You are viewing a free preview of this lesson.
Subscribe to unlock all 10 lessons in this course and every other course on LearningBro.
Why do we forget? The AQA specification focuses on two explanations for forgetting from long-term memory: interference theory (forgetting because similar memories compete) and retrieval failure (forgetting because the cues needed to access an available memory are absent). Both assume the information is in LTM; they differ on whether it has been disrupted by other learning or is simply inaccessible. Understanding the theories, their supporting studies and their limitations is essential for the exam.
Key Definition: Forgetting is the failure to recall or recognise information that was previously learned. A memory may be unavailable (lost from the store) or merely inaccessible (available but not retrievable without the right cue) — a distinction that lies at the heart of these two theories.
This lesson covers the AQA Paper 1 Memory content on explanations for forgetting: proactive and retroactive interference and retrieval failure due to the absence of cues. You are required to know the two types of interference (proactive and retroactive) and the role of similarity, together with the encoding specificity principle and the two forms of cue-dependent forgetting (context-dependent and state-dependent), plus the studies that support each — McGeoch and McDonald (1931), Baddeley and Hitch's (1977) rugby study, Godden and Baddeley (1975) and Carter and Cassaday (1998). The topic builds on the MSM and types-of-LTM lessons (these are explanations for forgetting from LTM) and links forward to eyewitness testimony (retrieval cues underpin the cognitive interview). You should be able to describe both explanations (AO1), apply them to everyday forgetting (AO2), and evaluate them through research support, ecological validity and the question of which explanation accounts for more forgetting (AO3).
Interference theory proposes that forgetting occurs because memories compete with and disrupt one another, and that this competition is greatest when the memories are similar. The information is not lost; rather, one memory blocks or distorts access to another. There are two types.
Key Definition: Proactive interference occurs when older learning disrupts the recall of newer learning. The old memory interferes forwards in time with the new.
Example: After years of using one phone number, you keep recalling the old number instead of your new one. (Direction: old disrupts new.)
Key Definition: Retroactive interference occurs when newer learning disrupts the recall of older learning. The new memory interferes backwards in time with the old.
Example: After learning your new number, you can no longer remember the old one. (Direction: new disrupts old.)
flowchart LR
subgraph PI["Proactive Interference"]
A["OLD learning"] -->|disrupts| B["NEW learning<br/>(harder to recall)"]
end
subgraph RI["Retroactive Interference"]
C["NEW learning"] -->|disrupts| D["OLD learning<br/>(harder to recall)"]
end
In both cases the key moderator is similarity: the more alike the two sets of material, the stronger the interference, because their cues and associations overlap and become confused.
Aim: To test whether the similarity of new material to original material affects how much is forgotten.
Procedure: Participants first learned a list of ten adjectives until they could recall it perfectly (a controlled baseline). They were then given new material to learn during a retention interval, before being tested on the original list. Six groups learned new material that varied in its similarity to the original list — from synonyms (most similar) through antonyms, unrelated adjectives, nonsense syllables and three-digit numbers, to a resting control group that learned nothing new.
Findings: Recall of the original list was worst when the new material was most similar.
| New (interfering) material | Similarity to original | Recall of original list |
|---|---|---|
| Synonyms | Most similar | Lowest (most forgetting) |
| Antonyms | High | Low |
| Unrelated adjectives | Moderate | Moderate |
| Nonsense syllables | Low | Higher |
| Three-digit numbers | Very low | Higher still |
| Rest (no new learning) | None | Highest (least forgetting) |
Conclusion: Retroactive interference is strongest when the competing material is most similar, demonstrating that similarity, not merely the act of learning something new, drives interference.
Underwood reviewed many earlier studies of serial learning and found that participants who had previously learned many word lists forgot considerably more of the most recent list than participants who had learned few or no prior lists. The more prior learning a person carried, the worse their recall of new material a day later. Conclusion: proactive interference accumulates — earlier learning progressively disrupts later learning — so forgetting reflects the quantity of competing prior material, not just the passage of time.
Aim: To test whether interference operates outside the laboratory, by examining whether forgetting depends on the number of intervening events rather than the time elapsed.
Procedure: Rugby union players were asked to recall the names of the teams they had played against over a season. Because injuries and other absences meant players had taken part in different numbers of games, the researchers could compare players for whom the same amount of time had passed but who had played different numbers of matches.
Findings: Recall depended on the number of games played in the interval, not on how much time had passed. Players who had played more matches (more intervening, similar events) recalled fewer team names; players who had missed games (fewer intervening events) recalled more, even over the same period.
Conclusion: This supports interference theory in a naturalistic setting — it is the intervening, similar experiences that cause forgetting, not simple time-based decay — and it strengthens the ecological validity of the laboratory findings.
Retrieval-failure theory proposes that information is not lost from LTM but is temporarily inaccessible because the cues present at encoding are absent at retrieval. The memory is available but cannot be accessed without the right trigger.
Key Definition: The encoding specificity principle states that a cue is most effective in prompting retrieval if it was present at the time of encoding. Recall improves when the cues at retrieval match the cues at encoding; forgetting results when they do not.
Cues may be external (features of the environment — context-dependent forgetting) or internal (the person's physiological or psychological state — state-dependent forgetting).
Aim: To test whether a mismatch between the external environment at learning and at recall causes forgetting.
Procedure: Deep-sea divers learned a list of words in one of two contexts — on land or underwater — and then recalled them in either the same or a different context. This produced four conditions in a repeated-measures design: learn land/recall land, learn land/recall water, learn water/recall water, learn water/recall land.
Findings: Recall was markedly better when the learning and recall contexts matched than when they mismatched — recall in the non-matching conditions was substantially lower (in the order of around 40% worse).
Conclusion: The external environment acts as a retrieval cue; when the cue is absent at recall, forgetting follows — strong support for the encoding specificity principle.
Aim: To test whether a mismatch between the internal state at learning and at recall causes forgetting.
Procedure: Volunteers learned lists of words and passages either after taking a mildly sedating antihistamine (creating an altered internal state of slight drowsiness) or after no drug, and then recalled the material in either the same or a different internal state. Again, four conditions were created (drug/drug, drug/no-drug, no-drug/no-drug, no-drug/drug).
Findings: Recall was significantly worse when the internal state at learning and recall did not match (e.g., learning on the drug but recalling without it) than when it matched.
Conclusion: Internal physiological state functions as a retrieval cue; a mismatch produces state-dependent forgetting, again consistent with encoding specificity.
The deepest conceptual point in this topic is the distinction between a memory being available (still present in LTM) and being accessible (retrievable at this moment). The two explanations sit on opposite sides of this line, and an examiner rewards candidates who make the contrast explicit.
| Interference theory | Retrieval failure | |
|---|---|---|
| Is the memory still in LTM? | The memory is disrupted/distorted by competing memories | The memory is fully available but temporarily inaccessible |
| Why can't we recall it? | A competing memory blocks or confuses it | The cues present at encoding are absent at retrieval |
| Key moderator | Similarity of the competing material | Match between encoding and retrieval cues |
| Can the right prompt recover it? | Not necessarily — the trace itself may be disrupted | Yes — supplying the missing cue restores access |
| Strongest evidence | McGeoch and McDonald; Baddeley and Hitch (rugby) | Godden and Baddeley (divers); Carter and Cassaday |
A single everyday episode shows how the two can apply to the same failure. Suppose you cannot remember the name of a person you met at a party last year. Interference would explain this as the many other names you have learned since (or before) competing with and disrupting that one — and predicts you may simply confuse it with a similar name. Retrieval failure would explain it as the absence of the cues present when you encoded the name (the room, the music, your mood, the conversation) — and predicts that returning to a similar context, or being reminded of the surrounding details, would suddenly bring the name back. Because the two theories make different predictions about what would help, they are genuinely distinct explanations and not merely two descriptions of the same process.
A further detail that deepens interference theory is the phenomenon of build-up and release from proactive interference, studied by Wickens and colleagues using the Brown-Peterson short-term forgetting task. When participants are given several successive trigrams (or word triples) drawn from the same semantic category, recall worsens across trials as proactive interference builds up — earlier, similar items increasingly disrupt later ones. But when the category is suddenly switched (e.g. from fruits to professions), recall jumps back up: this release from PI shows that interference depends on similarity of meaning, not merely on the number of prior items. This matters because it pins down why interference occurs — overlapping, similar representations compete — and provides a controlled demonstration that the content of the competing material, not just its quantity, governs forgetting.
A clear strength of interference theory is the body of controlled evidence behind it. McGeoch and McDonald demonstrated, with a tightly controlled baseline and systematically varied interfering material, that similarity increases retroactive interference, while Underwood showed the cumulative effect of prior learning (proactive interference). This matters because the use of laboratory experiments allowed precise manipulation of the relevant variable (similarity, or amount of prior learning) and control of confounds, supporting a genuine causal claim that competing memories disrupt one another. The implication is that interference is a real and demonstrable cause of forgetting, established under conditions rigorous enough to isolate it — a stronger evidential basis than mere correlation or anecdote.
A further strength is that interference is not confined to artificial word lists. Baddeley and Hitch's (1977) rugby study found that forgetting of team names depended on the number of intervening games rather than the time elapsed, exactly as interference theory predicts and contrary to a simple decay account. This matters because a common criticism of interference research — that it relies on meaningless laboratory material — is directly answered by a naturalistic study using a real, meaningful memory task. The implication is that interference has ecological validity and operates in everyday memory, which substantially strengthens the theory's external validity and its claim to explain real forgetting, not just performance on contrived tasks.
Set against this, much interference research does depend on artificial materials and conditions. The typical paradigm has participants learn lists of words or nonsense syllables, one after another, in a way that maximises competition — a situation rarely encountered in everyday life, where we seldom learn long sequences of highly similar material in quick succession. This matters because forgetting demonstrated under such artificial, interference-maximising conditions may overstate how often interference causes forgetting in real life, and may reflect demand characteristics of the laboratory. The implication is that, while interference is real (as the rugby study shows), its prevalence as an everyday cause of forgetting may be exaggerated by studies engineered to produce it — so the laboratory evidence should be interpreted alongside, not instead of, more naturalistic data.
A strength of retrieval-failure theory is its broad explanatory reach. Tulving argued that cue-dependent forgetting is the most common form of everyday forgetting — the familiar experience of a memory being "on the tip of the tongue", available but momentarily inaccessible until the right cue appears, fits the theory neatly. Godden and Baddeley's divers and Carter and Cassaday's antihistamine study provide controlled support for context- and state-dependent cues respectively. This matters because a theory that accounts for a wide range of common forgetting experiences, backed by experimental evidence, has considerable explanatory value. The implication is that much forgetting attributed to "loss" may in fact be retrieval failure, which reframes forgetting as a problem of access rather than storage — a conclusion with direct practical consequences (below).
Subscribe to continue reading
Get full access to this lesson and all 10 lessons in this course.