Most modern theories of indicative conditionals treat such conditionals as restricted epistemic necessity modals. I discuss two problems for such a view. First, indicative conditionals do not embed like necessity modals under other operators, in particular under possibility modals and probability operators: in these embedded contexts, conditionals do not seem to contribute a universal quantification over epistemic alternatives. Second, when we assess the probability of a conditional, we do not assess how likely it is that the consequent is epistemically necessary given the antecedent. I propose a semantics which solves the embedding problem and the probability problem, while still accounting for the data that motivated the necessity modal view. The account is based on the idea that the semantics of conditionals involves only a restriction of the relevant epistemic state, and no quantification over epistemic alternatives. The relevant quantification — if any — is contributed by the attitude being expressed towards the conditional. If the conditional is asserted, the relevant attitude is acceptance, which contributes a universal quantification, thus producing the overall effect of a restricted necessity modal.