Immunization and helplessness phenomena in the rat in a nonaversive situation

G. P. MULLINS 0 A. H. WINEFIELD 0 0 The University ofAdelaide , Ade/aide, South Australia 5001 The performance of rats on a visual discrimination task was found to be impaired following experience with an insoluble problem. The experiment was designed in order to preclude the possibility that the effect could be attributed to the development of an incompatible response. Prior experience with a soluble problem significantly reduced the deleterious effects of the insoluble problem. The results were interpreted as providing evidence of learned helplessness and behavioral immunization in a context where no aversive stimulation was employed. - In recent years, evidence has accumulated which seems to show that the phenomenon of learned help lessness (Seligman & Maier, 1967) is demonstrable in a variety of species and does not depend on ex perience with a noxious stimulus. Studies by Bainbridge (1973) and Hiroto and Seligman (1975) have been cited as evidence that experience in an insoluble problem situation may be sufficient to produce impaired performance in both humans and rats following transfer to a different, soluble, prob lem. Unfortunately, Bainbridge's results are also consistent with a different interpretation. Because his rats developed strong position preferences in response to the insoluble problem and because his transfer tasks involved spatial mazes, it is arguable that the impaired performance that he found was due to the development of an incompatible response rather than to a general lowering of motivation (learned helplessness). It has also been shown that prior experience of control over events may have an "immunizing" effect which offsets the effects of subsequent lack of control (Seligman, Rosellini, & Kozak, 1975). In the study reported here, the effects of experi ence in an insoluble problem situation were studied with and without prior experience in a soluble problem situation. On the basis of the learned he1p1essness hypothesis, it was predicted that prior experience with a soluble problem would produce less inter ference with subsequent learning because of the immunizing effect of the pretreatment. In order to rule out the possibility that the ob served effects of the experimental operations might be attributable to the development of an incom patible response rather than to a decreased level of motivation, subjects were trained in a circular dis crimination apparatus (Winefield & Jeeves, 1970). This apparatus enables the experimenter to vary the likelihood that rats will develop position prefer ences which are incompatible with the consistent choice of the positive stimulus. For example, where 12 doors are available (alternately positive and negative), subjects tend to choose doors in the vicinity of their initial position but they do not restriet their choices to particular doors or particu lar positions (Winefield & Mullins, 1976). By con trast, where subjects are only permitted access to two adjacent doors, they tend to develop preferences for the right or left door, which is incompatible with consistent choice of the positive stimulus (inasmuch as the latter is sometimes on the left and sometimes on the right). On the basis of the findings of Mullins, Winefield, and Levy (1973, Experiment 3), it was anticipated that performance would be independent of the nu mb er of doors available to the subjects. METHOD Experimental Design A 2 by 2 by 2 factorial design was used. The three factors were: prior experience with a soluble problem vs. no such prior ex perience (immunization factor); experience vs. no experience with an insoluble problem (insoluble factor); and training with 2 doors vs. training with 12 doors (incompatible response factor). The experiment was conducted in three phases. In Phase I, half the subjects were trained to criterion on an easy visual dis crimination problem. In Phase 2, half the subjects were placed in an insoluble problem situation in which they were randomly reinforced on SOOJo of trials. In Phase 3, all subjects were trained to criterion on a different visual discrimination. Throughout all three phases, half the subjects had access to only 2 adjacent doors and half had access to 12 doors. The experimental conditions administered to the eight groups are summarized in Table I. Subjects The subjects were 72 experimentally naive male Wistar strain rats, approximately 100 days old at the beginning of pretraining. The subjects were on a 24-h feeding schedule and were main tained at 85070 of their ad-lib body weight throughout the ex periment. Water was available at all times. Apparatus All subjects were run in a circular apparatus, which has been described in detail e1sewhere(Winefield. 1974; Winefield & Jeeves, 1970). Briefly, each subject was placed on a central starting platImmunization/Insoluble/2 Door Immunization/Insoluble/ 12 Door Immunization/No Insoluble/2 Door Immunization/No Insoluble/ 12 Door No Immunization/Insoluble/2 Door No Immunization/Insoluble/ 12 Door No Immunization/No Insoluble/2 Door No Immunization/No Insoluble/12 Door Random Reinforcement Random Reinforcement form surrounded by 12 goal compartments, each containing food. The compartments could be entered by the subject's crossing a 7.5-cm air gap onto a 11.5-cm-deep shelf and going through a top-hinged door. Each door could be locked by means of a rotat ing arm, and different discriminanda could be clipped onto the front. The entire apparatus was surrounded by curtains, and lighting was provided by means of an overhead globe (60 W) situated direct1y above the starting platform. In this situation, extraneous visual cues, which have been shown to be relied upon by rats for spatial orientation (Restle, 1957), were eliminated. For purposes of comparison with the normal two-choice situa tion, 10 of the compartments could be screened off by means of a flexible piece of metal which was clipped onto the partitions between the goal compartments. Procedure Because of the number of subjects involved, the experiment was run in two stages, with 40 subjects in the first and 32 subjects in the second stage, All experimental conditions were represented in each stage. Pretraining. After 7 days of handling, du ring which subjects were reduced to 85070 of their ad-lib weights, preliminary training was begun. Subjects were randomly allocated ro groups and underwent 8 days of training with mid-gray doors in 12- and 2-door situations, respectively. Pretraining was continued until all the rats were responding confidently in the sense that they would leave the starting platform, open the goal compartment doors, and eat in the goalbox. The subjects were given experience with locked doors; position responding in 2-door subjects was discouraged by locking the door on the preferred side. Phase 1: Initial discrimination (immunization). Half the sub jects from both 2- and 12-door groups (18 from each) were trained to a criterion of 22/24 correct responses over (...truncated)