Carlotta Pavese is currently an Assistant Professor of Philosophy at Duke University. In 2013-2014, she was a Bersoff Assistant Professor/Faculty Fellow of Philosophy at NYU. Before that, she was a graduate student at Rutgers. Her main areas of research are epistemology, philosophy of mind, philosophy of language, and philosophical logic. Her main project is on the nature of practical abilities and skills and she is working on book project on these topics — The Practical Mind — some of which is summarized below. She is also working on several other papers in epistemology, philosophical logic, and semantics. Since at Duke, she has been thinking seriously about the philosophy of cognitive sciences.

The Practical Mind

Carlotta Pavese

People differ in their abilities to do things — in their practical abilities, or skills. That is true both of basic skills and of complex skills. I can move my right ear back and forth, whereas Ale cannot. On the other hand, Ale is a talented chef, who can make the best risotto in North Carolina. By contrast, I am a terrible cook. What makes me different from Ale? In other words, how do people differ mentally and/or cognitively when they differ in their skills?

According to dispositionalism, skills are simply dispositions to behavior (Ryle 1949) and/or dispositions to be in certain mental states (for a defense of a view of this sort, see Stanley and Williamson (forthcoming)). For a dispositionalist, there is no further story to be told, from a mental and/or cognitive perspective, over and above the presence of those dispositions. There may be, of course, a lot more to say at the physical level. But at the mental/cognitive level, the only difference that there is between people differing in their skills has to do with the presence or absence of certain dispositions.

Contra dispositionalism, I argue that there are at least two facts about skills that call for an explanation at a cognitive/mental level. First, complex skills are composed out of basic skills in a distinct way; second, skills characteristically manifest in intentional action. In order to explain those two aspects of skills, the practical mind must be both very conceptual and very knowing.

Complex skills seem to be composed out of simpler skills. In full generality, for a complex task of φ-ing, the skill to φ requires the skill to perform all the parts of φ-ing. For example, one’s ability to make a dish requires one’s ability to execute parts of that dish. And one’s ability to play a difficult song requires one’s ability to play certain accords, in a certain succession — i.e., in accordance with a certain structure.

Any view of skills has to explain how complex skills arise from basic ones in this fashion. But this explanatory goal puts constraints on what skills can be and on how they can arise from simpler skills.

To see how, consider an analogy with a very common kind of explanation provided in the philosophy of language. Suppose we want to understand our ability to understand an arbitrary English sentence. It is natural to explain it in terms of simpler abilities — e.g. our ability to understand its parts, together with our ability to put them together according to the sentence’s structure. But it is commonly thought that, for that sort of explanation to be feasible, the meaning of a sentence must obey the so-called principle of compositionality: the meaning of a sentence must be compositional on the meaning of its parts, in accordance to its syntactic structure. Only in this case, can we explain our ability to understand an arbitrary English sentence in terms of our ability to understand its parts, together with our ability to put them together according to the sentence’s structure.

This kind of explanation puts constraints on what meanings can be. In particular, meanings of a sentence’s parts have to be the sort of things that compose with other meanings to give rise to the meaning of a whole sentence, if we are to explain our ability to understand any new arbitrary sentence. If the relevant candidates cannot compose in the relevant way, they do not count as the sort of things the grasp of which can explain our ability to understand an arbitrary English sentence.

As Fodor 1998 taught us, dispositions do not seem to compose. Moreover, it is totally mysterious what their ways of compositions would be. Because of this, dispositionalism (of all brands) is at loss when explaining the compositionality of meanings. By the exact same reasoning, dispositionalism is at loss when explaining how complex skills arise from more basic skills.

This sort of argument motivated me to look for an explanation of skills in terms of the compositionality of certain sorts of concepts (practical concepts) and in terms of our possessing those concepts. Concepts are combinatorial — they are supposed to be exactly the sorts of things that can combine with other concepts to make up propositions. Moreover, concepts possession plausibly explains our cognitive and perceptual abilities. For example, the concept of red enables one to discriminate red from other colors. And the concept of table enables one to tell table apart from chair. Why then not appeal to special kinds of concepts — practical concepts — to explain our distinctively practical abilities — our skills?

In the literature on know how, others have appealed to such things as practical modes of presentation (Stanley & Williamson (2001), Stanley (2011)). But many have complained that such practical modes of presentation are mysterious and in need of explanation. In “Practical Senses” (Philosophers’ Imprint, 2015), I respond to these concerns by elaborating a theory of practical modes of presentation, that I understand as sorts of concepts, within an explicitly Fregean framework. For a Fregean, concepts are ways of determining their referents — they are senses.

To those worried that anything like a practical sense actually exists, I respond that we actually find examples of practical senses in the semantic values that a certain sort of semantics, known as operational semantics, assigns to program texts. Operational semantic values are rules for mappings configurations of instructions to other configurations of instructions. I show that such operational semantic values satisfy the main requirements on Fregean senses — they are abstract, mind-independent, they determine algorithms as their referents and, finally, they have a distinctive kind of cognitive significance in that they are rules which enable those who grasp them with the ability to follow them.

So, operational semantic values are sorts of Fregean senses. In addition, because they can enter into composition with other things of the same kind, according to certain structural rules, they are the sorts of things that can explain the arising of complex skills from other simpler skills. Finally, they qualify as practical senses because their distinctive kind of cognitive significance guarantees that, if one understands them, then one is enabled with a practical ability — the ability to follow a rule. That is so because operational semantic values are themselves primitive recursive: they are definable in terms of other rules (their parts) that one can already follow, and those in turn are defined in terms of operations that are basic for one.

The sort of practical abilities operational semantic values endow one with are themselves rule-following abilities. That has led me to topic of rule-following and to a reflection on Lewis Carroll (1895)’s regress. Carroll asks us to imagine somebody (the Tortoise) who, given two premises, of the form A, and if A then B, cannot understand that B follows from them. The Tortoise seems unable to follow the rule of modus ponens. Now, here thinking of modus ponens as a logical principle — such as the principle that if A, and If A then B, then B — does not help. For if modus ponens were a logical truth, then using it in the course of an argument would amount to instantiating it as further premise. But that will not help the Tortoise see that B follows. Rather, it will embark her on the so-called regress of the premises.

Often philosophers suggest that this regress can be overcome by acknowledging that following a rule cannot be a matter of instantiating a logical truth as a further premise. But while this conclusion is certainly correct, note that it is only negative. It does not tell us what following a rule actually consists in.

A further question to ask is: if following a rule is not a matter of instantiating a logical truth as a further premise, what does it consist in? As before, just saying that following a rule is a disposition does not explain it. For, if we are to be guided by a rule, we must somehow represent it. But what kind of thing is an inferential rule, such that representing it can explain our ability to follow it?

It could not be simply a syntactic mapping, for we may be able to follow a syntactic mapping without being able to understand that a conclusion follows from the premises. Imagine a version of Searle (1980)’s Chinese Room — the Inferential Chinese Room — where a subject is asked to map premises to conclusions according to certain syntactic mappings. The subject may do it quite well and reliably, without necessarily understanding that the conclusion follows from the premises.

If syntactic mappings cannot explain our inferential abilities, what will? In some recent work (“Inferential Rules as Dynamic Semantic Values”), I propose that inferential rules can be thought as special sorts of dynamic semantic values — the sort of semantic values suggested by proponents of dynamic semantics.

Dynamic semantics is an approach to semantics more and more prominent in the philosophy of language. According to it, meanings are to be understood dynamically, as functions from context to context. Availing myself of the technical tools provided by dynamic semanticists, I show in detail that inferential rules can themselves be thought of as composite functions from context to context, made out of an update part — telling us how to update a context with the premises, and a test part — telling us how to check that the conclusion follows from the context so augmented.

Because, in contrast with logical truths, such dynamic semantic values are not propositions, following them is not a matter of instantiating them as further premises. Hence, so conceived, following a rule does not trigger the regress of the premises. And since being competent with such semantic values requires one to be able to check whether a conclusion follows from a certain set of assumptions, a view that appeals to dynamic semantic values in a theory of inferential competence does not face the problem of understanding that instead afflicts a syntactic conception of inferential rules.

So, if we think of following an inferential rule in this fashion — as a matter of being guided by the dynamic semantic values of the logical vocabulary — we make a lot of progress in understanding how to halt Carroll’s regress and what makes us different from the Tortoise. But we also make progress in understanding skills in general. For recall that practical senses and operational semantic values are themselves inferential rules of a certain sort — they are rules to move from configurations to configurations. If such rules were to be understood as syntactic mappings, we would be presented again with a version of the problem of understanding. In “Operational Semantics, Dynamic Semantics and Competences” (forthcoming in Philosophical Topics), I explain that operational semantic values and practical senses themselves are best thought of not as syntactic mappings, but rather themselves as special sorts of dynamic semantic values. That, I argue, completes my argument in “Practical Senses.”

Are practical senses all there is to a theory of skills and know how? In “Knowing a rule” (Philosophical Issues, A Supplement to Nous, 2015), I argue that they could not be. For our skills and know how characteristically manifest in intentional actions. However, the ability to follow a rule that one acquires by grasping a practical sense is blind and adrift without a propositional attitude that brings it to bear relevantly and intelligently. That is so because one may have the ability to follow a rule without knowing what task the rule is for, and so without being able to intentionally and relevantly deploy that ability. For example, one may have the ability to follow a method that is in fact a method to escape avalanches in virtue of being able to swim, for escaping avalanches requires making swimming movements. In this circumstance, one may nonetheless fail to be able to intentionally escape avalanches, if one does not know that one escapes avalanches by making swimming movements (this particular example is from Hawley (2003). For other examples, see “Knowing a rule”).

Some prominent theories do take the intentionality of an action to require a propositional attitude about the means to employ to perform that action (for example, see Goldman (1970)). Such a propositional attitude is usually taken to be a belief about how to perform that action. Gibbons (2001) argues, upon considerations of lottery cases, that it must be knowledge. According to these views of intentional action, in order for one to intentionally perform a task, one must have certain true beliefs, or even knowledge, about how to perform a task. In “Knowing a rule” (Philosophical Issues. A Supplement to Nous, 2015) and “Skill in Epistemology” (Philosophy Compass, forthcoming), I argue on similar grounds that, because skills and know how characteristically manifest in intentional actions, they must involve a propositional attitude about how to perform those actions.

Practical senses, just like ordinary senses, can then be thought of as components of the content of such propositional attitudes. By thinking of skills and know how in these terms, we are in position to explain not just one’s ability to follow a rule, but also our ability to intentionally perform task by deploying that rule.

The practical mind turns out to be, perhaps surprisingly, both very conceptual and very knowing.

References
Carroll, Lewis (1895) “What The Tortoise Said to Achilles.” Mind, 4(14): 278-80.
Fodor, Jerry (1998) Concepts: where Cognitive Science went wrong. Clarendon Press, Oxford.
Gibbons, John (2001) “Knowledge in Action.” Philosophy and Phenomenological Research 62.3: 579-600.
Goldman, Alvin (1970) A Theory of Human Action, Princeton University Press, Princeton.
Hawley, Katherine (2003) “Success and Knowledge-How.” American Philosophical Quarterly 40.1:19-31.
Pavese, Carlotta (2015) “Knowing a Rule,” Philosophical Issues, A Supplement to Nous, 25 (1): 165-188.
Pavese, Carlotta (2015) “Practical Senses,” Philosophers’ Imprint, 29(15): 1-25.
Pavese, Carlotta (forthcoming)“Operational Semantics, Dynamic Semantics, and Competences,” Philosophical Topics.
Pavese, Carlotta (forthcoming) “Skill in Epistemology,” Philosophy Compass,
Pavese, Carlotta (manuscript) “Inferential Rules as Dynamic Semantic Values,”
https://www.academia.edu/23462009/Inference_Rules_as_Dynamic_Semantic_Values
Ryle, Gilbert (1949) The Concept of Mind, Routledge.
Searle, J. R. (1980) “Minds, Brains, and Programs.” Behavioral and brain sciences, 3(03):417-424.
Stanley, Jason and Timothy Williamson (2001) “Knowing how.” The Journal of Philosophy 98.8: 411-444.
Stanley, Jason (2011) Know how. Oxford University Press.