<p>Subtle linguistic differences can shape beliefs about the social world. For example, the statement “Girls are just as good as boys at math” leads some people to endorse the stereotype that boys have more natural math skill compared with a statement with the positions of the groups reversed. Traditional accounts of linguistic framing characterize such effects as an irrational consequence of biased cognitive and emotional processes. In contrast, we hypothesized that framing effects of this sort depend on the ability to pick up on the pragmatic implications of subject-complement syntax, where the group framed as the complement (“boys”) is the implied standard or reference point. We investigated this possibility in two preregistered experiments (<i>N</i> = 1,593). Overall, participants who were better at <i>inferring</i> implicatures from subject-complement syntax were more likely to exhibit a framing effect by <i>endorsing</i> the implicature after reading subject-complement statements about math ability. This relationship held even when the statements referenced non-stereotyped groups and when controlling for other social-cognitive abilities associated with pragmatic competence. Framing effects were reduced for participants who explicitly recognized the statements as influencing their evaluations, but only when they invoked a stereotype to be discounted. These results suggest that pragmatic inference plays a crucial role in subject-complement framing but that people do not necessarily accede to what they infer. Our findings add to the growing body of evidence that many framing effects—far from being irrational—are a natural product of human communication.</p>