24-hour cooling-off rule

Parent: Draft Review Revise Loop · Published reference · snapshot 2026-09-24

Also known as: 24-hour cooling-off before sentence passes

Depth-first rabbithole dossier for 24-hour cooling-off rule; source-anchored research pack.

These notes link each claim to its source. A source may be a research report hosted on this site rather than the primary document. A published reference means the content is available; it does not certify independent review or accuracy.Read the editorial policy and follow the sources before relying on a claim.

Definitions

6. A meta-analysis of 117 studies found a positive incubation effect. Incubation means setting a problem aside before trying again. https://gwern.net/doc/psychology/writing/2009-sio.pdf 7. The effect is low-to-medium in size, with mean d ≈ 0.29. https://www.researchgate.net/publication/24001208_Does_Incubation_Enhance_Problem_Solving_A_Meta-Analytic_Review 8. Divergent-thinking tasks (producing many ideas) gained more from incubation than linguistic or visual insight tasks did. https://eric.ed.gov/?id=EJ827110 9. Longer preparation before the break gave a larger incubation effect. For a writer [source]

Structure and components

Framing note: no source found in this run tests "24 hours" as a specific interval for writing revision. The rule is a folk heuristic. The evidence below comes from three research lines that each test part of its mechanism: (a) incubation, (b) self-generated-text proofreading, and (c) whether sleep dampens emotion. [source]
**In scope:** the rule as a step in a writing and revision loop. The writer finishes a draft, holds it unsent for about 24 hours (usually one night's sleep), then re-reads it before deciding to send, revise, or discard it. The report covers the rule's parts, the causal mechanisms proposed for it, its invariants, and the limits of the evidence. [source]
1. Practitioner writing guidance states the rule as: after finishing a rough draft, leave it until the next day before revising, so you read it "with fresh eyes" and see errors you would miss right after writing. https://utminers.utep.edu/omwilliamson/univ1301/timemgmt.htm 2. The rule has four parts: (a) a trigger, which is a finished draft; (b) a hold, in which the draft is not released; (c) a delay interval of about 24 hours; (d) a re-read under a changed mental state, which ends in a decision to release, revise, or discard. This breakdown is synthesis drawn from claims 1, 5 and 16; no singl [source]
| Component | Grade | |---|---| | Irreversibility invariant (C05) | Sound: logical, and recorded since Horace | | Some separation beats none (C51) | Modest meta-analytic support that depends on conditions | | 24 hours specifically (C06) | **No evidence.** A convention (C08) | | Familiarity causes missed errors (C32–C37) | **Contested.** Three positions | | Sleep cools anger (C46–C50) | **Contested; leans against** | | Time without rumination cools anger (C43–C45) | Supported | | Delay improves structure more than typos (C53, C62, C66) | Plausible inference, untested | [source]
33. **Cost:** each draft cycle takes about a day longer. The evidence (claims 1, 11, 17) suggests that a shorter break, a nap or a night's sleep, or making the text look unfamiliar can deliver part of the benefit at lower cost. 34. **Benefit is conditional:** the gap helps most after real effort on the draft (claims 9, 15), when the break is spent on low-demand activity (claims 10, 26), and on idea-level or structural problems more than on catching typos (claims 8, 21–23). 35. **How to test it yourself:** compare revisions with and without the gap on matched drafts. Count structural (meaning-l [source]

How it works

- **C06** No primary study tests 24 hours against other intervals for draft or reply quality. This is a negative finding in all four reports. [M,H,E,P] - **C07** The earliest explicit "24-hour" wording found is Christopher Peterson in *Psychology Today*, 17 Aug 2011: "When an e-mail message pushes your buttons, do not answer it for at least 24 hours." He presents it as his own rule and cites no earlier source. [H] S23 - **C08** Peterson's reason for 24 hours: it is "a permissible delay before responding," and the writer "will likely have slept on the message." It is a social-acceptability conv [source]
In scope: the practice of letting a finished draft rest for about 24 hours before reviewing, revising, or sending it. That covers the evidence for the "rest" mechanism, where the evidence stops, the conditions under which the rule fails, and where sources disagree. [source]
24. Inference from claims 2 and 10: the evidence supports "some separation beats none" more than it supports 24 hours in particular. The one proofreading result that restored own-text performance used two weeks. https://gwern.net/doc/psychology/writing/2009-sio.pdf ; https://englelab.gatech.edu/articles/2022/Burgoyne%20et%20al.%20(2022)%20Revisiting%20the%20self-generation%20effect%20in%20proofreading.pdf 25. Inference from claims 18 and 19: if the goal is emotional de-escalation, an interval spent awake may do more than one that includes a night of sleep. The rule's "sleep on it" phrasing may [source]
11. RFC 1855 *Netiquette Guidelines* (S. Hambridge, Intel, October 1995), §2.1.1 "For mail," states: "Wait overnight to send emotional responses to messages." This is the earliest standards-body wording of the rule found in this run. — https://www.rfc-editor.org/rfc/rfc1855.txt 12. RFC 1855 is an Informational RFC (FYI 28), not a standards-track document. The rule is guidance, not a protocol requirement. — https://www.rfc-editor.org/rfc/rfc1855.txt 13. In October 2008, Gmail shipped "Mail Goggles," built by engineer Jon Perlow. When enabled, it made users solve arithmetic problems before sendi [source]
Horace: nine years before publishing (c. 19 BCE) → Quintilian: "until the love of invention cools," reread as a reader (c. 95 CE) → Jefferson: count to ten or a hundred when angry (1825) → Lincoln's unsent "hot letter" (1863) → RFC 1855: "wait overnight" for emotional email (1995) → Gmail Mail Goggles, friction before sending (2008) → the explicit "24-hour rule" in Peterson's Psychology Today post (2011) → writing centers: "a day—a few hours even" (current). [source]
1. I didn't write the dossier to disk because shell access was blocked, so I couldn't check how earlier runs name their output files. Should I save it as `~/.global-ai-hub/research-runs/frontier-2026-09-24/24-hour-cooling-off-rule/synthesis.md`? (Default: no, leave it in this response.) 2. Should I run one more deepening pass on the four open leads in §12, to try to reach SATURATED-DEPTH? (Default: no.) 3. Some claims (C70 and below) are shown as the incubation evidence says they should be. Fix C69's anchor before using it: it should read "S2 S45 S53 S33". [source]
1. Writing-process research that directly compares short and long delays between drafting and revising (the Hayes & Flower line of work), including whether a delay changes revision *depth*. 2. Pre-1995 print etiquette and business-writing manuals, to find earlier uses of "overnight" or "24 hours." 3. Full texts that returned 403 or were read only as abstracts: Baran, Denis, Cai, both Pilotti papers, the 8-hour detail in Bushman 2005, the LOC scan, and King's own text. 4. Settling which paper the Pilotti & Chodorow 2009 citation refers to (§10 item 1). [source]
Limits: - I read the full text of Sio & Ormerod 2009 (pp. 94–107) and the first two pages of Burgoyne 2022. - For Cai 2009, Nieuwenstein 2015, Baran 2012, Denis 2022, and both Pilotti papers, the claims rest on abstracts or search summaries, because the full text returned 403. Their Pilotti 2009 and Daneman & Stainton details come from Burgoyne 2022's literature review. - The gap that remains: no study tests a 24-hour delay against other intervals for writing revision quality. [source]
**In scope:** the practice of letting a finished draft or an emotionally charged reply rest for a fixed interval (conventionally 24 hours, or "overnight") before revising or sending it. That covers where the practice came from, how the interval changed over time, the primary sources behind it, and the evidence for and against the mechanism it relies on. [source]
17. The UNC Writing Center tells students to "Wait awhile after you've finished a draft before looking at it again. The Roman poet Horace thought one should wait nine years, but that's a bit much. A day—a few hours even—will work." The modern 24-hour guidance cites Horace directly as its ancestor. — https://writingcenter.unc.edu/tips-and-tools/revising-drafts/ 18. Stephen King's *On Writing* (2000) recommends that a novel draft rest for at least six weeks before the writer rereads it, so that it reads "like the work of someone else." This is the same rationale as Quintilian's, at a much longer [source]

Measurements and reference values

19. Daneman & Stainton (1993, *Reading and Writing*) found that writers detected fewer errors in their own just-written essays than in unfamiliar essays by other people. They attributed the gap to "extreme familiarity." — https://eric.ed.gov/?id=EJ498842 20. As summarized by Burgoyne et al. (2022), the Daneman & Stainton self-proofreading deficit was about 20% fewer errors detected. After a **two-week** delay, writers proofread their own essays about as accurately as a familiar essay by someone else. This is the closest experimental support for a cooling-off interval, and the interval tested w [source]
Verdict: **BUDGET_EXHAUSTED (soft stop).** Pass 4 returned 0%, but the stopping rule needs two consecutive passes under 5%, so saturation is not proven. One more pass through pre-1995 print etiquette and business-writing manuals could still date the "overnight / 24 hours" wording earlier. [source]
- **C51** Sio & Ormerod (2009) meta-analysis of 117 effect sizes: incubation has a positive effect. The mean effect is about 0.29 (95% CI 0.21–0.39), which is small to medium. [M,E,P] S44 S45 S46 - **C52** Longer preparation before the break predicted a larger effect (β = .03, p < .05). [M,H,E,P] S44 S45 - **C53** Divergent (many-answer) tasks gained more than linguistic or visual insight tasks. [M,H,E,P] S44 S45 - **C54** For linguistic problems, a low-demand filler task beat pure rest (β = .54). For creative problems, a high-demand task during the break reduced the effect. [M,H,E,P] S44 S45 [source]
Each report's own internal curve was also a soft stop: [M] 40% on its last pass; [H] 29% then 0% (one pass under 5%, but the rule needs two in a row); [P] 26%. [source]
8. Daneman & Stainton (1993) found that people detected fewer errors in essays they had written themselves than in unfamiliar essays by others. They detected more errors in familiar essays by others than in unfamiliar ones. The authors attribute the self-generation disadvantage to "extreme familiarity." https://eric.ed.gov/?id=EJ498842 9. In that study, students who proofread their own essays detected about 20% fewer errors than students who proofread a familiar essay by someone else (as summarized by Burgoyne et al. 2022). https://doi.org/10.1007/s00426-022-01699-3 10. The proposed pathway is [source]
14. Wagner et al. (2004) found that after sleep, more than twice as many people discovered a hidden task rule as after the same time awake. https://pubmed.ncbi.nlm.nih.gov/14737168/ 15. In the same study, sleep did not help when there had been no initial training. The gap helps only after real engagement with the task, which matches claim 9. https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=EXT_ID:14737168%20AND%20SRC:MED&resultType=core&format=json 16. A 2024 preregistered study describes the evidence that sleep aids insight as "mixed". It cites studies that found no benefit, or n [source]
- **Does a gap help you catch errors in your own text?** Daneman & Stainton (1993) found a self-generation penalty of about 20% (claim 20). Burgoyne et al. (2022) mostly failed to replicate it (claim 21). If the penalty is weak, the proofreading case for a 24-hour gap is weaker than folk advice claims. - **Is sleep special?** Wagner et al. (2004) found more than twice as much insight after sleep (claim 14). Later studies found no benefit, or no difference from waking rest (claim 16). The 2024 study suggests that sleep stage matters (claim 17). - **Does unconscious processing happen during the [source]

Problems, failure modes and limitations

- **C01** Practitioner form: finish a rough draft and leave it until the next day, so you reread it "with fresh eyes" and catch errors you would miss right after writing. [M] S1 - **C02** The UNC Writing Center puts it this way: "Wait awhile after you've finished a draft… The Roman poet Horace thought one should wait nine years, but that's a bit much. A day—a few hours even—will work." [H,P] S2 - **C03** *Synthesis:* the rule has two variants that share one mechanism. In the **revision variant**, you hold a draft and then revise it (S2). In the **send variant**, you hold a charged message and [source]
- **C19** Horace, *Ars Poetica* (c. 19–10 BCE): show the draft to critics, then "nonumque prematur in annum / membranis intus positis" (keep it back until the ninth year, the parchment kept at home). [M,H] S5 S6 - **C20** Classicists disagree on what Horace meant: store the finished poem for nine years, or spend nine years composing it. [H] S8 - **C21** Quintilian's letter to Trypho (*Inst.*, c. 95 CE) cites Horace. He says he rested his work "ut refrigerato inventionis amore diligentius repetitos tamquam lector perpenderem": so that, once his love of his own invention had cooled, he could wei [source]
1. A meta-analysis of 117 effect sizes found a positive incubation effect. The random-effects weighted mean was 0.29 (95% CI 0.21–0.39). https://gwern.net/doc/psychology/writing/2009-sio.pdf 2. In that meta-analysis, the length of the incubation period did **not** predict effect size (β < .001, n.s., across all 114 studies). This contradicts the idea that any particular rest length, 24 hours included, is what matters. https://gwern.net/doc/psychology/writing/2009-sio.pdf 3. The length of the preparation period (how long the person worked before the break) **did** predict a larger incubation ef [source]
24. Burgoyne et al. (2022, *Psychological Research*, https://doi.org/10.1007/s00426-022-01699-3) ran two eye-tracking experiments. Experiment 1 "failed to reveal a self-generation effect," and Experiment 2 found only "a weak but non-significant" effect. If own-text familiarity is not a robust handicap, the main rationale for a revision cooling-off period is weaker. — https://englelab.gatech.edu/articles/2022/Burgoyne%20et%20al.%20(2022)%20Revisiting%20the%20self-generation%20effect%20in%20proofreading.pdf 25. The same paper notes that Daneman & Stainton used very small samples (n = 10 per cond [source]
30. Outlook desktop can hold every outgoing message with a rule: "defer delivery by a number of minutes", up to 120 minutes. https://support.microsoft.com/en-us/outlook/mail/delay-or-schedule-sending-email-messages-in-outlook 31. The new Outlook for Windows does not offer this rule for all mail. You can only schedule each message by hand, so tooling cannot enforce a blanket delay there. https://learn.microsoft.com/en-us/answers/questions/1312123/how-to-delay-delivery-for-all-messages-in-new-outl 32. Standard mail-client tooling can enforce a gap of minutes to hours, but not 24 hours. A full 24 [source]
1. Sio & Ormerod (2009), Does incubation enhance problem solving? A meta-analytic review. Psychological Bulletin 135(1). https://gwern.net/doc/psychology/writing/2009-sio.pdf 2. Burgoyne et al. (2022), Revisiting the self-generation effect in proofreading. Psychological Research. https://englelab.gatech.edu/articles/2022/Burgoyne%20et%20al.%20(2022)%20Revisiting%20the%20self-generation%20effect%20in%20proofreading.pdf 3. Daneman & Stainton (1993), The generation effect in reading and proofreading. Reading and Writing 5. https://eric.ed.gov/?id=EJ498842 4. Pilotti & Chodorow (2009), Error detec [source]
18. According to psychologist Tom Stafford, people miss their own typos because they already know the meaning they intended. That expectation fills in what is missing on the page. https://kirkville.com/whats-up-with-that-why-its-so-hard-to-catch-your-own-typos-wired/ 19. Stafford's advice is to make the text look unfamiliar: change the font or colours, or print it and edit by hand. This substitutes a different kind of distance for time. https://kirkville.com/whats-up-with-that-why-its-so-hard-to-catch-your-own-typos-wired/ 20. Daneman & Stainton (1993) found that students detected about 20% fe [source]
**Met.** The report draws on more than 3 independent sources on different hosts: APA/Lancaster, Nature/PubMed, Springer Psychological Research, Cambridge JDM, SAGE PSPB, NCTE/ERIC, UNC, and Microsoft. It includes disconfirming evidence (claims 16, 21, 29). **Caveat:** none of these sources directly tests a 24-hour gap in revising prose. Every link from the evidence to the "24-hour" figure is an inference, and each one is marked as such above. [source]

Comparisons and alternatives

**Handoffs to concept-family-explorer** (sibling concepts, not researched here): incubation effect / Wallas stages; unconscious-thought theory; self-generation effect in proofreading; rumination vs. distraction; sleep-dependent emotional memory; construal-level theory; email emotional miscommunication; send-delay and undo-send tooling; read-aloud proofreading; the 24-hour purchase rule; the FTC Cooling-Off Rule (name collision only). [source]
17. The "overnight therapy" hypothesis (Walker & van der Helm 2009) proposes that REM sleep depotentiates the emotional charge of memories. That is the implied mechanism behind "sleep on an angry email." https://pmc.ncbi.nlm.nih.gov/articles/PMC2890316/ 18. **Disconfirming:** Baran et al. (2012, J. Neurosci.) found that emotional reactivity to negative pictures dropped greatly over *wake* but was relatively *preserved* over sleep. More REM time went with more preservation. https://www.jneurosci.org/content/32/3/1035 19. **Disconfirming:** Denis et al. (2022, PNAS) ran two large online experime [source]
- **Does familiarity with your own text hurt proofreading?** It hurts: Daneman & Stainton 1993 (https://eric.ed.gov/?id=EJ498842) and Pilotti & Chodorow 2012 (https://onlinelibrary.wiley.com/doi/10.1111/j.1467-9817.2010.01462.x). It helps: Pilotti & Chodorow 2009, larger n. No reliable effect: Burgoyne et al. 2022 (https://englelab.gatech.edu/articles/2022/Burgoyne%20et%20al.%20(2022)%20Revisiting%20the%20self-generation%20effect%20in%20proofreading.pdf). If there is no self-generation deficit, the proofreading rationale for the cooling-off rule mostly collapses. - **Does sleep cool emotion?** [source]
- **Met.** 20 claims, drawn from sources on 12+ distinct hosts. Primary sources include Psychological Bulletin, Psychological Research, Reading and Writing (via ERIC), Current Biology, J. Neuroscience, Nature, PSPB, Health Psychology, Psychological Review, and Horace via Perseus. - The run sought disconfirming evidence on purpose and found it for two of the four mechanisms (Burgoyne 2022; Baran 2012 and Lipinska 2022). - **Gap:** no primary study tests the 24-hour interval itself. The specific number rests on practitioner convention. - **Saturation:** pass 0 produced 12 claims and pass 1 produ [source]
- **C63** **The delay is spent ruminating.** If you reread and polish the angry draft during the wait, that is rumination, and the delay makes anger worse (from C43–C45). [M,H,P] S36 S38 - **C64** **The draft was thin.** Rest after a hurried, shallow draft yields little (from C52, C58, C59). [E,P] S45 S50 S51 - **C65** **Sleep keeps the grievance.** For de-escalation, a stretch of time awake may beat one that includes a night of sleep. "Sleep on it" may point at the wrong mechanism (inference from C48–C50). [M,E] S41 S43 - **C66** **No revision strategy.** Sommers (1980): students treated revi [source]
9. Daneman & Stainton (1993) found that writers detected fewer errors in their own just-written essays than in unfamiliar essays by others. They attributed the gap to "extreme familiarity." https://eric.ed.gov/?id=EJ498842 10. In the same study, the own-text deficit disappeared when participants proofread their essays **two weeks** after writing them. The only tested interval that restored performance was two weeks, not 24 hours. https://englelab.gatech.edu/articles/2022/Burgoyne%20et%20al.%20(2022)%20Revisiting%20the%20self-generation%20effect%20in%20proofreading.pdf 11. The Daneman & Stainto [source]
- **The proofreading effect may not replicate.** Burgoyne et al. (2022) ran two eye-tracking experiments. Experiment 1 found no self-generation effect. Experiment 2 induced overfamiliarity and found only a weak, non-significant effect. The original result used about 10 participants per condition. https://doi.org/10.1007/s00426-022-01699-3 - Pilotti & Chodorow (2009), with a larger sample, found that familiarity with one's own text **helped** error detection (as cited by Burgoyne et al.). https://doi.org/10.2466/pms.109.3.627-645 - **Sleep may preserve negative emotion instead of reducing it.** [source]
- Incubation / Wallas four-stage model - Sleep-dependent memory consolidation - Self-generation effect in proofreading - Rumination vs. distraction in emotion regulation - Scheduled / delayed send tooling - FTC Cooling-Off Rule (same name, different concept; disambiguation only) [source]

Facts and statements

**Scope.** In scope: holding a finished draft, or an emotionally charged message, for about 24 hours (or overnight) before you revise, send or discard it. Out of scope: the FTC "Cooling-Off Rule," which is a consumer-law right to cancel a sale and only shares the name [M,H,P]. Also out of scope: sibling steps of the Draft Review Revise Loop and the parent loop itself. [source]
**Out of scope:** legal consumer "cooling-off periods" (for example, statutory contract cancellation windows), 24-hour purchase-delay rules, sibling steps of the Draft Review Revise Loop (peer review, read-aloud, checklists), and the parent loop itself. [source]
- **In scope:** the rule itself. After you finish a draft, or write a message while upset, you wait about 24 hours before you revise or send it. This report covers how the rule is used, what it costs, how to test whether it works, and what that means in practice. - **Out of scope:** the Draft Review Revise Loop as a whole, peer review, other revision techniques, and the unrelated FTC "Cooling-Off Rule" for door-to-door sales. The FTC rule has the same name but is a consumer-law right to cancel a purchase, not a writing practice. - **Headline finding:** no primary study I found tests a 24-hour [source]
**Bottom line.** No primary study tests 24 hours against any other interval, for draft quality or for reply quality. All four reports agree on this. The number comes from convention and social acceptability. The rule's underlying mechanisms have partial support, and each one is contested. [source]
**Out of scope:** the Draft Review Revise Loop as a whole, peer review, other revision techniques, and the consumer-law "Cooling-Off Rule" (a separate FTC rule on door-to-door sales that only shares the name). [source]
Handoffs to concept-family-explorer (siblings surfaced, not researched): incubation effect, unconscious-thought theory, self-generation effect in proofreading, anger rumination, email emotional miscommunication, send-delay and undo-send tooling, and the FTC Cooling-Off Rule (name collision). [source]
1. A university writing guide recommends waiting after you finish a draft before you reread it. It says "A day—a few hours even—will work", and notes that Horace advised waiting nine years. https://writingcenter.unc.edu/tips-and-tools/revising-drafts/ 2. So one mainstream guide treats a few hours as enough and 24 hours as a comfortable default, not a threshold. https://writingcenter.unc.edu/tips-and-tools/revising-drafts/ 3. The emotional version of the rule goes back to Abraham Lincoln's "hot letters". He wrote the angry letter, put it aside until he had calmed down, and filed it as "Never se [source]
- **Interval length.** Sources prescribe nine years (Horace), "an interval" (Quintilian), six weeks (King), two weeks (the only experimentally tested interval, Daneman & Stainton), overnight (RFC 1855), 24 hours (Peterson), or "a few hours" (UNC). None of them derives its number from data. Positions are listed side by side, not averaged. - **Does own-text familiarity actually hurt revision?** Daneman & Stainton (1993) say yes. Pilotti & Chodorow (2009) found the opposite, and Burgoyne et al. (2022) could not replicate the effect. Unresolved. - **Is it the delay or the activity during the delay [source]
Limits: - The LOC, Perseus preface, and U. Michigan *Collected Works* pages returned 403 errors. The Lincoln endorsement text is therefore verified through the Dickinson House Divided edition, not the LOC scan. - The Stephen King claim relies on a secondary summary. - The Bushman 2005 8-hour detail was not checked against the full text. - No source earlier than RFC 1855 (1995) was found for "overnight" as applied to email, and none earlier than Peterson (2011) for "24 hours" by name. Earlier uses may exist in print etiquette guides that this run did not reach. [source]
16. Sio & Ormerod's meta-analysis (Psychological Bulletin, 2009) found a positive incubation effect: setting a problem aside before trying again helps. Divergent-thinking tasks benefit more than insight tasks. https://eric.ed.gov/?id=EJ827110 17. In that meta-analysis, longer preparation before the break increased the effect. Filling the break with high-demand tasks reduced it. For the rule, this means the effect depends on how much work went into the draft and on what the writer does during the delay. https://eric.ed.gov/?id=EJ827110 18. The mean incubation effect size is small to medium (d ≈ [source]
27. In Sommers (1980), student writers treated revision mostly as rewording. Experienced writers revised to find the overall shape of the argument, and imagined a reader while doing it. https://eric.ed.gov/?id=EJ240356 28. Implication: the gap gives a writer fresh eyes, but it does not give them a revision strategy. A writer who only rewords will still only reword after 24 hours. This is an inference from Sommers; it was not tested directly. https://files.eric.ed.gov/fulltext/ED220839.pdf 29. Research on the "unconscious thought advantage" is a close cousin of the incubation claim. It holds th [source]

Related concepts

24-hour — is a part of 24-hour cooling-off rule
rule — is a part of 24-hour cooling-off rule
cooling-off — is a part of 24-hour cooling-off rule

Children

No children recorded.

← the whole tree · 3D view· how to read this page