Different forms of the personal pronouns have different constraints on their interpretations. Chomsky described such differences syntactically, in terms of binding rules. Levinson and Huang propose pragmatic accounts. They describe the differences as differences in implicature, and claim to derive them from the neo-Gricean Q-, I-, and M-principles. Some explanations invoke the disjoint reference presumption (DRP), which Huang and Levinson derive from the I-principle. Following in their footsteps but taking a different path, Ariel proposes pragmatic explanations in terms of her "functional principle," which makes no reference to implicature, but does take expectations of conjoint or disjoint reference to be fundamental. I first consider these neo-Gricean accounts as synchronic explanations, arguing that they are unsuccessful for a variety of reasons. I conclude by considering whether the pragmatic principles instead explain pronominal differences diachronically. While more plausible, I present evidence that a diachronic account based on neo-Gricean principles is also unsuccessful.