Susanna Siegel
The Rationality of Perception
Oxford: Oxford University Press, 2017 (ISBN 978-0-198-7997081)
Reviewed by Anne Jaap Jacobson, 2019

The central concern of The Rationality of Perception is the epistemology of perception. How can perceptual experiences have epistemic merit or lack it? What epistemic roles can such experiences play? For Susanna Siegel, answering these questions involves examining the possible roles and clarifying the concepts. The result is a treatise in analytic philosophy.

Siegel lists the virtues of analytic philosophy (xxv): having an ear for new questions, respect for complexity, caution with common sense, and patience with the realm of the possible. Given the lack of receptivity feminists have experienced with analytic journals, Hypatia readers may be puzzled by the first item in the list. In addition, in my experience, and surely that of others, even new ideas far removed from anything to do with gender may well be met with scorn or ridicule.

Siegel's book possesses other features of analytic philosophy. One is that it is concerned with, and only with, an individual's knowledge. Hence, Siegel does not consider group knowledge or knowledge made possible by one's social position that would not be otherwise available (though see a discussion in the book's last section of how common beliefs in one's society can make a difference to the epistemic merit of an individual's belief). Another is the book's largely uncritical adoption of categories found in an ordinary, common-sense discussion of the mind. Thus, we find belief, action, desire, and (perceptual) experience all appearing without much critical attention. "Inference" is an exception, since there is a substantial discussion meant to clarify the concept on pages 77-98. Siegel does not, however, give the inferential/noninferential distinction the space it has received elsewhere. (For a recent discussion of the distinction, see Crary 2018.) We will consider some questions that arise about such categories below.

Siegel is concerned with a particular aspect of the epistemic role of perceptual experience. That perceptual experiences have some epistemic roles seems very obvious. For example, if it is a sunny day and one's experience of it is of a sunny day, then one knows it is a sunny day and, other things being equal, one is justified in saying that it is. However, Siegel is not particularly interested in the simple transition from experience to justified belief. Rather, she sees as routinely ignored the idea that experiences are themselves epistemically appraisable, and that fact is the focus of her attention. Central cases for her of appraisable experiences are hijacked experiences. In these cases, a psychological precursor affects the experience in a way that lessens or eliminates its justification. Thus, hijacking occurs when the perceptual processes give too much weight to one's prior outlooks and fail to give proper weight to the inputs (5). For example, one might report seeing someone's "angry face," when in fact the person's face is neutral. Here the input is outweighed by one's belief that the person is angry.

The focus of the book is on two principles that constitute the thesis of the rationality of perception (23):

The Epistemic Charge thesis: Experiences can be epistemically charged.

Inferential Modulation thesis: Experiences can be formed by inferences that can modulate their epistemic charge.

The discussion of charge defends the idea that perceptual experiences are epistemically appraisable. When an experience's background negatively affects its epistemic value, it is degraded. The element that degrades needs to be something of the experiencer's. As Siegel puts it, the element is part of one's outlook.

The next ingredient is inference. There are kinds of transitions from one belief to another that are not inferences (89). If I think someone has a beautiful dress on, I may think of a similar dress I once had. The transition from "That is a beautiful dress" to "I once owned a dress like that" need not be an inference at all; I am merely associating two things. It is far from easy to characterize what an inference is, but it is very important for Siegel to do so, since it is by inference that something from one's outlook affects the epistemic quality of an experience.

An inference is a kind of response to a state that in some sense carries information; the response results in a conclusion. The kind of response in question is very difficult to characterize directly, and Siegel approaches it by contrasting it with other kinds of responses. As we saw earlier, there are kinds of transitions that are not inferences. An inference is not, like our dress example, simply an association. There are other transitions that are not inferences, such as reckonings. Reckonings involve, among other things, an awareness of the starting point, which is not necessary for an inferential response.

Siegel also includes subconscious processes in her areas of interest. Here I think her discussion may inspire disagreement. Let me explain why with reference to a particular example. In the United States we regularly hear a doctor or a politician declare that the gifts they received from some source do not induce them to favor the source. However, recent research strongly suggests that they are simply wrong. There is good evidence to think that our judgments can be biased in favor of people who benefit us (Harvey 2010), even when the influence is not transparent to us. We seem subconsciously to go from someone's benefiting us to endorsing or supporting them.

This seems to me particularly true when one characterizes the transition in its proper neuroscientific terms. Dopamine is probably playing a role here, and while dopamine is very valuable, its appearance and employment does not constitute drawing a conclusion. As a consequence, Siegel's inclination to take such episodes as containing a conclusion about one's indebtedness is at least questionable.

The point just made can be illustrated with a simpler case. A bee's experience with syrup on an orange patch may mean that when it next sees an orange patch, dopamine will discharge to make the patch more vivid and salient. A motor program will also be set off and the bee will go to the orange patch. It seems acceptable to say that the bee expects to find syrup on the patch. But it is at least very odd to think of the bee as having drawn the conclusion that the patch has syrup on it. The vividness and the movement are merely stages in a transition.

The notion of an inference also features significantly in Siegel's interesting treatment of racism. However, it may be possible to characterize many instances of racism as transitions and not as inferences (for a recent discussion of the inferential/non-inferential distinction, see Crary 2018). What this sort of objection illustrates is a downside to following the ordinary ontology of philosophy of mind, at least as it has been practiced over the last half-century. There is much more recent material to make one concerned about Siegel's treatment of racism in the final section of the book. That is, there is a great deal of recent discussion about whether so-called implicit biases form a kind, even an informal one, that is of particular interest in discussions of racism. Siegel describes a number of cases of racism, and it is arguable that although many theorists have been inclined to describe them all in terms of individual features of active human beings, that is actually a misleading approach. Rather, as many have argued recently, the much more important factor in racism is social structure; this is another point of which Siegel takes inadequate notice.

We have just seen two problems with Siegel's discussion of racism. The first is its use of what can be considered to be a naive ontology. The second is a substantial neglect of structural factors in racism. Despite this, her treatment of racism can be extremely useful. Through extensive discussion, Siegel shows that justifications ordinarily offered for racist beliefs do not confer epistemic merit on them. Perhaps one can proceed in a discussion with most philosophy professors with an assumption that ordinary racism is epistemically unjustified, but one should not take such expectations to students. Rather than overriding students' convictions that, for example, statistics about crimes show minority members are more violent than whites, Siegel pursues such beliefs relentlessly down the closed alley in which they are eventually caught.

Let me mention finally some features of The Rationality of Perception that make it very accessible for readers at all sorts of levels. It may be too focused for general use in beginning classes, but it would be suitable for all levels after that. The accessibility comes in in part from the clarity with which it is written. And though it has many different principles and definitions, it is free of the reflections on them that give us, for example, PrI3* through PrI3***, or PrIB2. This means professors will not have the tempting exam questions that ask students what the difference between PrI3*** and PrIB2 is, for which I think students should be grateful. All in all, The Rationality of Perception is a very skilled and original work. I am glad I have read it and I would recommend anyone interested in the issues to do so.



Crary, A. (2018). The methodological is political: What's the matter with "analytic feminism"? Radical Philosophy 2 (2): 47-60.

Harvey, A., U. Kirk, G. H. Denfield, and P. R. Montague. 2010. Monetary favors and their influence on neural responses and revealed preference. Journal of Neuroscience 30 (28): 9597-9602.

Anne Jacobson received a BA from the University of California, Berkeley, and a BPhil and a DPhil from Oxford. She writes about early modern philosophy, feminist philosophy, and philosophy of mind, with a special emphasis on the philosophy of cognitive neuroscience, along with a commitment to much in the later Wittgenstein. Her recent work includes the co-edited Neurofeminism and her Keeping the World in Mind. She has taught at Oxford University and a number of American universities. She is currently Editor in Chief of the Wiley Blackwell 5-volume Encyclopedia of Women in Philosophy.

"Through extensive discussion, Siegel shows that justifications ordinarily offered for racist beliefs do not confer epistemic merit on them."