Trick wording

Trick wording puts the deception in the sentence. The control is ordinary and it does exactly what a control of its kind does: the checkbox stores a boolean, the toggle flips a setting, the button submits. What has been engineered is the label, so that a person who reads it at normal speed comes away believing the opposite of what their click will do. Untick this box to not stop receiving emails. Two negatives, one boolean, and a reader who has to parse a sentence like a logic exercise to keep their inbox.

Naming it precisely matters, because the neighbouring patterns are about different parts of the same row. A preselected opt-in is about the state the control starts in, and its label may be perfectly clear. Trick wording leaves the state alone and rewrites the sentence, which is why it survives audits that only check defaults, and why the fix is editorial rather than technical: one positive statement of what ticking does, no negatives, no exceptions clause.

It is one of the named types in Harry Brignull’s catalogue on deceptive.design, and it is regulated as well as catalogued. Consent obtained through a sentence a reasonable person would misread is not informed consent under the GDPR, and the same reasoning appears in unfair-practices enforcement elsewhere. That is a useful thing to be able to say in a review, because trick wording is often defended as tone or brevity rather than recognised as the dark pattern it is.

The tell is that the sentence needs re-reading. If the copy has more than one negative, or describes the outcome of the state the reader is not in, or leans on words like unless, except, and prefer not to, it is worth rewriting as the plainest possible sentence and seeing whether anyone objects to the plain version. Objections to plain phrasing are the proof: the same instinct produces confirmshaming, where the way out is worded to make refusing feel like a confession. One writes the refusal to hurt, the other writes the choice to be misread.