What is forgetting?

Both are types of interference theory — the idea that forgetting is caused by competing memories rather than decay — but they differ in direction. Proactive interference (PI): older learning interferes with the recall of newer material. Example: if you learned French for years, the French words may intrude when you try to recall new Spanish vocabulary. "Proactive" means the earlier learning acts forward in time into later recall. Retroactive interference (RI): newer learning disrupts recall of earlier memories. Example: after learning a new phone number, your old one becomes harder to recall. "Retroactive" means the newer learning acts backward in time. Underwood (1957) demonstrated PI at scale: participants who had previously learned many word lists in lab studies forgot a new list far faster than naive participants — the accumulation of prior lists produced substantial proactive interference.

Storage failure means the information was never durably encoded or has actually been lost from memory — the trace is gone. Retrieval failure means the information is stored but currently inaccessible because appropriate retrieval cues are absent. Tulving and Pearlstone (1966) provided a key demonstration: participants who could not freely recall category members could recall them when given the category names as cues. The information was "available" in storage but not "accessible" without cues — a distinction Tulving formalised as the availability-vs-accessibility distinction. This matters practically: many everyday "forgetting" episodes are retrieval failures. The classic tip-of-the-tongue experience — knowing that you know something but being unable to recall it — reflects retrieval failure rather than storage failure. Reinstating the original context (going back to the room where you had an idea) often enables retrieval.

Ebbinghaus (1885) memorised hundreds of lists of nonsense syllables and measured how much he retained after various delays using the "savings method" — measuring how much less time it took to relearn a list compared to the original learning time. His forgetting curve showed two key findings: 1. Forgetting is most rapid immediately after learning — the biggest drop occurs in the first hour. 2. The rate of forgetting slows progressively — after the steep initial decline, the curve flattens, with much slower loss over subsequent days and weeks. The forgetting curve is one of the most replicated findings in psychology. Cepeda et al. (2006) later showed it is well described by a power function. It is also the empirical basis for spaced repetition — the practice of reviewing material at expanding intervals to intercept the forgetting curve before substantial loss.

Forgetting is not simply a failure of memory — it serves important functions. Bjork (2011) proposed that forgetting old or outdated information is an adaptive mechanism that reduces interference with current knowledge. Inability to forget — as described in rare cases of highly superior autobiographical memory (HSAM) — does not produce superior overall cognition; instead it can produce difficulty in updating or revising memories. The finding that sleep actively clears metabolic waste from the brain (the glymphatic system) and consolidates important memories while allowing others to weaken suggests forgetting is an active, regulated process rather than passive decay. Selective forgetting of unwanted memories — studied in directed forgetting paradigms — shows that the brain has mechanisms to down-regulate access to specific memories. Transience (the progressive weakening of memory over time) may be adaptive by reducing the weight of outdated information on current decision-making.