Which of the following conclusions about occasion setters has been supported by data

Contexts have been viewed as either modulating the retrieval of conditioned stimulus–unconditioned stimulus (CS–US) and conditioned stimulus–no unconditioned stimulus (CS–NoUS) associations or having direct associations with the US (e.g., Urcelay & Miller, 2010). In the former framework, the context is regarded as modulating the association between two events (i.e., as a positive or negative occasion setter; Holland, 1983, 1992; Miller & Oberling, 1998). In the latter framework, the context is like a punctate CS that can compete with the target stimulus for behavioral control during training and summate with it at test. At least under some conditions, the context appears to control behavior like a CS and does not require additional assumptions about contexts having special properties. Although much of this work has used physical contexts because they are easily controlled, we should not forget that contexts can also be situated in time and the organism’s internal state (Bouton, 2010). For example, using alcohol as a negative discriminative cue has been shown to have inhibitory properties (Cunningham, 1979). This CS-like role of the context has been assessed in at least two ways: evaluating the influence of the context on responding to punctate cues in cue competition situations (e.g., Randich, 1981) and demonstrating that this competition is directly related to the measured excitatory value of the context (Urushihara & Miller, 2009). Thus, contexts are like punctate cues in that they can serve either as occasion setters or as CSs. Additionally, parameters that encourage occasion-setting properties in punctate cues could be viewed as instances of the potential context-like role of cues. Consequently, the distinction between occasion setters and simple Pavlovian CSs is not necessarily a matter of whether a stimulus is localized in time and space (i.e., a punctate cue) or diffuse (i.e., a context), but of how those stimuli are treated during training.

An extinction context, serving as a stimulus that enters into cue competition with a target punctate stimulus, has been shown to influence behavior on a test of extinction when testing occurs in an associatively neutral context (e.g., Laborda, Witnauer, & Miller, 2011). This suggests that, although the role of the context in extinction has been emphasized as one in which the context modulates retrieval of CS–US and CS–NoUS associations (Bouton, 1993), one should not overlook the potential influence of the context as a Pavlovian CS in these situations. The CS-like role of contexts is especially important for certain theoretical accounts of the renewal effect (e.g., Rescorla & Wagner, 1972). Renewal, the recovery of a conditioned response (CR) following extinction treatment (presentations of the CS-alone) when testing occurs outside the context of extinction, can be explained by viewing the extinction context as a conditioned inhibitor that protects the target cue from complete extinction (e.g., Rescorla, 2003) and suppresses responding to the target CS when testing occurs in the extinction context (e.g., when acquisition occurs in a different context from extinction and testing; an ABB control situation). Recovery is then expected when the assessment of responding occurs in a context in which extinction has not occurred (i.e., outside the inhibitory influence of the extinction context; e.g., when acquisition, extinction, and testing all occur in separate contexts, an ABC renewal situation). Simply put, if the context of extinction is a functional conditioned inhibitor, negative summation should occur between the target cue and the test context in the ABB condition, whereas if testing occurs in an associatively neutral context (i.e., ABC renewal), no such reduction in responding should occur, and greater responding is expected. However, it is generally assumed that renewal is not the result of contextual conditioned inhibition, but rather that the context modulates responding much like an occasion setter.

A major problem with an account of renewal based on contextual conditioned inhibition is that all previous attempts to detect the inhibitory properties of extinction contexts have failed. For example, Bouton and King (1983; Bouton & Swartzentruber, 1986; see also Nelson, Sanjuan, Vadillo-Ruiz, Pérez, & León, 2011) assessed whether the context of extinction could pass a summation test (Rescorla, 1969) with a transfer excitor that was not extinguished, and they also assessed the associative value of the context using context preference scores. In both assessments, no evidence for inhibitory properties of the context was found, despite their observing renewal effects. Thus, Bouton and King’s findings are problematic for the view that renewal can be entirely explained by the extinction context becoming a conditioned inhibitor. Certainly, their findings indicate that at least in some preparations an extinction context functions as a negative occasion setter rather than a Pavlovian inhibitor in producing renewal. However, there are a number of considerations suggesting that the role of the context as a conditioned inhibitor should not be categorically rejected as a potential explanation of some of the published renewal data. For instance, Bouton and King used a cue that had a relatively strong associative status as a transfer excitor in their summation test. If the context of extinction was working as an inhibitor that summated with the extinguished target CS in the ABB condition, then for maximum sensitivity the appropriate summation test should be conducted with a stimulus having an associative history parallel to that of the target cue. These demonstrations have led some to define renewal as recovery that occurs in the absence of inhibitory or excitory contexts (e.g., Nelson et al., 2011). This confuses the empirical phenomenon of renewal with one particular mechanism that is thought to contribute to the phenomenon under some circumstances (i.e., occasion setting by the extinction context). We propose that the extinction context becoming a conditioned inhibitor is yet another potential mechanism that under some circumstances can contribute to the renewal phenomenon.

The present experiments assessed whether extinction treatment outside of the acquisition context enables the extinction context to become a conditioned inhibitor, at least with some parameters. Specifically, the present experiments were designed to evaluate whether an extinction context could pass summation and retardation tests for inhibition using a design that approximates those commonly used to assess the conditioned inhibition properties of punctate cues. Additionally, we investigated whether the intertrial interval during extinction training plays a role in determining whether the extinction context becomes a conditioned inhibitor. According to the Rescorla–Wagner (1972) model, spaced intertrial intervals (like those used by Bouton & King, 1983) should allow any inhibitory properties of the context to extinguish (due to a negative US expectancy and an outcome during extinction that is zero), thereby producing a positive incremental shift in associative value (i.e., extinction of conditioned inhibition; Polack, Laborda, & Miller, 2011; but see Zimmer-Hart & Rescorla, 1974). In contrast, using massed extinction trials should limit the amount of intertrial time during extinction training, potentially preserving the inhibitory status of the context. Consequently, to manipulate the associative value of the context, we used massed and spaced extinction trials, which are analogous to the explicit context extinction procedures previously demonstrated to influence the associative value of contexts in extinction (Laborda et al., 2011).

Experiment 1

In the present experiment, we assessed whether a context in which a punctate excitor was extinguished can pass summation and retardation tests for conditioned inhibition. To understand prior observations of an extinction context failing to display inhibitory properties, we assessed whether total extinction-context exposure, which was manipulated using different extinction trial spacings, determines whether the extinction context becomes a conditioned inhibitor. Because we did not intend to directly compare inhibition to renewal in this series, we did not equate experience between the transfer excitor and the to-be-extinguished cue. In the introduction, we pointed out that failing to equate this experience might explain differences in sensitivity between those stimuli with respect to negative summation. In the present research, we selected parameters that would produce a moderate responding to our transfer excitor in order to achieve sensitivity to negative summation with the context. Future experiments designed to make direct comparisons between conditioned inhibition and renewal should control for experience with the tested stimuli, but that is not our present purpose. Here, we simply tried to identify whether an extinction context can ever function as a conditioned inhibitor under specific parameters.

Method

Subjects

The subjects were 24 male and 24 female experimentally naive Sprague-Dawley-descended rats obtained from our own breeding colony. Body-weight ranges were 243–306 g for males and 164–207 g for females. Subjects were randomly assigned to one of four groups (ns = 12), counterbalanced within groups for sex. The animals were individually housed in standard hanging stainless-steel wire-mesh cages in a vivarium maintained on a 16:8 h light:dark cycle. Experimental manipulations occurred near the middle portion of the light phase. The animals received free access to Purina Lab Chow, whereas water availability was limited to 30 min per day following a progressive deprivation schedule initiated 4 days prior to the start of the study. From the time of weaning until the start of the study, all animals were handled for 30 s, three times per week.

Apparatus

We used 24 experimental chambers, of three different types. Chamber O was 30 × 30 × 27 cm (l × w × h). The sidewalls of the chamber were made of stainless steel sheet metal, and the front wall, back wall, and ceiling of the chamber were made of clear Plexiglas. The floor was constructed of 0.3 cm diameter rods, spaced 1.3 cm center to center, and connected by NE-2 neon bulbs that allowed a 0.7-mA, 0.5-s constant-current footshock to be delivered by means of a high-voltage AC circuit in series with a 1.0-MΩ resistor. Each of 12 copies of Chamber O was housed in an environmental isolation chest that could be dimly illuminated by a houselight (1.12-W #1820 incandescent bulb) mounted on one wall of the experimental chamber.

Chamber R was rectangular, measuring 24.0 × 9.0 × 12.5 cm (l × w × h). The walls and ceiling were clear Plexiglas, and the floor was comprised of stainless steel rods measuring 0.5 cm in diameter, spaced 1.3 cm apart (center to center). The rods were connected by NE-2 bulbs, which allowed for the delivery of a 0.7-mA, 0.5-s constant-current footshock. Each of six copies of Chamber R was housed in separate light- and sound-attenuating environmental isolation chambers.

Chamber V was 27 cm long, 29.5 cm high, and 21.5 cm wide at the top, and 5.5 cm wide at the bottom. The floor was comprised of two 27-cm-long plates, 2 cm wide, with a 1.5-cm gap between the two plates. A 0.7-mA, 0.5-s constant-current footshock could be delivered through the metal walls and floor of the chamber. The ceiling was clear Plexiglas, the front and back walls were black Plexiglas, and the sidewalls were stainless steel. Each of six copies of Chamber V was housed in a separate sound- and light-attenuating environmental isolation chest.

Both R and V chambers could be equipped with a water-filled lick tube that extended 1 cm into a cylindrical niche, which was 4.5 cm in diameter, left–right centered, with its bottom 1.75 cm above the floor of the apparatus and 5.0 cm deep. In all chambers, there was a horizontal photobeam detector 1 cm in front of the lick tube that was broken whenever the subject licked the tube. Three 45-Ω speakers on the inside and back walls of all isolation chests could deliver a click train (6 Hz), a complex tone (450 and 550 Hz, presented simultaneously), and a white noise, all 6 dB above background. Ventilation fans in each enclosure provided a constant 76-dB background noise. All auditory cues were measured on the C-scale. A 10-s-duration click train served as our transfer excitor, X, and a 30-s-duration complex tone served as our extinguished cue, Z. A longer-duration stimulus was used for Z in order to facilitate a within-compound association with the extinction context, which we expected would favor making the context inhibitory. A 0.7-mA footshock of 0.5-s duration served as the US. The light intensities inside the three types of experimental chambers were approximately equal due to the difference in the opaqueness of the walls.

Context A consisted of an instance of Chamber O with the houselight (HL) off and a block of wood with two drops of methyl salicylate located inside the isolation chest. Context D consisted of an instance of Chamber O (different from the one used as Context A) with the HL on and a block of wood with two drops of banana concentrate located inside the isolation chest. The physical contexts used as Context B and C were counterbalanced between an instance of Chamber R and an instance of Chamber V. No odor cue was used for physical contexts B or C.

Procedures

Acclimation

On Day 1, all subjects received a 30-min session of acclimation to Context B and another to Context C, with a 2-h gap between sessions. The order of exposure was counterbalanced within groups. The lick tube was accessible in each context. Following this phase, the lick tubes were removed until the summation test.

Transfer excitor training

On Day 2, all subjects received four presentations of the transfer excitor X coterminating with the US, in a 60-min session in Context A. Stimulus onsets occurred at 6, 16, 36, and 51 min into the session.

Acquisition

On Days 3–5, all subjects received eight daily presentations of Z coterminating with the US, in a 120-min session in Context D. Stimulus onsets occurred at 4, 24, 39, 54, 64, 79, 89, and 109 min into the session.

Extinction

On Days 6–9, subjects in the massed condition (see Table 1) received 10 daily nonreinforced presentations of Z over a 6-min session in Context B (mean intertrial interval = 6 s). Stimulus onset occurred at 3, 39, 75, 111, 147, 183, 219, 255, 291, and 327 s into the session. The relatively short intertrial interval was intended to reduce extinction of the Context B–CS Z association during the intertrial interval. Subjects in the spaced condition received the same 10 daily nonreinforced presentations of Z in Context B, but over a 240-min session (mean intertrial interval = 23.5 min). Stimulus onset occurred at 12, 36, 60, 84, 108, 132, 156, 180, 204, and 228 min into the session.

Summation test

On Day 10, the lick tubes were reinserted into Contexts B and C, and all subjects were tested in either Context B or C for a summation test with transfer excitor X. The test stimulus (X) was presented immediately upon placing a subject into the context, and the time to drink for five cumulative seconds was recorded. We were interested in the response to the combined presentation of the excitatory punctate transfer cue (X) and the context. Consequently, we omitted any pre-CS period during the session in order to have simultaneous onset of the cue and the context. This procedure is an adaptation of that used by Urushihara and Miller (2009) to assess excitatory conditioning to the context alone. Subjects in the summation condition (Sum) were tested in Context B, whereas those in the retardation condition (Ret) were tested in Context C. The sessions were 15 min in duration, with an upper limit of 15 min recorded for subjects that did not complete five cumulative seconds of drinking. Higher lick latencies represent a behavioral measure of greater expectancy of the aversive shock (i.e., freezing). All test scores were converted to log scores to better approximate the within-group normal distributions necessary for parametric statistical analysis.

Retardation training

On Day 11, subjects in Condition Sum received four US presentations at 65, 110, 165, and 230 s into a 260-s session in Context C, which was associatively neutral (i.e., the experimental subjects from the summation test were now the control subjects for the retardation test). Subjects in Condition Ret received the same treatment in Context B, the extinction context. The lick tubes were not accessible during this phase.

Retardation test

Prior to Day 12, the lick tubes were reinserted into Contexts B and C. On Day 12, all subjects were tested in the context of retardation training from the previous phase. Upon placing a subject in the chamber, the time to drink five cumulative seconds was recorded. The sessions were 30 min in duration, with an upper limit score of 30 min recorded for subjects that did not complete five cumulative seconds of drinking. As in the summation test, all test scores were converted to log scores to better approximate within-group normal distributions.

Results and discussion

Summation test

As can be seen in Fig. 1, Condition Sum stopped suppressing faster than Condition Ret, indicating that negative summation occurred when subjects were tested in the extinction context relative to an associatively neutral context. Negative summation was more pronounced when the extinction trials had been massed relative to when they had been spaced. This suggests that under some conditions an extinction context can pass a negative summation test for conditioned inhibition and that extinction trial spacing has an influence on the amount of negative summation observed. The following statistics support these conclusions.

Fig. 1

Mean log times to complete five cumulative seconds of licking during the summation test of Experiment 1. Brackets represent SEMs. White bars represent testing of X in the extinction context (B). Black bars represent testing of X in an associatively neutral context (C). Massed and spaced manipulations refer to trial spacing during extinction. See the text for details and Table 1 for the experimental design

Full size image

A 2 (Sum vs. Ret) × 2 (massed vs. spaced) ANOVA was conducted on the log lick latencies from the summation test. A two-tailed criterion of .05 was used to identify significant differences for both summation and retardation data analyses. A significant interaction, F(1, 44) = 10.04, MSE = .05, Cohen’s f = .44, was detected. Planned comparisons were conducted in order to determine the source of the interaction. Within the massed condition, we found that summation with the extinction context (Condition Sum) resulted in weaker suppression than summation with the neutral context (Condition Ret), F(1, 44) = 52.00. Weaker suppression to the extinction context was also observed within the spaced condition, F(1, 44) = 7.46. Negative summation with the extinction context produced less suppression in Group Sum–Massed than Group Sum–Spaced, F(1, 44) = 12.89; however, suppression in Condition Ret was not significantly influenced by the spacing of the extinction trials, F(1, 44) < 1.

Retardation test

As can be seen in Fig. 2, the results of this test were not entirely consistent with the results of the summation test. Those groups that received retardation training in the extinction context (i.e., Condition Ret) exhibited weaker suppression at test than those groups that received training in the control context (C; i.e., Condition Sum), and whether the extinction trials were massed or not appears to have had little influence on the level of suppression to the context at test. The following statistics support these assertions.

Fig. 2

Mean log times to complete five cumulative seconds of licking during the retardation test of Experiment 1. Brackets represent SEMs. White bars represent testing in the control context (C). Black bars represent testing in the extinction context (B). Massed and spaced manipulations refer to trial spacing during extinction. See the text for details and Table 1 for the experimental design

Full size image

A 2 (Sum vs. Ret) × 2 (massed vs. spaced) ANOVA was conducted on the log lick latencies from the retardation test. This analysis revealed a main effect of context of testing (Sum vs. Ret), F(1, 44) = 9.84, MSE = .51, Cohen’s f = .43, but no effect of the extinction trial spacing, F(1, 44) = 1.36, p > .24, and no interaction between these variables, F(1, 44) < 1. Planned comparisons revealed that retarded acquisition to the extinction context was observed when extinction trials were massed, F(1, 44) = 4.62, and when they were spaced, F(1, 44) = 5.22. The suggestion that conditioned inhibition may be more pronounced when extinction trials were massed rather than spaced, supported by the summation results, was not supported by the present retardation data. However, the observation of retardation in the extinction context but not in Context C supports the view that the extinction context became inhibitory. It should be noted that the prior summation test location perfectly confounded the retardation test, but there is no theoretical or empirical reason to think that the location of the summation test would influence retardation test performance.

Conclusions

The results of the summation and retardation tests demonstrated that an extinction context can become a conditioned inhibitor, at least with some parameters. The summation test suggested that one of the parameters that influence the extent to which an extinction context becomes inhibitory is the spacing of the extinction trials. However, this claim was not supported by the retardation data.

One possible reason why retarded acquisition to the extinction context was observed in both the massed and spaced conditions is that the spaced trials may have produced latent inhibition due to the extra exposure to that context (Lubow & Moore, 1959; Takigasaki, 1993). Thus, Group Ret–Massed may have exhibited retarded acquisition due to the intended conditioned inhibition, whereas Group Ret–Spaced may have showed retarded acquisition due to latent inhibition. This provides a plausible explanation for why responding was similarly retarded following massed and spaced extinction. To control for any influence of latent inhibition, Experiment 2 equated total exposure to the two testing contexts.

Experiment 2

The results of Experiment 1 suggested that the extinction context became inhibitory, but this inhibition may have been due to discriminative inhibition training, rather than to Pavlovian conditioned inhibition training as predicted by the Rescorla–Wagner (1972) model. That is, in Experiment 1 rats received shocks in Contexts A and D, but nonreinforced experience in Context B. If the contexts were being processed like punctate cues, this treatment would be analogous to sequential differential inhibition training. To determine whether the extinction treatment produced Pavlovian conditioned inhibition or differential training to Context B, we equated the numbers of stimulus presentations in each test context (B and C), and to eliminate the potential differences in latent inhibition to the two test contexts, the total amounts of exposure to Contexts B and C were also equivalent in Experiment 2. Reduced responding in Context B relative to Context C would support the conclusion that the results of Experiment 1 were due at least in part to the context becoming inhibitory through Pavlovian conditioned inhibition, with CS Z serving as the excitor for inhibition training (i.e., not due to prior nonreinforced experience of Context B). If inhibition was produced by discriminative conditioning alone, then Context B and Context C should be equivalently inhibitory.

Method

Subjects and apparatus

The subjects were 48 male and 48 female experimentally naive Sprague-Dawley-descended rats obtained from our own breeding colony. Body-weight ranges were 196–395 g for males and 175–246 g for females. The larger number of subjects in Experiment 2 relative to Experiment 1 was required in order to obtain results that did not depend on marginal significance. Experiment 2 was run in two replications that used identical numbers of subjects and parameters. The housing, apparatus, and contexts were as stated in Experiment 1. The only difference from Experiment 1 was that the tone and the white noise served as the extinguished stimulus Z or the control stimulus Y, counterbalanced within groups.

Procedures

Acclimation, transfer excitor training, and acquisition

On Days 1–5, acclimation, transfer excitor training, and the acquisition phase were conducted as described in Experiment 1, with the exception that during acquisition nonreinforced presentations of Stimulus Y were interspersed with the reinforced presentations of Z during acquisition (see Table 2). Y presentations occurred at 10, 20, 35, 50, 70, 85, 105, and 115 min into the three 120-min sessions.

Extinction

On Days 6–13, half of the subjects in each group received 10 nonreinforced presentations of Z in Context B on even-numbered days and 10 nonreinforced presentations of Y in Context C on odd-numbered days, whereas the other half received the same treatment with the days reversed. All subjects received extinction treatment of Stimulus Z with the same intertrial intervals used in Experiment 1, and Stimulus Y was presented equivalently to Stimulus Z. The lick tubes were not accessible during this phase.

Summation test, retardation training, and retardation test

On Days 14–16, the summation test, retardation training, and retardation test were conducted exactly as described in Experiment 1.

Results and discussion

Summation test

As can be seen in Fig. 3, negative summation occurred in the massed condition, thereby replicating the findings of the massed condition in Experiment 1. However, in the spaced condition there appeared to be a lack of any observable negative summation, indicating that in Experiment 1 the relative novelty of Context C may have influenced the results of the summation test. Importantly, the summation test data from both Experiments 1 and 2 demonstrated that an extinction context, under some conditions (e.g., massed trials during extinction), can pass a summation test for conditioned inhibition. The following statistics support these conclusions.

Fig. 3

Mean log times to complete five cumulative seconds of licking during the summation test of Experiment 2. Brackets represent SEMs. White bars represent testing of X in the extinction context (B). Black bars represent testing of X in an associatively neutral context (C). Massed and spaced manipulations refer to trial spacing during extinction. See the text for details and Table 2 for the experimental design

Full size image

A 2 (replication) × 2 (Sum vs. Ret) × 2 (massed vs. spaced) ANOVA was conducted on the log lick latencies from the summation test. A main effect of replication (p < .05) was observed, which arose from reduced suppression in Replication 2 relative to Replication 1. This effect was likely due to the greater weight of the rats in Replication 2. Importantly, the replication did not interact with either of the critical manipulations; therefore, it is omitted from the rest of the discussion of the results. We observed an interaction between context of testing and extinction trial spacing, F(1, 88) = 5.60, MSE = .10, Cohen’s f = .22. Planned comparisons were conducted in order to identify the source of the interaction. We observed less responding in Group Sum–Massed than in Group Ret–Massed, F(1, 88) = 5.11, demonstrating negative summation when subjects were tested in the extinction context after massed extinction training. However, when the extinction trials were spaced, there appeared to be no appreciable difference in responding based on the context of testing (i.e., Groups Sum–Spaced vs. Ret–Spaced), F(1, 88) = 1.15. As in Experiment 1, Group Sum–Massed demonstrated weaker responding (i.e., more negative summation) than Group Sum–Spaced, F(1, 88) = 10.72, whereas there was no observable difference between Group Ret–Massed and Group Ret–Spaced, F(1, 88) < 1.

Retardation test

As can be seen in Fig. 4, the data from the retardation test of Experiment 2 are congruent with the summation tests from Experiments 1 and 2. The extinction context exhibited retarded behavioral control (relative to a group in which contextual latent inhibition was controlled) after massed extinction but not following spaced extinction trials. This finding supports our speculation that in Experiment 1 retardation was observed in Group Ret–Spaced largely due to strong latent inhibition. The following statistics support this interpretation.

Fig. 4

Mean log times to complete five cumulative seconds of licking during the retardation test of Experiment 2. Brackets represent SEMs. White bars represent testing in the control context (C). Black bars represent testing in the extinction context (B). Massed and spaced manipulations refer to trial spacing during extinction. See the text for details and Table 2 for the experimental design

Full size image

A 2 (replication) × 2 (Sum vs. Ret) × 2 (massed vs. spaced) ANOVA was conducted on the log lick latencies from the retardation test. Again, we observed a main effect of replication, with reduced suppression in all groups of Replication 2, but a lack of any interaction with replication. Therefore, we further analyzed only the effects of the critical manipulations, test context and extinction trial spacing. The analysis revealed an interaction between extinction trial spacing and test context, F(1, 88) = 5.08, MSE = .35, Cohen’s f = .21. Planned comparisons were conducted to assess the source of the interaction. These planned comparisons revealed that in the spaced condition appreciable retardation did not occur, F(1, 88) < 1; however, when extinction trials were massed, responding to the context of extinction (Group Ret–Massed) was retarded in acquiring excitatory control of behavior, as compared to responding to the control context (Group Sum–Massed), F(1, 88) = 11.15. Additionally, Group Ret–Massed demonstrated greater retarded acquisition than Group Ret–Spaced, F(1, 88) = 21.52, whereas Groups Sum–Massed and Sum–Spaced did not differ, F(1, 88) < 1.

Conclusions

The summation and retardation data from Experiment 2 suggest that extinction of a punctate cue outside of the acquisition context can result in that extinction context acquiring inhibitory properties when extinction is conducted with massed trials. Moreover, Experiment 2 demonstrated that trial spacing during extinction influences the extinction context’s inhibitory potential. Massed extinction trials resulted in the extinction context passing both summation and retardation tests, whereas spaced extinction trials seemingly attenuated the context’s becoming a conditioned inhibitor.

General discussion

The present observations suggest that the frequently cited, but usually rejected, explanation of renewal in terms of the extinction context becoming inhibitory may in fact be valid in some situations. The present data suggest that extinction contexts can become inhibitory, at least when the extinction trials are massed. These results have long been proposed as a potential account of renewal, but to our knowledge, the extinction context becoming an inhibitor has never before been clearly demonstrated. Bouton and King (1983) failed to observe conditioned inhibition in their extinction contexts, presumably because they used parameters comparable to our spaced condition, in which negative summation was much weaker and only observed relative to a context that was relatively novel in addition to being associatively neutral. Spaced extinction trials are commonly used in an attempt to limit any potential inhibitory role of the extinction context. A possible distinction between spaced relative to massed extinction trials is that spaced trials may be perceived as more like sequential presentations of the context and the extinguished cue, whereas massed trials are perceived more like simultaneous presentations. Lamarre and Holland (1987; see also Holland & Lamarre, 1984) observed that negative discriminative stimuli failed to transfer to novel excitors if the discriminative stimuli had been presented sequentially with respect to the excitatory CS during training, but did transfer if they were presented simultaneously with the excitatory CS during training. Essentially, sequential presentations during discrimination training produced occasion setters, whereas simultaneous presentations produced conditioned inhibitors. Additionally, attentional mechanisms have been proposed to account for renewal effects (Larrauri & Schmajuk 2008), and reduced attention to the context in the spaced condition could possibly account for the present findings.

According to the Rescorla–Wagner (1972) model, conditioned inhibitors generate negative expectancy of the US, so exposure to the inhibitor alone generates a negatively valued expectancy of the US that is incompatible with the actual outcome (zero value). Consequently, the error correction rule should reduce the negative expectancy generated by the inhibitor (i.e., extinction of an inhibitor). The long intertrial intervals used by Bouton and King (1983) during extinction can be viewed as context (inhibitor) alone trials, which according to the Rescorla–Wagner model should extinguish the inhibitory strength of the extinction context. It should be noted that there are numerous failures to observe attenuation of a punctate conditioned inhibitor through additional exposure to the inhibitor alone (for a review, see Williams, Overmier, & LoLordo, 1992); however, it has been recently observed that massive exposure to a punctate inhibitor alone can attenuate its inhibitory properties (Polack et al., 2011).

Extinction of conditioned inhibition may explain why extinction contexts are rarely observed to be conditioned inhibitors, but this alone cannot explain why renewal is still observed when inhibition is not. This is problematic for attempts to explain all renewal through an inhibition-based account. In Polack et al.’s (2011) series of experiments, after a punctate inhibitor was massively exposed, it continued to negatively summate with the initial training excitor, but not with a transfer excitor that had training equivalent to that of the training excitor. This implies that extinction of a conditioned inhibitor does not completely eliminate inhibition, but instead makes the conditioned inhibitor more stimulus specific. This stimulus-specific inhibition differs from occasion setting in that occasion setters modulate novel cues that have previously been occasion set by other stimuli, whereas a stimulus-specific inhibitor failed to generalize to a comparably treated cue. Thus, previous failures to show that an extinction context is a conditioned inhibitor may be misleading. Additional exposure to the extinction context (e.g., long intertrial intervals during extinction) potentially attenuates negative summation between the extinction context and a novel excitor but has little impact on negative summation with the specific cue that was extinguished in that context. The stimulus-specific inhibition that remains after extinction of a conditioned inhibitor is one of the attributes that define occasion setters. This parallel blurs the distinction between Pavlovian inhibitors and negative modulators.

Lastly, a confound in Bouton and King’s (1983) study is that the control group received extinction training in the acquisition context and was then shifted to a neutral context to test the transfer excitor, relative to the experimental group, which received extinction in a neutral context. Consequently, the two groups received different experiences due to extinction occurring in contexts of different associative values. Extinction treatment administered in the acquisition context is known to produce a strong reduction in behavioral control, even when testing occurs outside of the extinction context (Laborda et al., 2011; Tamai & Nakajima, 2000; Tamai, Nakajima, Kitaguchi, & Imada, 2001). Thus, Bouton and King’s observation of renewal may be explained simply in terms of enhanced extinction produced by extinguishing a cue in the presence of an excitatory context (Rescorla, 2000; but see Pearce & Wilson, 1991).

The present experiments demonstrate that extinguishing a punctate cue in a neutral context can cause that context to become a conditioned inhibitor. Although the present experiments did not assess renewal of the extinguished cue, as Bouton and King (1983) had, they did demonstrate that the context was inhibitory when extinction trials were massed. The finding that trial spacing is a determinant of whether the context becomes a conditioned inhibitor, in conjunction with recent findings that exposure to the conditioned inhibitor alone attenuates negative summation with transfer excitors (Polack et al., 2011), supports accounts of renewal based on contextual conditioned inhibition (e.g., Rescorla & Wagner, 1972). As extinction contexts tend to become inhibitory only with massed extinction trials, the present data do not provide a full account of renewal in terms of the extinction context becoming a conditioned inhibitor. However, the present data indicate that, at least under some circumstances, the inhibitory value of the extinction context can contribute to renewal.

References

• Bouton, M. E. (1993). Context, time, and memory retrieval in the interference paradigms of Pavlovian learning. Psychological Bulletin, 114, 80–99. doi:10.1037/0033-2909.114.1.80.

  Article  PubMed  Google Scholar 

• Bouton, M. E. (2010). The multiple forms of “context” in associative learning theory. In B. Mesquita, L. F. Barrett, & E. R. Smith (Eds.), The mind in context (pp. 233–258). New York: Guilford Press.

  Google Scholar 

• Bouton, M. E., & King, D. A. (1983). Contextual control of the extinction of conditioned fear: Tests for the associative value of the context. Journal of Experimental Psychology: Animal Behavior Processes, 9, 248–265. doi:10.1037/0097-7403.9.3.248.

  Article  PubMed  Google Scholar 

• Bouton, M. E., & Swartzentruber, D. (1986). Analysis of the associative and occasion-setting properties of contexts participating in a Pavlovian discrimination. Journal of Experimental Psychology: Animal Behavior Processes, 12, 333–350. doi:10.1037/0097-7403.12.4.333.

  Article  Google Scholar 

• Cunningham, C. L. (1979). Alcohol as a cue for extinction: State dependency produced by conditioned inhibition. Animal Learning & Behavior, 7, 45–52. doi:10.3758/BF03209656.

  Article  Google Scholar 

• Holland, P. C. (1983). Occasion-setting in Pavlovian feature positive discriminations. In M. L. Commons, R. J. Herrnstein, & A. R. Wagner (Eds.), Quantitative analysis of behavior: Discrimination processes (Vol. 4, pp. 183–206). New York: Ballinger.

  Google Scholar 

• Holland, P. C. (1992). Occasion setting in Pavlovian conditioning. In D. Medin (Ed.), The psychology of learning and motivation (Vol. 28, pp. 69–125). San Diego: Academic Press.

  Google Scholar 

• Holland, P. C., & Lamarre, J. (1984). Transfer of inhibition after serial and simultaneous feature negative discrimination training. Learning and Motivation, 15, 219–243. doi:10.1016/0023-9690(84)90020-1.

  Article  Google Scholar 

• Laborda, M. A., Witnauer, J. E., & Miller, R. R. (2011). Contrasting AAC and ABC renewal: The role of contextual associations. Learning & Behavior, 39, 46–56. doi:10.3758/s13420-010-0007-1.

  Article  Google Scholar 

• Lamarre, J., & Holland, P. C. (1987). Transfer of inhibition after serial feature negative discrimination training. Learning and Motivation, 18, 319–342. doi:10.1016/0023-9690(87)90001-4.

  Article  Google Scholar 

• Larrauri, J. A., & Schmajuk, N. A. (2008). Attentional, associative, and configural mechanisms in extinction. Psychological Review, 115, 640–676. doi:10.1037/0033-295X.115.3.640.

  Article  PubMed  Google Scholar 

• Lubow, R. E., & Moore, A. U. (1959). Latent inhibition: The effect of nonreinforced pre-exposure to the conditional stimulus. Journal of Comparative and Physiological Psychology, 52, 415–419. doi:10.1037/h0046700.

  Article  PubMed  Google Scholar 

• Miller, R. R., & Oberling, P. (1998). Analogies between occasion setting and Pavlovian conditioning. In N. A. Schmajuk & P. C. Holland (Eds.), Occasion setting: Associative learning and cognition in animals (pp. 3–35). Washington, DC: American Psychological Association.

  Chapter  Google Scholar 

• Nelson, J. B., Sanjuan, M. C., Vadillo-Ruiz, S., Pérez, J., & León, S. P. (2011). Experimental renewal in human participants. Journal of Experimental Psychology: Animal Behavior Processes, 37, 58–70. doi:10.1037/a0020519.

  Article  PubMed  Google Scholar 

• Pearce, J. M., & Wilson, P. N. (1991). Effects of extinction with a compound conditioned stimulus. Journal of Experimental Psychology: Animal Behavior Processes, 17, 151–162. doi:10.1037/0097-7403.17.2.151.

  Article  Google Scholar 

• Polack, C. W., Laborda, M. A., & Miller, R. R. (2011). Extinction of a conditioned inhibitor. Manuscript submitted for publication.

• Randich, A. (1981). The US preexposure phenomenon in the conditioned suppression paradigm: A role for conditioned situational stimuli. Learning and Motivation, 12, 321–341. doi:10.1016/0023-9690(81)90012-6.

  Article  Google Scholar 

• Rescorla, R. A. (1969). Pavlovian conditioned inhibition. Psychological Bulletin, 72, 77–94. doi:10.1037/h0027760.

  Article  Google Scholar 

• Rescorla, R. A. (2000). Extinction can be enhanced by a concurrent excitor. Journal of Experimental Psychology: Animal Behavior Processes, 26, 251–260. doi:10.1037/0097-7403.26.3.251.

  Article  PubMed  Google Scholar 

• Rescorla, R. A. (2003). Protection from extinction. Learning & Behavior, 31, 124–132. doi:10.3758/BF03195975.

  Article  Google Scholar 

• Rescorla, R. A., & Wagner, A. R. (1972). A theory of Pavlovian conditioning: Variations in the effectiveness of reinforcement and nonreinforcement. In A. H. Black & W. F. Prokasy (Eds.), Classical conditioning II: Current research and theory (pp. 64–99). New York: Appleton-Century-Crofts.

  Google Scholar 

• Takigasaki, T. (1993). Effects of contextual preexposure on the context–shock conditioning in rats. Japanese Psychological Research, 35, 148–152.

  Google Scholar 

• Tamai, N., & Nakajima, S. (2000). Renewal of formerly conditioned fear in rats after extensive extinction training. International Journal of Comparative Psychology, 13, 137–146.

  Google Scholar 

• Tamai, N., Nakajima, S., Kitaguchi, K., & Imada, H. (2001). Renewal of extinguished fear by context-shifting in rats’ conditioned lick suppression. Japanese Journal of Psychology, 71, 493–497.

  Article  Google Scholar 

• Urcelay, G. P., & Miller, R. R. (2010). Two roles of the context in Pavlovian fear conditioning. Journal of Experimental Psychology: Animal Behavior Processes, 36, 268–280. doi:10.1037/a0017298.

  Article  PubMed  Google Scholar 

• Urushihara, K., & Miller, R. R. (2009). Stimulus competition between a discrete cue and a training context: Cue competition does not result from the division of a limited resource. Journal of Experimental Psychology: Animal Behavior Processes, 35, 197–211. doi:10.1037/a0013763.

  Article  PubMed  Google Scholar 

• Williams, D. A., Overmier, J. B., & LoLordo, V. M. (1992). A reevaluation of Rescorla’s early dictums about Pavlovian conditioned inhibition. Psychological Bulletin, 111, 275–290. doi:10.1037/0033-2909.111.2.275.

  Article  Google Scholar 

• Zimmer-Hart, C. L., & Rescorla, R. A. (1974). Extinction of a Pavlovian conditioned inhibition. Journal of Comparative and Physiological Psychology, 86, 837–845. doi:10.1037/h0036412.

  Article  PubMed  Google Scholar 

Download references

What is occasion setting in psychology?

Is compound potentiation a counterintuitive result quizlet?

What is an example of latent inhibition?

Is the term given to the difference between what the subject expects and what actually occurs?