Between the Truths · 01
Christian Hardinghaus · October 2026
Sixty percent agree. Forty percent disagree. Which figure would you put in the headline?
Imagine a survey in which exactly 100 people answer a question with yes or no. Nobody abstains. Sixty say yes; forty say no. An editor could write either of these headlines.
Version A
“Clear majority supports the proposal”
Version B
“Four in ten reject the proposal”
Invented illustration; not the result of an actual survey.
Both statements fit the figures. Yet one tells a story of support and the other a story of resistance. Before we even know what the proposal is, a perspective has been established.
What framing means
Framing places an issue within an interpretive context. In his 1993 account, Robert M. Entman described how communication makes particular aspects stand out, encouraging a certain definition of a problem or an evaluation of it. This goes beyond loaded vocabulary: selection, context and emphasis matter too. [1]
No number has been invented in our example. But the choice of emphasis determines what we encounter first. Add “already” to the first headline, and another suggestion appears: “Already, 60 percent support the proposal.” That sounds like a trend. A single survey does not establish that support has risen. We would need a comparable earlier result.
Why the frame can matter
In decision experiments published in 1981, Amos Tversky and Daniel Kahneman demonstrated that different formulations of the same problem could shift participants’ preferences. Their research included choices in which outcomes were described as gains or losses. [2]
This does not prove that every headline produces the same response in every reader. Analysing a text tells us which perspective it offers. How particular people respond is a separate question.
A perspective is not automatically a deception
No report can tell everything at once. This explanation also selects examples and sets priorities. The useful question is whether a frame fits the evidence or distorts it.
For our invented survey, we would need to know who took part, what they were asked and how they were selected. Without that information, the respondents’ answers cannot simply be presented as the view of an entire population. A statement may be narrowly correct while suggesting much more than its evidence supports.
Nor are apparently opposing labels necessarily equally accurate descriptions. Whether an incident involved “protest” or “violence”, for example, depends on what actually happened. Critical reading does not require inventing an equally plausible opposite account for every statement. It requires checking the account against the evidence.
Three questions to ask
- What is foregrounded, and what recedes into the background?
- What judgment disappears when I remove words such as “already”, “only” or “finally”?
- Do the original data support the headline, even when it agrees with my own opinion?
A small experiment of your own
Choose a headline you find persuasive. Rewrite it without changing its verifiable information. Describe the same figure from the perspective of support and then from the perspective of opposition. Now read the full article. Which version does its contents greater justice?
This exercise will not make us perfectly impartial. But it can interrupt the quick passage from a piece of information to a ready-made story. That is a useful moment to look more closely.
Explore the mechanism
Framing Effect · Loaded Language
Sources
- Robert M. Entman (1993): Framing: Toward Clarification of a Fractured Paradigm. Journal of Communication 43(4), pp. 51–58.
- Amos Tversky and Daniel Kahneman (1981): The Framing of Decisions and the Psychology of Choice. Science 211(4481), pp. 453–458.
Examples without a source reference are original illustrations.
