<p>There are two competing and sometimes neglected assumptions in the philosophical discussion on the success and accuracy of episodic memory. In this paper, I articulate them and argue for one of them. Most philosophers assume that the success of a mnemic representation should be assessed holistically, with all of its elements having equal weight in the success of the memory. Some philosophers, in contrast, assume that the success of a mnemonic representation depends on specific relevant elements. I defend the latter assumption using the paradigmatic case of inquiries. I argue that social norms regarding memory prioritizes the accuracy of certain elements in inquiry contexts. In these contexts, the inquiry selects which elements are relevant to the success of episodic memory.</p>