Retrieval Failure

Interference Theory of Forgetting

Interference theory proposes that forgetting occurs not because memory traces simply decay or fade, but because other memories compete with the target memory, disrupting retrieval. There are two forms of interference, which differ in the direction of the disruption: proactive interference and retroactive interference. Both are most likely to occur when the competing memories are similar to each other.

Proactive Interference

Proactive interference (PI) occurs when older memories interfere with the recall of newer information. "Proactive" means that the older learning acts forward in time to disrupt what is learned later. A common everyday example is a person who has recently changed their PIN number: the old PIN keeps coming to mind and disrupts recall of the new one. Similarly, a language student who has previously learnt French may find that French vocabulary repeatedly intrudes when they are trying to recall their newly-learnt Spanish.

Underwood (1957) demonstrated PI in a classic study: participants who had learned many word lists prior to a target list recalled the target list far less well than those who had learned few prior lists. The more previous learning there was, the worse recall became — showing that accumulated prior memories progressively interfere with new ones.

Retroactive Interference

Retroactive interference (RI) occurs when newer learning interferes with the recall of older information. "Retroactive" means that more recent learning acts backwards in time to disrupt memories formed earlier. Continuing the language example: after intensively studying Spanish, the student may find that their French vocabulary — previously well-known — becomes harder to access, because Spanish words now intrude during retrieval.

McGeoch and McDonald (1931) provided key laboratory evidence for RI. Participants learned a list of adjectives, then either rested or learned new material (synonyms, antonyms, unrelated adjectives, nonsense syllables, or numbers) before recalling the original list. Recall was worst when the interpolated material was most similar (synonyms) — demonstrating that similarity is a crucial factor in the degree of RI.

Real-world Evidence

Most interference research uses artificial word-list tasks, raising ecological validity concerns. However, Baddeley and Hitch (1977) provided more naturalistic evidence using rugby players. Players were asked to recall the teams they had played against during a season. Crucially, players who had played more games (and thus had more intervening matches to cause interference) recalled earlier opponents less well than players who had missed games through injury. This suggests that retroactive interference operates in real-world memory, not just in the laboratory.

Evaluation

Interference theory is well-supported by controlled laboratory experiments, and the distinction between PI and RI has been replicated consistently. However, the theory has been criticised for relying too heavily on artificial word-list paradigms — real-life memories are typically richer, more distinctive, and better organised than word lists, which may make real-world interference less pervasive than laboratory studies suggest. The theory also struggles to explain why we forget information for which there is no obvious competing memory, and does not offer a neurological account of the interference mechanism.