Live·Open questions in longevity research
Questions

Find genuinely new, falsifiable hypotheses in empirical memetics and recommend the most promising theories and experiments. Interpret memetics as the transmission, transformation, competition and persistence of cultural information, including internet memes, narratives and cultural practices. Produce a research agenda, not a campaign to manipulate people. Identify the unresolved mechanisms using cultural evolution, cognitive science, network science, information theory and computational social science, including changes introduced by recommendation algorithms and generative AI. Distinguish established theories from new conjectures and check whether each proposed mechanism is already known under another name. Prioritize approximately five hypothesis families by substantive scientific novelty, explanatory power, discriminating testability, feasibility and expected information gain; rank the best first experiment. For each shortlisted theory give an operational definition of the transmitted unit, a causal mechanism or formal model, plausible competing explanations, contrasting quantitative predictions, a decisive experiment with manipulations and controls, measurable primary outcomes, power-analysis inputs rather than invented sample-size precision, major confounds and a result that would falsify the theory. Separate reach, copying fidelity, semantic change, adoption and persistence. Include both an affordable initial experiment and the stronger validation needed for a general claim. Assess what existing primary evidence actually establishes; do not call plausible extensions proven discoveries. User request in Russian: «хочу найти новые гипотезы в сфере меметиков; предложи самые перспективные эксперименты и теории». Return published question and hypothesis pages, complete Russian versions and English originals, with poster sheets.

Does earlier exposure change what people keep passing on?

The question as the research states itAfter recommendations reset and sharing opportunities match, does earlier exposure still change what people pass on?

Earlier encounters could change what people remember, and what they remember could in turn affect what they later pass on. Separately, software that stores earlier activity could keep showing particular material, creating more occasions for its continued transmission.

The whole reason

Those two routes can produce similar visible patterns of persistence while locating the influence of the past in different places. Attributing a software-dependent pattern to lasting changes in people would overstate how well the material can survive without that software history. Attributing a person-dependent pattern entirely to software would miss a possible route by which transmission continues after recommendations change.

The question in full

The question concerns whether earlier encounters with cultural material leave changes in people that affect how long they continue passing it on. Cultural material means information or practices that can move between people, such as stories, images, or ways of doing something. The comparison is between different earlier exposure histories after the recommendation software has had its stored history cleared and people have equal later chances to reproduce or share the material. It asks whether any continuing difference in transmission reflects something retained in people, or whether the apparent history effect depended entirely on information retained by the software. Resetting the software and equalizing later opportunities are stipulated conditions, not findings established by the supplied sources.

Competing hypotheses

These hypotheses propose different mechanisms. Comparing their predictions helps identify observations that could distinguish them.

  1. 01Confirmed understanding by an unreachable recipient may weaken the sender's memoryIn online narrative chains, confirmed understanding by an unreachable recipient may reduce the sender's memory while later tellings survive. Reject a distinct mechanism if receipt and archive conditions behave alike or ordinary offloading and goal completion predict the joint pattern.
  2. 02Exposure order may fragment private free time and block long cultural practicesPeople may retain exposure order as private time reservations, rejecting long cultural acts despite equal real opportunity after recommendation reset. Reject this distinct mechanism if calibrated ordinary time and choice models reproduce all predicted contrasts, or if private reservations are absent.
  3. 03Exposure history may bias sleep cueing to preserve cultural source lineagesAfter recommendation reset, exposure history may determine how source-linked sleep cues preserve details and cultural descendants; reject the distinct mechanism if ordinary memory and cueing models explain later transmission, or awakenings or unequal waking rehearsal explain the cue effects.
  4. 04Reassigning past effort to another cultural lineage may transfer its persistence advantageIn equally learned narrative-sharing practices after recommendation reset, crediting the same completed work to another lineage may move voluntary persistence without changing skill or memory fidelity; the distinct claim fails if ordinary accounting and effort justification predict the transfer
Each entry represents a published hypothesis. Where no hypotheses are published yet, the entries show possible answers to the scientific question.

What results would tell us about the hypotheses

Choose a possible result to see which hypothesis it would support, what the alternatives predict, and what would need to be tested next.

If we observe
Within reset/open-loop arms, require identical sender retellings and randomize subsequent receipt: independently verified comprehension by an unreachable recipient; equally successful private archival deposit with no later retrieval; or matched nonmnemonic task completion. Receipt salience, praise, elapsed time and output effort are matched. Actual receiver competence and sender belief are measured separately; pilot any credible receipt implementation before the main study. another hypothesis of the same gap predicts a negative receiver-receipt effect on later sender source-detail availability exceeding margin delta_m, despite the usual positive retrieval benefit, coupled with reduced repeated forwarding by that sender and preserved or increased survival through new recipients. Crucially, the contrast survives elimination of reciprocal access and survives controls for expected future usefulness, perceived task completion, effort and social approval. It need not preserve the sender's adoption. An ordinary allocator account predicts no receipt-specific sender-memory difference during common open loop; ordinary retrieval predicts benefit from the matched production; ordinary offloading predicts an effect tied to recoverable storage, and general goal completion predicts the matched closure control suffices. Primary decisive result is the joint sender-memory/descendant-survival sign pattern, not the number of shares alone. If both receipt and archive conditions behave alike, or a fitted offloading-plus-goal-closure model predicts the joint trajectories within the negligible-extra-effect margin on held-out content, delete this distinct family. If memory merely remains unobserved in one test, use source-cued recovery and recognition probes before claiming loss; no behavioral result licenses molecular erasure. Hypothetical result
Would support the hypothesis
Confirmed understanding by an unreachable recipient may weaken the sender's memory — In online narrative chains, confirmed understanding by an unreachable recipient may reduce the sender's memory while later tellings survive. Reject a distinct mechanism if receipt and archive conditions behave alike or ordinary offloading and goal completion predict the joint pattern.
Other hypotheses predict
  • Exposure order may fragment private free time and block long cultural practices — During the common open-loop probe give everyone the same objectively uninterrupted interval and the same menu of optional short and long enactments; all can physically finish either. Independently randomize a neutral reorganization of their pre-existing tentative plans that preserves planned duration, obligations, rewards and content, versus equally effortful plan inspection. Model planned and performed activities separately. Given externally equal opportunities, another hypothesis of the same gap predicts the history contrast interacts with indivisible enactment duration and is attenuated selectively by repacking, with effect magnitude predicted out of sample by available-start measure A(d). Example, purely mathematical rather than empirical: subjective free gaps of 4+4 versus 1+7 minutes both total eight, but a five-minute indivisible act has A(5)=0 versus 2 minutes of acceptable start positions. Splitting the same act into independently meaningful shorter parts should relieve rejection without extra source exposure, but does not by itself establish ancestry or adoption. A scalar resource-depletion model predicts equal rejection at equal total free time after controlling task cost; a boundary-task/time-contraction model predicts effects from current salient boundary distance, so fit it explicitly. The claimed new family is removed if independently calibrated boundary perception, planning-cost and ordinary opportunity-choice models reproduce all history-by-duration and repacking contrasts, or if measured private reservations are absent. If real external time or feasible action opportunities differ, this does not answer the L3 and the trial must be redesigned.
  • Exposure history may bias sleep cueing to preserve cultural source lineages — After the matched history/reset procedure, randomly assign counterbalanced benign source-associated cues to verified non-REM sleep, the same cues during quiet wake, or uncued control, while keeping total waking content exposure, elapsed interval and later opportunity schedule equal. Within sleep, target the historically prioritized source or its competitor with the same cue count and intensity, randomizing cue identity independently of exposure order. another hypothesis of the same gap predicts a prespecified history-by-targeted-source-by-sleep interaction on arbitrary-source-detail retention and subsequent qualified survival, with onset after sleep and persistence through the common open-loop interval. The specific candidate predicts that cueing the lower-priority source attenuates or reverses the history contrast beyond a consequence margin even when a preregistered ordinary memory-strength/TMR model calibrated in single-source controls predicts only an additive cue gain. Estimate that rival's nonlinearities and ceiling effects rather than assume additivity. A pure allocator model predicts no such open-loop, sleep-specific history interaction. The accounting, disposal and prospective-packing hypotheses do not predict source-cue specificity during verified sleep when their own states are held constant. Family-removing result: ordinary source memory plus the already established TMR/preference model predicts later choices and descendants within the negligible-extra-effect margin, or apparent cue effects are explained by awakenings or unequal waking rehearsal. A cue-induced preference shift alone is explicitly insufficient for novelty.
  • Reassigning past effort to another cultural lineage may transfer its persistence advantage — In a preregistered benign practice experiment, require equal amounts of neutral preparatory work that cannot improve skill, give identical monetary pay and practice opportunities, and randomize which lineage that work is recorded as supporting. Assignment of credit is made separately from actual variant exposure and order. After verified reset and common open-loop delivery, cross unchanged versus transparently reassigned effort credit with a matched relabeling-only control and an equal-value general acknowledgement control; no extra payment, audience or material advantage is introduced. another hypothesis of the same gap predicts that the sign of private post-support persistence follows the credited lineage, including a transfer to a different fully learned lineage, while blinded source competence and source-detail fidelity remain stable within their margins. The critical contrast is change in D after lineage-specific reassignment minus change after equally salient general acknowledgement or label change. It should survive explicit confirmation that prior work cannot be recovered and future costs are equal. Pure allocator memory cannot move common-open-loop private persistence this way; offline memory predicts a trace-related effect; time packing predicts duration-specific feasibility rather than credited ancestry; disposal predicts a recipient-comprehension contingency. Remove the distinct family if ordinary accounting plus effort justification predicts the transfer and later hazards out of sample, if mere labels suffice, or if the manipulation changes expected future reward or competence. A standard sunk-cost main effect is only a successful control, not discovery.

These are hypothetical results. Selecting one shows what would follow from it; it does not confirm a hypothesis or change its assessment.

Comparing hypotheses

Compare the proposed mechanisms, the predictions that distinguish the hypotheses, and the observations that would count against each one.

01

Confirmed understanding by an unreachable recipient may weaken the sender's memory

Recipient contingent mnemonic disposal
Proposed mechanism

In online narrative chains, confirmed understanding by an unreachable recipient may reduce the sender's memory while later tellings survive.

Full text

HERETICAL CANDIDATE — Successful transmission can be a private memory-disposal event. Exposure history builds an unfinished, source-specific retention demand in the person. The person continues reproducing that unit until another person has demonstrably acquired its meaning; that event then actively reduces the sender's source-specific accessibility even when the recipient is permanently unreachable and cannot serve as an external memory. Thus a lineage can persist through serial disposal while its successive carriers abandon private retention. The stored variable is an unresolved mnemonic demand, not durable liking, a social obligation, ownership, or recommendation weights. Equal-dose ordering matters if encounters occur before versus after such completion events. A hidden allocator reset cannot erase the demand, but a validated recipient-comprehension receipt can. The strong additional claim is a recipient-specific negative update to the sender's trace despite matched retrieval, matched successful task completion, equal future usefulness, and zero expected reacquisition. This goes beyond generic cognitive offloading and goal completion only if those controls fail to reproduce it. A schematic candidate update is m_after=m_before+a*rho-k*c, where m is independently measured source-specific accessibility, rho is the sender's matched retrieval operation, c is randomized evidence of another person's comprehension of that source, a is ordinary retrieval benefit and k is the proposed disposal effect. The risky regime is k>a*rho for a completed handoff; it is not assumed from current evidence. Lower sender retention together with continued qualified descendants targets SPV_7, not a claim that lower memory improves belief adoption.

What distinguishes its prediction

Within reset/open-loop arms, require identical sender retellings and randomize subsequent receipt: independently verified comprehension by an unreachable recipient; equally successful private archival deposit with no later retrieval; or matched nonmnemonic task completion.

Full text

Receipt salience, praise, elapsed time and output effort are matched. Actual receiver competence and sender belief are measured separately; pilot any credible receipt implementation before the main study. IH_01 predicts a negative receiver-receipt effect on later sender source-detail availability exceeding margin delta_m, despite the usual positive retrieval benefit, coupled with reduced repeated forwarding by that sender and preserved or increased survival through new recipients. Crucially, the contrast survives elimination of reciprocal access and survives controls for expected future usefulness, perceived task completion, effort and social approval. It need not preserve the sender's adoption. An ordinary allocator account predicts no receipt-specific sender-memory difference during common open loop; ordinary retrieval predicts benefit from the matched production; ordinary offloading predicts an effect tied to recoverable storage, and general goal completion predicts the matched closure control suffices. Primary decisive result is the joint sender-memory/descendant-survival sign pattern, not the number of shares alone. If both receipt and archive conditions behave alike, or a fitted offloading-plus-goal-closure model predicts the joint trajectories within the negligible-extra-effect margin on held-out content, delete this distinct family. If memory merely remains unobserved in one test, use source-cued recovery and recognition probes before claiming loss; no behavioral result licenses molecular erasure.

What would weaken the hypothesis

Exposure order may fragment private free time and block long cultural practices predicts instead: During the common open-loop probe give everyone the same objectively uninterrupted interval and the same menu of optional short and long enactments; all can physically finish either.

Full text

Independently randomize a neutral reorganization of their pre-existing tentative plans that preserves planned duration, obligations, rewards and content, versus equally effortful plan inspection. Model planned and performed activities separately. Given externally equal opportunities, IH_02 predicts the history contrast interacts with indivisible enactment duration and is attenuated selectively by repacking, with effect magnitude predicted out of sample by available-start measure A(d). Example, purely mathematical rather than empirical: subjective free gaps of 4+4 versus 1+7 minutes both total eight, but a five-minute indivisible act has A(5)=0 versus 2 minutes of acceptable start positions. Splitting the same act into independently meaningful shorter parts should relieve rejection without extra source exposure, but does not by itself establish ancestry or adoption. A scalar resource-depletion model predicts equal rejection at equal total free time after controlling task cost; a boundary-task/time-contraction model predicts effects from current salient boundary distance, so fit it explicitly. The claimed new family is removed if independently calibrated boundary perception, planning-cost and ordinary opportunity-choice models reproduce all history-by-duration and repacking contrasts, or if measured private reservations are absent. If real external time or feasible action opportunities differ, this does not answer the L3 and the trial must be redesigned.

Exposure history may bias sleep cueing to preserve cultural source lineages predicts instead: After the matched history/reset procedure, randomly assign counterbalanced benign source-associated cues to verified non-REM sleep, the same cues during quiet wake, or uncued control, while keeping total waking content exposure, elapsed interval and later opportunity schedule equal. Within sleep, target the historically prioritized source or its competitor with the same cue count and intensity, randomizing cue identity independently of exposure order. IH_03 predicts a prespecified history-by-targeted-source-by-sleep interaction on arbitrary-source-detail retention and subsequent qualified survival, with onset after sleep and persistence through the common open-loop interval. The specific candidate predicts that cueing the lower-priority source attenuates or reverses the history contrast beyond a consequence margin even when a preregistered ordinary memory-strength/TMR model calibrated in single-source controls predicts only an additive cue gain. Estimate that rival's nonlinearities and ceiling effects rather than assume additivity. A pure allocator model predicts no such open-loop, sleep-specific history interaction. The accounting, disposal and prospective-packing hypotheses do not predict source-cue specificity during verified sleep when their own states are held constant. Family-removing result: ordinary source memory plus the already established TMR/preference model predicts later choices and descendants within the negligible-extra-effect margin, or apparent cue effects are explained by awakenings or unequal waking rehearsal. A cue-induced preference shift alone is explicitly insufficient for novelty.

Reassigning past effort to another cultural lineage may transfer its persistence advantage predicts instead: In a preregistered benign practice experiment, require equal amounts of neutral preparatory work that cannot improve skill, give identical monetary pay and practice opportunities, and randomize which lineage that work is recorded as supporting. Assignment of credit is made separately from actual variant exposure and order. After verified reset and common open-loop delivery, cross unchanged versus transparently reassigned effort credit with a matched relabeling-only control and an equal-value general acknowledgement control; no extra payment, audience or material advantage is introduced. IH_04 predicts that the sign of private post-support persistence follows the credited lineage, including a transfer to a different fully learned lineage, while blinded source competence and source-detail fidelity remain stable within their margins. The critical contrast is change in D after lineage-specific reassignment minus change after equally salient general acknowledgement or label change. It should survive explicit confirmation that prior work cannot be recovered and future costs are equal. Pure allocator memory cannot move common-open-loop private persistence this way; offline memory predicts a trace-related effect; time packing predicts duration-specific feasibility rather than credited ancestry; disposal predicts a recipient-comprehension contingency. Remove the distinct family if ordinary accounting plus effort justification predicts the transfer and later hazards out of sample, if mere labels suffice, or if the manipulation changes expected future reward or competence. A standard sunk-cost main effect is only a successful control, not discovery.

02

Exposure order may fragment private free time and block long cultural practices

Resource and energy
Proposed mechanism

People may retain exposure order as private time reservations, rejecting long cultural acts despite equal real opportunity after recommendation reset.

Full text

CROSS-DOMAIN TRANSFER — Exposure order deposits private time reservations that reject otherwise possible cultural reproductions. A recipient who encounters several narrative/practice variants forms tentative plans for when to enact them. The first accepted plans reserve indivisible blocks in subjective future time; later plans fit into the remaining gaps. Reversing encounter order changes the arrangement of those reservations without changing total exposure, total planned minutes or external reproduction availability. After allocator reset the person still treats these private reservations as binding, so a long familiar practice may be privately unavailable while equally known short variants continue. The carrier is the person's prospective allocation of a time budget, not a network, spatial barrier, geometric neural representation or sensor. No actual opportunity is removed by the experiment. The candidate extension is an order-deposited, duration-specific fragmentation state that predicts willingness beyond total busyness, ordinary schedule salience, competence, expected value and the logged allocator. SPV_6 is stabilized for practices that fit surviving subjective gaps, while abandonment of nonfitting practices rises. Repacking a private plan without adding free time should selectively rescue long cultural acts.

What distinguishes its prediction

During the common open-loop probe give everyone the same objectively uninterrupted interval and the same menu of optional short and long enactments; all can physically finish either.

Full text

Independently randomize a neutral reorganization of their pre-existing tentative plans that preserves planned duration, obligations, rewards and content, versus equally effortful plan inspection. Model planned and performed activities separately. Given externally equal opportunities, IH_02 predicts the history contrast interacts with indivisible enactment duration and is attenuated selectively by repacking, with effect magnitude predicted out of sample by available-start measure A(d). Example, purely mathematical rather than empirical: subjective free gaps of 4+4 versus 1+7 minutes both total eight, but a five-minute indivisible act has A(5)=0 versus 2 minutes of acceptable start positions. Splitting the same act into independently meaningful shorter parts should relieve rejection without extra source exposure, but does not by itself establish ancestry or adoption. A scalar resource-depletion model predicts equal rejection at equal total free time after controlling task cost; a boundary-task/time-contraction model predicts effects from current salient boundary distance, so fit it explicitly. The claimed new family is removed if independently calibrated boundary perception, planning-cost and ordinary opportunity-choice models reproduce all history-by-duration and repacking contrasts, or if measured private reservations are absent. If real external time or feasible action opportunities differ, this does not answer the L3 and the trial must be redesigned.

What would weaken the hypothesis

Confirmed understanding by an unreachable recipient may weaken the sender's memory predicts instead: Within reset/open-loop arms, require identical sender retellings and randomize subsequent receipt: independently verified comprehension by an unreachable recipient; equally successful private archival deposit with no later retrieval; or matched nonmnemonic task completion.

Full text

Receipt salience, praise, elapsed time and output effort are matched. Actual receiver competence and sender belief are measured separately; pilot any credible receipt implementation before the main study. IH_01 predicts a negative receiver-receipt effect on later sender source-detail availability exceeding margin delta_m, despite the usual positive retrieval benefit, coupled with reduced repeated forwarding by that sender and preserved or increased survival through new recipients. Crucially, the contrast survives elimination of reciprocal access and survives controls for expected future usefulness, perceived task completion, effort and social approval. It need not preserve the sender's adoption. An ordinary allocator account predicts no receipt-specific sender-memory difference during common open loop; ordinary retrieval predicts benefit from the matched production; ordinary offloading predicts an effect tied to recoverable storage, and general goal completion predicts the matched closure control suffices. Primary decisive result is the joint sender-memory/descendant-survival sign pattern, not the number of shares alone. If both receipt and archive conditions behave alike, or a fitted offloading-plus-goal-closure model predicts the joint trajectories within the negligible-extra-effect margin on held-out content, delete this distinct family. If memory merely remains unobserved in one test, use source-cued recovery and recognition probes before claiming loss; no behavioral result licenses molecular erasure.

Exposure history may bias sleep cueing to preserve cultural source lineages predicts instead: After the matched history/reset procedure, randomly assign counterbalanced benign source-associated cues to verified non-REM sleep, the same cues during quiet wake, or uncued control, while keeping total waking content exposure, elapsed interval and later opportunity schedule equal. Within sleep, target the historically prioritized source or its competitor with the same cue count and intensity, randomizing cue identity independently of exposure order. IH_03 predicts a prespecified history-by-targeted-source-by-sleep interaction on arbitrary-source-detail retention and subsequent qualified survival, with onset after sleep and persistence through the common open-loop interval. The specific candidate predicts that cueing the lower-priority source attenuates or reverses the history contrast beyond a consequence margin even when a preregistered ordinary memory-strength/TMR model calibrated in single-source controls predicts only an additive cue gain. Estimate that rival's nonlinearities and ceiling effects rather than assume additivity. A pure allocator model predicts no such open-loop, sleep-specific history interaction. The accounting, disposal and prospective-packing hypotheses do not predict source-cue specificity during verified sleep when their own states are held constant. Family-removing result: ordinary source memory plus the already established TMR/preference model predicts later choices and descendants within the negligible-extra-effect margin, or apparent cue effects are explained by awakenings or unequal waking rehearsal. A cue-induced preference shift alone is explicitly insufficient for novelty.

Reassigning past effort to another cultural lineage may transfer its persistence advantage predicts instead: In a preregistered benign practice experiment, require equal amounts of neutral preparatory work that cannot improve skill, give identical monetary pay and practice opportunities, and randomize which lineage that work is recorded as supporting. Assignment of credit is made separately from actual variant exposure and order. After verified reset and common open-loop delivery, cross unchanged versus transparently reassigned effort credit with a matched relabeling-only control and an equal-value general acknowledgement control; no extra payment, audience or material advantage is introduced. IH_04 predicts that the sign of private post-support persistence follows the credited lineage, including a transfer to a different fully learned lineage, while blinded source competence and source-detail fidelity remain stable within their margins. The critical contrast is change in D after lineage-specific reassignment minus change after equally salient general acknowledgement or label change. It should survive explicit confirmation that prior work cannot be recovered and future costs are equal. Pure allocator memory cannot move common-open-loop private persistence this way; offline memory predicts a trace-related effect; time packing predicts duration-specific feasibility rather than credited ancestry; disposal predicts a recipient-comprehension contingency. Remove the distinct family if ordinary accounting plus effort justification predicts the transfer and later hazards out of sample, if mere labels suffice, or if the manipulation changes expected future reward or competence. A standard sunk-cost main effect is only a successful control, not discovery.

03

Exposure history may bias sleep cueing to preserve cultural source lineages

Sleep dependent source priority
Proposed mechanism

After recommendation reset, exposure history may determine how source-linked sleep cues preserve details and cultural descendants; reject the distinct mechanism if ordinary memory and cueing models explain later transmission, or awakenings or unequal waking rehearsal explain the cue effects.

Full text

SCOUT 1 — Sleep neuroscience: cultural exposure history survives in source-selective offline replay priority. An order of encounters can assign unequal future relevance to equally practiced source episodes. During subsequent non-REM sleep, replay consolidates one source and its arbitrary diagnostic details preferentially; recommendation reset therefore leaves a delayed human-carried lineage advantage. The strong extension is that the history-induced priority determines how a fixed amount of source-specific sleep cueing changes later voluntary reproduction and qualified lineage survival. This is not a new name for sleep consolidation, a claim that equal waking recall equals equal memory, or a universal cueing benefit. The hidden state is a source-specific eligibility for offline stabilization, assessed with experimentally associated neutral cues and sleep physiology. The candidate state changes only over verified sleep episodes in the relevant version of the hypothesis, rather than through a donor's handoff receipt, a reconfigured time budget or an effort account. It targets SPV_7 by preserving diagnostic source information available for later descendants after allocator support stops.

What distinguishes its prediction

After the matched history/reset procedure, randomly assign counterbalanced benign source-associated cues to verified non-REM sleep, the same cues during quiet wake, or uncued control, while keeping total waking content exposure, elapsed interval and later opportunity schedule equal.

Full text

Within sleep, target the historically prioritized source or its competitor with the same cue count and intensity, randomizing cue identity independently of exposure order. IH_03 predicts a prespecified history-by-targeted-source-by-sleep interaction on arbitrary-source-detail retention and subsequent qualified survival, with onset after sleep and persistence through the common open-loop interval. The specific candidate predicts that cueing the lower-priority source attenuates or reverses the history contrast beyond a consequence margin even when a preregistered ordinary memory-strength/TMR model calibrated in single-source controls predicts only an additive cue gain. Estimate that rival's nonlinearities and ceiling effects rather than assume additivity. A pure allocator model predicts no such open-loop, sleep-specific history interaction. The accounting, disposal and prospective-packing hypotheses do not predict source-cue specificity during verified sleep when their own states are held constant. Family-removing result: ordinary source memory plus the already established TMR/preference model predicts later choices and descendants within the negligible-extra-effect margin, or apparent cue effects are explained by awakenings or unequal waking rehearsal. A cue-induced preference shift alone is explicitly insufficient for novelty.

What would weaken the hypothesis

Confirmed understanding by an unreachable recipient may weaken the sender's memory predicts instead: Within reset/open-loop arms, require identical sender retellings and randomize subsequent receipt: independently verified comprehension by an unreachable recipient; equally successful private archival deposit with no later retrieval; or matched nonmnemonic task completion.

Full text

Receipt salience, praise, elapsed time and output effort are matched. Actual receiver competence and sender belief are measured separately; pilot any credible receipt implementation before the main study. IH_01 predicts a negative receiver-receipt effect on later sender source-detail availability exceeding margin delta_m, despite the usual positive retrieval benefit, coupled with reduced repeated forwarding by that sender and preserved or increased survival through new recipients. Crucially, the contrast survives elimination of reciprocal access and survives controls for expected future usefulness, perceived task completion, effort and social approval. It need not preserve the sender's adoption. An ordinary allocator account predicts no receipt-specific sender-memory difference during common open loop; ordinary retrieval predicts benefit from the matched production; ordinary offloading predicts an effect tied to recoverable storage, and general goal completion predicts the matched closure control suffices. Primary decisive result is the joint sender-memory/descendant-survival sign pattern, not the number of shares alone. If both receipt and archive conditions behave alike, or a fitted offloading-plus-goal-closure model predicts the joint trajectories within the negligible-extra-effect margin on held-out content, delete this distinct family. If memory merely remains unobserved in one test, use source-cued recovery and recognition probes before claiming loss; no behavioral result licenses molecular erasure.

Exposure order may fragment private free time and block long cultural practices predicts instead: During the common open-loop probe give everyone the same objectively uninterrupted interval and the same menu of optional short and long enactments; all can physically finish either. Independently randomize a neutral reorganization of their pre-existing tentative plans that preserves planned duration, obligations, rewards and content, versus equally effortful plan inspection. Model planned and performed activities separately. Given externally equal opportunities, IH_02 predicts the history contrast interacts with indivisible enactment duration and is attenuated selectively by repacking, with effect magnitude predicted out of sample by available-start measure A(d). Example, purely mathematical rather than empirical: subjective free gaps of 4+4 versus 1+7 minutes both total eight, but a five-minute indivisible act has A(5)=0 versus 2 minutes of acceptable start positions. Splitting the same act into independently meaningful shorter parts should relieve rejection without extra source exposure, but does not by itself establish ancestry or adoption. A scalar resource-depletion model predicts equal rejection at equal total free time after controlling task cost; a boundary-task/time-contraction model predicts effects from current salient boundary distance, so fit it explicitly. The claimed new family is removed if independently calibrated boundary perception, planning-cost and ordinary opportunity-choice models reproduce all history-by-duration and repacking contrasts, or if measured private reservations are absent. If real external time or feasible action opportunities differ, this does not answer the L3 and the trial must be redesigned.

Reassigning past effort to another cultural lineage may transfer its persistence advantage predicts instead: In a preregistered benign practice experiment, require equal amounts of neutral preparatory work that cannot improve skill, give identical monetary pay and practice opportunities, and randomize which lineage that work is recorded as supporting. Assignment of credit is made separately from actual variant exposure and order. After verified reset and common open-loop delivery, cross unchanged versus transparently reassigned effort credit with a matched relabeling-only control and an equal-value general acknowledgement control; no extra payment, audience or material advantage is introduced. IH_04 predicts that the sign of private post-support persistence follows the credited lineage, including a transfer to a different fully learned lineage, while blinded source competence and source-detail fidelity remain stable within their margins. The critical contrast is change in D after lineage-specific reassignment minus change after equally salient general acknowledgement or label change. It should survive explicit confirmation that prior work cannot be recovered and future costs are equal. Pure allocator memory cannot move common-open-loop private persistence this way; offline memory predicts a trace-related effect; time packing predicts duration-specific feasibility rather than credited ancestry; disposal predicts a recipient-comprehension contingency. Remove the distinct family if ordinary accounting plus effort justification predicts the transfer and later hazards out of sample, if mere labels suffice, or if the manipulation changes expected future reward or competence. A standard sunk-cost main effect is only a successful control, not discovery.

04

Reassigning past effort to another cultural lineage may transfer its persistence advantage

Irrecoverable investment accounting
Proposed mechanism

In equally learned narrative-sharing practices after recommendation reset, crediting the same completed work to another lineage may move voluntary persistence without changing skill or memory fidelity; the distinct claim fails if ordinary accounting and effort justification predict the transfer

Full text

SCOUT 2 — Behavioral accounting: exposure history persists as an unrecovered effort balance attached to a cultural lineage. Equal total exposure and work need not create the same balance if early versus late effort is mentally attributed to different variants. After reset, a person keeps voluntarily enacting a less preferred but well-known practice because abandoning it would close a subjective account at a loss. The substrate is a remembered allocation of sunk personal effort, not stronger competence, normative ownership or increased future production efficiency. The candidate extension is transferable accounting: a credible, payoff-neutral reassignment of the same completed effort from the old lineage to a new one should move the persistence advantage to the newly credited lineage without extra exposure or improved competence. That transfer must exceed predictions of ordinary sunk-cost, effort-justification, reference-point and labeling models fitted to matched noncultural choices. SPV_6 is stabilized for the credited lineage; reallocating its effort balance should raise its abandonment and lower abandonment of the newly credited competitor. This is not new merely because sunk costs occur in culture.

What distinguishes its prediction

In a preregistered benign practice experiment, require equal amounts of neutral preparatory work that cannot improve skill, give identical monetary pay and practice opportunities, and randomize which lineage that work is recorded as supporting.

Full text

Assignment of credit is made separately from actual variant exposure and order. After verified reset and common open-loop delivery, cross unchanged versus transparently reassigned effort credit with a matched relabeling-only control and an equal-value general acknowledgement control; no extra payment, audience or material advantage is introduced. IH_04 predicts that the sign of private post-support persistence follows the credited lineage, including a transfer to a different fully learned lineage, while blinded source competence and source-detail fidelity remain stable within their margins. The critical contrast is change in D after lineage-specific reassignment minus change after equally salient general acknowledgement or label change. It should survive explicit confirmation that prior work cannot be recovered and future costs are equal. Pure allocator memory cannot move common-open-loop private persistence this way; offline memory predicts a trace-related effect; time packing predicts duration-specific feasibility rather than credited ancestry; disposal predicts a recipient-comprehension contingency. Remove the distinct family if ordinary accounting plus effort justification predicts the transfer and later hazards out of sample, if mere labels suffice, or if the manipulation changes expected future reward or competence. A standard sunk-cost main effect is only a successful control, not discovery.

What would weaken the hypothesis

Confirmed understanding by an unreachable recipient may weaken the sender's memory predicts instead: Within reset/open-loop arms, require identical sender retellings and randomize subsequent receipt: independently verified comprehension by an unreachable recipient; equally successful private archival deposit with no later retrieval; or matched nonmnemonic task completion.

Full text

Receipt salience, praise, elapsed time and output effort are matched. Actual receiver competence and sender belief are measured separately; pilot any credible receipt implementation before the main study. IH_01 predicts a negative receiver-receipt effect on later sender source-detail availability exceeding margin delta_m, despite the usual positive retrieval benefit, coupled with reduced repeated forwarding by that sender and preserved or increased survival through new recipients. Crucially, the contrast survives elimination of reciprocal access and survives controls for expected future usefulness, perceived task completion, effort and social approval. It need not preserve the sender's adoption. An ordinary allocator account predicts no receipt-specific sender-memory difference during common open loop; ordinary retrieval predicts benefit from the matched production; ordinary offloading predicts an effect tied to recoverable storage, and general goal completion predicts the matched closure control suffices. Primary decisive result is the joint sender-memory/descendant-survival sign pattern, not the number of shares alone. If both receipt and archive conditions behave alike, or a fitted offloading-plus-goal-closure model predicts the joint trajectories within the negligible-extra-effect margin on held-out content, delete this distinct family. If memory merely remains unobserved in one test, use source-cued recovery and recognition probes before claiming loss; no behavioral result licenses molecular erasure.

Exposure order may fragment private free time and block long cultural practices predicts instead: During the common open-loop probe give everyone the same objectively uninterrupted interval and the same menu of optional short and long enactments; all can physically finish either. Independently randomize a neutral reorganization of their pre-existing tentative plans that preserves planned duration, obligations, rewards and content, versus equally effortful plan inspection. Model planned and performed activities separately. Given externally equal opportunities, IH_02 predicts the history contrast interacts with indivisible enactment duration and is attenuated selectively by repacking, with effect magnitude predicted out of sample by available-start measure A(d). Example, purely mathematical rather than empirical: subjective free gaps of 4+4 versus 1+7 minutes both total eight, but a five-minute indivisible act has A(5)=0 versus 2 minutes of acceptable start positions. Splitting the same act into independently meaningful shorter parts should relieve rejection without extra source exposure, but does not by itself establish ancestry or adoption. A scalar resource-depletion model predicts equal rejection at equal total free time after controlling task cost; a boundary-task/time-contraction model predicts effects from current salient boundary distance, so fit it explicitly. The claimed new family is removed if independently calibrated boundary perception, planning-cost and ordinary opportunity-choice models reproduce all history-by-duration and repacking contrasts, or if measured private reservations are absent. If real external time or feasible action opportunities differ, this does not answer the L3 and the trial must be redesigned.

Exposure history may bias sleep cueing to preserve cultural source lineages predicts instead: After the matched history/reset procedure, randomly assign counterbalanced benign source-associated cues to verified non-REM sleep, the same cues during quiet wake, or uncued control, while keeping total waking content exposure, elapsed interval and later opportunity schedule equal. Within sleep, target the historically prioritized source or its competitor with the same cue count and intensity, randomizing cue identity independently of exposure order. IH_03 predicts a prespecified history-by-targeted-source-by-sleep interaction on arbitrary-source-detail retention and subsequent qualified survival, with onset after sleep and persistence through the common open-loop interval. The specific candidate predicts that cueing the lower-priority source attenuates or reverses the history contrast beyond a consequence margin even when a preregistered ordinary memory-strength/TMR model calibrated in single-source controls predicts only an additive cue gain. Estimate that rival's nonlinearities and ceiling effects rather than assume additivity. A pure allocator model predicts no such open-loop, sleep-specific history interaction. The accounting, disposal and prospective-packing hypotheses do not predict source-cue specificity during verified sleep when their own states are held constant. Family-removing result: ordinary source memory plus the already established TMR/preference model predicts later choices and descendants within the negligible-extra-effect margin, or apparent cue effects are explained by awakenings or unequal waking rehearsal. A cue-induced preference shift alone is explicitly insufficient for novelty.

No test is published for this question yet

The hypotheses above state the observations that could distinguish them. A proposed experiment for this question has not yet been published.

Every proposed test

What the literature settles, and what it does not

The sources read against this question, the assumption it rests on, and the verdict that follows.

After recommendations reset and sharing opportunities match, does earlier exposure still change what people pass on?

What this question is asking

The question concerns whether earlier encounters with cultural material leave changes in people that affect how long they continue passing it on. Cultural material means information or practices that can move between people, such as stories, images, or ways of doing something. The comparison is between different earlier exposure histories after the recommendation software has had its stored history cleared and people have equal later chances to reproduce or share the material. It asks whether any continuing difference in transmission reflects something retained in people, or whether the apparent history effect depended entirely on information retained by the software. Resetting the software and equalizing later opportunities are stipulated conditions, not findings established by the supplied sources.

What the terms mean
Cultural material
Information or practices that can be learned or passed between people, such as a story, an image, or a way of doing something. It is a broad class rather than a single kind of object.
Cultural transmission
The passage of cultural material between people through sharing, copying, teaching, or learning. It can occur with changes to the material rather than requiring an exact copy.
Cultural persistence
The continued presence or transmission of cultural material over time. Here the question concerns continued reproduction or sharing, which is distinct from one person simply remembering the material.
Exposure history
The earlier sequence of encounters with material, potentially including their number, timing, and order. The question asks whether differences in that past continue to affect later transmission.
Allocator or recommendation software
The system that selects which material is shown and to whom. Its stored information about past activity can influence later selections; this is the software-side candidate for carrying history.
Recommendation state
The information retained by a recommendation system that affects its later choices. It can include records or learned settings, and does not refer to a single standardized storage component.
Recommendation-state reset
Clearing the recommendation system’s relevant stored history so that earlier activity no longer influences later selections through that state. In this question it is a stipulated condition, not an intervention evaluated by the supplied sources.
Reproduction opportunities
Occasions on which a person can repeat, copy, share, or otherwise pass on cultural material. Holding them constant means equalizing those chances, not forcing people to make the same sharing choices.
History effect
A later difference associated with different earlier encounters or activities. Whether the earlier history actually causes continued cultural sharing is part of what remains unresolved here.
Human memory
Processes by which people retain information and later use or recover it. Memory includes multiple processes rather than one uniform store; the question does not specify which process might affect sharing.
Retention
Information remaining available over time. A learning study may assess it by later recall, which does not measure whether the information is passed to another person.
Recall and retrieval
Bringing previously learned information to mind. These terms concern access to information already encountered, not a new encounter with a recommendation.
Retrieval practice and self-testing
Learning activities that involve trying to recall information, often by answering questions without looking at the material. They add an active remembering step and therefore are not equivalent to exposure alone.
Repeated studying or re-study
Looking over learning material again. It is the comparison activity in the supplied summaries reporting benefits of retrieval practice.
Selective retrieval practice
Practicing recall of some learned material while leaving other material unpracticed. This distinguishes the learning history of different items within the narrower task described in S7.
Retrieval-induced forgetting
Reduced later recall of some related, unpracticed material following practice retrieving other material. The name describes a pattern of performance; it does not by itself prove that memories were actively suppressed.
Generalized anxiety disorder
A clinical condition involving persistent, excessive anxiety and worry. S7 uses this diagnosis to identify the group whose recall differed between neutral and threatening material.
Neutral and threatening material
Categories of learned information distinguished by whether it carries threat-related meaning in the task. They are task categories, not a claim that every person experiences each item identically.
Concept mapping
Representing ideas and their connections in a diagram as a learning activity. It appears in S6’s educational review, not as a test of continued cultural transmission.
Fruit flies
The small insects studied in S3, where wasp exposure changed egg-laying behavior. Their results provide evidence from an animal system rather than from human recommendation users.
Memory-gene functions
Activities of genes implicated in the biological processes supporting memory. Disrupting them in S3 affected persistence and later teaching, while leaving the immediate behavioral response intact.
Communication in a brain structure
Signals exchanged by nerve cells within a region of the brain. S3 reports that inhibiting this communication in a relevant fly brain structure disrupted the later persistence and teaching outcomes.
Teaching ability in the fly study
The capacity, as described in the supplied study, for previously exposed flies to transmit the behavioral response socially. This usage does not establish human-like intention or explain human cultural sharing.
Rodents
A group of mammals that includes mice and rats. S4 reviews social transmission in these animals rather than the human setting asked about here.
Cognition
Processes involved in acquiring, using, and remembering information. S2 discusses these processes in birds as background, rather than testing the proposed comparison in people.
Abstract and full text
An abstract is a short summary of a study or review; full text is the complete article. Several supplied sources offer only abstracts, and the screening record for S6 warns that its full-text label overstates the available material.
What turns on the answer
  • The history effect remains If software history has been fully cleared and later sharing opportunities are equal, a remaining difference would not be explained by that cleared software history alone. It would be consistent with something retained in people affecting continued transmission, although the outcome alone would not identify which change in people carries the effect.
  • The history effect disappears If a previously observed difference disappears under the stated conditions, persistence would no longer show a detectable dependence on earlier exposure in that setting. This would be consistent with dependence on the earlier recommendation process, but disappearance alone would not prove that people retained nothing from their encounters.
  • The history effect is reduced but remains A smaller continuing difference would be compatible with earlier exposure influencing persistence through both people and the recommendation process. The remaining component would mean that clearing software history does not remove the whole pattern, while the reduction would limit a claim that people alone account for it.
Why it matters

Earlier encounters could change what people remember, and what they remember could in turn affect what they later pass on. Separately, software that stores earlier activity could keep showing particular material, creating more occasions for its continued transmission. Those two routes can produce similar visible patterns of persistence while locating the influence of the past in different places. Attributing a software-dependent pattern to lasting changes in people would overstate how well the material can survive without that software history. Attributing a person-dependent pattern entirely to software would miss a possible route by which transmission continues after recommendations change.

Could not be determined

The supplied evidence is too indirect to determine the literature status of the complete question. S1 identifies memory as a cultural process; S5 and S8 report benefits of retrieval practice for human retention; and S7 reports a more specific dependence of recall on retrieval history and material type. S3 establishes a memory-related requirement for persistence and teaching in fruit flies, while S2 and S4 supply animal-cognition background and S6 supplies academic-learning context. None tests continued human cultural transmission after recommendation history is cleared with equal later reproduction opportunities. The inference from these sources is that remembering and social transmission have relevant neighboring literatures, not that the proposed person-versus-allocator distinction has been settled or that its answer is absent from all literature. The sources marked as partly answering the question, S7 and S8, address only the narrower remembering component.S1S5S8S7S3S2S4S6

What the literature establishes
  • A review of cultural dynamics identifies processes within and between people, including production, interpretation, and memory of cultural information. The supplied abstract names human memory as part of cultural transmission; it does not report the controlled persistence comparison in the question.S1
  • The supplied abstract on brains and cognition discusses birds, tool manufacture, and transmission of tool design. It provides background on cognition and cultural transmission in another setting, without reporting the relevant comparison in people.S2
  • In fruit flies, the quoted study reports that disrupting specified memory-gene functions or blocking communication in a relevant brain structure left the immediate change in egg-laying behavior in the presence of a wasp intact. Those interventions abolished the behavior’s persistence after the wasp was removed and abolished subsequent teaching ability. This supports a requirement for functioning memory-related biological processes in that animal task.S3
  • The supplied rodent review describes models of social transmission of information. Its supplied abstract does not report a finding about exposure history and continued human transmission under recommendation resets.S4
  • A review reports that retrieval practice—actively trying to bring learned information to mind—can improve long-term retention compared with repeatedly studying it. This finding concerns later remembering rather than later cultural transmission.S5
  • The supplied material from a review of young adults’ academic learning describes self-testing, retrieval practice, and concept mapping as effective but underused. Although its metadata labels it full text, the supplied screening record says the available material consists of an abstract and page material, limiting what can be established from it.S6
  • In a laboratory study of selective retrieval practice, the group with generalized anxiety disorder showed retrieval-induced forgetting for neutral material but did not show it for threatening material. The supplied screening record describes this as a narrower case in which the accessibility of learned information depends on learning and retrieval history.S7
  • The supplied educational abstract states that testing or retrieval practice improves long-term retention more than studying the material again. It supports an effect of learning activity on human retention, without establishing continued transmission between people.S8
What it does not settle
  • Whether earlier exposure changes continued cultural transmission in people after recommendation state is reset and later reproduction opportunities are held equal is not established by any supplied source. None reports the full comparison.S1S2S3S4S5S6S7S8
  • Whether an observed persistence pattern is carried only by the allocator, only by changes in people, or by both remains unresolved. The sources on human remembering do not include the recommendation process whose stored history is central to this distinction.S1S5S6S7S8
  • An effect of actively recalling information does not establish the effect of exposure history alone. The supplied human learning sources do not settle whether different earlier encounters, apart from different retrieval activities, change later cultural sharing.S5S6S7S8
  • The connection between retaining information and repeatedly transmitting it is not established under the question’s conditions. Remembering material and passing it on are different measured outcomes.S5S6S7S8
  • The direction, size, duration, and dependence on the amount or sequence of exposure of any remaining human persistence effect are not established. Its applicability across populations, kinds of cultural material, and recommendation systems is also unresolved.S1S5S6S7S8
  • The fly finding does not establish that the same biological requirement explains human cultural persistence. The bird and rodent background likewise does not answer the comparison between human memory and stored recommendation history.S2S3S4
  • The supplied account of the selective-retrieval study notes explanations of retrieval-induced forgetting that do not require active suppression of memories. Even that narrower human result does not by itself establish a unique underlying memory mechanism.S7
  • The supplied set is insufficient to determine whether the complete question has been answered elsewhere in the literature. Several entries are abstracts, and the most relevant human findings concern learning and recall rather than the specified transmission comparison.S1S5S6S7S8
Where the sources disagree
  • No supplied source directly contradicts another on the stated human persistence question. There is a meaningful contrast in the narrower memory results: S5 and S8 describe improved retention following retrieval practice, while S7 concerns poorer later recall of related material and reports that this forgetting occurred for neutral but not threatening material in the anxiety group. These findings concern different material and outcomes, so they do not constitute opposing answers to the cultural-persistence question.S5S7S8
Sources read · 8

3 literature searches, 4 full texts, 6 abstract-only; 10 source(s) assessed against this question using the available text. A bounded search is not evidence of absence.

S1BackgroundAbstract only

Cultural dynamics. · Current opinion in psychology · 2016

“This article focuses on the micro-level mechanisms, which consist of the intrapersonal and interpersonal processes of production, grounding, interpretation, and memory of cultural information.”

Does not settle: The abstract identifies human memory as a cultural-transmission process but does not test exposure-history effects on persistence, reset recommendation state, or hold subsequent reproduction opportunities constant. It therefore does not distinguish memory in people from memory in an allocator.

S2BackgroundAbstract only

Evolution of brains, cognition, and consciousness. · Brain research bulletin · 2008

“tool manufacture and cultural transmission of tool design, theory of mind, and Piagetian object permanence to a high level.”

Does not settle: The abstract discusses avian cognition and proposed neural bases of consciousness. It does not test cultural persistence following exposure histories in people, reset recommendation state, or hold subsequent reproduction opportunities constant; it cannot distinguish memory in people from memory in an allocator.

S3Background

Social communication of predator-induced changes in Drosophila behavior and germ line physiology. · eLife · 2015

“Loss of memory gene functions, such as Adf1 , amn , dnc , dFmr1, rut , and Orb2 , or inhibition of MB synaptic transmission had no effect on the ability to change oviposition behavior in the presence of wasp, however, in each of these cases, persistence of this behavior after wasp removal, and subsequent teaching ability, was abolished.”

Does not settle: The text supports a neural memory requirement for persistent predator-induced behavior and teaching in Drosophila. It does not establish exposure-history effects on cultural persistence in people, distinguish human memory from recommendation-allocator memory, or test recommendation-state resets with subsequent reproduction opportunities held constant.

S4BackgroundAbstract only

Insights from social transmission of information in rodents. · Genes, brain, and behavior · 2019

“Here, we first review different models for social transmission of information in rodents.”

Does not settle: The supplied abstract reviews social learning in rodents. It does not establish whether cultural persistence in people retains exposure history after recommendation state is reset, or distinguish memory in people from allocator memory when subsequent reproduction opportunities are held constant.

S5BackgroundAbstract only

The critical role of retrieval practice in long-term retention. · Trends in cognitive sciences · 2011

“We review research that contradicts this traditional view by demonstrating that retrieval practice is actually a powerful mnemonic enhancer, often producing large gains in long-term retention relative to repeated studying.”

Does not settle: The abstract provides context on how retrieval practice affects long-term retention. It does not establish cultural persistence or retransmission, compare memory in people with memory in a recommendation allocator, or test whether exposure-history effects persist after recommendation state is reset and subsequent reproduction opportunities are held constant.

S6Background

Active recall strategies associated with academic achievement in young adults: A systematic review. · Journal of affective disorders · 2024

“Self-testing, retrieval practice, and concept mapping were also effective but under-utilized.”

Does not settle: The supplied text provides an abstract and page material, despite the full_text metadata label. It reviews academic performance and self-efficacy in higher education students; it does not test cultural persistence or retransmission after recommendation state is reset with subsequent reproduction opportunities held constant. It therefore does not distinguish exposure-history effects retained in people from allocator memory, or establish their duration or exposure dose under those controls.

S7Partly answers it

Impaired Retrieval Inhibition of Threat Material in Generalized Anxiety Disorder. · Clinical psychological science : a journal of the Association for Psychological Science · 2016

“Despite showing intact RIF of neutral targets, the GAD group failed to exhibit RIF of threat targets.”

Does not settle: The supplied text reports human recall after selective retrieval practice, supporting history-dependent accessibility of learned material under this narrower laboratory task. It does not test cultural persistence or retransmission, recommendation-state resets, or equal subsequent reproduction opportunities, so it cannot distinguish person-level exposure memory from allocator memory in the specified setting. Retrieval practice is not a direct test of exposure history alone; the text also notes non-inhibitory accounts of retrieval-induced forgetting.

S8Partly answers itAbstract only

Using Testing as a Learning Tool. · American journal of pharmaceutical education · 2019

“Testing or retrieval practice is superior to re-study for promoting long-term retention.”

Does not settle: The abstract supports human retention depending on learning history, but does not test cultural retransmission, recommendation-state resets, allocator memory, or equal subsequent reproduction opportunities. It does not establish whether cultural persistence retains exposure-history effects under those controls.

Every open question