Abstract
This paper criticizes two closely connected rationalist views about human agency. The first of these views, rationalism about agential control, claims that the capacities for agential control in normal adult human beings are rational capacities. The second view, rationalism about action, claims that the capacities for agential control in virtue of which the things we do count as our actions are rational capacities. The arguments of the paper focus on aspects of technical skills that control integral details of skillful action, like the details of a baseball player’s pitching technique. I argue that these aspects of technical skills are largely non-rational capacities, but are nonetheless capacities for agential control, and can help make the things we do count as actions. While rational capacities do have a central role in human agency, their importance should not lead us to neglect the significant constitutive role of non-rational capacities in our agency.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
The gerundive form of the word ‘reasoning’ suggests that it refers to an interruptible process. I use the word to refer to exercises of rational capacities to hold a commitment (like a belief or an intention) on the basis of reasons. I do not assume that exercises of these capacities must be interruptable processes.
There is, of course, a lot of diversity within the phenomenological tradition. This point about the mistake of construing reason as legalistic has also been made from within that tradition. For example, in developing his phenomenology-inspired account of perception, Alva Noë argues that phenomenologists (he focuses on Hubert Dreyfus) have tended to “over-intellectualize the intellect.” He suggests that Dreyfus mistakenly construes what we do as thoughtful, intellectual beings, in terms of deliberate acts of explicit rule following (pp. 188–120). His criticism of Dreyfus is similar to McDowell’s (2007b).
Noë’s phenomenology-inspired views about the role of bodily skill in perception intersect with my argument in this paper. According to Noë, perceptual experience depends constitutively on sensorimotor skills, like the kinds of skills that determine details of how you throw a ball. This view is anti-rationalist, if you allow that sensorimotor skills are non-rational capacities—that is, that if you allow that they are not, as such, manifestations of our status as intellectual beings capable of thought and understanding. (Noë does say these skills are non-conceptual, but he also says the sensory expectations that sensorimotor skills give us are a kind of understanding, so maybe he would deny that sensorimotor skills should be classified as non-rational.)
On the surface, this looks to be an anti-rationalist view about perception, but not about agency. However, Noë thinks that perception is active rather than passive. So that might make perceptions manifestations of agency. If so, then Noë’s view would imply that certain manifestations of agency consist partly in the exercise of non-rational sensorimotor skills. Interestingly, Noë seems to think that perception counts as active in partly because of the role of bodily skills in perception. However, he does not take up the task of addressing rationalist opposition to the view that bodily skills are active, and are part of our agency. That issue is the specific focus of this paper. I argue that “low-level” non-rational capacities, including sensorimotor skills, are part of our agency. If Noë’s way of constitutively linking perception and sensorimotor capacities is successful, my argument would strengthen his claim that perception is active. I thank an anonymous reviewer for pressing me to clarify the relation between my position and Noë’s.
One might object to the claim that “brutely physical” aspects of the body are not rational capacities, as based on a mistaken dualist view about mind and body. I discuss the constitutive role these aspects of the body in agency in (citation omitted for anonymity).
In my view, when your action is explained by an understanding of what to do, intention must be involved. If you hold a belief that doing the dishes would be good, there is still a further question as to whether actually to perform the action which you believe would be good to perform. In holding the belief, you do not thereby have a point of view concerning that distinctively practical question of whether to perform the action. An understanding of what to do is a point of view concerning that practical question. If you intend to do the dishes, you do thereby have a point of view concerning that practical question of what to do. (I find Hieronymi (2006) particularly helpful on this issue.)
A conscientious rationalist will allow that this sort of awareness is often present even when it is not occurrent in the individual’s consciousness.
Phenomenological criticisms of rationalism often argue for this claim.
Without entering into too much detail, I want to touch on two interesting concerns raised by an anonymous reviewer. First, in normal action, a complex array of concerns is at play. A corresponding array of intentions will seem applicable to any particular action. For example, if you do the dishes, the “operative” intention could be an intention to do the dishes, or an intention to lather, rinse and dry the dishes, or an intention to clean up, or an intention to do your fair share of the housework, or an intention to save yourself from having to do the dishes in the morning, and so on. Which of these actually explains your action? The reviewer worries that there is no principled way of specifying which of these candidate intentions is doing the real work in explaining a particular action. In my view, they could all be explanatory. In normal human action, your point of view about what to do and why is complex and multifaceted. Accordingly, actions flow from a complex array of interconnected intentions, and not from a single operative intention. Michael Bratman’s (2007) work on planning agency discusses these issues in a helpful way.
Second, the reviewer raises a helpful question of why one should accept that it is the individual’s rational capacities that determine which intentions she sees as relevant to a particular action. I do think that your rational capacities play an important role in determining how you act. But, as I see it, it can lead to misunderstanding to say that your rational capacities determine which intentions you see as relevant to an action. If you see some intentions as relevant to your action, but not others, that need not be because of some antecedent exercise of rational capacities which determines what is part of your practical point of view. Your point of view is as such an exercise of your rational capacities. For example, suppose you see washing the dishes as doing your fair share for the household, but you do not see your action as a form of relaxation. That is a fact about your point of view. Your point of view might be a result of antecedent reasoning. But, even if it is not, your point of view is still an exercise of rational capacities. Inasmuch as you have your point of view at the level of thought and understanding, it manifests your nature as a being capable of thought and understanding. In that way, your point of view is an exercise of your rational capacities as such—irrespective of the role of antecedent exercises of rational capacities, like capacities for reasoning.
One might point out that a kind of epistemic agential control is involved in theoretical reasoning, and not just reasoning that determines how we act. I set aside that kind of epistemic agential control here.
Stanley and Williamson (2001) argue that know-how is a type of propositional knowledge, on the basis of a semantic analysis of the term ‘knows how.’ It is not clear that this view commits them to a rationalist claim about agency. Arguably, propositional knowledge can consist partly in sub-personal representations, or in the exercise of non-rational representational capacities, like motor representations. Butterfill and Sinigaglia propose a view along these lines (see footnote 23). It is also worth noting that the Rylean view opposed by Stanley and Williamson is not necessarily anti-rationalist. A Rylean would claim that certain intellectual know-how, like the ability to do logic, is not reducible to knowing-that. But in making that claim, the Rylean would not have to deny that the ability to do logic is a rational capacity.
The policy also has to be “reflexive,” and the agent has to be “satisfied” with the policy.
McDowell does not discuss the tension between this concession and his rationalist conception of agency.
The claim that the thoughts involved in exercises of rational capacities must be accessible for use in conscious thinking does not mean that exercises of rational capacities must always, or even normally, be episodes of conscious thinking.
Jeannerod finds that the motor system is active when an individual consciously imagines herself performing a physical action (see, e.g., (1995),(1994), and (1999)). This finding might seem to contradict the claim that representations involved in motor control are inaccessible to conscious thought. Let me touch on three points to make in response to this objection, though I cannot develop the points in detail here. First, if motor representations are accessible for use in consciously imagining action, that does not mean motor representations are generally accessible for use in conscious thought. Second, motor representations may affect conscious imagery of performing an action by giving the imagery a certain “feel”, without the agent consciously representing the content of the motor represention in the act of imagination. Third, and most importantly, Jeannerod’s research does not indicate that all the specific details of physical action represented by the motor system are accessible for use in imagining action. Jeannerod acknowledges that motor representations become inaccessible to conscious imagery at functionally “lower” levels of the motor system, where the most specific details of how an action is performed are specified (1995, p. 1429).
Stanley and Krakauer (2013) claim that fine-grained details of action do not manifest skill, since they are determined by low-level abilities, and are not subject to direct voluntary control (which is a kind of rational control). Fridland (2014) argues that many such details can count as manifestations of skill, since one can learn to exert voluntary control over the details through repeatedly attending to the details in training. The point here is that many fine-grained details of action manifest skill even though they are determined by low-level abilities. Many fine-grained details that are integral to an agent's skillful action are typically determined by low-level, non-rational capacities, and are not subject to direct rational control.
By “general concepts,” I mean, roughly, a concept that represents a property or an individual, where the individual can use the general concept to represent that property or individual on separate occasions, with the same mode of presentation of the property or individual. See Burge (2009) for a detailed discussion of different types of conceptual generality.
See Burge (2003).
There are points of contact between this idea and McDowell’s (1998) reply to this non-conceptualist argument. Though McDowell claims that the perception content is already conceptual, and would deny that any “drawing up” is required.
See McDowell (1998) for a line of argument along these lines.
That is, within constraints set by the individual’s beliefs about the demonstratum, like her belief that it is a person. If the witness were seeing a hologram, her demonstrative might fail to represent.
Things are more complicated with indexicals, and especially with the I concept. But reference to details of skilled action is not accomplished through the exercise of indexicals. So reference through indexicals can be set aside for present purposes.
Butterfill and Sinigaglia (2014) suggest that intentions can demonstratively represent specific details of action, by deferring to low-level motor representations of the details. If so, the demonstrative representation gives the agent a certain kind of awareness of the details of her action at the level of thought. But it is the motor representation itself that actually guides the details of the action, rather than the demonstrative that defers to the motor representation.
However, it seems likely that technical skills that are attributable at the whole individual level are constituted in part by sub-capacities that are sub-personal, and are not genuinely attributable at the whole-individual level. This fact may seem puzzling. One might expect that a capacity that is partly constitutive of a whole-individual capacity must itself be a whole-individual capacity. To see that this thought is mistaken, it may help to consider capacities attributable to groups. An ensemble’s ability to play well is attributable to the ensemble as a whole. The ensemble’s ability to play well might consist partly in the cellist’s ability to play well. The cellist’s ability to play well is “modular,” in that it is not attributable to the ensemble as a whole. Nonetheless, the cellist’s ability is still partly constitutive of the ability of the ensemble as a whole. Though of course there are important differences between group abilities and abilities of individuals, the analogy helps show how abilities attributable at the whole-individual level might consist partly in sub-personal capacities.
See Strawson’s “Freedom and resentment” (2003) for discussion of the relation between freedom, agency, and attitudes like these.
This proposal is similar to one made by Randolph Clarke (2010) about the nature of intentional action. Clarke suggests that an event might count as an intentional action because of its relation to an intention, even if it is not specifically represented in the content of the intention. For example, a particular stroke made in the course of shaving might count as an intentional action because it is caused in a certain way (i.e. non-deviantly) by the agent’s intention to shave, even if the particular stroke is not represented in any intention. Clark’s proposal is about whether causal connection to intention is what makes an event count as an intentional action. He presents this view in response to a criticism of Davidsonian causal theories of action, due to David-Hillel Ruben (2003). Though related, this issue is not exactly the same as the issue under discussion here, which concerns the role of rational capacities in making an event count as a manifestation of agency. If intentions can have non-conceptual content, then the Davidsonian view that events count as actions because of their causal connection to intentions is compatible with the claim here—that events count as actions partly because of their connection to non-rational capacities for agential control.
References
Bratman, M. (2007). Structures of agency. Oxford: Oxford University Press.
Burge, T. (2003). Perceptual entitlement. Philosophy and Phenomenological Research, 67, 503–548.
Burge, T. (2009). Five theses on de re states and attitudes. In J. Almog & P. Leonardi (Eds.), The philosophy of David Kaplan. Oxford: Oxford University Press.
Butterfill, S., & Sinigaglia, C. (2014). Intention and motor representation in purposive action. Philosophy and Phenomenological Research, 88, 119–145.
Clarke, R. (2010). Skilled activity and the causal theory of action. Philosophy and Phenomenological Research, 80(3), 523–550.
Davidson, D. (1980). Actions, reasons and causes. Essays on actions and events. Oxford: Clarendon Press.
Dreyfus, H. (2002). Intelligence without representation–Merleau-Ponty’s critique of mental representation. Phenomenology and the Cognitive Sciences, 1, 367–383.
Dreyfus, H. (2006). Overcoming the myth of the mental. Topoi, 25, 43–49.
Dreyfus, H. (2007). Detachment, involvement, and rationality: are we essentailly rational animals? Human Affairs, 17, 101–109.
Ennen, E. (2003). Phenomological coping skills and he striatal memory system. Phenomenology and the Cognitive Sciences, 2, 299–325.
Frankfurt, H. G. (1978). The problem of action. American Philosophical Quarterly, 15(2), 157–162.
Frankfurt, H. G. (1988). Identification and externality. The importance of what we care about. Cambridge: Cambridge University Press, 58–66.
Franklin, D. W., & Wolpert, D. M. (2011). Computational mechanisms of sensorimotor control. Neuron, 72, 425–442.
Fridland, E. (2014). They’ve lost control: Reflections on skill. Synthese, 191, 2729–2750.
Gigerenzer, G. (2007). Gut feelings: The intelligence of the unconscious. New York: Viking.
Gilligan, C. (1985). In Hester, E., & Alice J., (eds) A different voice: Women’s conceptions of self and of morality. The future of difference. New Brunswick, NJ: Rutgers University Press,
Heck, R. (2000). Nonconceptual content and the “space of reasons”. Philosophical Review, 109, 483–523.
Herman, B. (2007). Moral literacy. Cambridge, MA: Harvard University Press.
Hieronymi, P. (2006). Controlling attitudes. Pacific Philosophical Quarterly 87
Jeannerod, M. (1994). The representing brain: Neural correlates of motor intention and imagery. Behavioral and Brain Sciences, 17, 2.
Jeannerod, M. (1995). Mental imagery in the motor context. Neuropsychologia, 33(11), 1419–1432.
Jeannerod, M. (1999). Mental imaging of motor activity in humans. Current Opinion in Neurobiology, 9, 735–739.
Kelly, S. D. (2000). Grasping at straws: Motor intentionality and the cognitive science of skilled behavior. In M. Jeff & W. Mark (Eds.), Heidegger, coping, and cognitive science: Essays in honor of Hubert L. Dreyfus (pp. 161–178). Cambridge: MIT Press.
McDowell, J. (1979). Virtue and reason. Monist, 62, 3.
McDowell, J. (1996). Mind and world. Cambridge, MA: Harvard University Press.
McDowell, J. (1998). Reply to commentators. Philosophy and Phenomenological Research, 58, 403–431.
McDowell, J. (2007a). Response to dreyfus. Inquiry, 50(4), 366–370.
McDowell, J. (2007b). What myth? Inquiry, 50(4), 338–351.
Pacherie, E. (2008). The phenomenology of action: A conceptual framework. Cognition, 107, 179–217.
Peacocke, C. (1998). Nonconceptual content defended. Philosophy and Phenomenological Research, 58, 381–388.
Raz, J. (1997). The active and the passive. Proceeding of the Aristotleian Society, 71, 221–227.
Reber, S. (1994). Parallel brain systems for learning with and without awareness. Learning and Memory, 1, 217–229.
Rietveld, E. (2008). Situated normativity: The normative aspect of embodied cognition in unreflective action. Mind, 117(468), 973–1001.
Ruben, D.-H. (2003). Action and its explanation. Oxford: Oxford University Press.
Shadmeh, R., & Mussa-Ivaldi, S. (2012). Biological learning and control. Cambridge: MIT Press.
Squire, L. R. (2004). Memory systems of the brain: A brief history and current perspective. Neurobiology of Learning and Memory, 82, 171–177.
Squire, L. R. (2009). Memory and brain systems: 1969–2009. The Journal of Neuroscience, 29, 12711–12716.
Stanley, J., & Krakauer, J. (2013). Motor skill depends on knowledge of facts. Frontiers of Human Neuroscience
Stanley, J., & Williamson, T. (2001). Knowing how. The Journal of Philosophy, 98(8), 411–444.
Strawson, P. (2003). Freedom and resentment. In Gary, W. (ed) Free will, Oxford University Press, 2nd ed.
Velleman, J. D. (2000). The possibility of practical reason. Oxford: Oxford University Press.
Wiggins, D. (1975–1976). Deliberation and practical reason. Proceeding of the Aristotleian Society 76:29–51.
Willingham, D. B. (1998). A neuropshycological theory of motor skill learning. Psychological Review, 105, 558.
Willingham, D. B., Nissen, M. J., & Bullemer, P. (1989). Development of procedural knowledge. Journal of Experimental Psychology, 15, 1047.
Acknowledgments
I want to thank Tyler Burge, Christopher Carlton, Pamela Hieronymi, Sean Kelsey, Gavin Lawrence, and Lucy O'Brien. I am also grateful to an anonymous reviewer for helpful comments.
