Correction order may bias later rule meanings through beliefs about why examples were chosen
In invented narrative worlds, correction order may carry inferred example-selection policies into new rule judgments despite equal fact recall. Reject an extra-history mechanism if ordinary sampling-based inference fully predicts the errors; a verified policy reset should make the excess effect negligible.
Stage of verification
- Hypothesis published2026-10-05
- Indirect evidenceAssessed at 4 of 10
- Direct testAwaited
Target map
Every target of every published hypothesis, each with the actions a hypothesis can propose on it. The targets and the actions of this hypothesis are drawn solid.
Rhythm or programme
Scope inference
The process of inferring whether an example supports a general rule or applies only to an exception
Where this hypothesis actsDuring source-grounded correction and human reconstruction of cultural rules
Hypotheses on this target 3
Inhibition2
Activation
Function preservation
Feedback restoration1
Rhythm restoration
Direct measurement
What is proposed
Feedback restoration
Align inferred scope with the source rule by correcting the inferred sampling policy
With whatChange of environment or regimen
HowMake example-selection policies observable and explicitly reset the current policy, retaining why the example was selected and the contextual relation defining its scope
Possible result
Expected reduction in excess correction-order effects on scope-transfer errors despite equivalent literal recall
From the recordan experimentally verified policy reset reduces the excess order effect below epsilon.
All targets of the lab
Every target read from the published hypotheses, each kind around its pictogram. A larger mark means more hypotheses act on that target. Point at a mark and the actions proposed on it branch out of it.
Solid and named: the targets of this hypothesis
Explore in depth
The logic
The train of thought that ends in this hypothesis. Each stage is the reason the next exists. The master question narrows to a goal, the goal to an unknown nobody has closed, the unknown to the hypothesis proposed here. Every step below says what it rests on and what carries it.
People can remember a correction exactly and still learn the wrong lesson about when it applies. The unexpected move is that receiving a correction after trying to retell something may make the corrected example seem like a specially selected exception, whereas receiving it first may make it seem representative. This is a proposal generated by the pipeline, not a measured result: it predicts that the better order depends on whether the source actually describes a general rule or a rare exception.
- Correction before reconstruction may make the example seem ordinarily selected; correction after reconstruction may make it seem selected to address an error.
- The learner's belief about why the example was chosen is proposed to change whether its true content is treated as a general rule or a limited exception.
- That interpretation is predicted to change private rule use, meaning the learner's own application of the rule, even when the learner remembers the corrective fact equally well.
- The learner is proposed to carry the inferred selection rule into later material, adding an effect of earlier order beyond what the current examples alone would explain.
- The predicted advantage changes direction between sources describing general rules and sources describing genuine exceptions; a successful reset of selection beliefs is predicted to remove the extra order effect.
- After correction stops, later recipients are proposed to inherit the interpretation through statements specifying when the rule applies and through choices that earlier learners actually transmit.
A photograph of a flooded street says something different about a town if the street was picked at random than if someone deliberately photographed the worst flooding. The photograph can be completely accurate in both cases.
Where the picture breaks: The picture illustrates why selection changes what a true example implies. It does not explain why correction order would change someone's belief about that selection, why the belief would carry into a new situation despite a reset, or how it would survive being passed to another person.
- Master questionstep 01 of 04
Cultural information changes as people pass it on, and the goal is to find new, testable explanations for its transmission, alteration, competition and persistence. The requested research agenda must distinguish existing explanations from new proposals and separate how widely material spreads, how accurately it is copied, how its meaning changes, whether people adopt it and whether it lasts.
Rests on: The goal itself defines cultural information broadly enough to include online images, narratives and practices. It explicitly requires competing explanations, experiments that separate them, measurable outcomes and results that would disprove each proposed mechanism.
Stated in the chain - Goal pillarstep 02 of 04
Practical experiments and progressively stronger checks are the chosen route to identifying what causes cultural information to change or persist.
Rests on: The master question calls for an affordable initial experiment followed by the stronger validation needed for a general claim, with deliberate changes and comparison conditions that distinguish competing explanations.
Stated in the chain - Gap questionstep 03 of 04
The focus narrows to whether correcting a person's account against its source before or after that person reconstructs it, meaning attempts to rebuild or retell it, preserves meaning differently. It also asks what information must survive for correction to work and whether any benefit continues after corrections stop, beyond predictions that treat each transmission process as unchanged by earlier events.
Rests on: The preceding pillar supplies the demand for experiments that test cause by deliberately changing a condition and successive validation, but it contains only that broad research direction.
LeapThe supplied transition does not explain why correction order is the particular unresolved mechanism selected from the broad experimental agenda, or specify the unchanged-process prediction it would exceed. This is a missing basis for the narrowing of scope, not evidence against studying correction order.
- Hypothesisstep 04 of 04
A true corrective example may be treated either as representative of a rule or as an exception selected because of a preceding mistake. The proposed mechanism is scope inference, meaning a judgment about which cases a rule covers: people reuse their belief about why an example was selected when interpreting later material. Correction before reconstruction is predicted to help when the source intends a general rule, but the direction is predicted to reverse when the source genuinely describes a rare exception. The necessary retained information concerns why the example was chosen and how its context limits the rule, rather than only who supplied it.S1S3
Rests on: The gap question supplies the targets of order, retained information and persistence after correction stops. The 2018 review in Wiley Interdisciplinary Reviews: Cognitive Science reports that changing beliefs about how examples were generated changes some judgments about how broadly a property applies; it does not test correction order, later reuse of those beliefs or transmission to new recipients. The supplied abstract of a 2019 Journal of Experimental Psychology: General paper reports broader application after adding dissimilar examples lacking a property and a model that can reproduce that result using assumptions about helpful example selection; it does not establish that participants inferred such a selection rule or that correction order changes it. These sources support a component the proposal borrows, while the order-dependent reuse and subsequent transmission remain proposed extensions.
Supported by literature
What is carried, and what is not. Two of the four screened sources speak to one shared component: how assumptions about example selection affect judgments about how broadly information applies; their specific limits are stated in the hypothesis step. None establishes the proposed sequence from correction order through reused selection beliefs and changed private rule use to persistence among later recipients after correction stops; the other two sources supply broader cultural-learning context.
Where the reasoning is carried by something unstated · 1
- Gap question. The supplied transition does not explain why correction order is the particular unresolved mechanism selected from the broad experimental agenda, or specify the unchanged-process prediction it would exceed. This is a missing basis for the narrowing of scope, not evidence against studying correction order. Establish the missing link before relying on this step.
How a result here could mislead · 3
- A difference between randomly chosen and error-selected examples could be mistaken for a new effect of correction history, even though the proposal treats ordinary sensitivity to example selection as an existing explanation. A remaining order difference after an announced reset could also reflect failure to understand the current selection rule rather than the proposed carryover, meaning reuse of a previous belief about selection. What closes it: The design requires separate items presented once to establish the ordinary response to selection, followed by fresh source material with the current selection rule explicitly reset. The model describing the ordinary response to example selection and the largest remaining difference between correction orders considered negligible must be fixed before results are examined. Comprehension and actual updating of the selection belief must be verified separately: an instruction announcing a reset does not establish that the reset worked.
- Worse outgoing accounts could reflect leaving out privately understood information, as the communication-plan rival predicts, rather than learning the wrong scope of a rule. Conversely, a private rule-use difference could arise because one group remembered the corrective fact less accurately, without any difference in beliefs about example selection. What closes it: Private judgments about where the rule applies, beliefs about why the example was selected, exact factual recall and omissions from outgoing accounts must be measured separately. The proposed comparison holds the realized example, its truth, the number of examples and available time equal. Equal recall requires a prespecified acceptable difference rather than merely a failure to detect a difference; the supplied design does not give that margin. A change confined to outgoing omissions with intact private scope favors the communication-plan rival rather than this mechanism. The attention-allocation rival concerns how effort is divided between jointly useful repairs, and the sleep rival concerns which learned episode is stabilized during sleep; the proposed comparison distinguishes these accounts only when those alternative causes are held fixed or separately manipulated.
- Meaning preserved in a chain of new recipients could be credited to an enduring selection belief even if later recipients keep receiving corrective information, or if only the wording survives. The reverse error would be to treat failure to recover an omitted rule as disproof, although the hypothesis does not predict recovery without an information path. What closes it: The later transmission phase must document that source correction has stopped and record what information about when the rule applies each recipient actually receives. Assessment must distinguish repeated wording from correct application of the rule to new cases. A persistence claim requires a traceable path through transmitted statements or choices; information absent from that path cannot silently be counted as available to later recipients.
What would make this wrong. The proposed extra mechanism would fail if, with example-selection comprehension and equal factual recall established, errors in applying rules from new sources were fully explained by the ordinary response to current selection and showed no additional dependence on earlier correction order. Its specific predictions would also fail if the order difference did not reverse between the specified general-rule and genuine-exception settings, or if an independently verified reset of the inferred selection rule left the extra order effect above the prespecified negligible bound. An effect confined to outgoing omissions while private scope remained intact would favor the communication-plan rival; an effect confined to the competing sleep-based process after selection comprehension was verified would favor that rival. Failure of later recipients to preserve scope despite a documented information path would defeat the persistence extension, without by itself disproving an effect in the original learner.
What it would change. If the proposed effect held beyond the ordinary influence of example selection, an account of cultural meaning would need to track people's beliefs about why information was supplied, alongside what they remember and repeat. Correction would have no universally best position: its effect would depend on whether those beliefs fit the source's intended general rule or exception, and later persistence would depend on what the resulting messages carry forward. Success in the artificial online task would still leave unestablished whether the mechanism operates in narratives without numerical rules, cultural practices, another language, or situations where people infer selection without an experimenter's explanation. Persistence through new recipients would require its own transmission evidence rather than following automatically from the first learner's private judgments.
Sources read · 4
Inductive reasoning 2.0. · Wiley interdisciplinary reviews. Cognitive science · 2018
“Notably, when steps are taken to undermine helpful sampling assumptions (i.e. by convincing reasoners that the premises were generated randomly) property reinforcement and negative evidence effects disappear.”
Does not settle: The review supports sampling assumptions as a determinant of property generalization, including effects of negative evidence. This excerpt does not test correction-before-reconstruction versus reconstruction-before-correction, error-selected exception interpretations, carryover of inferred sampling policies across later propositions, equal literal recall or source memory, functional application, reversal for genuine exceptions, or persistence through cultural descendants after correction withdrawal. It also does not establish that Bayesian inference and connectionist explanations are causally distinct.
Negative evidence and inductive reasoning in generalization of associative learning. · Journal of experimental psychology. General · 2019
“We show that this key qualitative result can be simulated by a Bayesian model that incorporates helpful sampling assumptions.”
Does not settle: The abstract reports increased generalization after adding dissimilar negative evidence and a Bayesian simulation incorporating helpful sampling assumptions. It does not establish that participants inferred an example-selection policy, or test reconstruction–correction order, error-selected exceptions versus ordinary samples, equal literal recall and remembered origin, reversal for genuinely rare exceptions, later cultural-rule scope judgments, or persistence through descendants after correction withdrawal.
Attachment, culture, and gene-culture co-evolution: expanding the evolutionary toolbox of attachment theory. · Attachment & human development · 2021
“I also review research pointing to a facilitating role of secure attachment relationships for social learning from caregivers among humans.”
Does not settle: The abstract discusses attachment as a context for cultural transmission and attachment-related working models as biases in religious transmission. It does not establish correction-order effects, inference about why corrective examples were selected, exceptional versus general rule scope, likelihood reweighting, equal literal recall with different functional application, or persistence of scope interpretations after correction withdrawal across descendants.
The natural selection of fidelity in social learning. · Communicative & integrative biology · 2010
“Social learning mechanisms are unlikely to be faithful enough to explain cultural stability because they are generally selected not for high fidelity but for generalization and adjustment to the individual's needs, capacities and situation.”
Does not settle: The abstract offers context on social learning fidelity, generalization and cultural stability. It does not test correction order, beliefs about example selection, exceptional versus general rule scope, later reuse of sampling assumptions, equal literal recall or source memory, private application, or persistence through descendants after correction withdrawal.
The gap this hypothesis explains
Which correction and retelling order preserves meaning beyond expected losses, depends on retained information, and outlasts correction?
Original wording · exactly as the pipeline generated it
Which ordering of source-grounded correction and human reconstruction preserves cultural meaning beyond fixed-channel predictions, what must be retained before correction works, and does the benefit survive correction withdrawal?
What this question is asking
The question concerns how the meaning of a story, message, or cultural practice survives when people retell it and receive corrections checked against an original or other specified reference. It asks whether correction before retelling works differently from correction after retelling, and whether either order preserves meaning better than a model that assumes each transmission step changes information in a fixed way would predict. It also asks which parts of the earlier material must remain available in memory for correction to help, and whether the advantage continues once corrections stop. The question appears to assume that such an extra advantage and a necessary memory condition exist; the supplied sources do not establish either assumption.
- Source-grounded correction
- A correction checked against an original or another explicitly specified reference. In this question it supplies information that could restore something changed or lost in a retelling; the input does not identify the reference or establish its authority.
- Human reconstruction or retelling
- A person's rebuilding of a message, story, or practice from what is remembered and understood, potentially changing it in the process. The question treats this as a separate activity whose position before or after correction may matter.
- Cultural transmission
- The passing of information, stories, or practices between people or across successive versions. It can involve copying and alteration, so successful transmission does not automatically mean that meaning remains unchanged.
- Cultural meaning and its preservation
- What a transmitted message or practice signifies, including the relationships and interpretations that make it understandable. Preservation means retaining the aspects counted as relevant, which is a matter of degree; the supplied question does not specify those aspects or a measure for them.
- Fixed-channel model and predictions
- An account that represents each transmission or correction step as changing information according to a fixed rule or pattern. Its predictions provide the comparison for the proposed extra benefit; fixed steps can themselves produce different outcomes when their order is reversed, and the input supplies no specific model.
- Retained information or memory prerequisite
- Information still available in memory before correction occurs. Calling it a prerequisite means that correction cannot work without it, which is a stronger claim than finding that people who remember more also change their beliefs more.
- Correction withdrawal and persistence
- Withdrawal means that an ongoing supply of corrections stops; persistence means that a previously observed benefit remains afterward. A later test does not by itself establish the effects of stopping ongoing correction.
- Misinformation correction, retraction, and correction resistance
- Misinformation is information treated as false or inaccurate, a correction supplies a replacement or challenge, and a retraction withdraws an earlier claim. Correction resistance means that the mistaken belief or its influence remains despite that intervention; S9 discusses this problem in its supplied abstract.
- Narrative and non-narrative correction
- A narrative correction presents information as a story, whereas a non-narrative correction uses another form of explanation or presentation. These are formats for delivering correction, not the same activity as asking a person to reconstruct material from memory.
- Myth-first and fact-first formats
- Myth-first presents a false claim before the correct information; fact-first presents the correct information first. S5 compares these internal presentation orders, not correction before versus after retelling.
- Baseline knowledge
- What participants already know before the activity being studied. The supplied S5 account describes a course with high starting knowledge, which bounds the setting of its reported format comparison.
- Discounting evidence and discounted statements
- Discounting evidence is information offered as a reason to believe an earlier statement less; a discounted statement is one that has received that challenge. In S7, the supplied account says that this evidence was invented, so its effects do not establish the effects of correction checked against a real reference.
- Recollection and gist recognition
- Recollection involves remembering previously encountered material, while gist recognition involves identifying its broad meaning without necessarily recovering its exact wording. S7 links these aspects of memory to belief change, but the supplied findings do not establish either as a necessary condition.
- Dysphoric group or depressed mood
- Dysphoric describes an unpleasant or depressed mood; S7 uses this label for one student group. It describes the group comparison reported in the source and should not be read here as establishing a clinical diagnosis or a general finding for everyone with depression.
- Memory-component estimates and replication
- Memory-component estimates are numerical quantities used in a study to describe proposed contributions of different memory processes. Replication means obtaining an earlier result again in another study or experiment; conceptual support from a related measure does not mean that the original numerical differences were reproduced.
- Exploratory analysis
- An analysis used to investigate a pattern rather than simply reporting the study's central planned comparison. The supplied S4 quotation explicitly labels the analysis against a narrative advantage exploratory, a limitation retained in the account of its finding.
- Abstract and review
- An abstract is a short summary of a publication and does not provide all of its methods or results. A review discusses existing work rather than necessarily reporting a new experiment; S9 is represented here only by the abstract of such a review.
An ordering of source-grounded correction and human reconstruction preserves cultural meaning beyond fixed-channel predictions, and some information must be retained before correction works.
Corrections checked against a reference supply material that can be used when a person remembers and retells a story or practice. The wording appears to assume that arranging these activities in a particular order can preserve meaning better than an account in which each step changes information in a fixed way, and that this extra benefit requires some earlier material to remain in memory. If those assumptions held, the question would concern the order and memory conditions responsible for an existing effect; otherwise, whether that effect exists is also unresolved.
The supplied search results do not establish this combined premise. S5 compares the order of false claims and correct facts within corrections, rather than correction before versus after human retelling, and reports no difference between those formats in its described course setting. S7 connects belief change with remembering corrective evidence, but its supplied account does not establish that such remembering is necessary, and the evidence presented to participants was invented rather than grounded in an independently checked source. S4 concerns narrative versus non-narrative corrections, and S9 supplies a review abstract about correction resistance. None establishes preservation beyond fixed-channel predictions or identifies a necessary retained element. This is an absence of support in the read material, not a demonstration that the premise is false; the question about persistence after withdrawal remains a question rather than an asserted premise.S4S5S7S9
The same question asked without the part nothing read establishes:
- Does placing correction checked against a reference before rather than after human retelling change how well cultural meaning survives, and does any difference exceed predictions based on fixed changes at each step?
- Does the information people retain before correction affect how well they preserve cultural meaning when retelling it?
- Does any improvement in preservation of cultural meaning continue after corrections checked against a reference stop?
- Correction before retelling has a lasting extra benefit If the required remembered material is available, correction could alter the information used to construct the next retelling, so later versions would inherit that change. If the advantage exceeds the specified fixed-channel predictions and remains after correction stops, preservation would extend beyond the period of external support, with order and retained information contributing to the explanation.
- Correction after retelling has a lasting extra benefit A retelling would first produce a version that can then be corrected against the reference before further transmission. If this sequence preserves meaning better than the reverse order and beyond the specified fixed-channel predictions, with the advantage surviving withdrawal, correcting the reconstructed version would have a different consequence from correcting the material used to construct it.
- An extra benefit depends on continuing correction An ordering advantage could arise while corrections repeatedly restore information that people would otherwise lose or change. If the advantage disappears after corrections stop, maintained access to the reference would account for ongoing support, and the observed preservation would not establish a lasting change in transmission.
- No benefit beyond fixed-channel predictions An ordering difference could still occur because two fixed information-changing steps need not produce the same result when reversed. If the specified model explains preservation, the result would not establish the additional memory-dependent benefit presupposed by the question; the read sources also leave open whether any ordering difference exists at all.
When people pass material on, later versions can carry different meanings from earlier versions. Under the mechanism contemplated by the question, a correction supplies information from a reference, and a person's retelling determines which corrected and remembered elements enter the next version. If the order changes what survives, judging a correction only by an immediate change in belief could miss its consequences for later transmission. If any advantage disappears when correction stops, treating it as lasting preservation would mistake continued external support for persistence of the meaning itself. Conversely, if ordinary fixed changes at each step explain the result, attributing the advantage to an additional memory-dependent process would overstate what has been learned.
The mechanism it proposes
The engine's own statement of the hypothesis, in full.
SCOUT 2 — From statistical evidence evaluation and ascertainment: people preserve the literal corrective fact but learn the wrong scope of the cultural rule because correction is sampled conditional on their preceding reconstruction. R-C makes a supplied corrective example look like an error-selected exception; C-R can make the same example look like an ordinary draw from the source's rule. Reusing that inferred sampling policy across later propositions yields history dependence even when the current sentence and its remembered origin are identical. The retained prerequisite is the rule describing why this particular source-grounded example was selected and the contextual relation that defines its scope. It is not sufficient to remember who said it. The mechanism is inferential likelihood reweighting, not changing a network, information-routing barrier, proofreader or transmitter's agenda. It predicts that C-R is superior when the source intends a general relation but R-C induces an exceptional-case interpretation; the direction reverses when the original source genuinely describes a rare exception. Thus there is no universal best order. SPV_4 is stabilized by appropriate scope inference, with private functional application potentially changing despite equal literal recall. After correction withdrawal, a scope interpretation can continue through descendants' explicit conditional statements and choices; the hypothesis does not allow recovery of an omitted arbitrary rule without a documented information path.
Testing and possible results
The prediction that would tell it apart
A hypothesis that predicts what its rivals predict is not worth running an experiment over. This is the observation on which this one differs.
Use invented narrative worlds with a general rule and explicitly indexed exceptions; correction supplies a true, partially diagnostic source example rather than simply spelling out the entire answer. Pilot that the example is understood and supports the prespecified source interpretation. Cross C-R/R-C with a transparent random-example selection rule versus error-triggered selection. Construct worlds where the exact observed corrective example has the same surface form and source truth under both sampling rules but different evidential relevance to the unsupplied generalization. Everyone receives the same number of source examples and equal time. Establish the ordinary sampling-conditioned response on independent one-step items. Then use new, exchangeable source units to test whether previously experienced correction order changes the inferred selection policy despite an explicitly reset current policy. Define D_Q as the order difference in a held-out scope-transfer error, with exact fact recall held equivalent. The candidate predicts D_Q changes sign between general-rule and genuine-exception worlds, and an experimentally verified policy reset reduces the excess order effect below epsilon. The extra-history term is required only if order-linked policy carryover predicts new-source errors beyond the known sampling-conditioned Bayesian baseline; a mere random-versus-selected example effect is established inference, not a new memetic result. Opening a communicative agenda cannot rescue the private scope error while leaving the selection interpretation unchanged; neither a bundle-allocation change nor a sleep cue has a necessary effect when its proposed state is clamped. Falsify in favor of IH_01 if only output omission changes with intact private scope, or in favor of IH_03 if only the physiological cue interaction remains after sampling comprehension is verified.
What testing it would take
The engine's own read on whether this is testable with methods that already exist.
An online causal-world task can randomize example-selection policies while exactly yoking the realized example, reading time and source truth. Use ethically benign artificial rules and make the policy observable; avoid deception about real-world claims. Independent pilots must show that the policies are understood and that examples differ in diagnosticity without creating a factual contradiction. Closed text relays with new recipients then test whether conditional scope, rather than literal wording, persists after source correction stops. A general claim needs nonnumeric narratives and practices, a new language, and spontaneous selection inference without experimenter explanations. The design is inexpensive, but a one-step sampling effect alone cannot clear the novelty gate.
Other explanations
Every other hypothesis the engine wrote for the same gap, and the observation that would separate the two.
Use invented narrative worlds with a general rule and explicitly indexed exceptions; correction supplies a true, partially diagnostic source example rather than simply spelling out the entire answer. Pilot that the example is understood and supports the prespecified source interpretation. Cross C-R/R-C with a transparent random-example selection rule versus error-triggered selection. Construct worlds where the exact observed corrective example has the same surface form and source truth under both sampling rules but different evidential relevance to the unsupplied generalization. Everyone receives the same number of source examples and equal time. Establish the ordinary sampling-conditioned response on independent one-step items. Then use new, exchangeable source units to test whether previously experienced correction order changes the inferred selection policy despite an explicitly reset current policy. Define D_Q as the order difference in a held-out scope-transfer error, with exact fact recall held equivalent. The candidate predicts D_Q changes sign between general-rule and genuine-exception worlds, and an experimentally verified policy reset reduces the excess order effect below epsilon. The extra-history term is required only if order-linked policy carryover predicts new-source errors beyond the known sampling-conditioned Bayesian baseline; a mere random-versus-selected example effect is established inference, not a new memetic result. Opening a communicative agenda cannot rescue the private scope error while leaving the selection interpretation unchanged; neither a bundle-allocation change nor a sleep cue has a necessary effect when its proposed state is clamped. Falsify in favor of another hypothesis of the same gap if only output omission changes with intact private scope, or in favor of another hypothesis of the same gap if only the physiological cue interaction remains after sampling comprehension is verified.
- What would separate them
Accurate correction may close the speaking agenda and cause cultural omissions predicts: Randomize source-equivalent correction versus restudy, C-R versus R-C, and a content-free outgoing-question agenda opened before versus after C. The agenda lists only which already-present relation must be conveyed; it does not supply its answer. Give an explicit new-recipient ignorance cue in every cell, match overt retrieval and elapsed time, and assess source knowledge on an independent unit set to avoid rehearsal contamination. Define D_A as the C-R minus R-C omission difference. The candidate predicts D_A>delta with no pre-opened agenda, but |D_A|<epsilon when the same agenda is opened before C; opening it only after C fails to rescue within the prespecified interval. Crucially, the correction arm has equal or higher private source accuracy but lower source-specific receiver accuracy. The correction-by-agenda interaction remains after matching trial-level attention, source-memory discrimination, total message length and ordinary given/new status. Uniform correction-time allocation, unchanged source-selection rules and a no-sleep immediate session do not remove it. Subsequent human-only descendants retain the between-lineage semantic advantage over a pilot-defined horizon; exact-artifact yoking abolishes predecessor-history differences. An absence of agenda timing specificity with a selective bundle-allocation effect favors another hypothesis of the same gap; a sleep-stage/cue reversal favors another hypothesis of the same gap; an explicit sampling-rule reversal favors another hypothesis of the same gap. If ordinary audience-design/retrieval models already predict the whole interaction, remove the new-family claim even if the best schedule remains R-C.
- What would separate them
Reconstruction may improve correction by revealing which repairs are useful together predicts: Cross C-R versus R-C with package-versus-separate-target allocation of an equal integration-time budget; everyone sees the same source cards for the same total time and completes the same number of model calls. Include targets whose experimentally defined functional value is additive and targets whose joint correction is complementary, while keeping the proposition rubric fixed. Estimate each participant's complementarity from incentive-compatible choices on independent calibration units. Let D_B be the R-C minus C-R source-specific functional accuracy gap. Predict D_B>delta under separate-target allocation for complementary relations, with a smaller |D_B| under package allocation; the package interaction is near zero for additive units. Manipulating retained relational identity changes measured bundle complementarity and predicts this interaction before seeing the final outcomes. The decisive replay control randomly yokes the exact per-target attention schedule and correction content from package winners to recipients who never bid: the auction-label effect and any unexplained order advantage should fall within epsilon once this allocation path is fixed. Agenda timing, sleep stage and evidence-sampling labels do not produce independent effects in the allocation-only model. Continued advantage after withdrawal must be predicted from what was actually corrected, not from the presence of old prices. If bids fail to forecast repair allocation, or fixed allocation leaves a reproducible agenda-, sleep-, or selection-rule-specific residual, reject this mechanism as the sufficient explanation in favor of the appropriate rival.
- What would separate them
Sleep may favor correction after reconstruction by stabilizing source-role bindings predicts: Independently tag source-relation encoding and the human reconstruction episode with counterbalanced nonsemantic cues. Cross C-R versus R-C with a matched nap/quiet-rest interval, then, in the physiological validation, deliver source-episode versus reconstruction-episode cues during verified slow-wave sleep, with sham and waking-cue controls. Hold clock time, interference, cue count, arousal and source information constant or explicitly counterbalance them. Define D_S as the post-interval R-C minus C-R probability of preserving the arbitrary source relation, subtracting each schedule's immediate baseline on independent items. Predict a stage-by-cue-by-order interaction exceeding delta: stabilizing a correctable source-bound episode benefits R-C, whereas selectively stabilizing a conflicting reconstruction reduces or reverses that advantage. The corresponding effect is weaker than delta in waking/sham controls under this candidate's stated scope. Entity-only retention weakens the interaction despite matched recognition. One-step kernels estimated separately for sleep and wake plus standard retrieval, spacing and targeted-memory-reactivation effects are the nearest strong rival; an extra order-by-binding state must improve held-out prediction beyond them. If only a nonspecific sleep main effect occurs, or a standard replay model already predicts the full contrast, remove the proposed distinct family. If awake agenda timing or package allocation predicts the advantage and verified stage-specific cueing adds no informative effect, prefer another hypothesis of the same gap or another hypothesis of the same gap. After withdrawal, test both original individuals and naive human chains; exact-message replay must eliminate inaccessible predecessor-sleep effects.
What stands behind it
Which of the figures above have a study behind them, which are the engine's own, and what it would take to refute the hypothesis. This audit never judges the idea.
Provenance audit: failed at enrich. Nothing below has been traced yet.