Bartels and May set out to clarify what a theory of knowledge-how should provide and begin to give the outlines of such a theory. In their view, a theory of knowledge-how should explain the difference between practical and theoretical knowledge, the former being characterized by a number of distinguishing features. The proposal, then, is to explain this difference in terms of the reliance on non-conceptual capacities (or representations) in the case of practical knowledge and on conceptual capacities in the case of theoretical knowledge, instead of using propositionality as the main criterion. Their account of what is to be captured by a theory of know-how, and their proposed solution, are preceded by an illuminating discussion of the shortcomings of each side of the intellectualism vs. anti-intellectualism debate.
In short, Bartels and May claim that the intellectualists are right to concede that the distinction between knowing-how and knowing-that cannot be made in terms of the propositionality of knowing-that. Three readings of propositionality are distinguished:
a representational reading, according to which the propositionality of some mental state depends on a sentence-like mental representation being tokened,
a conscious-availability reading, according to which propositional representations are consciously available and can be expressed linguistically, and
a semantic reading of propositionality, according to which the propositionality of some mental state depends on whether it is attributed as a propositional attitude.
It is argued that all three readings of propositionality are inapt for making the distinction between practical and theoretical knowledge. I take it that both the representational reading and the conscious-availability reading are implausible for independent reasons—the representational reading presupposes a language of thought, while the conscious-accessibility reading can arguably be undermined by considering cases in which someone would be said to know something she need not be able to express verbally, in terms of the proposition in question (this might involve some non-obvious logical consequences of one’s occurrent beliefs). In addition, the semantic reading is what our best intellectualist account of knowledge-how, namely Stanley’s, employs, and Bartels and May follow Stanley’s analysis here.
According to the semantic reading of propositionality, whether some psychological attitude is propositional depends on the semantics of the locutions used to ascribe such attitudes. And our best current theories of the semantics of knows-wh locutions—i.e., of locutions that involve the verb “know” and some question word such as “who”, “where”, “what”, “when”, or, to that effect, “how”—tells us that knowledge-how is propositional—just as knowledge-that is. But as a result, it is argued, intellectualists are not able to explain the respective peculiarities of practical and theoretical knowledge—both are propositional. This is identified as the major shortcoming of intellectualism.
The anti-intellectualists, on the other hand, lack a systematic criterion for the distinction between knowledge-how and knowledge-that. The introduction of different kinds of knowledge, based on different representational formats, by some anti-intellectualists is taken to be ad hoc (e.g., image-based knowledge and sensorimotor knowledge by Jung & Newen 2011). It is not based on an independently identified set of underlying representational formats that would explain the characteristic behavioral differences. Instead, it merely attempts to find alleged mental representational formats that intuitively fit the distinction (cf. Bartels & May this collection, p. 7). Further arguments to the effect that intellectualism is a non-starter are ineffective against Stanley’s (2011) version of intellectualism (cf. Bartels & May this collection, pp. 10-11). An attack from Toribio (2008, reference taken from Bartels & May this collection) to the effect that Milner & Goodale’s patient DF (cf. Milner & Goodale 1995) could not possibly have propositional knowledge of how to put a card into a slot presupposes that knowledge-how involves a conceptual grasp of how something is done or of what is acted upon. Roughly, Toribio argues that DF does not have propositional knowledge of how to put the card into the slot because she cannot report on the orientation of the slot. But Stanley acknowledges that some propositional attitudes involve the non-conceptual grasp of relevant states of affairs. In the case of DF, this involves the non-conceptual grasp of the orientation of the slot (cf. Stanley 2011, p. 172).
As a result, neither intellectualists nor anti-intellectualists provide a satisfactory account of knowledge-how. But both get some things right. The intellectualist is right in taking both knowledge-that and knowledge-how to be propositional. And the anti-intellectualist is right in requiring an explanation of the difference between these two kinds of knowledge, presumably in terms of underlying cognitive structures or kinds of mental representation.
Bartels and May, then, pick up on the idea that practical knowledge might involve non-conceptual capacities, while theoretical knowledge is conceptual. They list a number of received peculiarities of practical knowledge that are to be captured by a theory of practical knowledge. And it is proposed that these peculiarities are the same peculiarities that result from a reliance on non-conceptual representations. Among the differential features of practical knowledge are its being context-bound, implicit, and automatic and effortless. Non-conceptual capacities, it is argued, just have these features. The result is a position that is intellectualist in form, because all kinds of knowledge are propositional, but anti-intellectualist in spirit, as the distinction of practical vs. theoretical knowledge is maintained. Practical knowledge is not reduced to theoretical knowledge; rather, the former is a non-conceptual form of knowledge while the latter is conceptual.
One effect of drawing the distinction between practical and theoretical knowledge in terms of conceptuality is that Bartels and May must follow Stanley in accepting non-conceptual forms of propositional knowledge. Patient DF cannot report on the orientation of the slot, but nevertheless she non-conceptually grasps its orientation such that she is able to put the card into the slot. Due to her successful performance, she is said to know how to put the card into the slot, making this particular form of knowledge-how non-conceptual. This somewhat departs from tradition, where concepts are usually taken to be the constituents of thoughts, while thoughts are likely understood in a Fregean way as the intensions of sentences, i.e., propositions. It makes sense, though, because propositionality is understood semantically while conceptuality is not. Whether some cognitive capacity is conceptual or non-conceptual is thought to depend upon the kind of mental representation involved.