the description are the following. When we are deceived, it is because the thing really looks like what we take it as. At the same time, things will look differently once the deception is uncovered. And the deception is uncovered in the course of further perception/action/exploration of the world.

Let’s look at each of these features of the description in turn, and see what lessons are to be drawn from them.

For one thing, in many ordinary cases of perceptual deception, we are deceived because the thing we mistakenly perceive really does look like or sound like or taste like or feel like something else. The bush in the forest does, from such and such a vantage point, and in such and such light, look like a deer. The strawberry schmear does look in many respects like the salmon schmear. This is in direct contrast to some traditional modes of thinking about deception – modes Merleau-Ponty calls “sketchy reasoning.” If we start not with an appreciation of the positive character of deception but instead with an assumption that deception is a kind of negation, a departure from the objective world as it determinately presents itself to us, then the tendency is to see deception as the result of our erroneous contribution to what is truly given in experience. There is not, in fact, a deer on the path. And thus, the “sketchy reasoning” goes, we must associate what is there with some memory of or past experience of a deer. So the deception, on this account, is the result of the contributions of memory to what is actually experienced.

But, Merleau-Ponty points out, this way of thinking about deception in fact fails to accomplish what it sets out to, because the present experience must already have “form and meaning,” it must already look like something, in order to call forth just these memories as opposed to others (see PP 20–1). But that means that, in order to call forth the memory of a deer to make the bush seem like a deer, for example, the bush must already look like a deer. Otherwise, there is no reason why we would see it as a deer as opposed to a gorilla or the Shah of Iran, or anything else. Indeed, it is this looking like a deer that makes the deception deceptive – it “passes itself off as genuine perception precisely in those cases where the meaning originates in the source of sensation and nowhere else.” If that is so, then the supplement of memories comes too late to explain the deception (PP 20).

The phenomenology of deception, then, points us to the inherently meaningful structure of the perceptual world; indeed, it expands our understanding of it. It shows up as unmotivated the belief in a meaningless stratum of sensations, to which meanings subsequently are attached. Merleau-Ponty illustrates this through a discussion of Zöllner’s illusion, an optical illusion in which parallel lines are made to seem to be converging (Fig. 3.1).

For Merleau-Ponty, it is wrongheaded to start from the assumption that the lines must actually be given in perception as parallel, and then to try to explain how the lines end up being experienced as converging. Instead, the interesting question to ask about this illusion is