C1 What counts as learned

narrowed 7 attacks: 3 survived, 4 narrowed · serves Q1 · last change 2 October 2026

Encoding has succeeded when the learner can later access the target under cue conditions that match the practice criterion without new presentation of the material — accessibility under that cue set, not mere study exposure or storage-without-access. Under corrective feedback, that accessibility criterion is stricter than what C0's delayed retrieval-over-restudy contrast requires: the contrast can appear equally when unaided practice success is experimentally low.

In practice

Count something as learned only when it can be brought back later, without looking at the material again, from the same kind of prompt it will be needed for. Having read it, seen it, or followed it while it was in front of you does not count.

Where it holds

Inside the circle: the core of the claim. Outside: every bound and carve-out in the claim file, numbered, with the attack that forced it.

forced by a narrowing attackin the original wordingwritten at creation

Canonical retrieval-practice designs that motivate C0 (e.g. Roediger & Karpicke 2006 Exp. 1) typically gate the restudy/test contrast on study exposure rather than item-level verified retrieval; C1 asserts a stricter accessibility criterion than those designs verify.
Forced by attack #1 · Roediger & Karpicke 2006 · 14 September 2026
This bound first appeared in a change that applied several attacks at once; it is matched to #1 because it names that attack's source.
Success here is accessibility under the practice/criterion cue set (Tulving & Pearlstone), not storage availability alone and not “was on the page.”
Forced by attack #2 · Tulving & Pearlstone 1966 · 14 September 2026
This bound first appeared in a change that applied several attacks at once; it is matched to #2 because it names that attack's source.
Encoding success is criterion-relative (Morris, Bransford & Franks transfer-appropriate processing): it is defined relative to the practice/criterion task, not as task-free storage strength.
Forced by attack #5 · Morris, Bransford & Franks 1977 · 14 September 2026
Kill-condition prong 2 is met under corrective feedback (Gupta, Pan & Rickard 2022, DOI 10.3758/s13421-021-01236-4, full text verified 2026-10-01): with practice success experimentally ~0.22 versus ~0.55 via prior-study load, retrieval-with-feedback still beat restudy at ~48 h by a similar margin (Exp. 1 finals ~0.40 vs 0.24 at low success and ~0.60 vs 0.41 at high, no Training×Repetition interaction), so C0's delayed contrast with feedback does not require high unaided practice-phase accessibility success.
Forced by attack #6 · Gupta, Pan & Rickard 2022 · 1 October 2026
Added after the archived record of earlier wordings ends.
Racsmány, Szőllősi & Marián 2020 (DOI 10.3758/s13421-020-01041-5, full text verified 2026-10-02): no-feedback equality residual tested and failed — practice success experimentally varied via prior-study load with no practice feedback and a restudy arm
Forced by attack #6 · Gupta, Pan & Rickard 2022 · 1 October 2026
Added after the archived record of earlier wordings ends.
First-final 1-week TE present but not equal across levels (Exp 1 practice ~25–29%, TE d=0.43; Exp 2 ~66–75%, d=1.33; Exp 3 ~76–81%, d=1.71 / ~61% vs ~28%), so without feedback the contrast does not appear equally when accessibility fails for most items
Forced by attack #6 · Gupta, Pan & Rickard 2022 · 1 October 2026
Added after the archived record of earlier wordings ends.
Survived 2026-10-02.
Forced by attack #6 · Gupta, Pan & Rickard 2022 · 1 October 2026
Added after the archived record of earlier wordings ends.

7 of 7 bounds and carve-outs were forced by a specific attack. 3 attacks hit the boundary and did not move it: #3, #4, #7.

How it changed

Replay of how the wording of C1 changed, attack by attack. Earlier wordings come from the published history of the old learning-graph site, which kept a snapshot after every change. Added words and ~~removed words~~ are marked. Steps are in the order the verdicts took effect.

Starting wordingfirst published 14 September 2026, 22:19 CEST
Encoding has succeeded when the learner can later retrieve the target under conditions that match the practice criterion without new presentation of the material — the content is available in memory, not merely exposed.
No bounds were written yet.
Attack #1 + #2 narrowed Roediger & Karpicke 2006 · Tulving & Pearlstone 1966 · ledger date 14 September 2026 published 14 September 2026, 22:54 CEST
Attack: In Roediger & Karpicke 2006 Experiment 1, the restudy-versus-test contrast after one study episode admits all idea units without an item-level availability or retrieval gate, so “encoded” is operationalized as “was studied / was on the page.”
Attack: Tulving & Pearlstone 1966 show many studied items remain available in storage yet inaccessible without cues, so “can later retrieve under criterion-matched conditions” is accessibility, not the same as “available in memory.”
Attacks #1 and #2 were applied in the same cycle and published as one change. The record holds only their combined result, so they are shown together; no in-between wording is invented.
Encoding has succeeded when the learner can later ~~retrieve~~ access the target under cue conditions that match the practice criterion without new presentation of the material — ~~the content is available in memory, not merely exposed.~~ accessibility under that cue set, not mere study exposure or storage-without-access.
Bounds field after this step (changes marked):
Canonical retrieval-practice designs that motivate C0 (e.g. Roediger & Karpicke 2006 Exp. 1) typically gate the restudy/test contrast on study exposure rather than item-level verified retrieval; C1 asserts a stricter accessibility criterion than those designs verify. Success here is accessibility under the practice/criterion cue set (Tulving & Pearlstone), not storage availability alone and not “was on the page.”
Attack #3 survived Karpicke & Blunt 2011 · ledger date 14 September 2026 published 14 September 2026, 22:54 CEST
Attack: Karpicke & Blunt 2011 equated post-study learning via nearly identical idea production on initial concept maps (~0.78) and free recall (~0.81) before comparing retrieval practice to elaborative study, so at least one canonical design uses a stricter-than-exposure availability check.
Claim wording unchanged.
Encoding has succeeded when the learner can later access the target under cue conditions that match the practice criterion without new presentation of the material — accessibility under that cue set, not mere study exposure or storage-without-access.
Attack #4 survived Rowland 2014 · ledger date 14 September 2026 published 14 September 2026, 22:54 CEST
Attack: Rowland 2014 finds no significant testing-versus-restudy effect when initial retrieval success is ≤50% (g ≈ 0.03) but significant effects at higher success, so C0’s contrast does not appear equally when the stricter availability/retrieval test fails.
Claim wording unchanged.
Encoding has succeeded when the learner can later access the target under cue conditions that match the practice criterion without new presentation of the material — accessibility under that cue set, not mere study exposure or storage-without-access.
Attack #5 narrowed Morris, Bransford & Franks 1977 · ledger date 14 September 2026 published 14 September 2026, 22:55 CEST
Attack: Morris, Bransford & Franks 1977 transfer-appropriate processing implies encoding success is relative to a retrieval task, so “encoded” under C1 is criterion-relative rather than an absolute state of the memory trace independent of the cue set.
Claim wording unchanged.
Encoding has succeeded when the learner can later access the target under cue conditions that match the practice criterion without new presentation of the material — accessibility under that cue set, not mere study exposure or storage-without-access.
Bounds field after this step (changes marked):
Canonical retrieval-practice designs that motivate C0 (e.g. Roediger & Karpicke 2006 Exp. 1) typically gate the restudy/test contrast on study exposure rather than item-level verified retrieval; C1 asserts a stricter accessibility criterion than those designs verify. Success here is accessibility under the practice/criterion cue set (Tulving & Pearlstone), not storage availability alone and not “was on the page.” Encoding success is criterion-relative (Morris, Bransford & Franks transfer-appropriate processing): it is defined relative to the practice/criterion task, not as task-free storage strength.
Attack #7 survived Racsmány, Szőllősi & Marián 2020 · ledger date 2 October 2026
Attack: When practice success is experimentally lowered to ~25–29% with no practice feedback via a single prior study, retrieval still beats restudy at a 1-week first final (d = 0.43) but by far less than when practice success is ~76–80% (d = 1.71), so C0's retrieval-over-restudy effect does not appear equally at experimentally low practice success without corrective feedback.
Logged after the archived record ends; the wording at the time of this attack is not recorded.
Wording at this point: not recorded.
Attack #6 narrowed Gupta, Pan & Rickard 2022 · ledger date 1 October 2026
Attack: When prior study is reduced so practice retrieval success falls to ~0.22, retrieval with corrective feedback still beats equal-time restudy at a ~48 h test by as much as when practice success is ~0.55, so C0's retrieval-over-restudy effect appears equally when the stricter accessibility test fails for most items.
This change happened after the archived record of earlier wordings ends; any intermediate wordings were not recorded.
Encoding has succeeded when the learner can later access the target under cue conditions that match the practice criterion without new presentation of the material — accessibility under that cue set, not mere study exposure or storage-without-access. Under corrective feedback, that accessibility criterion is stricter than what C0's delayed retrieval-over-restudy contrast requires: the contrast can appear equally when unaided practice success is experimentally low.
Bounds field after this step (changes marked):
Canonical retrieval-practice designs that motivate C0 (e.g. Roediger & Karpicke 2006 Exp. 1) typically gate the restudy/test contrast on study exposure rather than item-level verified retrieval; C1 asserts a stricter accessibility criterion than those designs verify. Success here is accessibility under the practice/criterion cue set (Tulving & Pearlstone), not storage availability alone and not “was on the page.” Encoding success is criterion-relative (Morris, Bransford & Franks transfer-appropriate processing): it is defined relative to the practice/criterion task, not as task-free storage strength. Kill-condition prong 2 is met under corrective feedback (Gupta, Pan & Rickard 2022, DOI 10.3758/s13421-021-01236-4, full text verified 2026-10-01): with practice success experimentally ~0.22 versus ~0.55 via prior-study load, retrieval-with-feedback still beat restudy at ~48 h by a similar margin (Exp. 1 finals ~0.40 vs 0.24 at low success and ~0.60 vs 0.41 at high, no Training×Repetition interaction), so C0's delayed contrast with feedback does not require high unaided practice-phase accessibility success. Racsmány, Szőllősi & Marián 2020 (DOI 10.3758/s13421-020-01041-5, full text verified 2026-10-02): no-feedback equality residual tested and failed — practice success experimentally varied via prior-study load with no practice feedback and a restudy arm; first-final 1-week TE present but not equal across levels (Exp 1 practice ~25–29%, TE d=0.43; Exp 2 ~66–75%, d=1.33; Exp 3 ~76–81%, d=1.71 / ~61% vs ~28%), so without feedback the contrast does not appear equally when accessibility fails for most items; survived 2026-10-02.

The current claim file differs from the last archived wording; the difference is shown as the final step.

Attack record

#1 narrowed 14 September 2026 · Roediger & Karpicke 2006 · target: kill-condition
In Roediger & Karpicke 2006 Experiment 1, the restudy-versus-test contrast after one study episode admits all idea units without an item-level availability or retrieval gate, so “encoded” is operationalized as “was studied / was on the page.”
Strongest reading and verdict reasoning
Steelmans to: Kill-condition prong 1 is met for the canonical seed design that motivates C0; C1 cannot stand as if those studies already verified criterion-matched retrieval before the contrast.
Why: Named design fact — Exp. 1 assigns restudy vs test after one study period with no item-level retrieval gate; initial free recall when given is ~70%. Claim and Bounds rewritten same turn so C1 remains the graph’s stricter accessibility precondition, with an explicit Bound that seed designs often use study-exposure. Brief rewritten. Kill withheld: C1 is still tenable as the precondition C0 needs, once the operational gap is Bounded.
#2 narrowed 14 September 2026 · Tulving & Pearlstone 1966 · target: claim-text
Tulving & Pearlstone 1966 show many studied items remain available in storage yet inaccessible without cues, so “can later retrieve under criterion-matched conditions” is accessibility, not the same as “available in memory.”
Strongest reading and verdict reasoning
Steelmans to: The living sentence conflates storage availability with cue-dependent accessibility; C1 must define success as accessibility under the practice/criterion cue set.
Why: Named result — cued recall recovers items missed under free recall. Claim rewritten same turn from “available in memory” to accessibility under the practice/criterion cue set; Bounds note availability≠accessibility. Brief rewritten.
#3 survived 14 September 2026 · Karpicke & Blunt 2011 · target: kill-condition
Karpicke & Blunt 2011 equated post-study learning via nearly identical idea production on initial concept maps (~0.78) and free recall (~0.81) before comparing retrieval practice to elaborative study, so at least one canonical design uses a stricter-than-exposure availability check.
Strongest reading and verdict reasoning
Steelmans to: Kill-condition prong 1 is not universal across retrieval-practice designs; some verify post-study production before the critical contrast.
Why: Genuine kill-condition test: this design equates post-study idea production rather than treating encoding as page exposure alone. Prong 1 does not fire universally. Claim field unchanged; prior Bounds on seed-design exposure gates stand.
#4 survived 14 September 2026 · Rowland 2014 · target: kill-condition
Rowland 2014 finds no significant testing-versus-restudy effect when initial retrieval success is ≤50% (g ≈ 0.03) but significant effects at higher success, so C0’s contrast does not appear equally when the stricter availability/retrieval test fails.
Strongest reading and verdict reasoning
Steelmans to: Kill-condition prong 2 fails — the retrieval-over-restudy advantage shrinks or vanishes when initial retrieval success is low.
Why: Genuine kill-condition test: meta-analytic moderator shows nonsignificant testing effect at ≤50% initial success. Prong 2 (“appears equally when stricter test fails”) not met. Claim unchanged.
#5 narrowed 14 September 2026 · Morris, Bransford & Franks 1977 · target: bound
Morris, Bransford & Franks 1977 transfer-appropriate processing implies encoding success is relative to a retrieval task, so “encoded” under C1 is criterion-relative rather than an absolute state of the memory trace independent of the cue set.
Strongest reading and verdict reasoning
Steelmans to: C1’s match-to-practice-criterion clause must Bound that encoding success is defined relative to the practice/criterion task, not as task-free storage strength.
Why: Named TAP result. Bounds rewritten same turn to state criterion-relativity explicitly; Claim field already used “cue conditions that match the practice criterion” from cycle 1 — Bounds now pin that reading. Brief living sentence unchanged (already accessibility-under-cue-set). Status remains weakened.
#6 narrowed 1 October 2026 · Gupta, Pan & Rickard 2022 · target: kill-condition
When prior study is reduced so practice retrieval success falls to ~0.22, retrieval with corrective feedback still beats equal-time restudy at a ~48 h test by as much as when practice success is ~0.55, so C0's retrieval-over-restudy effect appears equally when the stricter accessibility test fails for most items.
Strongest reading and verdict reasoning
Steelmans to: Kill-condition prong 2 is met under feedback — TE vs restudy is experimentally independent of practice success lowered via prior study, contrary to Rowland's correlational ≤50% moderator and answering open question 5.
Why: Nightly-attacks 2026-10-01 04:11 Europe/Stockholm. Full OA text verified (eScholarship CC BY). Exp. 1 training success 0.217 (1×) vs 0.55 (4×); final restudy 0.238/0.411 vs test 0.404/0.601; Training×Repetition F(1,93)=0.0064, p=0.91. Exp. 3 training 0.231 vs 0.510; final restudy 0.286/0.429 vs test 0.434/0.60; interaction p=0.42. Genuine kill-condition test with restudy arm; prong 2 fires under corrective feedback. Claim and Bounds rewritten same turn: accessibility remains the definition of encoding success, but is scoped as stricter than what C0's delayed contrast requires when feedback is given; What-would-kill updated (remaining threat: same pattern without feedback, or collapse of accessibility to mere exposure/feedback). Brief rewritten. Status remains weakened. Distinct from Chan 2024 (no restudy) and from Rowland 2014 survive (opposite force).
#7 survived 2 October 2026 · Racsmány, Szőllősi & Marián 2020 · target: kill-condition
When practice success is experimentally lowered to ~25–29% with no practice feedback via a single prior study, retrieval still beats restudy at a 1-week first final (d = 0.43) but by far less than when practice success is ~76–80% (d = 1.71), so C0's retrieval-over-restudy effect does not appear equally at experimentally low practice success without corrective feedback.
Strongest reading and verdict reasoning
Steelmans to: Kill-condition residual prong 2 (no-feedback experimental case left after Gupta 2022) fails — TE vs restudy is present at low no-feedback success but not equal across experimentally manipulated success levels; without feedback, accessibility still matters for the size of C0's contrast.
Why: Nightly-attacks 2026-10-02 04:13 Europe/Stockholm. Full OA CC BY text verified (Europe PMC PMC7498445). Exp. 1 (n=29, 1× prior study, no practice feedback): practice success 25.3–28.7%; first final test > restudy t(28)=2.29, p=.03, d=0.43 (Fig. 3: ~23% vs ~16%). Exp. 2 (n=30, 3×): practice 66.1–75.0%; first final TE d=1.33. Exp. 3 (n=24, 6×): practice 76.1–80.9%; first final TE d=1.71 (Fig. 3: ~61% vs ~28%). Genuine kill-condition test of the residual no-feedback equality prong with restudy arm; kill fails because TE does not appear equally. Distinct from Gupta 2022 (feedback, equal TE), Chan 2024 (no restudy), Rowland 2014 (correlational ≤50%), and from the reversal-after-final-feedback framing not emitted 2026-10-01. Claim field unchanged; Bounds and What-would-kill note the survive and drop the answered residual. Status remains weakened.

Evidence

9 linked sources: 4 supports, 5 limits (first-pass draft, pending human review). supports: Karpicke & Blunt 2011, Rowland 2014, Kornell 2025, Racsmány et al. 2020 · limits: Morris et al. 1977, Roediger & Karpicke 2006, Tulving & Pearlstone 1966, Chan et al. 2024, Gupta et al. 2022

Full claim record in the evidence library →