In response to questions in which a "wh"-term interacts with a universal quantifier in object position, such as "Who picked every toy?," children as old as 5 years of age often provide a list, pairing toys with the people who picked each of them. This response pattern is unexpected, it has been claimed, because children appear to overproduce such pair-list answers in comparison to what would otherwise be expected in adults, therefore suggesting a non-adult grammar. However, not only have such comparisons been made to a hypothetical baseline of adult responses, but they also fail to take into account the range of possible answers that may be available for such questions, once certain syntactic and lexical manipulations are accounted for. We therefore lack sufficient evidence to fault the grammar for this response pattern. This article investigates this phenomenon from a fresh methodological and theoretical perspective, uncovering a more complex picture. We show, on the one hand, that children do overproduce pair-list readings to "which" questions with "every," in comparison to adults. On the other hand, they also "underproduce" pair-list answers in response to similar questions with "each." However, children are also sensitive to the syntactic position of the quantifier in the direction expected by a subject-object asymmetry. We therefore argue that a key part of the explanation for children's performance lies in immature lexical entries for the participating quantifiers.