Hostname: page-component-7b7567c8c-2n9c9 Total loading time: 0 Render date: 2026-11-14T08:22:19.862Z Has data issue: false hasContentIssue false
  • English
  • Français

Agency and Practical Abilities

Published online by Cambridge University Press:  16 May 2017

Will Small
Will Small*
Affiliation:
University of Illinois at Chicago
*
wsmall@uic.edu
Rights & Permissions [Opens in a new window]

Abstract

Though everyday life accords a great deal of significance to practical abilities – such as the ability to walk, to speak French, to play the piano – philosophers of action pay surprisingly little attention to them. By contrast, abilities are discussed in various other philosophical projects. From these discussions, a partial theory of abilities emerges. If the partial theory – which is at best adequate only to a few examples of practical abilities – were correct, then philosophers of action would be right to ignore practical abilities, because they could play no fundamental role in an account of human agency. For the idea that practical abilities do play a fundamental role in human agency to be worth considering, an alternative conception of them is needed. As a first step, I attempt some of the necessary ground-clearing work.

Information

Type
Papers
Information
Royal Institute of Philosophy Supplements , Volume 80: Philosophy of Action , July 2017, pp. 235 - 264
Copyright
Copyright © The Royal Institute of Philosophy and the contributors 2017 

Introduction

In our everyday thought and conduct, we place great weight on the acquisition, maintenance, and development of practical abilities – such as the ability to walk, the ability to sing, the ability to play the piano, and the ability to speak French. Practical abilities are an object of focus in many developmental, educational, and therapeutic contexts. The evidence of the everyday would thus seem to bear out Randolph Clarke's claim that ‘Abilities are fundamental to agency; we don't have a decent comprehension of agency without an understanding of them’.Footnote 1 It ought to be striking, then, that contemporary philosophy of action has remarkably little to say about practical abilities.Footnote 2 By contrast, philosophers working on other topics have quite a lot to say about them. From their work a partial theory of practical abilities seems to have emerged. But, I will argue, the partial theory is flawed: it at best works for some practical abilities only. Moreover, it presupposes that practical abilities are not, in fact, fundamental to agency. The partial theory and the status quo in the philosophy of action are thus such as to reinforce each other. My main goal is to bring into view some features of practical abilities the recognition of which may help us to see just what it would be to think that they are fundamental to agency.

Philosophers of action characteristically seek to understand what human action is by examining how it is explained.Footnote 3 And what strikes them as distinctive about their subject matter is that we very often explain human actions in terms of the reasons for which the agent acted: human action is such as to admit of this form of explanation. Philosophers disagree about whether the explanation of action in terms of reasons is a kind of non-causal explanation or a distinctive kind of causal explanation, but they agree that the fact that actions can be explained in terms of their agents' reasons for acting distinguishes them from other sorts of phenomena. Many of the questions philosophers of action are interested in concern the nature of these explanations: Are the reasons for which an agent acts facts, or known facts, or believed propositions, or beliefs? Can desires be reasons? Must a reason for φ-ing purport to show that φ-ing is good? And so on. The purpose of a reasons explanation is to lay out which among the various motives and reasons someone might have had for so acting were in fact the agent's, and thereby to render it intelligible why she should have acted as she did – and, perhaps, not in some other way. Such explanations presuppose that an action has occurred; they do not purport to tell us what it is for someone to have acted.Footnote 4 Thus inquiry into the form and elements of reasons explanations of action may reasonably ignore the question whether or not agential abilities are essentially involved in an account of what it is for someone to act. That Amy has the ability to play the piano would not, or at least not usually, figure in a specification of her reasons for playing the piano on an occasion: her reason might have been that she wanted to please Belinda, and knew that Belinda would be pleased if she played the piano for her. The question ‘Why are you playing the Moonlight Sonata on the piano?’ presupposes that the agent is playing the piano and that she has the ability to play the piano. That ability is hardly likely to be cited in her response to the reason-seeking ‘Why?’-question; it is more likely to figure in an answer to some such question as ‘How are you making music come out of that piece of furniture?’ A reasons explanation tells us why someone acted as she did, but more may belong in an adequate account of why and how some particular action occurred than belongs in an account of why someone φ-ed on an occasion: more may be involved in accounting for the why and how of a particular event than in explaining why an event of some particular type occurred.

By contrast, the so-called ‘standard story of action’ aims to take the materials of a reasons explanation and fashion them into a reductive account of (intentional) action. According to the standard story, what it is for someone to have raised her arm, say, is for appropriate mental states (for instance, a desire to attract the chair's attention and a belief that raising one's arm is a way to do so) to have appropriately caused her arm's rising. If an arm rising has the right kind of causal ancestry, it counts as an arm raising (an action, something the agent did); if an arm rising has some other kind of causal ancestry, it is a ‘mere’ arm rising (not an action, but something that merely happened to the person whose arm it is). The idea of an arm rising, or a ‘bodily movement’ more generally, is understood as something intrinsically non-agential and taken as fundamental. The standard story and its variants appeal only to mental states (such as belief, desire, intention), mental processes (such as deliberation or practical reasoning), bodily movements, and event-causation: abilities do not figure among the explanatory elements.Footnote 5

Practical abilities do not figure in the standard story. And it is sometimes objected that neither do agents: the standard story is said by some to ‘leave the agent out.’Footnote 6 I share Jennifer Hornsby's sense that these issues are conected:

when an account of the causal transaction in a case of agency is given in the claim that a person's believing something and a person's desiring something causes that person's doing something, … [t]he fact that a person exercises a capacity to bring something about is … suppressed. It is forgotten that the agent's causal part is taken for granted as soon as she is said to have done something. The species of causality that belongs with the relevant idea of a person's exercising her capacities is concealed.Footnote 7

Once we have in view someone's having done something, we have in view her having exercised some of her agential capacities; in enquiring after why she did what she did, we may explicate why she exercised those capacities, but not what it was for her to have so exercised them. But if we begin with a bodily movement that, for all we know, may or may not be an action (may or may not have been someone's exercising of some practical abilities of hers), and come to learn that it was an action (because it was appropriately caused by appropriate mental states), then we have not yet said anything about the role of the agent's capacities and their exercise in her acting.

Hornsby suggests that ‘the agent's causal part’ (her acting, her doing something) resides in her ‘exercising her capacities'. But this is not an idea that she further explicates. I think it would be worth explicating. But it will be difficult to do so if we fail to get clear on the basic shape of agential capacities of the relevant sort – what I am calling practical abilities. And the partial theory obscures this basic shape.

The partial theory emerges piecemeal from philosophical work on a variety of topics. For instance, in epistemology, virtue epistemologists such as Ernest SosaFootnote 8 and John GrecoFootnote 9 have attempted to account for the nature and value of knowledge in terms of its being an exercise of epistemic abilities or competences, while John HymanFootnote 10 has argued that one knows that p if and only if one has the ability to φ for the reason that p. In discussions of free will, moral responsibility, and determinism, something called ‘the ability to do otherwise’ is often at issue.  Kadri VihvelinFootnote 11 has recently argued that free will and moral responsibility do require the ability to do otherwise, but that this is compatible with determinism.Footnote 12 These philosophers very often appeal to practical abilities – such as the ability to drive a car, or an archer's ability to hit the bull's-eye – in order to shed light on (what they take to be) abilities that are less familiar to us, but which (they think) share something of the structure of practical abilities.Footnote 13 In their attempts to precisify the source of their analogies, elements of the partial theory emerge and coalesce. Another important source of the partial theory comes from work on the metaphysics of dispositions or powers.Footnote 14 Though work in this area typically focuses on such dispositions as fragility and solubility, it usually aspires to a systematic account of dispositionality or potentiality that encompasses human qualities such as the habit of smoking, the character trait of irascibility, and the ability to play the piano. The basic structure of an understanding of practical abilities is here taken to be supplied by an understanding of potentiality or the dispositional as such.Footnote 15

Philosophers working on these topics frequently provide a partial account of ability in order to make, or in application of, the claims they're really interested in making. This is, of course, perfectly legitimate: they have their own interests and agendas. But the conception of practical abilities that emerges thereby, and which to my mind seems to be becoming the received view on the matter, is deeply flawed. At the very least, it ought to be recognized as holding for (at best) some practical abilities only, and as embodying commitments that philosophers of action should reject. More significantly, it encourages a conception of abilities on which they are not ‘fundamental to agency’ because they play no genuine explanatory role in the theory of action. The claims of the partial theory that I wish to question are these: (1) A practical ability is a disposition of a special kind, distinguished by the fact that its subject is an agent and its manifestation is an action. (2) Practical abilities are to be specified in terms of act-types (that is, the ‘φ’ in ‘ability to φ’ ranges over act-types) and their manifestations are tokens of the relevant type.Footnote 16 (3) It is possible for an agent to φ despite lacking the ability to φ. This is because possessing the ability to φ requires more than the possibility (or even actuality) of φ-ing on some occasion, namely the possibility of φ-ing in a sufficiently wide range of circumstances; thus (4) an agent's ability to φ improves, or is better than another agent's, when she is such that her attempts to φ succeed more frequently, or across a wider range of circumstances.

1. Preliminaries

Before proceeding, however, two clarifications are in order. My topic is practical abilities, which I introduced by some examples. The characteristic exercises of practical abilities are intentional actions: someone exercises her ability to walk in walking to the shops or in going for a stroll, her ability to play the piano in playing Für Elise, and so on. But we ascribe to ourselves and others abilities that are not practical abilities in my sense: for instance, the ability to understand French, the ability to tell a hawk from a handsaw, and the ability to fall asleep on buses.Footnote 17 Whereas the ability to speak French is exercised in speech acts, which are intentional actions, the ‘acts of understanding’ in which someone manifests her ability to understand French are not intentional actions. The use of ‘act’ here is that of the contrast between power and act, a contrast which applies to passive powers as much as active ones. The acts of abilities to understand things, like the exercises of recognitional abilities in perception, are in an important sense passive; they are thus not practical abilities in my sense.Footnote 18 And though falling asleep on a bus is something one may be said to do, it might equally be said to be something that happens to one. There are various intentional actions one might perform in order to fall asleep, but falling asleep – on a bus or anywhere else – is not an intentional action (though it might be voluntary or involuntary), and the ability to fall asleep on buses is not a practical ability. Moreover, there may be things that someone can do, where her doing them would be an intentional action, but what is exercised is not a practical ability but rather some authority. By virtue of her authority, a registrar can – is able to – institute a marriage by saying things that anybody has the ability to say; but the ‘ability’ to institute a marriage, if one wishes to speak that way, is not a practical ability. The abilities under discussion in this paper are practical abilities.

With the scope of my topic thus sketched in some more detail, I pass to the second clarification. Some philosophers distinguish different kinds of ability on the basis of different senses or uses of expressions like ‘S is able to φ’ or ‘S has the ability to φ’. A distinction is frequently drawn between ‘general abilities' and ‘specific abilities’. A tennis player is said to have the ‘general ability’ to play tennis, something that someone who has never played it lacks; but in the absence of a racquet, or a ball, or an opponent (and so on) she is said to lack the ‘specific ability’ to play tennis: she can, or is able to, play tennis generally, but she can't, or isn't able to, play tennis here and now. An agent who has a specific ability to play tennis needn't actually be playing tennis, but, as John Maier puts it, ‘there is, as it were, nothing between her and the deed’.Footnote 19 Some philosophers distinguish a third kind of ability that sits between these two that corresponds to an agent's having ‘the necessary skills [i.e. the ‘general ability’] and the psychological and physical capacity to use those skills’Footnote 20 even though she may lack the necessary ‘external’ factors (such as access to equipment, etc.) required for ‘specific ability’.Footnote 21

It is at best misleading to speak of either two or three different kinds of ability on the basis of considerations such as these. There are three different kinds of factors that might explain why someone is not in a position to do something: she may lack the ability to do it, she may lack the opportunity to do it, or she may prevented from exercising her ability by inebriation, injury, fear, phobia, or unconsciousness (and so on). Of course, in ordinary English we may say that she is unable to do it – and perhaps that she lacks that ability to do it – for any of these reasons. But in our thought about the metaphysics of agency, we ought to be more precise: abilities, opportunities, and impediments are quite different sorts of explanatory factors. Abilities are ‘inherently general’, in Kenny's phrase,Footnote 22 whereas opportunities and impediments are constituted by particular situations, conditions, or occasions. Serena Williams has the ability to hit a forehand winner down the line, on the run. It is the same ability she exercises whenever she hits one, but the opportunities come and go, and some she does not take. When inebriated or injured, Williams does not lack a physical ability to exercise her ability to play tennis; rather, she has a condition that prevents her from exercising her tennis-playing abilities.Footnote 23 Of course, in saying that someone is able to do something we may be drawing attention to some combination of ability, opportunity, and lack of impediments to exercising ability; but this does not justify invoking a kind of ability that is constituted by the co-presence of these factors.Footnote 24 The ideas of ‘specific ability’ and ‘narrow ability’ are misnomers; moving forward, then, I shall assume that the only practical abilities are what many call ‘general abilities.’Footnote 25

2. Abilities and dispositions

If anything constitutes a default view about abilities, it is that they are dispositions – probably dispositions of some distinctive kind.Footnote 26 To say that Amy has the ability to play the piano, or to run a five-minute mile, or to cook spaghetti bolognese is not to say anything about what she is currently doing. Exactly what it is to say, on this prevailing view, depends on how dispositions are understood. According to some,Footnote 27 dispositions are to be understood in terms of conditionals; thus to say that someone has an ability is to say something about what she would do – if such-and-such either happened or were the case. For instance, it might be held that to have the ability to φ is to be such that, were you to have the opportunity to φ and try to φ, you would φ.Footnote 28 According to others,Footnote 29 dispositions are to be understood in terms of restricted possibilities; thus ability ascriptions do not say what an agent would do if certain conditions obtained, but simply what she can or could do. The idea is that we understand the claim that it is possible for someone to φ in terms of the idea that there is some possible world in which theyFootnote 30 φ, and then understand the more specific idea that someone has the ability to φ (that they can or could ‘in the ability sense’ φ) in terms of the idea that there is some possible world from among a restricted set of possible worlds in which they φ.Footnote 31

On the restricted possibility view of ability, what distinguishes an ability from a mere disposition is that the subject of an ability is an agent – or perhaps an agent insofar as she is an agent – and that the manifestation of the ability is an action. (More would need to be said, of course, in order to distinguish practical abilities from other sorts of agential dispositions to act – for instance, habits.) The conditional view of ability can appeal in addition to the distinctive role it grants to trying. In §5 below, I will return to this difference. But my main concerns abstract from any argument between proponents of conditional and restricted possibility views of ability. No doubt the question whether to prioritize conditionals or possibility in one's thought about dispositions is a significant one. My main focus, however, will be on problems that a conception of ability inherits when it is assumed that an account of dispositions of either of these sorts may be used as it basis: when it inherits what is common to both views.Footnote 32

Kenny writes: ‘A skill or ability is always a positive explanatory factor in accounting for the performance of an agent; an opportunity is often no more than a negative factor, the absence of circumstances that would prevent or interfere with that performance’.Footnote 33 How would philosophers of action have to conceive – or reconceive – of abilities for them to be candidates to play a genuinely explanatory role? If I'm right, a substantial rethink is in order. I propose to explore three issues, and the connections between them: the robustness or reliability of abilities; the gradability of abilities; and the individuation and specification of abilities.

3. Ability and success

Abilities are fallible, as a famous example from J.L. Austin illustrates:

Consider the case in which I miss a very short putt and kick myself because I could have holed it. It is not that I should have holed it if I had tried: I did try, and missed. It is not that I should have holed it if conditions had been different: that might of course be so, but I am talking about conditions precisely as they were, and asserting that I could have holed it. There is the rub. … [A] human ability or power or capacity is inherently liable not to produce success, on occasion, and that for no reason (or are bad luck and bad form sometimes reasons?).Footnote 34

An agent may have the ability and opportunity to φ, try to φ, and yet fail to φ. (According to Austin, it seems, this may be so even when nothing interferes.) But though they are fallible, abilities are also supposed to be reliable and robust: it is no accident if someone with the ability to φ succeeds in φ-ing. Thus accounts like the following are common:

S has the narrow ability at time t to do R as a result of trying [ = ‘in response to the stimulus of S's trying to do R’] iff, for some intrinsic property B that has S has at t, and for some time tʹ after t, if S had the opportunity at t to do R and S tried to do R while retaining property B until time tʹ, then in a suitable proportion of these cases, S's trying to do R and S's having of B would be an S-complete cause of S's doing R.Footnote 35

[A]bilities have the following structure: S has an ability A(R/C) relative to environment E = Across the set of relevantly close worlds W where S is in [conditions] C and in E, S has a high rate of success in achieving [result] R.Footnote 36

More straightforwardly, Kenny says that ‘There is this much truth in the conditional analysis of ability’:

‘I can φ’ entails: if I have the opportunity to φ, and if I do my best to φ, then I normally will φ.Footnote 37

What makes for a ‘suitable proportion of cases’ or an appropriately ‘high rate of success’, and what it takes to ‘normally’ succeed, will differ for different abilities. In some cases, a 50% rate of success would be evidence of lack of ability: it's luck, not ability, that accounts for my getting heads when I flip a coin.Footnote 38 But sport provides many examples of abilities for which a 50% success rate would be evidence of outstanding or even unthinkable ability. The basketball player Stephen Curry, an outstanding three-point shooter, made just over 45% of his three-point shots in 2015–16. The best hitters in Major League Baseball usually hit the ball in play between 29% and 35% of the time. And Dale Steyn – cricket's best bowler – has the ability to take wickets if anyone does, but he takes a wicket only once every 41 balls he bowls. So ability can tolerate quite low probabilities of success, depending on the difficulty of the task; and lack of ability can be consistent with relatively high probabilities of success.Footnote 39

Even though it may be difficult to say what counts as a normal or suitably high rate of success in different cases, views like those of Vihvelin, Greco, and Kenny seem to provide a natural way of saying what it is for an ability to come in degrees. The ‘gradability’ of  abilities is an idea well-entrenched in ordinary thought and practice. Amy is a better runner than Belinda, Brian a more able swimmer than Charles, Chloë shows greater ability at darts than David, and I am trying to get better at playing the piano. These accounts provide for at least two dimensions in which one agent's ability might be greater than another's. On the one hand, the more able agent's attempts might succeed at a higher rate (this is what baseball batting averages record, for instance). On the other hand, there might be more circumstances in which her attempts succeed: the more able the agent, the more circumstances constitute opportunities (twenty knot winds may prevent a beginner sailor from going out on the water while providing a more experienced sailor with an opportunity for an exciting sail).Footnote 40 Accounts that have a place for trying – depending on how it is construed – may provide for another dimension of gradability: Belinda has the ability to run a six-minute mile, but only if she really, really tries, whereas Amy, a better runner, can run a six-minute mile, without really trying. She barely breaks a sweat.

These criteria – higher rate of success and success in more circumstances – seem well-suited to account for the gradability of dispositions such as fragility and flammability (for ‘success’ read ‘manifestation’) and of some abilities. Stephen Curry's extremely high ability to make three-point shots is shown both by his making them at a much higher than average percentange and also by his succeeding with more difficult attempts than his peers. But the account looks much less promising when we think about those abilities whose exercises do not have the neat, binary conditions for success or failure (like making a basket, hitting the bull's eye, and so on) that match those of the metaphysicians’ favourite dispositions, fragility and solubility (it either breaks or it doesn't, or it dissolves or it doesn't).

Consider: Belinda can run a six-minute mile, but only if she really tries, whereas Amy can run a six-minute mile with ease. It is not that Amy is such as to succeed in running a six-minute mile with a higher rate of success than Belinda. (This might not be true: though Belinda really needs to try, perhaps she always gives 100% in everything she does.) Nor is it that Amy, unlike Belinda, can run a six-minute mile not only when it's warm and dry but when it's cold and wet, on grass, not only on the track, and so on. (This might not be true either: Amy might be unusually sensitive to such things.) Amy's greater ability as a runner doesn't have to do with how frequently, or under what conditions, she runs a six-minute mile. That she, unlike Belinda, can run a six-minute mile with ease suggests that she can run a mile in well under six minutes. She does not have the ability to run a six-minute mile better than Belinda; she is a better miler because she, unlike Belinda, can run a five-minute mile – this explains why when she decides to run a six-minute mile she can do so with ease. She is a better miler because she runs faster miles. Indeed, for many of the ‘physical’ or athletic abilities that we grade, being better at doing something will be understood in terms of physical magnitude: running faster, throwing things further, doing more push-ups, lifting heavier weights. Chloë’s ability to do push-ups is greater than David's, not because she is such as to succeed in doing a push-up with a higher degree of success than he is, or because she can do push-ups in a wider range of circumstances, but because she can do more push-ups.

There is another way in which the gradability of ability is not captured by the two dimensions provided for by views like Vihvelin's, Greco's, and Kenny's. Evgeny Kissin is a much, much better pianist than me. He can, indeed, play the piano faster than me – but that seems an unpromising way to account for his superiority. No doubt he can play more pieces than me; but more importantly, he can play more difficult pieces than me. He can play the Appassionata Sonata; I cannot. If I tried to play the Appassionata, I would fail. But I wouldn't fail to play the piano – the opening bars aren't very difficult. Playing the piano badly is still playing the piano. It's not that Kissin's ability to play the Appassionata is greater than mine because when he tries to play it, he succeeds, whereas were I to try, I would fail: I don't have the ability to play the Appassionata. We wanted to understand how my ability to play the piano is exceeded by Kissin's, yet it would be a rare occurrence indeed for me to try but fail to play the piano. Kissin is a better pianist than me not because he ‘succeeds’ in playing the piano with a higher rate of success or in a wider range of circumstances than I do, but because he plays more accurately and more musically than I can, and he can play more challenging pieces. Being better at φ-ing may often be a matter not, or not merely, of one's successes being more frequent or reliable, but of one's successes being better successes. Think of the craftsman or artist who turns out a better product: Brian is better than Charles at making pots, because the pots he makes are better pots. There are evidently normative or evaluative dimensions to the gradability of ability that do not belong to the accounts of what it is for one object to be more fragile than another, one substance to be more flammable than another, and so on.

As well as undermining the thought that the gradability of ability can be understood through those ideas – rate of manifestation in similar conditions, range of conditions of manifestation – that seem suited to accounting for the gradability of dispositions such as fragility and flammability, these examples raise a further question. In each of them, we find it natural to characterize the degree or grade of an agent's ability to φ in terms of whether or not she can exercise that ability by φ-ing in some determinate way. The fact that Amy can run a five-minute mile is a reason to praise her ability to run, while the fact that I cannot play the Appassionata may figure in a more muted evaluation of my ability to play the piano. But is it that Kissin's ability to play the piano exceeds in degree an ability I possess (the ability to play the piano) because he possesses an ability (the ability to play the Appassionata) that I lack? More generally, how are abilities to be specified or individuated?

4. Individuating and specifying abilities

The proper specification and individuation of abilities is not an issue that has received much philosophical attention. The following list of abilities is representative of the rather relaxed prevailing standards in the literature:

  • • to move one's limbs;Footnote 41

  • • to raise my hand; to sink a 3’ putt; to dress in a kilt;Footnote 42

  • • to stand on my head on commuter trains; to row a straight line in a crosswind;Footnote 43

  • • to sing; to sing when one's aunt is present;Footnote 44

  • • to read Emma; to read Persuasion; to read Werther;Footnote 45

  • • to speak French; to ride a bicycle; to sing in tune; to juggle;Footnote 46

  • • to kiss the Blarney Stone; to know that p;Footnote 47

  • • to pronounce the name ‘Luigi’; to marry for the sake of love; to marry for the sake of money; to φ for the reason that p.Footnote 48

The list contains some highly general abilities (e.g. the ability to move one's limbs – which ones? how?); some extremely specific abilities (e.g. the ability to read Emma, the ability to stand on my head on commuter trains); abilities that require particular objects, not just objects of particular kinds, for their exercise (the ability to kiss the Blarney Stone); and we find the ability to do something for one reason distinguished from the ability to do the same thing for a different reason. And of course we find such humdrum and everyday practical abilities as the ability to speak French, to sing in tune, and so on. Again, I don't propose to legislate on whether or not it is correct speech to ascribe any of these abilities: ordinary language is used for more than metaphysics. But we may ask: is someone who is reading Emma exercising an ability to read Emma – does an ability to read Emma belong in an explanation of her action – or is she reading it by exercising more general abilities, such as the ability to read English? And will the correct account of the abilities exercised in reading Emma parallel that of those exercised in playing the Appassionata?

To make progress, let us consider a rare concrete proposal for the individuation of abilities, due to John Hyman:

for all φ and ψ, the ability to φ and the ability to ψ are different abilities if it is possible to have one without having the other. It follows that the ability to run a mile in four minutes and the ability to run a mile in five minutes are different abilities, since it is possible to have the second without having the first, even though it is not possible to have the first without having the second.Footnote 49

By a continuation of the same reasoning, it would follow that the ability to run a mile is yet a third ability: Belinda has the ability to run a mile, but she can't run a five-, let alone a four-, minute mile; Amy can run a five- but not a four-minute mile; and Alex can run a four-minute (and thus a five-minute) mile. So by Hyman's lights we have three different abilities – of which Belinda possesses one, Amy two, and Alex all three. Parallel reasoning would suggest that we distinguish the ability to play the piano, the ability to play Für Elise, and the ability to play the Appassionata. Someone may have the ability to play the piano without having the ability to play Für Elise or the ability to play the Appassionata, and though possessing either of the more specific abilities entails the ability to play the piano, neither of them entails the other: an average pianist may have the ability to play Für Elise quite well, but not the difficult Appassionata, and a strong pianist who has learnt the Appassionata may simply never have had Für Elise in her repertoire.

Though he claims that such abilities are distinct, Hyman acknowledges that they are nevertheless ‘related’: ‘running a mile in five minutes … and … running a mile in four minutes … are both species of the generic act of … running a mile…, and the corresponding abilities are related in the same way’.Footnote 50 Presumably he would say the same thing in the other case: the ability to play the Appassionata and the ability to play Für Elise are species of the generic ability to play the piano. But it is not clear that this account of the relations is correct. It is true that running a mile takes time, so any act of running a mile will take a certain period of time; anyone who has the ability to run a mile will possess it to some degree, and that degree may be specified by a time. But the species of a genus, or the determinates of a determinable, exclude each other. Nothing can be a mammal without being either a human or a cow or a cat etc.; but equally nothing can be both a cow and cat. Nothing can be unicoloured all over without red all over or green all over or blue all over, etc.; but equally nothing can be red all over and green all over. Yet Alex has the ability run a four-minute mile and the ability to run a five-minute mile: the ‘species abilities’ do not exclude each other.Footnote 51 Neither the act-types nor the abilities are species of a generic act of running a mile or ability to run a mile.

Hyman might respond that though biological species exclude each other, a logical conception of the species-genus relation is available, one on which the species of a genus do not exclude each other. So there would be no problem in Alex having multiple species of the generic ability to run a mile: the ability to run a four-minute mile, the ability to run a five-minute mile, the ability to run a six-minute mile, and so on. But such a proposal still seems to multiply abilities unnecessarily, as Don Locke notes:

I can, for example, walk down Park Street, and I can also walk down State Street: I will succeed in doing these things if I try. But this surely does not mean that I have two abilities, the ability to walk down Park Street and the ability to walk down State Street; there is but one ability, the ability to walk, manifested in two different contexts. Again, the athlete who runs a four minute mile also, in the same performance, runs a five minute mile, but he does not thereby exercise or manifest two abilities. By that token he would be manifesting an infinity of abilities!Footnote 52

The ability to walk is something that has to be acquired. Once acquired, one can exercise it walking down a particular street. (Indeed, one cannot exercise it except by walking in some particular place, for some particular amount of time, in particular conditions – but surely there is such a thing as the ability to walk.) The ordinary human acquisition of the ability to walk is the acquisition of an ability that can be exercised in walking down Park Street or State Street or many other streets; on grass or on carpet or on tarmac; at dawn or at dusk; on Monday or on Tuesday. (It is not typically the acquisition of an ability that can be exercised in walking on hot coals or on ice or in space – these, if possessed at all, would usually be developed later. And it is never the acquisition of an ability that can be exercised in walking on water – contrast some insects.) Amy's ability to walk is sufficient to explain how it was possible for her (given the opportunity and motivation) to walk down State Street on Monday: there is no need to ascribe to her the ability to walk down State Street, the ability to walk on Mondays, or the ability to walk down State Street on Mondays in order to account for this; indeed, such ‘abilities’ would seem to trail after the performances they might be supposed to explain. So though particular streets may belong in the specification of what is done on some occasion when someone exercises her ability to walk (on Monday morning Amy did not merely walk; she walked down State Street from the river to Macy's), the streets do not belong in the specification of the abilities that are exercised in walking down them. However, this does not conflict with Hyman's principle for the individuation of abilities: one might accept that principle while taking a stricter stance than Hyman on the question of what are legitimate substitution instances for φ in ‘the ability to φ’.

Whereas the ordinary acquisition of the ability to walk would put someone in a position to walk down (say) State Street, the ordinary acquisition of the ability to run need not put someone in a position to run a mile, let alone a five-minute mile: stamina, strength, and training will characteristically be required. Going to a new city affords new opportunities for exercising one's ability to walk: there are new streets for one to walk down. It is the same thing that one is doing (walking) and the same ability that one is exercising (the ability to walk) on new and familiar streets alike. But the result of training is not like the result of travel: when, after months of training, Amy can finally run a five-minute mile, there is something she can do that she lacked the ability, and not merely the opportunity, to do before. And this would seem to be a reason for thinking that practice and training enlarge an agent's stock of abilities: Amy acquires the ability to run a five-minute mile, something she lacked before.

Why does Locke reject this? In running a five-minute mile (in running a mile inside of five minutes), Amy ran a six-minute mile (ran a mile inside of six minutes). But though she did many things – though her action exemplified many act-types – it would be absurd, Locke implies, to hold that she exercises or manifests an ability corresponding to each thing done. Of course there is no problem with an action manifesting many – perhaps infinitely many – properties. And there can be no problem with the idea that an agent might exercise more than one ability at the same time: doing so is surely the norm rather than the exception. (In writing this essay I am exercising both my ability to compose English prose and my ability to use a computer.) The problem seems to have to do with the idea that the agent might be exercising too many abilities, or at least too many abilities of the same sort. Though Amy ran a six-minute mile in running a five-minute mile, surely the only mile-running ability she exercised was the ability to run a five-minute mile. So perhaps one could agree with Locke that Amy does not exercise both the ability to run a five-minute mile and the ability to run a six-minute mile, while insisting with Hyman that they are distinct abilities: having recently acquired through training the ability to run a five-minute mile, this and this alone is the mile-running ability that Amy exercises when she sets out to run a (fast) mile.

A difficulty remains, however. If Amy tries to run a five-minute but for some reason succeeds only in running a six-minute mile, which ability did she exercise? One might hold that she exercised her ability to run a six-minute mile, the ability that she possessed prior to, and has retained since, her recent training regime led to her acquisition of the additional ability to run a five-minute mile. But this seems problematic: we seem to be picturing Amy as in the dark about which of her abilities she is exercising while she is acting. Alternatively, one might hold that when she acquired the ability to run a five-minute mile it replaced her prior ability to run a six-minute mile, and that it is the new ability that she exercised in running what turned out to be merely a six-minute mile on this occasion (practical abilities are fallible, after all). But this view seems like a less perspicuous notational variant on the idea that, through her training, Amy developed her ability to run: that ability used to be such that when she tried her best and things went well, she ran a six-minute mile; now it is such that when that happens, she runs a five-minute mile – though she can exercise that ability in running a mere six-minute mile if she so chooses. Amy's ability to run a mile is an ability to run a five-minute (and thereby a six-minute) mile.

Locke's principle for individuating abilities, which does so in a more coarsely-grained way than Hyman's, seems to capture this: ‘if those features from an agent's constitution and background which bring it about that he standardly succeeds in doing x are also sufficient to bring it about that he standardly succeeds in doing y, then we are dealing here not with two separate abilities, but with the one ability, to do x and do y’.Footnote 53 Amy can run a five-minute mile; a fortiori she can run a six-minute mile. Those features of Amy's constitution and background that underwrite her standardly succeeding in running a five-minute mile are also sufficient for her standardly succeeding in running a six-minute mile. By Locke's principle, it would seem that we are dealing here not with two separate abilities, but with a single ability: the ability to run a five-minute mile and the ability to run a six-minute mile are the same ability. Belinda, however, can run a six-minute mile but not a five-minute mile. Those features of her constitution and background that underwrite her standardly succeeding in running a six-minute mile are not sufficient for her standardly succeeding in running a five-minute mile: by Locke's principle, we are dealing with two separate abilities. And so when we put these results together, they suggest, absurdly, that the ability to run a five-minute mile both is and is not the same ability as the ability to run a six-minute mile.

The absurdity can be avoided by interpreting Locke's principle as yielding the verdict that Amy's ability to run a five-minute mile and her ability to run a six-minute mile are the same ability, and that Belinda's ability to run a six-minute mile is not an ability to run a five-minute mile. On this interpretation of Locke's principle, then, there is no answer to the question whether the ability to φ and the ability to ψ are the same ability or not; there are answers only to the question whether so-and-so's ability to φ and her ability to ψ are the same ability or not. But that seems to me just right: practical abilities are acquired, and some particular agent's acquisition of an ability is a particular process that particularizes the ability in her. Perhaps the point is better put like this: Amy and Belinda both have the ability to run, but in each of them that ability is differently specified in concreto. Amy's ability to run is such that she can exercise it in running a five-minute mile, whereas Belinda's is not – when Belinda exercises her ability to run in running a mile, she does six-minute miles. It lies within Amy's power to run a five-minute mile, whereas a six-minute mile is the best mile it is within Belinda's power to run. But in each of them it is the power to run, rather than a more narrowly specified power, that is exercised when they run a mile in however long it takes. The respective shapes of their abilities to run are different – though Amy is a better middle-distance runner than Belinda, perhaps Belinda is better at sprinting than Amy.Footnote 54

The different shapes taken by an ability in its being possessed by different agents may show themselves in a variety of ways. When the ability is properly specified by an action-type that captures what was done in a performance, the shape of the agent's ability in concreto may show itself in the proportion of successes to attempts, or in the range of circumstances that constitute opportunities in which the agent succeeds, or in the quality of the success itself (very often, of the product of the performance – whether it is a better or worse one of its kind), or – and surely this will often be the case – in some combination of these. But very often the proper specification of an ability is more coarsely-grained than the specification of what was done on an occasion of its exercise. This is clearly the case when the thing done was to walk down State Street on Monday morning, but the ability exercised was the ability to walk. Other cases are less clear. The ability to play the piano is certainly exercised in a performance of the Appassionata.Footnote 55 But is there an ability to play the Appassionata that is also exercised? Whereas someone who has the ability to walk is by that very fact in a position to walk down State Street when presented with the opportunity, it is hardly the case that someone who has the ability to play the piano is by that very fact in a position to play the Appassionata when presented with the opportunity (a working piano and the sheet music, say). A lot of practice is required. I incline to the view that the result of the practice (if it is successful) will be a development and determination of the agent's ability to play the piano, one in virtue of which she is able to play the Appassionata. (Kissin is able to play the Appassionata by exercising his ability to play the piano; my ability to play the piano is not such that I am able to play the Appassionata.) On this view, the specific shape of an agent's ability in concreto to play the piano or to run may be specified by saying which pieces are within her ability or repertoire, or how far she can run in what time. But such characterizations, of which there will be many for any given ability, are characterizations of the shape taken by a single ability, not the characterization of several different abilities.Footnote 56

I grant that there are other possible views in the vicinity; I hope to have opened up space for discussion of them. But I think that we may at least conclude that a specification of what it is within an agent's abilities to do is not thereby a specification of one of her abilities; or, as Locke puts it, ‘it seems possible to specify performances in more detail that is appropriate for abilities’.Footnote 57

This is a way of insisting on the generality of ability, which, though frequently noted, has been underestimated. Kenny and others correctly observe that abilities can be realized in distinct token actions of the type specified by the ability. But I have argued that abilities are more general than this: they can be realized in distinct token actions of distinct (though related) types. Therefore, the specification of an ability cannot be arrived at simply by reading it off the act-type exemplified by an concrete performance.Footnote 58 But though I have argued that abilities are in one respect more general (less specific) than is usually thought, I have also argued that they are in another respect less general (more particular, not more specific): in acquiring and possessing an ability, an agent particularizes it, thereby giving it a determinate shape.

Indeed, it is quite fundamental to abilities that they are exercised in performances more highly specified than the abilities themselves. It would otherwise be impossible to hold both that we can acquire new and novel abilities and that abilities play a genuine role in the explanation of their exercises. We are in a position to do novel things because we can do them by deploying abilities we already have in new ways, conditions, and combinations. We thus provide for an infinity of possible actions without requiring the postulation of an infinity of abilities. And surely this is required if we are to articulate a conception of ability on which abilities explain the performances in which they are exercised.Footnote 59 Fine-grained specifications of abilities make it hard to credit them with a genuine explanatory role.Footnote 60

There is some reason, then, to prefer coarsely-grained specifications of abilities. But this is not to say that more coarsely-grained specifications of abilities are always preferable. Sometimes a complex performance (for instance, talking while walking) does not require a correspondingly complex ability (the ability to talk while walking) over and above the constituent abilities (the ability to talk and the ability to walk), but sometimes it does: a child may have the ability to pat her head and the ability to rub her tummy without being in a position to pat her head while rubbing her tummy – here it does seem that she needs to acquire a further ability, the ability to pat her head while rubbing her tummy.Footnote 61

5. Ability and success, again

Opting for more coarsely-grained specifications of abilities makes it easier to see how abilities play an explanatory role: by exercising our abilities in different ways, situations, and combinations we can do more things than we have abilities. We do not postulate an ability for every thing done, and thus avoid the suspicion that it is the deed that ‘explains’ the ability, rather than the other way around. But to say only this would be to understate how fundamental abilities are to agency.

As I noted in §3, the fallibility of practical abilities is widely acknowledged. However, many hold not only that ability (in conjunction with opportunity and attempt) does not guarantee success, but also that successfully φ-ing does not suffice to establish the ability to φ – or in other words, that the ability is not necessary for successful action. Indeed, though there is some dissent – Austin says that ‘of course it follows merely from the premiss that he does it, that he has the ability to do it, according to ordinary English’Footnote 62 – the received view seems to be that one may successfully φ despite lacking the ability to φ.Footnote 63 Kenny gives an example that occurs frequently in the literature: ‘A hopeless darts player may, once in a lifetime, hit the bull, but be unable to repeat the performance because he does not have the ability to hit the bull’.Footnote 64 Even though the hopeless darts player hit the bull's eye, he cannot reliably hit it. Hitting it is not within his control: that he did hit it on this occasion was down to luck, not ability, as it will be if he ever hits it again. As Kenny notes, ‘Counterexamples similar to these will always be imaginable whenever it is possible to do something by luck rather than by skill’.Footnote 65

If it is possible to φ despite lacking the ability to φ, it is easy to see why many philosophers of action should think that the ability to φ plays no essential role in the explanation of an agent's φ-ing on some occasion. The example of the inept dart thrower who flukily hits the bull relies on a conception of what it is to act, or to act successfully, that is restricted to the bringing about of certain states of affairs. Evidently this conception of doing something or succeeding is independent of the idea of ability: a successful deed may be the exercise of an ability, but it may not. What is essential, on this conception, is that, but not how, the relevant state of affairs is brought about.Footnote 66 Vihvelin is quite explicit about this:

There are many ways to do something. One way is by having the ability. Another way is by accident or lucky fluke. Yet another way is by having one's brain and body moved, puppetlike, in the appropriate ways by a sorcerer. Doing something by accident or lucky fluke does not entail having the ability to do it; doing something due to direct manipulation by someone else does not entail having the ability to do it. This shows that the fact that S does A does not entail that S has the ability to do A.Footnote 67

Most philosophers of action would, I hope, deny that ‘having one's brain and body moved, puppetlike, in the appropriate ways by a sorcerer’ is a way of doing something: it is, rather, a way of suffering something – of having something done to one.Footnote 68 But there is certainly a conception of action on which one's accidental or flukey bringings-about are actions. Consider Davidson'sFootnote 69 famous example of (unintentionally) alerting the prowler by turning on the light. Davidson illuminated the room, which caused the prowler to be alerted; his deed could then be described in terms of what he brought about, as his alerting of the prowler. Nevertheless, I think that reflection shows that this conception of action (as the bringing about of results) is secondary or derivative: it depends on a conception of action as essentially the exercise of an ability. For the latter conception of action, but not for the former, Austin was correct: action suffices for ability.

The (not unconnected) influences of utilitarianism and a results-oriented mode of production may make the conception of action as essentially the exercise of an ability difficult to see. It may help to note that certain types of failures, as well as successes, may be manifestations of ability – as when an excellent goalkeeper gets a hand on an unstoppable shot that a lesser keeper would not have got near. Even though he failed to make the save, he certainly acted, and his action was an exercise of his goalkeeping abilities. So we cannot capture our concept of ability by relying solely on the thought that it is manifested in successful instances. Some ‘successful instances’, conceived of as mere ‘bringings about’, are not manifestations of ability, whereas some failures to bring about the effect that was desired or intended or constitutive of the ability may nevertheless be manifestations of ability. Sometimes all the skill in the world cannot yield success, but that means neither there was no action, nor that skill and ability were not on display.

Though this illustrates the difference between the two conceptions of action – action as the ‘successful’ bringing about of effects or states of affairs, and action as essentially the exercise of ability – it does not yet show why we should think of the latter conception as more fundamental than the former. But it is clear that to bring it about that p, one must bring p about in some way or other. Some ability or competence must be exercised, even if it is not an ability to bring it about that p. It is true that Kenny's hopeless darts player hits the bull without having the ability to hit the bull. But he hits it by doing things he does have the ability to do: grasp and grip and aim and throw a smallish object.Footnote 70

If this is right, then the idea that abilities can be distinguished from dispositions simply by characterizing their manifestations as actions is undermined. For if action, in the fundamental sense, is essentially the exercise of ability, then we will go around in an unilluminating circle if we appeal to the idea that abilities are manifested in actions to distinguish them from dispositions.

One might hope to break out of the circle by appealing to the idea of trying in order to distinguish abilities from dispositions – and thus to opt for the conditional view of ability (see §2 above). But this would be right only if trying were a causal condition on the exercise of ability – if an act of trying were conceived of as a trigger or stimulus of a disposition.Footnote 71 Yet this idea is surely misguided, conjuring up as it does an image of my body as a machine the passive operations of which are kickstarted by what I really do, namely try to act.Footnote 72 If – as indeed Kenny holds – it is a mistake to interpret the ‘if I tried’ in ‘I would if I tried’ as identifying a causal condition, and so a mistake to interpret the conditional as a causal counterfactual (as it is supposed to be in ‘the glass would break if it were struck’), then the point of drawing attention to the role of trying or attempt is not to claim that abilities have a particular triggering stimulus, where the disposition stimulated differs in no fundamental way from (e.g.) fragility, but rather to indicate that abilities are a formally distinctive kind of power. Practical abilities are volitional powers: they are exercised at will. This idea obviously requires elucidation and defence for which I do not have space here; I hope, however, to have motivated further investigation of it.Footnote 73

Conclusion

Everyday thought and practice accords greater significance to the role of practical abilities in human agency than do most philosophers of action, whose accounts are typically oriented around reasons for acting. I share Hornsby's sense that for the philosophy of action to give the agent her due, we must recognize that when an agent acts she exercises her practical abilities. But the conception of practical abilities that emerges piecemeal from a number of contemporary philosophical projects gets in the way of taking this point seriously, and obscures the possibility of crediting practical abilities with a significant role in the explanation of action. To get this possibility in view, an alternative conception of practical abilities is needed. My goal was not to develop this alternative, but to clear some of the ground for its development. It will require giving up the assumption that practical abilities can be distinguished from mere dispositions by virtue of some distinctively practical element in the disposition's structure, and instead working out what it is for an ability to be a power of a distinctive form. It will also require giving up the assumption that practical abilities are to be specified in terms of act-types, as these are commonly understood: a proper understanding of the acquisition, development, and exercise of practical abilities requires a better grip than we currently have on their specification and individuation.Footnote 74

Footnotes

1 Clarke, R., ‘Abilities to Act’, Philosophy Compass 10(12) (2015), 893.

2 For important exceptions, see Ryle, G. The Concept of Mind (London: Routledge, 1949) especially ch. 5; Kenny, A., Will, Freedom and Power (Basil Blackwell, 1975), especially ch. VII); Kenny, A., The Metaphysics of Mind (Clarendon Press, 1989), especially ch. 5; Baier, A.C., ‘Act and Intent’, The Journal of Philosophy 67(19) (1970), 648–58; Baier, A.C., ‘Ways and Means’, Canadian Journal of Philosophy 1(3) (1972), 275–93.

3 See A. Ford, ‘The Representation of Action’, this volume.

4 Cf. Hornsby, J., ‘Agency and Actions’, Royal Institute of Philosophy Supplements 55 (2004), 19: ‘when one has an action-explanation, one knows why someone did something, and that is to know why they played a particular causal role – why a brought it about that p, say. The explanation does not tell one what causal role a played: one already knows this when one knows what a did.’

5 For the origins of the standard story, see Davidson, D., ‘Actions, Reasons, and Causes’, The Journal of Philosophy 60(23) (1963), 685–700 ; for a recent defense, Smith, M., ‘Four Objections to the Standard Story of Action (and Four Replies)’, Philosophical Issues 22(1) 2012, 387–401 ; for the sorts of variants I have in mind, see e.g. Velleman, D., ‘What Happens When Someone Acts?’ Mind 101(403) (1992), 461–81, and Bratman, M., ‘Two Problems About Human Agency’, Proceedings of the Aristotelian Society 101(1) (2001), 309–26.

6 See e.g. Velleman, ‘What Happens When Someone Acts?’ and (for a more trenchant version of the criticism) Hornsby, ‘Agency and Actions’.

7 J. Hornsby, ‘Agency and Actions’, 22.

8 Sosa, E., Judgment and Agency (Oxford: Oxford University Press, 2015).

9 Greco, J., ‘The Nature of Ability and the Purpose of Knowledge’, Philosophical Issues 17(1) (2007), 57–69 .

10 Hyman, J., ‘How Knowledge Works’, The Philosophical Quarterly 49(197) (1999), 433–51; Hyman, J., Action, Knowledge, and Will (Oxford: Oxford University Press, 2015).

11 Vihvelin, K., Causes, Laws, and Free Will: Why Determinism Doesn't Matter (Oxford: Oxford University Press, 2013).

12 Vihvelin defends a version of the view that for one to have the ability to φ is for one to be disposed to φ if one has the opportunity to φ and tries to φ.

13 This is a venerable strategy: consider the uses Plato and Aristotle make of the concept of technē.

14 E.g. Mumford, S., Dispositions (Oxford: Oxford University Press, 1998); Vetter, B., Potentiality: From Dispositions to Modality (Oxford: Oxford University Press, 2015).

15 Practical abilities are frequently mentioned but rarely examined in recent discussions about the nature of knowledge how, where the following thesis is often discussed: one knows how to φ just in case one has the ability to φ. This thesis is attributed – mistakenly, in my view – to Ryle (The Concept of Mind, ch. 2), and rejected by, among others, J. Stanley and T. Williamson (‘Knowing How’, The Journal of Philosophy 98(8) (2001), 411–44). But neither those who attack nor those who defend the so-called Rylean thesis explain what it is to have the ability to φ.

16 Philosophy of action makes much use of the concept of an act-type or ‘thing done’ as contrasted with that of a token action or ‘doing (of a thing done)’. Much of this use is uncritical (for helpful criticism see Rödl, S., ‘Practice and the Unity of Action’ in Social Facts & Collective Intentionality (ed.) Georg Meggle (Frankfurt: Ontos Verlag, 2002), 323–42; Thompson, M., Life and Action: Elementary Structures of Practice and Practical Thought (Cambridge, MA: Harvard University Press, 2008), ch. 8). The conception of an act-type the partial theory's use of which I wish to criticize is broadly Davidsonian. Davidson, D., ‘The Logical Form of Action Sentences’ in The Logic of Decision and Action (Pittsburgh: University of Pittsburgh Press, 1967) argued that action sentences – such as ‘Shem kicked Shaun in the forum at noon’ – quantify over events. That sentence says that an event occurred, and that it had the properties of being a kicking, of being of Shaun, of being by Shem, of being in the forum, and of being at noon. Any token action exemplifies many act-types: among other things, what Shem did was to kick Shaun, to kick Shaun in the forum at noon, to kick someone, to do something at noon, etc. And any act-type could be (or at least could have been) instantiated by different token actions: Shem might do the same thing every day for a week – kick Shaun in the forum at noon – in which case there would be seven tokens of that act-type.

17 For more examples, see J. Hyman, Action, Knowledge, and Will, 177–8.

18 A contrast between active and powers is drawn by Aristotle at the beginning of Metaphysics Book Theta. A solvent has the active power to dissolve a solute; a solute has the passive power to be dissolved by a solvent. (Clearly not every active power is a practical ability.)  – In saying that acts of recognition and understanding are passive I do not mean to say that they are wholly passive; but I doubt that whatever form of activity may be present in them is that of intentional action. Though I distinguish recognitional and intellectual abilities from practical abilities, I do not think they are disjoint. For instance, the ability to recognize a wind shift is internal to the practical ability to sail, and the ability to understand French is internal to the ability to speak French. Crucially, however, whereas the abilities to sail and to speak French are exercised at will, the abilities to recognize a wind shift and to understand French are not – though one can do various things at will (e.g. close one's eyes, block one's ears) to prevent their exercise.

19 Maier, J., ‘The Agentive Modalities’, Philosophy and Phenomenological Research 90(1) (2015), 123.

20 K. Vihvelin, Causes, Laws, and Free Will, 11.

21 For two-fold distinctions between ‘general ability’ and ‘specific ability’, see Mele, A. ‘Agents' Abilities’ Noûs 37(3) (2003), 447–70; Whittle, A., ‘Dispositional Abilities’, Philosophers’ Imprint 10(12) (2010); Glick, E., ‘Abilities and Know-How Attributions’, in Knowledge Ascriptions (Oxford University Press, 2012); Maier, J., ‘Abilities’ in The Stanford Encyclopedia of Philosophy, (ed.) Zalta, Edward N. (Spring 2014); J. Maier, ‘The Agentive Modalities’. For three-fold distinctions, see Vihvelin, Causes, Laws, and Free Will, 7–16; Sosa, E., Judgment and Agency (Oxford: Oxford University Press, 2015), 22–23 .

22 A. Kenny, Will, Freedom and Power, 135.

23 Similarly, with psychological impediments to the exercise of one's general abilities: there can be psychological roadblocks to exercising abilities that need to be overcome. But overcoming them is not acquiring an ability to try to φ. Like removing a literal roadblock, removing a psychological roadblock such as grief, depression, anxiety, or a phobia is removing an obstacle to exercising one's abilities. Indeed, Vihvelin's characterization of what she calls ‘narrow ability’ to φ, or ‘having what it takes’ to φ – namely, ‘the necessary skills [i.e. the ‘general ability’] and the psychological and physical capacity to use those skills’ – makes it sound as if it consists of the conjunction of two abilities: the ability to φ and the ability (or capacity) to exercise one's ability to φ. This might suggest a regress: if one needs, in addition to the ability to φ, the ability to exercise one's ability to φ, why wouldn't one need the ability to exercise one's ability to exercise one's ability to φ (and so on)?

24 We would not expect acquiring or losing the ability to ride a bicycle to affect someone's possession of the ability to recite The Love Song of J. Alfred Prufrock. But we might well expect that if someone who can ride a bike and can recite Prufrock gets blind drunk, both abilities will be impaired; and when she sobers up, she will not have gained or regained a kind of ability (a ‘narrow ability’) to ride a bike – and with it, a ‘narrow ability’ to recite Prufrock – but she will no longer be impeded from exercising many of her ‘general abilities’.

25 As I noted above, someone may be able to do something because she possesses the authority to do it. Someone else may be unable to do it, because she lacks the relevant authority. But to lack authority is not to lack ability.

26 Cf. S. Mumford, Dispositions, 10: ‘I take it that the dispositional is a genus that can accommodate these subclasses [sc. tendencies, capacities and incapacities, powers and forces, potentialities and propensities, abilities and liabilities, etc.] as species’; Vihvelin, Causes, Laws, and Free Will, 171: ‘To have an ability to act is to have a disposition or bundle of dispositions’; E. Sosa, Judgment and Agency, 24: ‘Competences are a special case of dispositions….’

27 E.g. K. Vihvelin, Causes, Laws, and Free Will.

28 This is evidently closely related to the ‘Simple Conditional Analysis’ of dispositions, which faces well-known counterexamples (‘finks’ and ‘masks’ – see e.g. Martin, C.B., ‘Dispositions and Conditionals’, The Philosophical Quarterly 44 (1994); Lewis, D., ‘Finkish Dispositions’, The Philosophical Quarterly 47 (1997); Manley, D. and Wasserman, R., ‘On Linking Dispositions and Conditionals’, Mind 117(465) (2008)) that seem to show that satisfying the proposed conditional is neither necessary nor sufficient for possession of the disposition. There is much dispute over the correct response to such counterexamples: one might give a more complicated conditional analysis, or a non-conditional analysis, or take some conditional to be, not an analysis, but rather some sort of gloss that merely indicates the sui generis character of dispositions.

29 E.g. B. Vetter, Potentiality.

30 Or a counterpart of theirs.

31 To associate disposition (or ability) ascriptions with conditionals or with restricted possibilities is not to commit oneself to a reductive analysis of dispositions (or abilities) in terms of conditionals or restricted possibilities.

32 I am not alone in registering such concerns. Millikan writes that: ‘The modern philosophical tradition has unreflectively assimilated abilities to capacities and capacities to dispositions. This affords a slippery slope.’ Millikan, R., On Clear and Confused Ideas: An Essay About Substance Concepts (Cambridge: Cambridge University Press, 2000), 31. See also Baier ‘Ways and Means’, 285–6: ‘“Competence” is too pompous a word to use of simple skilled activities such as digging, stirring, walking, but it may be preferable to “ability” which has been debased by philosophical usage until it has lost any discriminatory power. When “x has the ability to φ” is equated with “x is able to φ” and that in turn is equated with “it is possible that x will φ” or “x is not causally necessitated not to φ” then the distinctive features of abilities are hopelessly lost.’

33 A. Kenny Will, Freedom and Power, 133.

34 Austin, J.L., ‘Ifs and Cans’, Proceedings of the British Academy 42(1956), 109–132 . Cited as reprinted in his Philosophical Papers (Oxford University Press, 1961), 166 n. 1.

35 K. Vihvelin, Causes, Laws, and Free Will, 186. ‘Narrow ability’ is Vihvelin's term for the so-called kind of ability that requires ‘the necessary skills and the psychological and physical capacity to use those skills’ (Causes, Laws, and Free Will, 11) but not the opportunity to use them (see §1 above for criticism). Her account transposes Lewis's ‘Finkish Dispositions’ conditional analysis of dispositions to the case of abilities. Though most writers prefer a uniform treatment of dispositions and abilities, Lewis himself proposed a restricted possibility view of abilities ( Lewis, D., ‘The Paradoxes of Time Travel’, American Philosophical Quarterly 13(2) (1976), 145–52).

36 J. Greco, ‘The Nature of Ability and the Purpose of Knowledge’, 61.

37 A. Kenny, Will, Freedom and Power, 142.

38 Cf. A. Kenny: ‘I cannot spell “seize”; I am never sure whether it is an exception to the rule about ‘i’ before ‘e’; I just guess, and fifty times out of a hundred I get it right. On each such case…it is the case that I am spelling “seize” correctly but it is not the case that I can spell “seize” correctly’ (Will, Freedom and Power, 136).

39 In basketball, making 40% of one's three-point shots would be outstanding, whereas making 60% of one's free-throws would be abysmal: it is much easier to make a free-throw than a three-pointer.

40 Cf. J. Maier, ‘The Agentive Modalities’, 128.

41 Hacker, P. M. S., The Intellectual Powers: A Study of Human Nature (Chichester, West Sussex: Wiley-Blackwell, 2013).

42 A. Mele, ‘Agents’ Abilities’.

43 R. Millikan, On Clear and Confused.

44 A. Whittle, ‘Dispositional Abilities’.

45 J.L. Austin, ‘Ifs and Cans’.

46 K. Vihvelin, Causes, Laws, and Free Will.

47 J. Spencer ‘Able to Do the Impossible’, Mind (forthcoming).

48 J. Hyman, Action, Knowledge, and Will.

49 J. Hyman, Action, Knowledge, and Will, 177–8.

50 Footnote Ibid., 178.

51 A related point is made by Vetter, B., ‘Multi-Track Dispositions’, The Philosophical Quarterly 63(251) (2013), 346.

52 Locke, D., ‘Natural Powers and Human Abilities’, Proceedings of the Aristotelian Society 74 (1973), 184.

53 Footnote Ibid., 187.

54 To distinguish again the issue of individuating abilities from that of specifying them, we should say that if Amy's ability is properly specified as the ability to run a five-minute mile, then Hyman's principle of individuation should be rejected, whereas if it is properly specified as the ability to run, then his principle may be accepted while acknowledging that what he claims follows from it in fact does not. (This bears on the cogency of Hyman's discussion as a response to the objection – by Hacker (The Intellectual Powers, 183) to Hyman's account of knowledge – that animates it.)

55 Obviously, in playing the Appassionata the agent is playing the piano. But, in the ordinary case, what is being done is playing the Appassionata: it is as a playing of the Appassionata that the agent's activity of piano-playing is evaluated for (among other things) whether or not it was completed.

56 Different such characterizations work in different ways. When we characterize Amy's ability to read English by saying that she has the ability to read Emma, we are using Austen's novel as a yardstick of (e.g.) the sort of vocabulary and prose that Amy can read in English – regardless of whether or not she has read Emma. But to characterize her piano-playing ability by saying that she has the ability to play the Appassionata would not ordinarily be to use the Appassionata as a mere yardstick of the sorts of technical and expressive demands that Amy can meet, regardless of whether she has played the Appassionata; ordinarily it would be to characterize the particular shape of her piano-playing ability as one that, as it were, contains the Appassionata.

57 D. Locke, ‘Natural Powers and Human Abilities’, 184.

58 The preceding discussion suggests that we need to reflect further on the internal structure of and relations between act-types as well as abilities.

59 For similar points about the specific cases of our linguistic and conceptual abilities, see Davidson, D. ‘Theories of Meaning and Learnable Languages’ in Proceedings of the International Congress for Logic, Methodology, and Philosophy of Science, (ed.) Yehoshua Bar-Hillel (North-Holland, 1965), 3–17 and Evans, G., The Varieties of Reference, (ed.) John McDowell (Oxford: Oxford University Press, 1982), 100ff.

60 Jack Spencer (‘Able to Do the Impossible’) argues for the provocative claim that ‘an agent might be able to do what it is metaphysically impossible tout court to do’ (msp. 1). A crucial concept in his argument is that of a ‘factive ability’, for instance the ‘abilities’ to know or discover or remember that p. Factive abilities are cases of ‘object-dependent abilities’ – ‘abilities to perform object-dependent actions, such as kissing the Blarney Stone or seeing the Statue of Liberty’ (msp. 4–5). But taking the concept of ability seriously ought to rule out the idea of object-dependent abilities, in which case Spencer's argument is over before it has begun. Spencer claims that ‘Object-dependent actions and abilities are common and familiar’ (msp. 5). But though object-dependent actions are surely familiar, the contention that object-dependent abilities are common seems to depend on the assumption that an ability can be specified by specifying the type(s) of action of which the ability's exercises are tokens – an assumption we have seen fit to reject. Object-dependent abilities are clearly explanatorily idle. To kiss the Blarney Stone, all that is needed, for someone who has the ability to kiss things, is the opportunity – and perhaps whatever is needed to overcome any psychological roadblocks to putting one's mouth on a stone that all and sundry are constantly kissing: no special ability needs to be acquired. And the truth of p is a necessary element of the opportunity to know that p, which opportunity can be taken only by those who have the ability to think that p (and the abilities to recognize reasons, deal appropriately with evidence, and so on, which may or may not be presupposed by or contained in the ability to think that p). Indeed, the ability to think that p may very well be a pseudo-ability, too: if p is the proposition that a is F, then there is no ‘ability to think that p’ over and above the abilities to think of a and to think F of things (cf. Evans (The Varieties of Reference) on what he calls the ‘Generality Constraint’).

61 Though I am sympathetic to the spirit that seems to me to animate Locke's discussion, the actual principle he offers may have to be rejected. It seems possible that one might acquire together two separable abilities the exercises of which involve exploiting the same features of the agent's constitution. Suppose that Sam, who previously lacked any cooking abilities, learns how to cook spaghetti bolognese. In learning how to do this, he learns – let us suppose simultaneously – how to cook bolognese sauce and how to cook and dress spaghetti. It would seem that the same features from Sam's constitution and background which bring it about that he standardly succeeds in cooking spaghetti are also sufficient to bring it about that he standardly succeeds in cooking bolognese sauce, and yet – pace Locke's principle – his ability to cook spaghetti and his ability to cook bolognese sauce are not the same ability (nor are they identical with his ability to cook spaghetti bolognese). It is not clear to me that we need or should expect of principle of individuation for abilities. Perhaps Millikan is correct: ‘The idea that one might count the number of a person's abilities, or count the abilities that go into a certain activity, often is not really coherent. Like patterns, however, or like patches of ground, abilities can be clearly distinguished and designated even when they have no clear criteria of individuation’ (On Clear and Confused Ideas, 64).

62 ‘Ifs and Cans’, 175.

63 See e.g. D. Locke, ‘Natural Powers and Human Abilities’, 185; A. Kenny, Will, Freedom and Power, 136; P.M.S. Hacker, The Intellectual Powers, 187–8; K. Vihvelin, Causes, Laws, and Free Will, 198; J. Hyman, Action, Knowledge, and Will, 180; B. Vetter, Potentiality, 222.

64 A. Kenny, Will, Freedom and Power, 136.

65 Footnote Ibid., 136.

66 Kenny in fact qualifies his claim that successful action does not entail ability: ‘a single performance, however successful, is not normally enough to establish the existence of ability. (I say ‘not normally’ because a single performance may suffice if the task is sufficiently difficult or complicated to rule out lucky success. Pushing one's wife in a wheelbarrow along a tightrope stretched across Niagara Falls would be a case in point)’ (Will, Freedom and Power, 136). But, to the extent that it is easy to imagine someone pushing his wife in a wheelbarrow along a tightrope stretched across Niagara Falls, it is easy to imagine his doing so being a case of lucky success, in which case the performance would not establish the ability to push his wife in a wheelbarrow along a tightrope stretched across Niagara Falls. What it may well establish (so long as the lucky success was not due to guardian angels, etc.) is that the agent has the ability to walk a tightrope, the ability to manoeuvre a wheelbarrow, and so on. Of course, someone who successfully navigated a short, low-stakes tightrope once would not thereby show herself to have the ability to walk a tightrope (as that is normally understood – which is to say, as that is understood by reference to the standards internal to the practice of tightrope walking).  It is a notable feature of Kenny's discussion that his awareness that the concepts of ability and luck come as a package co-exists with his taking for granted the idea of ‘doing something’ or ‘successful performance’; and yet surely that is just what ought to be at issue, if ability is the sort of ‘positive explanatory factor’ (Will, Freedom and Power, 133) he thinks it is.

67 K. Vihvelin, Causes, Laws, and Free Will, 198.

68 See e.g. Alvarez, M., ‘Agency and Two-Way Powers’, Proceedings of the Aristotelian Society 113(1) (2013), 101–21.

69 Davidson, D., ‘Agency’, in Agent, Action, and Reason (University of Toronto Press, 1971).

70 Cf. Baier's claim that there is an ‘asymmetric dependence’ of ‘effectings’ – bringings about – upon exercises of competences (abilities): ‘An action description such as “getting the feet wet” is non-specific with respect to how the feet were got wet, but implies that there was a way, either a wading or a washing or a hosing or …By contrast, a [description of an action as the exercise of an ability] need not, although it may, imply an effecting. This guarantees that whenever there is action, there is … exercise of a competence, and that all effectings will depend on competences’ (‘Ways and Means’, 289–90).

71 Such a view is explicitly advanced by K. Vihvelin, Causes, Laws, and Free Will.

72 For discussion of different conceptions of trying, see Hornsby, J. ‘Trying to Act’ in A Companion to the Philosophy of Action, (eds) Timothy O'Connor and Constantine Sandis (Chichester: Wiley-Blackwell, 2010), 18–25 . If trying were a causal stimulus of the exercise of abilities, then abilities would differ from many dispositions in having uniform stimuli – Vetter advances a restricted possibility view of dispositions in part because she thinks that ‘[s]timulus conditions play no part in individuating, or in giving the essence of, a disposition’, (‘Multi-Track Dispositions’, The Philosophical Quarterly 63(251) (2013), 349).

73 The idea is discussed – unsatisfactorily, in my view – by A. Kenny (Will, Freedom and Power, ch. VII); A. Kenny (The Metaphysics of Mind, ch. 5); Steward, H., ‘Sub-Intentional Actions and the Over-Mentalization of Agency’ in New Essays on the Explanation of Action (ed.) Sandis, Constantine (Houndmills, Basingstoke, Hampshire: Palgrave Macmillan, 2009), 295–312 ; Steward, H., ‘Responses’, Inquiry 56(6) (2013), 681–706 ; M. Alvarez, ‘Agency and Two-Way Powers’.

74 Earlier versions of this paper were presented at the Centre for the Study of Mind in Nature at the University of Oslo, the Royal Institute of Philosophy, and a meeting of the DFG Netzwerk »Praktisches Denken und gutes Handeln« at the University of Leipzig. Thanks to those who participated in those events. And special thanks to John Hyman for comments on a draft, from which I – and I hope the paper – benefitted a great deal.

Ibid
Ibid
Ibid