Three Senses of ‘End’

Ever since Aristotle, the terms that are translated ‘end’ (e.g. the Greek word telos and the Latin finis) have played a starring role in ethical theory. But in fact there are three crucially different things that can be meant by speaking of the “end for the sake of which” an agent is acting.

  1. In one sense, this “end” is the ultimate goal or end result that the agent is trying or intending to bring about.
  2. In a second sense, this “end” is the object of the fundamental wish or desire that motivated the action.
  3. In a third sense, this “end” is a state of affairs that the agent believes to be good, such that the agent believes the goodness of this state of affairs to explain what is good about the action.


In the first sense (1), to say that the action is done “for the sake of” this “end” is to say something about the structure of the plan or intention that the agent is executing in performing the action. This end is what guides your execution of the plan: carrying out the plan involves monitoring what is going on around you, and continually adjusting your behaviour in such a way that that this ultimate goal is achieved.

In the second sense (2), to say that the action is done “for the sake of” this “end” is to say something about what motivated the agent to adopt this plan or intention in the first place. What “motivated” the agent to adopt this plan is what plays a certain role in explaining why the agent adopted the plan; it corresponds to the agent’s motivating reasons for adopting the plan.

In the third sense (3), to say that the action is done “for the sake of” this “end” is to say something about what in the agent’s view justifies or counts as a normative reason in favour of the action.

It should be clear that these three things can come apart.

E.g. suppose that you decide to go for a run. You believe that this is a good thing to do because you think that it is a good way to keep fit. So your “end” in sense (3) – the state of affairs whose goodness you take to justify your action – is your keeping fit.

Suppose, however, that you are not actually motivated by any desire to keep fit; you are motivated solely by a desire to impress some of your fitness-loving friends. So your “end” in sense (2) – the object of the desire or wish that fundamentally motivates your action – is your impressing your friends.

Suppose, finally, that you know that there is only a low chance that your friends will even notice you going for a run, and so it would not be true to say that you intend to impress your friends. However, you do form a specific plan for the run – say, to run clockwise around Christ Church Meadow without stopping. So your end in sense (1) – the ultimate goal or end result that you are trying or intending to bring about, i.e. the goal that guides your behaviour – is simply your running clockwise all the way around Christ Church Meadow without stopping.

In fact, I believe that some of the arguments that Aristotle gives in the Nicomachean Ethics (e.g. in Book X, chap. 7) are vitiated by a failure to draw these distinctions as clearly as he could have done. But that is another story. The main point that I wanted to make here is simply that there are at least these three senses of the term ‘end’.

13 Replies to “Three Senses of ‘End’

  1. And I fear sometimes people use “end” to mean proper or fitting end, regardless of the psychological states of the person whose end it is.
    I guess I take it to be controversial that one can only intend a result if one thinks the result likely enough. But with that premise in place, I am tempted to agree with you that you have offered an example where the three different understandings come apart.
    But if I were to quibble, suppose that the runner learns that she certainly will not impress her friends with this run. She might well stop. Thus it is not obvious that running around Christ Church Meadow is the right description of the goal that guides her activity. Why not say the goal that guides her activity here too is trying to impress her friends?
    You write of the first sense of end that:
    carrying out the plan involves monitoring what is going on around you, and continually adjusting your behavior in such a way that that this ultimate goal is achieved.
    Seemingly “trying to impress your friends” is a better account of this than “trying to run around the Meadow” as the former but not the latter explains many of the things that this person would be responsive to in their behavior.

  2. Sorry, Ralph, but I don’t find this clear at all, and it does seem to me that you’ve offered three descriptions of the same thing.
    As David points out, the (provisionally!)ultimate end or goal in your example does seem to be impressing friends.
    Your last argument, that you can do something without being sure of success, would seem to establish your conclusion (that the ultimate end you’re trying to bring about can’t be impressing friends) only if trying to do something entails believing you will be successful. But that’s obviously false, so I don’t see why we can’t and shouldn’t say that the ultimate end you’re trying to bring about is impressing friends.
    It’s more plausible (though I think still false) to say that this can’t be the ultimate end you’re intending to bring about, since it can at least be argued that intention entails belief. But if you took that route, I think the appropriate thing to say would be that action for an end doesn’t always require an intention.
    The nonidentity of your third sense of “end” with senses 1 and 2 is more arguable, though (on the basis of the kind of analysis of normative judgment that I myself favor) I would deny it too. But you haven’t offered any argument here against an instrumental or end-relational account of normativity.

  3. There seems to be an assumption here that an end must be a final or ultimate end. But, why should that be so? Kant’s “Zweck”, for example, is usually translated as “end”. And all Kant means by “Zweck”, he says, is the object of a free choice or something for the sake of which an action is performed. This means that there can be lots of ends that are not final or ultimate ends, which I think is also Aristotle’s view (is it not?).

  4. David —
    On the contrary, I would say that “running around Ch.Ch. meadow” is an excellent description of the goal or end-result that you are intending to achieve! It is after all what guides how you put one foot ahead of another in the way in which you do; it determines which way you turn as you run along, and so on. So surely it is the end result that you are intending in this case.
    By contrast, I’m not sure that in this case you need be even “trying” to impress your friends at all. Why couldn’t the motivational story about why the agent adopts a plan of action could include a desire or wish for something that the agent believes this plan to be utterly incapable of achieving?
    Thanks to Rosalind Hursthouse, we are familiar with examples of people’s smashing crockery out of rage, and the like. Why couldn’t Michelangelo’s desire that Tommaso should reciprocate his adoration motivate him to write a love sonnet — even if Michelangelo knows full well that this poem will never make Tommaso reciprocate his love? So I don’t see why there has to be this tight connection between the motivating desires and the content of the motivated intention!
    Steve —
    I ‘m not sure that I know how to make it any clearer that these are not three descriptions of the same thing, but of three different things. But here goes…

    1. As it happens I do hold that if an agent rationally intend an end, it must be rational for her to believe that she will succeed in achieving that end. Presumably, we can just stipulate that you are rational in this case. But it could surely still be perfectly rational for you (or Michelangelo) to be motivated to act by a desire for something that you believe that action to be utterly incapable of promoting.
    2. Even on your view of normative judgment (which as you know I don’t accept), there will be many end-relative judgments, which the agent herself might make, that are not relativized to any end that is desired or intended by the agent who is acting. So it seems that you can also happily accept the distinction between (2) and (3).

    Sven —
    Kant may be the honourable exception here. Kant distinguishes sharply between (1) the “subjective end” (subjektiver Zweck) or “intention” (Absicht), (2) the motive (Bewegungsgrund), and (3) the “objective end” (objektiver Zweck). It would take a very careful study of Kant’s texts to substantiate the suggestion, but I’m tempted to think that this corresponds at least roughly to the threefold distinction that I have drawn here.

  5. That’s an interesting suggestion. I myself had in mind Kant’s use of “Zweck” in the Metaphysics of Morals, which seems somewhat less complicated than that in the Grundlegung. What I worry most about regarding your suggestion is the suggestion that “Bewegungsgrund” is “the object of the fundamental wish or desire that motivate[s] [an] action. But, then again, it is quite hard to know what he means by “Bewegungsgrund”, and your suggestion is a plausible one.

  6. Thanks Sven!
    You’re right that it’s hard to know what Kant means by “motive” (Bewegungsgrund). It seems that he would insist that when we act from duty, our fundamental “motive” is the moral law itself — which he wouldn’t be keen to describe as the object of a “desire” or “wish” (although it certainly is the object of the “moral feeling” of respect or reverence). So as I said, it was only a “rough correspondence” that I had in mind!

  7. If I am tracking, it seems so far you are only making a case that the different senses can come apart if the agent is quite irrational.

  8. I agree that one clear kind of case in which the three senses come apart is when the agent is irrational. But there are other cases as well.

    1. The distinction helps to reveal one of the defects in Smith’s “fetishism” argument for motivational internalism. Even if the fundamental desire that lay at the source of the process that motivated you to act was the desire to do what is right (and so acting rightly is your “end” in sense (2)) it does not follow that the ultimate goal that you intend to bring about (your end in sense (1)) is simply that you do what is right: the intended goal or end-result, which guides your behaviour in executing the intention, may be that those people are saved. Similarly, you may believe that the normative reason in favour of the act is not simply the act’s intrinsic rightness, but rather its having the saving of human lives as one of its consequences (which would then be its “end” in sense (3)). So the motivational externalist has resources that Smith failed to notice.
    2. There is also nothing irrational about the cases where we reason from means to ends (in sense (1)). E.g. suppose that you start out with the intention to go for a walk, and on that basis form an intention to walk to Iffley (so that the “end” that guides your behaviour is getting to Iffley by walking). Here the motivational story just starts with a desire for fresh air and exercise, but the intended goal or end-result is something much more specific.
    3. Finally, you might be a hedonist, and believe that your own pleasure is the only thing that can ground any normative reason for you to do anything; still, because of the “paradox of hedonism” you might avoid ever acting with the simple intention of getting pleasure, and you might also acquire fundamental desires for things other than your own pleasure. This too does not seem to me irrational.

    So the three senses of “end” can come apart in many different cases, it seems to me!

  9. Thanks for the explanations.
    Your view on the difference between (1) and (2) seems to be motivated by another premise/ assumption that I would reject. If I understand you, you’re claiming that if a desire is part of the causal etiology of an action, then it is a “motivation” for the action. I certainly agree that a desire for A can cause an action that is not aimed at A. But I think it’s incorrect to say in that case that the desire for A motivated the action. I would argue (and have argued) that motivation is an essentially teleological form of causation, in which case your proposed distinction between (1) and (2) fails. That’s a controversial theory of motivation, of course, but I would also deny that the kind of nonteleological causation by desire you’re talking about licenses us to talk about the action being done “for the sake of” the desired end. Take a Hursthouse kind of case: my desire to win our game of chess causes me to smash my king, upon losing. Even if we grant that my desire to win “motivates” me to smash my king, surely it isn’t right to say that smashing my king is something I do “for the sake of” winning?
    On the separateness of sense (3): true, I accept that agents can make judgments of goodness relativized to ends that they don’t themselves desire. But you were distinguishing between three kinds of case in which an agent acts “for the sake of” an end, and I think that in every case of normative judgment in which that is true, it will be a desired end–collapsing your distinction between (1)/(2) and (3).

  10. Hi Ralph,
    Could you use Velleman’s Freud case to defend the distinction between senses 1 and 2 (the case he uses to attack the belief-desire theory of intention)?
    Freud is motivated to knock the old inkstand onto the ground & thereby break it, by a desire for a new one and a belief that knocking it will lead to his sister getting him a new one.
    The motivation is not conscious and the action is not intentional, but the action is nonetheless teleologically motivated in a way that the Hurstouse actions are not. This looks like a case in which the agent acts for the sake of an end in sense 2 but not sense 1.

  11. Steve —
    Surely we can inquire what a person’s “reasons” were for having a certain belief or emotion (as well as for making a choice or performing an action)? These “reasons” are obviously motivating reasons — not normative reasons — because the person’s reasons might in fact be extremely bad reasons (instead of good, justifying reasons)!
    So it seems that attitudes — including beliefs and emotions — can be “motivated”, just as much as actions and choices. Of course, not every story about the causal aetiology of an action or attitude will cite the agent’s (motivating) “reasons”: reasons explanations have to make the explained action or attitude intelligible in a distinctive way. But all reasons explanations cite some truth about the “motivation” of the explained action or attitude.
    It seems implausible to me to claim that the motivation of beliefs and emotions must always have a “teleological” structure of the sort that you have in mind. So I’m afraid that I really doubt whether your teleological view gives the right account of all motivation as such.
    Brad —
    Velleman’s Freud example seems rather underspecified to me. Unlike Velleman, I firmly believe that there are unconscious intentions as well as conscious ones. However, if we add enough details to the example, then, yes, I believe that it would illustrate one of the distinctions that I have in mind!

  12. Ralph,
    Thanks – that makes sense. I take it that acting for the sake of an end in sense 1 requires more than an unconscious intention? Or is there a further distinction between cases involving conscious and unconscious guidance/monitoring?
    Any way, I thought the case would give you an esp strong response to Steve; you could grant, for the sake of argument, his claim about all motivation being teleological, and still have a case that supports your distinction. What do you think, Steve?

  13. Ralph,
    I’m reluctant to hijack the thread by arguing for a teleological account of motivation here; my basic point is that your case for the distinctions relied upon some implicit assumptions that could be challenged. However, I suppose the case at least has to be made that these challenges are reasonable!
    First, issues about motivation aside, you haven’t addressed my more general point that it isn’t appropriate to talk about doing things “for the sake of” some end if the causation isn’t teleological. Granting that believing for a reason isn’t motivated teleologically doesn’t have any bearing on that question, because presumably in this case there is no “end” in any of your three senses. The challenge to be blocked is that in the case of being motivated by a desire for an end, the “right kind” of connection is always teleological.
    Second, I think that the distinction between motivating and normative reasons is too simplistic, and hence that the fact that someone’s reason for believing is not a good normative reason does not entail that it must be (appropriately labelled) a “motivating” reason. Note that talk about “motivated beliefs” typically suggests beliefs formed for pragmatic rather than evidential reasons. But a teleological story can be told about belief for (evidential) reasons, by appealing to the (sub-agential) end of truth (very crudely). So a reason to believe that p is an indicator (or subjectively, something taken as an indicator) that to believe that p advances the end of thereby being in the right doxastic state with regard to p. I know you’ll disagree with that, but I think there is a defensible view in this vicinity.
    Brad, I don’t think Velleman’s Freud case is at all compelling, for the reason Ralph observes.

Comments are closed.