Voluntariness and habituation

‘Aristotle is not concerned, in his discussion of “the voluntary”, with intentional action, but with action for which one can be praised or blamed.’ Do you agree? (2000 Q3)

• Yes, broadly agree. Although Aristotle’s account of the voluntary is broader than intentional action, and narrower than the praiseworthy/blameworthy, his discussion is really in service of the latter.
• Context: Book III introduces his discussion of voluntariness, just after he’s talked about the doctrine of the mean, and just before going talking about decision & deliberation (and then the specific virtues of character).
  • Methodology: Aristotle starts from phainomena (appearances) and often appeals to endoxa (common beliefs) about what is praiseworthy and blameworthy, to determine what is virtuous.
  • So it’s important for him to establish what is indeed praiseworthy and blameworthy.
    • And this seems like it requires responsibility, which in turn requires voluntariness.
    • S’s freely doing φ is necessary for S to be morally responsible for φ
• Describe Aristotle’s analysis of the voluntary, how it maps on to intentional action:
  • His criteria for involuntariness: either principle not in agent, or ignorant of particulars.
    • So intentionality does come into it! Knowing who, what, with what instrument, etc, seems very much like a theory of intentionality.
    • But the reason he is interested in voluntariness is not really about intentionality per se.
  • “Principle in the agent” is an extremely weak condition. It does capture intentionality in a sense, but you need to be e.g. “blown by a gust of wind” for it to fail, rather than, say, coercion.
    • This does fit with the idea of intentional-action, arguably more so than it fits with praise/blameworthiness.
    • So we might think that Aristotle is concerned with intentionality, because (plausibly) only intentional actions are candidates for praise/blame – this links to how virtue is a state which decides, i.e. intentionality is key.
  • But ignorance of particulars isn’t just about lacking knowledge, needs to be from ignorance, not merely in. So actions from appetite/spirit are voluntary [does this fit here, or elsewhere?].
    • Here it comes apart from intentionality: if someone is drunk and kills another man senselessly, Aristotle doesn’t think this is the required sort of ignorance, even though evidently it’s unintentional.
• Other features pointing towards the focus as on praise/blame vs intentionality:
  • His discussion of putatively mixed acts and conclusion that they are indeed voluntary
    • Aristotle wants us to be able to ascribe praise/blame in these situations, and he thinks voluntariness is necessary for that, which is why he needs to conclude that they’re voluntary.
    • If the focus was on intentionality, there’d be no need for him to bring this up, because evidently they’re intentional.
    • Nielsen’s analysis: they might ordinarily be grounds for shame, but because of the circumstances we praise them for enduring something shameful (provided that the stakes were significant, not trifling).
  • The fact that he is careful to argue that habituation can be voluntary, in III.5
    • Dismisses idea that people can be ignorant of the mechanism, and so then implicitly via transitivity of responsibility, we’re responsible [hence blameable] for viciousness
    • But few would say you intentionally make yourself vicious. Again, he’s interested in voluntariness to secure conclusions about praise/blame.
• However, Aristotle doesn’t simply identify praise/blameworthy with voluntary – there are further conditions
  • In particular, children and animals act voluntarily, but aren’t worthy of praise/blame, because they don’t form a prohairesis.
  • So intentionality is important.
• Conclusion: Aristotle discusses the voluntary in terms of intentional action, and analyses with this as central, but the reason for that discussion is ultimately about praise/blame, via responsibility.

If mixed actions are voluntary then why are they ‘mixed’? Discuss whether the notion of mixed actions is helpful in understanding responsibility. (2017 Q4)

• Mixed actions are described as mixed because they appear to be mixed, even though they are not, on Aristotle’s analysis. This is consistent with his methodology of starting with appearances (phainomena), analysing puzzles (aporiai), and arriving at a conclusion that at least partly vindicates the common beliefs (endoxa).
• Discussion of mixed acts is in the first half of III, which covers voluntariness and decision.
  • Aristotle’s sufficient criteria for an action being voluntary: it must be done with knowledge of the particulars, and the principle in the agent.
  • An example of mixed acts: the captain of a boat throwing cargo overboard because they’re in a storm.
  • Or e.g. someone held at gunpoint handing over their wallet.
  • So these do straightforwardly meet his criteria. Nielsen: they’re just putatively mixed acts.
• Intuitively, we would think that it is absurd to hold someone as blameworthy for the consequences of an action like this where their hand has been forced. And that might lead us to conclude that Aristotle is wrong to take mixed actions as voluntary.
  • But this rests on a conflation of voluntariness, responsibility, and aptness for blame.
  • First, it seems like Aristotle thinks voluntariness isn’t sufficient for responsibility: he says that infants and animals can act voluntarily but don’t make decisions so cannot be responsible for their actions.
  • But for the mixed acts, they do meet this prohairesis condition. Where they help us understand responsibility for a seemingly-bad action is that it isn’t sufficient for blameworthiness.
  • The reason is that the action taken in these circumstances was the right one. It’s what the phronimos would have done, presumably, and as Aristotle says they are the standard and reference for what is right.
• This relates to discussion of shame: enduring shame (socially) for a fine thing – the aptness for shame attaches to the actions in usual circumstances; similarly one might take an action which is usually blameworthy but not when under “mixed” conditions.

Post-Claude & examiner report notes

• Aristotle calls the acts mixed in his own voice, not merely presenting the endoxon.
  • His considered verdict is that they are in the abstract perhaps involuntary, but overall more like voluntary.
  • The point here is that unqualifiedly throwing cargo overboard / etc is the sort of thing that nobody would choose, so it resembles the involuntary.
  • But in the circumstances, it’s both chosen and choiceworthy, so voluntary.
• It’s Nielsen, not Sauvé-Meyer, who has the “putatively” analysis.
  • The target here is to show that there’s no dirty hands – i.e., the right act in a mixed case is simply right, without any moral residue.
    • But she’s not denying that there’s something mixed about the acts in the sense described above, i.e. she accepts that unqualifiedly they’re bad
    • The captain is saving lives at the price of the cargo; it’s a tradeoff between two evils.
  • Voluntariness has fixed criteria, but pardon and praise vary on top of that, depending on the circumstances.
    • And there are some things we cannot be pardoned for, no matter how much the conditions might overstrain us – e.g. Alcmaeon killing his mother
• Consider whether the category of mixed acts is doing explanatory / theoretical work.
  • You can raise a redundancy objection: if the mixed acts really are just voluntary, then the label “mixed” isn’t really adding anything. The explanatory work is done by the voluntariness criteria plus the assessment of the situation.
  • But (as noted) they are useful as a test case / aporia that forces us to explain the gap between voluntariness and pardon
• Remember that Aristotle does think that some agents are blameworthy when doing mixed acts: if they endure a shame for a trifling end. Naturally the phronimos wouldn’t.

‘How can I be “in some way” responsible for my character? Either I’m responsible for it, or I’m not.’ Discuss whether this is a good objection to Aristotle. (2017 Q5)

• No, this isn’t a good objection to Aristotle. It’s entirely comprehensible how one could admit of degrees in responsibility, and in any case Aristotle’s main goal is to show that we bear nonzero responsibility for our character; nothing in his argument depends on the claim that we don’t bear full responsibility for it, and he just needs to establish the weaker point (since the stronger claim that we do bear full responsibility is unintuitive). A much more biting objection would be that in many cases Aristotle seems unable to establish that I am responsible for my character at all – specifically, if I have not been brought up in fine and noble habits.
• I will first make the general argument that responsibility admits of degrees. Then I’ll explain why Aristotle is interested in establishing responsibility for character at all, and demonstrate that even if it were true that responsibility were a binary concept, Aristotle’s argument would go through as intended. Finally, I’ll raise a better objection about Aristotle’s account of responsibility for character.
• It is true that you can collapse responsibility into a binary construct if you take some threshold. (one response to the quoted statement would be: well, define yourself to be responsible for your character if you bear any responsibility for it, and not responsible otherwise)
  • But the point is that the underlying concept admits of degrees.
  • Consider Fairground Ride:
    • You’re a fairground operator, and tasked with pressing the start button when everybody is buckled in, as indicated by a light on your dashboard, and press emergency stop if that light goes off. There are also signs and several announcements made instructing people to buckle in. You see the light illuminate and you press start. You’re then distracted and don’t notice that the light goes off, because an adult on the ride unbuckled their seat for fun. The ride continues, and they’re injured.
    • It looks like you’re in some way responsible for the injury. In a counterfactual analysis, had you acted differently (and rightly), the injury wouldn’t have occurred. But clearly other people also bear responsibility – specifically, the adult who unbuckled his seatbelt.
  • It’s beyond the scope of this essay to give a precise analysis of when the relationship of responsibility holds between an agent S and outcome X, but it seems to involve counterfactual impact, i.e. had S not performed/omitted some action, then X would not have (or would’ve been less likely to) obtain.
  • You can collapse this to something binary, but that just destroys information. What you might like to do is use possible-worlds machinery to determine how responsible you are.
• Aristotle cares about establishing responsibility for character because, given his views about habituation, he would otherwise struggle to argue that virtue is praiseworthy and vice blameworthy, which are endoxa relied upon throughout the Ethics.
  • It seems like for S to be justifiably blamed or praised for X, they must be responsible for it. (There may be further necessary conditions, such as [for blame] X not being a pardonable action, but we needn’t concern ourselves with them.)
  • But according to Aristotle’s view of habituation introduced in II, and applied throughout NE, we acquire stable states of character which determine the kinds of actions we perform, and our conception of the end (Book III).
  • So if we are not responsible for our state of character, then it’s hard to see why the virtuous agent is worthy of praise or the vicious one deserving of blame.
    • This is a problem for the methodology: in general Aristotle often identifies virtues by looking for what is praised. But this fails if praise doesn’t track virtue.
  • Conversely, if we are responsible for our character, and that is responsible for our actions, then (with the additional assumption of transitivity of responsibility), we can be held responsible for our actions, and thus it’s possible to blame/praise for them.
• Suppose that the objector is right, and that we are either fully or not-at-all responsible for our character. Then Aristotle is in the clear, for all the times where we aren’t not-at-all responsible.
  • And surely these coincide with the cases where we are not-at-all responsible under his theory that allows for us to be “in some way” responsible.
  • So the objection brings no new difficulties.
  • And it comes with the cost of denying the very intuitive idea that we can be only in some way responsible.
  • Consider Riches-to-Rags-to-Riches:
    • Robbie grew up in a comfortably-off family, but tragic misfortune meant that as a young adult he lost all his wealth and became destitute. He had to beg and scrounge in order to get by. One day, fortune smiled on him, and he received a large inheritance from a long-lost cousin. It is now several decades past then, and he has not had reason to be stingy for a long while. But his character is still disposed to that, and he’s less generous than would be virtuous.
    • It seems like Robbie is partly responsible for his character – surely in all that time that’s passed he could’ve reformed – but not wholly.
• A much more compelling objection is that Aristotle often fails to show that we are at all responsible for our character, if we haven’t been brought up finely.
  • Book X on moral education; does this mean that we aren’t to blame people at all whose parents were deeply morally flawed if they themselves also are?

Post-Claude & examiner report notes

• Probably too strong to say objection brings no new difficulties; seems like the threshold would not be at 0 but somewhere in between, so we would get different verdicts on whether responsible or not.
  • E.g. in Robbie case, probably binary account says: not responsible.
• Can push on transitivity, and in particular: if I’m only partly responsible for my character, does responsibility for resulting actions get diluted proportionally?
• How to read “in some way”?
  • The objector seems to be unhappy because they think it’s a hedge.
  • But really what Aristotle means is “via a specific causal route, to a specific extent”.

When, if ever, can someone be blamed for their actions according to Aristotle? (2020 Q10)

• Aristotle thinks there are a variety of circumstances in which someone’s actions are blameworthy; what they share in common is that the agent has done something wrong, bears responsibility for their actions (either at that moment or transitively via their character), and is not pardonable due to unusual circumstances.
• Naively, one might think that all wrong actions are blameworthy. But that is not Aristotle’s view; he is more forgiving.
  • Voluntariness is required for blameworthiness.
    • So if you act from ignorance, or the principle is not in you, then you wouldn’t be blameworthy.
    • In the latter case, it’s hardly “your action” anyway.
    • The thought here is that voluntariness is required for responsibility, which in turn is required for blameworthiness.
  • It’s not sufficient; there are cases which “overstrain human nature” where someone might voluntarily take an action but still be pardoned.
    • But some actions are so bad they never merit pardon – e.g., killing one’s mother like Alcmaeon. So certainly Aristotle thinks there are some actions which are worthy of blame!
• Voluntariness condition actually does not rule out many types of action.
  • The drunk, who acts in but not from ignorance.
  • The sailor on a sinking ship, whose hand is forced but nonetheless acts voluntarily.
    • Enduring shame for a trifling end is blameworthy.
• An objection: our conception of the good is not in our control; it’s habituated or our nature. But Aristotle has a two-pronged rebuttal.
  • Only totally insensible unaware of habituation; transitivity assumption (Sauvé-Meyer)
  • X.9 and II.2 on being brought up in fine habits; the intended audience of his book
  • If conception of good is in our nature, then no acts would be praiseworthy either. But that’s not true, from phainomena, so we must reject that.

Post-Claude & examiner report notes

• Spell out why responsibility is necessary for blame
  • Recall that children and animals act voluntarily but don’t deliberate, so aren’t blameworthy.
  • Blameworthiness requires: voluntary + deliberative agency + no overstraining + act reflecting a bad state of character.
• Could talk about the akratic
  • Why is she less blameworthy than the vicious person? Because her prohairesis is correct.
  • So blameworthiness tracks character, not just the wrongness of the act.

‘I choose how I act, but I don’t choose how I feel. Therefore, virtue and vice are not up to me.’ Discuss. (2022 Q5)

• On Aristotle’s view, it’s simply false that (provided suitable initial conditions are satisfied) you don’t choose how you feel; feelings can be cultivated and developed via habituation. But this is a rather implausible account of human psychology, and objections along the lines of the titular statement do pose a challenge for Aristotle.
• On the nature of virtue: it is a state which decides, consisting in a mean defined by reference to reason. The mean is a particular situation-specific response about both feelings and actions.
  • So the objection raised by the title is along the lines of: although I might be able to voluntarily control my actions, I have no control over what feelings I have. Thus it’s not in my control whether I attain virtue (on Aristotle’s account), since I can’t control whether I have the stable disposition about feelings the mean requires.
• Prima facie, it does look like a puzzle why we ought to care about someone’s feelings, rather than merely their actions. If someone consistently does the right thing, then what does it matter how they feel about it? More specifically, what is the failure of the continent agent (enkratês)?
  • Two possible resolutions.
    • Someone who does the right thing but contrary to their feelings is less reliable than the virtuous agent in arriving at the right action.
    • The virtuous agent has more harmony among the parts of their soul – the appetitive part and the rational part both generate desires that are in agreement, and this is good; they correctly take pleasure in the fineness of right action (Coope).
• In any case, Aristotle certainly thinks it is worse to be merely continent. If our feelings aren’t up to us, then that seems like a problem for him, in that there’s no path for a suitably-motivated but currently-feeling-incorrectly agent to attain eudaimonia.
• However, he can resolve the puzzle.
  • Provided you’re brought up in fine habits, then you will grow accustomed to the pleasure that accompanies fine action.
  • Habituation is all about pleasure and pain; it’s having the right emotional responses to your actions.
  • So it is up to you – if you engage in right actions, then (according to him) the pleasantness of the fineness will just come and shine through. [does he say more about the mechanism?]
• The trouble with this is twofold.
  • Has nothing to say about those not brought up in fine habits. But maybe he’s just not concerned with them at all.
  • Psychologically, not very plausible that we are able to control our feelings through sheer habituation. We’re very complex creatures with system 1 vs system 2 biases, often right action does involve careful reason despite intuitions / feelings.
    • E.g., scope insensitivity!
  • Also, it seems more intuitive that in some sense it’s especially admirable for the enkratic agent to do what she knows to be right despite not enjoying it.
• Maybe briefly mention worries about determinism.

After feedback from Claude

• Aristotle does think that we can exercise rational control over our emotions
  • VII.7 tickling example: you can anticipate what’s coming and are able to avoid laughing; similarly reason can pre-empt emotion
    • So it’s not only that habituation changes your emotions over the long-term; foresight can control them in the short term
  • Recall that there’s the part of the soul which isn’t rational but is responsive to reason, like a child to their father.
• Perhaps spell out more explicitly the similarities/differences between actions and emotions
  • Both cultivate over time via practice; partially controllable in moment via reason
  • But actions more directly under your control; emotions have some involuntary physiological components; it’s rather different to “try to feel X” than “try to do X”
  • Aristotle’s point is that over a long period, the dissimilarities diminish.
• Can engage with the technical terminology of what is “up to us” – Aristotle takes this as including things causally downstream of what we’ve chosen over time.
  • This just is further support to the habituation-responsibility-transitivity story.
• Maybe more natural examples of emotions we can’t really control: grief, romantic attraction.
• Talk about temperance in particular, as when you train your appetitive part to have correct desires (III.5).
• Idea that the enkratic is more admirable is of a Kantian flavour.

Does Aristotle’s view of habituation imply that virtue of character is an unreflective and routine-like disposition? Why or why not? (2024 Q3)

• No, he is clear that virtue of character must involve reason & deliberation – but it is a puzzle to explain how this fits with the notion that the virtuous person is well-habituated and takes pleasure in their fine actions.
• Three conditions for an action to be virtuous: decided on and done for the sake of the action, from a stable state, in knowledge.
  • So decision is crucial. And a decision (prohairesis) is a deliberative desire to do something that is up to us. Thus certainly it’s not unreflective.
  • At VI.12 Aristotle considers whether a state devoid of reflection might be virtuous, but he clarifies it cannot be – the distinction between a state in accord with (kata) reason vs involving (meta) it.
    • Ruling out natural virtue as full virtue.
• But we are owed an explanation of how it can be that the virtuous person needs to deliberate and reflect even despite their habituation.
  • One explanation: as noted, it is difficult to be virtuous and there are many ways to be in error. As Hursthouse notes, it’s a bit like hitting the bullseye.
  • So even someone entirely habituated must engage their intellectual faculties to arrive at the right action in each situation.
    • While virtue is a state, eudaimonia consists in the exercise of virtue; it’s an activity.
• A remaining doubt: given that virtue does involve decision, how is the virtuous person to act finely in situations where rapid action is called for? E.g., in the heat of battle.
  • Maybe this is where their habituation comes into it, coupled with eustochia (quickness).
    • [something to explain: why does the habituation matter elsewhere, given that we’re stipulating they do deliberate? → because it means they have the right conception of ends via wish?]

Post-Claude & examiner report notes

• You can mention Aristotle’s analogy with craft-goods: like how expert craftspeople pay attention to their product, the ethically excellent are highly reflective & adaptive to the situation.
  • (This is what Annas proposes.)
  • And actually, virtue is even more demanding than craft, because not only does the quality of the product matter, but also the motivational states of the agent producing it.
• Could be more specific about how the not-yet-virtuous agent differs in their actions from the prudent one.
  • E.g., in new situations / where they lack guidance, they’ll go wrong.
  • (Cf the stickler for legality vs the decent person.)

‘But presumably there are some things we cannot be compelled to do. Rather than do them we should suffer the most terrible consequences and accept death.’ (III.1, 1110a26-7) What role do these claims play in Aristotle’s account of the voluntary, and how might he argue for them? (2024 Q5)

• This comes from Aristotle’s discussion of voluntariness and decision in III.
  • He gives the example of Alcmaeon, who killed his mother.
  • It comes just after discussion of the conditions for voluntariness. When Aristotle says “compelled” he doesn’t really mean “there are some things that can only be done voluntarily”, because he has an extremely low bar for that.
    • i.e. even when the things are less bad than matricide, e.g. throwing cargo overboard, or wearing unusual clothes because a tyrant forces you to, these are voluntary (principle in agent and done in knowledge of particulars).
  • This is more about when an agent may be pardoned for their actions.
• Conditions which overstrain human nature: in some cases, even if an action is done voluntarily and is wrong, it may not be blameworthy.
  • This illustrates a hierarchy of concepts: voluntary is required for responsibility which is required for blame/praiseworthiness.
    • Children & animals act voluntarily but don’t bear responsibility because that requires decision.
    • And then even when responsible, we might be pardonable, because of overstraining.
    • This makes his account of voluntariness more plausible, since otherwise it would generate far more blame than seems reasonable. (Cf mixed acts too, which are unqualifiedly shameful but not given the circumstances if done for a noble end)
• The claims quoted are providing limits on the sorts of actions which can be pardoned.
  • Aristotle thinks that some are so beyond the pale that no amount of pressure can make them excusable.
  • Intuitively this might capture our idea of there being certain constraints that no moral agent can permissibly violate.
• Aristotle could defend it by saying that in suffering consequences and accepting death, we are sacrificing ourselves for the fine.
  • But this doesn’t explain why it is unpardonable to fail to do so – only that it’s admirable to not.
  • Maybe he thinks any minimally-moral person would be able to stand firm against such terrible things.
  • But under suitably extreme circumstances, surely it’s plausible that all our human reason and principles will break; it’s not at all clear why overstraining only applies to bad actions up to a certain level of badness.

Post-Claude & examiner report notes

• [Note that mixed actions are in III.1, and would be relevant to bring in.]
  • They fit into the dialectic very neatly: even though voluntary, they may be pardonable, and indeed perhaps even right.
  • And then the point is that this passage stops Aristotle from being too permissive.
• Remember that he thinks certain actions taken under overstrained conditions are even pitiable, not only pardonable.
  • Also remember that for an action to be involuntary it needs to involve pain and regret, as well as ignorance of the particulars. Otherwise it’s merely non-voluntary.
• To evaluate how Aristotle might argue for these claims, a few points:
  • Remember some acts are so intrinsically bad that “their names automatically include baseness”.
    • So perhaps these are the ones that nobody can be compelled to do – the virtuous person would always prefer death to performing them, because death for the fine is admirable.
    • Still, this doesn’t explain why it’s unpardonable rather than merely suboptimal to do those base acts.
    • Perhaps it’s connected to being bestial / subhuman.
  • His methodology of starting from commonly-held beliefs; it is presumably an endoxon that certain moral constraints can’t be overridden.
  • Since virtue is a stable state and not easily dislodged, the character of a virtuous person should be resilient even when under pressure.
    • Again, this doesn’t really go far enough to secure Aristotle’s non-compellable view though.