In this article, I explore Bernard Williams's notion of regarding one's past action ‘purely externally, as one might regard anyone else's action’ (Williams Reference Williams1981a: 27). The distinction between what I call taking an ‘external’ and ‘internal’ view on one's action occasionally arises in the moral luck and agent-regret literature, but this is often cursory, and there has been remarkably little work on what Williams himself means by this distinction, perhaps because Williams gives us little to work with.
I show that the external view precludes agent-regret and relates to agent-regret's particular expression. If we want to claim anything like a full understanding of agent-regret and its expression, we need a clear view of what exactly taking an external view involves, and this is important since some skepticism about agent-regret turns around this expression, for example Wallace (Reference Wallace2013: 42–45). I also show how we can link the external view to other aspects of Williams's work; for instance, I show how it differs from extrinsic luck, which throws that concept into relief, and I show how it might connect to Williams's distinction between the view from here and the view from there.
Understanding the external view should also be of interest aside from trying to understand Williams. The plausibility of taking an external view—regarding one's own action as if it were anyone else's—is important because our view of our own actions relates to our views of what we have done and what we are responsible for. Tony Honoré (Reference Honoré1999) holds, quite plausibly, that the outcomes we bring about—the things we do—shape our identities. If we can take an external view, then this adds another layer to the picture: sometimes what we do does not, at least in our own eyes, shape our identities. The external view also gives us an insight into how we understand the failure to take up other emotional reactions. For instance, one might look back on a past action and, despite seeing it as a worthwhile achievement or a shameful thing, feel no pride or shame. Understanding the external view will help us get a grip on this and deepen our understanding of the arousal of emotions. Turning to the philosophy of law, the external view might shed light on the reaction involved in cases of strict liability, where I can regret that a bad thing happened because of my action, but aside from the fact the law makes it matter in terms of legal reparations, not care about the fact that I caused the harm.
Still, all that is down the line. For this article, I focus on Williams and on the relation between the external view and agent-regret. I sketch this in the first section. In the second section, I offer a few constraints on any understanding of the external view. Then I turn to possible accounts of the external view. Given that it arises in Williams's ‘Moral Luck’ (Reference Williams1981a), one might associate it with intrinsic and extrinsic luck. Given that Williams wrote ‘Moral Luck’ for a symposium with Thomas Nagel, one might associate it with Nagel's internal and external views. I argue that neither of these elucidates the external view, and I offer two alternatives. One based on identification, the other based on what we care about.
1. The External View and Agent-Regret
Williams held that there is an important distinction between ordinary regret and agent-regret, and that our propensity for agent-regret reveals that the unintentional effects of our actions are important to our sense of who we are as agents (Williams Reference Williams1981a: 27–30). In short, when we regret something, we regret states of affairs; when we feel agent-regret, we regret our own action, partly due to unforeseen or inescapable consequences (Williams Reference Williams1981a: 27). Only the person who performed the action—famously, the lorry driver who runs over the child—can feel agent-regret, whereas anyone who knows about it can feel ordinary regret (Williams Reference Williams1981a: 27). Moreover, agent-regret has its own ‘particular kind of expression’ (Williams Reference Williams1981a: 27). For instance, Williams says that when we feel agent-regret, we will be motivated to act in different ways than we might when we feel ordinary regret: when I feel agent-regret, I might still feel the need to perform some reparative action even if an insurance payout has been made to any victims; whereas (although Williams is not explicit on this point) if I feel ordinary regret, we can suppose that I might want to help the victim but would be left satisfied by an insurance payout helping instead (Williams Reference Williams1981a: 28–29).
If I take an external view, I recognize that my action brought about the unfortunate state of affairs, and might even acknowledge that I should compensate the victim. I might, for instance, think that it would just be ‘unfair for the sufferer to bear the costs' when my actions brought about those costs (Williams Reference Williams1981a: 28). (Perhaps Williams means that given I would have received any benefits of my action, I should also bear the costs. See Honoré [Reference Honoré1999].) So, I do not take up exactly the same perspective as a bystander who feels regret, since I am willing to pay compensation. (If a bystander helps I suspect we see this as ‘assistance’ not ‘compensation’.) Yet Williams claims that if I take an external view, I would also be left content by an insurance payout (Williams Reference Williams1981a: 28–29). This means that an external view precludes agent-regret, given that agent-regret leaves one unsatisfied by an insurance payout. Taking an external view results in ordinary regret and extraneous factors (the unfairness of the victim bearing the cost when I would have received any benefits) explain why I am willing to pay compensation. It is not my agent-regret, or my concern for my own agency, that explains this willingness.
Given that someone who takes an external view takes an external view on her own action, this shows that ‘Agent-regret requires not merely a first-personal subject-matter’ (Williams Reference Williams1981a: 27). This is because of something central to agent-regret: as Marcia Baron (Reference Baron1988: 261) puts it, ‘That he was the agent … is ineliminably a part of what he feels.’ And Raimond Gaita (Reference Gaita2004: 53) says that this sort of regret is ‘directed not only to what we did or to its effects, but also at the fact that we did it’. In contrast, one who takes an external view regards the action ‘as one might regard anyone else's action’ (Williams Reference Williams1981a: 27). When I take an external view, that I was the agent is not ineliminably a part of what I feel; I regret that someone did such a thing (and I acknowledge ‘someone’ was me), not that I did such a thing. I regret that you got hurt, and I see that I hurt you so perhaps I should compensate, but I do not regret that I hurt you, and I am left perfectly content when you receive a large insurance payout that assuages the harm you suffered. When I feel agent-regret, I regret that I hurt you.
2. Characterizing the External View
It is not entirely clear what sort of cases Williams has in mind when he mentions the external view. He does not link the external view to any of his developed examples, so we have little to work with. But we have already encountered three features of taking the external view:
i. One regrets one's own action.
ii. One regrets it in the same way as one regrets anyone else's action.
iii. One might be willing to pay compensation, but would be satisfied were an insurance payout to cover the costs instead.
The task now is to see what account of the external view can capture these.
We must not associate taking an external view with just any case where agent-regret is in the same ball-park but does not arise. Consider the javelin-thrower Williams discusses in Shame and Necessity (Williams Reference Williams2008a: 61–62). He throws a javelin, and someone else runs in the way of it and dies. The javelin thrower does not regret what he has done, but that is not because he takes up an external view. Rather, he does not regret his action at all. He regrets that someone died, but he sees the case as one where the runner ran in the way of the javelin. The only action the thrower regrets is the runner's.
Taking an external view is also broader than just realizing that one was involved in what one regrets. Joseph Raz (Reference Raz2011: 233) attempts to distinguish agent-regret from taking an external perspective (though Raz admits that he is not sure that he captures the exact differences Williams was after). Raz imagines a case where he feels agent-regret about something, forgets about it, hears a story about JR doing that exact thing, and regrets that JR acted in that way (Raz Reference Raz2011: 233–34). Raz says that as soon as he realizes that he is JR he experiences ‘a different kind of regret … it is regret that I am such a person’ (Raz Reference Raz2011: 234). When Raz realizes that he is JR, he returns to feeling agent-regret, rather than ordinary regret.
We need not just construe this interpretation of the external view in terms of forgetting. Oedipus regretted that someone's actions brought a curse upon the city; the character of this regret changed once he realized the action was his own. Williams, in Shame and Necessity, brings out how Oedipus ‘moves to the discovery of just one thing, that he did it’, where he moves from regretting what was done to a thoroughly internal view involving agent-regret when he realizes that he was the one who killed Laius (Williams Reference Williams2008a: 69). We also find a case like this in Nathan's rebuke of David (2 Samuel:11–12). David lusts after Uriah's wife Bathsheba, so contrives to send Uriah to his death. Nathan, sent by God to hold David to account, tells David the story of a rich man who took a lamb from a poor man as a gift for a visitor, rather than slaughtering one of his own. David is enraged by this greed and says, ‘As the Lord liveth, the man that hath done this thing shall surely die’. But then Nathan says to David ‘Thou art the man’, and the full force of what David did to Uriah is brought home to him. He moves from regretting the selfish actions of the rich flock-owner to an internal view on such selfish actions by seeing that the story was an analogy for his own actions. Such an interpretation of the external view turns around not realizing that it was my agency involved, such that once one has an ‘it was I’ moment, one no longer takes an external view.
This approach certainly makes sense of some cases where one fails to feel agent-regret—but it does not do so in the right way. Williams says the regret is ‘directed’ at a past action, and this suggests that one is aware that it is one's own action (Williams Reference Williams1981a: 27). In Raz's example it would be infelicitous to say that, before he realizes he is JR, his regret is directed at his past action. So, it is not clear that ignorance involves taking an external view on my action. Yes, it captures that I take a view on an action, but it omits how I take a view on my action. Even if we can get past infelicitous expressions, the point remains that I can take up an external view and be willing to pay compensation even though an insurance payout would satisfy me. Surely this means that in some such cases I am fully aware that it is my action that is at issue—otherwise, it is not clear why I would be willing to pay compensation in the first place. Understanding the external view in terms of some revelation fails to account for this. I can take an external view on my action without being ignorant.
So what sort of cases do involve an external view? It seems to me, from Williams's brief discussion, that the external view is available in everyday cases. Take the following two cases, to which I return throughout this article:
• Maddie smashes Zack's vase.
• Arnold, who several decades ago carved graffiti into an old oak tree, learns that his grandson has done the same thing.
In the first case, Maddie, through no fault of her own, smashes Zack's vase. If Maddie (non-faultily) smashes Zack's vase, what might Maddie do in response? Well, perhaps she will try to fix the vase, or procure a replacement, or at least apologize to Zack. (For an interesting, but different, discussion of a similar case see MacKenzie [Reference MacKenzie2017].) How might a bystander, someone external to the situation, react? Max, another of Zack's friends, might try to buy a replacement vase, or help repair it; but we suppose that Max would be perfectly happy if an insurance company paid out and thus compensated Zack for the loss.
The idea introduced by the external view is that we can imagine Maddie taking a similar view to that of Max, even though she recognizes that she smashed the vase. She might think that it is bad that someone smashed the vase and try to fix it, or think that if it is between Zack and her, she should fix it; but she does not see anything important about the fact that she smashed it and is content if an insurance company pays out. We can certainly envisage—as much as it might annoy us—Maddie saying, ‘Oh, it's such a shame that your vase was smashed’, rather than ‘I'm so sorry I smashed it’. Of course, we might suggest to her that she is missing something important and that even if an insurance payout is available, she should do something else for Zack, not least apologize. Taking an external view is the sort of thing we might judge as inappropriate; in this case, there might be something unacceptable about Maddie ‘relegating [her] costs to … the insurance fund’ (Williams Reference Williams1981a: 29). Below I address why this might be, but the point here is just that we can easily imagine, and have all probably encountered, such a case where someone fails to—introducing another concept that is surely related to this discussion—take responsibility for what she has done and regards herself as no different from any bystanders.
In the second case, Arnold learns that his grandson Billy has carved graffiti into an old oak tree that is now withering and in need of attention. We can see Billy's action as regrettable. It is not just regrettable that the tree withers—that might be a natural occurrence that we are perfectly happy to see happen; rather, what is regrettable is that Billy unnaturally hastened the death of the tree. Likewise, I might not regret it if a lion dies, but do regret someone's shooting it.
Arnold might take an external view on how, several decades ago, he did the very same thing: he also carved into a tree as a child and the tree died because of his carvings. When he hears about Billy's actions, he remembers his own action. He regrets it in just the same way that he regrets Billy's actions. Arnold sees it as a silly thing a child did. But he does not attach any importance to the fact that he carved into the tree. (In contrast to Maddie, I suspect we are less likely to think poorly of Arnold for taking this view.)
Both cases are useful to us because it is not too much of a stretch to imagine Arnold or Maddie going either way: they might take an internal view and feel agent-regret, or they might take an external view and regret it only insofar as they regret anyone else carving into the tree or breaking the vase. This fits with Williams's (Reference Williams1981a: 28) idea that the external view is a ‘kind’ of recognition; we can suppose that the same agent might go one way or another. We might feel agent-regret, or might take an external view, in exactly the same case. (I say a little more on this below, in section 5.3.)
I take this as the basic idea behind taking an external view: in taking an external view, something comes between recognizing that I did something, and regretting that I did it—and this ‘something’ is not mere ignorance. But what is it that occasions such a breakdown between acknowledging it as my action and regretting that I did it?
3. Extrinsic Luck
One might think that the external view is linked to Williams's (Reference Williams1981a: 25–27) distinction between luck intrinsic to a project and luck extrinsic to a project (I label these ‘intrinsic luck’ and ‘extrinsic luck’). Williams (Reference Williams1981a: 25) says that when a project fails due to luck, ‘it matters how intrinsic the cause of failure is to the project itself’. We will see that an intrinsic failure tends to occasion agent-regret, while an extrinsic failure normally leads to ordinary regret. Thus, we can see why one might think this helps explain the external view: because both extrinsic luck and the external view preclude agent-regret. But when one fails due to extrinsic luck one's regret does not tend to attach to one's own action: so, this does not help to explain the external view, in which one does regret one's action (just not that it was one's own action).
I start by getting clear on what the intrinsic/extrinsic luck distinction involves, focusing only on cases of bad luck. Williams (Reference Williams1981a: 22–24) illustrates matters with the example of Gauguin, a competent painter who wants to take a new direction by painting in Tahiti, but who is unsure whether this will work out, or whether his skills are suited to such a style. If Gauguin fails to make it as a painter due to a lack of artistic creativity or a failure to take to the new style, he suffers from bad intrinsic luck since artistic creativity is a vital part of the project of becoming a painter; whereas if he suffers an injury that stops him from painting, the luck is extrinsic to his project of becoming a painter (Williams Reference Williams1981a: 25). We can see that artistic talent is far more closely related to the project of becoming a great painter than avoiding an injury is: it is more intrinsic to the project.
Another example might help to bring out this distinction: both Torino F.C.’s ‘Grande Torino’ side, and Manchester United's ‘Busby Babes' were successful football teams that could have gone on to great European victories. But both teams suffered serious air disasters in which many players were injured or killed. The teams did not, a few exceptions aside, push on to greater success. Contrast this with the ‘Total Football’ Dutch team of the 1970s, which should have won a major championship, but failed to do so: they just failed to win some important matches. The luck that afflicted the Dutch (winning football matches always requires a bit of luck) was closely related to the project of being a great football team; the luck that afflicted Grande Torino and the Busby Babes had nothing to do with football—it was a tragic external event.
• Intrinsic luck is luck that is related to one's project that leads to its failure.
• Extrinsic luck is luck that is unrelated to one's project but still leads to its failure.
Of course, some things are more intrinsic or less intrinsic. This is a nebulous distinction admitting of many vague cases. (Is it an intrinsic failure when injuries derail a footballer's career?) And it is hard to get a grip on what exactly is ‘related’ to a project—of course, they would not have died had they not been on the flight, and they were on the flight because they were footballers, and footballers competing at the highest level must travel—but the cases above should be clear enough to get a rough handle on matters.
Williams concludes his discussion of agent-regret by saying that Gauguin will only feel agent-regret if he is unsuccessful in his project (Williams Reference Williams1981a: 30–31; see Lang, Reference Lang, Chappell and van Ackerenforthcoming). (The discussion in Williams [Reference Williams1981a: 25–27, 36] is about Anna Karenina, a main character in Tolstoy's novel and a focus of Williams's discussion, but it applies just as much to Gauguin. Anna introduces irrelevant complications.) But things are subtler than that: in discussing the intrinsic/extrinsic distinction, Williams points to a difference in the thoughts that someone afflicted by each kind of luck will have. Gaugin, when he suffers bad intrinsic luck, will think that what he did was ‘insupportable’ and he will regret what he did (Williams Reference Williams1981a: 27). This is a clear case of agent-regret. But failure by extrinsic luck leads to a different set of thoughts: ‘He does not, and never will, know whether he was wrong’ and he need not think that ‘he failed’ (Williams Reference Williams1981a: 25; see Lang, Reference Lang, Chappell and van Ackerenforthcoming; Wallace Reference Wallace2013: 158). This Gauguin's thoughts do not attach to his decision. He does not seem to have enough material with which to form the thought that his action was regrettable. So—since things are unclear—he need not think that he should not have done what he did, or that there was anything regrettable in how he acted: he does not regret what he did, so he does not feel agent-regret.
Now, this understanding of the relationship between these forms of luck and regret leaves open that someone who is afflicted by extrinsic luck might regret setting off on the project in the first place—we can imagine an injured Busby Babe regretting ever becoming a footballer. But Williams's discussion suggests that such a character's regrets will most likely not attach to his actions, but to what happened. So, in the ordinary case, someone who is afflicted by extrinsic luck will regret how things turned out, just as anyone might but with the added fact that what turned out happened to him. He will not regret what he did. But someone afflicted by intrinsic luck will, in the ordinary case, regret what he has done. This is clearly agent-regret. Gauguin-injured might regret not making it as a painter, but Gauguin-the-failed-painter will regret what he has done: not adapting to the new style and thus failing to become a painter. Likewise, the surviving Busby Babes might regret not winning the European Cup, but they do not regret losing this or that match; whereas the Dutch side might regret their failures to win matches or to exercise fully their footballing abilities as well as they might have done. (In a remarkable display of perseverance two of the surviving Babes, Bobby Charlton and Bill Foulkes, won the European Cup a decade later, managed by Matt Busby.)
So, we have the following picture:
• Agents afflicted by intrinsic luck will regret what they did – they will feel agent-regret.
• Agents afflicted by extrinsic luck will regret how things turned out.
Despite how taking the external view and being afflicted by extrinsic luck both go hand in hand with regret rather than agent-regret, we should not equate external views with cases of extrinsic luck, and this is for two reasons. First, we have seen that extrinsic luck need not involve taking any sort of view on one's own agency. Gauguin-injured regrets not becoming a famous painter; as I have shown, it is far from clear that he has any regrets about what he did. The external view takes as its object one's own action, whereas when one suffers from extrinsic luck and feels regret, one might not even consider one's action. This account does not explain how one might regret one's own action but not feel agent-regret, so it is not a satisfactory account of the external view.
The second problem, which is not quite as apparent in the examples I have offered so far, is that the intrinsic/extrinsic distinction applies to the luck that afflicts projects. So, we need something that the luck can be intrinsic or extrinsic to. There does not seem to be any reason to restrict taking an external view to cases where projects are involved. Maddie and Arnold are plausible cases of taking up an external view, but it is not obvious that they had any relevant project with success conditions, such that we might say that there are certain lucky incidents intrinsic and others that are extrinsic to it. So, we might take an external view on something that does not admit of the intrinsic/extrinsic distinction and that thus cannot be classified in terms of extrinsic luck.
4. Nagel
One might think we will find a more promising path to understanding the external view if we remember that Williams's paper was part of a symposium with Thomas Nagel. In his essay ‘Moral Luck’, Nagel also explicitly mentions an ‘external view’, which he explores at great length elsewhere (Nagel Reference Nagel1986; Nagel Reference Nagel1979a: 37–38; Nagel Reference Nagel1979b). To echo Williams's own comment on a related issue, the internal and external in Nagel are not ways in which things exist; rather, they are each a ‘style of understanding’ (Williams Reference Williams2014: 262). We can, in principle, understand the same thing from two perspectives, akin to how Williams's internal and external views are kinds of recognition. And in Nagel's ‘Moral Luck’, he considers the internal and external views as styles of understanding agency. This approach seems promising, but I show below that Williams's understanding of an external view cannot be the same as Nagel's, despite some similarities. (For a critical overview of Nagel's work see Thomas [Reference Thomas2008], especially chapter 5.)
Nagel sees us as in tension between being things in the world subject to its causes and effects, and being free responsible agents (Nagel Reference Nagel1979a: 37–38). He says that we can take up two views corresponding to this tension: the internal view, and the external view. The external view corresponds to the idea that we are things in the world subject to its causes; under this view, we see ‘an act as something that happens' but this ‘seems to omit the doing of it’ (Nagel Reference Nagel1979b: 199). What Nagel means here is that the external view omits that someone—a free rational agent—has decided, has acted, or has somehow exercised agency. The external view sees what seemed to be a free action as entirely determined. Because of this total determination, ‘nothing of [the agent] is left to intervene in those circumstances' (Nagel Reference Nagel1986: 114). We find that the ‘responsible self’ is ‘swallowed up by the order of mere events' (Nagel Reference Nagel1979a: 36). Thus, the ‘doing’—the choosing, the exercise of free agency, and other such paradigmatically agential elements—is omitted, because there is no agent to do anything.
Contrast this with the internal view on an action: when I take an internal view, I see myself as responsible, freely choosing to act rather than determined by a prior chain of causes; and I understand the action as my own doing, as part of who I am (Nagel Reference Nagel1979a: 37). The internal view sees us as responsible agents, the external view removes this responsibility and agency from the picture:
• Nagel's internal view sees action as the exercise of free rational agency.
• Nagel's external view sees action in terms of predetermined cause and effect.
The internal view sees Frasier dating Mia as Frasier making a free choice to spend time with a woman whose company he enjoys; the external view says he had no choice in the matter and was compelled to date her because she looks just like his mother.
Nagel, in a move conducive to any attempt at reading Williams's external view as the same as his (albeit in a slightly different context), says that when one takes an external view of one's actions, one views oneself ‘as just one person among others' (Nagel Reference Nagel1979b: 205). This results from how it does not, on the external view, matter that I did it because I did not do it in any meaningful sense; the universe just caused it via my body. It is as if anyone had done it. This aspect accords with the feature of the external view that says when Maddie takes an external view she thinks that it is a bad thing that someone has smashed the vase but attaches no importance to the fact she was the one who did it. So, this account might go some way towards capturing Williams's understanding of the external view as engendering regret that is just the same as the regret we feel towards anyone else's actions.
But trying to understand Williams's external view in this way suffers from two problems. First, despite appearances, we cannot make sense of regretting it as if it were anyone else's action, nor can we make sense of regretting one's own action. Nagel's external view removes the sense that an action is any such thing; rather, it becomes a mere link in the chain of causes and events. The self is swallowed up, the doing is omitted, and nothing of the agent is left. As such, one regrets what happened in the same way as had anyone else done it, but also in the same way as had it just happened: had, say, an earth tremor knocked over the vase. The doing is omitted—so the doing cannot be regretted. So, we make sense of regretting an action as if it were anyone else's action only at the cost of no longer regretting an action in any meaningful sense, and we have seen that sometimes actions (carving into the tree, shooting the lion) are regretted whereas mere events (the tree's or lion's dying naturally) might not be. So, Nagel's external view does not capture that what is regretted is genuinely an exercise of agency. It does not explain why we might not feel agent-regret despite regretting our action because it does not let us regret the action.
Second, it is not as though Arnold or Maddie view what happened in a purely mechanistic way devoid of any agency: Arnold accepts he acted and was to some extent responsible for what happened and that some agency was involved. Nagel's external view does not capture how our agents view matters. (As Williams points out, it does not seem that we have to accept that the problems of moral luck boil down to Nagel's metaphysical distinction [Williams Reference Williams1995: 242].)
5. Identification and What We Care About
So, we cannot properly understand taking a view on our own action if we try to understand the external view either in terms of Williams's extrinsic luck or Nagel's external view. I now argue that Maddie, when she takes the external view, does not think it is important that she smashed the vase. I offer two ways of understanding this. The first concerns identifying with an action. On this view, when I take an external view of my action I admit that, in some sense, I did it; but I do not identify with performing it, so it does not reflect on me and does not serve to occasion agent-regret. The other account concerns caring about doing certain things. On this view, I admit that I did that thing, but although I might care that such a thing was done, I do not care that I did it.
5.1 Identification
Consider ‘distancing’ from one's own action. (The terminology is adapted from Dan-Cohen [Reference Dan-Cohen1991: 990]. Dan-Cohen [Reference Dan-Cohen2008] applies this to Williams.) As Meir Dan-Cohen says when he considers the self, it is able ‘to identify with various elements and thereby integrate them into itself, or to distance itself from them by objectifying them and holding them at arm's length’ (Dan-Cohen Reference Dan-Cohen1991: 966). We need not suppose that this identification is conscious or voluntary (Dan-Cohen Reference Dan-Cohen1991: 969), nor that we can end up distancing ourselves from each of our actions. But it seems clear that we do identify with some of our actions, and distance ourselves from others. I suggest that we can understand taking an external view in terms of distancing and an internal view in terms of identification, and I suggest that we can understand some forms of distancing in light of Williams's distinction between the view from here, and the view from there (Williams Reference Williams1981a: 35).
A little more about distancing and identification. There are various things that might see us distance ourselves from an action (Dan-Cohen Reference Dan-Cohen1991: sec. IV). Perhaps I was ignorant of what I was doing; this might explain why Maddie, who did not intentionally knock over the vase, takes an external view. Or I was acting in a role, as an executioner who might distance himself from the killings he performs on the job (Dan-Cohen Reference Dan-Cohen1991: 999–1001). Or I was under intense stress (Strawson Reference Strawson and Watson1982: 65–66). In each case, I might not identify with that action: I do not think the action says anything about me. It does not reflect my values or my ability to guide and shape what I do (of course, sometimes an inability to guide what I have done reflects on me, and I identify with it, but these also provide common cases of distancing). Instead, that action is somehow severed from my executive capacities. So, I do not see it as reflecting on me. Still, I might accept that in some sense—but not the sense relating to my executive capacities, my values, my deep self, however we cash it out—I did it. Perhaps it can properly be written into a chronicle of my life:
• The identification account understands the internal view as capturing the actions I identify with, and the external view as capturing actions from which I distance myself.
If I take an external view on my action then I do not identify with it, I do not see it as reflecting on me, and do not see it, in some important sense, as mine. In contrast, when I take an internal view I identify with that action and see it as mine, as an action that reflects on me. This understanding of the external view fits nicely with the idea that when I take an external view on my action, it does not occasion agent-regret: after all, if I do not think it reflects on me, I have no reason to regret the fact that I did it. As Raz puts it, ‘agent-regret relates to one's sense of who one is' (Raz Reference Raz2011: 233). When one takes an external view—where this is understood in terms of distancing—one might admit that the action has some role in a chronicle of one's life, but one does not see oneself as intimately related to that action; thus, it does not relate to one's sense of who one is, and we can see why it precludes agent-regret.
This is only a simplified picture of what it means to identify with something, but I hope the outline of the idea is clear enough, and we can see how taking an external view might be understood as distancing oneself from an action. In fact, Williams says something that supports this interpretation of the external view. In ‘The Actus Reus of Dr. Caligari’, he discusses Cesare, who is hypnotized to murder Dr. Caligari's victims; and here Williams claims that ‘In this state, [Cesare] cannot summon up, for instance, thoughts that would relate the killing to the rest of his life’—it is a ‘dissociation’ that is involved (Williams Reference Williams and Moore2008b: 106). (See also Williams's [Reference Williams2008a: 54] comments on Agamemnon: ‘[H]e is, so to speak, dissociating the action from himself”.) This is an extreme case, but we can see that if one takes an external view (though, Williams does not use this term in ‘Dr. Caligari’), the action does not relate to the rest of one's life: one does not identify with it. Yet the discussion in ‘Dr. Caligari’ is brief, and we surely cannot understand all cases of the external view in terms of hypnosis. There is another aspect of Williams's work that offers a way of understanding distancing by looking at our evaluative viewpoints and projects. The idea here is that sometimes one does not identify with an action because it does not relate to how one is now. This also does not cover all cases of distancing (but it covers more than the hypnosis involved in ‘Dr. Caligari’), but it allows us to see how the external view might relate to some important elements of Williams's work.
Williams says that when we make a choice, our perspective is ‘from here’, but when we assess that choice later, our perspective is ‘from there’, and between here and there we may have changed (Williams Reference Williams1981a: 35). The basic idea is that what I value, and how I evaluate things, changes over time. (For explorations of how a change of viewpoint occurs, and its broader importance in Williams, see Dan-Cohen [Reference Dan-Cohen2008], Lang [forthcoming], and Wallace [Reference Wallace2013].) For Williams, our evaluations are closely tied to who we are, and we can see this in terms of our projects. Williams holds that our projects are vital both to who we are as individuals with particular characters and to our evaluative take on the world (Williams Reference Williams1981b). In ‘Moral Luck’ he focuses on the idea that some projects are so important that whether or not they succeed vastly changes our evaluative viewpoint (Williams Reference Williams1981a: 35–36; Williams Reference Williams1995: 245–46). I will focus on more mundane cases rather than the grand projects of ‘Moral Luck’ and how the idea that our evaluations are tied to who we are can give us a Williamsian take on identification.
Williams says that often our regrets will attach to our decisions and this will not be affected by the perspective we end up with (Williams, Reference Williams1981a, p. 35). But it also strikes me that there will be many cases where our perspective has shifted such that we no longer identify with what we did, without the involvement of a momentous project. The perspective I take up when I make the assessment might be a very different perspective to the one I had when I acted because many of my projects and values have shifted a little. Given that our evaluations partly reflect who we are, the different evaluative viewpoint suggests that I will be a changed person. Arnold might regret that a tree was damaged beyond repair, but he was such a different person—he was only a child—with very different projects and goals. The action does not relate to his projects or ideals as he is now, and these projects and ideas partly make up who he is. Thus, he struggles to identify with the child he once was. And this shows itself in his evaluations: he does not feel agent-regret.
What makes this especially plausible is that regret, and agent-regret, are ‘retrospective’ emotions (Wallace Reference Wallace2013: 17, 182–83): agent-regret involves looking back on what one did, and this opens up space (and time) for one to be a quite different person when one evaluates matters. Although our evaluative viewpoints change, it is not so much the change of viewpoint that explains why we might identify (or not) with an action; rather, it is the associated change in us, the change in our projects and cares that results in that change of viewpoint, that explains whether or not we identify with the action, and this is then echoed in our evaluative viewpoint and affects whether or not our evaluative judgment ends up as ordinary regret, rather than agent-regret.
In Arnold's case, we might think that the distancing is because of how he has changed over time; but we can imagine a much more drastic change over a shorter time, and we can envisage these changes as radical and central to our lives—as Williams focused on—or in the more mundane manner that I have sketched. Of course, even if our projects have changed we can look back on an action and feel shame or agent-regret: we could imagine Arnold taking a negative view on the fact that he, as a child, was not composed enough to avoid the juvenile temptation to carve into the tree. And not every change in project will lead to such distancing. Nor will every case of distancing involve a change of projects—certainly Maddie's case cannot be explained in such a way, and we will have to look at fact that she did not intentionally smash the vase to explain why she might distance herself from it. But by considering a change of projects we can see a Williamsian way of explaining some aspects of identification and distancing that involves a divide between the view from here, and the view from there, which reflects our differing evaluative viewpoints, and the differences in who we are over time.
So, we can make sense of how identification and the external view fit together. Arnold or Maddie might not feel agent-regret because they do not identify with what they did, and so do not regret the fact that they did it.
5.2. Care
I think considering matters in terms of identifying with an action is on the right tracks. But saying that I did it but it was not really me can sound off base. Moreover, it is not clear that Maddie or Arnold need anything so complicated as a view of what is really them. The real explanation might be simpler, and might be more tightly tied to our understanding of agent-regret. So, I offer an alternative account. I now sketch how the external view might relate to what we care about, and link this to agent-regret.
We regret only the things we care about (Raz Reference Raz2011: 235; Wallace Reference Wallace2013: 20–26). What do I mean by ‘care about’? I use this notion fairly roughly, since there are many ways of cashing it out (Scheffler Reference Scheffler, Wallace, Kumar and Freeman2011). But I hope the following brief discussion will ensure my use of the term is clear enough and will bring out some key features of caring.
By ‘care’, I mean something akin to ‘valuing’ or Williams's notion of ‘importance’ (Williams Reference Williams1985: 182–83).Footnote 1 It seems to me that caring about something (and being susceptible to regret) turns around the ‘relative notion of importance’ at play when someone ‘finds a given thing important’ (Williams Reference Williams1985: 182). What we care about will link to our evaluative viewpoints and our projects: what I find important will often depend on my projects and relate to my desires and evaluations more broadly. An important feature of valuing comes out when R. Jay Wallace (Reference Wallace2013: 22) connects regret to ‘the phenomena of valuing and attachment’. On this picture, when we value something we do not just recognize that it is valuable, we ‘take an active interest’ in it (akin to Williams's notion of finding something important); and Wallace also notes that this ‘renders one emotionally vulnerable to how it fares' (Wallace Reference Wallace2013: 23; see Scheffler Reference Scheffler, Wallace, Kumar and Freeman2011). It opens one up to feeling regret or agent-regret. Moreover—and this will be important later—‘to value X involves seeing X as a source of reasons for action’; we want to help the things we value flourish, and to repair them if they suffer (Scheffler Reference Scheffler, Wallace, Kumar and Freeman2011: 31; see Gardner Reference Gardner2018: chap. 5). The idea, then, is that when we care about things, we see them as important and take an interest in them, which leaves us emotionally vulnerable and connects to our reasons for action.
Given the link between regret and care, if we feel agent-regret about something we have done, it follows that we care about the fact that we have done such a thing. We saw earlier that the external view precludes agent-regret. Since agent-regret involves caring about the fact I have done something, I suggest that when I take an external view I do not care about having done such a thing. When I take an external view, it might not be because I do not identify with some action of mine, but—a simpler thought—because I do not care that it was my action. On such an account, we do not need to say that I see it both as my action and not my action. Instead, it is my action, but I do not care that it was my action. I just do not care that I flipped the bird at the speeding driver, that I scored the own-goal, or that I ignored the irritating man at the party.
All we need for this to translate to the external view—which, remember, must make sense of me regretting an action, but not regretting that it was mine—is to see that I can care that someone acted, but not care that it was me who acted. I might care about an action in so far as I think it is bad that it happened, hence I can regret it; so, I can still think there needs to be some payout to help you after my blasting destroys your house. But I do not feel agent-regret because I do not care that it was my blasting activities as opposed to the other person's blasting activities: I see both as just as bad and I am happy enough with an insurance payout settling things for you. Or it might bother me that someone must put the animal out of its misery, but I do not care that I am the one to do it. I regret that someone killed the animal, I do not agent-regret doing it myself. Or I might care for my teammates and opposition, so think it is bad whenever someone else tackles them illegally; but, being a disruptive rugby player whose role is to get under the skin of the opposition, I recognize that sometimes this needs to be done, and am more than happy to get my hands dirty: I do not care if I tackle someone else illegally.
• The care account understands the internal view as capturing the actions I care about performing, and the external view as capturing the actions that I care were performed yet do not care about the fact that I performed them.
Someone who takes an external view on her own action cares that someone so acted, but does not care that she so acted.
Return to the idea that caring about something goes hand in hand with recognizing reasons to repair the thing we care about. Caring goes hand in hand with an emotional reaction: absent some special condition (one is terribly distracted, or emotionally overburdened), one who cares about something will feel something.Footnote 2 The care account might help us get clearer on the relationship between an insurance payout, agent-regret, and the external view.
Someone who takes an external view cares about the harm and wants it to be alleviated. This is like the position of one who feels ordinary regret. As I mentioned earlier, someone who takes an external view might be in a slightly more complicated position than that of an ordinary bystander since she is also willing to pay compensation. Still, this willingness to pay compensation does not suggest any emotional reaction to one's own action. Someone who takes up the external view might recognize that it was her action and that this means she must pay compensation. But her thought might just be that she must pay compensation because it is not fair that the victim has to bear the cost. She might think that in cases like this there is a general rule that whoever caused the damage should not leave the victim to suffer; but if there is an insurance payout, then the victim no longer has to suffer (and a wealthy insurance company bears the costs), so the situation is fair again.
Given the connection between caring and the external view, this suggests that when one takes an external view one recognizes but does not care about one's own role in occasioning the harm. The thoughts that spur a willingness to compensate involve some concern for fairness, and one sees it as fair that one must pay up because one recognizes one's role in bringing about the harm. But one can think that something is fair while thinking about it in a disinterested way. One might make a utility calculation or consider it as a maxim of fairness. One certainly does not need to care about what one has done in order to recognize the fairness of compensation. (One's care about the harm, though, may spur one to provide assistance.) Whereas someone who feels agent-regret directs an emotional reaction—agent-regret—to the fact that she did such a thing. She cares that she occasioned the harm, and so we can explain why she might want to make reparative steps in terms of what she has done: someone who feels agent-regret wants to make up for what she has done because she cares that she did it, and she wants to repair what she cares about, which is her role in the regrettable occurrence—and an insurance payout does not fix this. She wants not just to make things better, but to make up for what she has done. (Relatedly, in Shame and Necessity, Williams [Reference Williams2008a: 70] notes that sometimes an agent's attitude to his own life should be affected by what he did, and he adds that insurance does not reflect this.) Of course, this picture needs further exploration, and I cannot provide it here (see Gardner [Reference Gardner2018] for a compelling account) but we see that there are different explanations of why someone who feels agent-regret, and someone who takes an external view, are willing to pay compensation: one cares about what she has done, the other acts on considerations of fairness and she need not care about what she did to be astute to these considerations.
5.3. Two Satisfying Accounts
It should be clear that often identification and care will be tightly bound together: we tend to care about what we identify with. But they are separate accounts. One sees us as taking an external view when we fail to identify with an action, the other sees us taking an external view when—whether or not this means we do not identify with the action—we just do not care that we performed that action.
Both accounts give us a satisfactory understanding of taking an external view on one's own action:
i. One regrets one's own action. On such accounts, unlike Nagel's, one need not see the action as a mere chain in the causal story, but as a genuine action.
ii. One regrets it in the same way as one regrets anyone else's action. This is because the person does not see it as an action important to his identity, or does not care that he did it. So, we can make sense of how one might see it as if it were anyone else's action.
iii. One might be willing to pay compensation but would be satisfied were an insurance payout to cover the costs instead. For example, I might accept that if the victim will otherwise have to bear the costs, I should pay up, but I am satisfied if an insurance company pays up. I do not identify with what happened and so am more than happy if someone else pays up; or I care that you got hurt, but do not care that I did it, yet recognize that fairness demands that I do not leave you to suffer. Thus, I do not feel the need to do anything to make up for the fact that I did it, but I do see that something might need to be done about the bad thing that happened, so I might be willing to pay compensation but think that an insurance payout settles things just fine.
These accounts also let us make sense of why different people might feel agent-regret about different things: after all, we all care about different things, so we are all susceptible to regret different things. A child who does not care for football (likewise for a child who does not identify as a footballer) might not regret that he scored the own goal in his physical education lesson; a child who does care for football might care when he scores the own goal. We can even make sense of how the same person could take an internal or an external view on something. One can be more or less identified with something (Dan-Cohen Reference Dan-Cohen1991: 965–66; Williams Reference Williams1981a: 28), or care more or less about something, and this can vacillate (Dan-Cohen Reference Dan-Cohen1991: sec. III.B). Whether Maddie takes an internal or external view might depend on what she finds salient at that time. If she is stressed about moving house, she just does not have the mental space to care about smashing her friend's vase; but perhaps she will care if we point out how callous she is being. Or it might be that certain ways of characterizing the event bring out aspects she cares about or identifies with, and others leave her cold. We might care or not—or identify or not—depending on how we think of the action in question, and our situation when we are thinking about it.Footnote 3
Finally, Williams thinks there is a test of whether one can acceptably take up the external view. Does the agent think that insurance money would do just as good a job at making up for that action as would some action on his own part (Williams Reference Williams1981a: 28–29)? If so, he takes up an external view—as I have shown above. But Williams also points out that ‘It is an obvious and welcome consequence of this test that whether an agent can acceptably regard a given case externally is a function not only of his relations to it, but of what sort of case it is' (Williams Reference Williams1981a: 29). Williams then focuses on cases where insurance will not offer much, or any, succor to the victim. But we can regard an external view as unacceptable even if an insurance payout might help the victim, such as in the case where Maddie smashed the vase. My suggestion is that we can criticize someone for taking an external view because taking an external view might show that she does not care about or identify with what they should care about or identify with.
Whether we think someone should care or identify may well change with the situation and may well depend on who the agent is and his or her relation to the case. This gives force to what I say above about how we might criticize Maddie: perhaps what is really objectionable about her taking an external view and being happy enough if the insurance company pays for the vase is that she does not care about what she has done, and we think she should care (whereas we do not think that Arnold should care so much about what he did decades ago); but if we realize that Maddie is horribly stressed by her impending move, we might mitigate or withdraw our criticism, and this is partly because we realize she has more important things to care about. So, we can introduce normative standards concerning what we think people should care about or identify with, and we can see how there is room to criticize people for taking an external view.
6. Conclusion
I have argued that taking an external view depends on not identifying with or caring about one's own performance of an action. We thus can see why one will not feel agent-regret in such a case. Agent-regret involves regretting that I have done something, and when I take up an external view, I have no such regrets because I do not see the fact that I did it as latching on to my identity, or I do not care about the fact that I did it. So, my accounts—and I leave it open whether we should prefer one to the other, though I have suggested above that the care account might be simpler and truer to life—make sense of the external view and its relation to agent-regret. They do so without collapsing Williams's account of an external view into the sort of view taken up in Nagel's mechanistic external view. Furthermore, adopting either of these accounts steers us clear of a tempting error, that of identifying the external view with extrinsic luck.
My aim in this article has been to make plausible, by relating these accounts to other aspects of Williams's work, the idea that this might be what Williams had in mind. In doing this, I have sought to provide a clearer understanding of an admittedly minor part of Williams's work, the external view, which stands to have larger repercussions on more central elements of his work, such as agent-regret and moral luck. My hope is that future explorations of agent-regret will be clearer for it, and better able to capture Williams's rich discussion and his own view.
