During the poster session you can videochat with us via the following link: https://meet.jit.si/cuny_processingSIs
**Abstract**
The Scalarity Hypothesis (SH) posits that Scalar Implicatures (SIs) are cognitively costly insofar as they add a negative proposition into the meaning of the sentence. We tested SH by comparing the processing signature of negative, direct SIs arising from positively scalar terms (e.g., inferring "not all" from *some*) to that of positive, indirect SIs arising from their negated stronger scale-mate (e.g., inferring "some" from *not all*). While our response time results align well with SH, our dual task results challenge the idea that the polarity of an SI is the only or main factor explaining its potential cognitive cost.