RESEARCH DOSSIER
Relationship Trust, Betrayal, and Infidelity
Abstract. Relationship trust, betrayal, and infidelity are interrelated processes of expectation, relational obligation, boundary violation, and adjustment to new evidence. This critical narrative monograph reconstructs the history of trust research, distinguishes predictability, dependability, faith, risk regulation, attachment, and commitment, and establishes agreement-sensitive operational definitions of infidelity. It synthesizes population studies, motivation research, longitudinal and experimental evidence, psychometric methods, observational findings on disclosure, and small clinical intervention studies, including a 2026 dyadic pre-post report. The evidence supports heterogeneous motives and post-betrayal trajectories, context-sensitive outcomes of forgiveness, and the importance of sustained behavior for justified trust, while causal claims about universal treatments remain limited. The analysis specifies threats to construct validity, selection and recall bias, the distinction between relationship continuation and individual recovery, and prospective tests for digitally mediated and human–artificial relational boundary questions.
Keywords: relationship trust; betrayal; infidelity; extradyadic involvement; sexual fidelity; emotional infidelity; trust repair; risk regulation; interdependence; relationship commitment; attachment; disclosure; forgiveness; relationship dissolution; dyadic research; couples therapy; relationship boundaries; digital intimacy; human–artificial relations
The Scientific Object of Trust, Betrayal, and Infidelity
Relationship trust is an expectation that a particular partner will exercise care, keep consequential commitments, and respond dependably when one's interests are vulnerable to that partner's choices. Betrayal occurs when conduct violates a relational obligation or reasonably established expectation of loyalty, honesty, or protection. Infidelity is a class of relational transgression involving sexual, romantic, emotional, or digitally mediated involvement outside the relationship in circumstances where that involvement violates an agreed or mutually understood boundary. These definitions require attention to the actual relationship agreement, the partner's knowledge, and the behavior under examination. Suspicion is a psychological appraisal; an established violation is an event. Infidelity, betrayal, and low trust overlap without being interchangeable constructs. Research on close relationships provides the larger framework of interdependence, risk, and interaction histories within which these phenomena become intelligible Rempel et al. (1985) Finkel et al. (2017) Blow & Hartnett (2005a).
The unit of analysis changes with the question. Trust can be measured as a person's expectation, a partner-specific judgment, or a dyadic process sustained by repeated behavior. Betrayal can be assessed as an identified episode, a pattern of deception, or a retrospective interpretation following disclosure. Infidelity may be counted as an individual self-report of behavior, a mutually verified event, or a discrepancy between partners' reports. These measurements carry different evidentiary weight. A survey of extramarital sexual behavior cannot establish a rate of emotional infidelity, and a distressed respondent's suspicion cannot establish an affair. A scientific analysis must retain the distinction between behavioral occurrence, knowledge of the event, perceived intent, and subsequent relational consequences Blow & Hartnett (2005a) Allen et al. (2005).
This monograph is a critical narrative synthesis of original observational and experimental research, longitudinal relationship studies, psychometric investigations, nationally sampled surveys, clinical studies, and methodological reviews. Its central questions concern how trust becomes credible, how boundaries and secrecy shape the recognition of betrayal, what evidence explains extradyadic involvement, how discovery changes the relationship, and under what conditions repair is possible. It evaluates alternative explanations and study-design limitations while keeping safety, coercion, and independent decision-making within the research frame. It reports no newly collected data or pooled meta-analysis. The synthesis treats relationship continuity, individual recovery, and restoration of trust as separate possible outcomes Karney & Bradbury (1995) Fincham & May (2017).
Historical Foundations of Relationship Trust Research
Systematic trust research developed through social psychology's study of interdependence, attribution, and the willingness to become vulnerable to another person's actions. The trust framework proposed by Rempel, Holmes, and Zanna in 1985 distinguished predictability, dependability, and faith in close partners. Their survey and scale work showed that these components could be investigated separately, giving relationship researchers a more discriminating language than a single question about whether someone was trustworthy. Predictability concerns anticipated behavior based on prior observation; dependability involves reliability under pressure; faith involves confidence that care will persist under uncertain future circumstances. These dimensions remain analytically useful even when specific scale structures require evaluation in new populations Rempel et al. (1985).
By the 1990s, longitudinal work had made relationship change a principal object of explanation. Karney and Bradbury's 1995 review argued for studying the course of marital quality and stability through vulnerabilities, stressful contexts, and adaptive processes rather than inferring temporal dynamics from one assessment. Commitment research then examined the behavioral information people receive when partners sacrifice, accommodate, or depart from immediate self-interest. Two longitudinal studies reported by Wieselquist and colleagues associated such prorelationship conduct with trust development. In this tradition, trust becomes a learned expectation arising from interaction, yet its formation can also be biased by attachment history, perceived regard, and present vulnerability Karney & Bradbury (1995) Wieselquist et al. (1999).
Research on infidelity developed alongside this broader relationship science but confronted unusually severe measurement obstacles. Concealment, stigma, disagreement over definitions, and the timing of disclosure complicate representative sampling and retrospective recall. The paired reviews by Blow and Hartnett in 2005 documented limitations in how infidelity was defined and measured and summarized evidence concerning relational, individual, and contextual factors. These reviews are historical anchors of the field, not definitive estimates of contemporary prevalence. Digital communication, diversified partnership arrangements, and changed disclosure practices require periodic reevaluation of instruments and recruitment frames Blow & Hartnett (2005a) Blow & Hartnett (2005b) Fincham & May (2017).
Predictability, Dependability, and Faith
Predictability grows when repeated circumstances reveal stable features of a partner's conduct. It can make interactions efficient, although a reliably self-interested person may be predictable without deserving confidence as a caring partner. Dependability requires information about how the partner acts when incentives conflict and one person bears the consequences of the other's choice. Faith extends beyond observed cases to uncertain futures; it is especially consequential when illness, relocation, financial strain, or vulnerability make constant verification impossible. The multidimensional view of trust therefore asks what a partner is expected to do, under which conditions, and on what evidence Rempel et al. (1985) Wieselquist et al. (1999).
A trustworthy partner's behavior can be operationalized through kept agreements, truthful communication, responsiveness to requests, respect for boundaries, and repair after errors. These observations become more informative when opportunities to violate expectations are present. A partner who has never encountered a conflict of interest offers less evidence about dependability under temptation than a partner whose costly behavior has repeatedly supported the relationship. At the same time, retrospective accounts are vulnerable to halo effects: global satisfaction can color memories of earlier reliability. A strong design combines partner reports, time-stamped events, independent descriptions, and repeated assessments instead of treating generalized impressions as direct behavioral records Wieselquist et al. (1999) Murray et al. (2006).
Trust also includes calibration. Extremely low confidence can prevent cooperation despite a consistent record of responsible behavior; excessive confidence can leave a person vulnerable to repeated violations. The scientific goal is to explain how people adjust expectations in light of relevant evidence and the costs of being mistaken. This is particularly important after betrayal, when the injuring act is known but future conduct cannot yet be established. No fixed number of apologies, disclosures, or monitoring episodes mechanically restores faith. Any increase in justified confidence depends on the consistency and diagnostic quality of subsequent actions and on the autonomy of the person asked to trust again Murray et al. (2011) McNulty & Fincham (2012).
Trust as Risk Regulation
Vulnerability is the functional condition that makes interpersonal trust consequential. Murray, Holmes, and Collins described a risk-regulation system in which confidence in a partner's positive regard helps people pursue intimacy despite possible rejection. In this account, cognitive expectations, affective reactions, and behavioral rules influence whether a person approaches or withdraws when relational dependence becomes salient. Trust functions within a continuing tension between securing connection and protecting oneself against exploitation. The model makes specific predictions about the context in which expressions of care can support closeness Murray et al. (2006).
Subsequent work differentiated reflective trust, consisting of articulated beliefs about the partner's caring, from impulsive trust, represented by relatively automatic evaluations. An experimental and observational research program examined situations where these forms diverged, including the regulation of self-protection. A person can report deliberate confidence while responding automatically with alarm, or retain automatic warmth while deliberately identifying an unacceptable violation. This distinction offers a plausible framework for understanding the unsettled period after disclosure, although applying findings from ordinary trust-threat tasks directly to severe infidelity remains a theoretical extension requiring targeted tests Murray et al. (2011) Murray et al. (2013).
Risk regulation is reciprocal. One person's protective withdrawal may be interpreted by the other as rejection; an attempt at reassurance may be understood as care or as pressure depending on interaction history. Murray and Holmes's motivation-management framework places mutual responsiveness and the management of self-interested incentives at the center of close relationships. This perspective explains why relationship repair requires more than a private change of belief. Trust is reorganized through a process in which responses to uncertainty, accountability for conduct, and new evidence become visible to both partners over time Murray & Holmes (2009) Wieselquist et al. (1999).
Commitment, Interdependence, and Behavioral Evidence
Commitment is the intention to maintain a relationship; trust concerns expectations about a partner's conduct under vulnerability. The constructs are associated, yet commitment can persist because of investments, barriers to leaving, or limited alternatives even when confidence in the partner has deteriorated. The Investment Model Scale distinguishes satisfaction, alternatives, investment size, and commitment, offering a framework for identifying why a person remains in a relationship. Its validation studies provide evidence that the components can be measured separately and can predict later dyadic adjustment and relationship persistence Rusbult et al. (1998).
Commitment influences trust through behavior when it motivates accommodation or sacrifice that communicates a partner's priorities. Wieselquist and colleagues' two longitudinal studies advanced a model in which prorelationship actions provide diagnostic evidence of benevolent motives. The causal interpretation remains bounded: commitment may help produce these actions, but unmeasured relationship quality and reciprocal influence can also link reported commitment and trust. Separating the actor's intention, the observer's interpretation, and the concrete consequences of action improves causal inference Wieselquist et al. (1999) Finkel et al. (2017).
After an affair, partners may report continuing attachment, shared responsibilities, and high investment while also reporting damaged trust. Researchers should distinguish the decision to stay, the felt ability to leave, financial or caregiving dependence, relationship satisfaction, and emotional safety. Persistence alone is not evidence of repair. Likewise, separation does not imply that the betrayed partner failed to recover psychologically. A theoretically careful analysis represents relationship status as one outcome among several and considers the unequal resources and constraints that can shape whether continuation is freely chosen Rusbult et al. (1998) McNulty & Fincham (2012).
Attachment and Expectations of Partner Availability
Attachment theory offers a complementary account of why some relational threats are especially destabilizing. Hazan and Shaver's 1987 work translated attachment concepts into the study of adult romantic relationships, identifying expectations concerning closeness, availability, and abandonment. Attachment-related anxiety can heighten vigilance toward rejection, while avoidance can support distancing and reduced reliance on a partner. These tendencies are probabilistic patterns derived from assessments and interaction histories; they cannot reliably identify whether any particular partner has been unfaithful Hazan & Shaver (1987) Murray et al. (2006).
Following confirmed betrayal, prior attachment concerns may influence the intensity and form of threat appraisal, efforts to seek reassurance, anger, withdrawal, or ambivalence about reconciliation. A person with a history of insecurity can respond reasonably to actual evidence, and a person previously secure can experience profound distress after serious deception. The central methodological obligation is to measure the pre-event relationship, the occurrence and severity of the violation, and the subsequent response separately. Otherwise a post-event attachment score may partly reflect the experience being studied, generating a misleading inference that insecurity caused the event or the reaction Allen et al. (2005) Karney & Bradbury (1995).
Attachment concepts are also relevant when studying relationship repair. Requests for transparency, closeness, time, or distance can serve different functions depending on the context and partners' prior experiences. A careful analysis examines whether these requests increase safety, mutual understanding, and behavioral reliability or amplify coercion and monitoring. The literature does not establish any single attachment-style formula for forgiveness, reconciliation, or permanent separation. Attachment therefore belongs in a multilevel explanatory model alongside explicit agreements, power, opportunity, communication, and observed conduct Hazan & Shaver (1987) Fincham & May (2017).
Defining Infidelity Through Agreements and Conduct
Infidelity is operationally best understood as a violation of a relationship's understood terms governing intimacy or exclusivity. Sexual contact can constitute a violation in one partnership and permitted behavior in another. Emotional involvement, repeated intimate disclosures, secret dating, or sexually explicit digital exchanges may also violate a specific agreement. The relevant dimensions include the form of involvement, concealment, prior negotiation, knowledge of the partner's expectations, and whether an actionable commitment was broken. Researchers must report the definition supplied to participants instead of silently converting self-defined infidelity into a universal behavioral category. Blow and Hartnett identified definitional inconsistency as a primary source of noncomparability among studies Blow & Hartnett (2005a) Fincham & May (2017).
A useful measurement protocol begins with an explicit behavioral inventory and a separate agreement inventory. Questions about sexual behavior, romantic investment, personal secrecy, online interaction, and emotional attachment should not be merged into one undifferentiated item. Researchers can then ask whether the behavior was disclosed, prohibited, ambiguous, or negotiated. This separation reduces the risk of interpreting consensual nonmonogamy as cheating and of treating a consensually established boundary violation as harmless merely because it did not involve sexual intercourse. Qualitative descriptions can reveal contexts that standardized categories omit, but their interpretation still requires transparent coding rules Allen et al. (2005) Blow & Hartnett (2005a).
Behavior, disclosure, and moral evaluation are analytically distinct. A participant may report an event considered acceptable by both partners, an event concealed despite known restrictions, or an ambiguous interaction that the partners never discussed. A study measuring participants' moral judgments of hypothetical affairs cannot directly estimate actual behavior rates. Conversely, a survey restricted to extramarital sexual intercourse can describe that restricted form without determining the full prevalence of betrayal. Reliable construct definition is therefore the first requirement for interpreting any apparent association among gender, personality, relationship dissatisfaction, and infidelity Treas & Giesen (2000) Blow & Hartnett (2005b).
Forms of Betrayal Beyond Sexual Infidelity
Betrayal in close relationships can involve disclosure of confidential information, financial deception, broken commitments, concealed decisions, or an action that exposes a partner to preventable harm. Infidelity is one specific domain within this broader class. The emotional meaning of a transgression depends on the obligation at issue, its consequences, whether deception was planned, the recurrence of the behavior, and how the involved person responds when confronted. A single behavioral label cannot encode all of these dimensions. Close-relationship research identifies expectations and responsiveness as central to judging whether another person's action expresses care or disregard Rempel et al. (1985) Murray & Holmes (2009).
Secrecy can independently alter the injury. Concealment affects the betrayed person's ability to make informed decisions and can reshape memories of prior events when new information emerges. Research by Afifi, Falato, and Weiner examined the way infidelity was discovered and the relational outcomes reported by 115 college students with relevant experiences. Their identity-management approach suggests that the circumstances of discovery matter, although the retrospective student sample cannot establish how discovery method causally determines long-term recovery across relationships Afifi et al. (2001).
Public betrayal, private deception, and gradually revealed partial truths create different informational environments. A partner may need to revisit the chronology of previously shared activities and evaluate which assurances were accurate. Researchers should measure the information available at each time point and avoid presuming that repeated questioning necessarily represents pathology. The same behavior may reflect an effort to establish factual consistency, a response to uncertainty, or coercive control, depending on its function and context. Event-specific assessment and partner perspectives are necessary to distinguish these mechanisms Afifi et al. (2001) Gordon et al. (2005).
Motivations for Extradyadic Involvement
Motivational accounts include dissatisfaction, perceived neglect, anger, attraction, desire for novelty, sexual opportunity, low commitment, self-esteem concerns, and circumstances that lower restraint. These categories are often correlated and can be inferred retrospectively through narratives that protect self-image. Selterman, Garcia, and Tsapelas revisited motivations for extradyadic infidelity using a multidimensional approach, challenging models that compress all affairs into one reason or a single continuum of relationship unhappiness. Their work supports investigating motives as differentiated self-reports rather than treating participants' justifications as demonstrated causal mechanisms Selterman et al. (2019) Allen et al. (2005).
A subsequent study of 495 U.S. undergraduate and internet-recruited respondents associated reported motives with affair characteristics and outcomes. Partner-directed motives such as anger and reduced love were linked with longer affairs and relationship dissolution, whereas situational motives were associated with other patterns of behavior, disclosure, and experience. These associations help generate more precise hypotheses. They do not establish causal effects of motives because participants retrospectively reconstructed motives and events, selection was nonprobabilistic, and relationship outcomes may influence how motives are remembered Selterman et al. (2021).
An adequate model combines individual dispositions, relationship history, available alternatives, opportunity structures, and the local social environment. Predicting an affair from any single trait or relationship dissatisfaction score is unreliable. Personality and relationship measures are often collected from only one partner, and cross-sectional designs make directionality difficult to resolve. The field needs preregistered prospective studies that assess agreements and temptation before a possible violation, record observable circumstances, and allow for relationships in which attraction to others never leads to agreed-boundary violations Atkins et al. (2001) Fincham & May (2017).
Opportunity, Context, and Selection Effects
Opportunity is more than proximity to an alternative partner. Work schedules, travel, social networks, digital communication, privacy, and access to independent social spaces structure encounters and the chance to conceal them. In the national survey analysis by Atkins and colleagues, income and work status were among variables associated with reported extramarital sex, alongside interactions involving marital satisfaction, religious behavior, and prior divorce. These associations do not show that employment or income independently causes infidelity; they may index opportunity, social position, or unmeasured differences between respondents Atkins et al. (2001).
Opportunity effects must be distinguished from selection. People who enter certain environments differ in relationship histories, attitudes, economic resources, and partner expectations. Digital platforms similarly combine greater access to social interaction with different levels of anonymity and visibility. Simply comparing platform users with nonusers cannot isolate an effect of technology. Study designs should identify baseline commitment, preexisting relationship conditions, explicit boundaries, and actual behavior before attributing change to an app or a communication channel Blow & Hartnett (2005b) Allen et al. (2005).
Context also includes family arrangements and the potential consequences of disclosure. A participant who depends financially on a partner may face different incentives to conceal an affair or to report it in a survey than someone with independent resources. Relationship science requires analysis of power and material constraints without interpreting them as universal psychological motives. A precise account distinguishes who has opportunity, who bears the risk, what behavior occurred, and which reports are observable to investigators. These distinctions are essential for translating correlational population findings into valid developmental or contextual hypotheses Karney & Bradbury (1995) Atkins et al. (2001).
Gender, Culture, and the Limits of Generalization
Reported gender differences in infidelity depend on the study period, age, partnership type, definition of infidelity, and social costs of disclosure. Earlier studies often privileged heterosexual marriage and sexual intercourse, making their findings difficult to apply to younger dating cohorts or consensually nonmonogamous and LGBTQ+ relationships. A classification based only on the respondent's sex cannot capture the agreements defining the partnership or the gendered pressures that shape both behavior and reporting. The methodological and substantive reviews of Blow and Hartnett identify such scope and sampling limitations as persistent concerns Blow & Hartnett (2005a) Blow & Hartnett (2005b).
Culture influences the moral language of loyalty, the meaning of exclusivity, privacy expectations, and the availability of family or institutional support after betrayal. Cross-national comparisons require measurement equivalence and transparent sampling rather than assuming that the same English-language item has an identical meaning everywhere. Different communities can recognize distinct behaviors as disloyal while sharing a concern for consent and credible commitment. Treating culture as one fixed characteristic of a nationality risks replacing mechanism with stereotype; a stronger approach measures specific norms, agreements, constraints, and relationship practices Finkel et al. (2017) Allen et al. (2005).
Empirical synthesis should also account for cohort and historical change. Findings from 1990s surveys cannot silently be projected onto 2026; new forms of communication require updated behavioral inventories, while the underlying need to know whether a partner keeps meaningful commitments remains pertinent. Large population samples can estimate conditional frequencies, intensive dyadic studies can explain how trust changes over time, and qualitative studies can identify categories overlooked by prior instruments. Together these approaches offer a more adequate foundation than one unqualified claim about how often or why people cheat Treas & Giesen (2000) Blow & Hartnett (2005a).
Discovery, Disclosure, and the Reorganization of Meaning
The discovery of a previously concealed affair can alter the interpretation of an entire relationship history. The harmed partner receives new information about past behavior, the reliability of previous assurances, and the degree to which mutual expectations were shared. This is a process of belief revision involving facts, autobiographical memory, identity, and future planning. Afifi, Falato, and Weiner studied discovery method as a potential correlate of relational outcomes among college students who had experienced infidelity; their findings make the manner of disclosure a legitimate research variable rather than a mere logistical detail Afifi et al. (2001).
Discovery occurs through several pathways: voluntary disclosure, confrontation, a third-party report, digital evidence, or gradual accumulation of inconsistencies. Each pathway changes the information available and the participant's perceived ability to make an informed choice. Voluntary disclosure can signal accountability, but its meaning depends on completeness, timing, prior deception, and subsequent conduct. Confrontation can yield information but may also intensify defensiveness or fear. Retrospective samples capture how people narrate these experiences and cannot establish that one discovery pathway will produce a particular outcome in every couple Afifi et al. (2001) Gordon et al. (2005).
An event history approach should track when the extradyadic involvement began, when each partner became aware of relevant information, the dates of consequential conversations, and later decisions. Such a design distinguishes initial injury from the effects of repeated misinformation. It also makes clear that knowing the facts and accepting them emotionally can proceed on different timescales. Scientific explanations of disclosure must be attentive to safety: disclosing sensitive information in relationships marked by coercion or violence can carry specific risks that cannot be reduced to generic advice about honesty Allen et al. (2005) Fincham & May (2017).
Emotional and Cognitive Consequences of Betrayal
Confirmed betrayal can evoke anger, humiliation, sadness, fear, disgust, grief, intrusive questioning, changes in sleep, and renewed appraisal of the partner's reliability. These responses vary in duration and intensity, and the same event may produce conflicting impulses toward contact and separation. Clinical research has investigated depression and anxiety symptoms in the aftermath of humiliating marital events. Cano and O'Leary compared 25 women who had recently experienced such events with 25 women in a control condition with similar reported marital discord, finding an association with greater risk of major depressive episodes and certain distress symptoms. The small, selected sample and combined event category constrain generalization Cano & O'Leary (2000).
Descriptions of betrayal-related distress sometimes use language associated with trauma, but a severe relational violation does not by itself establish a specific psychiatric diagnosis. Researchers must assess symptoms, functional impairment, relevant diagnostic criteria, prior vulnerability, and alternative explanations before applying clinical labels. Self-reports of intrusive memories or hypervigilance can indicate meaningful suffering without meeting a diagnostic threshold. For comparable reasons, an observed improvement in distress does not necessarily mean the relationship has become safe or that the individual has forgiven the offender Gordon et al. (2005) Cano & O'Leary (2000).
Cognitive consequences include reduced confidence in recalled events, altered assumptions about exclusivity, attention to inconsistencies, and difficulty forecasting the future. These reactions often follow from uncertainty about which parts of a shared history remain reliable. Longitudinal research is needed to separate transient post-disclosure changes from durable patterns of distress, while dyadic methods can examine whether one partner's coping affects the other's behavior. Care must be taken not to reclassify evidence-based caution as insecurity merely because the person who violated the agreement finds it uncomfortable Murray et al. (2011) Karney & Bradbury (1995).
Attribution, Responsibility, and Moral Appraisal
Explanations for a betrayal involve perceived intent, controllability, repetition, awareness of boundaries, and the degree of harm. A partner may attribute an event to a situational lapse, a persistent preference for alternative relationships, hostility, or indifference to shared obligations. These interpretations influence whether an explanation is considered credible and whether renewed trust appears justified. Explanatory statements offered by the actor are data about the actor's account; they cannot establish causation or erase the effects of conduct. The multi-domain literature on extramarital involvement supports examining individual, interpersonal, and contextual factors together Allen et al. (2005) Selterman et al. (2019).
Attributions are also potentially dynamic. Early after disclosure, the betrayed partner may have limited information, and the offending partner may provide incomplete accounts. The eventual narrative can change as records, admissions, and subsequent behavior become available. Hall and Fincham analyzed attributions, forgiveness, and dissolution among 87 individuals reporting experiences of infidelity and found associations consistent with a mediation model. Because the study is observational and retrospective, a statistical mediation pathway does not prove the temporal or causal sequence by which interpretations produce separation or forgiveness Hall & Fincham (2006).
Responsibility concerns whether a person knowingly violated a commitment and what steps the person takes following the violation. Explaining preconditions such as loneliness, sexual dissatisfaction, or opportunity is compatible with assigning responsibility for a freely chosen act. Scientific analysis should therefore separate causal description from moral exculpation. It also needs to distinguish isolated violations from patterned manipulation, coercive control, and situations in which one partner faces threats for asserting boundaries. These distinctions affect the ethics and interpretation of any program designed to support reconciliation McNulty & Fincham (2012) Gordon et al. (2004).
Forgiveness as a Psychological Process
Forgiveness is commonly studied as change in a person's motivations toward someone who caused an injury, including reductions in revenge and avoidance and, in some accounts, the development of more constructive intentions. McCullough, Worthington, and Rachal's work proposed an empathy-related motivational model, while later measurement research operationalized avoidance and revenge responses to transgressions. This conceptual history provides instruments for studying response to betrayal without identifying forgiveness with simple forgetting, restored confidence, or a decision to remain in the relationship McCullough et al. (1997) McCullough et al. (1998).
Commitment can influence motivation to forgive. Finkel, Rusbult, Kumashiro, and Hannon tested this proposition using a priming experiment, survey data, and interaction records; results associated stronger commitment with forgiveness, in part through interpretations of betrayal. Multiple methods strengthen the plausibility of the mechanism, although experimental manipulations of commitment-related ideas and observations of everyday betrayals do not by themselves determine outcomes following severe, prolonged infidelity. Careful synthesis therefore differentiates laboratory transgressions, actual affair disclosure, and patterns of repeated harm Finkel et al. (2002).
Forgiveness has no fixed behavioral sequel. People may release revenge motives while ending the partnership; they may continue the partnership without having forgiven the injury; and they may decide that rebuilding the relationship would be unsafe or unacceptable. Contextual research by McNulty and Fincham shows that positive processes such as forgiving can be associated with better or worse outcomes depending on the surrounding relationship conditions. Accordingly, forgiveness should be evaluated in relation to future harm, accountability, agency, and the person's own values rather than used as an imposed endpoint of treatment McNulty & Fincham (2012) Hall & Fincham (2006).
Trust Repair as a Process of Evidence Accumulation
Rebuilding trust requires a revised expectation supported by the partner's sustained behavior. The person responsible for a violation may acknowledge the event, supply truthful information, accept boundaries, and demonstrate reliable conduct across situations where deception would again be possible. The process takes place within an interpersonal system: one person generates new evidence, while the other evaluates its consistency and significance. This conceptualization follows from models of dependability, prorelationship behavior, and risk regulation, although infidelity-specific causal trials have not established a universal sequence of repair steps Rempel et al. (1985) Wieselquist et al. (1999) Murray et al. (2006).
An apology and a promise are communicative events, not proof of future reliability. Their evidentiary value depends on whether statements are accurate, whether behavioral changes are observable, and whether the offending partner continues to respect limits after immediate pressure subsides. Transparency can support verification when freely negotiated and proportionate to the specific injury. Constant surveillance, forced access to accounts, threats, and restrictions on independent movement require a different safety analysis. Researchers must distinguish mutually agreed information sharing from coercive monitoring and should avoid treating the latter as a standard clinical technique Gordon et al. (2004) McNulty & Fincham (2012).
Repair can also involve updating relationship rules. Couples may discover that their earlier understandings of exclusivity, online intimacy, or disclosure were poorly specified. Explicit agreement clarifies what conduct will count as a future violation, while genuine repair of a known past violation requires responsibility for what was already understood. The outcome should be measured with repeated trust judgments, actual behavioral consistency, psychological distress, and the quality of mutual decision-making. The possibility of improved satisfaction must be considered separately from whether the injurious behavior has ceased Gordon et al. (2005) Atkins et al. (2005).
Relationship Continuation, Separation, and Independent Recovery
The aftermath of infidelity includes several trajectories. Some couples separate quickly, some maintain the relationship while experiencing prolonged distrust, and others report improvement after disclosure and negotiated change. No single outcome can be inferred from the presence of an affair alone. The relationship's earlier quality, commitment, investments, safety, economic circumstances, disclosure pathway, and pattern of subsequent conduct all shape what researchers observe. Longitudinal study of these variables is preferable to treating relationship status at one follow-up as a universal indicator of success or failure Karney & Bradbury (1995) Rusbult et al. (1998).
Relationship dissolution and forgiveness have been examined together, but observational links do not establish that forgiveness should be increased to prevent separation. Hall and Fincham found associations among attributions, forgiveness, and termination in a relatively small retrospective sample. Their results support studying mechanisms of interpretation and interpersonal motivation, while leaving open whether remaining together improves personal well-being. Research on context-dependent effects of positive marital processes offers a caution against adopting partnership preservation as the default goal for every participant Hall & Fincham (2006) McNulty & Fincham (2012).
Independent recovery may involve restoring self-trust, making sense of available evidence, regaining ordinary routines, maintaining social support, and establishing boundaries for future relationships. These are reasonable domains for measurement whether a couple reconciles or separates. Clinical studies frequently focus on couples attending treatment and may miss people who choose to leave, have no access to care, or are exposed to coercion. More representative prospective cohorts should follow both partners and former partners over time, with outcomes that include personal agency, safety, mental health, and social functioning Atkins et al. (2005) Gordon et al. (2004).
Methods for Measuring Trust and Relational Transgression
Trust measurement combines global judgments, partner-specific expectations, behavioral indicators, and repeated reports. The Rempel, Holmes, and Zanna framework offers items organized around predictability, dependability, and faith, while interdependence and commitment instruments measure adjacent constructs that must remain distinct. For any questionnaire, internal consistency alone cannot establish that it measures trustworthy behavior: a scale may reliably capture insecurity, relationship satisfaction, or an idealized impression. Construct validity requires convergent and discriminant tests, partner reports, and prediction of responses under independently characterized relational risks Rempel et al. (1985) Rusbult et al. (1998).
Infidelity measurement requires a behavioral definition and explicit reference period. Structured instruments should separately assess sexual, romantic, emotional, and digital involvement, the relevant relationship agreement, concealment, disclosure, and perceived harm. Anonymous reporting can reduce some social-desirability pressures but does not eliminate selection or recall bias. Prospective design is stronger for estimating temporal order; event-contingent sampling may capture actual conflicts, while partner interviews illuminate discrepancies in interpretation. Researchers should register the primary outcome and specify whether the study estimates occurrence, suspicion, perceived betrayal, or psychological reaction Blow & Hartnett (2005a) Afifi et al. (2001).
Repeated measures have particular value because trust changes through interactions. Multilevel modeling can distinguish differences between couples from fluctuations within the same couple, and actor-partner interdependence models can estimate how each person's prior assessment relates to the other's subsequent outcome. Missing data are rarely random when disclosure prompts separation or withdrawal from research. Statistical plans must handle dependent observations, attrition, measurement invariance, and competing pathways to relationship dissolution. None of these techniques automatically makes observational associations causal, and research reports should identify the assumptions necessary for a causal interpretation Karney & Bradbury (1995) Peluso et al. (2026).
Experimental and Longitudinal Evidence
Experimental studies can manipulate situational cues to rejection, commitment, or partner regard and observe changes in judgments and behavior. The risk-regulation and impulsive-trust programs use such methods to probe psychological mechanisms under controlled conditions. Their strengths include temporal ordering and clearer contrasts between specified conditions. Their limits arise when brief laboratory tasks are generalized to deception sustained over months or years, with real household, sexual-health, financial, and social consequences. A measured response to an imagined betrayal is evidence about that task, rather than direct evidence of how individuals recover from a confirmed affair Murray et al. (2011) Murray et al. (2013).
Longitudinal partner studies track adaptations and changes in satisfaction, trust, commitment, and behavior across time. Wieselquist and colleagues' studies support the idea that observed prorelationship actions and trust covary dynamically, and relationship-development reviews establish the importance of preexisting vulnerabilities and contextual stress. Prospective measures reduce some retrospective distortion, although they still require careful handling of reciprocal influence and omitted variables. Event-based designs could improve this evidence by collecting baseline measures before an occurrence, then separately recording discovery and post-event behavior Wieselquist et al. (1999) Karney & Bradbury (1995).
Large cross-study modeling offers an additional perspective on which self-report variables are robustly associated with relationship quality. Joel and colleagues pooled information across 43 longitudinal couples studies and used machine-learning methods to assess predictors, illustrating the value of cumulative datasets. These results speak primarily to general relationship-quality prediction, not a validated model for identifying infidelity or certifying that trust has been restored. Predictive performance and causal explanation remain distinct goals, particularly where sensitive labels, privacy, and unequal base rates make automated classifications risky Joel et al. (2020) Finkel et al. (2017).
Evidence on Couples Therapy After Infidelity
Clinical treatment research examines whether structured couple interventions are associated with changes in distress, communication, and relationship outcomes after infidelity. Gordon, Baucom, and Snyder reported a replicated case-study investigation of an integrative intervention with six couples who completed treatment. The program addressed emotional and cognitive reactions to affairs, and several couples showed improvement. A case series can document feasibility and clinically meaningful change, but lacks random assignment or a counterfactual group able to determine whether the intervention caused the improvement Gordon et al. (2004).
Atkins, Eldridge, Baucom, and Christensen analyzed outcomes among 19 couples with infidelity histories who were part of a broader randomized marital therapy trial involving 134 couples. Infidelity couples began more distressed, and those who disclosed an affair before or during treatment showed improvements in satisfaction relative to comparison patterns described in the study. This was an exploratory subgroup analysis, not a trial that randomly assigned betrayed couples to independent infidelity-specific treatments; subgroup size and disclosure selection complicate causal interpretation. The findings establish that change is possible in selected treatment-seeking couples without furnishing an unconditional recovery rate Atkins et al. (2005).
More recently, Peluso, Irvine, and Marquez reported a dyadic pre-post analysis of 35 couples receiving Gottman Method Couples Therapy after infidelity. Both partners showed improvements on measured outcomes, and actor-partner models identified selective associations across trust, commitment, and conflict behavior. The absence of a randomized untreated or alternative-treatment comparison prevents attribution of change to the named therapy rather than to time, selection, concurrent life changes, or regression to the mean. The sample concerns couples who completed treatment and does not establish general effects for separated couples or those in unsafe relationships. This 2026 report nonetheless advances dyadic description of the recovery process Peluso et al. (2026).
Clinical Assessment, Safety, and Ethical Boundaries
Clinical assessment after infidelity should identify the event and timing, each partner's immediate priorities, current mental-health symptoms, access to social support, and the presence of fear, threats, coercion, or other interpersonal harm. A clinician cannot responsibly assume that a joint session is appropriate solely because two partners appear together. When controlling behavior or violence is present, the safety of both disclosure and confrontation may require a different assessment pathway and specialized services. Published treatment descriptions provide useful conceptual approaches, but the research base does not justify universal prescriptions for disclosure timing or reconciliation Gordon et al. (2005) Allen et al. (2005).
Depressive and anxiety symptoms following betrayal deserve valid assessment without presuming that everyone exposed to the same event develops a mental disorder. Cano and O'Leary's case-control findings support considering serious distress while also demonstrating the small and selected samples on which some historical clinical claims rest. Researchers and practitioners should distinguish the psychological impact of betrayal from individual diagnostic categories and avoid importing trauma nomenclature as a blanket description. Clinical recommendations for a specific person must consider diagnosis, risk, context, and professional judgment Cano & O'Leary (2000) Gordon et al. (2005).
Ethically, the well-being and agency of each participant matter independently of whether the partnership survives. Therapy outcome measures should include distress, personal functioning, autonomy, and safety alongside couple satisfaction. The contextual literature on forgiveness indicates that apparently prosocial processes can have different consequences in harmful conditions. An intervention that increases superficial compliance while reducing the harmed partner's choices should not be counted as successful trust repair. Medical or psychological treatment claims require more rigorous testing than a theory of interpersonal responsiveness by itself can supply McNulty & Fincham (2012) Peluso et al. (2026).
Digital Communication and the Artificial Era
Digital communication creates additional environments in which relationship boundaries are negotiated, experienced, and sometimes violated. Private messaging, repeated intimate exchanges, image sharing, dating platforms, and persistent conversational histories can influence the opportunity for alternative relationships and the visibility of conduct. The central classification question remains whether the involved behavior conflicted with an agreement or sufficiently clear mutual expectation. Device access alone does not determine fidelity, and a partner's decision to protect ordinary personal privacy does not prove deception. Historical infidelity research establishes the importance of operational definitions, while contemporary technologies make their specification more demanding Blow & Hartnett (2005a) Fincham & May (2017).
The Artificial Era adds interactions with artificial conversational systems to the study of relational expectations. An individual may seek emotional support, imaginative intimacy, companionship, or sexualized exchange through an artificial interface. Whether a partner interprets that conduct as betrayal depends on negotiated boundaries, knowledge, secrecy, and the function the interaction performs within the relationship. Existing evidence about human–human extradyadic involvement cannot be transferred wholesale to these interactions: agency, reciprocal obligation, disclosure, and persistent memory have different operational forms. Research should identify the actual properties of the artificial system and measure human responses rather than assert equivalence by analogy Finkel et al. (2017) Murray & Holmes (2009).
Human–Artificial Relations therefore presents a research program concerning trust calibration, expectations of disclosure, partner conflict over digital behavior, and changes in the distribution of emotional attention. Prospective dyadic studies can measure agreements before use, interaction categories, perceived violation, and subsequent relationship outcomes. Ethical design requires meaningful consent and privacy protection because storing intimate conversations for research creates additional vulnerabilities. Artificial Sapiens and the philosophical horizon From Homo to Artificial supply a broader conceptual frame in this publication; empirical claims remain accountable to observable variables and replicable tests. This distinction keeps philosophical interpretation, operational hypotheses, and measured results explicit Joel et al. (2020) Fincham & May (2017).
Critical Synthesis and Unresolved Research Questions
Evidence converges on a multilevel account. Trust is shaped by prior interaction, beliefs about care, observed responsiveness, and the management of vulnerability. Betrayal changes the evidential basis of those expectations. Infidelity adds the need to define the relevant relationship agreement, the precise extradyadic behavior, and the circumstances of concealment or disclosure. Motives and outcomes vary, and associations between dissatisfaction and affairs do not support a single causal pathway. Trust repair involves the generation and interpretation of new behavioral evidence; it cannot be inferred from a momentary declaration of forgiveness or the decision to remain partnered Rempel et al. (1985) Wieselquist et al. (1999) Selterman et al. (2021).
Several weaknesses constrain stronger conclusions. Most infidelity studies use retrospective self-report; behavioral definitions vary; longitudinal pre-event data are scarce; clinical intervention samples are small and selective; and much historical work centers heterosexual marriage in Western societies. Dyadic dependence and attrition often complicate analysis. Controlled experiments illuminate general mechanisms of trust but rarely reproduce the informational and material stakes of severe betrayal. An adequate causal account needs representative prospective cohorts, pre-agreement measures, repeated event histories, independent replication, preregistered hypotheses, culturally validated instruments, and intervention studies that record harms as well as benefits Blow & Hartnett (2005a) Allen et al. (2005) Peluso et al. (2026).
Priority questions concern how people update expectations after contradictory evidence, which behaviors demonstrate trustworthy change, and how different disclosure pathways affect autonomy and distress. Additional studies should examine agreements across diverse partnership structures, the interplay of opportunity with choice, trajectories after separation, and the distinction between compassionate forgiveness and unsafe accommodation. Digital and human–artificial interactions require fresh concepts and measurements, not borrowed rates of traditional extramarital sex. Research that treats both partners as interdependent but individually situated persons will yield more informative explanations and more ethically adequate endpoints McNulty & Fincham (2012) Finkel et al. (2017).
Conclusion
Relationship trust, betrayal, and infidelity form a coherent research object when defined through credible reliance, consequential relational obligations, observable boundary violations, and the revision of expectations across time. The historical literature establishes multidimensional trust, interdependence, commitment, and risk regulation as major explanatory resources. Infidelity research adds evidence about behavior, opportunity, motivation, secrecy, and discovery, while demonstrating persistent difficulties of definition and observation. The most dependable conclusions concern the need to measure these distinct components directly rather than to treat a relationship label as a substitute for evidence Rempel et al. (1985) Finkel et al. (2017) Blow & Hartnett (2005a).
The empirical record shows that reported motives and relationship conditions are heterogeneous, that disclosure can carry substantial psychological consequences, and that some selected couples improve during treatment. It does not establish a universally effective repair protocol, a single cause of affairs, or an outcome that all affected partners should pursue. Commitment, forgiveness, continued partnership, restored confidence, and personal recovery must each be evaluated with separate measures. Future work should prioritize prospective and dyadic designs, stronger clinical comparisons, consistent operational definitions, and outcomes attentive to autonomy and safety Selterman et al. (2021) Atkins et al. (2005) Peluso et al. (2026).
The synthesis therefore identifies trust as a dynamically calibrated expectation of responsible partner conduct, betrayal as a consequential breach of relational obligation, and infidelity as a specifically defined boundary violation whose meaning depends on the relationship's actual terms. These concepts remain relevant as digitally mediated and human–artificial relationships diversify the contexts of intimacy. Scientific progress will depend on methods capable of tracing agreements, actions, interpretations, and outcomes within each relationship while preserving the evidentiary distinction among theory, association, experiment, and clinical intervention Murray et al. (2006) Fincham & May (2017).
References
- Afifi, W. A., Falato, W. L., & Weiner, J. L. (2001). Identity concerns following a severe relational transgression: The role of discovery method for the relational outcomes of infidelity. Journal of Social and Personal Relationships, 18(2), 291–308. https://doi.org/10.1177/0265407501182007
- Allen, E. S., Atkins, D. C., Baucom, D. H., Snyder, D. K., Gordon, K. C., & Glass, S. P. (2005). Intrapersonal, interpersonal, and contextual factors in engaging in and responding to extramarital involvement. Clinical Psychology: Science and Practice, 12(2), 101–130. https://doi.org/10.1093/clipsy.bpi014
- Atkins, D. C., Baucom, D. H., & Jacobson, N. S. (2001). Understanding infidelity: Correlates in a national random sample. Journal of Family Psychology, 15(4), 735–749. https://doi.org/10.1037/0893-3200.15.4.735
- Atkins, D. C., Eldridge, K. A., Baucom, D. H., & Christensen, A. (2005). Infidelity and behavioral couple therapy: Optimism in the face of betrayal. Journal of Consulting and Clinical Psychology, 73(1), 144–150. https://doi.org/10.1037/0022-006X.73.1.144
- Blow, A. J., & Hartnett, K. (2005a). Infidelity in committed relationships I: A methodological review. Journal of Marital and Family Therapy, 31(2), 183–216. https://doi.org/10.1111/j.1752-0606.2005.tb01555.x
- Blow, A. J., & Hartnett, K. (2005b). Infidelity in committed relationships II: A substantive review. Journal of Marital and Family Therapy, 31(2), 217–233. https://doi.org/10.1111/j.1752-0606.2005.tb01556.x
- Cano, A., & O'Leary, K. D. (2000). Infidelity and separations precipitate major depressive episodes and symptoms of nonspecific depression and anxiety. Journal of Consulting and Clinical Psychology, 68(5), 774–781. https://pubmed.ncbi.nlm.nih.gov/11068963/
- Fincham, F. D., & May, R. W. (2017). Infidelity in romantic relationships. Current Opinion in Psychology, 13, 70–74. https://doi.org/10.1016/j.copsyc.2016.03.008
- Finkel, E. J., Rusbult, C. E., Kumashiro, M., & Hannon, P. A. (2002). Dealing with betrayal in close relationships: Does commitment promote forgiveness? Journal of Personality and Social Psychology, 82(6), 956–974. https://doi.org/10.1037/0022-3514.82.6.956
- Finkel, E. J., Simpson, J. A., & Eastwick, P. W. (2017). The psychology of close relationships: Fourteen core principles. Annual Review of Psychology, 68, 383–411. https://doi.org/10.1146/annurev-psych-010416-044038
- Gordon, K. C., Baucom, D. H., & Snyder, D. K. (2004). An integrative intervention for promoting recovery from extramarital affairs. Journal of Marital and Family Therapy, 30(2), 213–231. https://doi.org/10.1111/j.1752-0606.2004.tb01235.x
- Gordon, K. C., Baucom, D. H., & Snyder, D. K. (2005). Treating couples recovering from infidelity: An integrative approach. Journal of Clinical Psychology, 61(11), 1393–1405. https://doi.org/10.1002/jclp.20189
- Hall, J. H., & Fincham, F. D. (2006). Relationship dissolution following infidelity: The roles of attributions and forgiveness. Journal of Social and Clinical Psychology, 25(5), 508–522. https://doi.org/10.1521/jscp.2006.25.5.508
- Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as an attachment process. Journal of Personality and Social Psychology, 52(3), 511–524. https://doi.org/10.1037/0022-3514.52.3.511
- Joel, S., Eastwick, P. W., Allison, C. J., Arriaga, X. B., Baker, Z. G., Bar-Kalifa, E., Bergeron, S., Birnbaum, G. E., Brock, R. L., Brumbaugh, C. C., Carmichael, C. L., Chen, S., Clarke, J., Cobb, R. J., Coolsen, M. K., Davis, J., de Jong, D. C., Debrot, A., DeHaas, E. C., ... Wolf, S. (2020). Machine learning uncovers the most robust self-report predictors of relationship quality across 43 longitudinal couples studies. Proceedings of the National Academy of Sciences, 117(32), 19061–19071. https://doi.org/10.1073/pnas.1917036117
- Karney, B. R., & Bradbury, T. N. (1995). The longitudinal course of marital quality and stability: A review of theory, method, and research. Psychological Bulletin, 118(1), 3–34. https://doi.org/10.1037/0033-2909.118.1.3
- McCullough, M. E., Rachal, K. C., Sandage, S. J., Worthington, E. L., Jr., Brown, S. W., & Hight, T. L. (1998). Interpersonal forgiving in close relationships: II. Theoretical elaboration and measurement. Journal of Personality and Social Psychology, 75(6), 1586–1603. https://doi.org/10.1037/0022-3514.75.6.1586
- McCullough, M. E., Worthington, E. L., Jr., & Rachal, K. C. (1997). Interpersonal forgiving in close relationships. Journal of Personality and Social Psychology, 73(2), 321–336. https://doi.org/10.1037/0022-3514.73.2.321
- McNulty, J. K., & Fincham, F. D. (2012). Beyond positive psychology? Toward a contextual view of psychological processes and well-being. American Psychologist, 67(2), 101–110. https://doi.org/10.1037/a0024572
- Murray, S. L., & Holmes, J. G. (2009). The architecture of interdependent minds: A motivation-management theory of mutual responsiveness. Psychological Review, 116(4), 908–928. https://doi.org/10.1037/a0017015
- Murray, S. L., Gomillion, S., Holmes, J. G., Harris, B., & Lamarche, V. (2013). The dynamics of relationship promotion: Controlling the automatic inclination to trust. Journal of Personality and Social Psychology, 104(2), 305–334. https://doi.org/10.1037/a0030513
- Murray, S. L., Holmes, J. G., & Collins, N. L. (2006). Optimizing assurance: The risk regulation system in relationships. Psychological Bulletin, 132(5), 641–666. https://doi.org/10.1037/0033-2909.132.5.641
- Murray, S. L., Pinkus, R. T., Holmes, J. G., Harris, B., Gomillion, S., Aloni, M., Derrick, J. L., & Leder, S. (2011). Signaling when (and when not) to be cautious and self-protective: Impulsive and reflective trust in close relationships. Journal of Personality and Social Psychology, 101(3), 485–502. https://doi.org/10.1037/a0023233
- Peluso, P. R., Irvine, T., & Marquez, A. (2026). Couples satisfaction, relationship status, and symptom change after Gottman method couples therapy for infidelity: A dyadic analysis. Psychotherapy Research. Advance online publication. https://doi.org/10.1080/10503307.2026.2689634
- Rempel, J. K., Holmes, J. G., & Zanna, M. P. (1985). Trust in close relationships. Journal of Personality and Social Psychology, 49(1), 95–112. https://doi.org/10.1037/0022-3514.49.1.95
- Rusbult, C. E., Martz, J. M., & Agnew, C. R. (1998). The Investment Model Scale: Measuring commitment level, satisfaction level, quality of alternatives, and investment size. Personal Relationships, 5(4), 357–387. https://doi.org/10.1111/j.1475-6811.1998.tb00177.x
- Selterman, D., Garcia, J. R., & Tsapelas, I. (2019). Motivations for extradyadic infidelity revisited. The Journal of Sex Research, 56(3), 273–286. https://doi.org/10.1080/00224499.2017.1393494
- Selterman, D., Garcia, J. R., & Tsapelas, I. (2021). What do people do, say, and feel when they have affairs? Associations between extradyadic infidelity motives with behavioral, emotional, and sexual outcomes. Journal of Sex & Marital Therapy, 47(3), 238–252. https://doi.org/10.1080/0092623X.2020.1856987
- Treas, J., & Giesen, D. (2000). Sexual infidelity among married and cohabiting Americans. Journal of Marriage and Family, 62(1), 48–60. https://doi.org/10.1111/j.1741-3737.2000.00048.x
- Wieselquist, J., Rusbult, C. E., Foster, C. A., & Agnew, C. R. (1999). Commitment, pro-relationship behavior, and trust in close relationships. Journal of Personality and Social Psychology, 77(5), 942–966. https://doi.org/10.1037/0022-3514.77.5.942