Art history has lost and rediscovered semiotics several times since the 1950s, and at the moment writers employ an eclectic mixture of theories derived mainly from Ferdinand de Saussure and Charles Sanders Peirce. Of the two, Peirce is possibly the more influential model, on account of his tripartite theory of iconic, indexical, and symbolic signs. (The Saussurean model is identified more closely with the poststructuralist moment in art history, beginning in the 1970s.) The question of semiotics is once again topical now that visual culture is consolidating as a discipline, because the new field can choose from a wide range of semiotic practices - or it can choose to bypass semiotics altogether. This essay is a contribution to that current state of affairs. My principal point is that Peirce is much stranger than he is taken to be: he is idiosyncratic and demanding, and at times outlandishly hermetic. For most of what art history and visual studies aim to do, Peirce is simply not necessary; and when he is pertinent, he is so mainly as a model for concertedly logical thinking of a sort that is rare in visual studies or art history.