<p>Logical inferentialism is often characterized as the view that the meaning of logical expressions is given by their rules of inference. This position is standardly contrasted with truth-theoretic semantics, according to which meaning is understood in broadly referentialist terms by appealing to truth conditions. Logical moderate inferentialism calls into question this familiar opposition by articulating a hybrid picture: while the rules of inference matter for determining the meaning of the connectives, their meaning is still to be understood in terms of their contribution to the truth conditions of sentences containing them. This view, however, faces Carnap’s challenge, according to which standard inferential rules generally fail to exclude unintended interpretations of logical expressions. Several recent proposals aim to overcome this difficulty. We argue that these proposals systematically fall short of settling the issue in favor of moderate inferentialism, as they rely on substantive, often unexamined assumptions that compromise their effectiveness. We conclude by articulating the challenge that the logical moderate inferentialist must address before the project can be considered in good standing.</p>