How much is given in perceptual experience, and how much must be constructed? John Locke's answer to this question contains two prima facie incompatible strands. On the one hand, he claims that ideas of primary qualities come to us passively, through multiple senses: the idea of a sphere can be received either by sight or touch. On the other hand, Locke seemingly thinks that a faculty he calls "judgment" is needed to create visual ideas of threedimensional shapes. How can these accounts be made consistent? The problem comes to a head in the discussion of Molyneux's problem: can a person born blind and then made to see identify a sphere and a cube, when he had only touched them in the past? Locke's answer is no, but, as George Berkeley points out, it is hard to see why: if the ideas of shapes come to us through sight as well as touch, nothing should stop Molyneux man from instantly recognizing sphere and cube. Pace much of the existing literature, I argue that although Locke does think we receive visual ideas of primary qualities, the faculty of judgment is required to "inflate" those ideas into three dimensions and to correct for other perspectival distortions. I then show how this answer is consistent with the rest of his theory of perception. 1 | INTRODUCTION The opening eight chapters of Book II of the Essay concerning Human Understanding suggest a simple picture of sensory experience. On this view, the mind passively awaits the given of experience: ideas of primary and secondary qualities. Unlike an idea of a color, an idea of a primary quality such as shape comes to us through multiple senses. The ninth chapter threatens this simple picture. There, Locke tells us that some visual ideas of primary qualities are