Shock Contingency Research Articles

Neural correlates of conceptual-level fear generalization in posttraumatic stress disorder.

Posttraumatic stress disorder (PTSD) may develop when mechanisms for making accurate distinctions about threat relevance have gone awry. Generalization across conceptually related objects has been hypothesized based on clinical observation in PTSD, but the neural mechanisms remain unexplored. Recent trauma-exposed military veterans (n = 46) were grouped into PTSD (n = 23) and non-PTSD (n = 23). Participants learned to generalize fear across conceptual categories (animals or tools) of semantically related items that were partially reinforced by shock during functional magnetic resonance imaging. Conditioned fear learning was quantified by shock expectancy and skin conductance response (SCR). Relative to veteran controls, PTSD subjects exhibited a stronger neural response associated with fear generalization to the reinforced object category in the striatum, anterior cingulate cortex, amygdala, occipitotemporal cortex, and insula (Z > 2.3; p < 0.05; whole-brain corrected). Based on SCR, both groups generalized the shock contingency to the reinforced conceptual category, but learning was not significantly different between groups. We found that PTSD was associated with an enhanced neural response in fronto-limbic, midline, and occipitotemporal regions to a learned representation of threat that is based on previously established conceptual knowledge of the relationship between basic-level exemplars within a semantic category. Behaviorally, veterans with PTSD were somewhat slower to differentiate threat and safety categories as compared with trauma-exposed veteran controls owing in part to an initial overgeneralized behavioral response to the safe category. These results have implications for understanding how fear spreads across semantically related concepts in PTSD.

Speculation and explicit identification as judgmental standards for positive or negative reinforcement: A comment on Baron and Galizio (2005)

In a recent article published in The Behavior Analyst, Baron and Galizio (2005) revisited the question proposed by Michael (1975) on the distinction between positive and negative reinforcement (Skinner, 1953). They claimed that “the operational specification of which reinforcers are positive and which are negative contains an essential ambiguity” (p. 95). The conundrum presented by Baron and Galizio is represented in a case of a rat pressing a lever to turn on a heat lamp in a cold room (Weiss & Laties, 1961). This case is conventionally treated as positive reinforcement, because the lever pressing causes heat. Baron and Galizio, however, argue that one may regard it as negative reinforcement, because the lever pressing terminates coldness. Although this argument is logically sound, I prefer the first view, because the heat is a physical entity in the rat's environment, but “coldness” is a hypothetical state of the rat and is also a secondary concept meaning “lack of heat.” Other ambiguous cases of positive reinforcement presented by Baron and Galizio (2005) include a rat pressing a lever for delivery of a food pellet (or removal of a state of food deprivation, i.e., “hunger”), people working for delivery of money (or ending of a moneyless period, i.e., “poverty”), a child turning on the TV to watch a cartoon program (or termination of “boredom”). Baron and Galizio also presented ambiguous cases of negative reinforcement such as an animal making response for removal of shock (or the onset of the stimuli correlated with safety). I agree that there exists an essential ambiguity in specification of positive and negative reinforcement. But, we should have general standards for decision concerning the positive and negative types of reinforcement, and I feel that Baron and Galizio (2005) neglect them in the major part of their article. The judgmental standards distinguish between speculation and identification. As behavior analysts, we should prefer less speculation and more explicit identification (Skinner, 1950). In the case of the rat making a lever press for a food pellet, the pellet is countable and we can directly manipulate it. However, the “hunger” is a hypothetical construct (i.e., a guess of the rat's internal state). Money is also material for people, but “poverty” is secondarily defined. The cartoon program is on the air, but the termination of a child's “boredom” is not identified (i.e., if it exists, its termination is attainable by other TV programs). The removal of shock for a rat in a shuttle box is easily controlled and tested by an experimenter, but the stimuli correlated with safety are harder to identify specifically unless we preliminarily arranged a cue–shock contingency. Thus, categorizing this case as positive reinforcement is awkward (I do not mean that such analysis is meaningless, however; see below). Taking the “less speculation, more explicit identification” approach to the categorization conundrum has a practical advantage, because it permits behavior analysts to manipulate the environment directly rather than to burden such changes with speculation. It is noteworthy here that Baron and Galizio (2005) acknowledged this point: “When we are talking about reinforcers we are referring to environmental events rather than cognitive or physiological happenings” (p. 98). They seem to overlook that perspective in their discussion of the positive–negative distinction. Of course, there is some merit to the “two-sides” approach. Both positive and negative types of reinforcement may be operative in a single case. For example, as noted by Baron and Galizio (2005), self-injurious behavior (Iwata et al., 1994) can be maintained by escape from task demands (negative reinforcement) and attention of others (positive reinforcement). I disagree, however, with Baron and Galizio's claim that attention from others may be equivalent to relief from loneliness. “Relief from loneliness” is speculative, in that it depends on an inference about a behavioral or cognitive state. Similarly, it is not simple to equate escape from an aversive task with access to an alternative activity, because the latter is not well identified. Admittedly, examining a case from two perspectives might bring better understanding of a particular case and provide additional options for therapeutic interventions in the field of applied behavior analysis. I recommend, nevertheless, a “single-side” view as a first approach. Without basic training in viewing a case from a fixed standpoint, students are likely to have a hard time learning to take a two-sided view. Thus, I think that we must keep the distinction between positive and negative reinforcement, especially in the introductory course of behavior analysis.

Intrathecal infusions of anisomycin impact the learning deficit but not the learning effect observed in spinal rats that have received instrumental training

Previous research has shown that spinally transected rats will learn to maintain a flexion response when administered shock contingent upon leg position. In short, a contingency is arranged between shock delivery and leg extension so that Master rats exhibit an increase in flexion duration that lasts throughout the training session. Furthermore, when Master rats are later tested they reacquire the flexion response in fewer trials, indicative of some savings. As a control, a second group of spinal rats (Yoked rats) are given shock irrespective of leg position (noncontingent shock). These animals fail to show the same increase in leg flexion duration. Interestingly, when Yoked rats are later tested with a shock contingency in place, they still fail to learn (learning deficit). The present experiments were designed to determine whether both forms of instrumental learning in spinal animals require de novo protein synthesis. As such, we administered various doses of anisomycin intrathecally prior to training. Additionally, spinal rats were trained and tested either immediately or 24 h after test. We found that only the highest dose of anisomycin (125 μg/μl) had an effect in Yoked animals that were tested 24 h after training. Specifically, the highest dose of anisomycin reversed the learning deficit in those animals. Moreover, anisomycin had a similar effect when administered prior to training and immediately following training, but not 6 h after training. Finally, the results demonstrated that the observed effect of anisomycin was not due to state-dependency.

Lorazepam reinstates punishment-suppressed remifentanil self-administration in rats

We recently described a reinstatement procedure that models relapse to drug abuse in cases where abstinence results from aversive consequences of drug use. The potential value of this punishment-based model of relapse depends on its sensitivity to relapse-inducing events that are ineffective in the widely used extinction-based model. It is known that certain drugs can have anti-punishment effects, but these drugs have not been tested in the punishment-based reinstatement procedure. Therefore, the effects of the benzodiazepine, lorazepam, were examined using punishment-based and extinction-based reinstatement procedures. Rats self-administered the opioid, remifentanil (4 microg/kg per infusion). Two punishment groups were trained with response-contingent footshock that suppressed baseline rates of responding to zero. In an extinction group, remifentanil delivery was discontinued, and baseline responding stabilized at a low rate (mean=0.06 responses/min). Lorazepam (0.08-10 mg/kg, IP) was given during test sessions with the shock contingency discontinued for both punishment groups. Remifentanil delivery was maintained during testing in one punishment group but not the other. Lorazepam reinstated self-administration responding in both punishment groups but not in the extinction group. Priming injections of heroin reinstated responding in both the punishment and extinction groups, but combining heroin and lorazepam did not enhance reinstatement. This is the first demonstration that a trigger for relapse may have different effects depending on whether aversive conditioning contributed to the achievement of abstinence. It may be important to consider potential anti-punishment effects of both abused drugs and therapeutic agents in the treatment of individuals with a history of drug abuse.

Subregion-specific differences in hippocampal activity between Delay and Trace fear conditioning: an immunohistochemical analysis

Lesions of the hippocampus attenuate acquisition of the tone–shock contingency in Trace, but not in Delay fear conditioning. These findings suggest that hippocampal regions are differentially involved in these two forms of fear conditioning. The present study was aimed at testing the hypothesis that hippocampal neurons are differentially activated during acquisition and retrieval of Delay versus Trace fear conditioning. Male C57BL/6J mice were exposed to eight tone–shock pairings (in Trace conditioning the shock came 30 s after the tone), and tested for immobility upon reexposure to contextual stimuli or to one tone presentation. Ten brain regions were analyzed by immunohistochemistry for inducible transcription factors (ITF) c-Fos and Zif268 1.5 h after training, context test or tone test. Acquisition of both Delay and Trace fear conditioning produced significant induction of c-Fos in the majority of brain regions analyzed compared to naive control animals. Importantly, Delay fear conditioning caused a higher increase of c-Fos expression in the CA3 region of the hippocampus compared to Trace-trained animals. After cue reexposure, Zif268 levels in the dentate gyrus of the hippocampus were higher in Trace-conditioned than in Delay-conditioned animals. In addition, reexposure-related c-Fos expression in the anterior cingulate cortex was significantly higher in Delay-conditioned animals than in Trace-conditioned animals. The present study confirms differential activation of hippocampal subregions in Delay and Trace fear conditioning.

Avoidance and classical conditioning of leg flexion in dogs

This study of leg flexion conditioning in dogs, which were trained under both avoidance and classical contingencies consecutively, confirmed and extended the results from the Wahlsten and Cole study (In Classical Conditioning II: Current Research and Theory, Appleton Century-Crofts, New York, 1972). Briefly, dogs were trained to asymptotic behavioral levels under either avoidance or classical contingencies with a CS-US interval of either three (3) or five (5) seconds where the unconditioned stimulus (US) was shock to the foreleg and the conditioned stimulus (CS) was 1000 Hz tone. The dogs were then switched to the other contingency (without any modification in the stimulus situation other than the shock contingency) and trained to asymptotic behavioral levels. The CS remained on for the entire CS-US period and terminated with the end of the scheduled interval. Under the classical contingency, the US occurred as scheduled on every trial regardless of the dog's behavior. Under the avoidance contingency, the US was prevented from occurring if the subject responded with a criterion leg lift during the CS-US interval. The only feedback to the dog of a successful performance was the leg lift itself. The results indicated that there were two different conditioned responses produced, one just after CS onset under the avoidance contingency, and one just before US onset under the classical contingency for both CS-US intervals. The findings were interpreted as supporting a single-factor informational view of learning and a neural model was presented

Therapeutic shock device (TSD): Clinical evaluation with self-injurious behaviors

A man with profound mental retardation and multiple topographies of severe self-injurious behavior (SIB) had been receiving contigent shock for SIB for 2 years before the study started. Shock was being delivered with a handheld shock stick (Hot Shot Sabre Six), which produced burns to the man's skin. SIB rate, without a shock contingency, was 10 responses/min. The Therapeutic Shock Device (TSD), worn by the client and remotely operated by a radio frequency signal, provided superior control of SIB (0.02 responses/min) compared with the shock stick (0.06 responses/min) without causing tissue damage. TSD treatment was introduced in a mixed multiple baseline design across times, settings, and behaviors. The client did not appear to find the TSD aversive. To the contrary, his behaviors indicated that he preferred to wear it. The TSD appeared to provide a substitute for restraint, the hypothesized reinforcer for the man's most frequent form of SIB.

Reconsideration of the role of competing responses in demonstrations of the interference effect (learned helplessness).

In Experiment 1a, rats trained to escape shock by performing a 2-s inactive response were less impaired on a subsequent 2-way shuttle response than their yoked counterparts that received inescapable shock. In contrast, in Experiment 1b, rats trained to escape shock by performing a longer duration inactive response were more impaired on the subsequent escape task than their inescapably shocked counterparts. In Experiment 2, the results of Experiments 1a and 1b were replicated, and the inactive responses performed during pretreatment by both the escapable and inescapable shock groups were assessed and correlated with test stage 2-way shuttle escape performance. These activity data indicate that inactivity during pretreatment shock in both escapable and inescapable shock groups was a highly reliable predictor of subsequent 2-way shuttle performance, irrespective of the pretreatment shock contingency to which these Ss were exposed.

Effect of experimentally induced chronic renal failure upon the behavior of rats.

Two groups of rats were submitted to a 2-hour training session under a free operant avoidance of shock contingency 23 weeks after surgery, which provoked a state of chronic renal failure (experimental group) or after a sham operation (control group). Subjects in the control group had superior performances than those from the experimental group when total of shocks received and total numbers of avoidance and escape responses were compared. Results of biochemical blood analyses obtained from samples taken immediately after the avoidance training session indicate that the experimental group showed higher plasma urea concentration, higher levels of creatinine and osmolarity. The data show that the procedure used to provoke chronic renal failure was successful, and results in general confirm and extend findings relative to the effects of acute renal failure on behavior.

Free-operant avoidance of alcohol: An analogue study of alcohol aversion

An experimental analysis of the effects of peripheral electric shock on free-operant avoidance of alcohol was performed on five hospitalized alcoholics. A single-case design in which the aversive stimulus was systematically presented and withdrawn was used with each subject. There was little or no avoidance responding by any of the subjects in the initial baseline (no shock) condition. During conditioning, four of the subjects required relatively high shock intensities before making any avoidance responses. Removing and reinstating the shock contingency demonstrated that the shock exerted precise control over two subjects' avoidance behavior, some control over one subject and little or no control over the other two. After 100% avoidance was established, each subject continued to avoid alcohol even when the shock contingency was removed. Although there appeared to be some generalization of the avoidance to the subjects' natural environment, there are serious questions about the treatment potential of the procedure because of the high shock intensities that were necessary to condition the avoidance.

Free-operant avoidance of alcohol

An experimental analysis of the effects of electric shock on free-operant avoidance of alcohol was conducted with a single subject design. The subject made no avoidance responses during the baseline condition, where no shock was delivered. By removing and reinstating the shock contingency, it was found that the shock did exert some control over the avoidance behavior. However, after the subject was conditioned to avoid 100 per cent of the shock-alcohol, he continued to avoid all of the available alcohol even when there was no threat of shock. This 100 per cent avoidance of alcohol in the experimental setting has been observed in follow-up sessions up to 6 months after conditioning, and there appeared to be generalization of the conditioning effects to the subject's everyday environment.

Lick-shock contingencies in the rat.

Hungry rats were allowed to lick an 8% sucrose solution and then one of four lick-shock contingency conditions was superimposed on the licking baseline. These conditions were: free-operant avoidance, free shock, punishment, and no shock. From highest to lowest response rates, the groups fell in the order-avoidance, no shock, free shock, and punishment. Lick rates adjusted rapidly to introduction and removal of the contingencies. Post-shock responding was lowest in the punishment condition and highest in the free shock condition. No method was found simultaneously to equate shock frequency and separate response rates for the three shock contingency conditions. Only small, or no, reductions in shock rate occurred over sessions under the free-operant avoidance schedule when the shock-shock interval was 10 sec but large reductions occurred when the shock-shock interval was reduced to either 1 or 2 sec.

Noninstrumental punishment and shock intensity in observational learning.

It was hypothesized that direct negative reinforcement to observers, solely contingent upon errors in a model's behavior, would facilitate or interfere with the observer's subsequent efforts to learn the task as a function of intensity of punishment (electric shock). 80 Ss were run in a factorial design varying shock intensity, the number of model errors and shock contingency. The task was a complex temporal maze which explicitly required the learning of a series of correct switch selections and implicitly required avoidance of punished switches. The findings indicated that relatively severe noninstrumental punishment disrupted vicarious acquisition of the task insofar as avoidance of punished decisions was concerned. The effects were independent of whether the shock was contingent or noncontingent. Marginal support was found for the contention that mild punishment would facilitate performance beyond the level to which no punishment would influence performance.

Reply to Bignami

There appears to be no dispute between BigJJami and me concerning what I believe to be the major points about muscarinic anticholinergic agents and passive avoidance in the rat. A particular demonstration of the absence of an effect on passive avoidance by such an agent (Meyers & Koenig, 1967) occurring within a context of studies using different behavioral situations and reporting a disruptive effect for these agents (Bignami, 1967; Bovet, Robustelli, & Bignami, 1965; Meyers, 1965; etc.) strongly suggests that anticholinergic drugs produce no general inability to avoid passively. Rather, it indicates that only in certain situations do anticholinergic drugs impair passive avoidance. The conditions under which a deficit will or will not be observed remains to be elaborated and, of course, the efforts to do this, which Bignami remarks are now underway in his laboratory, are to be welcomed. It is obvious, as Bignami indicates, that the presence of an external signal for the shock contingency does not guarantee the lack of an effect by anticholinergic agents on passive avoidance behavior. Thus, the discrepant findings observed in various studies which used externally-cued passive avoidance tasks must have