Most words are impossible to define. But for logical vocabulary, there's an embarrassment of riches: we could write down infinitely many good definitions for a given quantifier. This becomes interesting if these possibilities aren't just notational variants but specify distinct hypotheses about what concepts and operations are recruited by a given meaning. In this talk, I'll add to the mounting experimental evidence for a transparent relationship between meanings and the cognitive systems used to evaluate them. Then, I'll discuss experiments that leverage this relationship to argue for particular lexical specifications of each, every, and all.