Questions Are Control: Notes and References

Notes

1. Mats Ekström and Melisa Stevanovic, “Conversation Analysis and Power: Examining the Descendants and Antecedents of Social Action,” Frontiers in Sociology 8 (2023), article 1196672, doi:10.3389/fsoc.2023.1196672. The article connects control over conversational agendas and initiating actions with constraints on what others can do next. ↩

2. John Heritage, “Epistemics in Action: Action Formation and Territories of Knowledge,” Research on Language and Social Interaction 45, no.1 (2012): 1–29, doi:10.1080/08351813.2012.646684. Public author copy, inspected September 10, 2026. ↩

3. Lisa M. Miles et al., “Asking Questions Changes Health-Related Behavior: An Updated Systematic Review and Meta-Analysis,” Journal of Clinical Epidemiology 123 (2020): 59–68, doi:10.1016/j.jclinepi.2020.03.014. Public full text. ↩

4. Elizabeth F. Loftus and John C. Palmer, “Reconstruction of Automobile Destruction: An Example of the Interaction Between Language and Memory,” Journal of Verbal Learning and Verbal Behavior 13, no. 5 (1974): 585–589, doi:10.1016/S0022-5371(74)80011-3. Public full text. ↩

5. Tatiana Tkačuková, “The Power of Questioning: A Case Study of Courtroom Discourse,” Discourse and Interaction 3, no. 2 (2010): 49–61, public full text. ↩

6. William B. Swann Jr., Blair Stephenson, and Thane S. Pittman, “Curiosity and Control: On the Determinants of the Search for Social Knowledge,” Journal of Personality and Social Psychology 40, no. 4 (1981): 635–642, doi:10.1037/0022-3514.40.4.635. Public author copy. ↩

7. James R. Detert and Amy C. Edmondson, “Implicit Voice Theories: Taken-for-Granted Rules of Self-Censorship at Work,” Academy of Management Journal 54, no. 3 (2011): 461–488, doi:10.5465/amj.2011.61967925. Public full-text record. ↩

8. Aled Jones et al., “Interventions Promoting Employee ‘Speaking-Up’ within Healthcare Workplaces: A Systematic Narrative Review of the International Literature,” Health Policy 125, no. 3 (2021): 375–384, doi:10.1016/j.healthpol.2020.12.016. Public repository record and full text. ↩

9. U.S. Government Accountability Office, Protections for Whistleblowers and Others: Selected Agency Actions Regarding Reports of Potential Wrongdoing, GAO-26-107650 (March 2026), public report and report highlights. ↩

10. Karen Huang et al., “It Doesn’t Hurt to Ask: Question-Asking Increases Liking,” Journal of Personality and Social Psychology 113, no. 3 (2017): 430–452, doi:10.1037/pspi0000097. Public author copy. The observational speed-dating analysis has a critique, an official correction to that critique, and an author reply, documented in Appendix C; it is distinct from the conversational experiments. ↩

11. Alison Wood Brooks, Francesca Gino, and Maurice E. Schweitzer, “Smart People Ask for (My) Advice: Seeking Advice Boosts Perceptions of Competence,” Management Science 61, no. 6 (2015): 1421–1435, doi:10.1287/mnsc.2014.2054. Public author copy. Author-reported data availability limits reproduction; Appendix C distinguishes that limitation from demonstrated unreliability. ↩

12. Arthur Aron et al., “The Experimental Generation of Interpersonal Closeness: A Procedure and Some Preliminary Findings,” Personality and Social Psychology Bulletin 23, no. 4 (1997): 363–377, doi:10.1177/0146167297234003. Public full text. ↩

13. Jean-Philippe Laurenceau, Lisa Feldman Barrett, and Paula R. Pietromonaco, “Intimacy as an Interpersonal Process: The Importance of Self-Disclosure, Partner Disclosure, and Perceived Partner Responsiveness in Interpersonal Exchanges,” Journal of Personality and Social Psychology 74, no. 5 (1998): 1238–1251, doi:10.1037/0022-3514.74.5.1238. Public author copy. ↩

14. Michael Kardas, Amit Kumar, and Nicholas Epley, “Overly Shallow? Miscalibrated Expectations Create a Barrier to Deeper Conversation,” Journal of Personality and Social Psychology 122, no. 3 (2022): 367–398, doi:10.1037/pspa0000281. Public author copy. ↩

15. Guy Itzchakov, Harry T. Reis, and Netta Weinstein, “How to Foster Perceived Partner Responsiveness: High-Quality Listening Is Key,” Social and Personality Psychology Compass 16, no. 1 (2022): e12648, doi:10.1111/spc3.12648. Public author copy. ↩

16. Paul Schrodt, Paul L. Witt, and Jenna R. Shimkowski, “A Meta-Analytical Review of the Demand/Withdraw Pattern of Interaction and Its Associations with Individual, Relational, and Communicative Outcomes,” Communication Monographs 81, no. 1 (2014): 28–58, doi:10.1080/03637751.2013.813632. Public author-uploaded full text. ↩

17. Christina Nicolaidis et al., “‘Respect the Way I Need to Communicate with You’: Healthcare Experiences of Adults on the Autism Spectrum,” Autism 19, no. 7 (2015): 824–831, doi:10.1177/1362361315576221. Public author manuscript. ↩

18. Katie Maras et al., “The Witness-Aimed First Account (WAFA): A New Technique for Interviewing Autistic Witnesses and Victims,” Autism 24, no. 6 (2020): 1449–1467, doi:10.1177/1362361320908986. Public full text. ↩

19. Catherine J. Crompton et al., “Autistic Peer-to-Peer Information Transfer Is Highly Effective,” Autism 24, no. 7 (2020): 1704–1712, doi:10.1177/1362361320919286. Public full text. ↩

20. Catherine J. Crompton et al., “Neurotype-Matching, but Not Being Autistic, Influences Self and Observer Ratings of Interpersonal Rapport,” Frontiers in Psychology 11 (2020): 586171, doi:10.3389/fpsyg.2020.586171. Public full text. ↩

21. Danielle M. Beaton, Fuschia Sirois, and Elizabeth Milne, “Experiences of Criticism in Adults with ADHD: A Qualitative Study,” PLOS ONE 17, no. 2 (2022): e0263366, doi:10.1371/journal.pone.0263366. Public full text. ↩

22. Katie L. Maras and Dermot M. Bowler, “Brief Report: Suggestibility, Compliance and Psychological Traits in High-Functioning Adults with Autism Spectrum Disorder,” Research in Autism Spectrum Disorders 6, no.3 (2012): 1168–1175, doi:10.1016/j.rasd.2012.03.013. Public accepted manuscript, Methods, Results, and Table 2. The historical title is the authors’ terminology; the essay does not use it as a classification of people. ↩

23. Callie M. Ginapp et al., “The Experiences of Adults with ADHD in Interpersonal Relationships and Online Communities: A Qualitative Study,” SSM—Qualitative Research in Health 3 (2023): 100223, doi:10.1016/j.ssmqr.2023.100223. Public full text. ↩

24. Russell Golman, David Hagmann, and George Loewenstein, “Information Avoidance,” Journal of Economic Literature 55, no. 1 (2017): 96–135, doi:10.1257/jel.20151245. Public author copy. ↩

25. Linh Vu et al., “Ignorance by Choice: A Meta-Analytic Review of the Underlying Motives of Willful Ignorance and Its Consequences,” Psychological Bulletin 149, nos. 9–10 (2023): 611–635, doi:10.1037/bul0000398. Public author copy. ↩

26. Stephanie Beckwith et al., Coercive Control Literature Review—Final Report (Melbourne: Australian Institute of Family Studies, 2023), public report. ↩

27. Vahid Malekpour et al., “The Experience of the Coercive Control Scale: Factor Structure and Psychometric Properties,” Iranian Journal of Psychiatry and Behavioral Sciences 17, no. 1 (2023): e119016, doi:10.5812/ijpbs-119016. Public full text. ↩

Appendix A: Scope, Method, and Claim Boundaries

Working research question

Is there an empirical and theoretical basis for the claim that questions are a form of control and that seeking knowledge produces consequences?

Scope

This essay examines eight overlapping literatures:

• conversation analysis and institutional questioning;

• question wording, memory, and measurement reactivity;

• control motivation, curiosity, and information seeking;

• psychological safety, employee voice, and self-censorship;

• question asking, reciprocal disclosure, responsiveness, and interpersonal closeness;

• high-quality listening and demand–withdraw conflict patterns in close relationships;

• autistic communication, neurotype matching, accessible interviewing, and attention-deficit/hyperactivity disorder (ADHD) lived experience;

• information avoidance, moral consequences of knowledge, and coercive monitoring.

The review prioritizes peer-reviewed research, meta-analyses, systematic or narrative reviews, official research reports, exact digital object identifier (DOI) records, and publicly inspectable full text.

Claim boundaries

• Question includes information-seeking questions, confirmation requests, challenges, advice seeking, and some question-shaped assertions. These are not interchangeable.

• Control is separated into power-over—constraining another person’s options—and power-to—the capacity to understand or act.

• Consequences includes interactional obligations, social evaluation, institutional response, changes in agency, and moral demands that follow from knowledge.

• Natural consequences is treated as a hypothesis under examination. Most penalties for asking are socially produced, not inevitable features of inquiry.

• Rarely neutral means rarely interactionally inert. It does not mean every question has a positive or negative moral valence.

• Workplace voice research is relevant but does not directly measure every kind of question asking.

• Genuine questioning is evaluated through voluntary participation, responsiveness, reciprocity, premise correction, and whether the answer can alter the asker. Sincerity alone is insufficient.

• Neurodivergent is not treated as a single comparative population. Autism and ADHD findings are separated, within-group variation is assumed, and missing evidence for other neurodivergent groups is recorded.

• Neurotype-matching studies concern rapport and information transfer in particular short interactions. They do not establish that matched-neurotype relationships are always better.

• Coercive-control research applies to patterns of domination. It must not be used to label an isolated, ordinary question as abuse.

• Demand–withdraw research concerns a reciprocal conflict pattern involving pressure and retreat. It does not establish that question asking is demand or that privacy, processing time, or a bounded refusal is withdrawal.

• Private conversation records are used only to recover and test first-person observations. They are selection-biased, not publicly accessible, and cannot support population claims or conclusions about another person’s motives.

• Names and identifying details in personal passages are changed or omitted. The essay reports the author’s behavior and experience rather than diagnosing the other people involved.

Verification labels

• Verified: The full source or a sufficiently complete official/public research report was directly inspected.

• Limited Verification: Abstract Only: The publisher’s abstract and preview were inspected, but a public full text was not located.

• Limited Verification: Publisher Preview: A detailed publisher preview and/or institutional record was inspected, but a complete public text was not located.

• Context Only: A source helps define or locate the issue but does not directly establish the essay’s central claim.

Appendix B: Claim Audit

Claim 1: Questions can create normative pressure to answer

Status: Supported.

Conversation-analysis scholarship describes initiating actions as making particular responses relevant and making an absent response accountable. This is social pressure, not literal compulsion.

Claim 2: Questions are inherently domination

Status: Rejected.

The same interaction literature distinguishes power-over from power-to and repeatedly shows that form alone is not decisive. Questions can support another person’s goal, express care, or expand agency.

Claim 3: Control deprivation can increase information seeking

Status: Suggestive, narrow evidence.

Swann, Stephenson, and Pittman experimentally observed more diagnostic information seeking after a control-deprivation manipulation. The sample was small, demographically narrow, artificial, and old. The study supports a mechanism, not a population law.

Claim 4: People are punished for asking questions

Status: Context-dependent.

Inspected organizational research supports anticipated costs of speaking up, and the GAO report documents reprisals against some whistleblowers. These are not measured penalties for every question. Inspected question-asking experiments also find benefits in liking and perceived competence within particular settings. Neither penalty nor reward is universal; the leader-humility source remains an unverified research lead.

Claim 5: Seeking knowledge has unavoidable moral consequences

Status: Partially supported and philosophically extended.

Information-avoidance research shows that people sometimes avoid facts that could threaten preferred choices or self-image. The further claim that knowledge creates an obligation to act is normative, not established by the behavioral data alone.

Claim 6: Demanding information can become coercive control

Status: Supported at the pattern level.

Coercive-control research includes monitoring, surveillance, and insistence on access to a partner’s activities or whereabouts. The evidence supports analysis of repeated patterns, retaliation, and erosion of autonomy—not diagnosis from an isolated question.

Claim 7: The consequences of asking are natural

Status: Mostly rejected.

Some informational consequences follow directly from learning: uncertainty changes, contradictions become visible, and some beliefs become harder to maintain. Social penalties and rewards are contingent responses produced by relationships and institutions.

Claim 8: Questions are rarely neutral

Status: Supported only after narrowing.

Conversation analysis supports the claim that questions alter sequential expectations. Question–behavior research finds a small average effect in a defined health-behavior literature, and eyewitness experiments show that wording can change immediate answers and later reports. These findings support “rarely interactionally inert,” not “always morally positive or negative.”

Claim 9: Genuine questions can deepen connection

Status: Supported with conditions.

Question asking, escalating reciprocal disclosure, perceived responsiveness, and deeper conversation are associated with liking, temporary closeness, intimacy, and greater-than-expected connection. The evidence does not support depth through one-way extraction. Voluntariness, reciprocity, and experienced responsiveness are central boundary conditions.

Claim 10: Neurotypical and neurodivergent people experience questions differently

Status: Plausible but too broad as stated.

Autistic adults report question-relevant communication needs, including processing time, accessible language, written options, and difficulty with vague prompts. Experimental work shows benefits from bounded non-leading recall structure and lower rapport or information transfer in some mixed-neurotype interactions. The fully inspected ADHD studies document participants’ histories of criticism and experiences in relationships and communities. The formal-questioning comparison remains preview-only and supplies no independently verified result here. There is no basis for one universal neurotypical-versus-neurodivergent effect.

Claim 11: Open questions are inherently more autonomy-preserving

Status: Rejected.

Open prompts can preserve narrative control but impose high retrieval and relevance-selection demands. Closed or leading questions can overconstrain the answer. Evidence from Witness-Aimed First Account interviews supports a third option: the answerer helps define bounded topics and retains free recall inside them.

Claim 12: Neurodivergent adults are generally more suggestible

Status: Rejected.

Maras and Bowler (2012), added to the source library in September, found no significant group difference in the measured suggestibility or compliance outcomes in a small autistic comparison study. The retained ADHD comparison remains preview-only and is not used to establish an independently verified null. Specific vulnerabilities must be measured rather than presumed; neither finding justifies a universal claim or an equivalence conclusion.

Claim 13: People who do not ask questions care less

Status: Rejected.

Question asking, particularly responsive follow-up, can communicate interest and care. High-quality listening research identifies several other possible signals, including reflection, paraphrasing, attention, nonverbal response, and behavior that demonstrates understanding. Persistent non-engagement can still be experienced as unresponsiveness, but question absence alone does not establish motive or care.

Claim 14: Demand–withdraw is a question-asking pattern

Status: Rejected.

Demand–withdraw research measures a broader reciprocal conflict pattern: pressure through demands, criticism, or complaints paired with defensiveness, silence, avoidance, or inaction. Questions may occur within the pattern, but the evidence does not classify questioners as demanders or answer refusal as withdrawal. Its relevance is interactional: each person’s control-seeking response can intensify the other’s.

Appendix C: Source Library

Ekström and Stevanovic (2023)

Verification: Verified.

Type: Peer-reviewed hypothesis-and-theory article synthesizing conversation-analysis research on power.

Contribution: Connects agenda control, sequential constraints, accountability for nonresponse, authority, power-over, and power-to.

Limit: The article develops theory from conversation-analysis literature; it is not an experiment estimating the frequency or size of question effects.

Access: DOI record and public full text.

September 2026 inspection: PDF pp. 3–5, especially Capacities to act in social interaction. Relevant full-source body inspected; no raw-data replication claimed.

Heritage (2012)

Verification: Verified.

Type: Peer-reviewed conversation-analysis article.

Contribution: Shows how relative knowledge, rights to know, and epistemic status shape whether an utterance functions as a request, assertion, challenge, or confirmation.

Limit: The evidence is qualitative and interactional; it does not classify questions as ethical or unethical.

Access: DOI record and the 29-page public author copy, recovered through the UCLA author index and inspected September 10, 2026. The August access failure is historical.

September 2026 inspection: pp. 4–6 on epistemic status, rights and territories; pp. 16–25 on mismatches between form and action. Relevant full-source body inspected; no raw-data replication claimed.

Tkačuková (2010)

Verification: Verified.

Type: Peer-reviewed linguistic case study of courtroom discourse.

Contribution: Demonstrates how institutional questioners control topic, turn type, response length, and the narrative made available to an audience.

Limit: Courtroom cross-examination is a highly asymmetrical setting and cannot stand in for ordinary conversation.

Access: Public full text.

September 2026 inspection: pp. 49–51, 54–60, especially section 7 Conclusions. Relevant full-source body inspected; no raw-data replication claimed.

Swann, Stephenson, and Pittman (1981)

Verification: Verified.

Type: Laboratory experiment.

Sample: 61 female undergraduates at the University of Texas at Austin.

Contribution: Participants exposed to uncontrollable outcomes later selected more diagnostic social information and diagnostic questions.

Limit: Small, narrow, dated sample and artificial manipulation; the design does not establish a broad causal law about everyday questioning.

Access: DOI record and public author copy.

September 2026 inspection: pp. 637–639, Method and Results; experimental procedures and diagnosticity measures. Relevant full-source body inspected; no raw-data replication claimed.

Detert and Edmondson (2011)

Verification: Verified.

Type: Four-study organizational research program using interviews and surveys.

Contribution: Identifies learned beliefs that make speaking up to authority feel risky or inappropriate and links them to workplace silence.

Limit: Voice includes suggestions, concerns, challenges, and other speech acts; it is not a direct measure of question asking alone. Later studies used modest samples.

Access: DOI record and public full-text record.

September 2026 inspection: Full author-uploaded article: Studies 1–4; Study 4 methods/results, Table 4, pp. 477–481. Relevant full-source body inspected; no raw-data replication claimed.

Frazier et al. (2017)

Verification: Limited as of September 10, 2026; full methods/results were not inspected. Retained as a research lead, excluded from the narrative’s empirical support.

Type: Meta-analysis.

Coverage: 136 independent samples, more than 22,000 individuals, and nearly 5,000 groups.

Abstract/preview report, not independently verified evidence: Establishes a broad empirical relationship between psychological safety and learning, information sharing, task performance, and related outcomes.

Limit: Most included research is correlational, measures vary, and psychological safety is broader than the safety of asking a particular question.

Access: DOI record, public full text, and institutional repository record.

Citation retained as research lead: M. Lance Frazier et al., “Psychological Safety: A Meta-Analytic Review and Extension,” Personnel Psychology 70, no. 1 (2017): 113–165, doi:10.1111/peps.12183. Public full text and institutional repository record.

September 2026 inspection: Publisher abstract and institutional repository metadata only. Only the visible material is corroborated; historical full-text access is not presumed.

Jones et al. (2021)

Verification: Verified.

Type: Systematic narrative review of workplace speaking-up interventions in healthcare.

Coverage: 34 included studies.

Contribution: Finds recurring importance of hierarchy, interdisciplinary and cultural relationships, and psychological safety.

Limit: Most intervention findings were inconclusive; healthcare workplaces have distinctive stakes and professional structures.

Access: DOI record and public repository record and full text.

September 2026 inspection: Publisher pp. 377–382: section 3, intervention synthesis, section 4.3. Relevant full-source body inspected; no raw-data replication claimed.

Huang et al. (2017)

Verification: Verified.

Type: Three reported study settings using live dyadic conversation, experiments, and an observational speed-dating field dataset.

Contribution: More questions, particularly follow-up questions, were associated with greater liking through perceived responsiveness.

Limit: The work focuses on get-to-know-you interactions and does not test coercive, accusatory, or institutionally compelled questions.

Access: DOI record and public author copy.

September 2026 inspection: Studies 1–3, especially Study 2B and Study 3; 2019 author reply pp. 1139–1142. Relevant full-source body inspected; no raw-data replication claimed.

Methodological exchange: The observational speed-dating analysis was debated by Kluger and Malloy (2019) and Yeomans et al. (2019). The critique itself was corrected after confusing second dates offered with second dates received; the correction affected tables, a figure, and conclusions. The corrected critique, official correction, and author reply must be read together. This exchange does not retract the original paper or erase its separate experimental evidence. See the additional source records below.

Brooks, Gino, and Schweitzer (2015)

Verification: Verified.

Type: Multi-study experimental research on advice seeking.

Contribution: Advice seekers were often seen as more competent, especially for difficult tasks and when asking someone with relevant expertise.

Limit: Advice seeking is a cooperative and relatively narrow type of question. The effect can reverse when the chosen advisor is clearly not an expert.

Access: DOI record and public author copy.

September 2026 inspection: pp. 1424–1432; Studies 1–5, especially Study 5 sections 10.2–10.3. Relevant full-source body inspected; no raw-data replication claimed.

Reproducibility limitation: The author-completed Many Coauthors study-68 record reports gaps in original/raw or reproduction data. This limits independent reproduction; it is not a finding of fraud, invalidity, or retraction. The reported expert/nonexpert boundary was checked in the full article. See the provenance record below.

Cojuharenco and Karelaia (2020)

Verification: Limited as of September 10, 2026; full methods/results were not inspected. Retained as a research lead, excluded from the narrative’s empirical support.

Type: Four experiments plus a study of management practitioners’ beliefs.

Abstract/preview report, not independently verified evidence: Finds context-dependent competence penalties and more pervasive humility premiums when leaders ask genuine questions, with positive downstream effects on trust and helping.

Limit: The studies focus on leaders, scenario-based judgments, and humble information-seeking questions; strategic or manipulative questions may operate differently.

Access: DOI record and public author copy.

Citation retained as research lead: Irina Cojuharenco and Natalia Karelaia, “When Leaders Ask Questions: Can Humility Premiums Buffer the Effects of Competence Penalties?” Organizational Behavior and Human Decision Processes 156 (2020): 113–134, doi:10.1016/j.obhdp.2019.12.001. Public author copy.

September 2026 inspection: Publisher abstract, extended preview and conclusion; author PDF identified but body retrieval failed. Only the visible material is corroborated; historical full-text access is not presumed.

Hart, VanEpps, and Schweitzer (2021)

Verification: Limited as of September 10, 2026; full methods/results were not inspected. Retained as a research lead, excluded from the narrative’s empirical support.

Type: Five experiments.

Coverage: 1,427 participants.

Abstract/preview report, not independently verified evidence: Askers overestimated the discomfort and negative impressions produced by sensitive questions across several conversational settings.

Limit: Full text was not publicly located for this review. The findings do not license questioning where refusal is unsafe or privacy is being overridden.

Access: Publisher abstract and preview and DOI record.

Citation retained as research lead: Einav Hart, Eric M. VanEpps, and Maurice E. Schweitzer, “The (Better Than Expected) Consequences of Asking Sensitive Questions,” Organizational Behavior and Human Decision Processes 162 (2021): 136–154, doi:10.1016/j.obhdp.2020.10.014. Limited verification: the publisher’s detailed abstract and preview were inspected; a public full text was not located.

September 2026 inspection: Final publisher abstract/preview; an earlier 2019 four-study preprint also identified. Only the visible material is corroborated; historical full-text access is not presumed.

Golman, Hagmann, and Loewenstein (2017)

Verification: Verified.

Type: Interdisciplinary review in economics and psychology.

Contribution: Synthesizes evidence that people actively avoid available information for hedonic, identity-protective, attentional, and strategic reasons.

Limit: The review spans heterogeneous domains and mechanisms. Avoidance is not always irrational or unethical.

Access: DOI record and public author copy.

September 2026 inspection: pp. 96–103 defining active avoidance; pp. 110–118 on motives, identity, dissonance and moral wiggle room; pp. 123–129 on implications. Relevant full-source body inspected; no raw-data replication claimed.

Vu et al. (2023)

Verification: Verified.

Type: Meta-analysis of a specific experimental paradigm involving willful ignorance about effects on others.

Coverage: 33,603 decisions, 6,531 participants, and 56 treatment effects.

Contribution: Estimates that 40 percent avoided available consequence information and that ignorance availability reduced altruistic choice by 15.6 percentage points compared with provided-information conditions.

Limit: The treatment effects were heterogeneous, and the paradigm does not represent all reasons for declining information.

Access: DOI record and public author copy.

September 2026 inspection: pp. 619–625, especially main results p. 622 and Table 4 p. 624. Relevant full-source body inspected; no raw-data replication claimed.

Beckwith et al. (2023)

Verification: Verified.

Type: Australian government research institute literature review.

Contribution: Synthesizes coercive control as a cumulative pattern involving reduced autonomy, monitoring, surveillance, intimidation, isolation, and other tactics.

Limit: The search prioritized recent Australian domestic-and-family-violence research and supplemented gaps with international and grey literature. The report explicitly notes measurement and prevalence limitations.

Access: Public report.

September 2026 inspection: Full AIFS report: Executive Summary/Definitions, Review methodology, tactics and technology-facilitated abuse sections. Relevant full-source body inspected; no raw-data replication claimed.

Malekpour et al. (2023)

Verification: Verified.

Type: Cross-sectional psychometric study.

Contribution: Identifies aggression and spying/monitoring as two components in the tested coercive-control scale and helps distinguish mutual information sharing from constant insistence.

Limit: Convenience sample, female-heavy participation, one national context, and author-added items limit generalizability.

Access: DOI record and public full text.

September 2026 inspection: Methods, Results/Table 1 and Discussion of scale components and whereabouts. Relevant full-source body inspected; no raw-data replication claimed.

Source consistency: The original paper contains inconsistent tabulated denominators. Its conceptual/scale contribution is retained, but those totals are not silently reconciled or used as population estimates.

Miles et al. (2020)

Verification: Verified.

Type: Updated systematic review and meta-analysis of randomized trials.

Coverage: 43 measurement-versus-no-measurement studies with about 104,000 participants. The abstract gives 104,388, while Results 3.1 gives 104,096 for that comparison set; the exact count remains unresolved.

Contribution: Finds a small average question–behavior effect for health-related outcomes, showing that measurement questions can sometimes alter the behavior they appear only to record.

Limit: The reported standardized mean difference is .06, 95% CI [.02,.09], with substantial heterogeneity and evidence of publication bias. The conflicting participant totals are preserved rather than silently reconciled. The result concerns defined health-behavior measurement contexts, not all questions.

Access: DOI record and public full text.

September 2026 inspection: PMC abstract; Results sections 3.1 and main meta-analysis; forest plot and publication-bias discussion. Relevant full-source body inspected; no raw-data replication claimed.

Loftus and Palmer (1974)

Verification: Verified.

Type: Two laboratory experiments on question wording and eyewitness reports.

Coverage: 45 students in Experiment 1 and 150 in Experiment 2.

Contribution: Demonstrates that changing a verb in a speed question changed estimates and was associated one week later with reports of broken glass that had not appeared.

Limit: Old student samples, artificial film-viewing tasks, uncertain numerical judgments, and a narrow memory context. It does not show that questions can implant arbitrary beliefs.

Access: DOI record and public full text.

September 2026 inspection: pp. 585–587: Experiment 1/Table 1 and Experiment 2/Table 2. Relevant full-source body inspected; no raw-data replication claimed.

Burris (2012)

Verification: Limited as of September 10, 2026; full methods/results were not inspected. Retained as a research lead, excluded from the narrative’s empirical support.

Type: One field study and two experiments on challenging versus supportive employee voice.

Abstract/preview report, not independently verified evidence: Reports that managers judged challenging voicers as worse performers and endorsed their ideas less, with threat and loyalty perceptions mediating responses.

Limit: Full text was not publicly located. Employee voice is broader than questioning, and managerial judgments in workplace studies do not estimate retaliation in the general population.

Access: Publisher abstract and DOI record.

Citation retained as research lead: Ethan R. Burris, “The Risks and Rewards of Speaking Up: Managerial Responses to Employee Voice,” Academy of Management Journal 55, no. 4 (2012): 851–875, doi:10.5465/amj.2010.0562. Limited verification: the publisher’s abstract and study summary were inspected; a public full text was not located.

September 2026 inspection: Publisher abstract only after author/repository searches. Only the visible material is corroborated; historical full-text access is not presumed.

Morrison (2023)

Verification: Limited as of September 10, 2026; full methods/results were not inspected. Retained as a research lead, excluded from the narrative’s empirical support.

Type: Integrative review of employee voice and silence research published during the decade after a 2014 review.

Coverage: The prior source record reports 158 selected papers spanning 2014–2021, but that count and screen were not visible in the abstract and could not be confirmed from full text in September. Treat them as historical unverified extraction, not an established audit result.

Abstract/preview report, not independently verified evidence: Frames upward voice as discretionary, improvement-oriented communication that can benefit organizations while challenging higher-status actors or the status quo.

Limit: The organizational-behavior literature uses varied measures, is frequently correlational, and does not isolate grammatical questions as a speech act.

Access: DOI record and public full text.

Citation retained as research lead: Elizabeth Wolfe Morrison, “Employee Voice and Silence: Taking Stock a Decade Later,” Annual Review of Organizational Psychology and Organizational Behavior 10 (2023): 79–107, doi:10.1146/annurev-orgpsych-120920-054654. Public full text.

September 2026 inspection: Publisher abstract and references; full-text panel failed to load. Only the visible material is corroborated; historical full-text access is not presumed.

U.S. Government Accountability Office (2026)

Verification: Verified.

Type: Official report based on documents and testimonial information from selected federal agencies and organizations.

Contribution: Documents the public value of disclosures, employer-reprisal risk, uncertainty around nondisclosure agreements, and mechanisms used to protect or incentivize people who report potential wrongdoing.

Limit: This is a selected-agency policy report, not a prevalence survey and not research on ordinary interpersonal questioning.

Access: Public report and report highlights.

September 2026 inspection: GAO-26-107650 full report, What GAO Found; agency protection actions; stakeholder NDA issues; Scope and Methodology. Relevant full-source body inspected; no raw-data replication claimed.

Aron et al. (1997)

Verification: Verified.

Type: Three laboratory studies developing a closeness-generating procedure.

Contribution: Study 1 found that gradual, reciprocal, personal self-disclosure produced greater temporary closeness than structured small talk. The procedure supplies a model of questions as supportive scaffolding.

Limit: College classroom samples and a temporary subjective outcome. The method was not intended to create an ongoing relationship, and later manipulations of attitude matching, expected liking, and an explicit closeness goal did not add significant effects.

Access: DOI record and public full text.

September 2026 inspection: pp. 365–367: Study 1 method and results; General Discussion. Relevant full-source body inspected; no raw-data replication claimed.

Laurenceau, Feldman Barrett, and Pietromonaco (1998)

Verification: Verified.

Type: Two event-contingent interaction-diary studies conducted across one or two weeks.

Contribution: Supports intimacy as a process involving self-disclosure, partner disclosure, and perceived partner responsiveness; emotional disclosure had a stronger relationship with intimacy than factual disclosure.

Limit: Observational diary associations do not establish that asking deeper questions causes intimacy, and the samples were college students.

Access: DOI record and public author copy.

September 2026 inspection: Studies 1 and 2 Methods/Results; disclosure and perceived responsiveness models. Relevant full-source body inspected; no raw-data replication claimed.

Kardas, Kumar, and Epley (2022)

Verification: Verified.

Type: Multi-experiment research program comparing expected with experienced deep conversation.

Contribution: Finds that strangers’ conversations were less awkward and more connecting than expected; participants especially overpredicted the relative awkwardness of deep conversation and underestimated others’ care.

Limit: The studies used bounded, consented conversations and operationalized depth through personal disclosure. They do not establish safety in coercive, dependent, or high-stakes relationships.

Access: DOI record and public author copy.

September 2026 inspection: Overview; Experiments 1a–7b; Experiment 5 pp. 380–384 and General Discussion. Relevant full-source body inspected; no raw-data replication claimed.

Itzchakov, Reis, and Weinstein (2022)

Verification: Verified.

Type: Peer-reviewed integrative review and theoretical model linking high-quality listening with perceived partner responsiveness.

Contribution: Identifies follow-up questions as one of several observable listening behaviors that can communicate attention, understanding, validation, and care. The model also includes reflection, paraphrasing, nonverbal attention, and later reciprocal listening, preventing question frequency from becoming a complete measure of responsiveness.

Limit: The article synthesizes and links two research traditions; it does not experimentally compare “questioners” with “non-questioners” as stable conversational types. Some proposed pathways remain research questions.

Access: DOI record and public author copy.

September 2026 inspection: Sections 2–4 on perceived responsiveness/listening behaviors; discussion and future research. Relevant full-source body inspected; no raw-data replication claimed.

Schrodt, Witt, and Shimkowski (2014)

Verification: Verified.

Type: Meta-analysis of demand–withdraw interaction research.

Coverage: 74 studies representing 14,255 participants.

Contribution: Finds a moderate overall association between demand–withdraw behavior and the outcomes examined (r = .360), with larger average associations for relational (r = .423) and communicative (r = .418) outcomes. Similar overall associations appeared regardless of which gender occupied the demand or withdraw role. The review foregrounds the interaction pattern rather than assigning the problem to a fixed kind of person.

Limit: Most included relationships were heterosexual marriages, measures and outcome categories varied, most associations were reciprocal or correlational rather than causal, and some studies analyzed individuals rather than dyads. Demand includes criticism and complaints as well as pressure for discussion; it is not a synonym for asking questions.

Access: DOI record and public author-uploaded full text.

September 2026 inspection: Full author-uploaded article: Methods; results/Table 3; Discussion, pp. 42–49. Relevant full-source body inspected; no raw-data replication claimed.

Nicolaidis et al. (2015)

Verification: Verified.

Type: Community-based participatory qualitative research.

Coverage: Semi-structured interviews with 39 autistic adults and 16 supporters.

Contribution: Documents question-relevant healthcare needs including processing time, written communication, accessible language, avoidance of vague or overly open prompts, and providers’ willingness to accommodate individual communication.

Limit: Healthcare is high-stakes and institutionally asymmetric. Participants were adults able to participate in an interview study, and their recommendations should not be treated as a universal autistic profile.

Access: DOI record and public author manuscript.

September 2026 inspection: PMC Methods/recruitment and accommodations; Tables 1–2; Results/Recommendations; Discussion limitations. Relevant full-source body inspected; no raw-data replication claimed.

Maras et al. (2020)

Verification: Verified.

Type: Experimental mock-witness study of an accessible, non-leading interview procedure.

Coverage: 33 autistic and 30 allistic adults.

Contribution: The Witness-Aimed First Account procedure let witnesses generate bounded event topics and then recall freely within them, improving detail and accuracy in both groups.

Limit: Participants recalled videos under research conditions rather than experiencing real police pressure. Benefits to both groups do not erase possible group differences in other interview tasks.

Access: DOI record and public full text.

September 2026 inspection: Methods/Participants and Design; Results Table 2; Discussion. Relevant full-source body inspected; no raw-data replication claimed.

Crompton et al. (2020), information transfer

Verification: Verified.

Type: Between-groups diffusion-chain experiment.

Coverage: 72 adults allocated across autistic, allistic, and mixed-neurotype chains.

Contribution: Detail retention declined more steeply and rapport was lower in mixed chains, while autistic and allistic matched chains did not significantly differ in detail retention.

Limit: A story-retelling chain is not a full model of everyday conversation, and the sample was modest. The study supports a mismatch account in this task, not neurotype separatism.

Access: DOI record and public full text.

September 2026 inspection: Methods pp. 1706–1708; Tables 1–2 and Results. Relevant full-source body inspected; no raw-data replication claimed.

Sample independence: The same interacting participants are analyzed in the companion rapport paper. No detected autistic/allistic difference is not a formal equivalence finding.

Crompton et al. (2020), rapport

Verification: Verified.

Type: Two studies of self-rated and observer-rated rapport.

Coverage: Study 1 uses the same 72 adults and interaction procedure as the information-transfer paper. Study 2 adds 80 autistic/allistic observers who rate clips of 18 of those interacting participants; it is not an independent replication of the interaction sample.

Contribution: Autistic pairs rated rapport higher than mixed pairs, and both autistic and allistic observers rated mixed interactions lower than matched interactions.

Limit: Allistic pairs reported the highest self-rated rapport, interactions were brief and semi-structured, and matching does not determine the quality of any individual relationship.

Access: DOI record and public full text.

September 2026 inspection: Study 1 Participants/Materials pp. 3–5; Study 2 pp. 6–8; Tables 1–3. Relevant full-source body inspected; no raw-data replication claimed.

Beaton, Sirois, and Milne (2022)

Verification: Verified.

Type: Qualitative thematic analysis of open-text responses.

Coverage: 162 adults: 96 self-reporting an ADHD diagnosis and 66 with high ADHD traits on a screening measure.

Contribution: Participants reported recurring criticism of inattention, organization, and impulsive social behavior, with perceived effects on self-worth and wellbeing.

Limit: Self-selected online sample, many participants identified through traits rather than a verified diagnosis, no non-ADHD comparison group, and no direct experiment on question framing.

Access: DOI record and public full text.

September 2026 inspection: Methods/Participants, Procedure and Analysis; Results themes; Discussion limitations. Relevant full-source body inspected; no raw-data replication claimed.

Gudjonsson, Young, and Bramham (2007)

Verification: Limited as of September 10, 2026; full methods/results were not inspected. Retained as a research lead, excluded from the narrative’s empirical support.

Type: Comparative study using a standardized measure of interrogative suggestibility.

Coverage: 36 adults diagnosed with ADHD and 36 general-population controls.

Abstract/preview report, not independently verified evidence: Reports no significant group difference in yielding to misleading questions or shifting answers after negative feedback. The ADHD group showed lower recall and more “don’t know” responses.

Limit: Small sample, formal experimental questioning, dated diagnostic context, and incomplete public access. It cannot settle how adults with ADHD experience ordinary questions.

Access: Publisher preview, institutional record, and DOI record.

Citation retained as research lead: Gísli H. Gudjonsson, Susan Young, and Jessica Bramham, “Interrogative Suggestibility in Adults Diagnosed with Attention-Deficit Hyperactivity Disorder (ADHD): A Potential Vulnerability During Police Questioning,” Personality and Individual Differences 43, no. 4 (2007): 737–745, doi:10.1016/j.paid.2007.01.014. Limited verification: the publisher’s detailed preview and an institutional record were inspected; a public full text was not located.

September 2026 inspection: Publisher detailed preview and King’s College institutional record. Only the visible material is corroborated; historical full-text access is not presumed.

Ginapp et al. (2023)

Verification: Verified.

Type: Nine semi-structured online focus groups using interpretative phenomenological analysis.

Coverage: 43 adults aged 18–35 reported clinical ADHD diagnosis and met the screening threshold. For 27, the researchers obtained provider/diagnostic documentation or used stimulant-prescription monitoring records as a diagnostic proxy; this does not mean 27 independently verified diagnoses through one uniform method.

Contribution: Participants described feeling different from neurotypical people, masking, communication difficulty, being perceived as “too much,” easier connection with some ADHD peers, and benefits and risks of ADHD online communities.

Limit: Self-selected, 84 percent female, 72 percent White, highly educated, predominantly North American, and mostly recruited from online ADHD communities. The study does not directly compare observed conversations by neurotype.

Access: DOI record and public full text.

September 2026 inspection: Methods sections 2.1–2.2; demographic table; Results relationship and community themes. Relevant full-source body inspected; no raw-data replication claimed.

Maras and Bowler (2012), suggestibility and compliance

Verification: Relevant full accepted manuscript inspected September 10, 2026: Methods 2.1–2.2, Results, Discussion, and Table 2.

Citation: Katie L. Maras and Dermot M. Bowler, “Brief Report: Suggestibility, Compliance and Psychological Traits in High-Functioning Adults with Autism Spectrum Disorder,” Research in Autism Spectrum Disorders 6, no.3 (2012): 1168–1175. https://doi.org/10.1016/j.rasd.2012.03.013.

Coverage: 32 autistic and 30 nonautistic adults participated overall. Instrument-specific samples were smaller: interrogative-suggestibility task, 20 autistic/19 comparison; self-reported compliance, 28/26. Recruitment and analysis totals are distinct.

Contribution: No significant group differences in yielding, answer shifting, total suggestibility, recall, or self-reported compliance were detected. These are separate outcomes. The study supplies the previously missing direct autism comparison, rather than relying on an eyewitness-recall-format study to answer a suggestibility question.

Limit: Small, selected adult sample; outcome-specific participation and self-report limits. Nonsignificance does not establish equivalence, remove individual vulnerabilities, or generalize to all autistic people, intellectual disabilities, or ADHD. Confidence: medium for the bounded reported result.

Access: Public accepted manuscript and repository record.

Huang methodological exchange — additional source records

Critique: Avraham N. Kluger and Thomas E. Malloy, “Question-Asking as a Dyadic Behavior,” Journal of Personality and Social Psychology 117, no.6 (2019): 1127–1138, doi:10.1037/pspi0000156, public manuscript. Inspected September 10, 2026. Role: methodological counterargument concerning the observational speed-dating analysis. It does not test the conversational experiments. The linked manuscript may predate the correction and must be read with the official notice; a separately corrected full manuscript was not verified here.

Official correction: Correction to Kluger and Malloy, doi:10.1037/pspi0000355, Journal of Personality and Social Psychology 119, no.6 (2020): 1477, inspected in full September 10, 2026 in the final page of this public journal PDF. The notice identifies the offered-versus-received date outcome error and affected results. It corrects the critique, not Huang et al.; an uncorrected reanalysis is not reliable evidence of refutation.

Author reply: Michael Yeomans, Alison Wood Brooks, Karen Huang, Julia Minson, and Francesca Gino, “It Helps to Ask: The Cumulative Benefits of Asking Follow-Up Questions,” Journal of Personality and Social Psychology 117, no.6 (2019): 1139–1144, doi:10.1037/pspi0000220, public full text. Full six-page reply inspected September 10, 2026. Role: response to the reanalysis, preserving experimental and observational distinctions. It is the original authors’ argument, not an independent replication.

Assessment: The dispute and correction qualify the speed-dating inference. The essay retains bounded experimental liking findings, does not claim retraction, and does not independently reproduce the dyadic analyses. Confidence is high in the existence and stated scope of the exchange; resolving all model-dependent conclusions would require reanalysis.

Author-reported data provenance — additional source records

Brooks study record: Many Coauthors, study 68, author-completed data-provenance report, inspected September 10, 2026. Brooks and Schweitzer report gaps in original/raw or reproduction data. Role: reproducibility limitation; this is neither an independent replication nor a misconduct adjudication.

Huang study record: Many Coauthors, study 45, author-completed data-provenance report, inspected September 10, 2026. The record states that Gino did not collect the data. This separate record prevents transferring conclusions from a coauthor’s other studies by association. Role: provenance context, not proof that every result is correct.

Appendix D: Counterevidence and Pruned Claims

Pruned: Questions are inherently coercive

This collapsed sequence pressure, care, learning, advice seeking, accountability, and surveillance into one moral category. The draft now argues that questions redistribute control and evaluates the surrounding conditions.

Pruned: Seeking knowledge naturally leads to punishment

The organizational literature supports real and anticipated penalties in some hierarchies. Inspected experiments also find that question asking can increase liking or competence judgments in specific settings. The leader-humility and sensitive-question studies are retained as limited-access research leads, not independently verified narrative evidence. “Natural” now identifies an open problem rather than a settled mechanism.

Pruned: Silence proves concealment

Nonresponse can reflect privacy, uncertainty, fear, fatigue, resistance to a false premise, or a healthy boundary. A question makes silence socially noticeable; it does not make one interpretation true.

Pruned: Knowledge always increases agency

Information can widen action, but it can also overwhelm, expose, constrain, or become another person’s instrument of control. The information-avoidance literature includes rational and protective reasons not to know.

Pruned: Refusal to answer is evidence of guilt

This is unsupported and incompatible with the essay’s distinction between legitimate inquiry and compelled disclosure.

Pruned: Questions are never neutral

The evidence supports sequence effects, framing, salience, and small average behavioral effects in some domains. It does not support assigning every question a positive or negative moral valence. The draft now distinguishes “rarely inert” from “never morally neutral.”

Pruned: Open questions always maximize agency

Very open prompts can preserve narrative freedom while creating high retrieval and relevance-selection demands. Bounded, answerer-directed structure can improve access without becoming leading.

Pruned: Neurodivergent communication is deficient communication

Matched-neurotype studies show effective autistic peer-to-peer information transfer and higher rapport than in mixed interactions. The draft treats at least some difficulty as relational mismatch rather than locating the entire failure in one person.

Pruned: Neurodivergent adults are generally more vulnerable to suggestion

The fully inspected Maras and Bowler autism comparison found no significant group difference in its measured suggestibility outcomes. It does not establish equivalence or generalize to all autistic adults. The ADHD comparison remains a disclosed research lead pending full text. Suggestibility, compliance, recall, processing, and response style remain distinct.

Counterevidence retained in the narrative

• follow-up questions can increase liking;

• advice seeking can increase perceived competence;

• deep conversations can feel less awkward and more connecting than participants predict;

• structured, reciprocal disclosure can generate temporary closeness;

• supportive question structure can improve recall for both autistic and allistic adults;

• no significant detail-retention difference was detected between autistic and allistic matched chains; this was not a formal equivalence test;

• some adults with ADHD report easier connection and recognition with ADHD peers;

• a small inspected autism comparison did not detect increased interrogative suggestibility; the ADHD source remains limited and does not establish a verified cross-diagnostic conclusion;

• high-quality listening can communicate responsiveness through questions, reflections, paraphrasing, attention, and other behaviors;

• demand–withdraw evidence locates harm in a reciprocal pressure-and-retreat pattern rather than in one gender or a fixed conversational identity;

• not all information avoidance is irrational, selfish, or harmful.

Appendix E: Resolution Record

Resolved for the base draft

• The thesis has been narrowed from “questions are control” to “questions move control.”

• “Rarely neutral” has been narrowed to “rarely interactionally inert,” with moral valence evaluated separately.

• The subtitle is retained as the proposition the essay tests rather than a conclusion it assumes; the opening and ending make clear that many consequences presented as natural are socially chosen.

• Repeated summaries and evidence disclaimers were consolidated or rewritten when they did not add pressure, precision, or a new boundary; deliberate recurrence of the central claim remains.

• Attention-deficit/hyperactivity disorder (ADHD) is introduced before abbreviation in both the primary narrative and the independently readable scope appendix, then abbreviated consistently.

• Power-over and power-to are separated.

• Social consequences are distinguished from consequences intrinsic to learning.

• Question effects are mapped across asker, answerer, relationship, institution, and knowledge.

• Questioning authority is distinguished from being questioned by authority.

• Genuine connection is modeled as voluntary reciprocal disclosure plus experienced responsiveness, not question quantity alone.

• Autism and ADHD evidence is separated; neurodivergence is not treated as one population trait.

• Accessibility and reciprocity/responsiveness have been added to the final evaluative test.

• Direct evidence, adjacent evidence, and normative inference are labeled separately.

• Counterevidence appears in the main narrative rather than only in supporting material.

• Coercive-control evidence is limited to patterns, retaliation, surveillance, and loss of autonomy.

• Documented personal scenes now test both sides of the thesis: inquiry as an attempt to understand, and certainty as a way of taking away another person’s authority to correct the model.

• Asking and non-asking are no longer treated as moral personality types; perceived responsiveness and conflict loops provide the stronger unit of analysis.

Reviewed, non-blocking extensions

• Whether a later draft should more fully distinguish asking another person from asking a system, archive, institution, or oneself.

• Whether the moral-obligation section needs a dedicated philosophy source in addition to behavioral information-avoidance research.

• Whether experiences specific to dyslexia, dyspraxia, Tourette syndrome, intellectual disability, traumatic brain injury, and other neurodivergent groups can be supported with question-specific research rather than generalized accommodation guidance.

• Whether the six-part test—entitlement, refusal, framing, accessibility, reciprocity/responsiveness, and consequence—should become the essay’s final organizing structure.

Appendix F: Historical Publication Review — August 2026

The statements below record the August review and release. They are preserved as history, not current verification or authorization. September corrections and access findings supersede them where they conflict; the current narrative requires a new exact-hash authorization.

Publication Review Report

• Draft: Questions Are Control
• Review date: August 3, 2026
• Final status: Publish.
• Release state at that review: Published on the canonical Dicatania site on August 4, 2026; that historical narrative was bound to the listed hash.

Source audit

• Complete: Yes.
• Inventory: Twenty-nine sources directly inspected; three sources retained with limited verification.
• Issues: No unsupported central claim was identified. The limited-access sources remain explicitly labeled and do not carry the argument alone.

Citation verification

• Exact URLs or stable identifiers present: Yes.
• Public accessibility checked: Yes, August 3, 2026.
• Direct source inspection confirmed: Yes.
• Issues: The stale Frazier full-text link was replaced with the current institutional full text. Heritage's stable DOI remains, but its previously public author copy is no longer available; that access limitation is disclosed in the note and source record.

Claim traceability

• Complete: Yes.
• Unsupported claims: None identified. First-person observations remain labeled as private, selection-biased, and nongeneralizable.

Contradictions and uncertainty

• Contradictory evidence covered: Yes.
• Unresolved uncertainty disclosed: Yes.
• Confidence impact: The essay rejects the strongest thesis, limits “rarely neutral” to interactional force, separates autism and ADHD evidence, and records missing evidence for other neurodivergent groups.

Readability

• Audience fit: General readers comfortable with a long evidence-led personal essay.
• Terms to define: None outstanding; specialized concepts are defined when introduced.
• Repetition pass: Approved through author override after pruning repeated source-access disclosures from the narrative.
• Acronym pass: Complete; unfamiliar acronyms are introduced before abbreviation and used consistently afterward.
• Academic fog to remove: None that warrants changing the candidate narrative.
• Primary narrative readable without appendices: Yes.
• Supporting materials independently verify significant claims: Yes.
• Chicago Notes and Bibliography ready: Yes.

Voice authorization

• Voice profile: v2, approved.
• Primary narrative SHA-256: sha256:28c3e1903cd642133856432383256c7b11c1754d04ab87d0c5649e4174be784f.
• Review record: content/editorial/voice-reviews/questions-are-control.md. The superseded v1 record is preserved at content/editorial/voice-reviews/archive/questions-are-control-v1.md.
• Authorization basis: Author override following explicit acceptance and lock of the final candidate; independent editorial review is not claimed.
• Authorization matched the August narrative: Yes; it does not authorize the revised September narrative.
• Detector scores used: No.
• Issues: None.

Release assets and controls

• Cover: Approved and linked in front matter (cover-v2.png).
• Social card: Approved, linked in front matter (social-card-v2.png), and matched to the cover at 1200 × 630 pixels.
• Personal narrative: Names and identifying details remain omitted; private records support first-person reflection only.
• Audio: The existing Ava system-voice narration is not approved for publication and has been removed from deployable paths. A public audiobook requires a separately licensed voice and documented commercial rights.

Required changes

• Primary narrative or evidence: None.
• Release completed: On August 4, 2026, draft was changed to false and status to published without altering the hash-bound primary narrative. The final strict build and canonical deployment completed the release.
• If the primary narrative changes: Compute a new hash and repeat independent voice review before release.

September 10, 2026 Correction and Verification Addendum

Corrected the reversed social-risk synthesis, qualified universal and speculative mechanisms, distinguished bounded experiments from coercive disclosure, acknowledged the inclusive healthcare sample, and identified the Crompton papers’ shared participants. Ginapp’s diagnostic-documentation and prescription-proxy routes are now distinguished. Miles’ inconsistent participant totals remain explicit, with the published effect and confidence interval retained.

Of the original 32 source records, 26 received full relevant-source inspection in this audit. Six remain limited: Frazier; Cojuharenco and Karelaia; Hart; Burris; Morrison; and Gudjonsson, Young, and Bramham. They remain in the library as disclosed research leads, with their citation metadata preserved, and no longer supply the narrative’s empirical claims. This is a verification limitation, not evidence that those papers are false.

Added one fully inspected autism comparison, the three records in the Huang critique/correction/reply exchange, and two author-completed provenance records. The new autism evidence does not generalize to ADHD. The methodological exchange does not erase the original experimental findings; data availability gaps do not establish misconduct or invalidity.

The dated September record supersedes the historical August readiness statements. Personal observations remain private and nongeneralizable; the six practical tests remain an ethical proposal rather than a validated instrument. Original participant-level analyses and meta-analyses were not reproduced, and targeted correction/retraction checks are not exhaustive clearance.