![Sample our Humanities journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5939/TOC-Subject-Sample-2021-Humanities.jpg"})
ABSTRACT
In ‘Knowing What Things Look Like,’ I argued against the immediacy of visual objectual knowledge, i.e. visual knowledge that a thing is F, for an object category F, such as avocado, tree, desk, etc. Joseph Shieber proposes a challenging dilemma in reply. Either knowing what Fs look like requires having concepts such as looks or it doesn’t. Either way my argument fails. If knowing what Fs look like doesn’t require having such concepts, then he claims we can give an immediacy-friendly anti-intellectualist account of knowing what Fs look like, one that I neglected. If it does require having such concepts, then knowledge of what Fs look like plays no important role in ordinary cases of knowing things to be F by their looks. I argue for several claims. First, Shieber's anti-intellectualist proposal fails for independent reasons. Second, I give reasons for thinking that knowing what Fs look like doesn't require having a general concept of looks, which lessens worries about nonhuman animals having such knowledge. Finally, I consider the possibility, important to Shieber's argument, that nonhuman animals are simply incapable of knowing what Fs look like. I argue the implications for human knowledge are far from clear.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1 In what follows, unless otherwise indicated, all references to my work are references to this paper (McGrath Citation2017).
2 Cf. Bonjour Citation1985 and Wright Citation2007.
3 Cf. Cohen Citation2002.
4 Adherents of such liberal views include epistemologists sympathetic to versions of ‘phenomenal conservatism’ and ‘dogmatism,’ such as Huemer (Citation2001), Tucker (Citation2010), and Pryor (2001), Siegel (Citation2012), as well as reliabilists such as Goldman (Citation2008) and Lyons (Citation2009).
5 This is Shieber Citation2017.
6 I modify these conditions slightly in the final version given in the paper (6), but they will do for our purposes here.
7 Two clarifications here. First, by using ‘de dicto’ I do not mean to imply that the knowledge in question is knowledge of a dictum, i.e., a proposition. Recognitional abilities, which aren’t themselves knowledge of propositions can be de dicto vs. de re in the sense I’m describing. One can use ‘S has an ability to recognize Fs by sight’ in a way in which ‘Fs’ occurs transparently, but also use it in a way in which it occurs opaquely. In the first case, the truth of the ascription requires that S have the ability, of an Fs S sees, to recognize it’s one of them/that sort; in the second, what is required is that S have the ability, of an F that S sees, to recognize that it is an F. Second, I use ‘de re’ with respect to the object kind, not with respect to the look. (How the look is conceived or cognized is an important issue under ‘intellectualist’ theories of knowing what Fs look like (see McGrath Citation2017, 16–17), but not the present concern.)
8 This point is noted by Sainsbury and Tye (Citation2012;, 167). See also McGrath Citation2017 (17 n19).
9 There is a third way to use ‘know what Fs look like’, one in which ‘what Fs look like’ is used as a free relative. (See Stoljar (Citation2016) for discussion.) When used in this way, the knowledge attributed – call it free relative knowledge – is knowledge of a property, knowledge of a certain look, a look that Fs in fact have. Compare knowing ‘where Jon is’ merely by knowing that place. If you know Houston, and Jon is in Houston, then even if you don’t have any knowledge about Jon, you would count as having this sort of knowledge of where Jon is. This free-relative knowledge is even further removed from the only-because argument than the de re knowledge is. When we explain someone’s being positioned to know that a thing is F by referring to their knowing what Fs look like – as done in step 1 of the only-because argument – we’re not offering up as explanans of their being so positioned the fact that they know a certain look. Knowledge of the look, unlike de re knowledge of what Fs look like, doesn’t even suffice for knowledge concerning Fs at all.
10 Anti-intellectualist accounts of knowing what Fs look like take such knowledge to be neither identical to or grounded in factual knowledge. The intended comparison is anti-intellectualism about knowing how. See McGrath Citation2017 (section 5) for discussion.
11 As I note in McGrath Citation2017 (23), you can know what a perfect ringer for an avocado looks like without having any such disposition. Here F is ringer for an avocado. Moreover, the view that knowing what Fs look like is such a disposition fails for more ‘normal’ categories as well. Consider F = powered sugar. Suppose a person learns what powdered sugar looks like. Must they be disposed to respond to this distinctive look with a belief or seeming that it is powdered sugar, in the absence of evidence it’s not? Not plausibly. They might leave open the possibility that it is flour. Indeed, consider F= avocado. One might learn what avocados look like but lack the disposition to respond to this look with a belief or seeming that it isn avocado because one leaves open the possibility that other kinds of things in her normal environment might have the same appearance. Maybe they shouldn’t leave this possibility open if they have no evidence for it, but their leaving it open is no bar to their learning and knowing what avocados looks like.
12 Watanabe, Sakamoto, and Wakita (Citation1995) taught pigeons to distinguish Monet paintings from Picasso paintings. I very much doubt that these pigeons have the concept of a Monet painting, or that they know that the painting is a Monet, as opposed to knowing it has a certain complex shape/color feature (or appearance). For more on pigeons’ visual abilities, see the references at the end of Shieber’s paper.
13 I thank an anonymous referee for this question. The question fits well with Shieber’s comments in footnote 6 of his paper (747).
14 One might object that I’m presupposing an overly fine-grained account of propositions here. However, I don’t think fineness of grain is a crucial issue. If the objects of belief and factual knowledge are coarse-grained propositions, then guises will enter into beliefs/knowledge, and the distinction between mediate and immediate knowledge will be sensitive to these guises. One knowledge state will be capable of mediating another even if they have the same proposition as object. For instance, suppose Lois Lane knows that Superman can fly. This is knowledge of a proposition <x flies> under a ‘Superman’ guise. She then learns that Superman is Clark Kent. This is knowledge of an identity proposition under a pair of guises. These together give her knowledge of <x flies> under a ‘Clark Kent’ guise. The latter is mediated by the former, despite their having the same proposition as object.
15 It's unproblematic that the implication would go through if one also knew that diamonds and cubic zirconia stones look alike, or even if one knew, of diamonds and cubic zirconia stones, that the former look like the latter. But the implication does not go through when such knowledge is absent.
16 You might worry that this same objection would undermine the Ringers Principle, interpreted so as to apply to de re knowledge of what Fs look like. I don’t think it would. Suppose I understand what ringers are, and suppose I know, of Fs, what they look like. Understanding ringers, I’ll also know, of Fs, that they look the same as ringers for them. But then it seems I’ll know, of ringers for Fs, that they look this way, too. The converse works as well, I think. If I know of ringers for Fs, what they look like, and understand what ringers are, I’ll know that they look the same as ringers for them. But then I’ll know, of ringers for ringers of Fs – i.e., of Fs –what they look like.
If we try to apply the same reasoning to subclasses of ringers, it fails. Suppose I know, of diamonds, what they look like. I know, of them, that they look the same as ringers for them. So, I’ll know, of ringers for them, i.e., ringers for diamonds, what they look like. But if I don’t know, of cubic zirconia stones and ringers for diamonds, that the first is a subclass of the second, I won’t know, of cubic zirconia stones, what they look like. It can be an epistemic accomplishment to learn, of a pair of classes, that one is a subclass of the others. This is so even under a liberal conception of what it takes to know, of a class of things, that it has a feature.
17 Notice, as well, that when you evaluate the counterfactual, you certainly don’t suppose yourself to be severely impaired or somehow abnormal in such a way that you can’t distinguish avocado-looking things from non-avocado-looking things. In the relevant situation, you still have the Shieber-style response ability for avocados.
18 One might wonder whether the response ability version of anti-intellectualism2 would perform better if construed as an account of free-relative knowledge of what Fs look like. I will not explore the matter. See note 9 for reasons for setting aside free-relative knowledge of what Fs look like in our discussion.
19 Shieber writes that anti-intellectualism2 ‘doesn’t place any conceptual requirements on the possession of knowledge of what things look like’ (475). The emphasis is mine. The current interpretation does place a conceptual requirement on this knowledge: possession of the concept of the relevant object kind F.
20 Could Shieber explain anti-intellectualism2 in terms of an ability to have this knowledge of conditionals? No. The ability to have that knowledge will not issue in immediate knowledge in paradigm cases: to arrive at knowledge that this is a human from knowledge of the conditional, you would also need knowledge that one of them is a human, and the incorporation of this knowledge into the epistemizer would make it mediate.
21 For convenience, in this discussion I have ignored issues about the identity and determinateness of looks. I have proceeded as if object kinds have a single characteristic look. This is an oversimplification. In my 2017 (36-38), I discuss some of the complexities. This topic is relevant here, because certain species of nonhuman animal will of course be unable to perceive some appearance features that normal humans can perceive (and vice versa), e.g., certain color properties. In my 2017, I noted the context-sensitivity of ‘what Fs look like.’ But I suspect that this doesn’t go far enough. In addition, different subjects can know what Fs look like in virtue of associating distinct looks, which are all genuine looks of Fs, with the concept F.
22 Evans writes: ‘ … if a subject can be credited with the thought that a is F, then he must have the conceptual resources for entertaining the thought that a is G, for every property of being G of which he has a conception.’ (1982, 104)
23 For instance, we might think the pigeons in experiments by Watanabe, Sakamoto, and Wakita (Citation1995) have concepts of certain complex appearance properties (ones distinctive of Monet paintings) and have the ability to recognize and thereby know them when present.
24 Even independently of recombination considerations, is it plausible to think that nonhuman animals have the concept avocado, or even tree? I have assumed so here, but the assumption is inessential. They might lack these concepts but still have concepts of the same categories – concepts that they associate with expectations concerning properties of the categories’ members, e.g., not only their appearance properties but their affordances.
25 These issues are discussed in McGrath Citation2018 (126-129) and briefly in McGrath Citation2017 (26n26).
26 For further discussion, see McGrath Citation2021.
27 I am grateful to a referee for Inquiry for excellent comments which improved the paper in many ways, as well as to Joseph Shieber for writing such a thought-provoking paper.
![Supercharge Your Next Research Paper: Research and write your next paper with Jenni AI.]({"type":"image" , "src" : "/sda/112444/Skyscraper-humanities.png"})