Knowledge is a nicely written and engaging introduction to epistemology focused on the contemporary debates surrounding the concept of knowledge. It is very broad in its attention, touching on an extensive number of epistemological issues, while consistently engaging the contemporary literature. The book is very accessible with key principles clearly defined, as well as central arguments and examples carefully presented and explained. Knowledge also goes beyond introducing the reader the contemporary epistemological scene. Evans and Smith also present and motivate their own novel analysis of knowledge, complemented with a focus on non-human animal knowledge.
In chapter 1, Evans and Smith provide an introduction to the theory of knowledge and begin their conceptual analysis. Although much of setup is to help the reader fix in on the appropriate target concept and to get some early practice wielding intuitions, Evans and Smith also argue from some non-traditional theses from the get-go. First, they argue that knowledge-that need not be propositional knowledge. Motivation for this claim is found from our willingness to attribute knowledge-that to non-human animals. Evans and Smith argue that non-human animals perhaps do not represent the world propositionally, so if non-human animals have knowledge-that, knowledge-that must not be (at least exclusively) propositional. The authors prefer to refer to knowledge-that as informational knowledge (rather than propositional knowledge), and they have informational knowledge as their target concept (as opposed to knowledge-how, knowledge by acquaintance, etc.). Second, and for similar reasons, Evans and Smith dispense with truth as a necessary condition for informational knowledge (hereafter simply ‘knowledge’). If truth is a property of propositions alone, and non-human animals have knowledge despite not being able to represent the world propositionally, truth is not a necessary condition for knowledge. Knowledge, however, must still be accurate (not just anything can be known!), but accuracy is a broader notion, extending beyond (though including) propositions and their truth. Third, Evans and Smith maintain that belief is not a propositional attitude. The authors stick with tradition in claiming that belief is a necessary condition of knowledge, so if non-human animals have knowledge and yet non-human animals do not represent the world propositionally, at least not all beliefs are propositional attitudes.
It is clear from the start that (non-human) animal knowledge is taken quite seriously by the authors. Throughout the book, Evans and Smith challenge various accounts of knowledge on the basis of consequences that they see those accounts as having for the possibility of animal knowledge. More traditional epistemologists might want more motivation than they are given here for taking animal knowledge so seriously and (perhaps more importantly) for thinking that the possibility of animal knowledge requires such drastic revisions to more traditional analyses of knowledge. In the very least, this book provides ample opportunity for examining traditional theories of knowledge in light of contemporary ethology.
In the second chapter, Evans and Smith tackle the skeptical problem with both clarity and care. They take the time to make many distinctions that often get missed in other introductions to epistemology (metaphysical vs. epistemic possibility, ambitious vs. modest responses to skepticism). They guide the reader through various skeptical scenarios and skeptical arguments, including the recent skeptical arguments stemming from disagreement—a nice addition (although they incorrectly label the Right Reasons View as ‘the Total Evidence View’). The book moves to epistemic contextualism as the focus of the third chapter. The authors do a good job of clearly presenting and motivating the view as well as applying it to the skeptical puzzle. Evans and Smith consider the leading objections to contextualism, though they do not come to a final verdict on this semantic thesis. The nearby issues involving pragmatic encroachment are also discussed in this chapter. Here too, the authors do well in integrating the current literature and pointing the reader toward further research questions. Including these topics allows this introduction to capture a great deal of the epistemological landscape of the last 20 years that often gets left out of other introductions.
The book then turns to examining various accounts of warrant (that which turns true/accurate belief into knowledge). Internalist accounts of warrant are first up, and Evans and Smith divide the conceptual space by identifying different ways that one might respond to the infinite regress argument concerning justification and proceeding to test these accounts of justification against the problem of easy knowledge. In the end, the authors defend a holistic coherentist account of justification accompanied by a higher order requirement for justification. For instance, they claim that in order to be justified in believing a perceptual proposition, one must be justified in believing that perception is reliable, though they deny that this higher order justification must be antecedent to the first-order perceptual justification. This account hopes to avoid the familiar problems of a higher order requirement since the justification for the higher order belief comes by way of its coherence relations with one’s total belief set. The authors claim that such a higher order belief will typically cohere well with one’s belief set so long as that set includes a number of perceptual beliefs (beliefs which they claim presuppose the reliability of perception). Holistic coherentism avoids the regress argument by denying that justification is linear (and that there are inferentially justified beliefs), while avoiding the problem of easy knowledge by not allowing for justified perceptual beliefs without a justified belief in the reliability of perception. The chapter concludes by applying this account of justification to non-human animals as well as to the literature on testimonial justification.
While there is much to like about their discussion of alternative accounts of justification, Evans and Smith’s discussion of holistic coherentism is regrettably brief. The authors do not do much to unpack the notion of coherence, and they do not consider objections to holistic coherentism (save one in the following chapter) or explain what they think is mistaken with the extant objections to it. Does coherentism offer a suitable truth connection? After all, every belief in a coherent set of beliefs could be false. How (if at all) do experiences figure in? After all, coherent sets of beliefs can be isolated from reality and one’s experiences of it.
Chapter 5 delves into defeat and the basing condition. Evans and Smith only discuss defeat in relation to knowledge, which is somewhat unfortunate for two reasons. First, a discussion on how justification can be defeated (in the sense that the subject is no longer justified in believing something that she was justified in believing) would further elucidate the concept of justification. Second, many internalists do not analyze Gettier cases as cases where one’s justification is defeated (though some do). After all, in such cases, the subject remains justified in believing the target proposition, so it is in some sense misleading to claim that the subject’s justification has been defeated. The chapter then turns to the often-neglected topic of basing. The authors give this topic much more attention than it typically gets, which is a good thing. They briefly describe and criticize both causal and doxastic theories of basing before giving a sketch of their own novel dispositional account. They conclude the chapter by considering whether knowledge requires a properly based belief and whether a holistic coherence account of justification can make sense of basing. The authors claim that doxastically justified beliefs must be based only on some subset of beliefs within a coherent system of beliefs. The challenge of which subset(s) matter and why, however, is not addressed here.
Externalist accounts of warrant are the topic of chapter 6. Evans and Smith tackle a couple causal accounts, Nozick’s truth-tracking account, as well as process reliabilism. Each account is laid out clearly and saddled with several objections though the authors advise caution regarding some of the more ‘far out’ intuitions cited against these accounts. It is nice to see these externalist accounts measured against Gettier problems along with their internalist competitors, and the authors assess the costs and advantages of each of these theories nicely.
In chapter 7, Evans and Smith direct the discussion toward epistemic evaluations. The authors nicely distinguish between instrumental and categorical norms and argue that epistemic norms are best seen as a kind of categorical norm. The authors then explicate deontological, consequential, and virtue theoretic approaches to epistemic evaluations. Along the way, they turn to some prominent accounts of knowledge/warrant in the neighborhood: Sosa’s virtue account and Plantinga’s proper function account. Sosa and Plantinga’s accounts get relatively less attention than their externalist rivals did in the previous chapter, and the evaluation of these accounts is quite quick. In particular, it is surprising that Sosa’s account of animal knowledge is not given more attention, given that it might be seen as a close competitor to the authors’ own account.
Evans and Smith begin to advance their own positive account of knowledge in chapter 8. They start by outlining desiderata for a successful account of knowledge gleaned from the earlier chapters, and they nicely integrate the earlier topics in advancing their own account. The account itself brings together elements from a number of epistemological accounts with its closest affinity being to a proper functioning account of knowledge. Perhaps most notably, the offered account relativizes the conditions of knowledge to species—what it takes for a human to know is not what it takes for a cat to know. Evans and Smith insist that knowledge is not an equivocal concept, but rather a generic concept—one that is satisfied differently in different kinds of beings, depending on the capacities of beings of that kind (much like health). In brief, their account claims that a subject knows when he/she veridically uses reliable cognitive processes that are functioning properly in an environment to which they are well suited (203).
Evans and Smith proceed to defend their account as suitably naturalistic. They respond to ‘Darwin’s Doubt’ and give reason to think that their account is superior to both Plantinga’s proper functioning account as well as its internalist competitors. Evans and Smith’s account shares features with both Plantinga and internalist accounts, since they argue that human knowledge typically requires internalist justification, though they claim that a proper functioning story is required to turn a justified belief into knowledge.
Such hybrid accounts typically fail to satisfy the parties from both sides, though this analysis of knowledge is worthy of attention (much more than I can give it here). I will briefly point out two issues. First, while Evans and Smith’s account draws from both reliabilism and a proper functioning account, they give no reason to think that their account avoids some of the well-worn problems that inflict those accounts. For instance, there is not even a gesture as to how to answer the generality problem. Without a solution here, we are simply unable to assess the account on offer. Along those same lines, a lot of work is being done by ‘normal environments’, which only makes this problem worse. All we get from the authors is that normal environments are those “which depend on the sorts of environments the processes evolved in” and which “are the sort of environment within which [the subject’s] cognitive capacities function properly” (175). The former is hardly helpful, and the latter seems to give way to a circular account. Second, while requiring internalist justification for human knowledge avoids some of the problems that afflicted externalist accounts, Evans and Smith maintain that other species need not be justified in what they believe in order to know. So, it seems that clairvoyant-type objections to reliabilism and the problem of easy knowledge will return, at least when the account is applied to non-human knowledge.
Regardless of the merits of their positive account, Evans and Smith’s book should be praised as an extremely accessible and up-to-date introduction to epistemology that canvasses a host of epistemological hot topics. It is surely a welcomed addition to the literature.