentall & Hogan, 1976, 1978, or different achromatic brightness values in training and distinctive colors in transfer, Zentall & Hogan, 1974, 1975), significant evidence for positive transfer can be demonstrated.
Similar results have been found with a successive matching task in which pigeons are trained to peck when the two halves of the response key are the same color and not to peck when they are colored differently (matching) or the reverse (mismatching), and transfer involves the same or the reverse task with novel stimuli (Zentall & Hogan, 1975). Thus, under a variety of conditions pigeons show clear evidence of forming emergent reflexive (identity) relations (see also Zentall, Edwards, Moore, & Hogan, 1981).
Although there is some evidence for symmetry or backward associative learning in animals based on Pavlovian-conditioning experiments (see e.g., Hearst, 1989; Spetch, Wilkie, & Pinel, 1981), there is little evidence for symmetry when conditional discriminations have been used (see D'Amato, Salmon, Loukas, & Tomie, 1985; Gray, 1966; Hogan & Zentall, 1977; Lipkens, Kop, & Matthijs, 1988; Richards, 1988; Rodewald, 1974; Sidman, Rauzin, Lazar, Cunningham, Tailby, & Carrigan, 1982). For example, when Hogan and Zentall (1977) trained pigeons to peck one comparison stimulus in the context of a particular conditional (or sample) stimulus and to peck the other comparison in the context of a different sample, little evidence of an emergent symmetric relation was found when the sample and comparison stimuli were interchanged.
Unfortunately, with these procedures not only is the temporal relation of the samples and comparisons reversed (during original training, the comparisons were presented only after the sample had been removed, whereas during test, it was the reverse), but also the location of those cues is switched (i.e., the samples that originally appeared on the center key now appear on the side keys and vice versa). In some animals, location of the sample stimulus in conditional discriminations appears to be an important attribute of the samples (see Iversen, Sidman, & Carrigan, 1986). Altering the location of training stimuli may thus account for the failure to find evidence for backward associative learning in monkeys (D'Amato et al., 1985) and pigeons (Gray, 1966; Hogan & Zentall, 1977; Richards, 1988; Rodewald, 1974). But even when pigeons have experienced individually presented stimuli in the locations appropriate to the transfer test, evidence for emergent symmetric relations has not been found (Lipkens et al., 1988). However, as these authors note, the context created by the individual presentation of stimuli may be quite different from that of a conditional discrimination.
In the most carefully designed study of this type, conditional discrimination training included preexposing monkeys to the transfer stimuli in locations appropriate to symmetry test and in the context of a conditional identity discrimination (Sidman et al., 1982). Yet, here too, little evidence for backward associative learning was found.
It is possible, however, that in many nonhuman species, the development of strong backward associations depends on the presence of a meaningful (e.g., hedonic) stimulus as one of the two associated events. We have recently tested this hypothesis by training pigeons on an identity matching task with differential outcomes. Correct responses to a red comparison were followed by a food outcome and correct responses to a green comparison were followed by a no-food outcome (Zentall, Sherburne, & Steirn, 1992). This training should have established red-food and green-no-food associations. In test, the red and green samples were replaced by food and no-food samples. Although the use of hedonic events as samples may seem unusual, pigeons learn this task quite readily (see Grant, 1991; Maki & Hegvik, 1980; Sherburne & Zentall, 1993). The design of this experiment is presented in Table 1.
TABULAR DATA OMITTED
As can be seen in Figure 2, evidence for backward associative learning was found in this experiment (i.e., food samples were associated with a preference for the red comparison, whereas no-food samples were associated with a preference for the green comparisons).
Historically, evidence for the demonstration of emergent transitive or mediated relations in nonhumans has come from two lines of research, sensory preconditioning and second-order conditioning. In sensory preconditioning research, original training involves the pairing of two "neutral" stimuli. In a second training phase, one of those stimuli is paired with an unconditioned stimulus. Finally, there is a test for the conditioned response to the remaining stimulus from original training (see Seidel, 1959, for a review). Second-order conditioning is similar to sensory preconditioning but the temporal order of the two training phases is reversed (see Rescorla, 1980, for a review).
More recently, evidence for emergent transitive relations has been reported in a number of studies involving either Pavlovian conditioning (Holland, 1981; Holland & Forbes, 1982) or conditional discriminations (Steirn, Jackson-Smith, & Zentall, 1991). Steirn et al. (1991) first trained pigeons in a simple successive (response independent) discrimination to associate a red stimulus, for example, with food and a green stimulus with no food. The pigeons were then trained on a conditional discrimination, to respond to a vertical comparison if the sample was food, and to a horizontal comparison if the sample was no food. Finally, subjects were presented with red and green samples and a choice between vertical and horizontal comparisons. The design of this experiment is presented in Table 2.
A significant difference in savings (Steirn et al., 1991, Experiment 1) and in transfer performance (Steirn et al., 1991, Experiment 2) was found between the positive transfer group (for which correct responses were consistent with the hypothesized mediated associations) and the negative transfer group (for which correct responses were inconsistent with those associations). The results (Steirn et al., 1991, Experiment 1) are presented in Figure 3. Thus, although no association between the hues and lines was explicitly trained, an emergent relation (apparently mediated by the food and no-food events) developed between them.
The kind of transitive relations described here is what D'Amato et al. (1985) referred to as associative transitivity (i.e., if Stimulus A is associated with Stimulus B, and Stimulus B is associated with Stimulus C, one may find evidence of an emergent relation between Stimulus A and Stimulus C). This kind of transitivity should not be confused with what D'Amato et al. call inferential transitivity (i.e., if Stimulus A is "greater than" Stimulus B and Stimulus B is "greater than" Stimulus C, subjects might correctly conclude that Stimulus A is "greater than" Stimulus C). We will discuss evidence for inferential transitivity in a later section.
To summarize, the research cited in this section indicates that pigeons are capable of demonstrating emergent relations involving each TABULAR DATA OMITTED of the three components of stimulus equivalence: reflexivity, symmetry, and transitivity. Although these phenomena have been demonstrated in separate experiments, in principle there is no reason to believe that more than one of these could not be demonstrated in pigeons in a single experiment. For example, Sidman and Tailby (1982) have suggested that symmetry and transitivity can be evaluated simultaneously using a matching design in which two samples (S1 and S2) are each associated with one comparison (C1) and two other samples (S3 and S4) are each associated with the other comparison (C2). According to Sidman and Tailby, the demonstration of an emergent relation between the two samples associated with the same comparison (i.e., many-to-one matching) would be evidence for both transitivity and symmetry. Transitivity would be indicated because the relation between samples must be mediated by the comparison common to both. Symmetry would be indicated because the path from one sample to the other must involve a backward association between the common comparison and the second sample.
Many-to-One Sample-to-Comparison Mapping
To demonstrate the formation of a stimulus class between two samples associated with the same comparison, one could ask if a direct association exists between the two samples. Spradlin and Saunders (1986) found that human subjects trained on many-to-one matching readily matched samples within each set to one another. Unfortunately, as already mentioned, replacing comparisons with previously trained samples generally leads to the severe disruption of matching performance in pigeons.
There are, however, alternative methods of assessing emergent relations between samples associated with the same comparison. One might expect, for example, that if many-to-one delayed-matching training results in an emergent relation between the pairs of samples associated with the same comparison, then training a new association involving one of those samples and a new comparison would generalize, without explicit training, to the other sample. Spradlin, Cotter, and Baxley (1973) trained human subjects on many-to-one matching. Then, after learning to match some of the samples from each purported stimulus class to new comparisons, the subjects proceeded to match the remaining samples to the new comparisons, as well.
In a similarly-designed experiment with pigeons, we asked if trained associations involving a subset of samples from a presumed stimulus class would transfer to the remaining samples in that class. The design of this experiment is presented in Table 3.
The results of this experiment show that performance on the untrained sample-comparison associations was better than chance when the tested associations were consistent with the presumed stimulus class, whereas performance was worse than chance when those associations were inconsistent with the presumed TABULAR DATA OMITTED class (Urcuioli, Zentall, Jackson-Smith, & Steirn, 1989). Thus, many-to-one matching training can result in an emergent relation between samples associated with the same comparison. The fact that a new association trained to one of those samples can generalize to the other, suggests that original training resulted in the formation of a stimulus class involving the two samples.
In a related experiment, Zentall, Sherburne, and Urcuioli (in press) trained pigeons on many-to-one delayed matching and then, as in the Urcuioli et al. (1989) study, trained them to associate two of the original samples (one associated with each of the original comparisons) with a new pair of comparisons. Following training with the new comparisons, delays were inserted between the offset of the sample and the onset of the comparisons. The remaining samples from original training were then inserted in the retention interval, and facilitation of delayed matching performance was found when the interpolated stimulus was compatible with the presumed stimulus class established during original training, whereas performance was disrupted when the interpolated stimulus and the stimulus class were incompatible. Furthermore, it was found that those pigeons that had learned the delay task slowly (median split) showed a general disruption in performance whenever either stimulus (compatible or incompatible) was inserted in the retention interval, whereas those pigeons that had learned the delay task quickly showed the expected facilitation and disruption of performance relative to that on control trials. Thus, there TABULAR DATA OMITTED may be individual differences in the degree to which emergent stimulus relations develop. The results of this experiment are presented in Figure 5.
The development of emergent relations between samples associated with the same comparison following many-to-one training can be shown with a number of other procedures, as well (Zentall, Sherburne, Steirn, Randall, Roper, & Urcuioli, 1992; Zentall, Sherburne, & Urcuioli, submitted; Zentall, Steirn, Sherburne, & Urcuioli, 1991; Zentall, Urcuioli, Jagielo, & Jackson-Smith, 1989). For example, matching tasks involving hue samples are generally acquired faster and are remembered better (i.e., show flatter retention functions when delays are inserted between the sample and comparison stimuli) than matching tasks involving line-orientation samples. Differences in the slopes of the retention functions for line-orientations versus hues have been demonstrated under a variety of between-group and within-subject conditions (Zentall et al., 1989). However, if following many-to-one matching training in which one line orientation and one hue are each associated with the same comparison those two samples come to form a stimulus class, one might expect the delay functions for those samples to have similar slopes. In fact, Zentall et al. (1989) reported that similar slopes were found for the line-sample and hue-sample delay functions following many-to-one matching training (but not following training with other mapping arrangements). See Table 5 for the design of this experiment. The results of this experiment are represented in Figure 6.
For related findings in which samples associated with the same comparison result in delay functions with similar slopes, see Zentall et al. (submitted).
One-to-Many Sample-to-Comparison Mapping
According to Sidman and Tailby (1982), emergent relations should be found not only between two samples associated with the same comparison (many-to-one mapping), but also between to comparisons associated with the same sample (one-to-many mapping). Evidence for emergent relations following one-to-many training has been inconsistent. On one hand, Wetherby, Karlan, and Spradlin (1983) reported strong evidence for emergent relations among comparisons associated with the same samples with human subjects, as did Zentall et al. (1992) using rate of reversal learning to assess emergent comparison relations in TABULAR DATA OMITTED pigeons. On the other hand, strong evidence for emergent stimulus relation between comparisons has not always been found in humans (Saunders, Wachter, & Spradlin, 1988; Spradlin & Saunders, 1986) and pigeons (Urcuioli & Zentall, 1993: this issue). Whether differences in emergent relations between samples and comparisons reflect methodological or theoretical differences, is not clear at this time.
Other Evidence for Emergent Relations in Pigeons
The research described here has focussed on evidence for emergent relations of the type required by a formal definition of stimulus equivalence. Two other kinds of emergent relations should be mentioned: functional stimulus class formation and inferential transitivity (or transitive inference).
Functional stimulus classes. Vaughan (1988) trained pigeons to respond to one (arbitrarily selected) set of stimuli (A+) and to abstain from responding to another stimulus set (B-). Following acquisition, the contingencies associated with each set were reversed (i.e., A-, B+), and then reversed again, and again, repeatedly. After a large number of reversals, the pigeons responded differentially to the positive and negative stimuli on any reversal following presentation of the first few stimuli in each set. Apparently, the pigeons had developed two functional stimulus classes involving the stimuli in each set, and once the first few stimuli in each set had been presented, thus allowing the current valence of each set to be determined (i.e., + or -), appropriate responding occurred to the remaining members of the set.
Hayes (1989) concluded that these data indicate the formation of functional stimulus classes but not equivalence sets because all the relevant associations had been trained previously. Whether this distinction is a psychologically meaningful one remains to be seen.
Transitive inference. Although transitive inference has been studied with a variety of procedures, considerable research conducted with a number of species, including children (e.g., Bryant & Trabasso, 1971), chimpanzees (Gillan, 1981), monkeys (McGonigle & Chalmers, 1977), rats (Davis, 1992), and pigeons (Fersen, Wynn, Delius, & Staddon, 1991) has involved training with at least four simultaneous stimulus discriminations of the type, A+ B-, B+ C-, C+ D-, D+ E-. Evidence for transitive inference has come from the finding that on test trials, Stimulus B is preferred over Stimulus D. Although the relation between associative and inferential transitivity needs to be clarified, they both clearly involve the demonstration of emergent stimulus relations.
The research described here demonstrates that, under a variety of conditions, pigeons can demonstrate emergent (untrained) relations among stimuli, and that such a demonstration suggests the formation of stimulus classes. It has been suggested that nonhumans may not be capable of the kinds of complex processes necessary for the development of stimulus equivalence, unless, perhaps, as in the case of certain chimpanzees, they have had language training (e.g., see Hayes, 1989, p. 391). Taken as whole, however, the research reported here suggests that such learning also can occur in a variety of nonhuman species.
BRYANT, P. E., & TRABASSO, T. (1971). Transitive inferences and memory in young children. Nature, 232, 456-458.
D'AMATO, M. R., SALMON, D. P., LOUKAS, E., & TOMIE, A. (1985). Symmetry and transitivity of conditional relations in monkeys (Cebus apella) and pigeons (Columba livia). Journal of the Experimental Analysis of Behavior, 44, 35-48.
DAVIS, H. (1992). Logical transitivity in animals. In W. K. Honig & J. G. Fetterman (Eds.), Cognitive aspects of stimulus control (pp. 405-429). Hillsdale, N J: Erlbaum.
FERSEN, L. von, 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.
GILLAN, D. J. (1981). Reasoning in the chimpanzee: II Transitive inference. Journal of Experimental Psychology: Animal Behavior Processes, 7, 150-164.
GRANT, D. S. (1991). Symmetrical and asymmetrical coding of food and no-food samples in delayed matching in pigeons. Journal of Experimental Psychology: Animal Behavior Processes, 17, 186-193.
GRAY, L. (1966). Backward association in pigeons. Pychonomic Science, 4, 333-334.
HAYES, S. C. (1989). Nonhumans have not yet shown stimulus equivalence. Journal of the Experimental Analysis of Behavior, 51, 385-392.
HEARST, E. (1989). Backward associations: Differential learning about stimuli that follow the presence versus the absence of food in pigeons. Animal Learning and Behavior, 17, 280-290.
HOGAN, D. E., & ZENTALL, T. R. (1977). Backward associations in the pigeon. American Journal of Psychology, 90, 3-15.
HOLLAND, P. C. (1981). Acquisition of representation-mediated conditioned food aversions. Learning and Motivation, 12, 1-18.
HOLLAND, P. C., & FORBES, D. T. (1982). Control of conditional discrimination performance by CS-evoked event representations. Animal Learning and Behavior, 10, 249-256.
IVERSEN, I. H., SIDMAN, M., & CARRIGAN, P. (1986). Stimulus definition in conditional discriminations. Journal of the Experimental Analysis of Behavior, 45, 297-304.
LIPKENS, R., KOP, P. F., & MATTHIJS, W. (1988). A test of symmetry and transitivity in the conditional discrimination performance of pigeons. Journal of the Experimental Analysis of Behavior, 49, 395-410.
McGONIGLE, B. O., & CHALMERS, M. (1977). Are monkeys logical? Nature, 267, 694-696.
MAKI, W. S., & HEGVIK, D. K. (1980). Directed forgetting in pigeons. Animal Learning and Behavior, 8, 567-574.
RESCORLA, R. A. (1980). Pavlovian second-order conditioning. Hillsdale, NJ: Lawrence Erlbaum Associates.
RICHARDS, R. W. (1988). The question of bidirectional associations in pigeons' learning of conditional discrimination tasks. Bulletin of the Psychonomic Society, 26, 577-580.
RODEWALD, H. K. (1974). Symbolic matching-to-sample by pigeons. Psychological Reports, 34, 987-990.
SAUNDERS, R. R., WACHTER, J., & SPRADLIN, J. E. (1988). Establishing auditory stimulus control over an eight-member equivalence class via conditional discrimination procedures. Journal of the Experimental Analysis of Behavior, 49, 95-115.
SEIDEL, R. J. (1959). A review of sensory preconditioning. Psychological Bulletin, 56, 58-73.
SHERBURNE, L. M., & ZENTALL, T. R. (1993). Coding of feature and no-feature events by pigeons in a delayed conditional discrimination. Animal Learning and Behavior, 21, 92-100.
SIDMAN, M. (1986). Functional analysis of emergent verbal classes. In T. Thompson & M. D. Zeiler (Eds.), Analysis and integration of behavioral units (pp. 213-145). Hillsdale, NJ: Erlbaum
SIDMAN, M., RAUZIN, R., LAZAR, R., CUNNINGHAM, S., TAILBY, W., & CARRIGAN, P. (1982). A search for symmetry in the conditional discriminations of rhesus monkeys, baboons, and children. Journal of the Experimental Analysis of Behavior, 37, 23-44.
SIDMAN, M., & TAILBY, W. (1982). Conditional discrimination vs. matching to sample: An expansion of the testing paradigm. Journal of the Experimental Analysis of Behavior, 37, 5-22.
SPETCH, M. L., WILKIE, D. M., & PINEL, J. P. J. (1981). Backward conditioning: A reevaluation of the empirical evidence. Psychological Bulletin, 89, 163-175.
SPRADLIN, J. E., COTTER, V. W., & BAXLEY, N. (1973). Establishing a conditional discrimination without direct training: A study of transfer with retarded adolescents. American Journal of Mental Deficiency, 77, 556-566.
SPRADLIN, J. E., & SAUNDERS, R. R. (1986). The development of stimulus classes using matching-to-sample procedures: Sample classification versus comparison classification. Analysis and Intervention in Mental Disabilities, 6, 41-58.
STEIRN, J. N., JACKSON-SMITH, P., & ZENTALL, T. R. (1991). Mediational use of internal representations of food and no-food events by pigeons. Learning and Motivation, 22, 353-365.
URCUIOLI, P. J., & ZENTALL, T. R. (1993: this issue). A test of comparison-stimulus substitutability following one-to-many matching by pigeons. The Psychological Record, 43, 745-760.
URCUIOLI, P. J., ZENTALL, T. R., JACKSON-SMITH, P., & STEIRN, J. N. (1989). Evidence for common coding in many-to-one matching: Retention, intertrial interference, and transfer. Journal of Experimental Psychology: Animal Behavior Processes, 15, 264-273.
VAUGHAN, W. Jr. (1988). Formation of equivalence sets in pigeons. Journal of Experimental Psychology: Animal Behavior Processes, 14, 36-42.
WETHERBY, B., KARLAN, G. R., & SPRADLIN, J. E. (1983). The development of derived stimulus relations through training in arbitrary-matching sequences. Journal of the Experimental Analysis of Behavior, 40, 69-78.
ZENTALL, T. R., EDWARDS, C. A., MOORE, B. S., & HOGAN, D. E. (1981). Identity: The basis for both matching and oddity learning in pigeons. Journal of Experimental Psychology: Animal Behavior Processes, 7, 70-86.
ZENTALL, T. R., & HOGAN, D. E. (1974). Abstract concept learning in the pigeon. Journal of Experimental Psychology, 102, 393-398.
ZENTALL, T. R., & HOGAN, D. E. (1975). Concept learning in the pigeon: Transfer of matching and nonmatching to new stimuli. American Journal of Psychology, 88, 233-244.
ZENTALL, T. R., & HOGAN, D. E. (1976). Pigeons can learn identity, difference, or both. Science, 191,408-409.
ZENTALL, T. R. & HOGAN, D. E. (1978). Same/different concept learning in the pigeon: The effect of negative instances and prior adaptation to the transfer stimuli. Journal of the Experimental Analysis of Behavior, 30, 177-186.
ZENTALL, T. R., SHERBURNE, L. M., & STEIRN, J. N. (1992). Development of excitatory backward associations during the establishment of forward associations in a delayed conditional discrimination by pigeons. Animal Learning and Behavior, 20, 199-206.
ZENTALL, T. R., SHERBURNE, L. M., STEIRN, J. N., RANDALL, C. K., ROPER, K. L., & URCUIOLI, P. J. (1992). Common coding in pigeons: Partial versus total reversals of one-to-many conditional discriminations. Animal Learning and Behavior, 20, 373-381.
ZENTALL, T. R., SHERBURNE, L. M., & URCUIOLI, P. J. (in press). Common coding in a many-to-one delayed matching task by pigeons as evidenced by facilitation and interference effects. Animal Learning and Behavior.
ZENTALL, T. R., SHERBURNE, L. M. & URCUIOLI, P. J. (submitted). Common coding of hedonic and nonhedonic samples in a many-to-one delayed matching-to-sample task.
ZENTALL, T. R., STEIRN, J. N., SHERBURNE, L. M., & URCUIOLI, P. J. (1991). Common coding in pigeons assessed through partial versus total reversals of many-to-one conditional and simple discriminations. Journal of Experimental Psychology: Animal Behavior Processes, 17, 194-201.
ZENTALL, T. R., URCUIOLI, P. J., JAGIELO, J. A., & JACKSON-SMITH, P. (1989). Interaction of sample dimension and sample-comparison mapping on pigeons' performance of delayed conditional discriminations. Animal Learning and Behavior, 17, 172-178.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Special Issue: Stimulus Equivalence|
|Author:||Zentall, Thomas R.; Urcuioli, Peter J.|
|Publication:||The Psychological Record|
|Date:||Sep 22, 1993|
|Previous Article:||Similarity- and nonsimilarity-based conceptualization in children and pigeons.|
|Next Article:||Stimulus relations in comparative primate perspective.|