Monday, March 23, 2009

Disambiguating 'Autonomy'

Another thing I like about Arpaly's book is her careful disambiguation of eight senses of 'autonomy':

(1) Agent-autonomy: self-control, or acting on your explicit 'best judgment' (in contrast to akrasia).

(2) Personal efficacy or the ability to take care of yourself (in contrast to domestic 'dependents').

(3) Independence of mind: in contrast to blind 'followers' and brainwashing victims.

(4) Normative, moral autonomy: what's violated when one is hit or stolen from.

(5) Authenticity: being true to yourself, or acting on what are in fact your deep concerns.

(6) Self-identification: feeling unconflicted, and acting on what you perceive as your deep concerns. (This differs from authenticity since one's self-image could be inaccurate.)

(7) 'Heroic' autonomy: supposedly possessed only by ubermensch. (I'm a little unclear on this one.)

(8) Reasons responsiveness: the ability to recognize and respond to reasons.

Arpaly suggests that only the last of these is strictly necessary for moral responsibility. (It's also true that someone incapable of agent-autonomy won't be a moral agent, but she suggests that this is because both have the same precondition: being a reflective creature.) In any case, Arpaly seems right in observing that mere lapses in agent-autonomy don't excuse: there are plenty of blameworthy akratic actions (and some praiseworthy ones too, cf. Huck Finn).

A couple more notes:
* Authenticity is also relevant to (the degree of) moral responsibility, insofar as we are more praiseworthy or blameworthy, on her account, when the morally significant concerns behind our actions are deeper concerns of ours.

* Historicism about responsibility (the view that whether S is morally responsible in phi-ing depends on extrinsic historical facts about how S came to be the way she is) may partly be motivated by confusing 'independence of mind' [which is an uncontroversially historical notion] with other -- more important -- ahistorical senses of 'autonomy'.

* In light of all this ambiguity, we might do better to retire the word 'autonomy' in favour of whichever precise sense we have in mind: self-control, mental independence, authenticity, reasons-responsiveness, or whatever.


  1. I am uneasy saying that these are eight discrete senses of autonomy, as they all seem to be applications of an underlying generative concept. Strictly speaking, autonomy means "self-governance," or "self-law-making." It seems right that these different senses come up, but I wonder if there is more than family resemblance to bind them. Or maybe "autonomy" is an elliptical term that acquires its specific content from what one is (tacitly) negating. Whether there is a common concept or the term is a trouser-word, a taxonomy bothers me. In any case, I think you are right that it is time that we retire the ambiguous "autonomy" in favor of more precise, and explicit formulations.


  2. These kinds of taxonomies are always interesting, but I'm suspiscious of the claim that these eight are so unrelated. In particular, I think you could reduce them all either to (a) reasons responsiveness, which I tend to think of as autonomy in the normal sense of the word or else to (b) what I believe is generally referred to as political autonomy.

    (I ignore (7) in the following. I'm not familiar enough with Nietzsche to comment.)

    In particular, I suspect that reasons responsiveness covers (1), (3) and (5), since each of those is a matter of responding to, and only to, the reasons one has. This is less obvious for (1) and (5), but I take it that we have reasons to be internally consistent in our attitudes.

    Political autonomy covers (2) and (4), for obvious reasons. If there's a distinction between these two at all, isn't it the classic distinction between positive and negative liberty?

    That leaves (6), which I understand to be a matter of feeling that one is autonomous. But that isn't a way of being autonomous any more than feeling that one is rich is a way of being rich. So I'm a little perplexed by what that is doing on the list at all.

    So perhaps you could say something more about why you think that (6) belongs on the list at all, and why you think the remaining six senses of "autonomy" are so different from one-another?

  3. I think Alex raises some good points. In fact, reason responsiveness and political autonomy themselves don't seem too far-removed. An argument could be made that political autonomy is presupposed in reason responsiveness: in order to follow out your own reasons you must have the external as well as internal capacity to legislate your own actions. I do think that (6) is reasonable if we think of self-identification of autonomy as a way of individuating one's actions as one's own, produced with reason and intent.

  4. Alex - I'm not sure where you're coming from here. The eight senses are very clearly conceptually distinct (no two are synonymous), and even metaphysically distinct in that for any pair it's possible to have one without the other. There may be correlations or other connections between them (perhaps reasons-responsiveness is a precondition for some of the others, if you think it is a precondition for agency itself), but not anything recognizable as "reductions", I should think.

    Here are some examples to prove distinctness:

    1. The possibility of rational akrasia (e.g. Huck Finn) shows that it's possible to have reasons responsiveness without self-control.

    3. Take the standard Ann and Beth case: the Dean has neuroscientists rewire Beth's brain so that she is as dedicated an academic as Ann. Ann is reasons-responsive, hence so is the New Beth, but New Beth lacks historical independence of mind, due to the interference of the neuroscientists.

    Even ongoing mental influence looks like it could be compatible with reasons responsiveness, at least in non-extreme cases. Consider someone who becomes a [non-blind] 'follower' precisely because they are sensitive to the reliability of their leader. Sometimes people say that followers ipso facto lack mental independence. But this (weak) sense of mental 'dependence' or influence is compatible with reasons responsiveness.

    5. Unless we're subjectivists about reasons, authenticity can presumably come apart from reasons-responsiveness. And in any case it's obvious that it can come apart from self-control (since people may resolve to act in ways that fail to reflect their deepest concerns). So the transitivity of identity prevents you from holding that both self-control and authenticity are identical to reasons-responsiveness.

    [If you're merely making the far weaker claim that an act's being authentic or being self-controlled are pro tanto reasons in favour of so acting, then I'm not sure how that counts against the point that these various senses of 'autonomy' need to be distinguished. Though I also think even the weak claim is false, due to the standard worries about bootstrapping.]

    2 and 4 are clearly distinct: I could make you more efficacious while violating your moral autonomy, say by surgically enhancing you against your will. (E.g. Wolverine.)

    You could say that political autonomy "covers" both, but that's just another way of saying that common talk of "political autonomy" is ambiguous, since there are two very distinct ideas here.

    Finally: #6 (self-identification) belongs on the list because this is a list of things that actual people sometimes seem to have in mind with their 'autonomy'-talk, not a list of things that ideal people would call by a single term if they were thinking clearly.

    P.S. "feeling that one is autonomous" -- in which sense? :-p

  5. Thanks for the helpful response Richard. I fear this is going to very quickly going to become a very wide-ranging discussion, but a brief response:

    I take it that reasons responsiveness means not only that one is responding to the reasons that one has, but also that one would be responding differently if those reasons were different. One isn't reasons responsive merely by being lucky. On this conception of reasons responsiveness, Huck Finn, New Beth, and the blind follower all fail to be reasons responsive. They don't, in nearby possible worlds, do the reasonable thing. (e.g. if the neuroscientists had wired Beth to be a committed glamour model instead, she'd be doing that, even if there are no good reasons to do such a thing.)

    I suspect that wide-scoping the relevant reasons can deal with authenticity, and also self-control, without leading to bootstrapping. One ought (to stop caring about one's deepest concerns or else act on them), and one ought (to stop believing that one should do X or else do it).

    I'll put aside the rest: I think there's more to be said, but I can understand where you're coming from a bit more clearly now.

  6. I agree that a blind follower probably isn't (very) reasons responsive. But I meant to raise the possibility of a non-blind follower in my previous comment.

    re: Huck Finn, you seem to be assuming that one can only track reasons at the level of explicit consciousness and deliberation. Why think that? Huck might have reliably good subconscious dispositions: his compassion leads him to genuinely track right-making features (it's no accident that he helps Jim). Hence the term 'rational akrasia', rather than 'lucky akrasia'. (The latter is possible too, but less interesting.)

    Finally, in the Beth case I think you're assessing the wrong counterfactuals. Reasons-responsiveness is surely an intrinsic property of agents, depending only on their actual dispositions, and not what their dispositions might easily have been (say if they only narrowly avoid being subject to massive external manipulation). New Beth is an intrinsic psychological duplicate of Ann, and hence reasons-responsive if Ann is. The relevant counterfactual is not whether New Beth would still respond to reasons appropriately if the neuroscientists had given her a different character. The question is simply whether New Beth, given her actual character, would still respond to reasons appropriately if faced with different reasons. And the answer is that she would.

  7. Richard,

    Good points. This is more complex than I thought. I still suspect that what I'm saying has merit, and that automony is more unified than you're imagining. Very briefly: I'm not sure in what sense one is a follower is it's not being done blindly, Huck Finn fails to be reasons responsive in his explicit judgements, and New Beth, as you're phrasing the case, I now think is free, but just a different person.

    But I certainly think (and I have no doubt you agree!) that there's a lot more to be said on each of these cases, and to establish what I want to say would take a lot of work that isn't really appropriate in a comments thread. Perhaps I'll post on it myself sometime. Thanks for the discussion.


Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)

Note: only a member of this blog may post a comment.