Human freedom and agency (Aquinas).pdf

Human freedom and agency Thomas Williams (forthcoming in the Oxford Handbook to Thomas Aquinas) The end of human action Aquinas opens the second part of the ST by arguing, in a series of careful steps, that there is one and only one ultimate end for all human actions. The placement of this a...

Human freedom and agency Thomas Williams (forthcoming in the Oxford Handbook to Thomas Aquinas) The end of human action Aquinas opens the second part of the ST by arguing, in a series of careful steps, that there is one and only one ultimate end for all human actions. The placement of this argument is no accident, since the notion of an end is of fundamental importance not only in Aquinas’s theory of human action but in his accounts of practical reasoning, law, and the virtues. Yet the interpretation of Aquinas’s argument in ST 1a2ae, q.1, is a matter of considerable controversy. I shall first follow the argument through its successive steps and then briefly consider three possible ways of understanding Aquinas’s claim that all human actions have exactly one ultimate end. Aquinas first argues that every human action is for the sake of some end. That is, every human action is purposive in some way; it is done for the sake of attaining some goal or realizing some desired state of affairs. To the obvious objection that we do things all the time without any purpose at all, Aquinas replies by distinguishing between a human action (actus humanus) and an action of a human being (actio hominis). Human actions, properly so called, are those that proceed from human beings in virtue of their distinguishing power, which is to be in control of their own actions (dominus suorum actuum) through reason and will. Anything else that a human being does can be called the action of a human being, but not (in the proper sense) a human action. Human actions, then, are those that are willed on the basis of rational deliberation. And since “the object of the will is an end and a good, it follows that all human actions are for the sake of an end.”1 What, precisely, is meant by the “end” of a human action? Aquinas tells us in a. 2 that an end is something cognized as good. It must be something cognized because otherwise we just have “natural appetite,” the sort of built-in directedness by which heavy objects are moved 1 ST 1a2ae, q. 1, a. 1. All translations are my own. 1 toward the center of the earth. And an end must be cognized as good because the will is moved only by what is good. As Eleonore Stump emphasizes, the will is not a neutral steering wheel; it is “an inclination for what is good, where the phrase ‘what is good’ is used attributively and not referentially.”2 And because only beings with the capacity for abstract thought can cognize what is good qua good (as opposed to merely cognizing something that is in fact good), only they can have will, the appetite that follows upon intellectual cognition of something as good. “The object of the will,” Aquinas says, “is the end and good universally.” But the end of any particular action will not be good-in-general, but some particular thing cognized as good. This end is what gives each particular action its species (a.3), what makes the action the kind of action it is. In a.4 Aquinas argues further that every human action has an ultimate end: that is, an end that, with respect to that particular action, is not itself for the sake of some further end. It is impossible for there to be an infinite series of ends, each of which moves the appetite only instrumentally, as a means to some more ultimate end. There has to be some mover that moves the appetite on its own account. This non-instrumental mover of the appetite is the ultimate end for the sake of which a particular action is performed. And it is not merely that every action must have at least one ultimate end. Aquinas argues further that, at any given time, a human being can have only one such ultimate end. The most persuasive of his three arguments for this claim runs as follows: Since every thing desires (appetat) its own perfection, what someone desires as an ultimate end is what he desires qua perfecting and completing him.... For that reason, an ultimate end must so fulfill a person’s whole desire that nothing more is left for him to desire; and that cannot be so if something additional is required for him to be perfect. And that’s why desire cannot aim at two [distinct] things as if each were its perfect good.3 (ST 1a2ae, q. 1, a. 5) This single ultimate end can be an aggregate of goods that the agent regards as collectively 2 Stump, Eleonore (2003), Aquinas (London and New York: Routledge), 278. 3 One might well think that the argument as a whole is question-begging, since the claim that “what one desires as an ultimate end is what he desires qua perfecting and completing him” follows from the claim that “every thing desires its own perfection” only if we interpret the latter as meaning something like “every thing desires only its own perfection” – but on that interpretation, the argument already presupposes the very unity of desire that it purports to establish. For further analysis of Aquinas’s arguments in the opening questions of the 1a2ae, see Ryan, Peter F. (2001), “M ust the Acting Person Have a Single Ultimate End?”, Gregorianum 82, 325-356. 2 constituting his perfection; the arguments of a. 5 do not commit Aquinas to understanding the ultimate end as a unitary good. This ultimate end, whether unitary or aggregate, must be (Aquinas argues) the ultimate explanation for all of a given person’s actions. For whatever the agent desires, she desires sub ratione boni: under the notion or aspect of the good. Now if something is not desired as the complete, all-inclusive good (that is, the ultimate good), it must be desired as aiming at or leading to the complete, all-inclusive good. The ‘as’ does not mean that an agent will always be thinking of the ultimate end in every action she performs, any more than someone on a journey must be consciously thinking about her destination in order to be truly said to be aiming at that destination. But the desire for the ultimate end is what explains every other desire. The ultimate end is the first mover of desire, and secondary movers – that is, any desirable objects other than the ultimate end – move desire only in virtue of their relation to the ultimate end, whether the agent is thinking explicitly about that relation or not. As Aquinas puts it, “the power of the first intention, which is of the ultimate end, remains in any desire of any object whatever, even if one is not actually thinking of the ultimate end” (ST 1a2ae, q. 1, a. 6, ad 3). Finally, Aquinas argues that this ultimate explanation for all actions is the same for every human being. All human beings have the same ultimate end, for all human beings desire their own perfection, though different people will have different ideas about what perfection consists in. So Aquinas says that all human beings have the same ultimate end as far as the intelligible formula for the ultimate end goes (secundum rationem ultimi finis), but not in terms of the object that people think is aptly described by that formula (secundum id in quo finis ultimi ratio invenitur). To summarize: Aquinas argues in the first question of the 1a2ae that (a) all human actions are performed for the sake of an end; (b) any given human action has one and only one ultimate end; (c) this one ultimate end is the same for all of a given human being’s actions; and indeed (d) all human beings have the same ultimate end. Interpreters of Aquinas have disagreed about how to construe each of these claims, but the most widely divergent readings concern (b) and (c). The most obvious way of taking (b) and (c) is as descriptive claims about the psychology of human actions. But when taken as descriptive claims, (b) and (c) strike many interpreters as flatly 3 indefensible. We may take Jean Porter’s objection as representative: Not many of us have the self-possession, or the fanaticism, to shape our whole lives around devotion to some one object, cause, or ideal. And even those few of us who do cannot be said literally to direct all our actions toward one end. Even the most dedicated individuals indulge in an ice-cream cone or a joke once in a while.... Let me say at once that if Aquinas does indeed claim that each person always directs all her actions and activities toward some one goal or idea, then I do not see how that claim could be defended.4 In light of such reservations about (b) and (c) taken as descriptive claims, some interpreters argue that what Aquinas sets forth in the opening question of the 1a2ae is normative rather than descriptive: his claims “state criteria of fully rational action, and the arguments for them draw attention to what is required by the concept of rational action,”5 as Scott MacDonald puts it. On this interpretation Aquinas is not denying that human beings sometimes act without having a single ultimate end in view, but instead saying that such actions are not fully rational and therefore not fully human. As Jean Porter, another defender of the normative reading, says, this analysis “leaves room to admit that some behaviors which do not meet this ideal can still be said to be rational, and hence truly human actions, albeit in a derivative sense.”6 A third interpretation is possible, according to which Aquinas’s claims in the opening question of the 1a2ae are neither descriptive nor normative, but explanatory. That is, Aquinas is analyzing the metaphysical preconditions for human action. His arguments for a single ultimate end are akin to his arguments for a single unmoved mover or uncaused cause; they are at bottom neither psychological descriptions nor conceptual analysis, but quia arguments intended to trace the phenomena of human action to their ultimate explanation in a good that awakens desire, and 4 Porter, Jean (1990), The Recovery of Virtue: The Relevance of Aquinas for Christian Ethics (Louisville: W estminster/John Knox Press), 72-73. 5 MacDonald, Scott (1991), “Ultimate Ends in Practical Reasoning: Aquinas’s Aristotelian Moral Psychology and Anscombe’s Fallacy”, The Philosophical Review 100, 31-66, at 40. Ryan, Peter F. (2001), “A Single Ultimate End Only for ‘Fully Rational’ Agents? A Critique of Scott MacDonald’s Interpretation of Aquinas,” American Catholic Philosophical Quarterly 75, 433-438, argues that Aquinas is neither making empirical claims nor analyzing the concept of fully rational action, but “trying to articulate a necessary condition for human action as such, a condition that he thinks follows from just what a human act is” (438). Ryan emphasizes, however, that Aquinas’s view is empirically falsified if there are any cases of human action that do not meet the necessary condition, and he argues in “Must the Acting Person Have a Single Ultimate End?” that there are such cases. 6 Porter 1990 (73). 4 thereby initiates action, in its own right and not in the power of any other good. The natural desire for perfection explains why we desire anything at all. Since every human being has this desire, every human action is ultimately to be explained with reference to one and the same ultimate end – though, owing the variabilities of human reason, different people will seek their fulfillment in different objects or states. The structure of human action A human action, as we have seen, is one over which the agent has control in virtue of possessing reason and will. Since the object of will is an end – something cognized as good – every human action will begin with cognition: an intellectual apprehension of some object or state of affairs as good in the agent’s circumstances. (The cognition of the circumstances is also the work of the intellect, which means that even before there is any act of will, there are two ways in which human agents can go astray: by taking something to be good that is not in fact good, and by misconstruing their circumstances.) As a consequence of this intellectual apprehension, there is an act of will called simply voluntas (will) in which the agent desires the end as such.7 In order for action to result, however, there must be a further act of the will in which one desires the end not merely as a good in itself but as a good that can be attained by means of action. This act, called intentio, presupposes an intellectual apprehension of the end as attainable. It also sets in motion the work of the intellect in consilium or deliberation. Consilium is an investigation of the means by which the intended end may be attained or realized. As the intellect turns up various means, the will can endorse them as desirable, or at least tolerable, means to the desired end; this endorsement is called consensus or consent. Consilium terminates with an act of iudicium, the intellect’s judgment that one of the means that have received the will’s consent is the best, all things considered. The will then selects the means that the intellect has judged best; this act of will is called electio, commonly (if somewhat misleadingly) translated “choice.”8 After electio, the intellect orders that the elected means be carried out in action, and 7 Aquinas denies that voluntas is in general a necessary consequence of the intellectual apprehension of something as an end. I take up the issue of necessity and contingency in human action in the last section. 8 For an incisive explanation of the problems with translating electio as “choice,” see Stump (2003), 288- 289. 5 the will carries out the intellect’s command (imperium) by making use (usus) of the external bodily powers in the way that the intellect has prescribed, until the intellect apprehends that the end has been attained; only then is the will is at rest in the enjoyment (fruitio) of that end.9 The previous paragraph lays out the basic structure of a fully articulated and complete human action.10 In practice, as Aquinas recognizes, human actions may not involve all of these steps. For example, if consilium turns up only one acceptable means, consensus will collapse into electio.11 If the means to the desired end is perfectly obvious, there will be no need for consilium to begin with, and the agent can proceed directly from intentio to iudicium.12 And the intellect’s failure to stick to its command – a manifestation of what Aquinas calls inconstancy – can interrupt usus before it comes to completion in the external act by which the agent means to accomplish her end, thereby forestalling fruitio.13 Moreover, the individual acts of will and intellect that go into making up a complete human action need not be separated by any perceptible time. Nor should we think that all these acts will be apparent to the agent’s consciousness: as Alan Donagan observes, “the components of simple human acts are ascertained, not by introspecting what happens when we perform them, but by examining cases in which an act is begun but not completed.”14 Intellectualism and voluntarism Note that in Aquinas’s account of the structure of human acts, every act of will is 9 This is what Aquinas calls “perfect fruitio,” as opposed to the “imperfect fruitio” that one can have regarding an end that is “possessed only in intention.” See ST 1a2ae q. 11, a. 4. 10 Or rather, it sets out one possible, and fairly mainstream, reading of the Treatise on Human Acts. For a sampling of the variety of interpretations in recent scholarship, see Gilby, Thomas, trans. and ed. (1970). Summa theologia vol. 17 (Cambridge: Blackfriars), Appendix 1; Donagan, Alan (1982). “Thomas Aquinas on Human Action,” in The Cambridge History of Later Medieval Philosophy, ed. Norman Kretzmann, Anthony Kenny, and Jan Pinborg (Cambridge: Cambridge University Press), 642-654; McInerny, Ralph (1992). Aquinas on Human Action: A Theory of Practice (W ashington, DC: The Catholic University of America Press), chapter 3; W estberg, Daniel (1994). Right Practical Reason: Aristotle, Action, and Prudence in Aquinas (Oxford: Clarendon Press), Part III; Brock, Stephen L. (1998). Action and Conduct: Thomas Aquinas and the Theory of Action (Edinburgh: T & T Clark); Stump (2003), 287-294. 11 ST 1a2ae q. 15, a. 3, ad 3. 12 ST 1a2ae q. 14, a. 4, especially ad 1. 13 ST 2a2ae q. 156, a. 1. 14 Donagan (1982), 654. 6 preceded by an act of intellect. This is exactly what we would expect, given Aquinas’s understanding of will as intellectual appetite. To recur to Eleonore Stump’s point, the will is not, on Aquinas’s view, a neutral steering wheel; it is “an inclination for what is good, where the phrase ‘what is good’ is used attributively and not referentially.”15 Hence, the content of any act of will – what is willed – has to depend on some intellectual apprehension of something as good in some way and under some description. Aquinas expresses this claim by saying that the intellect is the formal cause of the will’s acts. So stated, Aquinas’s view appears straightforwardly intellectualist: the ultimate explanation for every human act rests in the intellect, for the will has no independent capacity to shape the exercise of human agency. Yet Aquinas also speaks of ways in which the will controls the intellect: by directing it to consider alternatives, to stop deliberating, to attend to different features of a possible object of will. In doing these things, the will moves the intellect as an efficient cause. Many interpreters of Aquinas have appealed to the will’s efficient-causal control over the intellect as evidence that Aquinas is not a thoroughgoing voluntarist after all: the ultimate explanation for at least some human acts, they argue, will rest in the will’s direction of the intellect in shaping the exercise of human agency.16 The fundamental dispute between intellectualist and voluntarist interpretations of Aquinas’s action theory is whether the will’s efficient-causal control over the will must always be preceded by an act of intellect in which the agent judges that exercising such efficient-causal control is all-things-considered best in the circumstances. For example, Aquinas notes that the will can direct the intellect to consider a proposed course of action under different descriptions. I may be leaning toward spending the day at the beach because it would be pleasant and relaxing, but my will can command my intellect to think of it instead as a shameful retreat from pressing obligations. Must such an act of will itself be preceded by an intellectual judgment that the good 15 Stump, Eleonore (2003), Aquinas (London and New York: Routledge), 278. 16 For interpretations that find some degree of voluntarism in Aquinas, see, for example, Gallagher, David (1994), “Free Choice and Free Judgment in Thomas Aquinas,” Archiv für Geschichte der Philosophie 76, 247-277; Stump, Eleonore, and Norman Kretzmann (1985), “Absolute Simplicity,” Faith and Philosophy 2, 353-82; Finnis, John (1991), “Object and Intention in Moral Judgments according to Aquinas,” The Thomist 55, 1-27; Kent, Bonnie (1995), Virtues of the Will: The Transformation of Ethics in the Late Thirteenth Century (Washington, DC: The Catholic University of America Press). 7 thing to do, all things considered, is to attend to the ways in which a day at the beach is not good? Or can the will of its own power direct the intellect to reconsider? Space does not permit a careful examination of all the relevant texts, so I will simply express my own judgment that Aquinas is an intellectualist. As Jeff Hause explains, On Aquinas’s view, the will cannot, by any innate capacity, direct the intellect’s attention, keep the intellect from issuing judgments about what one ought to do, or keep itself from willing what the intellect has determined one ought to will. Nor can it select one from among a variety of alternatives unless the intellect has first settled on that one as the alternative to be pursued. Which, if any, of a set of objects the will wills, and whether it wills anything at all, depends not on any voluntaristic capacity of the will, but on how the intellect judges the object in question.17 In the process that leads to action, the intellect is not in the business of making judgments of the form “Here’s an ice-cream cone: take it or leave it – or look around for something else” and then leaving it up to the will to make the choice. The intellect offers a verdict: a iudicium or sententia. And the will, being nothing more than the appetitive arm of the intellectual soul, elects in accordance with the intellect’s judgment. As in a one-party totalitarian state, by the time there’s an election, there’s only one name on the ballot. Necessitation, causation, and freedom One reason that many scholars have resisted interpreting Aquinas’s account as intellectualist is that they believe an intellectualist account must be deterministic.18 And since (the argument continues) Aquinas holds that causal determinism is incompatible with freedom and moral responsibility, he must reject intellectualism. This argument fails for at least two reasons. First, an intellectualist account need not be deterministic; if the intellect itself operates indeterministically, Aquinas’s theory as a whole will be indeterministic even if it is also intellectualist.19 Second, and more important, it is not at all clear that Aquinas affirms the 17 Hause, Jeffrey (1997), “Thomas Aquinas and the Voluntarists,” Medieval Philosophy and Theology 6, 167-182, at 168. See also M acDonald, Scott (1998), “Aquinas’s Libertarian Account of Free Choice,” Revue Internationale de Philosophie 2, 309-328. 18 See Gallagher (1994) for an argument of this kind, and Hause (1997) for a rebuttal. 19 MacDonald (1988) defends this sort of reading of Aquinas. 8 incompatibility of freedom and causal determinism. There is, to be sure, a trivial sense in which it is obvious that Aquinas never affirms the incompatibility of freedom and determinism: for determinism in its contemporary acceptation is “the thesis that only one continuation of the state of things at a given moment is consistent with the laws of nature,”20 and Aquinas does not speak in terms of laws of nature. Aquinas speaks instead of the active and passive causal powers of created things. This point is certainly worth making, because interpreters have been much too quick to speak of determinism as though Aquinas’s metaphysics of causation and necessitation can be mapped unproblematically on our own. But the claim that Aquinas never affirms the incompatibility of freedom and determinism is not terribly interesting if we mean it only in this trivial sense, so from here on out I shall use ‘determinism’ as the name of an analogous thesis: that only one continuation of the state of things at a given moment is consistent with the ordinary, non-miraculous working of the active and passive causal powers of created things. Equipped with a more suitable understanding of determinism, we can now ask what Aquinas has to say about the compatibility or incompatibility of freedom and determinism. And the answer, surprisingly, is nothing. What Aquinas address instead is the compatibility or incompatibility of freedom and necessitation. Aquinas identifies several distinct senses in which an act might be described as necessary. Not all of them are inconsistent with freedom, and none of them (or so I shall argue) is equivalent or even roughly analogous to determinism. Consider first Aquinas’s arguments in DM 6, where the question is whether human beings have free choice (libera electio) of their acts or instead choose from necessity. Aquinas says that human beings are never necessitated with respect to exercise (that is, with respect to whether we will or do not will), and that we are necessitated with respect to specification (that is, with respect to willing this or that particular object) only in the case of happiness itself. But everything Aquinas says in reaching these conclusions is perfectly compatible with determinism. For example, he argues that the will is not necessitated with respect to specification by anything other than happiness on the grounds that one can think of any such thing either under a description under which it is suitable or under a description under which it is not. This can come about, he says, (1) because 20 van Inwagen, Peter (2002), Metaphysics, 2 nd ed. (Boulder: W estview Press), 203. 9 one description outweighs the other, (2) because some interior or exterior prompting causes one to attend to the object under one description rather than under the other, or (3) because one is in a state of mind that makes the object salient under one description rather than another. If the will’s not being necessitated with respect to specification is explained entirely in terms of these three possibilities, as the text strongly suggests, then non-necessitation is perfectly compatible with determinism. What Aquinas denies is that the intellect, the will, and their objects are such that in the abstract only one choice is possible; he does not deny, and indeed seems to lack the metaphysical resources necessary to deny, that the nature of the intellect and the will and their objects are such that in every particular, concrete instance only one choice is possible.21 Nor does Aquinas’s denial that the will can be subject to necessity of coercion have any tendency to show that Aquinas rejects the compatibility of freedom and determinism, despite the fact that many interpreters appeal to that denial to establish precisely that conclusion. Aquinas argues in QDV 22, a. 5 and ST 1a q. 82, a. 1 that the will cannot be subject to necessity of coercion (necessitas coactionis), the kind of necessity that would obtain if the will were compelled by some external agent in such a way that it could not act otherwise. But (contrary to what some interpreters suggest) Aquinas does not argue in these passages that necessity of coercion is incompatible with freedom. Instead, he argues that it is conceptually impossible for the will to be subject to necessity of coercion, because something that is coerced is moved in a way that is entirely against its natural inclination; since the will is itself an inclination, it is impossible for it to will something unless there is in the will some inclination to that very thing. Moreover, it is quite clear that necessity of coercion is not coextensive with determinism. Although every instance of necessity of coercion is a case in which determinism holds true, the converse does not hold: there can be cases of determinism in which there is no necessity of coercion. Necessity of coercion obtains only when a thing is moved in a way that is contrary to its natural inclination. If I open my hands to allow the baseball I’m holding to fall toward the 21 Note that in DM 6 ad 10 Aquinas says that the will’s not being necessitated with respect to specification by any incomplete good is just like the intellect’s not being necessitated with respect to specification by any contingent truth. No one takes it that Aquinas is denying determinism with respect to any particular intellectual act in which someone affirms a contingent truth. W hy then should any take it that Aquinas is denying determinism with respect to any particular volitional act in which someone elects a partial good? 10 earth, the ball falls in accordance with its natural inclination, not by necessity of coercion; yet clearly its downward path is also in accordance with determinism. So Aquinas’s denial that the will can be subject to necessity of coercion is not equivalent to, and does not entail, the claim that freedom is incompatible with determinism. Now it might be objected here that I have stolen a base against libertarian readings of Aquinas (that is, readings according to which Aquinas affirms both that human beings sometimes act freely and that freedom is incompatible with determinism). By defining determinism in the way I have done, I have, it appears, committed libertarians to accepting the Principle of Alternative Possibilities (PAP): S does A freely only at t only if it is possible that S not do A at t. Some libertarians deny PAP, and one might argue that Aquinas is just that sort of libertarian.22 But a PAP-denying libertarian who might wish to enlist Aquinas as an ally would first have to show that Aquinas denies the compatibility of freedom and determinism, or something sufficiently like determinism, in order to count as a libertarian at all. And since (as I have argued) the usual arguments for Aquinas’s incompatibilism fail, there is no good reason to think that Aquinas is a libertarian of any kind. Moreover, the PAP-denying libertarian will typically say that on Aquinas’s view, it is sufficient for an act’s being free that the act proceeds from the agent’s own intellect and will, whether the agent has alternative possibilities or not. If that is the case, however, it is perfectly possible for there to be free acts of will even if, given the nature of the will, the intellect, and the external world, no intellectual agent ever has alternative possibilities. I cannot myself see how such a view can be accounted a version of libertarianism, but I will not quibble over words. The passages I have considered, and others like them, do not of course show that Aquinas affirms that freedom and determinism are compatible, just that he fails to give us any reason to think they are incompatible. In the present state of scholarship it is difficult to say anything more definitive on this subject; Aquinas’s modal theory and his theory of causation remain woefully understudied. 22 Stump (1997) 591-594 and (2003) 299-306 argues for this understanding of Aquinas’s views. 11