Perspective taking in conversation is the process of representing what another person may know, believe, intend, feel, or experience while interpreting and responding to that person. It is central to everyday communication because speakers rarely possess identical information, social histories, expectations, or goals. The crucial scientific distinction is between constructing a plausible account of another mind and determining whether that account is accurate. An inference may feel empathetic and remain mistaken; a brief question may supply the missing information. The practical objective is therefore to make interpersonal understanding progressively more accurate through attentive inference, curiosity, direct information, and correction.
Research distinguishes perspective taking, perspective getting, empathic accuracy, and perceived understanding. Perspective taking uses available knowledge and imagination to reason about another person's viewpoint. Perspective getting obtains relevant information from that person or, where appropriate, from direct experience of the situation. Empathic accuracy assesses correspondence between inferred mental content and a criterion, commonly the person's own reported thoughts or feelings. Perceived understanding concerns how understood someone feels. These processes can reinforce one another, but they measure different outcomes (Eyal, Steffel, and Epley, 2018; Vorauer, Hodges, and Hall, 2025).
What Perspective Taking Means in a Conversation
A working psychological definition
Perspective taking involves tracking distinctions between one's own and another person's access to information, beliefs, interpretations, preferences, and emotional significance. In a conversation about a missed deadline, one participant may perceive a lack of respect while another is describing an unexpected emergency. Understanding requires recognizing that each is acting on a different information set and that each interpretation remains open to revision. This is a context-dependent cognitive and interpersonal process, rather than a personality diagnosis or a guarantee of goodwill.
The process can be deliberate, as when a person asks how an event might look from a colleague's position, or relatively spontaneous, as when listeners adjust a reference to what their conversation partner can see. In both cases, the mind uses incomplete cues. The speaker's actual mental state remains partially private; the listener's inference is a hypothesis supported to varying degrees by language, context, prior knowledge, and subsequent clarification.
Perspective taking, empathy, mentalizing, and accuracy
Cognitive perspective taking concerns inferences about another person's beliefs, intentions, knowledge, and interpretation. Affective perspective taking concerns inferences about another person's emotional experience. Emotional empathy and empathic concern refer to partly different affective responses to another person's experience. Mentalizing or theory of mind encompasses the broader capacity to represent mental states, whereas empathic accuracy names the degree of correct inference in a particular task. The distinctions matter because emotional resonance, confidence, and correctness can diverge (Eyal, Steffel, and Epley, 2018; Vorauer, Hodges, and Hall, 2025).
A person might correctly recognize that a colleague feels disappointed without sharing the disappointment; might feel compassion while incorrectly identifying the cause; or might accurately predict what a customer prefers while expressing little warmth. Neither warmth nor verbal fluency is a substitute for evidence of accuracy. Likewise, an outwardly reserved or differently expressive participant may possess considerable knowledge of the other person's point of view.
Whose perspective is being represented?
At least three perspectives can matter in a single exchange: the speaker's private perspective, the listener's private perspective, and the common ground jointly established through conversation. The common ground includes mutually available facts and an evolving record of what has been clarified. It is not equivalent to assuming that both participants interpreted the facts identically. In referential communication experiments, people sometimes initially consider objects invisible to the speaker and only subsequently correct their interpretation using common-ground information (Keysar, Barr, Balin, and Brauner, 2000).
Consider the instruction, “Move the small glass,” when several glasses are visible to the listener but only some are visible to the speaker. Efficient comprehension requires representing which referents the speaker could have intended. In ordinary discussion, the equivalent ambiguity may concern which meeting, promise, insult, or proposed remedy is being referenced. An apparently simple statement can have several plausible meanings until participants establish a shared referent.
How People Infer Another Person's Perspective
Egocentric anchoring and adjustment
One influential account proposes that people begin from their own interpretation and then adjust for what another person might know or experience. Across five studies, Epley and colleagues found patterns consistent with initial egocentric anchoring and often insufficient correction. Time pressure increased certain egocentric biases, whereas incentives for accuracy reduced them in the studied tasks (Epley, Keysar, Van Boven, and Gilovich, 2004).
The model explains an ordinary communication failure: because one's own reasons are vivid and readily accessible, it is easy to underestimate how little of that reasoning another person has heard. A manager knows why a decision was made and overestimates how self-evident the decision should be to an employee. A friend remembers a subtle change of tone and assumes everyone noticed its cause. The error arises partly from the unequal availability of information, and correction requires deliberate attention to that asymmetry.
The anchoring-and-adjustment framework is useful, but it should not be treated as the only explanation of conversational behavior. Some research emphasizes the coordinated, sequential work of actual interacting partners, which can be missed when experiments isolate millisecond-level processing. Immediate egocentric attention and later successful joint understanding can coexist. A laboratory eye movement is evidence about one phase of comprehension, not a complete description of relationship competence (Keysar et al., 2000; Perspective taking in conversation: a defense of speaker non-egocentricity, 2013).
Social prediction and the illusion of transparency
People can overestimate how visibly their private thoughts and feelings are expressed. Research on the illusion of transparency found that participants sometimes believed their internal states could be detected more readily than observers actually detected them (Gilovich, Savitsky, and Medvec, 1998). This bias has two conversational consequences: speakers may withhold explicit explanation because they expect to be understood, while listeners may assume that visible behavior reveals more than it can reliably reveal.
A long pause can reflect uncertainty, fatigue, careful reflection, disagreement, processing differences, or an effort to formulate a safe response. None is uniquely encoded in the pause. Similarly, eye contact, vocal intensity, facial expression, and conversational speed are probabilistic, culturally and situationally variable cues. Treating a single cue as a diagnostic window into hidden motives can amplify error. The defensible inference incorporates multiple sources and remains revisable.
Metaperception: what I think you think about me
Conversation also involves metaperception: judgments about how another person sees us. After presenting an idea, a speaker may think, “She considers me incompetent,” even though the listener is trying to understand a technical detail. A 2025 integrative scientific review connects accuracy about others' thoughts and feelings with the accuracy of these self-relevant metaperceptions, highlighting systematic biases that can disrupt interaction (Vorauer, Hodges, and Hall, 2025).
Metaperceptive uncertainty matters especially when a situation involves status, exclusion, reputation, or the possibility of rejection. An attempt to infer another person's motives can quietly become an attempt to resolve one's fear of being judged. Asking “What concerns do you have about the proposal?” yields more relevant information than mentally completing the sentence “You must think I am careless.” Even an accurate answer may be emotionally uncomfortable, yet the inference can now be evaluated against an actual response.
Motivation and selective attention
People do not seek accuracy with equal intensity in every situation. Attention is distributed according to goals, incentives, self-protection, perceived threat, and relationship expectations. Studies of couples indicate that accuracy about a partner's thoughts can have different consequences depending on whether those thoughts are relationship-threatening. In one videotaped marital-conflict study, higher accuracy about threatening content was associated with lower immediate subjective closeness, whereas accuracy about nonthreatening content showed the opposite pattern (Simpson, Oriña, and Ickes, 2003).
These findings caution against equating maximum information about another person's private mind with immediate relational benefit. Accuracy may expose a genuine problem that requires attention, boundaries, or time. It does not follow that deliberate misunderstanding is an appropriate general strategy. The relevant distinction is between being able to receive accurate information, choosing a safe time to seek it, and deciding how to respond after receiving it.
Does Perspective Taking Make People More Accurate?
The strongest direct experimental challenge
The popular advice to “put yourself in their shoes” suggests that simulation should improve the ability to predict what someone thinks or wants. Eyal, Steffel, and Epley directly tested this proposition across 25 experiments involving a range of mental-state judgments and interpersonal prediction tasks. Instructions to take another person's perspective produced no consistent gain in accuracy and sometimes increased confidence without improving performance. In a final experiment, directly acquiring the other person's perspective through conversation improved accuracy, whereas simply imagining the perspective did not (Eyal, Steffel, and Epley, 2018).
This result has a precise scope. It challenges the general claim that an instruction to imagine someone's viewpoint reliably increases factual insight. It does not establish that perspective taking has no emotional, motivational, cooperative, or relational value. It also does not imply that every question is safe, candidly answered, or sufficient. The studies distinguish an imagined account from information obtained from a real person, and the latter can alter the evidence available for the judgment.
Perspective getting and the information advantage
Perspective getting consists of asking someone what they think, feel, prefer, or mean and using the answer to update an inference. In three experimental studies, Damen, Pollmann, and Grassow examined willingness to seek another person's perspective, predictive accuracy, and obstacles to directly asking. Their research reinforces the value of obtaining person-specific information and documents circumstances in which people fail to use that opportunity (Damen, Pollmann, and Grassow, 2021).
In daily life, asking “What was most difficult about that conversation?” can uncover concerns that “I know exactly how you felt” prematurely closes. Information seeking is most effective when the question is understandable, the other person has a meaningful choice about disclosure, and the listener genuinely updates their response. A question asked only to confirm an existing verdict adds little. A question answered under intimidation may provide distorted evidence because the person cannot safely disagree.
Experiencing the situation versus reading expression
Perspective information can also be obtained by encountering some of the same circumstances. In five experiments, Zhou, Majka, and Epley compared inferring strangers' reactions from facial expressions with seeing the images those strangers viewed. Access to the situation often yielded better predictions in that experimental design, while perceivers underestimated its informational value (Zhou, Majka, and Epley, 2017).
The implication is methodological: situation-specific evidence may be richer than a snapshot of visible behavior. It would be an overgeneralization to assume that similar external exposure produces identical subjective experience. History, physical conditions, culture, neurodevelopment, incentives, and social roles still differ. Direct experience can supplement conversation, but the person who lived through an event remains a crucial source concerning its personal meaning.
Accuracy, Confidence, and Feeling Understood
A correct inference has a criterion
Empathic accuracy research often asks one participant to infer thoughts or feelings that another participant reports having at particular moments in an interaction. For example, partners may watch a recording, mark significant thoughts, and later compare inferred mental content with the target's account. The method is more informative about judgment accuracy than simply asking whether the observer considers themself empathic (Simpson, Oriña, and Ickes, 2003; Vorauer, Hodges, and Hall, 2025).
The criterion has limits. Self-reported thoughts can be forgotten, selectively disclosed, revised after an event, or shaped by social expectations. Agreement with a report offers a usable but imperfect benchmark for private experience. Behavioral prediction tests and independently observable facts offer different kinds of accuracy criteria. Good research states what was predicted, how the benchmark was constructed, when measurement occurred, and whether the task resembled real conversation.
Confidence and subjective understanding can diverge
A person may speak with great confidence about another person's emotional state and be incorrect. Conversely, a cautious listener may accurately understand important parts of an experience while openly naming uncertainty. Accuracy and perceived responsiveness are therefore distinct outcomes. A statement that sounds validating can help someone feel heard even when details are wrong, while an analytically accurate statement can be delivered without care or permission. The conversational task requires both attention to evidence and respect for the other person's experience.
Confidence is particularly vulnerable to familiarity. Knowing someone for years gives the perceiver useful historical information, but it can also strengthen expectations that no longer fit the person's current situation. A long relationship does not eliminate the need to ask about new circumstances, preferences, values, or pressures. An accurate model of a person is a model that can change when the person changes.
Why errors can feel relationally costly
Some interpersonal consequences depend on how a mistake is framed and repaired. Underestimating someone's distress may be interpreted as indifference, while overestimating it may be experienced as intrusive or patronizing. There is no universally correct degree of expressed sympathy independent of the target's goals and feedback. A useful repair acknowledges the inferential error and requests correction: “I assumed you were angry. What was actually going on for you?”
Understanding can also reveal genuine incompatibility. Accurate discovery of a conflict of values, a boundary, or a refusal can make agreement less likely while improving the honesty of the exchange. The value of perspective taking cannot be measured solely by immediate harmony. Sound outcomes include clearer expectations, informed disagreement, cooperative adjustment, and safe disengagement where appropriate.
Benefits That Research Supports—and Their Boundaries
Negotiation and recognition of interests
Perspective taking can sometimes help people identify what another party needs from a negotiation. In three studies using negotiation tasks, Galinsky and colleagues found that cognitive perspective taking improved the discovery of hidden opportunities for agreement and some negotiation outcomes; emotionally sharing the counterpart's perspective did not show the same advantage in those tasks (Galinsky, Maddux, Gilin, and White, 2008).
The difference between positions and interests is central. A colleague's demand for a particular schedule may reflect childcare obligations, concentration needs, access requirements, or a fairness concern. If both parties repeat only their proposed schedules, they may miss arrangements that satisfy important interests. Perspective taking can motivate hypotheses about those interests, while asking the colleague tests them. The negotiation studies establish task-specific benefits; they do not license claims of uniform advantage in high-stakes, coercive, or structurally unequal negotiations.
Greater openness and less premature certainty
Considering alternative interpretations can interrupt the assumption that one's own perception is the only plausible one. It may support curiosity, less dismissive questioning, and more flexible interpersonal judgments. These are credible process benefits even when the listener's initial guesses prove inaccurate. A systematic review of interventions addressing empathy and perspective taking in controversial discussions identifies multiple approaches and situational considerations, including offline and online formats, while underscoring the need for stronger evidence about mechanisms and transfer (Vaassen, van den Broek, Aaldering, and Beersma, 2026).
A crucial distinction is between a changed attitude toward an interlocutor and increased accuracy about that person's particular beliefs. Less hostility, greater willingness to listen, and greater predictive accuracy are separate outcomes. Research demonstrating one should not be cited as though it demonstrates all three.
Cooperation without agreement
Understanding someone's reasons can improve the quality of disagreement. A listener can explain the other's position fairly, identify a shared decision point, or separate a factual misunderstanding from a real difference in values. Cooperation may then involve an explicit trade-off, a limited compromise, an agreed procedure, or an acknowledgment of enduring disagreement. Perspective taking supplies information relevant to such choices; it does not determine which choice is ethically or practically appropriate.
In some cases, accurate understanding increases an individual's ability to negotiate strategically at another person's expense. The same skill that helps discover a mutually beneficial trade-off can help predict vulnerabilities. Ethical evaluation therefore requires examining consent, honesty, use of power, and effects on both parties. Psychological accuracy is a descriptive achievement; prosocial behavior is a separate normative and behavioral achievement.
When Perspective Taking Misleads or Backfires
Projection disguised as empathy
An imaginative answer to “How would I feel in that situation?” may primarily reveal the speaker's values, resources, prior experiences, and vulnerabilities. It may be useful as an initial hypothesis, but another person's circumstances can generate a very different reaction. The mistake becomes stronger when the observer announces certainty or overrides the other person's actual account. An especially common projection substitutes one's preferred solution for what the other person requested.
A productive alternative is to separate three statements: “This is what happened”; “Here is my tentative interpretation”; and “What does it mean to you?” That sequence protects the difference between observation, inference, and the target's self-description. The inferential step remains useful because it generates questions and possible explanations, but the conversation provides the correction mechanism.
Stereotypes and false consensus
Group stereotypes may influence what seems plausible about someone's motives, anger, attention, communication style, or loyalty. An observer may believe they have taken the other's perspective while reproducing a cultural narrative about that person's identity. Similarly, the false-consensus tendency can make one's own preferred reaction appear typical. Research on anchoring and adjustment gives one account of how self-relevant information can dominate judgments; experiments about intergroup interaction show that perspective-taking instructions can even produce counterproductive behavior under certain conditions (Epley et al., 2004; Vorauer, Martens, and Sasaki, 2009).
An appropriate response emphasizes individual evidence over category assumptions. A shared language, diagnosis, age group, nationality, or cultural background does not determine a person's emotional meaning. When group-level experience is relevant, it should be used to frame a respectful question about the individual rather than as a shortcut that replaces the person's voice.
Cross-group encounters and impression management
In four studies of intergroup encounters, Vorauer, Martens, and Sasaki found conditions in which perspective-taking instructions led lower-prejudice participants to behave less positively, apparently in connection with self-focused concerns during interaction (Vorauer, Martens, and Sasaki, 2009). The result complicates the assumption that actively imagining another's thoughts always improves behavior. Attention can shift toward anxiety about how one appears, leaving fewer resources for responding to the actual person.
Successful interaction depends on more than an internal mental exercise. The other person's feedback, the ability to tolerate uncertainty, and the structure of the exchange matter. Asking permission, acknowledging what one knows, and allowing the interlocutor to correct an interpretation are observable alternatives to covert speculation.
Threat, intimacy, and relational consequences
When someone identifies a partner's fear, resentment, or contemplated departure accurately, the information may be consequential and painful. The marital-interaction findings of Simpson, Oriña, and Ickes illustrate that the short-term subjective benefit of accuracy depends on the content being understood (Simpson, Oriña, and Ickes, 2003). This issue connects to the broader psychology of intimate relationships, but the present article concerns the general inference process rather than couple therapy, attachment diagnosis, or relationship-repair programs.
The practical implication is pacing and consent. A person can be curious without demanding disclosure and can clarify a disagreement without insisting that the other person reveal every private feeling. Some information is appropriately confidential. Reciprocal openness is an interpersonal choice that develops within a trustworthy relationship, not a universal obligation of accurate communication.
Power, Safety, and the Right Not to Disclose
Power affects whether a person can speak frankly, ask questions, or challenge an inference. A supervisor, clinician, teacher, interviewer, or public authority may control resources that matter to the other party. The apparent availability of a conversation does not prove that disclosure is voluntary. Perspective getting is most meaningful when the listener can hear unwanted answers without retaliation and the respondent can set boundaries.
Ordinary disagreement must be distinguished from intimidation, coercive control, discriminatory threats, stalking, and interpersonal violence. In unsafe contexts, questions framed as empathic exploration may expose sensitive information or invite further manipulation. A person facing threats is entitled to prioritize safety, document relevant facts, seek trusted professional or institutional support, and disengage as appropriate. Mediation or reciprocal vulnerability should never be presumed safe merely because the conflict is described as a communication problem.
The demand to “see both sides” can also conceal asymmetrical responsibility. Understanding why someone is coercive does not make coercion acceptable. A survivor's accurate prediction of another person's anger may serve immediate self-protection and should not be interpreted as consent or equal participation in the harm. Perspective taking remains ethically situated within rights, boundaries, and accountability.
Culture, Development, Language, and Neurodevelopment
Cultural and conversational variation
People differ in what they state explicitly, how they manage disagreement, how directly they ask questions, and what counts as respectful interruption or silence. Conversational style is shaped by language background, institutions, relationships, power distance, and immediate setting. A listener who treats one cultural norm as a universal indicator of sincerity or competence is likely to misinterpret some partners. The general process of building shared meaning is widely relevant, but specific scripts need negotiation with the individuals concerned.
Perspective taking must therefore include uncertainty about the rules of the interaction itself. “Would you prefer a direct question or time to reflect?” can provide information that nonverbal observation alone cannot. Shared ground may be established through spoken language, writing, assistive communication, translation, or other accessible modalities. Neither fluent speech nor sustained eye contact is a universal requirement for valid understanding.
Development and learning
Children gradually develop the ability to track different viewpoints, while adults continue to encounter errors in rapid comprehension. An eye-tracking experiment comparing children and adults suggested that both groups could initially interpret a reference egocentrically, with adults correcting more efficiently under the examined conditions (Epley, Morewedge, and Keysar, 2004). This supports a developmental focus on flexible correction and communication context, rather than treating childhood performance as a permanent trait.
Learning includes noticing when a misunderstanding occurs, seeing that another person can possess different information, and practicing appropriate repair. Yet measured success in a constrained experimental task does not automatically establish successful adaptation across school, work, family, friendship, and digital settings. Development also depends on opportunity, accessibility, reciprocal accommodation, and social safety.
Neurodevelopment and the double-empathy discussion
Differences between autistic and non-autistic communication partners can produce mutual misunderstanding. Research on the double-empathy problem treats cross-neurotype communication as a bidirectional interaction rather than assuming that every difficulty belongs to one participant. In a study of observers rating autistic, non-autistic, and mixed-neurotype conversations, mixed interactions were evaluated less favorably by outside observers; observers' judgments also differed from participants' own evaluations (Jones, Botha, Ackerman, King, and Sasson, 2024).
A 2024 empathic-accuracy study found that non-autistic participants had more difficulty judging the feelings of autistic narrators under the study's conditions, adding evidence that difficulty interpreting another person need not imply indifference or a unilateral deficit (Cheang, Skjevling, Blakemore, Kumari, and Puzzo, 2024). The evidence remains varied across experimental tasks, populations, and outcome definitions. No single study establishes a universal rule that same-neurotype pairs always understand each other better. Social-communication differences and clinical questions require their own context-sensitive assessment; difficulty taking another's perspective is not itself a psychiatric diagnosis.
Digital conversation and missing cues
Text messages, voice notes, video calls, and asynchronous platforms change the information available for inference. Delay can be mistaken for hostility, brevity for rejection, and punctuation for emotional intensity. The absence of particular cues creates ambiguity but does not license a reliable decoding system. The relevant method is to distinguish what the message literally states from how it is being interpreted, request clarification when proportionate, and consider that response time can reflect practical constraints.
A systematic review of interventions in controversial discussions describes the promise and design challenges of supporting empathy and perspective taking across online and offline settings; interventions differ and cannot be collapsed into one proven digital technique (Vaassen et al., 2026). The transferable principle is correction through better information and communication conditions, rather than confidence generated by limited signals.
How Perspective Taking Is Studied and Measured
Behavioral and laboratory methods
Experimental studies can instruct participants to adopt another perspective, ask them to predict preferences or feelings, and score predictions against predetermined criteria. Visual-world language tasks track gaze as people interpret ambiguous instructions, showing when listener and speaker viewpoints diverge in real time. Negotiation simulations test whether perspective-related instructions help identify hidden interests and improve agreements. Such designs allow specific causal comparisons, but the manipulated task and selected outcome define the scope of the result (Keysar et al., 2000; Galinsky et al., 2008).
Performance on a task that requires recognizing a facial expression is different from recognizing the personal meaning of a long conflict. Accurately naming a visible emotion is different from inferring its cause. Predicting another person's choice is different from understanding their private justification. Each study must be read according to its operational definition of perspective or accuracy.
Dyadic interaction and empathic-accuracy paradigms
A dyadic paradigm observes two people interacting rather than treating judgments as context-free individual scores. In video-recall procedures, targets identify what they thought or felt at moments in a real conversation, while perceivers attempt to infer those contents. Researchers can examine timing, relationship context, expressed cues, perceived threat, and accuracy. These methods better approximate interpersonal dynamics than isolated vignettes, while also introducing recall and self-report limitations (Simpson, Oriña, and Ickes, 2003; Vorauer, Hodges, and Hall, 2025).
Important research questions include whose accounts define accuracy, whether consent and disclosure affect those accounts, how participants respond to correction, and whether gains persist in a later encounter. Comparing self-reported empathy with performance measures is valuable because strong identification with the idea of being empathic may coexist with systematic prediction errors.
What counts as evidence that training works?
A perspective-taking exercise can improve an immediate answer on a hypothetical vignette without producing a measurable change in the next difficult conversation. Evidence is stronger when investigators measure behavior, include relevant comparison conditions, observe actual interactions, follow participants over time, and test whether gains transfer across partners and contexts. Intervention reviews suggest possibilities for improving openness and interaction, but variation in populations, intervention components, and outcomes limits claims of an all-purpose training method (Vaassen et al., 2026).
The strongest established finding for this article is the existence of predictable perspective-inference errors and the demonstrated informational advantage of directly obtaining relevant evidence in tested settings. The evidence for stable, long-term, broad generalization from brief perspective-taking instruction is more limited. Findings about negotiation and intergroup discussion are context-specific, and individual differences do not support labeling all people who communicate differently as deficient.
A Research-Informed Method for Improving Understanding
First, identify the actual question
Before inferring someone else's mind, define the uncertainty. Are you trying to know what they mean by a word, which facts they know, what outcome they want, whether they understood your proposal, or how they experienced an event? These are different questions with different evidence requirements. An assumption about feeling does not resolve an ambiguity about instructions, and an agreement about facts does not prove agreement about values.
For example, if a teammate responds “This plan won't work,” ask whether the concern is timing, cost, access, risk, or a different intended outcome. This transforms a vague interpersonal judgment into a testable question. Accurate listening is closely related but has its own skills and evidence base; see Active Listening: Skills, Examples and Psychological Evidence.
Second, label observations and interpretations separately
Describe the evidence in terms that both people could potentially verify. “You said the timeline is too short” reports the utterance. “You think I am incompetent” reports an inference. Ask what would distinguish competing interpretations and what evidence would change your mind. The goal is calibrated inference, not an elaborate internal story about hidden intentions.
The distinction helps especially when nonverbal cues are ambiguous. Understanding Social Cues: Context, Ambiguity and Interpretation examines those interpretive uncertainties in greater detail. In the present context, nonverbal information contributes a hypothesis that spoken clarification can strengthen or revise.
Third, invite perspective getting
Where safe and appropriate, ask an open question that permits correction: “Which part of this mattered most to you?” “What did you understand my request to mean?” “What outcome would feel workable?” “Have I missed a constraint?” Avoid requiring the person to confirm your emotional label. Direct information is especially useful when personal experience differs from the observer's assumptions (Eyal, Steffel, and Epley, 2018; Damen, Pollmann, and Grassow, 2021).
An invitation respects privacy: “You can tell me as much or as little as you want.” If a participant declines, the listener can still clarify practical arrangements without forcing disclosure. The value of a question depends on whether the respondent can answer honestly, without undue consequence, and whether the questioner is willing to learn.
Fourth, check comprehension without claiming ownership of the answer
A summary can be offered tentatively: “It sounds as though the unpredictability of the schedule is harder than the workload. Is that right?” A correction should be welcomed as useful information rather than resisted as a challenge to empathy. If the person says, “No, the workload itself is the problem,” the accurate response is to update the model. Feeling understood often depends on whether correction changes the conversation.
Validating someone's reported experience can coexist with disagreement about an explanation or proposed solution. Emotional Validation in Communication: Meaning and Examples addresses acknowledgment as a distinct communicative process. Perspective taking contributes to validation when it helps identify the actual experience being acknowledged.
Fifth, translate understanding into proportionate action
After clarification, ask what follows from the information. Sometimes it requires a practical change; sometimes an apology, a renegotiated expectation, a clearer explanation, or a boundary. An accurate inference that has no effect on response quality may be of limited conversational value. Conversely, a respectful decision to disagree can still demonstrate that the person's concerns were understood.
In repeated relationships, revisit prior models rather than treating them as permanent personality descriptions. Notice when a prediction was wrong and what evidence corrected it. A small record of genuine corrections is a better learning resource than a growing collection of confident interpretations. This procedure is a research-informed synthesis of accuracy findings, not a tested universal multistep clinical intervention.
Examples Across Everyday Contexts
A workplace disagreement
A department head concludes that an employee resists a software change because they dislike learning. Instead of relying on the stereotype, the head asks what parts of the proposed transition are risky. The employee identifies a missing accessibility feature and a deadline conflict. The head now possesses new information and can decide whether to change the implementation plan. The essential gain was correcting the interpretation of resistance; no claim is made that the employee must approve the final decision.
A friendship misunderstanding
One friend interprets several delayed replies as withdrawal. The other explains that unexpected caregiving responsibilities have reduced phone use. The explanation does not prove that every delay has the same cause, but it changes the evidence for the current interpretation. The friends can agree on a practical signal for urgent communication without insisting on identical expectations about response speed.
A disagreement about emotions
A listener says, “You are furious because I criticized your idea.” The speaker replies, “I am worried about the effects on the client, and I need the risk documented.” The correction shifts discussion from presumed anger to a stated concern. The listener can acknowledge the concern and work on the substantive risk. The original emotional inference was plausible but inaccurate; its usefulness lay in prompting a question rather than becoming a verdict.
When asking is unsafe
A person experiencing intimidation knows that a direct question may provoke retaliation. The scientifically useful distinction between imagined and obtained perspectives does not override safety. The person may rely on observable actions, seek external help, limit disclosure, or disengage. The relevant goal is informed and safe action, not extracting the other person's private reasons at any cost.
Evidence Strength, Limitations, and Unresolved Questions
Established: humans regularly make systematic errors when inferring others' knowledge and mental states; subjective confidence is an imperfect indicator of correctness; language interpretation depends on information asymmetries and shared context. Controlled experiments provide strong evidence for these mechanisms in specified tasks (Epley et al., 2004; Keysar et al., 2000; Eyal et al., 2018).
Established within tested paradigms: merely imagining another's perspective does not reliably improve mental-state prediction; directly acquiring person-specific information can improve accuracy. These results should be applied with attention to the reliability, consent, and context of disclosure. Negotiation tasks also demonstrate specific benefits of perspective-related reasoning when hidden interests can be discovered (Eyal et al., 2018; Galinsky et al., 2008).
Preliminary or context-dependent: interventions to encourage perspective taking in controversial discussions can improve some social outcomes, but study designs and outcome measures differ. Effects of perspective taking in intergroup contact, intimate conflicts, and mixed-neurotype communication depend on task, motivation, and participants. More work is needed on actual dialogue, disclosure safety, learning transfer, and longer-term outcomes (Vaassen et al., 2026; Vorauer et al., 2009; Jones et al., 2024).
Limited or contested: whether a short, generic perspective-taking exercise creates durable, generalized interpersonal accuracy; whether task-specific advantages can be translated into reliable universal scripts; and how best to score accuracy when private states are fluid or strategically disclosed. Contrasting research traditions sometimes emphasize individual inference speed and at other times the gradual accomplishments of interacting pairs. Progress requires measuring both moment-to-moment cognition and negotiated shared meaning (Vorauer, Hodges, and Hall, 2025; Perspective taking in conversation: a defense of speaker non-egocentricity, 2013).
Frequently Asked Questions
Is perspective taking the same as empathy?
Perspective taking is the attempt to represent another person's viewpoint, while empathy can include emotional resonance or concern. Someone can take a cognitive perspective without sharing an emotion, and can feel compassion while misidentifying what another person thinks. Empathic accuracy measures how well a particular inference matches an appropriate criterion.
Why can putting yourself in another person's shoes be wrong?
The imagined situation may be filtered through your preferences, past experiences, beliefs, and fears. The person you are trying to understand may have different information or priorities. In controlled research, instructions to imagine another viewpoint did not reliably increase accuracy, whereas directly getting information sometimes did (Eyal, Steffel, and Epley, 2018).
How can I check whether I understood someone?
Separate the person's stated words from your interpretation, ask a nonleading question if appropriate, and invite correction in your own summary. Treat the correction as evidence and update what you believe. Some experiences remain private and some uncertainty appropriately remains unresolved.
Can perspective taking improve conflict resolution?
It can help identify interests or misinterpretations and may support constructive negotiation in particular settings. The benefit depends on context, power, safety, and whether information actually becomes available. Understanding someone else's motive does not require yielding to it or tolerating coercion.
Is difficulty understanding social cues a mental disorder?
No. A difficulty in a conversation describes a behavior or experience and is not a stand-alone psychiatric diagnosis. Language, context, culture, disability, anxiety, sensory access, interpersonal history, and the other participant's communication can all influence understanding. Clinical assessment, when relevant, evaluates a broader pattern according to applicable professional criteria.
Does knowing someone well eliminate perspective-taking errors?
Familiarity supplies useful historical information but can also generate overconfident assumptions. Understanding improves when longstanding expectations are checked against new facts and the person's current account.
Related Articles
Cluster Communication, Conflict and Social Skills — The scientific foundation for conversational meaning, interpersonal understanding, and conflict processes.
Active Listening: Skills, Examples and Psychological Evidence — Attentive responding and clarification as complementary skills.
Understanding Social Cues: Context, Ambiguity and Interpretation — Why observable signals permit multiple interpretations.
Emotional Validation in Communication: Meaning and Examples — Acknowledging another person's reported experience without assuming agreement.
Social Perception and Misinterpretation in Everyday Conversation — Review attribution, expectancy, bias and interpersonal interpretation.
References
Bezuidenhout, A. (2013). Perspective taking in conversation: A defense of speaker non-egocentricity. Journal of Pragmatics, 48(1), 4–16. https://doi.org/10.1016/j.pragma.2012.11.007.
Cheang, R. T. S., Skjevling, M., Blakemore, A. I. F., Kumari, V., and Puzzo, I. (2025). Do you feel me? Autism, empathic accuracy and the double empathy problem. Autism, 29(9), 2315–2327. https://doi.org/10.1177/13623613241252320.
Damen, D., Pollmann, M. M. H., and Grassow, T.-L. (2021). The benefits and obstacles to perspective getting. Frontiers in Communication, 6, 611187. https://doi.org/10.3389/fcomm.2021.611187.
Epley, N., Keysar, B., Van Boven, L., and Gilovich, T. (2004). Perspective taking as egocentric anchoring and adjustment. Journal of Personality and Social Psychology, 87(3), 327–339. https://doi.org/10.1037/0022-3514.87.3.327.
Epley, N., Morewedge, C. K., and Keysar, B. (2004). Perspective taking in children and adults: Equivalent egocentrism but differential correction. Journal of Experimental Social Psychology, 40(6), 760–768. https://doi.org/10.1016/j.jesp.2004.02.002.
Eyal, T., Steffel, M., and Epley, N. (2018). Perspective mistaking: Accurately understanding the mind of another requires getting perspective, not taking perspective. Journal of Personality and Social Psychology, 114(4), 547–571. https://doi.org/10.1037/pspa0000115.
Galinsky, A. D., Maddux, W. W., Gilin, D., and White, J. B. (2008). Why it pays to get inside the head of your opponent: The differential effects of perspective taking and empathy in negotiations. Psychological Science, 19(4), 378–384. https://doi.org/10.1111/j.1467-9280.2008.02096.x.
Gilovich, T., Savitsky, K., and Medvec, V. H. (1998). The illusion of transparency: Biased assessments of others' ability to read one's emotional states. Journal of Personality and Social Psychology, 75(2), 332–346. https://pubmed.ncbi.nlm.nih.gov/9731312/.
Jones, D. R., Botha, M., Ackerman, R. A., King, K., and Sasson, N. J. (2024). Non-autistic observers both detect and demonstrate the double empathy problem when evaluating interactions between autistic and non-autistic adults. Autism, 28(8), 2053–2065. https://doi.org/10.1177/13623613231219743.
Keysar, B., Barr, D. J., Balin, J. A., and Brauner, J. S. (2000). Taking perspective in conversation: The role of mutual knowledge in comprehension. Psychological Science, 11(1), 32–38. https://doi.org/10.1111/1467-9280.00211.
Simpson, J. A., Oriña, M. M., and Ickes, W. (2003). When accuracy hurts, and when it helps: A test of the empathic accuracy model in marital interactions. Journal of Personality and Social Psychology, 85(5), 881–893. https://doi.org/10.1037/0022-3514.85.5.881.
Vaassen, C. E., van den Broek, T. A., Aaldering, H., and Beersma, B. (2026). Increasing empathy and perspective-taking in controversial discussions: A systematic review and future research directions. Computers in Human Behavior Reports, Article 101062. https://doi.org/10.1016/j.chbr.2026.101062.
Vorauer, J. D., Hodges, S. D., and Hall, J. A. (2025). Thought-feeling accuracy in person perception and metaperception: An integrative perspective. Annual Review of Psychology, 76, 413–441. https://doi.org/10.1146/annurev-psych-011624-024416.
Vorauer, J. D., Martens, V., and Sasaki, S. J. (2009). When trying to understand detracts from trying to behave: Effects of perspective taking in intergroup interaction. Journal of Personality and Social Psychology, 96(4), 811–827. https://doi.org/10.1037/a0013411.
Zhou, H., Majka, E. A., and Epley, N. (2017). Inferring perspective versus getting perspective: Underestimating the value of being in another person's shoes. Psychological Science, 28(4), 482–493. https://doi.org/10.1177/0956797616687124.