Back to the Field Note

Field Note No. 002 · Companion evidence note

Research & Perspectives: Gaslighting Is Not a Difference of Opinion

Companion evidence note for Field Note No. 002, Gaslighting Is Not a Difference of Opinion. This note records the research informing the article, the populations and contexts represented, access levels for each source, important limitations, and the points where the article moves from published evidence into Sage & Atlas editorial interpretation. It is intended for readers who want to explore the evidence in greater depth; the Field Note itself remains the primary reading experience.

Attribution: Lori / Sage & Atlas editorial.

Access labels and sample/co-abuse bounds describe the source versions reviewed for this companion. Evidence descriptions reflect the source versions reviewed for this note, not a new review of all underlying scholarly full texts. Access labels name the specific source versions available; they are not evidence-quality ratings. Access counts remain 10 full texts (Klein = author preprint only) and 2 abstracts.

Citation years: References use the issue/publication year when volume, issue and pages are assigned; online-first dates are recorded separately.

Why This Note Exists

Field Notes articles are written for readers, not as literature reviews. Research informs what an article says and where it stops, while study-by-study detail lives here. This note records what informed Field Note No. 002, what the evidence does and does not support, who was studied, and where the article moves from findings into editorial interpretation.

Working Definition Used in the Field Note

For the Field Note, Sage & Atlas uses a studio working definition synthesized from:

  1. Hailes & Goodman (2025) — FULL TEXT. Qualitative interviews with 14 IPV survivors. Two-part conceptualization: tactics that undermine another person’s sense of reality, and resulting epistemic self-doubt (erosion of confidence in oneself as a knower). Domains of knowing and lasting impacts on self-trust as knower are described.
  2. Sweet (2019) — FULL TEXT. Sociological framework grounded in interviews with 43 women from domestic violence support groups (Chicago area). Gaslighting as consequential in power-laden intimacy; stereotypes and institutions; isolation; intent not required as a definitional necessity in Sweet’s account; caution that over-broad political uses of the term “go too far.”

Labeled as studio synthesis. The Field Note does not claim that all scholars agree on a single definition. Intent debates, pattern-versus-incident debates, and whether a power imbalance is strictly necessary remain open across the field (see Darke et al., 2026; Klein et al., 2023).

Experience Distinctions Framework

The Field Note presents an Experience Distinctions Framework (Sage & Atlas editorial interpretation; not a validated typology or diagnostic tool). Categories are overlapping distinctions for clarity, not stages. One category does not inevitably become the next.

CategoryWhat it isGuardrail
Ordinary disagreementDifferent views, priorities, or interpretationsNot abuse by itself
Differing recollectionsDifferent accounts of a shared event; editorial distinction, not a memory-science claimNot proof of gaslighting; never used to impugn survivors
InvalidationDismissing feelings, reactions, or meaningsNot automatically gaslighting
Broader psychological abuse / coercive control tacticsDenial, minimization, blame shifting, isolation, intimidation, control of resources or narrativesOverlaps with gaslighting sometimes; constructs are not identical (Dokkedahl; Lohmann)
Gaslighting patternRecurring behavior undermining another person’s sense of reality and eroding confidence in their own knowing; power can compound harmNot every fight or lie establishes this pattern

A diagnosis is neither required nor evidence of gaslighting.

The framework uses equal-weight distinctions, not a staircase, score or diagnostic assessment. The textual distinctions stand on their own.

Evidence by Research Area

A. Understanding Gaslighting

Claim / questionBest sourcesAccessStrength / notes
Definitions vary (intent, repetition, outcome, power)Darke (2026); Adair (2026)Darke: FULL TEXT; Adair: FULL TEXTWell-supported as field observation
Power-laden sociological definitionSweet (2019)FULL TEXTSweet’s framework; n = 43 (not 48)
Two-part tactics + epistemic self-doubtHailes & Goodman (2025); Adair (2026)FULL TEXTSolid for labeled working definition
Intent not always requiredSweet (2019); Darke (2026)Sweet: FULL TEXT; Darke: FULL TEXTMixed across fields
Power necessary? Open questionSweet vs Klein et al. (2023)Sweet: FULL TEXT; Klein: author-preprint FULL TEXTPresent both sides
Gold-standard validated gaslighting instrument——Not established as a gold-standard validated instrument in sources reviewed for this note; March scale measures acceptance

B. Intimate Relationships

Claim / questionBest sourcesAccessNotes
Survivors describe denial, minimization, blame shifting, memory/judgment attacksSweet; Hailes & Goodman; KleinSweet: FULL TEXT; Hailes & Goodman: FULL TEXT; Klein: author-preprint FULL TEXTElements of a pattern — not one-to-one proof
Isolation / credibility / institutionsSweet (2019)FULL TEXTSample-bound
Distinguish from broader psych abuseHailes two-part; Dokkedahl terminologyHailes & Goodman: FULL TEXT; Dokkedahl: FULL TEXT (publisher HTML basis)These are distinct constructs

C. Personality Research Considered, Not Used to Substantiate the Revised Article Section

The article’s section “Personality, Diagnosis, and Harmful Behavior” emphasizes accountability without treating a diagnosis as a prediction. Its focus is what someone repeatedly does and the harm that behavior causes, not an invitation to diagnose them. This is Sage & Atlas editorial framing, not a summary of the studies below. Neither March nor Oliver establishes that particular diagnoses cause gaslighting, and neither supports a claim about how common gaslighting is among people with personality disorders. These studies remain background considered during development, not evidence for a diagnosis-based explanation of abuse. It is editorial and ethical wording, not a research finding from the studies reviewed here. Those studies do not establish the prevalence of these behaviors across personality-disorder diagnoses.

Narcissistic Traits / NPD

  • March et al. (2025) — FULL TEXT (OA HTML); previously considered, now unused in the article. Preregistered cross-sectional survey of 315 Australian community adults; outcome = acceptance of gaslighting tactics (author-developed Gaslighting Questionnaire), not observed perpetration.
    • Zero-order: Dark Tetrad traits associated with greater acceptance.
    • Unique multivariate predictors: primary psychopathy, Machiavellian tactics, sadism.
    • Narcissism (grandiose and vulnerable) is NOT a unique predictor after shared variance.
    • Sex × vulnerable narcissism interaction concerns acceptance, not observed perpetration.
  • Oliver et al. (2024) — FULL TEXT (public PMC article); previously considered, now unused in the article. Across 22 studies from 21 articles, N = 11,520, the overall pooled correlation was r = .15, 95% CI [.12, .19], with high heterogeneity (I² = 86.42%). This is a weak association between trait narcissism and reported IPV perpetration, not a percentage, probability or causal estimate. The authors reported stronger associations for psychological and cyber forms than physical forms. It is not gaslighting-specific and does not establish a link to clinical NPD. These numbers are recorded here for transparency; they are not used in the article.
  • Clinical NPD × gaslighting: No direct empirical association established in sources reviewed for this note; not a universal absence claim.

BPD

  • No direct empirical association between BPD (diagnosis or traits) and gaslighting was verified in sources reviewed for this note.
  • This R&P does not expand into BPD symptom lists or speculative mechanisms. The absence statement is limited to sources reviewed for this note.

Other Constructs

  • Primary psychopathy, Machiavellian tactics, and sadism appear as unique predictors of acceptance in March (2025).
  • Miano et al. (2021) — ABSTRACT only. At most, a young-adult study reports associations involving gaslighting and personality-pathology dimensions. No coefficients, diagnostic, causal, or observed-perpetration claims are supported by the available access. Unused for article claims.
  • Darke et al. (2026) caution against over-focusing on individual differences as the primary explanation of gaslighting.

D. Effects

OutcomeGaslighting-specificBroader psych IPV / CCGuidance
Epistemic self-doubt / loss of self-trust as knowerHailes & Goodman (2025) FULL TEXT—Prefer gaslighting-specific wording
Broader mental-health associations, not gaslighting-isolated effectsLimited gaslighting-specificDokkedahl: FULL TEXT (publisher HTML); Lohmann: FULL TEXT (complete public PMC article)Lohmann pooled PTSD/depression symptom severity; do not imply both reviews pooled every outcome
Isolation / dependence on other’s interpretationSweet; HailesCC literaturePattern language; not universal

E. Recognition and Recovery

TopicEvidenceLimit
Recognizing patterns without diagnostic labelsQualitative + definitional workNo checklists
Healing / coping themesHailes FULL TEXTNo promised methods
Documentation / confrontation as restorers of self-trustNot established as tested gaslighting interventions in sources reviewed for this noteSafety first; optional reflection only

Intimate Relationships Evidence Detail

Sweet (2019): FULL TEXT. Power-embedded sociological framework; stereotypes, institutional leverage and isolation. Interviews with 43 women recruited from four Chicago-area domestic violence support groups; 33 interviews in English and 10 in Spanish. All but two identified a male partner as primary abuser; the other two identified a father or brother. This is not an exclusively intimate-partner sample or a general population sample. See Sweet methods pp. 857–858 and note 9. Primary N is 43, not the secondary figure of 48.

Hailes & Goodman (2025): FULL TEXT. Two-part conceptualization; domains of knowing; self-trust and survivor-described coping themes. The study included 14 participants: 13 identified as women and one as genderqueer and nonbinary; discussed harming partners were male. None remained in the gaslighting relationship, which had ended one to eight years earlier. English-speaking, retrospective, no partner data; survivors treated as truthtellers by design. Gaslighting co-occurred with physical, sexual and other psychological abuse, limiting separation of harms. See methods p. 272 and limitations pp. 279–280.

Klein et al. (2023): FULL TEXT for the accessible author preprint only; the journal version of record was not checked. Qualitative analysis of gaslighting in romantic relationships; in the preprint, power imbalances were common but not described as necessary in every reported case. That bound belongs to this sample and the authors’ interpretation.

What behaviors can / cannot prove: Recurring contradiction, minimization, blame shifting, and attacks on judgment appear in survivor accounts of gaslighting patterns, but no single behavior proves gaslighting out of context. The Field Note uses pattern language, not a diagnostic checklist.

Direct vs Indirect Evidence Register

“Direct” means relevance to a gaslighting-named construct, not an evidence-quality ranking, proof of enacted behavior, or demonstrated perpetration. Darke (2026) is a narrative interdisciplinary review, not a systematic review.

LabelSources
Direct (gaslighting-named)Sweet (2019); Hailes & Goodman (2025); Klein et al. (2023); Darke et al. (2026); Adair (2026); March et al. (2025) (acceptance outcome); Miano et al. (2021) (access-limited)
Indirect (related broader constructs)Dokkedahl et al. (2022); Lohmann et al. (2024); Oliver et al. (2024)
Studio interpretation (must be labeled)Experience Distinctions Framework; precision thesis; recognition prompts; personality framing and recovery conclusion

Methodological Limitations (Required)

  • The sources reviewed for this note include selected samples with these language and setting limits: Hailes and Klein recruited English-speaking participants; Sweet’s interviews were conducted in English and Spanish (33 English and 10 Spanish) with women recruited from four Chicago-area domestic violence support groups (Hailes methods p. 272; Sweet methods pp. 857–858; Klein author preprint sample). This describes the reviewed evidence, not a quantified field-wide language or geographic skew.
  • Definitional inconsistency across disciplines.
  • A gold-standard validated gaslighting instrument was not established in sources reviewed for this note; March’s questionnaire measures acceptance, not perpetration.
  • The Sweet and Hailes samples have the distinct population limits detailed above; survivor-centered accounts are not prevalence estimates. Hailes reports co-occurring abuse, so harms cannot be isolated to gaslighting alone.
  • Dokkedahl: broader psychological IPV; predominantly cross-sectional evidence; low/very-low GRADE certainty and substantial heterogeneity; planned adjustment for other physical/sexual IPV and prior trauma could not be performed. Lohmann: meta-analytic effects from cross-sectional data, PTSD and depression symptom severity, high heterogeneity and no causal inference; only one eligible CPTSD study, so no pooled CPTSD estimate. Neither review isolates gaslighting. See both reviews’ methods, results and limitations.
  • Personality studies: correlational, self-report, community samples; not clinical diagnoses.
  • Power necessity unresolved (Sweet vs Klein).
  • Access limits for this note: Miano (2021) and Graves & Samp (2021) remain abstract-only after lawful attempts. Oliver et al. (2024) is now full text via the public PMC article. Klein is verified from the author preprint, not a journal-version check. The historical memory leads did not supply verified primary support.
  • Memory science remains omitted. The technical sentence “Ordinary memory is reconstructive for everyone” stays removed: its provenance is untraced, and no verified supporting passage from those historical memory leads was obtained. This is not a finding that reconstructive-memory science is false.

Findings Informing the Article and Sage & Atlas Interpretation

What the article saysKind of claimBasis
Gaslighting involves undermining reality and eroding confidence in one’s own knowingWorking definition (studio synthesis of empirical qualitative work)Hailes & Goodman (2025); Sweet (2019)
Pattern can be made more consequential by power, stereotypes, institutions, isolationQualitative / theoreticalSweet (2019); Hailes & Goodman (2025)
Power imbalance may be common but necessity is openOpen scholarly disagreementSweet vs Klein et al. (2023)
Word has broadened in public use; precision mattersCultural observation + Sweet’s scope cautionSweet (2019); Merriam-Webster (2022) background
Epistemic injury / loss of self-trust as knowerQualitativeHailes & Goodman (2025)
Broader psych IPV / CC linked with serious mental health outcomesMeta-analytic / review (indirect for gaslighting)Dokkedahl (2022); Lohmann (2024)
Personality section: acknowledge harmful behavior without diagnostic predictionRevised editorial and ethical wording; no empirical citation in section“May use” does not state prevalence; the reviewed studies do not establish prevalence across personality-disorder diagnoses
Experience Distinctions Framework (five overlapping distinctions)Sage & Atlas interpretationEditorial framework; no study tests it
Disagreement / differing memories / invalidation are not automatically gaslightingSage & Atlas interpretation grounded in definitional workEditorial reading
Recognition prompts, safety-first stance, and educational boundarySage & Atlas interpretationEditorial and ethical guidance, not tested interventions in sources reviewed for this note
Rebuilding self-trust may be winding and take timeEditorial, non-prescriptive conclusion; qualitative resonance, not a treatment findingHailes survivor accounts inform lasting self-doubt; no recovery timetable or safe/effective strategy established in sources reviewed for this note; distance is not always safe and none of these is required

Selected References (APA 7)

  • Adair, J. (2026). Defining gaslighting in gender-based violence: A mixed-methods systematic review. Trauma, Violence, & Abuse, 27(4), 1409–1424. https://doi.org/10.1177/15248380251344316 (Online first: 12 July 2025)
  • Darke, L., Paterson, H., & van Golde, C. (2026). Illuminating gaslighting: A comprehensive interdisciplinary review of gaslighting literature. Journal of Family Violence, 41(6), 1325–1341. https://doi.org/10.1007/s10896-025-00805-4 (Online first: 13 January 2025)
  • Dokkedahl, S. B., Kirubakaran, R., Bech-Hansen, D., Kristensen, T. R., & Elklit, A. (2022). The psychological subtype of intimate partner violence and its effect on mental health: A systematic review with meta-analyses. Systematic Reviews, 11, Article 163. https://doi.org/10.1186/s13643-022-02025-z
  • Hailes, H. P., & Goodman, L. A. (2025). “They’re out to take away your sanity”: A qualitative investigation of gaslighting in intimate partner violence. Journal of Family Violence, 40(2), 269–282. https://doi.org/10.1007/s10896-023-00652-1 (Online first: 5 October 2023)
  • Klein, W., Li, S., & Wood, S. (2023). A qualitative analysis of gaslighting in romantic relationships. Personal Relationships, 30(4), 1316–1340. https://doi.org/10.1111/pere.12510 (accessed as author manuscript/preprint; journal version not checked)
  • Lohmann, S., Cowlishaw, S., Ney, L., O’Donnell, M., & Felmingham, K. (2024). The trauma and mental health impacts of coercive control: A systematic review and meta-analysis. Trauma, Violence, & Abuse, 25(1), 630–647. https://doi.org/10.1177/15248380231162972 (Online first: 13 April 2023)
  • March, E., Kay, C. S., Dinić, B. M., Wagstaff, D., Grabovac, B., & Jonason, P. K. (2025). “It’s all in your head”: Personality traits and gaslighting tactics in intimate relationships. Journal of Family Violence, 40(2), 259–268. https://doi.org/10.1007/s10896-023-00582-y (Online first: 23 June 2023; considered; unused in article)
  • Merriam-Webster. (2022). Word of the year 2022. https://www.merriam-webster.com/wordplay/word-of-the-year-2022
  • Miano, P., Bellomare, M., & Genova, V. G. (2021). Personality correlates of gaslighting behaviours in young adults. Journal of Sexual Aggression, 27(3), 285–298. https://doi.org/10.1080/13552600.2020.1850893 (ABSTRACT)
  • Oliver, E., Coates, A., Bennett, J. M., & Willis, M. L. (2024). Narcissism and intimate partner violence: A systematic review and meta-analysis. Trauma, Violence, & Abuse, 25(3), 1871–1884. https://doi.org/10.1177/15248380231196115 (FULL TEXT; public PMC article. Previously considered; unused in the article.)
  • Sweet, P. L. (2019). The sociology of gaslighting. American Sociological Review, 84(5), 851–875. https://doi.org/10.1177/0003122419874843

Sources Not Used in the Article

  • Loftus (2005); Gabbert et al. (2003); Johnson et al. (1993) — general memory science; ethics-bound; omitted from the article. These historical leads did not provide verified primary support for the article.
  • Darke Memory (2025) — metadata/existence only; unidentified lead; no verified primary text was reviewed; not the interdisciplinary review cited above. Not used in the article.
  • Graves & Samp (2021) — ABSTRACT only after a lawful attempt; the primary abstract concerns experience of gaslighting in a recalled dating disagreement, not established observed perpetration. Deferred from article use.

Source-Specific Access Register

Labels describe the versions reviewed for this companion. Exactly one access label is used per source. Full text names the accessible version, not a quality grade or an audit of every underlying study.

SourceRecorded accessBasis and boundary
Sweet (2019)FULL TEXTArticle reviewed for this companion; methods pp. 857–858 and note 9; primary N = 43
Hailes & Goodman (2025)FULL TEXTComplete text reviewed for this companion; methods p. 272; limitations pp. 279–280
Adair (2026)FULL TEXTComplete public PMC article; review concerns definitions, not measurement validation
Darke (2026)FULL TEXTComplete publisher HTML; narrative interdisciplinary review, not systematic; not Darke Memory
Klein (2023)FULL TEXTAccessible author preprint only; journal version of record not checked
March (2025)FULL TEXTComplete publisher HTML for this companion; acceptance, not observed perpetration; unused in article
Dokkedahl (2022)FULL TEXTComplete publisher HTML; broader psychological IPV
Lohmann (2024)FULL TEXTComplete public PMC article; coercive control, not gaslighting
Oliver (2024)FULL TEXTComplete public PMC article for this companion; unused in article; bounded numbers retained here
Miano (2021)ABSTRACTLawful full-text attempts unsuccessful; unused for article claims
Merriam-Webster (2022)FULL TEXTPrimary webpage, cultural context only
Graves & Samp (2021)ABSTRACTDeferred; primary abstract concerns experience, not established observed perpetration

Editorial Safety and Educational Boundaries

The cautions about distance, confrontation, documentation, and privacy; the educational/support boundary; and the qualified final passage are Sage & Atlas editorial and ethical language. They are not research findings, tested interventions, legal advice, or claims that a particular recovery path is safe or effective. The reflection prompts are optional and need not be written down. The article does not recommend a named support resource.

Return to the Field Note: Gaslighting Is Not a Difference of Opinion

All Research & Perspectives notes