01 · Executive synthesis

Durable language memory is built by reconstruction, not by comfortable exposure

Second-language study often feels productive when a word, rule or sentence looks familiar. The memory literature repeatedly shows why that feeling is dangerous: availability during study is not the same as later accessibility without the answer in front of you.112531

Three mechanisms need to be separated. Repetition increases the number of encounters. Retrieval requires the learner to reconstruct a target from a cue. Spacing changes the temporal distance between encounters or retrievals. Repetition is therefore the dosage; retrieval is the memory operation; spacing is the timing policy. The strongest learning architecture uses all three, but it does not assume that more repetition, harder retrieval or longer spacing is always better.791027

62Works reviewed

Meta-analyses, reviews, landmark memory experiments and L2 studies.

48L2 experiments in one meta-analysis

Kim & Webb’s synthesis provides the strongest field-specific estimate of spacing.

3Open research figures

Published figures reproduced from CC BY 4.0 Cambridge articles.

Bottom line

Remembering a language is a control problem, not a calendar problem. A good system watches whether the learner can retrieve, how confidently and how recently; it lengthens the gap after success, shortens it after failure, and periodically changes the cue so memory becomes usable rather than merely familiar.

02 · Review method

An evidence-focused integrative review, with deliberate emphasis on delayed retention

This review searched publisher pages, scholarly indexes, bibliographic records and open research materials available on the web through August 7, 2026. Priority was given to (1) quantitative syntheses and systematic reviews of retrieval or spacing, (2) landmark experimental work establishing mechanisms, and (3) peer-reviewed second-language studies that directly manipulated retrieval opportunities, repetition count or practice distribution.

Selection and interpretation rules

Inclusion. A paper had to contribute directly to retrieval practice, repeated encounters, distributed practice, successive relearning, or a language-learning boundary condition relevant to those mechanisms. The final set contains 62 scholarly works, including 30 L2-specific studies.

Outcome hierarchy. Delayed recall/production was treated as stronger evidence of durable memory than immediate posttest performance. Recognition, cued recall, free recall, productive vocabulary, grammar judgement, pronunciation and speech fluency were not treated as interchangeable outcomes.

Figure policy. We did not copy figures from ordinary copyrighted articles. Three figures below come from Cambridge University Press articles explicitly licensed under Creative Commons Attribution 4.0. All other diagrams are author syntheses that summarize patterns without recreating a source figure.

Limit. This is not a preregistered PRISMA systematic review. It is a deep, evidence-focused literature review designed to explain converging mechanisms, conflicting findings and practical implications; the 62-paper matrix makes every included source and review conclusion auditable.

03 · Memory problem

A learner can “know” a French word at 8 p.m. and fail to retrieve it next week

Memory performance depends on the match between what was encoded, the cue available later, and the accessibility of the target at that moment. Repeated exposure can make processing fast and familiar while leaving recall fragile. That is why restudy often performs well immediately but loses to retrieval after a delay.111325

For a second language, this problem multiplies. A lexical item includes orthography, sound, meaning, grammatical behavior, collocations and the ability to retrieve it during comprehension or production. The repetition literature shows that different dimensions of word knowledge develop at different rates; even many encounters do not produce “complete” lexical mastery.555758

Figure 1 · Author synthesis

The durable L2 memory loop

EncodeAttach form to meaning, sound, context and a usable cue.
RetrieveAttempt the target without seeing the answer; production is stronger evidence than recognition.
CorrectFeedback repairs errors and restores the target after a failed or partial retrieval.
Return laterLet accessibility decline enough that the next successful retrieval is not a trivial echo.
Original synthesis based on retrieval-practice, spacing, feedback and successive-relearning evidence.710272937

04 · Retrieval

The act of trying to remember is itself a learning event

Retrieval practice is not simply measurement. When a learner reconstructs peindre from “to paint,” produces a verb form from a sentence cue, or recalls a phrase after hearing its meaning, the retrieval changes the memory that will be available later.131823

01

Keep retrieving after the first correct answer

In a foreign-vocabulary paradigm, Karpicke & Roediger found that additional study after initial mastery contributed little to delayed recall, whereas repeated testing after mastery produced a large benefit.13

02

Make retrieval effortful—but still successful

More difficult successful retrieval can strengthen later memory, but difficulty that causes repeated failure is not automatically desirable. Cue support should fade as the learner stabilizes.14

03

Use productive recall, not recognition alone

Meta-analytic and classroom evidence indicates that recall/production tests often create larger learning gains than recognition-only tests.1830

04

Give corrective feedback

Feedback protects against reinforcing errors and can improve later recall after testing. In language learning, feedback timing is a moderator rather than a universally fixed rule.294142

L2-specific evidence

Barcroft created a tiny retrieval window by delaying the Spanish target word after the picture. That opportunity to try recall improved immediate, 2-day and 1-week vocabulary performance.37 Folse likewise found that repeated retrieval-oriented exercises could outperform a single more elaborate exercise.59

05 · Repetition

Ten encounters are not ten equivalent learning events

“Repetition” hides several variables: Is the answer visible or generated? Is the word encountered in the same sentence or a new one? Is the learner reading, hearing, writing or speaking? Does the repetition occur ten seconds later or ten days later? Research on word learning shows a dose–response relation, but also that encounter count alone cannot describe what is learned.9555658

Passive re-exposure

  • Useful for establishing form and meaning.
  • Can create familiarity without independent recall.
  • Naturalistic reading/viewing gains are moderated by frequency, proficiency, modality and cognateness.576061

Retrieval repetition

  • Requires reconstructing the target before feedback.
  • Repeated successful recall has stronger delayed-memory evidence than repeated study.1318
  • Can be cued, productive, spoken or written.

Contextual variation

  • Changes the cue structure around the same language item.
  • Helps prevent learning from being bound to one prompt.
  • Transfer is not guaranteed; it weakens as new tasks diverge from practice.20

Skill repetition

  • Speaking-task repetition can reduce planning burden and alter fluency.
  • Immediate fluency components can improve while other components worsen.47
  • Procedural targets can prefer shorter gaps than declarative targets.49

06 · Spacing

Forgetting is not merely the enemy; a little forgetting can make the next retrieval productive

Spacing separates learning events so the next encounter is less supported by short-lived activation. The benefit is not “the longer the better.” If the gap is too short, the learner can coast on residual accessibility; if it is too long, retrieval collapses. The best region shifts outward as the desired retention horizon lengthens.123

The field-specific evidence is substantial. Kim & Webb’s 2022 meta-analysis found a medium-to-large overall L2 spacing effect. Short and long spacing performed similarly on immediate tests, but longer spacing did better on delayed tests. Expanding schedules were not statistically superior to equal schedules, and effects varied by learning target, session count, practice type, activity, feedback and retention interval.41

Published Figure A · Rogers, Nakata & Chiu (2025), Figure 4

A recent L2 replication: spaced schedules beat massing after a 10-day delay

Published Figure 4 from Rogers, Nakata and Chiu 2025 comparing posttest performance across intersession intervals with the original Cepeda et al. study
What the figure contributes. The online conceptual replication broadly reproduced the central spacing pattern: spaced conditions significantly outperformed massed study, and adjusted analyses favored several longer gaps over the shortest gap. The authors also observed lower absolute posttest performance than the original study, highlighting context and online-data differences.45 Source article. Reproduced under CC BY 4.0.

There is no evidence-based universal “1–3–7–14” schedule

A fixed sequence can be a useful interface default, but the science supports adaptive intervals. The desired retention interval, success on the last retrieval, knowledge type, number of sessions, age and learning context should all influence the next gap.24143444649

07 · Retrieval × spacing

The highest-value combination is successive relearning

Retrieval and spacing solve different problems. Retrieval strengthens access routes; spacing ensures those routes have to be reconstructed rather than merely reactivated from working memory. Successive relearning formalizes the combination: retrieve to a criterion now, then retrieve to criterion again in later sessions.102627

Figure 2 · Author synthesis

A language-specific successive-relearning cycle

1 · LearnMeaning + form + sound + example. Do not leave encoding ambiguous.
2 · RetrieveHide the answer. Recall or produce it. Use a cue strong enough to permit success.
3 · FeedbackShow the correct target, repair form/pronunciation, then retrieve again.
4 · SpaceLeave the item until some accessibility has faded. Interval depends on horizon and performance.
5 · RelearnRetrieve to criterion again, then increase cue variability and productive demands.
This is an instructional synthesis, not a claim that one exact sequence was tested as a package. It combines the successive-relearning framework with L2 retrieval and spacing evidence.27374041

08 · Language targets

Vocabulary has the deepest evidence base; grammar, pronunciation and fluency complicate the rule

VocabularyStrongest evidence. Long-term benefits appear in classic French/Spanish studies, modern meta-analysis, flashcard/fill-in tasks, contextual reading and online replication.33343537414245
GrammarDistributed syntax/grammar instruction can improve delayed outcomes even when immediate tests show little difference.3648
CollocationsDistributed collocation practice produced gains on studied phrases and transfer to unstudied collocations sharing a node.50
Speech skillsSpacing effects are skill-specific: long and short spacing can tie for fluency, while shorter spacing can favor procedural tone production.4649

Published Figure B · Kakitani & Kormos (2024), Figure 1

The fluency study explicitly separated 1-day and 7-day training schedules

Research design figure from Kakitani and Kormos 2024 showing short-spaced and long-spaced fluency training schedules and posttests
What the figure contributes. The design makes the schedule comparison transparent: four practice sessions were distributed either with 1-day or 7-day intersession intervals, followed by delayed posttests. Despite different learning trajectories, the two spacing schedules produced similar fluency gains at 7 and 28 days.46 Source article. Reproduced under CC BY 4.0.

Published Figure C · Suzuki & Hanzawa (2022), Figure 4

Massed speaking practice can improve one fluency dimension while degrading others

Published results figure from Suzuki and Hanzawa 2022 showing fluency outcomes across massed, short-spaced and long-spaced task repetition groups
What the figure contributes. The study is an antidote to one-dimensional “more practice is better” thinking. Massed task repetition reduced some pauses most on an immediate novel task, yet participants also showed slower articulation and more verbatim repetition, and delayed transfer effects were limited.47 Source article. Reproduced under CC BY 4.0.

09 · Boundary conditions

The literature contains real contradictions—and they are informative

A trustworthy review should explain where the headline effect weakens. Several L2 studies show that spacing schedules interact with age, classroom ecology, knowledge type and the kind of performance being trained.

01

Children in authentic classrooms

Rogers & Cheung first found a short 1-day gap better than an 8-day gap, then in a conceptual replication found no significant difference. Both studies caution against importing adult laboratory ratios directly into primary classrooms.4344

02

Procedural knowledge

Li & DeKeyser found a shorter spacing advantage for procedural tonal-production knowledge even while longer spacing showed some benefit for declarative knowledge.49

03

Fluency

Kakitani & Kormos found 1-day and 7-day schedules similarly effective at delayed tests. Suzuki & Hanzawa found that massed repetition improved some pause outcomes immediately but carried costs for other fluency measures.4647

04

Direct counterevidence

Yan & Zhou reported conditions in which massed vocabulary practice was competitive or superior, including delayed receptive outcomes. Spacing is a strong average effect, not an inviolable law.52

05

Pronunciation schedule

Carpenter & Mueller found blocked practice better than interleaving for French pronunciation rules across four experiments, showing that a “desirable difficulty” imported from another domain can reverse for a specific language target.62

What this means for claims

  • Do not advertise a single spacing interval as scientifically optimal for everyone.
  • Do not equate a better immediate test with better long-term retention.
  • Do not equate receptive recognition with productive speaking ability.
  • Do not infer a cognitive or neurological benefit from spacing beyond the learning outcomes actually measured.
  • Measure forgetting, not merely end-of-lesson accuracy.

10 · Evidence-informed practice architecture

Design the review engine around memory state, not fixed calendar dates

The literature can support an adaptive architecture even though it cannot dictate one universal schedule. A practical system should combine (a) initial meaningful encoding, (b) retrieval before restudy, (c) feedback, (d) repeated retrieval to criterion, (e) spacing across sessions, and (f) cue/context variation for transfer.7132027293741

Ten design commitments supported by the review

  1. Require an attempt before revealing the answer. Recognition can be a scaffold, not the endpoint.
  2. Continue retrieval after first success. First correct recall is not durable mastery.13
  3. Use feedback after errors or uncertain recall. Corrective feedback reduces persistence of wrong answers.29
  4. Distribute review across sessions. A single intensive session is efficient for exposure, not necessarily retention.141
  5. Adapt the next interval to performance and retention horizon. Longer desired retention generally justifies longer optimal gaps, but skill type changes the equation.249
  6. Do not overengineer expanding schedules. Equal spacing often performs similarly to expanding spacing.123841
  7. Track productive and receptive knowledge separately. “I recognize it” and “I can say it” are different states.5558
  8. Vary cues after the memory stabilizes. Transfer decreases as practice and final use diverge; varied retrieval should be built intentionally.20
  9. Use different spacing policies for declarative and procedural targets. Vocabulary recall, grammar judgement, pronunciation and fluency do not share one optimum.46474962
  10. Evaluate with delayed tests. Immediate success can conceal later forgetting and can favor massing.113348

Illustrative scheduling defaults can be used in software, but they should be labeled as defaults—not as a universal scientific prescription—and should adapt after each retrieval.

An illustrative adaptive sequence—not a “scientific magic interval”

Session 0: encode + retrieve to criterion → next review: roughly 1 day if fragile → then 3–4 days → 7–10 days → 2–4 weeks → longer maintenance intervals. After a failure, shorten the gap and restore cue support; after easy repeated success, lengthen it. The exact values are interface heuristics constrained by the evidence, not values directly established as universally optimal.2341

11 · Evidence matrix

All 62 sources, with the conclusion used in this review

The table makes the review auditable. Each “main conclusion” is a concise paraphrase of the finding or synthesis relevant to retrieval, repetition or spacing; it is not a quotation from the paper.

Literature reviewed and review-level conclusion
#StudyDomain / designMain conclusion used hereWhy it matters to this review
01Cepeda, Pashler, Vul, Wixted & Rohrer (2006)
Distributed practice in verbal recall tasks: A review and quantitative synthesis		Meta-analysis · spacingAcross hundreds of comparisons, distributed practice reliably outperformed massed practice; the size of the benefit depended strongly on the delay to the final test and the spacing interval.Foundation for the spacing effect and lag–retention interaction.
02Cepeda, Vul, Rohrer, Wixted & Pashler (2008)
Spacing effects in learning: A temporal ridgeline of optimal retention		Large experiment · spacingThe relation between study gap and later retention is non-linear: an intermediate gap is best, and the optimal gap becomes longer as the desired retention interval becomes longer.Shows why there is no universal “best” number of days.
03Cepeda, Coburn, Rohrer, Wixted, Mozer & Pashler (2009)
Optimizing distributed practice: Theoretical analysis and practical implications		Experiment · spacingFor foreign vocabulary and other materials, very short gaps were inferior for delayed retention; performance followed a non-monotonic function in which the best gap depended on the test delay.Directly influenced later L2 spacing replications.
04Carpenter, Cepeda, Rohrer, Kang & Pashler (2012)
Using spacing to enhance diverse forms of learning: Review of recent research and implications for instruction		Review · spacingSpacing benefits extend beyond simple word lists to many kinds of educational material, although the precise schedule should be selected in relation to the intended retention period.Generalizability of spacing to instruction.
05Kang (2016)
Spaced repetition promotes efficient and effective learning: Policy implications for instruction		Review · spacingA large experimental literature supports distributing learning over time rather than cramming, especially when retention beyond the immediate test is the goal.Translates spacing research into educational policy.
06Smolen, Zhang & Byrne (2016)
The right time to learn: Mechanisms and optimization of spaced learning		Neuroscience review · spacingSpaced training is robust across species and tasks; molecular and systems-level evidence suggests that separated learning events can interact with time-dependent consolidation processes.Mechanistic support without claiming a single biological clock for language study.
07Carpenter, Pan & Butler (2022)
The science of effective learning with spacing and retrieval practice		Review · retrieval + spacingRetrieval practice and spacing are among the most consistently supported learning strategies, and their benefits often persist across educational materials, learners and delays.Modern synthesis of the two core mechanisms.
08Dunlosky, Rawson, Marsh, Nathan & Willingham (2013)
Improving students’ learning with effective learning techniques: Promising directions from cognitive and educational psychology		Review · learning techniquesPractice testing and distributed practice received the highest utility ratings among the ten techniques reviewed, whereas rereading and highlighting were judged much less dependable.Places retrieval and spacing above passive restudy.
09Toppino & Gerbier (2014)
About practice: Repetition, spacing, and abstraction		Review · repetition/spacingRepeated events are not equivalent: timing, retrieval of prior presentations, encoding variability and abstraction all influence what repetition produces.Explains why “number of repetitions” is an incomplete dosage metric.
10Latimier, Peyre & Ramus (2021)
A meta-analytic review of the benefit of spacing out retrieval practice episodes on retention		Meta-analysis · spaced retrievalSpacing retrieval episodes improves later retention, supporting the combination of testing and distributed practice rather than treating them as separate interventions.Evidence for spaced retrieval as a joint mechanism.
11Roediger & Karpicke (2006)
Test-enhanced learning: Taking memory tests improves long-term retention		Experiment · retrievalRestudying can look better on an immediate test, but retrieval practice produces superior retention after longer delays.Classic demonstration that short-term fluency can mislead.
12Karpicke & Roediger (2007)
Expanding retrieval practice promotes short-term retention, but equally spaced retrieval enhances long-term retention		Experiment · retrieval scheduleExpanding retrieval can help early performance, but it was not consistently superior for long-term retention; equal spacing can match or exceed it.Counters the idea that expanding intervals are always best.
13Karpicke & Roediger (2008)
The critical importance of retrieval for learning		Experiment · foreign vocabularyAfter a foreign-language item had first been recalled correctly, additional study produced little delayed benefit, whereas continued testing produced a large delayed-recall advantage.Direct evidence that successful retrieval should continue after first mastery.
14Pyc & Rawson (2009)
Testing the retrieval effort hypothesis: Does greater difficulty correctly recalling information lead to higher levels of memory?		Experiment · retrieval effortWhen retrieval remained successful, more effortful retrieval generally produced stronger later memory than very easy retrieval.Supports desirable difficulty—but only when success remains attainable.
15Pyc & Rawson (2010)
Why testing improves memory: Mediator effectiveness hypothesis		Experiment · retrieval mechanismRetrieval practice can strengthen useful mediators linking cues to targets, helping explain why an act of recall changes later accessibility.Mechanistic account relevant to form–meaning links.
16Butler (2010)
Repeated testing produces superior transfer of learning relative to repeated studying		Experiment · retrieval transferRepeated retrieval improved not only retention of practiced material but also transfer to new questions more than repeated study.Supports varied retrieval when the goal is flexible use.
17Karpicke & Blunt (2011)
Retrieval practice produces more learning than elaborative studying with concept mapping		Experiment · retrievalRetrieval practice produced strong gains on comprehension and inference tests, indicating that retrieval can support meaningful learning rather than rote recall alone.Reframes testing as a learning event.
18Rowland (2014)
The effect of testing versus restudy on retention: A meta-analytic review of the testing effect		Meta-analysis · retrievalAcross experiments, retrieval practice improved delayed retention over restudy; effects were larger for recall-based initial tests than for recognition-only tests.Supports productive recall over recognition-only practice.
19Adesope, Trevisan & Sundararajan (2017)
Rethinking the use of tests: A meta-analysis of practice testing		Meta-analysis · retrievalPractice testing showed a robust positive effect across educationally relevant conditions, with moderator effects related to test format, feedback and delay.Broad educational evidence for retrieval practice.
20Pan & Rickard (2018)
Transfer of test-enhanced learning: Meta-analytic review and synthesis		Meta-analysis · retrieval transferRetrieval-practice benefits can transfer beyond the exact practiced question, but transfer declines as the final task becomes less similar to the retrieval activity.Warns that vocabulary flashcards alone do not guarantee conversational transfer.
21Agarwal, Nunes & Blunt (2021)
Retrieval practice consistently benefits student learning: A systematic review of applied research in schools and classrooms		Systematic review · classroomsAcross 50 classroom experiments, retrieval practice was usually beneficial, although the evidence base was geographically uneven and implementation varied.Supports classroom use while highlighting external-validity gaps.
22Moreira, Pinto, Starling & Jaeger (2019)
Retrieval practice in classroom settings: A review of applied research		Review · classroomsApplied classroom studies generally support retrieval practice, but outcomes depend on test format, feedback, content and implementation.Bridges lab evidence to authentic instruction.
23Carrier & Pashler (1992)
The influence of retrieval on retention		Experiment · retrievalAttempting and succeeding at retrieval strengthened later retention beyond additional exposure, establishing an early causal basis for the testing effect.Foundational retrieval evidence.
24Slamecka & Graf (1978)
The generation effect: Delineation of a phenomenon		Experiment · generationInformation generated by the learner was remembered better than information merely read, across several experimental conditions.Precursor to retrieval/generation approaches in vocabulary learning.
25Bjork, Dunlosky & Kornell (2013)
Self-regulated learning: Beliefs, techniques, and illusions		Review · metacognitionLearners often mistake immediate processing fluency for durable learning and therefore underuse difficult but effective strategies such as testing and spacing.Explains why cramming can feel more effective than it is.
26Rawson, Vaughn, Walsh & Dunlosky (2018)
Investigating and explaining the effects of successive relearning on long-term retention		Experiment · successive relearningRepeatedly reaching a correct-recall criterion across spaced sessions produced strong long-term retention; benefits reflect the combination of retrieval success and relearning across time.Operational model for durable language review.
27Rawson & Dunlosky (2022)
Successive relearning: An underexplored but potent technique for obtaining and maintaining knowledge		Review · successive relearningThe authors define successive relearning as recalling to criterion and then relearning to criterion in later spaced sessions, emphasizing both potency and time efficiency.Best compact architecture for combining retrieval and spacing.
28Pashler, Rohrer, Cepeda & Carpenter (2007)
Enhancing learning and retarding forgetting: Choices and consequences		Review · learning/forgettingInstructional choices about spacing, testing and feedback alter both acquisition and forgetting; optimizing learning requires considering retention over meaningful delays.Shifts evaluation from immediate scores to forgetting curves.
29Butler & Roediger (2008)
Feedback enhances the positive effects and reduces the negative effects of multiple-choice testing		Experiment · feedbackFeedback after testing increased later correct recall and reduced persistence of wrong multiple-choice lures.Supports corrective feedback after retrieval attempts.
30McDaniel, Roediger & McDermott (2007)
Generalizing test-enhanced learning from the laboratory to the classroom		Classroom experiments · retrievalAcross educational materials, testing improved retention; production tests often yielded larger benefits than multiple choice, and feedback added value.Supports low-stakes productive retrieval in courses.
31Karpicke, Butler & Roediger (2009)
Metacognitive strategies in student learning: Do students practise retrieval when they study on their own?		Survey + memory evidenceStudents reported rereading more often than self-testing and showed limited awareness of retrieval practice as a learning tool.Design implication: systems must prompt retrieval rather than expect learners to choose it.
32Kornell (2009)
Optimising learning using flashcards: Spacing is more effective than cramming		Experiment · flashcardsUsing larger flashcard stacks and spacing study across time improved learning over smaller, more massed stacks and cramming; learners often misjudged the better method.Directly relevant to vocabulary-app review algorithms.
33Bloom & Shuell (1981)
Effects of massed and distributed practice on the learning and retention of second-language vocabulary		L2 vocabulary · classroomMassed and distributed study were similar immediately, but distributed study produced substantially better retention several days later in high-school French vocabulary learning.Early classroom L2 demonstration of delayed spacing benefit.
34Bahrick & Phelps (1987)
Retention of Spanish vocabulary over 8 years		L2 vocabulary · very long-termThe spacing used during original Spanish vocabulary learning predicted memory many years later; longer intersession spacing was associated with more durable retention.Rare multi-year evidence for language memory.
35Bahrick, Bahrick, Bahrick & Bahrick (1993)
Maintenance of foreign language vocabulary and the spacing effect		L2 vocabulary · longitudinalLonger spacing slowed initial acquisition but substantially improved later retention; widely spaced relearning could achieve comparable durability with fewer sessions.Shows the acquisition–retention trade-off.
36Bird (2010)
Effects of distributed practice on the acquisition of second language English syntax		L2 grammar · longitudinalDistributed grammar practice produced superior long-term outcomes to massed practice even when short-term differences were smaller.Extends spacing beyond vocabulary to syntax.
37Barcroft (2007)
Effects of opportunities for word retrieval during second language vocabulary learning		L2 vocabulary · retrievalIntroducing a brief delay that allowed learners to attempt Spanish-word retrieval improved immediate, 2-day and 1-week vocabulary scores relative to presentation without a retrieval opportunity.Direct L2 retrieval-practice evidence.
38Nakata (2015)
Effects of expanding and equal spacing on second language vocabulary learning: Does gradually increasing spacing increase vocabulary learning?		L2 vocabulary · scheduleThe amount of spacing strongly affected learning, while expanding spacing had limited and inconsistent advantages over equal spacing.Prioritize adequate spacing over a fetish for expanding schedules.
39Nakata & Elgort (2021)
Effects of spacing on contextual vocabulary learning: Spacing facilitates the acquisition of explicit, but not tacit, vocabulary knowledge		L2 vocabulary in contextSpacing supported explicit vocabulary knowledge from contextual reading, but evidence for tacit/implicit lexical knowledge was weaker.Knowledge type matters.
40Nakata, Suzuki & He (2023)
Costs and benefits of spacing for second language vocabulary learning: Does relearning override the positive and negative effects of spacing?		L2 vocabulary · registered reportLong spacing during both initial learning and relearning produced better long-term retention and higher words-learned-per-trial efficiency despite taking more elapsed time.Spacing can improve efficiency when measured as durable learning, not session length.
41Kim & Webb (2022)
The effects of spaced practice on second language learning: A meta-analysis		L2 meta-analysis · spacingAcross 98 effect sizes from 48 experiments (N=3,411), spacing had a medium-to-large effect; longer spacing was better on delayed tests, and equal versus expanding schedules were statistically equivalent overall.Highest-level L2-specific synthesis in this review.
42Kim & Webb (2023)
Does spaced practice have the same effects on different second language vocabulary learning activities? Fill-in-the-blanks versus flashcards		L2 vocabulary · activity × spacingWith 150 Korean learners, spaced practice outperformed massed practice on a two-week delayed test for both flashcards and fill-in-the-blanks; immediate effects varied by activity and feedback timing had no overall effect.Practice type moderates short-term spacing effects.
43Rogers & Cheung (2020)
Input spacing and the learning of L2 vocabulary in a classroom context		Child L2 classroom · spacingIn authentic Hong Kong primary classrooms, a 1-day interval outperformed an 8-day interval after four weeks, suggesting that laboratory lag ratios do not simply transfer to young learners and real lessons.Important ecological boundary condition.
44Rogers & Cheung (2021)
Does it matter when you review? Input spacing, ecological validity, and the learning of L2 vocabulary		Child L2 classroom · replicationA conceptual replication found large learning gains but no significant difference between 1-day and 8-day schedules after four weeks.Replicates the caution that age/context can flatten lag differences.
45Rogers, Nakata & Chiu (2025)
Optimizing distributed practice online: A conceptual replication of Cepeda et al. (2009)		L2 vocabulary · online replicationAll spaced schedules significantly outperformed massing after a 10-day delay; after controls, several longer gaps also outperformed the shortest gap, broadly replicating the spacing function online.Recent direct replication with an openly licensed published figure.
46Kakitani & Kormos (2024)
The effects of distributed practice on second language fluency development		L2 oral fluency · spacingFour fluency-training sessions separated by either 1 day or 7 days produced similar gains on 7- and 28-day posttests despite different training trajectories.For procedural fluency, longer spacing was not automatically better.
47Suzuki & Hanzawa (2022)
Massed task repetition is a double-edged sword for fluency development: An EFL classroom study		L2 oral fluency · task repetitionMassed repetition reduced some pause measures most strongly on an immediate novel task, but also slowed articulation and increased repetitions; delayed transfer differences were limited.A vivid example of immediate performance gains not equaling broad durable fluency.
48Miles (2014)
Spaced vs. massed distribution instruction for L2 grammar learning		L2 grammar · classroomImmediate grammar gains were statistically similar, but the spaced group outperformed the massed group on delayed error analysis/correction.Spacing may emerge only on delayed or demanding grammar measures.
49Li & DeKeyser (2019)
Distribution of practice effects in the acquisition and retention of L2 Mandarin tonal word production		L2 tone production · declarative/proceduralSpacing interacted with knowledge type: longer spacing showed some advantage for long-term declarative knowledge, whereas shorter spacing favored procedural tonal-production knowledge.Strong evidence that skill type changes the optimal schedule.
50Yamagata, Nakata & Rogers (2023)
Effects of distributed practice on the acquisition of verb-noun collocations		L2 collocations · spacing/transferA collocation-spaced schedule produced the largest gains for both studied and novel collocations sharing the same node verb.Spacing can support phrase learning and limited transfer.
51Çekiç & Bakla (2019)
The effects of spacing patterns on incidental L2 vocabulary learning through reading with electronic glosses		L2 vocabulary · spacing patternWith the same overall number of encounters, distributing encounters across more weekly sessions produced better retention than concentrating them into fewer sessions.Schedule matters even when encounter count is held constant.
52Yan & Zhou (2023)
Effects of massed and distributed practice on L2 vocabulary learning		L2 vocabulary · counterevidenceIn this university sample, massed practice was competitive and in some analyses more effective, including delayed receptive knowledge, showing that a spacing advantage is not inevitable in every design.Required counterexample against overgeneralization.
53Muqaibal, Kasprowicz & Tissot (2026)
Evaluating the impact of spaced practice using CALL on vocabulary learning in the classroom		L2 vocabulary · CALL classroomBoth 1-day and 7-day computer-assisted review schedules improved vocabulary over control, but the two spaced schedules did not differ significantly.Technology can make spacing effective without revealing one optimal interval.
54Rogers (2023)
Spacing effects in task repetition research		L2 conceptual review · task repetitionTask-repetition research must distinguish intersession interval, retention interval, task similarity and the type of fluency outcome; failure to do so can make spacing results look contradictory.Framework for procedural speaking practice.
55Webb (2007)
The effects of repetition on vocabulary knowledge		L2 vocabulary · repetition dosageMore encounters produced greater gains across multiple dimensions of word knowledge, but even ten encounters did not guarantee complete lexical mastery.Repetition count helps, but mastery is multidimensional.
56Rott (1999)
The effect of exposure frequency on intermediate language learners’ incidental vocabulary acquisition and retention through reading		L2 vocabulary · reading repetitionMore repeated encounters during reading increased incidental vocabulary learning and retention, with higher-frequency conditions generally producing larger gains.Repeated contextual exposure matters.
57Waring & Takaki (2003)
At what rate do learners learn and retain new vocabulary from reading a graded reader?		L2 vocabulary · extensive readingIncidental vocabulary gains from reading were real but modest and subject to attrition; encounter frequency predicted learning, yet repeated exposure alone did not ensure durable recall.Highlights limits of exposure-only repetition.
58Pellicer-Sánchez (2016)
Incidental L2 vocabulary acquisition from and while reading: An eye-tracking study		L2 vocabulary · repeated readingRepeated encounters altered online processing and produced substantial but incomplete form/meaning learning; different lexical measures showed different levels of acquisition.A word is not simply “known” after N exposures.
59Folse (2006)
The effect of type of written exercise on L2 vocabulary retention		L2 vocabulary · repetition/retrievalCompleting three retrieval-oriented fill-in exercises produced better retention than a single exercise condition, indicating that number of meaningful retrieval opportunities can outweigh nominal depth of one task.Supports repeated retrieval opportunities in practice design.
60Vidal (2011)
A comparison of the effects of reading and listening on incidental vocabulary acquisition		L2 vocabulary · input modality/frequencyReading generally produced larger incidental vocabulary gains and retention than listening, with proficiency and frequency of occurrence moderating outcomes.Repeated input interacts with modality and proficiency.
61Peters & Webb (2018)
Incidental vocabulary acquisition through viewing L2 television and factors that affect learning		L2 vocabulary · audiovisual repetitionViewing a full-length television program produced incidental meaning gains; frequency of occurrence, prior vocabulary and cognateness predicted learning.Naturalistic repetition is conditional on lexical and learner factors.
62Carpenter & Mueller (2013)
The effects of interleaving versus blocking on foreign language pronunciation learning		L2 pronunciation · scheduleAcross four experiments, blocked practice produced better learning of French pronunciation rules than interleaving, contrary to a common generalization from category-learning research.Another reminder that practice schedules depend on the target skill.

12 · Conclusions

The science favors a memory loop, not a repetition counter

The strongest conclusion from this literature is simple but consequential: durable second-language learning is better predicted by what the learner has successfully retrieved across time than by how many times the material has merely been seen.

Repetition matters because language knowledge requires multiple encounters.5556 Retrieval matters because rebuilding the answer strengthens later accessibility and often outperforms restudy.131837 Spacing matters because it prevents every repetition from becoming a near-duplicate while tuning difficulty to the desired retention horizon.1241 The combination—successful recall, feedback, then later relearning—is the most defensible general architecture.1027

But the boundary conditions are equally important. Longer spacing is not always better; expanding schedules are not reliably superior; children and authentic classrooms can behave differently from adult laboratory samples; and procedural pronunciation or fluency targets may respond differently from declarative vocabulary.384344464962 A scientifically credible language-learning system should therefore be adaptive, outcome-specific and tested on delayed transfer—not marketed around a fixed “forgetting curve” schedule.

Practical takeaway

See it once. Retrieve it many times. Space those retrievals. Change the cue. Test it later.

That is the most defensible one-line translation of more than four decades of modern retrieval research and more than four decades of second-language spacing research.

13 · References

Literature reviewed

62 works are listed in the same numbering used throughout the article and evidence matrix. DOI links open the publisher or DOI record.

  1. Cepeda, Pashler, Vul, Wixted & Rohrer. 2006. Distributed practice in verbal recall tasks: A review and quantitative synthesis. Psychological Bulletin 132(3), 354–380. doi:10.1037/0033-2909.132.3.354
  2. Cepeda, Vul, Rohrer, Wixted & Pashler. 2008. Spacing effects in learning: A temporal ridgeline of optimal retention. Psychological Science 19(11), 1095–1102. doi:10.1111/j.1467-9280.2008.02209.x
  3. Cepeda, Coburn, Rohrer, Wixted, Mozer & Pashler. 2009. Optimizing distributed practice: Theoretical analysis and practical implications. Experimental Psychology 56(4), 236–246. doi:10.1027/1618-3169.56.4.236
  4. Carpenter, Cepeda, Rohrer, Kang & Pashler. 2012. Using spacing to enhance diverse forms of learning: Review of recent research and implications for instruction. Educational Psychology Review 24, 369–378. doi:10.1007/s10648-012-9205-z
  5. Kang. 2016. Spaced repetition promotes efficient and effective learning: Policy implications for instruction. Policy Insights from the Behavioral and Brain Sciences 3(1), 12–19. doi:10.1177/2372732215624708
  6. Smolen, Zhang & Byrne. 2016. The right time to learn: Mechanisms and optimization of spaced learning. Nature Reviews Neuroscience 17, 77–88. doi:10.1038/nrn.2015.18
  7. Carpenter, Pan & Butler. 2022. The science of effective learning with spacing and retrieval practice. Nature Reviews Psychology 1, 496–511. doi:10.1038/s44159-022-00089-1
  8. Dunlosky, Rawson, Marsh, Nathan & Willingham. 2013. Improving students’ learning with effective learning techniques: Promising directions from cognitive and educational psychology. Psychological Science in the Public Interest 14(1), 4–58. doi:10.1177/1529100612453266
  9. Toppino & Gerbier. 2014. About practice: Repetition, spacing, and abstraction. Psychology of Learning and Motivation 60, 113–189. doi:10.1016/B978-0-12-800090-8.00004-4
  10. Latimier, Peyre & Ramus. 2021. A meta-analytic review of the benefit of spacing out retrieval practice episodes on retention. Educational Psychology Review 33, 959–987. doi:10.1007/s10648-020-09572-8
  11. Roediger & Karpicke. 2006. Test-enhanced learning: Taking memory tests improves long-term retention. Psychological Science 17(3), 249–255. doi:10.1111/j.1467-9280.2006.01693.x
  12. Karpicke & Roediger. 2007. Expanding retrieval practice promotes short-term retention, but equally spaced retrieval enhances long-term retention. Journal of Experimental Psychology: Learning, Memory, and Cognition 33(4), 704–719. doi:10.1037/0278-7393.33.4.704
  13. Karpicke & Roediger. 2008. The critical importance of retrieval for learning. Science 319(5865), 966–968. doi:10.1126/science.1152408
  14. Pyc & Rawson. 2009. Testing the retrieval effort hypothesis: Does greater difficulty correctly recalling information lead to higher levels of memory?. Journal of Memory and Language 60(4), 437–447. doi:10.1016/j.jml.2009.01.004
  15. Pyc & Rawson. 2010. Why testing improves memory: Mediator effectiveness hypothesis. Science 330(6002), 335. doi:10.1126/science.1191465
  16. Butler. 2010. Repeated testing produces superior transfer of learning relative to repeated studying. Journal of Experimental Psychology: Learning, Memory, and Cognition 36(5), 1118–1133. doi:10.1037/a0019902
  17. Karpicke & Blunt. 2011. Retrieval practice produces more learning than elaborative studying with concept mapping. Science 331(6018), 772–775. doi:10.1126/science.1199327
  18. Rowland. 2014. The effect of testing versus restudy on retention: A meta-analytic review of the testing effect. Psychological Bulletin 140(6), 1432–1463. doi:10.1037/a0037559
  19. Adesope, Trevisan & Sundararajan. 2017. Rethinking the use of tests: A meta-analysis of practice testing. Review of Educational Research 87(3), 659–701. doi:10.3102/0034654316689306
  20. Pan & Rickard. 2018. Transfer of test-enhanced learning: Meta-analytic review and synthesis. Psychological Bulletin 144(7), 710–756. doi:10.1037/bul0000151
  21. Agarwal, Nunes & Blunt. 2021. Retrieval practice consistently benefits student learning: A systematic review of applied research in schools and classrooms. Educational Psychology Review 33, 1409–1453. doi:10.1007/s10648-021-09595-9
  22. Moreira, Pinto, Starling & Jaeger. 2019. Retrieval practice in classroom settings: A review of applied research. Frontiers in Education 4, 5. doi:10.3389/feduc.2019.00005
  23. Carrier & Pashler. 1992. The influence of retrieval on retention. Memory & Cognition 20, 633–642. doi:10.3758/BF03202713
  24. Slamecka & Graf. 1978. The generation effect: Delineation of a phenomenon. Journal of Experimental Psychology: Human Learning and Memory 4(6), 592–604. doi:10.1037/0278-7393.4.6.592
  25. Bjork, Dunlosky & Kornell. 2013. Self-regulated learning: Beliefs, techniques, and illusions. Annual Review of Psychology 64, 417–444. doi:10.1146/annurev-psych-113011-143823
  26. Rawson, Vaughn, Walsh & Dunlosky. 2018. Investigating and explaining the effects of successive relearning on long-term retention. Journal of Experimental Psychology: Applied 24(1), 57–71. doi:10.1037/xap0000146
  27. Rawson & Dunlosky. 2022. Successive relearning: An underexplored but potent technique for obtaining and maintaining knowledge. Current Directions in Psychological Science 31(4), 362–368. doi:10.1177/09637214221100484
  28. Pashler, Rohrer, Cepeda & Carpenter. 2007. Enhancing learning and retarding forgetting: Choices and consequences. Psychonomic Bulletin & Review 14(2), 187–193. doi:10.3758/BF03194050
  29. Butler & Roediger. 2008. Feedback enhances the positive effects and reduces the negative effects of multiple-choice testing. Memory & Cognition 36(3), 604–616. doi:10.3758/MC.36.3.604
  30. McDaniel, Roediger & McDermott. 2007. Generalizing test-enhanced learning from the laboratory to the classroom. Psychonomic Bulletin & Review 14(2), 200–206. doi:10.3758/BF03194052
  31. Karpicke, Butler & Roediger. 2009. Metacognitive strategies in student learning: Do students practise retrieval when they study on their own?. Memory 17(4), 471–479. doi:10.1080/09658210802647009
  32. Kornell. 2009. Optimising learning using flashcards: Spacing is more effective than cramming. Applied Cognitive Psychology 23, 1297–1317. doi:10.1002/acp.1537
  33. Bloom & Shuell. 1981. Effects of massed and distributed practice on the learning and retention of second-language vocabulary. Journal of Educational Research 74(4), 245–248. doi:10.1080/00220671.1981.10885317
  34. Bahrick & Phelps. 1987. Retention of Spanish vocabulary over 8 years. Journal of Experimental Psychology: Learning, Memory, and Cognition 13(2), 344–349. doi:10.1037/0278-7393.13.2.344
  35. Bahrick, Bahrick, Bahrick & Bahrick. 1993. Maintenance of foreign language vocabulary and the spacing effect. Psychological Science 4(5), 316–321. doi:10.1111/j.1467-9280.1993.tb00571.x
  36. Bird. 2010. Effects of distributed practice on the acquisition of second language English syntax. Applied Psycholinguistics 31(4), 635–650. doi:10.1017/S0142716410000172
  37. Barcroft. 2007. Effects of opportunities for word retrieval during second language vocabulary learning. Language Learning 57(1), 35–56. doi:10.1111/j.1467-9922.2007.00398.x
  38. Nakata. 2015. Effects of expanding and equal spacing on second language vocabulary learning: Does gradually increasing spacing increase vocabulary learning?. Studies in Second Language Acquisition 37(4), 677–711. doi:10.1017/S0272263114000825
  39. Nakata & Elgort. 2021. Effects of spacing on contextual vocabulary learning: Spacing facilitates the acquisition of explicit, but not tacit, vocabulary knowledge. Second Language Research 37(2), 233–260. doi:10.1177/0267658320927764
  40. Nakata, Suzuki & He. 2023. Costs and benefits of spacing for second language vocabulary learning: Does relearning override the positive and negative effects of spacing?. Language Learning 73(3), 799–834. doi:10.1111/lang.12553
  41. Kim & Webb. 2022. The effects of spaced practice on second language learning: A meta-analysis. Language Learning 72, 269–319. doi:10.1111/lang.12479
  42. Kim & Webb. 2023. Does spaced practice have the same effects on different second language vocabulary learning activities? Fill-in-the-blanks versus flashcards. Modern Language Journal 107(4), 944–964. doi:10.1111/modl.12879
  43. Rogers & Cheung. 2020. Input spacing and the learning of L2 vocabulary in a classroom context. Language Teaching Research 24(5), 616–641. doi:10.1177/1362168818805251
  44. Rogers & Cheung. 2021. Does it matter when you review? Input spacing, ecological validity, and the learning of L2 vocabulary. Studies in Second Language Acquisition 43(5), 1138–1156. doi:10.1017/S0272263120000236
  45. Rogers, Nakata & Chiu. 2025. Optimizing distributed practice online: A conceptual replication of Cepeda et al. (2009). Studies in Second Language Acquisition 47(1), 417–439. doi:10.1017/S0272263124000706
  46. Kakitani & Kormos. 2024. The effects of distributed practice on second language fluency development. Studies in Second Language Acquisition 46(3), 770–794. doi:10.1017/S0272263124000251
  47. Suzuki & Hanzawa. 2022. Massed task repetition is a double-edged sword for fluency development: An EFL classroom study. Studies in Second Language Acquisition 44(2), 536–561. doi:10.1017/S0272263121000358
  48. Miles. 2014. Spaced vs. massed distribution instruction for L2 grammar learning. System 42, 412–428. doi:10.1016/j.system.2014.01.014
  49. Li & DeKeyser. 2019. Distribution of practice effects in the acquisition and retention of L2 Mandarin tonal word production. Modern Language Journal 103, 607–628. doi:10.1111/modl.12580
  50. Yamagata, Nakata & Rogers. 2023. Effects of distributed practice on the acquisition of verb-noun collocations. Studies in Second Language Acquisition 45(2), 291–317. doi:10.1017/S0272263122000225
  51. Çekiç & Bakla. 2019. The effects of spacing patterns on incidental L2 vocabulary learning through reading with electronic glosses. Instructional Science 47, 353–371. doi:10.1007/s11251-019-09483-4
  52. Yan & Zhou. 2023. Effects of massed and distributed practice on L2 vocabulary learning. System 115, 103049. doi:10.1016/j.system.2023.103049
  53. Muqaibal, Kasprowicz & Tissot. 2026. Evaluating the impact of spaced practice using CALL on vocabulary learning in the classroom. Language Teaching Research 30(1), 365–402. doi:10.1177/13621688221146146
  54. Rogers. 2023. Spacing effects in task repetition research. Language Learning 73, 445–474. doi:10.1111/lang.12526
  55. Webb. 2007. The effects of repetition on vocabulary knowledge. Applied Linguistics 28(1), 46–65. doi:10.1093/applin/aml048
  56. Rott. 1999. The effect of exposure frequency on intermediate language learners’ incidental vocabulary acquisition and retention through reading. Studies in Second Language Acquisition 21(4), 589–619. doi:10.1017/S0272263199004039
  57. Waring & Takaki. 2003. At what rate do learners learn and retain new vocabulary from reading a graded reader?. Reading in a Foreign Language 15(2), 130–163. doi:10.64152/10125/66776
  58. Pellicer-Sánchez. 2016. Incidental L2 vocabulary acquisition from and while reading: An eye-tracking study. Studies in Second Language Acquisition 38(1), 97–130. doi:10.1017/S0272263115000224
  59. Folse. 2006. The effect of type of written exercise on L2 vocabulary retention. TESOL Quarterly 40(2), 273–293. doi:10.2307/40264523
  60. Vidal. 2011. A comparison of the effects of reading and listening on incidental vocabulary acquisition. Language Learning 61(1), 219–258. doi:10.1111/j.1467-9922.2010.00593.x
  61. Peters & Webb. 2018. Incidental vocabulary acquisition through viewing L2 television and factors that affect learning. Studies in Second Language Acquisition 40(3), 551–577. doi:10.1017/S0272263117000407
  62. Carpenter & Mueller. 2013. The effects of interleaving versus blocking on foreign language pronunciation learning. Memory & Cognition 41(5), 671–682. doi:10.3758/s13421-012-0291-4