The case against decay

A term coined by accident

The vocabulary begins in Göttingen in 1900, when Georg Elias Müller and Alfons Pilzecker published a long monograph on the learning of nonsense syllables. Their observation was that material learned shortly after an earlier list was recalled less well than material learned after a rest, and that the earlier list appeared to suffer for it too. They named the effect retroactive inhibition, and they proposed that a memory requires a period of undisturbed settling — a process they called consolidation. The term outlived the theory that produced it and is still in use.

What Müller and Pilzecker described was not fading. It was disruption: the second list did something to the first. That distinction set up an argument that ran for the next half-century.

McGeoch's objection

In 1932 John A. McGeoch published a paper arguing that the passage of time explains nothing at all. His analogy has outlasted his data. Iron left in a shed develops rust, but time is not the cause — oxygen and moisture are. Time is the interval during which causes operate. A theory of forgetting that appeals to disuse, McGeoch argued, names the interval and mistakes it for the mechanism.

The alternative he proposed was that forgetting reflects competition among responses. When a cue has been associated with more than one response, recall becomes a contest, and the losing response is unavailable rather than absent. The prediction that follows is testable: fill an interval with material that competes, and retention should fall; empty the interval of competing material, and retention should hold.

The sleep comparison, and its sample

The experiment most often cited in support of that prediction was published in 1924 by John G. Jenkins and Karl Dallenbach at Cornell. Two participants learned lists of nonsense syllables and were tested after intervals of one, two, four and eight hours. In some conditions the interval was spent asleep; in others it was spent awake and going about ordinary activity. Retention after sleep was substantially better, and the gap widened as the interval lengthened. Jenkins and Dallenbach concluded that forgetting is less a matter of decay than of interference from ongoing experience.

The result is durable, widely replicated in broad outline, and much weaker as evidence than its reputation suggests. There were two participants. The sleep intervals were, by necessity, at night, and the waking intervals during the day, which leaves the comparison confounded with time of day and with everything that varies across the circadian cycle. Later work by Bruce Ekstrand and colleagues in the late 1960s and early 1970s complicated the picture further by reporting that when in the sleep period the interval falls affects retention — a result that a pure interference account does not obviously predict, and that pushed some researchers back toward a consolidation explanation of the kind Müller and Pilzecker had proposed.

John Wixted, reviewing this literature in 2004, argued that the sleep findings are better read as evidence about consolidation than as a clean demonstration that interference alone drives forgetting. The two accounts are not mutually exclusive, and the modern literature largely treats them as complementary rather than rival.

How much forgetting was the laboratory itself?

The sharpest version of the interference argument arrived in 1957, when Benton Underwood delivered an analysis of decades of published verbal-learning studies. Participants in these experiments typically learned many lists over many sessions. Underwood pointed out that recall after twenty-four hours varied enormously across studies, and that the variation tracked how many lists a participant had already learned. Those learning their first list retained a great deal. Those who had learned dozens retained very little.

The implication was uncomfortable. A large share of what the field had been measuring as forgetting was produced by the experiments themselves — prior lists interfering with later ones. Underwood's estimate of how much has been revised since, and the reanalysis rested on comparisons across studies that differed in many other ways, but the basic observation reoriented the field toward proactive as well as retroactive sources of interference.

Where the argument stands

Contemporary research does not treat decay and interference as a settled contest. Reviews by Oberauer and Lewandowsky and by Barrouillet and colleagues continue to disagree about whether any purely time-based loss occurs in short-term memory, and the disagreement turns on details of experimental design that resist easy adjudication. What has held up is McGeoch's narrower point: an interval by itself is not an explanation, and any account of forgetting has to specify what happens during it.

• Müller, G. E. & Pilzecker, A. (1900). Experimentelle Beiträge zur Lehre vom Gedächtnis. Zeitschrift für Psychologie, supplement 1.
• Jenkins, J. G. & Dallenbach, K. M. (1924). Obliviscence during sleep and waking. American Journal of Psychology, 35, 605–612.
• McGeoch, J. A. (1932). Forgetting and the law of disuse. Psychological Review, 39, 352–370.
• Underwood, B. J. (1957). Interference and forgetting. Psychological Review, 64, 49–60.
• Ekstrand, B. R. (1972). To sleep, perchance to dream. In C. P. Duncan et al. (Eds.), Human Memory.
• Wixted, J. T. (2004). The psychology and neuroscience of forgetting. Annual Review of Psychology, 55, 235–269.

Editorial note: this article summarises published findings. It is not advice, and it makes no claim about any individual reader.