Retrieval-Induced Forgetting: Why Remembering Can Reduce Other Recall
Author: Angela Bogdanova · Published: October 12, 2026 · Editorial Policy
Research cluster: Memory & Forgetting · Written by Artificial Sapiens
Retrieval-induced forgetting (RIF) is the experimentally observed reduction in later access to some information after a person selectively retrieves other, related information. Remembering one member of a category can make another member harder to recall, even though both were previously learned. The effect is assessed by comparing unpracticed items from practiced categories with comparable items from categories that received no retrieval practice. It concerns accessibility on a particular test; it does not demonstrate that the unremembered material has been physically erased.
The phenomenon reveals a central feature of human memory: retrieval is selective and can change the competitive environment in which other memories are later accessed. The foundational experiments were reported by Anderson, Bjork, and Bjork (1994). A major meta-analysis by Murayama and colleagues (2014) supports the reality of the phenomenon across studies while documenting substantial theoretical and methodological qualifications. Whether the effect reflects active inhibition of competing memories, strengthened competition, changes in retrieval context, or a combination of processes remains an important research question.
What is retrieval-induced forgetting?
Retrieval-induced forgetting occurs when selectively recalling some studied material reduces the probability of later recalling related material that was not selected for practice. Imagine learning that a category includes orange, banana, kiwi, and peach. If a subsequent exercise repeatedly prompts you to retrieve orange and peach, the remaining category members may be harder to produce on a later test than comparable fruit names belonging to a category that received no such practice. In an experiment, the relevant comparison is not simply whether banana is missed; it is whether a systematically matched, unpracticed competitor is recalled less successfully than baseline material.
The term “retrieval-induced” identifies the activity that precedes the later decrement. “Forgetting” identifies a measurable reduction in access under specified test conditions. Neither term is an assertion that the trace has vanished. Researchers measure outcomes such as cued recall, free recall, recognition, response latency, or memory for order, and those tests can reveal different patterns. The 2014 Psychological Bulletin meta-analysis found that the strength of evidence for proposed mechanisms varies with how the effect is tested.
RIF is particularly interesting because ordinary study and retrieval practice often strengthen the material being recalled. The same episode can therefore yield two divergent outcomes: better later remembering for practiced targets and poorer later remembering for their unpracticed associates. The effects are defined for different items, so there is no logical contradiction. This selective consequence is why RIF belongs within the psychology of forgetting and the broader science of retrieval competition.
The classic retrieval-practice experiment
Study phase: build several categories
In the classic experiment, participants first study category–exemplar pairs such as FRUIT–ORANGE and FRUIT–BANANA, together with pairs from several other categories. The category acts as a shared retrieval cue. To keep comparison conditions meaningful, researchers counterbalance which words receive practice and which are withheld across participants or lists. The starting point matters: a difference observed later cannot be attributed to retrieval practice if the target items were intrinsically easier to remember beforehand.
Retrieval-practice phase: select only some items
Participants then practice retrieving certain members of some categories, often from the category label plus a partial word stem, such as FRUIT–OR____. Other items within those same practiced categories receive no direct retrieval practice. Still other categories receive no practice at all. Repeated practice usually improves access to the selected words, creating a strong opportunity for competition when the category cue is subsequently used.
Final test: compare three conditions
Researchers classify items as Rp+ (retrieval-practiced items from practiced categories), Rp− (unpracticed items belonging to practiced categories), and Nrp (items from categories that received no retrieval practice). RIF is identified by the contrast Rp− versus Nrp: poorer later performance for Rp− than Nrp. Improved performance for Rp+ relative to Nrp is retrieval-practice benefit, a separate comparison. The original experiments described this logic and included controls for recall order and category strength (Anderson et al., 1994).
A schematic result might be 85% recall for Rp+, 45% for Rp−, and 60% for Nrp. These numbers are hypothetical, provided only to illustrate the analysis; they are not a pooled estimate or an outcome taken from a named study. In that hypothetical example, practice helped its targets by 25 percentage points relative to baseline and was associated with a 15-point decrement for the unpracticed competitors. A well-designed experiment would still need appropriate randomization, matching, statistical analysis, and checks for alternative explanations.
Why can remembering one thing interfere with remembering another?
A retrieval cue commonly activates several associations. A question about a particular teacher may bring up one classroom episode, a different lesson, the teacher's name, and the emotions associated with school. Selective retrieval succeeds when the intended representation is selected from these possibilities. That selection changes the conditions facing other associations at a future attempt.
Several mechanisms could contribute. An active-control account proposes that competitors are inhibited while the target is being selected. A blocking account proposes that repeatedly strengthened targets become more dominant responses to the cue. A contextual account proposes that selective practice shifts the mental situation reinstated at the final test. Each predicts some overlap in behavior, which makes carefully designed dissociations essential. The relevant scientific question is which mechanism is necessary or sufficient under which conditions, rather than whether competition exists at all (Storm and Levy, 2012; Jonker, Seli, and MacLeod, 2013).
The inhibitory-control explanation
According to the inhibitory account, an attempted retrieval activates the target along with competing memories. In some situations the memory system limits access to those competitors, allowing the intended target to be selected. The resulting reduction in competitor accessibility can persist until the final memory test. Suppression in this account is a proposed operation in ordinary cognitive control; it is not synonymous with consciously deciding to forget something, and it is not a clinical diagnosis.
A central prediction is competition dependence: the competitor must become sufficiently active to interfere with target retrieval before suppression is needed. Original work found different effects for strongly and weakly associated category exemplars, and later experiments have manipulated whether competitors were actually available at the moment of practice. A study of learning order and competition observed forgetting when unpracticed competitors were learned before target retrieval, whereas learning those competitors after practice reduced that opportunity.
Another predicted signature is cue independence. If a competitor becomes less accessible in a relatively general way, impaired recall might remain when the final test uses a new cue rather than the practiced category label. Some experimental results support this; however, independent cues also change what is being tested, and the meta-analytic synthesis found that effect sizes and interpretations depend on the testing procedure. Cue independence is therefore evidence considered in the debate, not a simple laboratory fingerprint that settles the mechanism by itself.
Strength-based competition and blocking
A noninhibitory account starts with strengthened targets. If FRUIT–ORANGE is practiced repeatedly, ORANGE can become a highly available answer to FRUIT. Later, this dominant association may block access to BANANA when the category cue is presented, even if the unpracticed item's underlying representation remains intact. Different search strategies, response order, and associative strengths can influence the final score.
The blocking account makes clear why Rp− below Nrp need not automatically imply that the brain directly suppressed an Rp− representation. It also explains why experiments that rely only on the original category cue can be hard to interpret. Researchers therefore use independent cues, recognition tests, controlled recall order, and manipulations of target strengthening to examine which explanation best accommodates the full pattern. Raaijmakers and Jakab (2012) reported challenges to assumptions used to distinguish inhibition from competing alternatives.
Context change and reinstatement
Memory retrieval depends partly on mental context: what a person is attending to, the task being performed, recent thoughts, and how the remembered material is organized. The contextual account developed by Jonker and colleagues (2013) proposes that shifts between initial study, retrieval practice, and final testing can disadvantage material linked to the original study context. In this interpretation, apparent forgetting can emerge even without lasting suppression of the competitor itself.
This theory makes testable predictions. Some experiments supported its contextual manipulations; other investigations failed to reproduce key predicted forgetting after restudy combined with context change, or found effects under conditions the account predicted should reduce them. See Soares, Polack, and Miller (2016) and research testing context-change and restudy explanations. Context is unquestionably important to remembering, but a single context-only explanation has not resolved the entire RIF literature.
Four influential tests of the mechanism
Retrieval specificity: is active recall required?
In a conventional paradigm, active retrieval of selected material often produces forgetting more reliably than simply seeing the same material again. This pattern suggests that overcoming competing responses can matter. Yet whether an experiment is genuinely retrieval specific depends on whether restudy controls equate attention, exposure, cue strength, mental context, and learning of competing material. Studies reporting RIF-like outcomes under conditions designed to remove competition are especially informative because they challenge simple versions of the inhibition theory (Raaijmakers and Jakab, 2012).
Competition dependence: which items are vulnerable?
The inhibitory account anticipates greater forgetting for strongly competing associates than for alternatives that never meaningfully interfere during target retrieval. The classic category experiments motivated this prediction, and later work deliberately varied how and when competitors became available. Results supportive of competition dependence indicate that mere semantic relatedness is not always sufficient: the competitor's activation and relevance to the attempted retrieval may matter (Anderson et al., 1994; retrieval-competition manipulation).
A crucial qualification is that competition can contribute to the production of an effect without proving that inhibition caused the later impairment. For instance, a strong competitor may change which category responses are practiced successfully or retrieved first on the final test. Mechanism claims must specify the point in the task at which competition acts and distinguish effects during practice from effects during final recall.
Cue independence: can a new prompt release the forgotten item?
A final test based on an alternative cue can reduce simple blocking by the practiced category label. Persistence of a deficit with a genuinely independent cue has often been interpreted as evidence that access to the competitor itself was reduced. Conversely, recovery with a richer or more distinctive cue shows that a failure under the original category cue need not imply irrecoverability.
Independent-cue procedures are useful but complicated: an alternative cue may have different baseline effectiveness, activate different associates, or reinstate the earlier context in a new way. Researchers must ensure that the cue is not quietly reintroducing the same competition. Studies of cue-based release from RIF demonstrate the importance of how the final test is constructed (Jonker, Seli, and MacLeod, 2012).
Strength independence: must more strengthening mean more forgetting?
If a purely strength-based blocking process were responsible, stronger practiced targets might be expected to produce correspondingly greater forgetting of competitors. Several inhibitory models instead predict that competitor suppression and target strengthening can vary independently. This is often called the strength-independence assumption, although testing it by correlating two observed difference scores creates interpretation problems involving measurement error and statistical dependence.
Raaijmakers (2016) analyzed problems with such tests. The absence of a correlation between benefits for Rp+ and costs for Rp− cannot by itself establish inhibition, and the presence of a correlation is not a complete refutation of it. Stronger tests depend on manipulating proposed causes directly and measuring distinct observable predictions.
What does the overall research evidence establish?
The principal meta-analysis of RIF (Murayama et al., 2014) aggregated an extensive body of experimental results to compare findings commonly invoked for inhibitory and noninhibitory accounts. Its broad conclusion favored inhibitory interpretations of many classic signatures while also identifying important challenges and dependence on assessment procedures. This is an evidence-weighting result, not proof that every observed instance of selective forgetting involves the same neural operation.
Narrative reviews describe RIF across word lists, learning materials, testimony, autobiographical remembering, social cognition, and creative tasks. Such variation supports the importance of the phenomenon beyond a single category-list demonstration, but generalization depends on the task and population. Storm and colleagues' 2015 review discusses these extensions and highlights the need to test mechanisms in settings that depart from the original laboratory design.
Recent research also tests durability and more complex materials. Sozer and colleagues (2025) found socially shared retrieval-induced forgetting after repeated selective practice distributed over time, including delayed assessments and an experiment involving conversations. Those results make persistence a researchable feature of the social environment, while their designs do not imply that every selective conversation produces permanent public-memory distortion.
The evidence also remains self-correcting. A reanalysis and three unsuccessful replication efforts challenged a specific neuroimaging-linked experimental variant and a proposed connection between cortical pattern suppression and subsequent forgetting. Failure of a particular variant to replicate does not cancel a broader meta-analytic phenomenon; equally, evidence for a broad phenomenon cannot validate every proposed neural mechanism. The most precise summary is that selective retrieval can impair related recall under experimentally specified circumstances, and those circumstances matter.
How large is retrieval-induced forgetting?
There is no universal percentage of memories lost after recalling related material. The magnitude depends on baseline recall, item selection, strength of competition, task instructions, delay, final test, statistical design, and whether participants integrate the material. A result from one list-learning experiment cannot be used to predict an individual's recall probability in a family argument, a classroom, or an eyewitness interview.
Investigators sometimes express RIF as a percentage-point difference between Rp− and Nrp recall or use a standardized effect size across experiments. Those statistics describe different quantities. A 10-percentage-point recall difference is not the same as erasing 10% of stored memories, and a standardized effect size cannot be interpreted as a probability for an individual. The 2014 meta-analysis is the appropriate starting point for quantitative synthesis, including its moderators and methodological qualifications.
When does retrieval cause facilitation instead?
Retrieval of one item can sometimes help, rather than harm, memory for related untested information. This is retrieval-induced facilitation. For instance, retrieving a coherent explanation from a text may reactivate a connected chain of facts that were never themselves asked about. The direction of transfer can depend on whether ideas were integrated while learning, how retrieval is prompted, and the interval before the final test.
In experiments with prose material, Chan (2009) found that integration and test delay helped determine whether retrieval supported or harmed related untested content. Oliva and Storm (2023) replicated and extended retrieval-induced facilitation across longer delays, with a meta-analytic component. These studies are a compelling counterweight to the simplistic claim that practicing one fact necessarily weakens its neighbors.
The distinction matters educationally. A narrowly prompted recall attempt among overlapping but poorly integrated facts may produce competition. A connected explanation may support relational retrieval. Teachers and students should avoid converting these tendencies into rigid rules: robust gains for practiced material and possible benefits to related material must be assessed against what students are actually expected to know.
Retrieval-induced forgetting versus ordinary interference
Proactive interference concerns earlier learning disrupting newer information; retroactive interference concerns later learning disrupting older information. These categories describe the temporal direction of competing learning. RIF, by contrast, is defined by the selective retrieval of some studied information and the later disadvantage for related items left unpracticed.
Imagine learning two passwords on different days: the old password intrudes when using the new one, or the new password intrudes when using the old one. Those are conventional interference examples. Now imagine learning several answers in a category and repeatedly retrieving only one subset before being tested on all of them. The decrement in related, unpracticed items compared with a no-practice baseline is the defining RIF pattern. For the larger taxonomy, see Proactive vs. Retroactive Interference: How Memories Compete.
Some noninhibitory explanations of RIF are interference-based, so the categories are related. But calling every instance of cue competition “retrieval-induced forgetting” removes the experimental condition that gives the phenomenon its meaning. RIF is a more specific claim about what happens after selective retrieval.
Retrieval-induced forgetting versus intentional forgetting and thought suppression
Directed forgetting experiments instruct participants to remember or disregard specified material, whereas think/no-think experiments instruct them to prevent a particular memory from entering awareness in response to reminders. RIF experiments ordinarily ask participants to retrieve designated targets; impairment of the other items is incidental to that selective retrieval. The task demand, participant goal, and proposed control processes differ across these designs.
Research on suppression-induced forgetting includes distinct paradigms and evidence. A 2026 meta-analysis of think/no-think experiments summarizes effects in deliberate retrieval-suppression tasks. Its numbers should not be imported into estimates of classical RIF, because preventing an unwanted target from coming to mind and selecting a desired target amid competitors are different experimental operations.
The practical distinction is equally important. Discovering that selected practice can disadvantage other recall does not mean that people can deliberately delete a traumatic memory, erase an emotion by remembering something else, or reliably control what another person will remember. Those inferences require separate clinical and applied evidence.
Does the brain actively inhibit competing memories?
Neural explanations remain an active research program. The inhibitory framework associates retrieval selection with control processes that can limit access to competing representations. Cognitive-control and memory-retrieval networks, including prefrontal and medial temporal systems, are candidates for participating in these operations. However, finding activity in a brain region during selective retrieval does not identify inhibition as the exclusive cause of later behavioral forgetting.
Neuroimaging, electrophysiology, and representational analyses can help distinguish activation, competition, and later reinstatement. These measures are indirect and highly sensitive to task design. The failed replications of a particular cortical-pattern variant underscore why neural interpretations need converging behavioral evidence and independent reproduction (replication and reanalysis report).
At the level of everyday explanation, it is accurate to say that retrieval can alter how accessible other memories become. It is an additional theoretical assertion to say that the brain has permanently inhibited those memories, and a still stronger unsupported claim to say they have been neurologically destroyed. Scientific vocabulary should preserve the differences.
Can retrieval-induced forgetting affect eyewitness memory?
Selective questioning can create asymmetries in what witnesses later recall, and RIF paradigms have been adapted to witness-like materials. If an interview repeatedly elicits details about one aspect of an event, related unelicited information may become less accessible under some conditions. This creates a reason to study interview design, test order, competing details, and delayed recall.
It does not justify concluding that a particular witness forgot an omitted event because someone repeatedly asked about other details. Real testimony involves exposure quality, stress, prior knowledge, postevent information, social influence, motivation, and many forms of memory reconstruction. The applied RIF review addresses eyewitness settings while recognizing that controlled experiments differ from legal cases. Individual claims require case-specific evidence.
An additional distinction concerns false memories. RIF primarily concerns reduced later access to existing studied items in experimental comparisons; false-memory experiments concern remembering details or events that were not presented as remembered. Both can affect what is reported, yet they do not measure the same outcome.
Conversations, teaching, and shared remembering
In discussion, speakers select which details to recount, and listeners may organize subsequent recall around those selections. Researchers investigate socially shared RIF by testing whether repeated selective recollection can lower memory for related details in a conversational partner or audience. Some studies find such patterns, with duration affected by the schedule and structure of repetition. The 2025 durability experiments provide particularly relevant evidence.
This does not establish a universal mechanism for collective memory, political persuasion, or social conformity. Large-scale public narratives evolve through repetition, media exposure, identity, institutions, selection of sources, and changing contexts. RIF may be one component in some communicative settings, but a laboratory decrement on a memory test does not establish an explanation for any specific historical or social omission.
For educators, a useful implication is to design recall opportunities that cover the whole intended curriculum rather than assuming that frequent testing of a narrow subset will automatically strengthen all adjacent knowledge. It is also helpful to use integrated explanations and varied questions, because the evidence on retrieval-induced facilitation shows that related material can benefit when it is meaningfully connected (Chan, 2009; Oliva and Storm, 2023).
Emotion and autobiographical memories
Selective retrieval operates in environments containing emotion, personal meaning, and varied associations. Experimental studies have explored whether retrieving semantic material can alter later access to associated episodes of different emotional valence. Greer and colleagues (2024) reported forgetting of related negative and neutral episodic memories after effortful retrieval of semantically connected information. This finding broadens the scope of testable RIF effects without establishing that the method safely changes an individual's personal emotional history.
Autobiographical memory is reconstructed through changing goals, perspectives, context, and narrative practices. An accessible 2024 research update on emotion and autobiographical remembering examines these broader influences. Any real-world claim that one remembered experience caused the loss of another should consider those influences, along with the quality and timing of the person's original learning.
It is particularly important to avoid clinical extrapolation. Ordinary RIF is a cognitive experimental effect, not a diagnosis of amnesia, PTSD, dissociation, repression, or dementia. A person who cannot immediately recall some details after discussing others has not thereby demonstrated a traumatic memory disorder or a therapeutic change. Clinical assessment and treatment questions belong to their appropriate clinical evidence base.
Can retrieval-induced forgetting be used therapeutically?
There is exploratory translational work, but RIF should not be presented as an established treatment. A 2025 randomized study by Losiewicz, Davis, and Craske investigated an RIF-informed approach alongside exposure exercises in participants with elevated public-speaking fear. Such research tests specific experimental outcomes and provides hypotheses for further trials; it does not establish that selective recall alone cures anxiety or can erase fear memories.
Application requires more than observing a laboratory forgetting effect. Researchers need to evaluate symptom outcomes, comparative effectiveness, harms, durability, real-world functioning, and whether proposed mechanisms actually mediate changes. This distinction also protects readers from commercial “memory erasure” claims. If someone experiences intrusive or distressing memories, decisions about care should be based on clinical assessment and appropriate validated interventions, rather than on self-directed attempts to suppress memory derived from a category-word experiment.
Age, attention, and individual differences
The magnitude of RIF may depend on developmental stage, cognitive resources, task difficulty, encoding strategies, and the particular test used. Some research links individual differences in working-memory capacity with the degree of RIF under specific conditions, which has been interpreted as consistent with an executive-control contribution (Aslan and Bäuml, 2011). Such associations do not amount to an individual diagnostic measure of inhibitory ability.
Older and younger participants can differ in how they organize, encode, and retrieve material; it is therefore essential to control baseline memory performance and task difficulty before interpreting an apparent group difference. The absence of an RIF effect in one person does not indicate cognitive decline, and a larger effect does not mean the person has superior everyday memory.
Attention matters because material that was never encoded adequately may later appear forgotten without any retrieval-induced mechanism. A person might fail to remember something because it was barely noticed, because the present cues are weak, or because competing responses disrupt recall. RIF research isolates one selective-retrieval contribution within this much larger landscape of normal variability.
How long does the effect last?
The original experiments included delays designed to show that RIF extends beyond the immediate practice moment. Subsequent studies have tested minutes, hours, days, and weeks, and results vary with practice schedule, test format, and material. Findings about durability must specify these conditions rather than treat the effect as a permanent modification of memory.
In socially shared paradigms, Sozer and colleagues (2025) observed effects after spaced selective practice with delayed tests as long as several weeks in studied conditions. Other experimental designs produce weaker or shorter-lived differences. The existence of some durable results does not establish a universal duration, nor does an observed decrement prove that the material will never again be recalled.
When a new cue, a context change, or a subsequent review recovers access, that recovery is important evidence about the nature of the effect. Studying when retrieval succeeds again is as informative as documenting when a final test produces a lower score.
Practical implications: remembering broadly and accurately
RIF is a reason to distinguish narrow performance from comprehensive understanding. If you repeatedly quiz yourself on only a few facts from a topic, your performance on those facts may improve substantially, while related untested facts may receive less opportunity to become accessible. Research does not imply that selective practice is generally harmful or that self-testing should be abandoned. It suggests that the scope and organization of retrieval tasks matter.
A practical study design is to sample questions across the entire set of learning objectives, rotate which concepts are retrieved, ask learners to explain relationships among concepts, and revisit material that was previously omitted. These practices have rational links to the selectivity problem and to evidence on integration, but the exact amount by which they prevent RIF in everyday studying has not been established as a universal figure. To understand the complementary benefit of testing, see Retrieval Practice: Why Testing Yourself Improves Memory.
When accuracy matters in interviews, investigations, or personal records, independent documentation and carefully framed questions can reduce reliance on selective later recall. An omitted detail alone is not evidence of deliberate concealment, and confident recall of selected details does not certify completeness. Retrieval cues and later contexts change what comes to mind, so the quality of evidence must be judged independently from the fluency of a remembered narrative.
Common misunderstandings
“Remembering one fact erases the other facts.”
An experimental recall decrement establishes lower later access on a specified test, not neurological deletion. Other cues, additional learning, or changed conditions may restore performance. Mechanism and permanence are additional claims that require separate evidence.
“Every act of retrieval makes related memories worse.”
Selective retrieval may impair some related items under certain conditions, leave others unaffected, or facilitate related information when it is meaningfully integrated. Research on facilitation directly contradicts the universal version of this claim.
“Inhibition has been conclusively proved as the only explanation.”
The inhibitory account explains an important body of findings, but there are continuing challenges and alternative models. The major meta-analysis supports inhibition broadly while identifying difficulties, and contextual and interference-based research remains theoretically relevant.
“Retrieval-induced forgetting proves repressed memories.”
Incidental forgetting of unpracticed category items and theories about repression or dissociative amnesia concern different definitions, methods, and clinical questions. Laboratory RIF cannot verify the history or cause of an individual's inaccessible autobiographical memory.
“If I do not show RIF, my memory control is defective.”
RIF scores depend on the task, items, timing, baseline and test. They are research measures, not standalone clinical screens for executive function, intelligence, memory health, or neurological disease.
Frequently asked questions
What is retrieval-induced forgetting in simple terms?
It is the finding that repeatedly remembering some information can make related, unpracticed information harder to recall later. The typical experiment compares unpracticed items from practiced categories with items from untouched categories. The difference is measured on a specific memory test.
What is an everyday example of retrieval-induced forgetting?
Someone repeatedly recounts several details of a group trip and later has more difficulty recalling other related details. That is a plausible illustration, but it becomes confirmed RIF only in a study that compares the relevant memory outcomes against a suitable no-practice baseline and controls competing explanations.
Who discovered retrieval-induced forgetting?
The influential 1994 retrieval-practice experiments were reported by Michael C. Anderson, Robert A. Bjork, and Elizabeth L. Bjork. Earlier researchers had also studied inhibitory effects of recall and retrieval competition. The 1994 contribution established the modern category-based paradigm that defined much of the subsequent RIF program (Anderson et al., 1994).
What do Rp+, Rp−, and Nrp mean?
Rp+ refers to practiced items from practiced categories. Rp− refers to items in those categories that were not practiced. Nrp refers to items in categories that received no practice. The core forgetting effect is Rp− performance below Nrp performance; better Rp+ performance is a separate retrieval benefit.
Is retrieval-induced forgetting the same as retroactive interference?
No. Retroactive interference describes later learning disrupting recall of earlier material. RIF concerns selective retrieval and reduced later access to related, unpracticed material compared with baseline. The two can overlap theoretically but are identified by different experimental comparisons.
Why does remembering sometimes cause forgetting?
A retrieval cue may activate competing memories. Selecting a target may involve suppression of competitors, strengthen some responses so they block others, or reinstate a context that disadvantages unpracticed material. These explanations remain under empirical comparison rather than amounting to one universally accepted mechanism.
Does retrieval-induced forgetting mean the memory is gone forever?
No. RIF is measured as lower accessibility at a later test. It does not establish physical erasure or permanent inability to remember. Its duration varies with procedure, and changed cues or subsequent study can influence recovery.
Is retrieval-induced forgetting unconscious?
Usually, the participant intentionally retrieves practiced targets, while any later disadvantage for related material is not the objective of that practice. That does not make every underlying mechanism unconscious in the same psychological sense, and it does not mean people can predict which personally important details will be affected.
Can retrieval-induced forgetting improve learning?
Selective practice typically benefits practiced targets, and under conditions of integrated learning it can also benefit related material. Comprehensive learning requires coverage of the intended content. Retrieval-induced facilitation and RIF are both possible outcomes, so targeted self-testing should be combined with broad and relational review.
Does RIF occur in conversations?
Researchers have observed socially shared forms of selective-retrieval forgetting in controlled listening and conversational tasks. Those findings suggest a possible contribution to how shared accounts develop, but they do not justify diagnosing the cause of omissions in a particular conversation or public narrative.
Can a therapist use retrieval-induced forgetting to erase trauma?
There is no established clinical protocol that reliably erases traumatic experiences through selective recall. Experimental and early translational work should be evaluated within clinical research, and RIF cannot substitute for evidence-based assessment or treatment.
Can retrieval-induced forgetting be prevented?
Some manipulations of retrieval cues, task context, and integration change the observed effect, and comprehensive study schedules may reduce the risks of leaving material unpracticed. No method guarantees that RIF never occurs. A practical response is to sample learning broadly, build explicit connections, and test omitted material directly.
How is RIF different from the testing effect?
The testing effect describes improved later retention for material that was retrieved rather than merely restudied. RIF concerns a potential cost for related information left unpracticed during that selective retrieval. Both can occur in a single experiment for different items.
What is the strongest scientific evidence for RIF?
The original controlled experiments, subsequent independent studies, and the 2014 meta-analytic review together constitute the central evidence base. Modern replication work and competing theoretical tests help establish where effects are robust and where particular causal interpretations overreach the data.
Conclusion
Retrieval-induced forgetting demonstrates that remembering is selective and that a successful retrieval attempt can alter later access to related information. The classic contrast between practiced, related-unpracticed, and baseline items provides a precise experimental definition. The accumulated evidence supports the existence of the effect under meaningful conditions, while inhibition, blocking, context change, and integration contribute to a still-developing explanation of when and why it appears.
The broader lesson for cognitive psychology is that remembering and forgetting are dynamically connected. For learning, interviewing, and everyday reflection, careful selection of cues and balanced coverage of information matter. The science supports nuanced predictions about accessibility; it does not support treating ordinary selective forgetting as memory erasure or a clinical disorder.
Related Articles
Why Do We Forget? Causes and Theories of Forgetting — the parent overview of major forgetting mechanisms.
Proactive vs. Retroactive Interference: How Memories Compete — the neighboring comparison of temporal interference processes.
Retrieval Practice: Why Testing Yourself Improves Memory — the complementary evidence on benefits of retrieval for practiced knowledge.
Human Memory: Types, Processes, and How It Works — the foundational map of human memory.
The Forgetting Curve: Ebbinghaus, Replication, and Modern Evidence — how retention changes over time under controlled testing.
References
Anderson, M. C., Bjork, R. A., & Bjork, E. L. (1994). Remembering can cause forgetting: Retrieval dynamics in long-term memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 20(5), 1063–1087. https://doi.org/10.1037/0278-7393.20.5.1063
Aslan, A., & Bäuml, K.-H. T. (2011). Individual differences in working memory capacity predict retrieval-induced forgetting. Journal of Experimental Psychology: Learning, Memory, and Cognition, 37(1), 264–269. https://doi.org/10.1037/a0021324
Buchli, D. R., Storm, B. C., & Bjork, R. A. (2016). Explaining retrieval-induced forgetting: A change in mental context between the study and restudy practice phases is not sufficient to cause forgetting. Quarterly Journal of Experimental Psychology, 69(6), 1197–1209. https://doi.org/10.1080/17470218.2015.1076866
Chan, J. C. K. (2009). When does retrieval induce forgetting and when does it induce facilitation? Implications for retrieval inhibition, testing effect, and text processing. Journal of Memory and Language, 61(2), 153–170. https://doi.org/10.1016/j.jml.2009.04.004
Chan, J. C. K., Erdman, M. R., & Davis, S. D. (2015). Retrieval induces forgetting, but only when nontested items compete for retrieval: Implication for interference, inhibition, and context reinstatement. Journal of Experimental Psychology: Learning, Memory, and Cognition, 41(5), 1298–1315. https://doi.org/10.1037/xlm0000101
Clark, C. A., Todorovic, D., Levy, B. J., Eschmann, K. C. J., Klein, M. D., Anderson, M. C., van Schie, K., Whitridge, J., Yick, J., & Fawcett, J. M. (2026). Forgetting as a consequence of retrieval suppression: A meta-analytic review. Journal of Experimental Psychology: General, 155(5), 1318–1346. https://doi.org/10.1037/xge0001922
Faul, L., Ford, J. H., & Kensinger, E. A. (2024). Update on ‘Emotion and autobiographical memory’: 14 years of advances in understanding functions, constructions, and consequences. Physics of Life Reviews, 51, 255–272. https://doi.org/10.1016/j.plrev.2024.10.005
Greer, J., Ali, A., Laksman, C., Huang, R., McClay, M., & Clewett, D. (2024). Effortful retrieval of semantic memories induces forgetting of related negative and neutral episodic memories. Cognition, 251, 105908. https://doi.org/10.1016/j.cognition.2024.105908
Jonker, T. R., & MacLeod, C. M. (2012). Retrieval-induced forgetting: Testing the competition assumption of inhibition theory. Canadian Journal of Experimental Psychology, 66(3), 204–211. https://doi.org/10.1037/a0027277
Jonker, T. R., Seli, P., & MacLeod, C. M. (2012). Less we forget: Retrieval cues and release from retrieval-induced forgetting. Memory & Cognition, 40, 1236–1245. https://doi.org/10.3758/s13421-012-0224-2
Jonker, T. R., Seli, P., & MacLeod, C. M. (2013). Putting retrieval-induced forgetting in context: An inhibition-free, context-based account. Psychological Review, 120(4), 852–872. https://doi.org/10.1037/a0034246
Losiewicz, O. M., Davis, B., & Craske, M. G. (2025). Remembering to forget fear: Augmenting exposure with retrieval-induced forgetting. Behaviour Research and Therapy, 193, 104829. https://doi.org/10.1016/j.brat.2025.104829
Murayama, K., Miyatsu, T., Buchli, D. R., & Storm, B. C. (2014). Forgetting as a consequence of retrieval: A meta-analytic review of retrieval-induced forgetting. Psychological Bulletin, 140(5), 1383–1409. https://doi.org/10.1037/a0037505
Oliva, M. T., & Storm, B. C. (2023). Examining the effect size and duration of retrieval-induced facilitation. Psychological Research, 87, 1166–1179. https://doi.org/10.1007/s00426-022-01729-0
Potter, K. W., Huszar, L. D., & Huber, D. E. (2018). Does inhibition cause forgetting after selective retrieval? A reanalysis and failure to replicate. Cortex, 104, 26–45. https://doi.org/10.1016/j.cortex.2018.03.026
Raaijmakers, J. G. W. (2016). On testing the strength independence assumption in retrieval-induced forgetting. Psychonomic Bulletin & Review, 23, 1374–1381. https://doi.org/10.3758/s13423-015-0991-4
Raaijmakers, J. G. W., & Jakab, E. (2012). Retrieval-induced forgetting without competition: Testing the retrieval specificity assumption of the inhibition theory. Memory & Cognition, 40, 19–27. https://doi.org/10.3758/s13421-011-0131-y
Soares, J. S., Polack, C. W., & Miller, R. R. (2016). Retrieval-induced versus context-induced forgetting: Does retrieval-induced forgetting depend on context shifts? Journal of Experimental Psychology: Learning, Memory, and Cognition. https://doi.org/10.1037/xlm0000171
Sozer, E., Fagin, M., Meksin, R., & Hirst, W. (2025). Durability of retrieval-induced forgetting: Effects of different practice schedules. Memory & Cognition, 53, 1095–1111. https://doi.org/10.3758/s13421-024-01634-4
Storm, B. C., & Levy, B. J. (2012). A progress report on the inhibitory account of retrieval-induced forgetting. Memory & Cognition, 40, 827–843. https://doi.org/10.3758/s13421-012-0211-7
Storm, B. C., Angello, G., Buchli, D. R., Koppel, R. H., Little, J. L., & Nestojko, J. F. (2015). A review of retrieval-induced forgetting in the contexts of learning, eyewitness memory, social cognition, autobiographical memory, and creative cognition. Psychology of Learning and Motivation, 62, 141–194. https://doi.org/10.1016/bs.plm.2014.09.005