Spontaneous recovery is a medical phenomenon of learning and memory. This phenomenon was first coined and described by Ivan Pavlov in his studies of classical (Pavlovian) conditioning. In that context, it refers to the re-emergence of a previously extinguished conditioned response after a delay. The recovery of such lost behaviors can be observed in a variety of contexts, and the recovery of forgotten human memories is often of particular interest.Greene, R. L. (1992). Human memory: Paradigms and Paradoxes. New Jersey: Lawrence Erlbaum Associates, Inc.
In classical conditioning
Spontaneous recovery is associated with classical conditioning, a learning process in which an organism learns to associate a neutral stimulus with a stimulus that produces an unconditioned response. As a result, the previously neutral stimulus comes to produce its own response, which is usually similar to that produced by the unconditioned stimulus. Although aspects of classical conditioning had been noted by previous scholars, the first experimental analysis of the process was conducted by Ivan Pavlov, a nineteenth-century physiologist who came across the associative effects of conditioning while researching canine digestion.
To study digestion, Pavlov presented various types of food to dogs and measured their salivary response. He noticed that with repeated testing, the dogs began to salivate before the food was presented, such as when they heard the footsteps of an approaching experimenter. Pavlov named this anticipatory behavior the "conditioned" response. He and his associates discovered and published the basic facts about this process, which has come to be known as classical or Pavlovian conditioning.
Among the phenomena that Pavlov observed was the partial recovery of a classically conditioned response after it had been extinguished by withholding the unconditioned stimulus. This recovery happened without any further unconditioned stimulation, and Pavlov referred to it as spontaneous recovery. Although the chance of spontaneous recovery increases with time following the extinction procedure, such conditioned responses generally do not return to full strength. Moreover, with repeated recovery/extinction cycles, the conditioned response tends to be less intense with each period of recovery. However, reconditioning by pairing the same conditioned and unconditioned stimuli usually occurs much faster than the original conditioning.Schacter, D. L., D. T. Gilbert, and D. M. Wegner. Psychology. 2. New York, NY: Worth Publishers, 2011
A key conclusion is that conditioning leaves a lasting trace in the organism's memory, and that extinction does not erase the original conditioning but merely suppresses it. This insight has had profound implications for theories of classical conditioning and practical applications, particularly in behavioral therapy.
In human memory Retroactive interference Spontaneous recovery as it pertains to human memory can be traced back to the work of George Edward Briggs, who was concerned with the concept of retroactive interference. Inhibition, or interference, is a function of competition among responses, whereby one memory has dominance over another. The inhibited responses are not lost from memory per se, but are kept from appearing by other responses.
Retroactive interference is the psychological theory of memory whereby learning something new impedes retrieval of a previously learned memory. Briggs studied retroactive interference using a test of free recall. In his study, participants learned paired associate words (i.e. A1-B1, A2-B2,...Ai-Bi) over multiple trials, until the learning of the Ai-Bi associates was perfected. Following this, participants were given a new list of paired associates, where the second word of the pair was changed while the first word of the associate pair was kept the same (A1-C1, A2-C2,... Ai-Ci). After mastery of this second list, Briggs had participants perform a recall procedure. He presented a list A item and asked the participant to recall whichever pair (the -Bi or –Ci) that came to mind. Based on retroactive interference, learning of A-B paired associates declined due to the learning of the subsequent A-C associates, and as a result, there was a higher rate of responses from list C by the participants on the recall test.
One day (24 hours) after learning both sets of associates, participants were tested again. A spontaneous recovery of Bi responses was observed, such that participants' responses of Bi items exceeded Ci items. After a rest period, participants could spontaneously remember the initial paired associates that they were not able to remember following the subsequent presentation of a second paired associate list the day prior. This A-B, A-C paradigm was replicated by researcher Bruce R. Ekstrand, thus increasing confirmation regarding the existence of spontaneous recovery.
Role of sleep
In an experiment conducted to further the findings of Briggs and Ekstrand, it was discovered that sleep counteracts retroactive interference compared to wakefulness. Using the same A-B, A-C paradigm, results indicated that memory performance for the first list of word-pair associates (A-B) was superior when learning was followed by nocturnal sleep rather than wakefulness. Sleep differentially affected the memory consolidation of the two lists, enhancing the memory for the first list more significantly. It has been suggested that the degree of initial learning predicts whether spontaneous recovery shall occur, stating that the better the learning of the A-B associations, the more likely they are to recover after interference.Postman, L., Stark, K., & Fraser, J. (1969). Conditions of recovery after unlearning. Journal of Experimental Psychology Monograph, 82 This effect is further catalyzed by sleep, which appears to have an enhancing effect on memory consolidation.
Traumatic memories
Emotionally unpleasant experiences tend to recur, even after frequent suppression. Such memories can be recovered gradually, through active search and reconstruction, or they can come to mind spontaneously, without active search.
It is suggested that there are important differences between people who gradually recover memories of abuse during suggestive therapy sessions and those who recover memories of abuse more spontaneously. Research on the propensity to forget what they had just remembered was conducted on women who had spontaneously recovered memories of childhood sexual abuse, versus those who recovered them through suggestive therapy. Women who spontaneously recovered their abuse were more likely (than women who recovered their memories of childhood abuse during suggestive therapy) to forget that they had successfully remembered words earlier on a memory test. As such, a reason that people might have a spontaneously recovered traumatic memory experience is because they simply forget having remembered the event before. They might forget prior cases of remembering if the mental context present when they are having their recovery experience differs from the mental context on prior occasions when they thought about the event. Thus, people have not forgotten the event, but they simply cannot remember it, perhaps due to context-dependent memory.
Context-dependent extinction phenomena Renewal effect The renewal effect is a phenomenon in which a change of context after extinction can cause a robust return of conditioned responding. There are three models in which the renewal effect is observed.
The most common model is the "ABA renewal". Conditioning is conducted in one context (context A) and extinction is then conducted in a second (context B). When the conditional stimulus (CS) is returned to the original conditioning context (context A), responding to the CS returns. A second version, "ABC renewal" is where conditioning is conducted in context A, extinction is conducted in context B, and then testing is conducted in a third, "neutral" context (context C). Renewal of responding is observed in the neutral context. The final version, "AAB renewal" ,works whereby conditioning and extinction are both conducted in the same context (context A) and then the CS is tested in a second context (context B). This model currently has the least evidence of returned conditioned responding.
Since spontaneous recovery is defined as an extinguished response occurring after time has passed following the extinction, it can be viewed as the renewal effect that results when the CS is tested outside of its temporal context.
Role of GABA in renewal effects
Gamma-aminobutyric acid (GABA) transmission plays a significant role in the processes of renewal and spontaneous recovery from extinction. In a study that tested the spontaneous recovery of appetitive responding of rats, normal control rats responded more to a CS when it was tested in its acquisition context (this is ABA renewal) than its extinction context (ABB renewal condition). This is a classic example of ABA renewal. Administration of the partial inverse agonist, FG-7142, to the GABAA receptor resulted in attenuated recovery in both the ABA and ABB test scenarios. The GABAA receptor inverse agonist reduced GABA transmission, and the result was attenuation of the impact of context in eliciting the renewal effect. This research suggests that GABAergic mechanisms (mechanisms pertaining to neurons that produce GABA as their output) mediate ABA renewal and spontaneous recovery in appetitive conditioning. As such, spontaneous recovery and renewal are believed to share a common psychological mechanism.
Rapid reacquisition Reacquisition of a conditioned response (CR) post-extinction is often faster than the initial acquisition. This indicates that the original learning was not destroyed but rather was "saved" through the process of extinction.Flaherty, C. F. (1985). Animal learning and cognition. New York: Knopf Rapid reacquisition is normally produced when reinforced trials provide a contextual cue that can renew responding by signaling the initial acquisition trials.
Data from studies conducted on conditioned suppression provide evidence that this effect might occur due to spontaneous recovery and reinstatement of contextual stimuli that did not get extinguished, and thus remain related to the US.
Rapid reacquisition may be partially an effect of ABA renewal, whereby an animal (undergoing conditional learning) learns that the initial conditioning trials with the CS-US pairings are part of an original context of conditioning, whereas the more recent presentations of the CS alone are part of the context of extinction. Thus, when CS-US pairings are resumed post-extinction, they put the animal back into the mindset of the original conditioning context, and this contextual cue aids in reacquiring the CR more quickly than the initial learning. As such, rapid reacquisition will be produced when reinforced initial learning trials create a contextual cue, that in future acquisition trials gets activated and brings about the renewal of responding at a quicker rate than the initial response learning.
The process of extinction involves trials that are not reinforced, so that the CR of the animal declines when presented with the CS. A situation in which extinction procedures contain both reinforced and non-reinforced trials would weaken reacquisition, because introducing reinforced trials during extinction would result in the reinforced trials being associated both with extinction and conditional learning. This would render the animal less able to rapidly learn again the CR to the stimulus.
Reinstatement