Examining the efficacy of a common response to stereotypical generics

People often respond to stereotypical generic statements such as “girls wear makeup” with counter-stereotypical claims like “boys can wear makeup too.” But are such responses effective in reducing the stereotypical inferences people draw from the original claim? On the one hand, they may seem effective because they directly challenge the idea that a property (e.g., wearing makeup) applies only to one group (e.g., girls). On the other hand, listeners may reason that inserting the modal ‘can’ implies that the second-mentioned group (e.g., boys) is less strongly associated with the property than the first. Here, we present experimental evidence from adults (N = 405) and 4- to 7-year-old children (N = 365) showing that such can generics are ineffective at extinguishing stereotypical inferences. Although can generics appeared helpful relative to leaving the other group completely unmentioned (i.e., not mentioning boys after hearing “girls wear makeup”), they nevertheless resulted in thinking that the group described with can generics is less associated with the relevant property than the group described without the additional modal term. The ineffectiveness of can generics stems from reasoning over alternative options: because both groups could have been described in the same way, ‘can’ signals a weaker relationship between the property and the group described with the additional modal term. We discuss implications for how to effectively respond to stereotypical statements.