Stanley and Williamson on Ryle: "Knowing How"

[In July and August, 2010, I made some significant revisions and deleted some questionable portions of this post.]


In "Knowing How" (2001), Jason Stanley and Timothy Williamson (S&W) defend intellectualism against Gilbert Ryle. Their paper was selected by The Philosopher's Annual as one of the ten best papers of 2001. Yet, as I will argue, they profoundly misrepresent Ryle (and so fail to make a sound critique of his project). This suggests that there has been a widespread and severe misunderstanding of Ryle among academic philosophers.

Despite the problems with their response to Ryle, S&W's formulation of knowledge-how as a species of knowledge-that is a stand-alone argument and invites criticism of its own. As I aim to show, a clarification of some relevant issues makes it difficult to fully accept their analysis.

In section I, I present intellectualism. In section II, I correct S&W's misrepresentation of Ryle's argument against intellectualism. In section III, I point out an important difference between Ryle and S&W's conceptions of knowledge-that. In section IV, I defend the ability hypothesis against S&W's criticisms. In sections V, I make a case against S&W's analysis of knowledge-how.


I. The Intellectualist Legend

In chapter 2 of The Concept of Mind, Ryle argues that knowledge-how cannot be reduced to knowledge-that.* This argument is presented within the context of his critique of the "intellectualist legend." This is the view that intelligent behavior is always accompanied or preceded by a mental act, such as an act of theorizing. Intellectualism propagates the "myth" (to use Ryle's term) that what makes a behavior intelligent is its causal relation to a mental act which precedes or accompanies it.

Ryle summarizes his point (pp. 49-50):

"The central point that is being laboured in this chapter is of considerable importance. It is an attack from one flank upon the category-mistake which underlies the dogma of the ghost in the machine. In unconscious reliance upon this dogma theorists and laymen alike constantly construe the adjectives by which we characterise performances as ingenious, wise, methodical, careful, witty, etc. as signalising the occurence in someone's hidden stream of consciousness of special processes functioning as ghostly harbingers or more specifically as occult causes of the performances so characterised. They postulate an internal shadow-performance to be the real carrier of the intelligence ordinarily ascribed to the overt act, and think that in this way they explain what makes the overt act a manifestation of intelligence. They have described the overt act as an effect of a mental happening, though they stop short, of course, before raising the next question--what makes the postulated mental happenings manifestations of intelligence and not mental deficiency."

Ryle aims to show that intelligence is a matter of behavior, and that it does not always involve an antecedent act or process of intellection. He regards theorizing as the most exemplary form of intellection, which he describes as "one practice among others" (p. 26). The goal of theorizing, he says, is usually knowledge-that, which he regards as being in possession of facts. Knowledge-how, in contrast, is defined in terms of intelligent behavior.


II. The Reductio

S&W correctly identify the form of Ryle's argument against intellectualism, which is a Reductio, or "vicious regress." According to S&W, Ryle begins his Reductio by adopting the following two premises:

Premise 1: If one F's, one employs knowledge-how to F.
Premise 2: If one employs knowledge that p, one contemplates the proposition that p.

Yet, Ryle adopts neither of these premises.

Ryle never suggests anything like premise 1. Admittedly, S&W go on to explain that this premise must be qualified, because it would not make sense if F stood for something like digestion, for example. They say F must at least be restricted to intentional behavior. Curiously, they also say that Ryle "hints" that such a restriction is called for. Yet, Ryle hints at no such thing, because he never suggests this premise to begin with. Ryle states explicitly and repeatedly, from the very beginning of his discussion, that he is talking about intelligent behavior. He leaves no room for doubt. He is not talking about intentional behavior in general, nor is he talking about reflexive behaviors such as digestion.

More importantly, Ryle would not accept premise 2. Ryle does not commit to the position that employments of knowledge-that always entail contemplation. He claims that knowledge-that is the goal, not the condition, of theorizing. Still, even if we suppose that contemplation generally employs knowledge-that, contemplation is not the only, or even the most paradigmatic, example of employing knowledge-that. One may exhibit knowledge-that by publicly or privately stating propositions, such as when one recites the rules of chess. Knowledge-that entails acknowledgment of a proposition, and not necessarily contemplation.

Ryle clearly does not make the argument S-W suggest. In contrast to S&W's formulation, Ryle's Reductio can be better understood as follows. As noted earlier, Ryle arguess that acts of intellection, such as theorizing, are one species of intelligent behavior. It follows that, if intelligent behavior must be accompanied or preceded by some other act of intellection, then an infinite number of acts must occur to produce any intelligent behavior. To put it more succinctly, since intellection is itself a form of intelligent behavior, it cannot be regarded as a necessary antecedent to intelligent behavior without calling for an infinite regress.


III. Knowledge-That

Given Ryle's dispositionalism, and the way he clarifies the knowledge-how/knowledge-that distinction, it is clear that he regards knowledge-that in terms of abilities. For example, Ryle explains the knowledge-how/knowledge-that distinction with the example of learning how to play chess (p. 40-41): being able to recite the rules of chess is not the same as being able to play the game.** He also refers to knowledge-that with the phrase "propositional competence" (p. 49), a phrase which draws our attention directly towards abilities. While Ryle does regard knowledge-that as being in possession of facts, this should be regarded dispositionally, as a set of abilities to do with the use of a language.***

Yet, S&W introduce Ryle's distinction by claiming that knowledge-that "is not an ability, or anything similar." They attribute to Ryle a more nebulous, though perhaps more common, view of knowledge-that as "a relation between a thinker and a true proposition." Unfortunately, it is not clear what it means for a person to stand in a relation to a true proposition, and S&W offer no clarity on the matter.

In any case, after presenting a complex semantic and syntactic analysis, they conclude that statements about knowledge-how contain embedded questions the answers to which take the form of Russellian propositions, even if those propositions are only entertained "under a practical mode of presentation." In other words, if we can say that somebody knows how to do X, we are saying they know that w is a way to do X in a practical way. S&W conclude, knowledge-how entails propositional knowledge, or knowledge-that.

While S&W's formulation of knowledge-how poses no apparent threat to Ryle's Reductio against intellectualism, it does challenge his view that knowledge-how is logically prior to knowledge-that.

In the remained of this paper, I present a comparative analysis of how S&W's and Ryle's formulations of knowledge-how play out, both in general and in specific relation to the ability hypothesis. I begin by defending the ability hypothesis against S&W's objections.


IV. The Ability Hypothesis

In addition to critiquing Ryle and defending intellectualism, S&W use their analysis to defend Frank Jackson's knowledge argument against the ability hypothesis (hereafter "AH").

Briefly put, the knowledge argument claims that a scientist called Mary is able to learn every physical fact about color vision without ever living in the world of colorful objects, and that she learns some new facts about color when she finally sees colorful objects for the first time. The conclusion is that Mary's newly aquired facts are not physical facts, and so physicalism (the view that all of the physical facts are all of the facts) is false.

AH responds to this challenge by arguing that Mary gains new abilities--specifically, the abilities to recognize, remember, and imagine color experiences--and not new factual knowledge. This is sometimes expressed by saying that Mary gains know-how, and not propositional knowledge, or knowledge-that. (Considering the confusion over how to define terms such as "knowledge-how" and "knowledge-that," this may not be the best way to formulate AH.)

S&W argue that AH depends upon the mistaken belief that knowledge-how is distinct from knowledge-that. Yet, AH need not reject the view that Mary's new abilities are propositional knowledge "under a practical mode of presentation." It need only reject the claim that the facts involved are new to Mary. AH may thus stipulate that Mary knew the same facts while inside her black-and-white room, though under a different mode of presentation. Thus, even if S&W's formulation of knowledge-how is adopted, it does not force a rejection of AH. (See Yuri Cath, "The Ability Hypothesis and the New Knowledge-How" [2009].)

S&W also attempt to undermine AH by rejecting the claim that knowledge-how can be understood in terms of abilities. For example, they claim that a concert pianist who loses her arms may still be said to know how to play the piano, even though she has lost the ability.

There is something strange and uneasy about insisting that the pianist still knows how to play the piano but is just unable to do so. We cannot in full confidence say that she knows how to play the piano in the ways she once knew. Surely she still knows something of what it is like to play the piano; she can remember it in some respects; and she can probably use her toes or other devices to pick out a tune. She may also have retained the ability to instruct others in the art of performance. She has retained some abilities, but not others. There is no reason to conclude that she retains the know-how, but not any of the abilities. (See Laurence Nemirow, "So This Is What It's Like: A Defense Of The Ability Hypothesis" [2006]).****

The relevance of S&W's analysis to AH is independent of its relation to both Ryle and intellectualism. While S&W may not have successfully critiqued Ryle, they may still be said to have furthered our discussion of AH. However, it is not clear that they have done so in a way which is of decisive value. Even if one accepts S&W's formulation of knowledge-how, one may still retain AH. Yet, there are reasons for rejecting their formulation--reasons which may provide a clearer understanding of AH.


V. Knowledge-How

According to S&W, when we say that Hannah knows how to ride a bicycle, we mean that Hannah knows of some w that it is a way for her to ride a bicycle. We have little trouble specifying ways of riding a bicycle, and it is hard to imagine anybody knowing how to ride a bicycle without knowing of such ways that they are ways to ride a bicycle. We may thus be tempted to accept S&W's formulation. However, activities like bicycle riding are quite advanced and involve some degree of intellectual awareness. This is why we have little trouble describing and analyzing them. As Ryle says, "intellectual development is a condition of the existence of all but the most primitive occupations and interests" (p.317). For more primitive cases of knowledge-how, it is impossible to specify ways. This makes S&W's formulation far less appealing.

One way of approaching more primitive varieties of knowledge-how is to consider Ryle's distinction between achievement and task verbs. It appears that S&W's formulation of knowledge-how ignores this distinction. They take "Hannah knows how to ride a bicycle" as a paradigmatic case of knowledge-how attribution, even though "ride a bicycle" is a task, not an achievement. This profoundly limits the scope of their analysis.

Task and achievement verbs differ in their "logical force" (p. 150). Unlike task verbs, achievement verbs imply abilities which are not identified by indicating specific events or processes. Ryle further notes that achievement verbs differ from task verbs in so far as that they do not admit of the qualifications, "correctly" or "incorrectly." One can ride a bicycle correctly or incorrectly, but one cannot win a race correctly, or see the color red correctly.

Ryle says it is a category error to suppose that, when one observes the winning of a race, one observes two distinct events: the crossing of the finish line and the act of winning. Achievement verbs "do not stand for perplexingly undetectable actions or reactions, any more than 'win' stands for a perplexingly undetectable bit of running, or 'unlock' for an unreported bit of key-turning" (p. 152). What is observed is one's crossing of the finish line before the rest of the competitors, and this is interpreted as an act of winning. Thus, while "knowing how to cross the finish line" presumably entails knowledge of some way to cross the line, "knowing how to win" does not entail knowledge of any particular way of winning.


So with perception of the color red, what is observed is the act of identification, and--based on its relationship to other observations--this is interpreted as an act of seeing, remembering, recognizing, and/or imagining.
One cannot see, imagine, recognize, or remember redness correctly or incorrectly. One can, however, remember orange and mistakenly identify it as red. One can see red and mistakenly identify it as orange. One can correctly or incorrectly identify instances of redness.

Knowing what it is like to see the color red is a good candidate for knowing-how which is not based on intellection. In Ryle's terms, the verbs "see," "remember," "recognize," and "imagine" are achievement verbs, not task verbs. We have these abilities, and know that we have these abilities, but we cannot specify ways of performing these actions. While we can say somebody remembers, recognizes, or imagines red, we cannot indicate any way of doing so. We cannot even identify the way we do these things ourselves. Which is to say, nothing counts as a way of seeing red.

The ability to identify redness is a behavioral disposition, and it is implied when a person says, "I know what it is like to see red." However, saying "I know what it is like to see red" does not indicate anything it is like to see red at all. It does not pick out anything about a separate act of seeing as distinct from the act of identification. There is knowledge-how associated with "knowing what it is like to see red" which cannot be defined in terms of propositional competence, or knowledge of particular ways.

To take another example, Ryle notes that being able to tell jokes does not entail knowing that there is any particular way of telling jokes. When we say somebody knows how to tell jokes, we do not imply that there is any particular way of telling them. We do not imply that, when he tells a joke, he is employing any particular method which he applies to joke-telling in general. We could not indicate any such method if pressed, and neither could the person who knew how to tell jokes.

S&W claim that when we say a person knows how to tell a joke, we mean they know that some w is a way to tell a joke. This suggests that the person who knows how to tell jokes knows that one act of telling a joke is an example of telling jokes. But, then, in attributing knowledge-how to the person, we are claiming only that they know that one particular joke-telling exercise is a joke-telling exercise. Clearly this is not what we mean when we attribute knowledge-how. When we say a person knows how to tell a joke, we are attributing abilities which extend beyond any particular set of jokes.

One can know how to tell jokes well without there being any fact of the matter about what defines good joke-telling. Similarly, one can know what it is like to see red without there being anything one could identify as what it is like to see red. This counter-intuitive fact has made resolution of the knowledge argument rather difficult.


V
I. Conclusion

S&W approach a Rylean understanding of knowledge when they discuss Carl Ginet. Following Ginet, they note that "it is simply false that manifestations of knowledge-that must be accompanied by distinct actions of contemplating propositions." Ryle makes exactly this point, and it is why he rejects the intellectualist legend. Yet, S&W use Ginet's observation as a point against Ryle. This is explained by the fact that they have fundamentally misunderstood Ryle's project.

Intelligent behavior is not always dependent upon propositional competence. Contrary to S&W's formulation, knowledge-how does not always entail knowing that any particular w is a way to X.


See also:

Notes

* Stanley and Williamson reference another of Ryle's texts, originally published earlier though now available in a collection published in 1971. I unfortunately do not have this other text at my disposal. However, that text seems to be in line with what can be found in The Concept of Mind, and S-W do not draw attention to any significant differences between the two texts.

** If this is not clear, imagine a person who can recite all the rules of chess, but who demonstrates no strategic competence whatsoever. When they play, they follow the rules, but do not make any moves which might be regarded as intelligent. For example, they might only move their pawns. This sort of playing is what we would expect of somebody in the very beginning stages of the learning process; and, if such a student were asked if they could play chess, they would reasonably respond, "No, I'm just starting to learn."

*** We might suppose that knowledge-that is a species of knowledge-how, though Ryle does not put it this way. He regards knowledge-that as a peculiar sort of ability which does not easily translate into the language we use for knowledge-how. For example, he notes that we normally speak of a person having only slight knowledge of how to perform actions, but not of having slight knowledge that something is true.

**** Nemirow (2006) identifies Noam Chomsky and Torin Alter as supporters of this argument against AH. However, Alter has recently recanted his position on the matter. See Alter, "Phenomenal Knowledge Without Experience" (2008), footnote 14.

Comments

Popular posts from this blog

An Argument For Compatibilism

Coco, The Book of Life: What's the Difference?

The Unintended Irony of Birdman and Big Hero 6