Naive Realism: Why Your Own View Feels Like Plain Fact

Naive realism is the assumption that your perception of a situation is a window rather than an interpretation. Lee Ross and Andrew Ward, who gave the pattern its name in a 1996 chapter, set it out as three linked beliefs: that you see events as they are; that other open-minded people exposed to the same events will come to see them the way you do; and that anyone who still disagrees is therefore either missing information, incapable of reasoning it through, or distorting the picture to serve an interest. Philosophers use the same phrase for a position about perception, the view that the senses give direct access to the world; this page is about the social psychology sense, which is about disagreement rather than about seeing.

The mechanism is that interpretation is invisible to the person doing it. You have access to the conclusion your mind produced, not to the steps that produced it, so the conclusion arrives feeling like a report of the facts. Two people reading the same paragraph bring different background knowledge, different ideas about which details matter, and different guesses about what the writer left out, and each ends up with an account that feels like simple reading comprehension. Neither is aware of having filled anything in.

The first two of the three beliefs are mostly harmless. The damage is in the third. Once a disagreement can only be explained by ignorance, stupidity or bad faith, the useful questions stop: there is no reason to ask what the other person is looking at, because the answer is already known. Emily Pronin, Thomas Gilovich and Lee Ross traced this to an asymmetry in what people have access to. Judging yourself, you consult your own intentions and find them clean; judging someone else, you have only their conclusion, and a conclusion you disagree with looks like evidence of a motive.

The clearest fingerprint of naive realism is the hostile media effect: partisans on both sides of an issue watch the same coverage and each side reports that it was slanted against them. The effect is strongest in the people most involved with the issue, which is also where it does the most harm, because those are the people negotiating, voting and arguing. It does not require anyone to be unreasonable. It only requires each side to treat its own reading as the baseline.

In 1985 Robert Vallone, Lee Ross and Mark Lepper published "The hostile media phenomenon" in the Journal of Personality and Social Psychology. They recruited Stanford students who had identified themselves as pro-Israeli, pro-Arab or neutral, and showed all of them the same set of television news segments covering the September 1982 massacre at the Sabra and Shatila refugee camps in Beirut. Everyone watched the identical tape, in the same room, with the same instructions.

Afterward, each group reported that the coverage had been hostile to its own side. Pro-Israeli students said the segments had treated Israel unfairly; pro-Arab students said the same segments had been slanted toward Israel. Asked to count the references to Israel that had been favorable and unfavorable, the two groups returned tallies that barely overlapped: the pro-Israeli viewers reported 57 percent of the references as unfavorable and 16 percent as favorable, while pro-Arab viewers reported 26 percent as unfavorable and 42 percent as favorable. Both groups also predicted that a neutral viewer would come away with a worse impression of their own side than they had held going in.

The precedent was older. In 1954 Albert Hastorf and Hadley Cantril screened a film of a rough Princeton versus Dartmouth football game to students at both colleges and asked them to count the infractions. Princeton students saw Dartmouth commit roughly twice as many as Dartmouth students saw. A separate questionnaire asked which team had started the rough play: 86 percent of Princeton students said Dartmouth, while a slim majority of Dartmouth students said both sides had. Hastorf and Cantril concluded that the two groups had not seen the same game in any useful sense.

Emily Pronin, Thomas Gilovich and Lee Ross brought the pieces together in a 2004 paper in Psychological Review. Across several studies, people rated themselves as less susceptible to common biases than the average person, and explained their own views by reference to evidence while explaining opposing views by reference to the other side's motives. Richard Perloff's 2015 review of three decades of hostile media research found the effect across many topics and many countries, and found it largest among the people most involved with the issue at hand.

Naive realism turns up wherever two people with different starting points read the same material.

The counter is not to doubt everything you perceive. It is to treat your reading as a reading, and to make the other person's reading visible before judging it.

• Ask what they are looking at, not why they are wrong. "What in the text made you read it that way?" produces the other person's evidence, which is the only thing that can change either view.
• State the other position well enough that someone holding it would accept the statement. If you cannot, you do not yet know what you disagree with.
• Count before you characterize. In a disputed article, a document or a meeting, tally claims, sources and numbers per side. A count is harder to read in two ways than an impression is.
• Expect the asymmetry. You will feel more objective than the other person, and so will they. Treating that feeling as information about the situation is the error; treating it as a reliable feature of being human is useful.
• Watch for matched complaints. When both sides call the same referee, article or manager biased against them, what that establishes is that each side is reading the same material differently. It does not establish that the thing was even. Go to the record and compare the competing claims against it.

Several biases describe the gap between how people see themselves and how they see others. They differ in what is being misjudged.

• It is not the bias blind spot, which is the specific finding that people rate themselves as less biased than others. The blind spot is one measurable consequence of naive realism, not the whole assumption.
• It is not the false consensus effect, which is overestimating how many people share your view. Naive realism is about what you conclude when you discover that they do not.
• It is not naive cynicism, the expectation that other people are more self-interested than they are. Naive cynicism explains the other side by motive as a default; naive realism arrives there only after the disagreement turns up.