Myles Burnyeat has argued that in De Anima II.5 Aristotle marks out a refined kind of alteration which is to be distinguished from ordinary alteration, change of quality as defined in Physics III.1-3. Aristotle's aim, he says, is to make it clear that perception is an alteration of this refined sort and not an ordinary alteration. Thus, it both supports his own interpretation of Aristotle's view of perception, and refutes the Sorabji interpretation according to which perception is a composite of form and matter where the matter is a material alteration in the body. I argue that Burnyeat's interpretation of II.5 should be rejected for a number of reasons, and offer a new interpretation of the distinctions drawn in the chapter, and the relations between them. I conclude that the chapter provides no evidence against the Sorabji view or for Burnyeat's view. Aristotle's assertion that perception is a refined kind of alteration means that it is the kind of alteration that preserves and is good for the subject of that alteration. There is no inconsistency in the thought that perception is a refined alteration of this sort while it, or its matter, is an ordinary alteration.