Perspective-taking: seeing it their way

**Perspective-taking** is the deliberate attempt to model another person's mental state — their beliefs, needs, and constraints — accurately enough to respond to what is actually there rather than to your assumption of it. It is not the same as agreeing, validating, or excusing. The Arbinger Institute distinguishes between an **outward mindset** (genuinely seeing others' needs first) and an **inward mindset** (treating others as obstacles or irrelevancies). Perspective-taking is the cognitive move that makes the outward mindset possible. Without it, every 'empathetic' response is aimed at a caricature.

**Perspective-taking** is imagining someone else's viewpoint from the inside — reasoning about what they probably think or feel. **Perspective-getting**, a term used by **Eyal, Steffel & Epley (2018)**, means simply asking the other person directly. Their research found that perspective-taking alone performs only marginally better than chance when predicting how someone actually thinks or feels, while perspective-getting — asking — closes the gap dramatically. The practical upshot: your mental model of another person is usually more wrong than you think, and _a single honest question corrects more than an hour of introspection_.

Yes, and the evidence is clear on this. **Eyal, Steffel & Epley (2018)** showed that trying to imagine another's view often produces **overconfidence** — you feel more certain you understand, while actually being no more accurate. This is especially dangerous in long relationships, where familiarity breeds the illusion of knowing. The more comfortable you are with someone, the more you rely on your mental model and the less you check it against reality. Perspective-taking without verification is a **self-sealing error**: it feels like understanding but insulates you from feedback.

**Naive realism**, a concept associated with psychologist **Lee Ross**, is the tendency to believe that you perceive reality directly and objectively — and therefore that anyone who sees things differently is either uninformed, irrational, or biased. It fuels unnecessary conflict because it makes disagreement feel like a character flaw in the other person rather than a natural consequence of two people filtering experience differently. **Jones-Fosu** (*I Respectfully Disagree*) argues that most relational conflict is a naive-realism problem in disguise: both parties are certain they hold the neutral, factual view, and the conversation stalls because neither examines that assumption.

The **Arbinger Institute** (*The Outward Mindset*) describes two fundamental orientations: an **inward mindset** treats other people as tools, obstacles, or irrelevancies — their needs only register insofar as they affect you. An **outward mindset** genuinely sees others' needs, objectives, and constraints as real and worth accounting for _before_ asserting your own. Perspective-taking is the mechanism that makes the outward mindset real rather than performative. Without an accurate read of what someone actually needs, you can claim to care about them while responding entirely to your invented version of them.

Because conflict activates the brain's threat-detection system, which narrows attention and pulls focus inward. **William Ury** (*Possible*) describes this as 'going to the balcony' — strategic empathy requires the most effort precisely when it feels least natural: in moments of high tension, frustration, or personal stakes. When you most need an accurate read of the other person, your capacity to form one is at its lowest. This is why Ury recommends treating the pause itself as a skill — stepping back _before_ responding is not avoidance, it is the prerequisite for any useful response.

Everyone processes incoming experience through a set of internal filters. In NLP terminology, we **delete** most of what is happening (attention is selective), **generalise** from past experience (one bad outcome shapes future expectations), and **distort** what we perceive to fit existing beliefs. No two people's filters are identical, so no two people occupy exactly the same version of a shared event. **Pat Wadors** (*Unlock Your Leadership Story*) frames it simply: any single perspective is necessarily incomplete. A map is not the territory — and your map of someone else's experience is several steps further removed than your map of your own.

No. This is the most important boundary to hold. **Perspective-taking is an accuracy tool, not a moral position.** Understanding why someone acted the way they did — their pressures, their history, their fears — does not commit you to endorsing the action, excusing the harm, or abandoning your own position. In negotiation, **G. Richard Shell** (*Bargaining for Advantage*) argues that identifying what the other side actually wants, before proposing anything, is a purely strategic move that makes your own position stronger, not weaker. You can hold a firm line and understand the other side simultaneously; in fact, the firm line is more defensible when it is responding to something real.

Before responding, run three checks: (1) **What do I actually know** versus what am I assuming? Name the assumption explicitly. (2) **What might they be trying to protect** — a boundary, a value, an identity? People rarely defend a position; they defend what the position represents to them. (3) **Have I asked them** rather than guessing? A direct question like 'help me understand what matters most to you here' costs almost nothing and, per **Eyal, Steffel & Epley (2018)**, returns far more accuracy than any amount of imagining. See also [how to stop assuming what others think](/en/blog/stop-assuming-what-others-think) for a fuller step-by-step.

A personal CRM like Endearist lets you record what someone told you mattered to them — their concerns, their circumstances, their stated preferences — so you can refer back to their actual words rather than your reconstructed memory of them. Memory compresses and distorts over time, which means even the best perspective-taker is working from a degraded mental model after a few months. Writing it down creates a stable external record of what the person actually said, not what you later decided they probably meant. This is perspective-getting preserved across time: you asked, they answered, and the record holds the answer honest.