by Martha Stark, MD / Faculty, Harvard Medical School



Historically, memory consolidation (the term proposed more than 100 years ago) referred to the process whereby a memory, labile after its initial acquisition and therefore sensitive to being modified, would become increasingly stabilized over the long term and resistant to interference from competing or disrupting factors. In other words, once a short-term memory had become fixed, or consolidated, in long-term memory (that is, once the memory had undergone long-term potentiation), the memory would be permanently installed in the brain (Dewar et al. 2007).


The idea that deeply entrenched traumatic memories could be erased if certain conditions were met was actually the result of a cleverly designed study conducted over 50 years ago by the behavior therapists Rubin, Fried, and Franks (1969) – again, all the more remarkable because, back in the 1960s, most neuroscientists still embraced the idea that once a deep emotional learning had been acquired, it was forever.


The experiment devised by Richard Rubin and his team of investigators is now being widely quoted by proponents of memory reconsolidation; but it was not appreciated, at the time, for the significance that it was later to assume.


Rubin and his group came up with the ingenious idea of capitalizing upon the well-known retrograde amnesia “produced by the disruptive effect of electroconvulsive therapy (ECT) on memory trace consolidation” at the level of the memory trace’s “structural encoding” (Rubin et al. 1969, p. 37).


Earlier studies involving the use of ECT had generally been conducted on subjects who were anesthetized and therefore unconscious during their treatments. But Rubin and his colleagues formulated the following, testable hypothesis: “If the patient’s attention is strongly directed … to his most disturbing feelings and imagery … and if he is instantly given ECT (awake), there should result a significantly greater amelioration and reduction of symptoms than that obtained when ECT is given in the usual way” (Rubin et al. 1969, p. 39).


The investigators (Rubin et al. 1969) specifically selected subjects who were suffering from obsessions, delusions, and hallucinations. Then, in order to reactivate the neural mechanisms encoding the “psychopathological imagery” underlying these distorted perceptions of reality, the subjects were instructed to focus their attention on their symptoms. Rubin reasoned that having them focus their attention in this way would return the neural networks fueling the symptoms to a malleable state, which would render those circuits vulnerable to being disrupted. ECT was immediately administered with the subjects being kept awake throughout the treatment.


Of the fifteen patients in Rubin’s study (Rubin et al. 1969), seven had been previously treated anywhere from five to twenty-eight times, with routine ECT (and anesthesia) – but those treatments had been ineffective. These seven patients therefore served as their own controls.


Following a single treatment with ECT, all fifteen of the patients “improved dramatically for periods of three months to three years <the duration of the study>. One relapsed after nine months but recovered after another treatment” (Rubin et al. 1969, p. 40). The authors go on to write, “The probability of the effectiveness of this treatment resulting from chance factors only is less than 0.1%. This estimate is based on the number of previous, ineffective treatments” (Rubin et al. 1969, p. 40). To calculate this percent, they used Fisher’s exact test, more accurate than the chi-square test and recommended when the total sample size is less than 1000.


Rubin and his team astutely concluded that their study was proof that, at least in principle, the mental schemas fueling the symptoms must have been entirely obliterated because treatment with ECT prompted complete remission of symptoms only when those mental schemas were reactivated in subjects who were awake while given the electroconvulsive shock – and not when those mental schemas were lying dormant with the subjects unconscious.


Parenthetically, although Rubin’s brilliantly conceived study appeared to demonstrate eradication of old bad, it did not specifically address the introduction of new good. In other words, although the patients appeared to be released from the tyranny of their obsessions, delusions, and hallucinations, it was not clear what, if anything, got locked in, or reconsolidated, in the place of those pathological perceptions of reality.


That notwithstanding, Rubin’s trailblazing experiment was ingenious and, although largely ignored by the neuroscientific community until the 1990s, laid the groundwork for future research efforts.


Dewar M, Cowan N, Della Sala S. 2007. Forgetting due to retroactive interference: A fusion of Muller and Pilzecker’s (1900) early insights into everyday forgetting and recent research on anterograde amnesia. Cortex Jul;43(5):616-634.

Rubin RD, Fried F, Franks CM. 1969. New application of ECT. In RD Rubin & C Franks (eds.), Advances in Behavioral Therapy, 1968 (pp. 37-44). New York, NY: Academic Press.