The Loop Case and Kamm’s Doctrine of Triple Effect
Forthcoming in Philosophical Studies
© S. MATTHEW LIAO
Faculty of Philosophy, Oxford University, Littlegate House, 16/17 St. Ebbes St., Oxford OX1 1PT, UK; e-mail: email@example.com; www.smatthewliao.com
The Loop Case and Kamm’s Doctrine of Triple Effect
Judith Jarvis Thomson’s Loop Case is particularly significant in normative ethics because it calls into question the validity of the intuitively plausible Doctrine of Double Effect, according to which there is a significant difference between harm that is intended and harm that is merely foreseen and not intended. Recently, Frances Kamm has argued that what she calls the Doctrine of Triple Effect (DTE), which draws a distinction between acting because-of and acting in-order-to, can account for our judgment about the Loop Case. In this paper, I first argue that even if the distinction drawn by DTE can be sustained, it does not seem to apply to the Loop Case. Moreover, I question whether this distinction has any normative significance. The upshot is that I am skeptical that DTE can explain our judgment about the Loop Case.
The Loop Case and Kamm’s Doctrine of Triple Effect
Most nonconsequentialists intuitively believe that it is morally permissible to redirect the trolley in the
Original Trolley Case: A runaway trolley is headed toward five people who will be killed unless something is done. The trolley, however, can be redirected onto another track where an innocent bystander sits, whereby the five would be saved but the one would be killed.
At the same time, most nonconsequentialists intuitively believe that it would be impermissible to do the following:
Push Case: A runaway trolley is headed toward five people who will be killed unless something is done. We can push an innocent bystander in front of the trolley, which would be stopped by hitting the bystander, thereby saving the five but killing the one.
A principle that seems neatly to capture the intuitive judgments in these cases is the
Doctrine of Double of Effect (DDE): There is a moral constraint on intending evil (such as harm), even when the evil will be used as a means to a greater good. However, we may be permitted to employ neutral or good means to promote a greater good, even though we foresee the same evil as a side effect, if (a) the good is proportionate to the evil, and (b) there is no better way to achieve this good.
According to DDE, because we merely foresee the death of the one in Original Trolley but do not intend him to be hit as a means to saving the five, it is permissible for us to redirect the trolley. But in Push, because we intend the one be hit by the trolley as a means to stopping it from hitting the five, it is not permissible for us to push him into the trolley’s path.
Judith Jarvis Thomson has famously challenged DDE on the basis of the
As Thomson explains, in
More recently, Frances Kamm has argued that one can modify DDE so as to
Party Case: We intend to throw a party in order to have fun. We foresee though that this will result in a big mess, and we will not have a party if we will be left to clean up that mess by ourselves. However, we foresee that if we throw the party, our friends will feel indebted to us and this will cause them to help clean up. Hence, we have the party because we believe that our friends will feel indebted and because we will not have a mess to clean up. However, we do not give the party in order to make our friends feel indebted or to clean up for us (p. 103).
Using the because/in-order-to
distinction, Kamm argues that in
In this paper, I
question whether DTE applies to
Kamm’s analysis of
The second step is her introduction of the because/in-order-to distinction, using Party as an example. On the basis of this distinction, Kamm also rejects a common test for the presence of an intention, known as the
Counterfactual Test: If we decline to perform an act (e.g. redirect the trolley) because the act would not bring about the effect (e.g. hitting a bystander), this means that we intended to produce the effect.
According to Kamm, if we do an act
(e.g. give a party) because we believe that it will have a certain effect (e.g.
inducing feelings of indebtedness), then we would not do that act if it would
not have that effect. We can thus satisfy the Counterfactual Test without
intending the effect, which shows the test to be inadequate. If this is
right, Kamm argues, this test cannot be used to show that we intend to hit the
In the third step
of her analysis of
1) if we intend to bring about X, then we seek means to accomplish the end of bringing it about.
2) if we intend to bring about X, then we pursue X. This means that if one way fails to produce X, we would adopt another way.
3) if we intend to bring about X, and our intentions should be consistent in so far as we are rational, then we will filter out intentions that conflict with intending X.
Kamm argues that we do not meet
these conditions in
Moreover, we fail to meet 2), according to Kamm, because in
Extra Push Case: Suppose the redirected trolley would jump over the one person on the track and loop back toward the five. We need not be committed to giving the trolley an extra push, which would be unnecessary to get the trolley away from hitting the five from the front, but which would be necessary to get the trolley to hit the one person, supposing that hitting the one is what is causally necessary to prevent the trolley’s jumping over him and looping back toward the five (p. 97).
Kamm argues that if we give the
extra push in Extra Push, then we would be doing something that was undertaken
especially to accomplish hitting the one, which would clearly involve intending
to hit the one. However, so she argues, in
fail to meet 3), Kamm argues, since “it is consistent for someone who redirects
the trolley because he believes that it will hit the one to then try to rescue
the one from being hit” (p. 98). If so, and if intending to rescue the
one is inconsistent with intending to hit the one, then, according to Kamm,
this would show that we did not intend to hit the one. To motivate this
idea, Kamm proposes that when we redirect the trolley in
We will redirect the trolley and then try to push the one out of the way. For if we have an opportunity to save the one and do not do so, we would be intending his being hit. If we fail to rescue the one, we still get the advantage of the five’s being saved because the one will be hit. If we succeed in rescuing the one, the five are no worse off and they had some chance of being saved (p. 98).
According to Kamm, if these are our
In the fourth step of her analysis of
By analogy, in
While the distinction between
acting because-of and in-order-to may be clear in some contexts, it is harder
to draw in
Kamm tries to make
the same distinction in Loop by claiming that the agent’s initial end in
diverting the trolley is to stop it from hitting the five from the front,
an objective that would be defeated – deprived of its value –
if the trolley hits the five from the back, as it would do
in Loop without the convenient presence of the one. But unlike Party in which one can tell a
plausible story in which the friends’ indebtedness is not a means to the
agent’s end to have fun, it simply seems implausible that the agent’s end is
“preventing the trolley from hitting the five from the front”. The
In addition, Kamm’s appeal to Bratman’s conditions for intention is problematic. In particular, Bratman’s first two conditions, as Kamm interprets them, are too strong, excluding some clear-cut cases of intending.
Let’s begin with the first condition: if we intend to bring about X, then we seek means to accomplish the end of bringing it about. Consider the following:
Mortal Enemy Case: At a party, we notice by chance that we could kill our mortal enemy, who has asked us to pass him a tray of nuts. Unbeknownst to him, the tray contains nuts to which he is deathly allergic. We pass the tray at his request. Previously, we had no intention of killing our enemy.
Bratman’s second condition, as Kamm interprets it, is also too strong. We can intend to bring about X without its being the case that we would adopt other ways of bringing about X should one way fail to produce it. Consider
Mortal Enemy Case II: Everything is the same as Mortal Enemy. Further, suppose that if our enemy decides not to eat the nuts, for whatever reason, we would not try to kill him by other means.
It should be clear that in Mortal Enemy
II, we still intend to kill our enemy even though we are not committed to
killing him by other means if he decides not to eat the nuts. If so, Mortal
Enemy II shows that one can intend to bring about X without being committed to
trying again – without satisfying Bratman’s second condition. Just as we can
intend to kill our enemy without surveying possible means or being committed
trying again, we can intend to hit the one in
Moreover, even if the because-of/in-order-to
distinction, and the corresponding distinction between primary and secondary
reasons, could be applied to
Even in Party, where the because-of/in-order-to distinction is easier to draw and where our friends would benefit from the party, this seems to be the case. To be sure, our end in Party is in part the fun our friends will share, and we believe their fun will outweigh their feelings of indebtedness and their burden of cleaning up. Nevertheless, in relying on their feelings and efforts for the success of our plan, we seem to be manipulating them, to take advantage of their sense of social propriety, even if we do so in part for their good. That is, the success of our plan seems to depend on knowingly doing something to induce feelings of indebtedness in our guests, something that we perhaps as good hosts, and certainly as friends, should not do. Indeed, it would seem more respectful if we had asked them explicitly to help with the clean-up.
To make the problem more vivid, consider
Dinner Case: We want to have dinner with a friend at a nice restaurant in order to have a good time with our friend. We foresee, however, that dinner will be very expensive, and we would not arrange it if we expect to have to pay. But we also foresee that our friend, who is extremely wealthy, will be so happy to be invited that he will insist on paying.
Similar to Party, we have the dinner at the restaurant because we believe that our friend will pay, but we do not arrange the dinner in order to get him to pay. Our primary reason for the invitation is not to cause our friend to lose money, but to have a good time with our friend. We nevertheless seem to manipulate our friend, to take advantage of his wealth. It might have been more respectful to him, if perhaps less likely to succeed, to be upfront about the expected arrangements; to invite him to take us out to dinner.
In response to the idea that we are taking advantage of our friends in Party, Kamm explains that our guests would make us worse off than we would be without the party (as the mess they produce overrides the fun for us): “Hence, when they serve [our] interests, they merely eliminate the threat of the mess they themselves present to [us]. This is unlike improving [our] position from what it would be without their presence” (p. 103). But arguably, this misses the point of friendship and further suggests that in Party, we are in fact taking advantage of our friends. In friendship, we do not just aim to break even, so to speak. We often undertake additional burdens for the sake of the friendship.
Kamm also argues that we do not treat the one as “available” for our purposes, because we may refuse to do anything else to him in support of our primary or secondary reasons for acting: “we need not be willing to give the extra push in the Extra Push Case” (p. 103). But our unwillingness to give the extra push in Extra Push does not show that we are not taking advantage of our friend. Consider
Dinner Case II: Everything is the same as Dinner, except that we have agreed that we will take the bill when it comes, and will not do anything to induce our friend to pay it, such as remarking on its size. Moreover, if our friend insists on paying, we will initially refuse, giving him ample opportunity to change his mind.
Just because we are willing to take a risk of getting stuck with the bill does not mean that we do not regard our friend as a means paying the bill (that is, as a walking check book). While it is not clear whether a person can be treated as a mere means by acts whose primary reason would also benefit the person in some sense, our manipulation of our friend still seems disrespectful in these circumstances.
We can make analogous points regarding
Here it is worth making a note about the debate between Warren Quinn and Kamm,
on which the previous discussion is based. Quinn has distinguished between harmful
direct agency and harmful indirect agency. The former involves treating
other people as available for our purpose in a way that is harmful to them
(without their consent); whereas the latter does not involve such intention. Kamm has argued that
Kamm has argued that DTE, that is, the
distinction between acting because-of and in-order-to, can explain the
permissibility of redirecting the trolley toward the one in
 See, e.g., Thompson, J. J. "The Trolley Problem." The Yale Law Journal 94 (1985): 1395-415. It might be worthwhile stipulating that if the five were not present, the trolley would not go around and hit the one, but would carry on harmless down the track.
 Thompson, "The Trolley Problem," p. 1403.
 Thompson, "The Trolley Problem," p. 1403.
See, e.g., Kamm, F. M. Intricate Ethics: Rights, Responsibilities, and
 Unless otherwise specified, all references to Kamm refer to Intricate Ethics.
For some other recent discussions of
 Kamm says that the primary/secondary reason distinction is “another way of understanding the cases I discussed” (p. 102).
 It might be worth mentioning that it is hardly a conceptual truth that one could not throw a party in order to make someone else feel indebted. For example, consider the
Election Party Case: Suppose you are running to be elected to the board of your local country club. You could throw a party in order that your friends feel indebted and would thereby vote for you in the upcoming election.
Kamm addresses the question of why the problems should be differentiated this
way by treating it as a question of whether different threats must have
different objects doing the threatening. She responds that
Wagon: The trolley is headed toward killing the five. We divert it and, as it goes on the side track, it stops. But we know that it will stop on a button whose depression will cause a wagon on the side track to be set in motion. The wagon will head around the loop toward killing the five, but it will be stopped by hitting one person on the side track (pp. 94-95).
But the similarity of Loop and Wagon for moral
purposes, which let us assume for the sake of argument, may cut either way: It
may be taken to show, as Kamm intends, that a second threatening object isn’t
required for there to be a second threat, since Loop is morally like Wagon. But
it may also be taken to show that a second threatening object need not create a
second threat, since Wagon is morally like
 I would like to thank an anonymous reviewer for suggesting the following point.
 Quinn, W. "Actions, Intentions, and Consequences: The Doctrine of Double Effect." Philosophy and Public Affairs 18, no. 4 (1989): 334-51, p. 343.
 Quinn, W. "Actions, Intentions, and Consequences: The Doctrine of Doing and Allowing." Philosophical Review 98, no. 3 (1989): 287-312, p. 288.
 I would like to thank David Wasserman, Mike Otsuka, Wibke Gruetjen, an anonymous reviewer at Philosophical Studies for their very helpful comments on earlier versions of this paper. Thanks are also due to participants of the Kamm Reading Group at www.ethics-etc.com for very fruitful discussions regarding this topic.