Does moral theory drive philosophy of action? Here’s what I mean. A main question of philosophy of action is what an action is. It seems to me that the answer to this question is strongly influenced by what type of moral theory one accepts.
For example, an act-utilitarian does not need anything more out of actions than for them to be events with effects. She can rank voluntary and involuntary behavior on the same scale; she can rank the behaviors of agents and non-agents on the same scale. This is likely to produce a very thin account of actions.
A deontologist needs more. For a deontological theory to work, actions have to fall into types (e.g. promises, assertions/lies, thefts, murders). These types of action generally are intentional actions. They also are not individuated by the causes or effects of the action. The deontologist may also need a distinction between acts and omissions, and/or between intended and merely foreseen actions. So a deontologist needs an action theory that can make these distinctions, and will think of actions in terms of them. Similarly, for a Kantian, actions are essentially the sort of thing that have a “maxim” behind them.
A virtue theorist also has different needs from a theory of action. Actions have to be behaviors of a creature whose life is partially constituted by its actions, a creature with an end to which the ends of its actions are related. So the virtue theorist will think of actions as limited to a certain kind of creature and as having teleological properties which the other two viewpoints can dispense with.
I am assuming in all this that the main thing we do with our concept of action is evaluate actions, either prospectively in deliberation or retrospectively in assigning responsibility, praise, and blame, and that therefore the main constraint on the concept of action is the nature of the evaluation.
All this suggests that Anscombe, in “Modern Moral Philosophy,” got something important backwards. She argued that due to the alleged decrepitude of current moral theory, we needed to do some more foundational work in philosophy of mind and action. But I think she has the foundations the wrong way around: we are not likely to do interesting philosophy of action without some direction, at least in general terms, from a moral theory.
Heath,
I think I really disagree with your approach to this issue. Just because an act-utilitarian believes that the consequences of an action are the only directly morally relevant features of it does not mean that he must begin with a different account of actions than do his adversaries. Act-utilitarians say that consequences of actions are all that matter; but in order to say this with any meaning, they must first have a clear idea of what actions are and how they are to be distinguised from non-actions. So, like Anscombe, I believe that we must first begin with a coherent concept of action in order to develop a sound moral theory, and not the other way around.
I think some comments of Mill reinforce this idea. He writes:
I submit that he who saves another from drowning in order to kill him by torture afterwards, does not differ only in motive from him who does the same thing from duty or benevolence; the act itself is different. The rescue of the man is, in the case supposed, only the necessary first step of an act far more atrocious than leaving him to drown would have been. Had Mr Davies said, “The rightness or wrongness of saving a man from drowning does depend very much”–not upon the motive but–“upon the intention”, no Utilitarian would have duiffered from him (Util. II, 19, footnote).
Let me end by asking a question: why do you suppose that just because different moral theories find different aspects of actions to be morally relevant, this means that they are using different conceptions of actions in their theories?
Heath,
I’m not sure I follow. Just a quick comment on your characterisation of AU. You write that:
“For example, an act-utilitarian does not need anything more out of actions than for them to be events with effects. She can rank voluntary and involuntary behavior on the same scale; she can rank the behaviors of agents and non-agents on the same scale. This is likely to produce a very thin account of actions.”
So, AUian I take it that an act is wrong if it doesn’t maximize utility or something like that. Does that really range over all events that have effects? Over behaviour of non-agents? I find it quite absurd that lightning strikes, flowing of the river, growing of the tree, buzzing of the fly, flying of the bird, a butterfly’s flapping of the wings, ticking of a clock and so on could be morally wrong. I don’t think many utilitarians want to say this.
Also, I’m not sure why a deontologist needs more from a theory of action. What she might need is some more nuanced way of classifying actions but I’m not sure that makes a difference in the question what it takes for an event to be an action. Besides, not all deontoligists are generalists anyway.
Jussi says “I find it quite absurd that lightning strikes, flowing of the river, growing of the tree, buzzing of the fly, flying of the bird, a butterfly’s flapping of the wings, ticking of a clock and so on could be morally wrong. I don’t think many utilitarians want to say this.”
The thing is, though, that for an act utilitarian the claim that an act is wrong is simply the claim that it is one of the acts, among the various alternatives, that has less than the best consequences. Heath’s point, I take it, is that there is nothing in this characterization to prevent us generalizing and similarly evaluating non-acts. And some consequentialists do just this. Here’s Derek Parfit in Reasons and Persons:
“C [Consequentialism]’s central claim is:
(C1) There is one ultimate moral aim: that outcomes be as good as possible
C applies to everything. … Consequentialism covers, not just acts and outcomes, but also desires, dispositions, beliefs, emotions, the color of our eyes, the climate, and everything else. More exactly, C covers anything that could make outcomes better or worse. According to C, the best possible climate is the one that would make outcomes best.” (pp.24-25)
In other words, the property possessed by the best climate–the property of being the climate, among the various alternatives, that makes outcomes best–is a property that looks a lot like moral wrongness as act utilitarians conceive of moral wrongness.
Scott,
You ask why I think moral theories use different concepts of action. My main reason is that there are few constraints on the analysis of the concept of action, other than those provided by the evaluation of actions. That is, there is not much else to *do* with the concept action, except engage in the evaluation of actions. Granted, there may be a few other constraints. But, I think, not much. This implies, not so much that moral theories *do* employ different concepts of action, but that they *can*, and that they often will. For example, we could expect consequentialists to be willing to criticize and dispense with the doing/allowing distinction, or the intended/foreseen distinction, whereas deontologists will find ways to support them.
But here are some other strands of evidence: first, Anscombe’s own work on action, particularly the later stuff, is clearly influenced by her Catholic moral commitments. Second, in all the work that has been done since Intention on action theory, I am not aware that very much if any of it has had any effect on moral theory. (I would be interested to hear counterexamples to this claim.) Third, the philosopher who most closely follows Anscombe’s ideas about modern moral philosophy, Alasdair MacIntyre, does not rest much if any of his argument on analyses of action concepts. His method is entirely different.
I’m slightly confused. I’m not sure if the point was that all events count as acts (not very plausible – I recall some Humean saying that all acts need to be desired at least) or that, as you say, we shouldn’t care about acts at all but rather about all events, whether or not acts, that can have the same moral properties (not very plausible either).
I’m not surprised that the property which global consequentialists ascribe to all kinds of things if they are not optimific is *a lot like moral wrongness* *as conceived by act-utilitarians*. I would be slightly surprised if act-utilitarians were right about which *acts* have the property of being morally wrong but even more surprised if all events, no matter how disconnected from agency, were candidates for bearing the moral properties we talk about in the actual moral discourse. Smells like changing the subject to me.
Heath,
I think you are right to emphasize the connections between action theory and moral theory, but I am not sure why you think Anscombe was wrong about how to proceed. Our views about how to live are much more likely to be distorted by bias, parochialism, and other social psychological factors than are our views about the metaphysics of action. Scientists would be rightly miffed if they were told that because their best theories of the mind or the world were phrased in terms not well suited to our favored moral theory, they should give them up. Rather, it seems like moral theorizing itself should be constrained by our best descriptions of ourselves and the world.
W/r/t Michael’s point–
I’ve always found Parfit’s description of consequentialism refreshingly willing to take the theory to its natural conclusion–that any event with good or bad consequences can be morally right or wrong. However, this seems to me to be a reductio of Consequentialism; if, for example, the climate is eligible to be morally right or wrong, then I don’t understand what we are talking about any more when we talk about morality.
If the property of rightness is the same as the property of having the best consequences, then, rather trivially, that property will be shared by the action of saving the lives of a hundred children in Darfur (in a suitable context of alternatives) and the Mediterranean climate (in a suitable context of alternatives). So must the act-utilitarian say they’re both right? Well, assuming that she definess rightness in that way (and I’m not sure why she would have to do so, but let’s assume she does), it will be a consequence of her theory. But she need not *say* so. Perhaps the *concept* of rightness or wrongness has more to it than picking out the property, or perhaps calling something right/wrong has the *implicature* that it is to be praised/blamed. In these cases, it would be misleading to call the climate “morally right”, though it would be strictly speaking true that it had the property. This might go some way towards explaining away the manifest absurdity of calling the song of the cuckoo wrong.
Heath,
You say:
there are few constraints on the analysis of the concept of action, other than those provided by the evaluation of actions. That is, there is not much else to *do* with the concept action, except engage in the evaluation of actions.
I guess this is where you and I disagree. One question we can ask is, “What are actions, and how do they differ from mere bodily movements or events such as hurricanes, etc?” Another question we can ask is, “How are we to morally evaluate actions?” You seem to be saying that the second question is prior to the first; I however, have a hard time understanding this, for the second question relies on a concept that stands in need of explication and analysis before it can even get off the ground.
I think that moral philosophy presumes distinctions that the action theorist is interested in determining. Consider intentionality; (not claiming anyting new)it seems that we should make a distinction between an event and an action based on the intentionality of the agent. An action is an event that is intentional, or where it would make sense to apply the concept of intentionality as part of the explantation of why something occurred. Therefore, actions are a species of events such that all actions are events, but not all events are actions. The quesion is essentially an ontological one pertaining to the structure of actions/events.
In ethics we already presume this distinction, regardless of our normative persective, in so far as we do not hold people who do something unintentionally, or by accident, as responsible for the event/action as we do if it were intended. The concept of intentionality is applicable in non-intentional or unintentional, actions and part of the concept of intentionality implies the existence of conditions (excusing conditions to borrow from Austin) that would limit responsiblity.
Now, having said this, it seems to me that Heath is making (at least in part) an empirical claim and if so, this claim can be tested by developing an experiment to test whether, or not, moral theory drives action theory. I must admit that I do not know what this experiment would look like, but I do think it would be interesting if one could be developed.
John,
I’m not sure that we do not hold people responsible for intentional actions. Here’s a case from Al Mele. Take a drunk driver whose driving home from bar at night. At some point she suddenly untintentionally sways on the road. As it happens she hits a pedestrian and kills him. She did not see him and neither was her swaying intentional in any way. So, she did not intentionally kill the pedestrian. Neither did she do any action intentionally that lead to the killing of the pedestrian. My intuition though is that we can hold her responsible for the killing.
I’m also slightly puzzled that you first say that actions are intentional events and then talk about non-intentional or unintentional actions.
Jessi
Thanks for your comments.
We do hold people responsbile for actions they intentionally chose to do. I said we do not hold people AS responsible if their actions are unintentional. There are degrees of responsibility that are determined, in part, by the level of intentionality displayed by the agent. Re your example from Mele; I think we would hold her responsible for getting drunk in the first place with her knowing full well that she needed to get home and not making alternative plans for doing so.
The reason I speak of unintentional actions is that I think that we distingusih between those events that are actions from those that are not based on whether or not we can apply the concept of intentionality, which includes the criteria for performing or doing something unintentionally as part of the overall concept. Again re your example, if we want to distiguish why she did something wrong we need to be able to say that even though her action re driving the car after leaving the bar was unintentional, the prior conditions, or at least some relevent prior conditions, where the result of intentional decisions on her part, or should have been. I think we need some way of distinguishing events such as a pen failing from a table where the concept of intentionality or uninentionality does not make sense re the falling aspect, from actions such as those in your example. I may intentionally push the pen off the table, but the reason it falls when pushed off the table is outside any analysis that would include intentionality. Your example needs intentionality to make sense of what is occuring from a moral point of view.
I hope this clarifies what I was trying to say earlier.
Jussi
I apologize for misspelling your name. It was unintentional, but I should have been more careful.
John A.
John,
no worries. It’s pretty hilarious with that kind of a name anyway to see people try to pronounce it over here. I’ve thought about changing my name to Michael Smith… For what it’s worth, I think much of the problems in this ball-park can be avoided by recognising that responsibility is an ambiguous notion. There is a sense of attributability and the sense of blameworthiness. Attributability is the sense in which we attribute acts to agents and distinguish acts from events. Responsibility under this sense may require that you intend to do the action at least under some description. In this sense, we do not want to attribute responsibility to the drunk driver of the act of killing (but of course we may want to do so of the act of drinking). This allows to accept that there are unintentional actions that we do not attribute to the agent and her agency but rather explain them as accidental. What distinguishes those from mere events is another interesting question.
The second sense of responsibility is when we hold agents blameworthy for their actions. Now, my intuition of the drunk driver case is that the agent is blameworthy for killing the pedestrian and this blameworthiness goes beyond the responsibility of starting to drink earlier in the pub. However, if we accept that responsibility is ambiguous, then holding the agent responsible, i.e., blameworthy, does not necessitate that we attribute the act to the intentional agency of the agent. Maybe, it is enough for this sort of responsibility that the agent could have earlier on avoided putting herself under the risk of the relevant kind of accidents.
Jussi
By an large I would agree with your analysis of attributability and blameworthiness. I agree with you that there is ambiguity in the notion of responsiblity which makes these types of interchanges very important. This is one reason why I mentioned degrees of responsibility. At one level, responsibility simply means that someone did something. At this level intentionality would not enter into the analysis. However, I am not sure that we would want to suggest that simply because someone did something that they are to be blamed, or praised, for the action.
As you do, I want to hold the drunk driver responible for her actions to the point that we can blame her for what occured. At this point I do think that blame or praise is dependent on the level of intentionality that the agent displayed, or should have displayed, when performing the action. The problem as I see it is if we do not utilize the concept of intentionality what other criterion can be used to blame her for her actions. Would we blame her if the accident was a result of her not being drunk, but having a heart attack? I think we would not. I think we should argue along the lines that in order to be blamed for her action she should have been in the right frame of mind to perform the actions she did, but was not. She determined the conditions within which her actions would unfold. Because the accident occured as a result of her not being in the right frame of mind as a result of choices she made, she is responsible (blameworthy) for what occurs.
While thinking about your comments, I came up with the following questions: 1)to what extent is intentionality volitional? To what extent is the driver being drunk a result of her failure to properly exercise her volitional capabilites? 2) can someone be coerced or deceived into performing an action that we would then consider intentional even if we knew she was coerced or deceived?
I am enjoying and learning from this exchange. What this has demonstrated to me is the importance of trying to reach an agreement on meaning thru dialogue. Or, the importance of developing a complete ontology of what is going one when we say that someone did something and is morally responsbile for what she did.
Relating this to Heath’s original comments, I think this exchange does suggest, if not cdemonstrate, that philosophy of action drives moral theory.