One common reason to use negation is to express that something differs from our expectations. For instance, uttering a negated sentence such as “John is not violent” is adequate if the speaker indeed believed that John is a violent person (e.g., since John is a member of a criminal gang). In our research, we investigated whether recipients infer from negated utterances that the speaker must have had good reasons to assume that the information under the scope of negation is a plausible alternative, potentially affecting the recipient’s attitude formation towards the individual talked about. Participants were presented with photos of individuals and short verbal descriptions varying in polarity (negated vs. affirmative) and valence (negative vs. neutral vs. positive). If negation-based pragmatic inferences really occur, we hypothesized that (1) individuals described with negated sentences involving negative attributes (e.g., “Sam is not cruel”) should get more negative likeability ratings than individuals described with negated sentences involving neutral attributes (e.g., “Sam is not right-handed”) and (2) individuals described with negated sentences involving positive attributes (e.g., “Isabelle is not honest”) should get more positive likeability ratings than individuals described with negated sentences involving neutral attributes (e.g., “Isabelle doesn’t have many neighbors”). However, the results did not match these predictions and rather resembled prior findings from semantics suggesting that negating an adjective conveys a mitigated meaning of its antonym. This challenges the idea that pragmatic reasoning of the kind we propose is involved in attitude formation from negated utterances.