Introduction
Aristotle holds that both in nature and in the domain of human agency goals are explanatory of things that are for their sake. For example, he thinks that your goal of being healthy explains why you are taking a walk if you are taking a walk in order to be healthy. Likewise, he thinks that the goal of fully realized feline form (for instance) explains why kittens develop in the ways they generally do. My interest in the present paper is specifically in goals in the domain of human agency. I want to argue that according to Aristotle there are two importantly different kinds of goals of human actions and that one of these two kinds has been frequently, though not universally, overlooked.Footnote 1 Apart from psychological goals of actions, goals that agents adopt as their purposes, there are also, I submit, goals or ends that actions have by being the kinds of actions they are (and, in some cases, by occurring in the circumstances in which they do). Those latter goals belong to their actions whether or not agents adopt them as purposes and whether or not agents are aware of them.
This result should be of interest from at least two general points of view. First, it may enable us to see Aristotle's teleology as more unified than it would otherwise appear. If in the domain of human agency the goals of actions are always purposes that agents adopt, then the goals of actions and the goals of natural processes may seem to be of two fundamentally different kinds. After all, the goals of natural processes do not depend for their status as goals on being adopted as purposes by anyone. However, if the main claim of the present paper is correct, and it is not the case that Aristotle takes all goals of human actions to be purposes adopted by agents, then we may be in a better position to capture the unity of Aristotle's concept of the final cause. Second, in reconstructing Aristotle's position on the natural goals of actions, I articulate on his behalf a conception of how reasons for action are present in the world according to which the presence of reasons for action does not depend on suitable beliefs or desires the agent in question has. This conception may well be of interest to philosophers working on the nature and status of reasons.
I will proceed as follows. In section 1, I discuss three passages in Aristotle's ethical treatises that employ the notion of being ignorant of a goal of one's action. I propose to make sense of them by appealing to goals that certain actions are directed to by being the kinds of actions they are. (I call such goals ‘natural goals of actions’.) In sections 2 and 3, I discuss Aristotle's account of chance in Physics II.4–6, arguing that it provides strong evidence that Aristotle's teleology of human agency operates with a distinction between natural goals of actions on the one hand and psychological goals of actions on the other.
1. Being Ignorant of What One's Action is for
Aristotle seems to think that agents can be ignorant of goals of their actions. He holds that if an action is done because the agent is ignorant of certain kinds of particular facts, then the action, or (perhaps more precisely) the action conceived of in the relevant way, is not voluntary. For example, since Oedipus was ignorant of the fact that the man he killed on the road was his father, his killing of his father was not a voluntary action of his, though he did voluntarily kill a man.
When Aristotle enumerates the kinds of particular facts about an action, such that ignorance of them makes actions count as not voluntary, he includes goals of actions:
Presumably, then, it is not a bad idea to delimit these particulars, and say what they are, and how many. They are: who is doing it; what he is doing; about what or to what he is doing it; sometimes also what he is doing it with—with what instrument, for example; for the sake of what—for example, safety; in what way, for example, gently or hard. (Nicomachean Ethics III.1, 1111a3–6)Footnote 2
My concern is with ignorance of what it is that one's action is ‘for the sake of’. Some scholars think that what Aristotle has in mind here is not ignorance of a genuine goal of one's action, but rather ignorance of the result of one's action.Footnote 3 Goals of actions, these scholars think, are purposes adopted by the agents in question, and Aristotle does not seem to countenance purposes of which the agents themselves are ignorant. Thus, it is sometimes held that in this passage and some others to which we will shortly turn, Aristotle uses the expression hou heneka and related expressions in a nonteleological way, purely to denote a result (Ross Reference Ross1936: 517–18; Lennox Reference Lennox2001: 250–58; Judson Reference Judson and Judson1991: 73–99). On this way of thinking, that which is, in those passages, presented as the hou heneka of something is not meant to operate as a final cause or as a teleological explanatory factor.
It is worth stressing that this is a highly unusual way of translating these Greek expressions. In Aristotelian usage, if something is said to be (or to come to be) heneka something else, that second item, standardly, is thought of as that for the sake of which the first item is or comes to be; in other words, the second item is thought of as the goal for the sake of which the first item is or comes to be. For what it is worth, a use of heneka + genitive in which the item denoted by the genitive expression is meant to be only a result without being a goal is not recognized in Bonitz's Index.
The passage from NE III.1 in which Aristotle envisages agents who are ignorant of what their actions are for is not an isolated anomaly. He employs the same idea in two other passages in the two ethical treatises. In NE V.8, he says that
actions done with ignorance are errors if someone does neither the action he supposed, nor to the person, nor with the instrument, nor for the sake of what he supposed. For he thought, for instance, that he was not hitting, or not with this, or not hitting this person, or not for the sake of that; but as it happened that for the sake of which was not what he thought (for example, he acted not in order to wound, but in order to spur on), or the victim or the instrument was not the one he thought. (NE V.8, 1135b12–16)
‘The cases where ignorance is a factor are mistakes’, Rowe translates the first sentence of that passage, ‘i.e. when the person affected, or the action itself, or the instrument, or the effect is not what the agent supposed, because he did not think he was hitting the other person, or not with this, or not this person, or not with this effect’ (Broadie and Rowe Reference Broadie and Rowe2002: 170).
A third passage from Aristotle's ethical writings in which he invokes ignorance of a goal of one's action is in EE II.9. This is the passage from the EE in which he specifies the forms of ignorance that he thinks make an action count as done involuntarily. The passage is useful for our purposes not only because it features the notion of being ignorant of a goal of one's action. It also offers an example that might help us get clear on the general notion.
The voluntary seems to be the opposite of the involuntary, and acting knowing either whom or with what or for the sake of what (thus sometimes a man knows that it is his father, but he acts not in order to kill, but in order to save, as in the case of the daughters of Pelias; or he knows with what he acts—that this is a drink—but treats it as a love-potion, and wine, when it was hemlock) is opposed to acting in ignorance of whom and with what and what, because of ignorance, not incidentally; but what is done because of ignorance of what and with what and whom, is involuntary; so its opposite is voluntary. (EE II.9, 1225b1–8)
In this passage Aristotle seems to offer the case of the daughters of Pelias as an example of ignorance of a goal of one's action. Medea persuaded the daughters of Pelias that they could rejuvenate their aging father by chopping him up and then boiling the pieces. They followed Medea's instructions and, in doing so, killed their father.Footnote 4
This example enables us to make sense of the notion of an action having a goal of which the agent is ignorant. Some actions, just by being the kinds of actions that they are, are suited as means to ends that it may on occasion be appropriate to pursue. Dismembering a human being is such an action, as is decapitation or (to offer a less gruesome example) cardiopulmonary resuscitation (CPR). These are actions that in themselves promote certain outcomes, outcomes that at least on occasion may be goals that it is appropriate to pursue. The goal to which dismembering people is suited is killing them. The goal to which CPR is suited is keeping alive a person whose breathing or heartbeat has stopped. These are actions, then, that by their nature are suited to certain goals. Goals to which actions are by their nature suited I will call their natural goals. To say that decapitation is for killing—in other words, that a natural goal of decapitating people is to kill them—is not to say anything about what it is that a given agent who engages in that action is trying to achieve. It is, in other words, to say nothing about the psychological goal of the agent. As a result, the agent's psychological goal in acting and a natural goal of the action they perform can come apart, as they do in the case of the daughters of Pelias. The fact that Aristotle's example lends itself so remarkably well to making this point suggests that he chose the example precisely to make this point.
With this understanding of how an action can have a goal of which the agent, at the time of action, is ignorant, we may avoid a nonteleological interpretation of the language of goals in the passages about ignorance of particulars that we have looked at. We can make sense of the passages from NE III.1 and V.8 by relying on the distinction between a natural goal of an action and a psychological goal, or purpose, of the agent.
You might know that a given potion is hemlock, and hence poisonous, but for some reason think that, in some situation, drinking it will save a person rather than kill her (NE III.1, 1111a13–14), much as the daughters of Pelias thought that they could rejuvenate their father by chopping him up and boiling the bits. After all, if you did not know that the potion is poisonous, this would be a case of ignorance of an object used in doing the action (1111a5), rather than a case of ignorance of a goal of the action. Or again, you might know that the whip with which you are lashing a horse is a multitailed leather whip equipped with lead tips, but think that using it to whip a horse is an appropriate and effective way of spurring it on. What you do not know, then, is that applying this kind of whip is for wounding, not spurring (NE V.8, 1135b15–16).
2. ‘Things for the Sake of Something’ in Aristotle's Account of Chance
As we have seen, several scholars hold that in the passages from Aristotle's ethical treatises we considered in the previous section, he uses hou heneka language in a nonteleological way, purely to denote results of actions, as opposed to goals of actions. This interpretation of the ethical texts is used to support interpretations of Aristotle's discussion of chance in Physics II.4–6 according to which in that discussion, too, Aristotle operates with a nonteleological notion of a hou heneka (Lennox Reference Lennox2001: 257; Judson Reference Judson and Judson1991: 78). I hope that in the previous section, I managed at least to throw in doubt the idea that in the ethical treatises Aristotle operates with such a notion. I now turn to the discussion of chance in the Physics.
When properly understood, that discussion provides no evidence for a nonteleological notion of a hou heneka and instead offers strong evidence for an Aristotelian distinction between natural goals of actions and their psychological goals. In section 3, I will present and discuss the evidence for this distinction between two kinds of goals. However, to be in a position to appreciate that evidence, we must first address a major interpretive difficulty about Aristotle's account. Here too, the distinction between natural and psychological goals of actions will be important for making sense of Aristotle's view.
In Physics II.4–6 Aristotle offers an analysis of chance and luck. Luck is the more specific notion. Roughly speaking, lucky events are desirable chance events that happen to people. Aristotle begins his analysis of chance by locating things that are by chance in the domain of things that do not come to be in the same way always or for the most part (196b10–17). He then introduces the notion of things that are for the sake of something and offers the following initial explication of what it is for something to occur by chance: ‘Whatever would be done from thought or from nature are things for the sake of something. Whenever such things come to be incidentally, we say that they are by chance’ (Physics II.5, 196b21–24).The things that are by chance, Aristotle thinks, satisfy two conditions: they are things for the sake of something, and they come to be incidentally. It has proven difficult to work out precisely what kinds of things are supposed to satisfy those conditions.Footnote 5 Just what kinds of things are Aristotelian chance events?
One might think that the things that are by chance, so far as cases of human agency are concerned, are desirable results (or undesirable ones, to make room for bad luck). To use Aristotle's own example (Physics II.5, 196b33–34; cf. II.4, 196a3–5), such a desirable result might be collecting a debt after you ran into your debtor in the marketplace when he had just received enough money to return what he owed you. So it was by chance, and specifically by luck, that you collected what you were owed.
This answer to our question, that the things that are by chance are desirable results (or, in cases of bad luck, undesirable ones), may seem plausible at first sight, but there is good reason to reject it. Aristotle describes the desirable result of collecting the debt as something that the creditor's action was for the sake of and as a goal of the creditor's action (196b34–36, 197a1–2). Below in section 3, I will argue for a robustly teleological reading of that language, but the thing to note for now is just that Aristotle describes that desirable result in this way, as a thing for the sake of which and as a goal. The chance event, by contrast, must be a thing that is for the sake of something. In standard Aristotelian usage, things that are for the sake of something are the explananda of teleological explanation, whereas goals or things for the sake of which are the explanantia. Things that are for the sake of something are means or steps that are or come to be for the sake of the goals in question (cf. 197b26–7). They are not themselves those goals.Footnote 6
So might the relevant things that are by chance be just the means or steps taken that, in cases of luck, lead to the desirable result? In our example, the means would be the creditor's going to the marketplace. As Ross points out, we might well describe what happened in the example by saying that the creditor went to the marketplace by luck (516–17). Furthermore, Aristotle does note that the creditor, in those kinds of circumstances, is said to have gone to the marketplace by luck (197a3).Footnote 7 And as I have already observed, the expression ‘things for the sake of something’ is standardly used by Aristotle to refer to means or steps for the sake of some goal. However, when Aristotle specifies what it is that in his example came about incidentally—that is, what satisfies the second condition pinpointed in his initial explication of chance events—he describes it as the creditor going and collecting his money (197a15–16). He seems to specify the chance event as the combination of a means (going to the market) and a goal or quasi-goal (collecting the money). Below in section 3 I will address the question of whether collecting the money is meant to be a genuine goal of the creditor's action—I will argue that it is. But the question for now is whether the chance event is meant to be the combination of going to the market and collecting the money, rather than just the means or quasi-means of going to the market.
Collecting the money is a goal or goal-like outcome, which Aristotle, as we have seen, refers to as a goal and thing for the sake of which. The creditor's going to the marketplace is an action that, as it happened, promoted the goal or goal-like outcome of collecting the money. From a linguistic point of view, it should be possible for the expression ‘thing that is for the sake of something’ to denote a combination of a goal-promoting means and an end or goal: such a combination is a case of X that is for the sake of Z.Footnote 8 There is a third possible interpretation, then: the ‘things that are for the sake of something’ that are supposed to include the things that are by chance are meant to be combinations of goal-promoting means and goals or goal-like outcomes
There is, in fact, good reason to think that this third interpretation is correct. To see this, we need to get clear about what it means for things to come to be incidentally. Aristotle makes plain that in saying that things that are by chance come about incidentally, he has in mind that they are caused incidentally, rather than by a per se cause.Footnote 9 He explicates this idea as follows:
Just as a thing is something either per se or incidentally, so it may be a cause. For instance, the art of housebuilding is a per se cause of a house, whereas the pale or the musical is an incidental cause. That which is per se a cause is determinate, but the incidental cause is indeterminate; for the possible attributes of an individual are innumerable. (Physics II.5, 196b24–29)
And again:
Things do, in a way, occur by chance, for they occur incidentally and chance is an incidental cause. But it is not the cause without qualification of anything; for instance, a housebuilder is a cause of a house; incidentally, a flute player may be so. And the causes of the man's going and getting the money (when he did not go for the sake of that) are innumerable. He may have wished to see somebody, or may have been litigating as plaintiff or defendant, or may have gone to see a spectacle. (Physics II.5, 197a12–18)
When a flute player builds a house, Aristotle thinks, he is an incidental cause of the house. In this kind of case, there is also a per se cause of the house: the housebuilder, who happens also to be a flute player. By contrast, to revisit Aristotle's example of a chance event, nothing is the per se cause of the man's going to the marketplace and getting his money. There is, however, an incidental cause of the event: the man may, for instance, have gone to the marketplace because he wanted to meet a friend there (197a17). His decision to meet his friend at the marketplace, in that case, is an incidental cause of going there and getting the money.
Note that what this decision is the incidental cause of—going to the market and getting the money—is what I have called a combination of a goal-promoting means and a goal or quasi-goal. It is not just a goal-promoting means, for example, going to the marketplace. In fact, the goal-promoting means in the example does have a per se cause. After all, the creditor who goes to the marketplace from a decision to (say) meet a friend there does want to go to the marketplace. His decision to go there is a per se cause of his going there. In other words, there is a per se cause of the goal-promoting means that features in Aristotle's example. By contrast, the combination of that goal-promoting means and the goal or quasi-goal of getting his money back does not have a per se cause. It is this combination that is caused incidentally, as Aristotle says it is. The creditor's decision to go to the marketplace in order to meet a friend there is a per se cause of his going to the market, but it so happens that in going there he runs into his debtor and gets his money back. This is parallel to one of Aristotle's examples of incidental causation: a chef is a per se cause of gustatory pleasure, and it so happens that in providing that pleasure he makes someone healthy (Metaphysics E 2, 1027a3–5; cf. Meyer Reference Meyer1993: 105). It is clear, then, that chance events, conceived of as things that are for the sake of something and as things that come to be incidentally, are meant to be combinations of means and goals or quasi-goals.
3. Natural Goals, Not Quasi-Goals, of Actions
We are now ready to address the question whether the goals in the means-goal combinations that Aristotle identifies as chance events are meant to be genuine goals or only quasi-goals. I will argue that they are meant to be genuine but nonpsychological goals.
We should begin by confronting the difficulty that in discussing the example of the creditor who runs into his debtor and gets his money back, Aristotle seems both to deny and to assert that getting the money counts as a goal of the creditor's action:
The man would have come for the sake of getting back the money when his debtor was collecting contributions, if he had known. In fact, he did not come for the sake of this, but it so happened that he came and did what was for getting back the moneyFootnote 10 . . .. The goal, the recovery, is not one of the causes in him, but is an object of decision and an outcome of thought. (196b33–197a2)
The creditor, Aristotle says, did not come for the sake of getting the money. Yet getting the money, Aristotle also says, was a goal of what he did when he went to the market (similarly at Physics II.8, 199b18–22). To put things mildly, it is not immediately obvious that Aristotle can be consistent in saying both of these things.
One response to the difficulty is to distinguish between a teleological use of telos and hou heneka language and a nonteleological use, in which the relevant expressions denote results with no implication that those results are genuine teleological goals. An alternative response, and the one that I favor, is to distinguish between psychological goals of actions—that is, goals that agents adopt as their purposes—and natural goals of actions, that is to say, goals that actions have by being the kinds of actions they are and, in some cases, by occurring in the circumstances in which they occur. (I will explain in section 3.2 why in some cases the circumstances in which the action occurs may play a role in making some outcome a natural goal of an action.) Both of these responses allow Aristotle to say that getting the money was and was not a telos of the creditor's action without being inconsistent. On one interpretation, getting the money was not the goal of the action but it was its result. On the other interpretation, it was not the creditor's psychological goal, but it was a natural goal of the action, something that the action was for.
The second interpretation, which distinguishes between natural and psychological goals of actions, has the advantage that it does not posit a nonteleological use of hou heneka language. Such a use would, as we have seen, be nonstandard and unparalleled, except possibly for the passages from Aristotle's ethical writings that we looked at in section 1 and that I argued do not, in fact, feature such a use.
In the three subsections of the present section, I argue for the natural goals interpretation by calling attention to a number of details in Aristotle's discussion in Physics II.4–6 that show that he recognizes natural, nonpsychological goals of actions.
3.1 The Appendix on the Expression ‘in vain’ in Physics II.6
In an appendix to his discussion of chance, Aristotle says this:
An indication [sc. of the correctness of Aristotle's account] is the expression ‘in vain’, which we use when a goal fails to come to be, though the thing that is for its sake does come to be.Footnote 11 For example, suppose walking is for the sake of loosening the bowels, and a man walks without having this come to be: we say that he walked in vain and that his walk was in vain, suggesting that this is what is in vain: something which is by nature such as to be for the sake of something else, when it does not accomplish that which it is for and which it is by nature such as to be for—since if someone said that he had taken a bath in vain because the sun did not go into eclipse, he would be ridiculous. For solar eclipses are not what taking a bath is for. (Physics II.6, 197b22–29)
It is not immediately clear, and not important for our purposes, how exactly it is that this appendix on the expression ‘in vain’ is supposed to support Aristotle's analysis of chance. But there are a few points of detail that, given our purposes, deserve special attention. For something to be in vain, Aristotle holds, it must be by nature such as to be for the sake of something else, and it must fail to accomplish ‘that which it is for and which it is by nature such as to be for’ (197b26–7). It seems reasonable to construe Aristotle's references to nature as references to the nature of the action in question, so that the idea is that actions like taking a walk are directed toward suitable goals by their nature, that is, by being the kinds of actions that they are.
Once we construe the references to nature in this way, we can see that the seemingly verbose expression ‘that which it is for and which it is by nature such as to be for’ may in fact be precise and economical. Aristotle may want to distinguish between what an action is for by being the kind of action that it is—a natural goal of the action—and a goal an agent has in taking the action: a psychological goal of the action. In fact, given Aristotle's purposes in this appendix, there is good philosophical reason to distinguish between these two notions of what an action is for. To see this, we should begin by noting that many actions are by their nature directed toward more goals than one. Taking a walk, for example, has a number of natural goals, a number of goals to which it is by its nature suited: these include (say) loosening the bowels after a meal, clearing one's head after intellectual work, and, quite simply, the pleasure one might get from taking a walk.
For it to be appropriate to say that someone takes a given action in vain, it should be the case that the goal that fails to come about is both a natural and a psychological goal of the action. If the goal that fails to come about is only a psychological but not also a natural goal of the action, then it would, Aristotle reasonably thinks, not be appropriate to say that the action was in vain: Aristotle's example of such a case is a person who says that they took a bath in vain because it did not effect a solar eclipse. Effecting a solar eclipse is not a natural goal of taking a bath, and that is why it would be ridiculous for someone to say that he or she took a bath in vain because no solar eclipse came about. This would be a ridiculous thing to say even, and especially, if effecting a solar eclipse was a psychological goal of the action, a goal that the person in question had adopted as a purpose in taking a bath. That, I take it, is Aristotle's point when he says, at the end of our passage, that ‘if someone said that he had taken a bath in vain because the sun did not go into eclipse, he would be ridiculous’.
However, if the goal that fails to come about is only a natural but not also a psychological goal of the action, it would not be appropriate to say that the action was done in vain. To see this, consider a kind of action that has a number of natural goals, such as taking a walk. Suppose you want to loosen your bowels and take a walk for that purpose, and for no other specific, proximate purpose. (In speaking of specific, proximate purposes, I mean to exclude more general, remoter goals such as being healthy, being successful in one's career, having a life that goes well.) And suppose you achieve your goal of loosening your bowels. There plainly are other natural goals of the action of taking a walk, other goals to which this action is by its nature suited and directed, such as the pleasure that one might get from taking a walk. If one or more of those other natural goals did not come about, that does not make it appropriate to say that you took a walk in vain. After all, you only had one proximate purpose in taking a walk, which you attained with complete success, and you were not interested in any of the other proximate natural goals of your action.
Thus, there is good philosophical reason for distinguishing between natural goals and psychological goals of actions and for insisting that for an action to be in vain, it must be the case that the goal that fails to come about is both a natural and a psychological goal of the action in question. On what I think is the most natural and philosophically plausible reading of our passage, this is precisely what Aristotle is insisting on.
On this reading, our passage makes clear that Aristotle thinks that actions such as taking a walk are directed toward certain goals by being the kinds of actions that they are. We can and should make use of this result in thinking about the question whether the combinations of means and goals or quasi-goals in which Aristotle thinks chance events consist are meant to include genuine goals or only quasi-goals. The reason for thinking that chance events in the category of luck include only quasi-goals is that the relevant outcomes (e.g., collecting a debt) are not psychological goals of the agent in question. But this is not a good reason since Aristotle in the appendix on the expression ‘in vain’ explicitly recognizes nonpsychological, natural goals of actions: goals that actions have by being the kinds of actions they are.
However, one might wonder whether this notion of natural goals of actions—the notion of goals that actions have by being the kinds of actions they are—applies straightforwardly to Aristotle's example of the creditor going to the market and collecting his debt. One might be struck by the fact that while (say) decapitation is a generally reliable way of killing someone, going to the marketplace is not in the same way a generally reliable way of collecting what one is owed by some debtor. This is a legitimate concern. In light of it, I want to complicate the picture a bit by suggesting that at least in some cases the status of a nonpsychological goal of an action as a goal depends in part on the circumstances in which the action occurs.
3.2 Natural Goals and Circumstances of Actions
As I have presupposed already, there is no need to ascribe to Aristotle the view that for any given action, there is precisely one specific type of proximate goal toward which it is by its nature directed. People take walks with a view to several proximate goals, and reasonably so. So Aristotle may think, and should think, that taking a walk has several proximate natural goals, including (say) loosening one's bowels after a meal, but also clearing one's head after intellectual work, or simply the pleasure one might get from taking a walk.
One might distinguish between action-types that have privileged natural goals, in such a way as to make it appropriate to speak of the natural goal of the action-type, and action-types that do not have such privileged natural goals, but are directed toward a number of different things. The natural goal of decapitating people is to kill them. By contrast, loosening one's bowels is only one goal among many to which taking a walk is naturally suited; these goals all seem to be on a par with one another. Similarly, going to the marketplace is suited to several goals. People go to marketplaces not only to collect money from debtors, but also to shop, to meet friends, to see spectacles, and so forth. All of these are reasonable goals to pursue in going to the marketplace; all of them are goals to which the action-type of going to the marketplace is naturally suited.
Distinguishing between action-types that do and those that do not have privileged natural goals may turn out to be important if Aristotle's account of chance is to be philosophically defensible. To see this, let us briefly revisit the daughters of Pelias. We would not want to say that it was by chance, or by bad luck, that they ended up killing their father when they cut him up and boiled the pieces. Saying that this was a case of chance, or bad luck, seems inappropriate because chopping people up always leads to their death. Furthermore, it was a matter of reasonable expectation that cutting Pelias up would result in his death, and Aristotle seems right in holding that what is by chance is contrary to, or beyond, reasonable expectation.Footnote 12 There is fairly clear indication that Aristotle wants to rule out cases of this kind when he says, with regard to his example of the creditor meeting his debtor, that this would not be a case of something happening by chance if the creditor got his money back for the most part (197a2–5), which would seem to be shorthand for saying that the episode would not count as a case of chance if the creditor got his money back for the most part whenever he went to the marketplace.
We should also recall that Aristotle begins his discussion of chance by locating things that are by chance in the domain of things that do not come to be in the same way always or for the most part (196b10–17). Aristotle is not as explicit about this as we would want him to be, but his view seems to be that the combinations of goal-promoting actions and natural goals that are relevant to his account of luck are meant to exclude ones in which the two members are related to one another in such a way that the action realizes its goal always or for the most part.Footnote 13 If so, his analysis seems in fact to focus mainly, and perhaps exclusively, on those kinds of goal-promoting actions that are directed toward a number of natural goals, such as taking a walk or going to the market.
The idea would then be that such combinations of goal-promoting actions and suitable goals, when they come about incidentally, count as things that are by chance. For example, the combination of going to the market and getting one's money back, in cases in which it is caused by a desire to meet a friend there, is something that comes to be in the same way neither always nor for the most part and something that is by chance. By contrast, the combination of dismembering people and killing them, when it is caused by a desire to save them, is something that always comes about in the same way (that is, invariably in all cases) and is not something that is by chance because dismembering people is always followed by their death, no matter what the cause of the action may be.
However that may be, the key point for our purposes is that Aristotle clearly recognizes a general connection between the action of taking a walk and goals such as loosening one's bowels. There is a sense in which taking a walk is for loosening one's bowels whether or not in some specific situation one is taking a walk with the purpose of loosening one's bowels. On this basis we may ascribe to Aristotle the view that there is likewise a general connection between, say, going to the marketplace and goals such as collecting debts, meeting friends, seeing spectacles, and the like.
It seems obvious that there is a general connection between going to public places and meeting and interacting with members of one's community. Going to public places at suitable times pretty much inevitably leads to interactions with others. Moreover, people go to public places in order to reap a large variety of benefits that result from interacting with one another. The action of going to some public place, in general, is worthwhile in large part because of the benefits of interacting with others: of doing business, socializing, enjoying artistic performances together, and so forth. On that basis it seems reasonable to say that the action-type of going to a public place is for the sake of interacting with members of one's community and perhaps also, more specifically, that it is for the sake of doing business, socializing, experiencing cultural events, and so forth. Aristotle may think that given general connections of this kind, there is a sense in which going to the marketplace is for collecting money whether or not in some specific situation one is going to the marketplace with the purpose of collecting money.
But he might also think that connections between actions and such rather specific outcomes depend in part on the circumstances in which the action takes place. The fact that your debtor is currently in funds and is having a cup of coffee in a neighborhood café makes available to you a course of action that you have reason to pursue. It makes available a combination of a goal that is suited to being adopted by you (in that in achieving it you would attain some good) and an action that is such as to promote the accomplishment of that goal, and that is therefore choiceworthy. Aristotle might think that the connection between the action of now going to that particular café and the goal of getting your money back depends in part on the circumstances, namely, on the fact that your debtor is currently in funds and that you may easily get hold of him in the café. One might want to distinguish, on Aristotle's behalf, between natural goals of actions and other nonpsychological but genuine goals of actions: the former would be goals that actions are directed to just by being the kinds of actions that they are (e.g., killing for decapitation); the latter would be goals that actions are directed to in part by occurring in the circumstances in which they occur. For simplicity, I refer to both of these kinds of nonpsychological but genuine goals as natural goals of actions.
The important point for our purposes is this: there is good reason to think that one thing that happens when something comes about by luck, according to Aristotle, is that someone performs an action that is directed toward some given goal by being the kind of action that it is and perhaps by occurring in the circumstances in which it occurs. That goal is a nonpsychological but genuine goal of the action in question.
3.3 Objects of Decision
I now want to add a point that pertains specifically to the natural goals Aristotle thinks are realized when something comes about by luck. These goals are natural goals not only in that certain actions are naturally directed toward them. It is not just that they are natural goals of those actions. They are natural goals also, I suggest, in that attaining them realizes some good or avoids incurring something bad. In other words, they are natural goals in that they are suitable goals for people to adopt and pursue.Footnote 14
To see that this is so, we should note that in his final statement of what chance and luck are, Aristotle says that things that are by luck are among the objects of decision (197b20–22). Here is that final statement:
Plainly, then, in the domain of things which, without qualification, come to be for the sake of something, if the things whose cause is external come to be, but not for the sake of that which comes about,Footnote 15 we say that it is by chance. And if such an outcome is for creatures capable of decision and is among the objects of decision we say that it is by luck. (Physics II.6, 197b18–22)
This picks up the earlier claim that chance (meaning luck) is an incidental cause of things that are for the sake of something in the domain of things that depend on decision (197a5–6). That claim in turn follows the observation, made with regard to the example of the creditor getting his money back by luck, that the goal—that is, the recovery of the money—is an object of decision and an outcome of thought (197a1–2). In saying this, Aristotle does not claim that in the situation he has in mind the recovery of money was an object of a decision the creditor had made or that it featured in the creditor's thoughts as a goal to be accomplished.
Far from it: had there been such a decision or such an exercise of thought, it would not have been by luck that the creditor went to the marketplace and collected his debt (as Aristotle notes at 197a3–5). In saying that the recovery of money is an object of decision and an outcome of thought, Aristotle must have in mind that recovering money that is owed to one is something that is suited to being adopted as an object of a decision and as a goal of action. Thus, in saying that things that are by luck are among the objects of decision, Aristotle means to say that they are among the things that are suited to being objects of decision: things that are such as to be decided on. I will say more about what this is supposed to mean in a bit.
One might be puzzled by the fact that Aristotle in our discussion treats the natural goal of recovering money as an object of decision, given that in his ethical writings decision is associated specifically with means selected with a view to accomplishing some goal (for example, in NE III.5, 1113b3–5). However, Aristotle's conception of decision in the ethical writings is such that decisions are individuated by reference to both the means decided on and the goals with a view to which those means are selected. If you know what someone's decision is, Aristotle thinks, you know both the means and the goal that he or she has selected (‘We discern from decision what sort of person someone is [i.e., good, bad, or in between]: that is a matter of what it is that he acts for the sake of, not of what he does’ [EE II.11, 1228a2–4; cf. NE III.2, 1111b5–6; insertion added]). And for a decision to be good, the means decided on must be such as to promote the goal in question, and that goal must be worthy of pursuit (according to EE II.11, virtue of character makes decision correct by making correct the goals of decisions [1228a1–2; cf. 1227b12–13]). Given this orientation of decision both toward means and toward ends, we can make good sense of the fact that, in our discussion, Aristotle treats a natural goal, such as recovering money, as an object of decision.
One might also worry that in associating decision with the selection of appropriate goals and then saying that things that are by luck are among the objects of decision, Aristotle is in effect restricting the things that are by luck to incidentally caused realizations of goals only, rather than conceiving of them, in the way I suggested, as incidentally caused combinations of goal-promoting actions and suitable goals. This worry, too, should dissolve once one realizes that Aristotle conceives of decision as oriented both toward means and toward ends, so that, in a way, the objects of Aristotelian decision are means-end combinations. Thus, in locating things that are by luck in the domain of objects of decision, he can plausibly be seen as locating them precisely as on my interpretation we expect him to, namely, in the domain of combinations of goal-promoting actions and suitable goals.
The domain of objects of decision is the domain of objects that are suited to being decided on. The members of that domain, I am suggesting, are pairs of goal-promoting actions and appropriate goals. But for such pairs to be suited to being decided on, or to be such as to be decided on, it should not only be the case that the action that is one constituent of such a pair is such as to promote the goal in question. It should also be the case that the goal that is the other constituent of the pair is such that in accomplishing it, one attains some good or avoids something bad. It might be a matter of collecting a debt, paying ransom for a friend, seeing an enjoyable theatrical performance, or whatever.
There are of course action-types that are directed toward bad outcomes. Setting one's own house on fire or throwing one's cargo overboard are such action-types. But as Aristotle says, ‘no one willingly throws cargo overboard, without qualification, but anyone with any sense throws it overboard to save himself and the others’ (NE III.1, 1110a9–11). No one with any sense would throw his or her cargo overboard just to be rid of it. And I take it that in speaking of things as being objects of decision, Aristotle has in mind things that a person with some sense would or might decide on. No one with any sense would decide to pursue a goal by accomplishing which he or she would not attain anything good or avoid something bad. So, for a given goal to be suited to being decided on, it must be such that in accomplishing it one attains some good or avoids something bad. The objects of decision, then, are complex action-types that exhibit a means-end structure, where the end in question is in some way good. They are therefore, and in this sense, instances of ‘things that are for the sake of something’, that is, they are cases of one thing being for the sake of another.Footnote 16
Conclusion
On the interpretation of Aristotle's discussion of chance in Physics II.4–6 that I have offered, that discussion contains no evidence for a nonteleological notion of a hou heneka. Rather, we see Aristotle operating with two distinct notions of goals and things for the sake of which: on the one hand, there are the familiar psychological goals, goals adopted by people as their purposes; on the other hand, there are goals that it is natural for people to pursue and that particular actions are directed toward by being the kinds of actions that they are and perhaps by occurring in the circumstances in which they occur. I have called goals of the latter type natural goals of actions. Both of these notions are explanatory and teleological: psychological goals contribute to explaining why given individuals take given courses of actions in particular circumstances; the natural goals of actions contribute to explaining the choiceworthiness of the actions whose natural goals they are. For example, going to marketplaces is generally something worth doing because it enables one to do other worthwhile things, such as meet people and do business. Furthermore, there is no good reason to think that the passages about ignorance of a goal of an action that I discussed in section 1 provide evidence for a nonteleological notion of a hou heneka. Those passages can be made good sense of without positing such a notion once we see that Aristotle takes some types of actions to be by their nature directed to certain goals, the way decapitation is by its nature for killing.
Let me conclude by noting that although the notion of a natural goal of an action with which we have been concerned surfaces rarely in the Aristotelian corpus, it is very much of a piece with Aristotle's general conception of how value is present in the world. Aristotle thinks that for-the-sake-of relations are part of the objective structure of reality not just in the domain of things that happen by nature, but also in the domain of human agency. For example, when he ranks goods, such as wealth, honor, and virtue, in terms of how end-like or final they are, as he does in NE I.7 (1097a25–b6), he means to assign to these various things their proper places in the objectively correct hierarchy of human goods (cf. the discussion of Aristotle's ‘hierarchy of ends’ in Kraut [Reference Kraut1989: ch. 4]). When he says that wealth is not a good that is final, that has the character of an end, on the grounds that we pursue it on account of something else (1097a25–28)—that is, for the sake of something else— Aristotle means to make a point about how it is proper to value wealth. He is aware of the fact that there are plenty of people who value wealth as the highest and most final good (cf. NE I.4, 1095a20–23). But he thinks that this is an improper way of valuing wealth, a way that fails properly to reflect the place of wealth in the objectively correct hierarchy of human goods. That hierarchy is constituted by a system of goods, structured by for-the-sake-of relations that exist in the nature of things. If all goes well, we learn what those goods are and how they are related to one another. In doing so, we adopt as our own psychological goals things that in themselves have the character of a goal. Learning to be good, for Aristotle, is in important part making one's own psychology reflect the objectively correct hierarchy of human goods.
