Playing to win in order to lose

Let’s say I have a friend who needs cheering up as she has had a lot of things not go her way. I know that she is definitely a better badminton player than I. So I propose a badminton match. My goal in doing so is to have her win the game, so as to cheer her up. But when I play, I will of course be playing to win. She may notice if I am not, plus in any case her victory will be the more satisfying the better my performance.

What is going on rationally? I am trying to win in order that she may win a closely contested game. In other words, I am pursuing two logically incompatible goals in the same course of action. Yet the story makes perfect rational sense: I achieve one end by pursuing an incompatible end.

The case is interesting in multiple ways. It is a direct counterexample to the plausible thesis that it is not rational to be simultaneously pursuing each of two logically incompatible goals. It’s not the only counterexample to that thesis. A perhaps more straightforward one is where you are pursuing a disjunction between two incompatible goods, and some actions are rationally justified by being means to each good. (E.g., imagine a more straightforward case where you reason: If I win, that’ll cheer me up, and if she wins, that’ll cheer her up, so either way someone gets cheered up, so let’s play.)

The case very vividly illustrates the distinction between:

1. Instrumentally pursuing a goal, and

2. Pursuing an instrumental goal.

My pursuit of victory is instrumental to cheering up my friend, but victory is not itself instrumental to my further goals. On the contrary, victory would be incompatible with my further goal. Again, this is not the only case like that. A case I’ve discussed multiple times is of follow-through in racquet sports, where after hitting the ball or shuttle, you intentionally continue moving the racquet, because the hit will be smoother if you intend to follow-through even though the continuation of movement has no physical effect on the ball or shuttle. You are instrumentally pursuing follow-through, but the follow-through is not instrumental.

Similarly, the case also shows that it is false that every end you have you either pursue for its own sake or it is your means to something else. For neither are you pursuing victory for its own sake nor is victory a means to something else—though your pursuit of victory is a means to something else.

Given the above remarks, here is an interesting ethics question. Is it permissible to pursue the death of an innocent person in order to save that innocent person’s life? The cases are, of course, going to be weird. For instance, your best friend Alice is a master fencer, and has been unjustly sentenced to death by a tyrant. The tyrant gives you one chance to save her life: you can fence Alice for ten minutes, with you having a sharpened sword and her having a foil with a safety tip, and you must sincerely try to kill her—the tyrant can tell if you are not trying to kill. If she survives the ten minutes, she goes free. If you fence Alice, the structure of your intention is just as in my badminton case: You are trying to kill Alice in order to save her life. Alice’s death would be pursued by you, but her death is not a means nor something pursued for its own sake.

If the story is set up as above, I think the answer is that, sadly, it is wrong for you to try to kill Alice, even though that is the only way to save her life.

All that said, I still wonder a bit. In the badminton case, are you really striving for victory? Or are you striving to act as if you were striving for victory? Maybe that is the better way to describe the case. If so, then this may be a counterexample to my main thesis here.

In any case, if there is a good chance the tyrant can’t tell the difference between your trying to kill Alice and your intentionally performing the same motions that you would be performing if you were trying to kill Alice, it seems to me that it might be permissible to do the latter. This puts a lot of pressure on some thoughts about the closeness problem for Double Effect. For it seems pretty plausible to me that it would be wrong for you to intentionally perform the same motions that you would be performing if you were trying to kill Alice in order to save people other than Alice.

Pursuit of victory

I wonder if it's rational to choose to play a game precisely in order to win. Winning is a practice-internal goal. Does this goal make sense when one isn't already decided on playing the game? Of course people do choose to play in order to win. But I wonder if those aren't cases where the victory is a means to something else, like money, fame or satisfaction?

The value of victory

Is winning a game always worthwhile? Consider this solitary game: I guess a number, and if my number is different from the number of particles in the universe, then my score equals the number of particles in the universe. I can play this over and over, winning each time. If it's good for me to win at a game, I continue to rack up benefits. So for reasons of self-interest, I should play this game all the time. I could even set myself up as playing it by default: I announce that each time I breathe, the length of my inspiration in milliseconds counts as my guess. I will be racking up benefits every day, every night. But that's silly.

Pursuing a goal because it's good to pursue the goal

A standard case of goal-directed activity is where I pursue a goal because the goal is worth achieving either intrinsically or in light of further ends. But not all cases of goal-directed activity are like that. Imagine (there may well be something like it) a parking system for a car that uses a variety of sensors to generate on a screen an animated 2D overhead view of the car, nearby obstacles and the parking space, so that you can park the car simply by looking at the screen. After I got used to the system, instead of thinking about moving my car, I would be thinking of moving the little animated car on the screen, much as in a video game. I would thus engage in end-directed activity whose goal would be that the little animated car move into its on-screen parking spot. As a result of my engaging in this activity, the real car would move into the real parking spot. But note that the on-screen movement of the animated car isn't a means to the movement of the real car. Rather, it is my pursuit of the goal of the on-screen car moving into the on-screen spot that accomplishes the movement of the real car, and it is the movement of the real car, not the on-screen movement, that has the relevant value. Moreover, in this system I would accomplish the movement of the real car more effectively by not thinking about the real car, and only focusing my goal-directed reasoning on the on-screen car.

This is a fancy example, but whenever we use computers such things happen. For instance, when you send an email on a non-touchscreen computer, you think about how to click an on-screen "Send" button. Your means to that is to move a pointer on the screen to the button (say, with a mouse) and to click. So you have as your goal the movement of a pointer on the screen to a particular rectangular area on the screen. But in fact the movement of a pointer on the screen to a rectangular area on the screen does nothing to accomplish the sending of the email. The actual means to the sending of the email is the changing of a pair of behind-the-scenes coordinate variables to values in the intervals corresponding to the coordinates of the "Send" button, followed by the pressing of the mouse button. But you don't think about the coordinate variables. You may not even not know that that's how the system works. You think about moving the little arrow on the screen. But the little arrow is only a helpful visualization of the two coordinates. If the screen turned off, or the software stopped updating the displayed arrow, but the underlying coordinate variables continued to track the mouse position, you'd still send the email (but it would be hard to aim). In this case, people accomplish the goal of sending the email by aiming to move the on-screen arrow rather than by aiming to change the underlying coordinate variables. The pursuit of the goal of moving the arrow helps you send the email, but the fulfillment of that goal does not.

So we now have three kinds of goal-directed activity. In the first sort, the goal is pursued for its own sake. In the second, the goal is instrumentally valuable for the sake of something else. In the third, the one I want to think about, what is instrumentally valuable is not the goal that is being pursued but my pursuit of that goal. For it is the pursuit of that goal, rather than that goal itself, that promotes my further end.

The third case is actually a pretty common phenomenon. In the two cases I gave above, the way this worked was that the achievement of a goal that had no relevant instrumental or ultimate value (the movement of an animated car or a pointer) was correlated with the achievement of another goal that was valuable instrumentally or not. Another kind of case is where the the focus is not so much on the achievement on the first goal, but where the focus is on the pursuit of it. Games develop all sorts of human excellences. Some of these excellences are developed precisely through the pursuit of victory. Striving to win a race or climb to the end of a route provides one with healthy physical exercise, develops some aspects of strength of will, etc. In these cases, it isn't so much the achievement of victory that is correlated with the valuable things, as it is the striving for victory that gives rise to the valuable things.

Because of these considerations, it valuable for us to be able to set goals for ourselves, goals that are not otherwise valuable. The reasons why it is valuable that I have discussed so far are based in our cognitive and moral limitations. We can better focus on parking or clicking if we just think of moving the on-screen car or pointer. We are better motivated to exercise body or mind by pursuing victory (understood broadly to include non-competitive cases, like climbing to the end of a route or doing a jigsaw puzzle). I don't think all the goals achieved in this way, however, arise out of our limitations. For in the case of games, there are aesthetic goods that are achieved by the honorable pursuit of victory. (Note: some games have rules of honor that go beyond the rules adherence to which is logically necessary for one to count as having won rather than cheated.) The goods may be achieved whether or not one achieves victory, though they tend to be fuller when one does achieve victory. But, again, one may well be more effective--this may differ from game to game--at achieving these aesthetic goods when one's means-end reasoning isn't actually aiming at them, when one is aiming at victory (subject to side-constraints of honor, if applicable).

Are these cases an exception to the idea that we always act for the sake of the good? In one sense they're not, since one pursues the goal because pursuit of the goal is valuable in some way. But these cases seem to be an exception to the idea that when we engage in goal-directed action, the goal must be instrumentally or intrinsically valuable. Well, maybe not. Maybe the way this works is as follows. You see that it would be good for you to pursue a certain innately largely valueless goal (say, moving a pointer to a rectangle on the screen, or climbing over a sequence of holds marked with pink tape). Because it would be good to pursue the goal, you use your normative power to adopt goals, and when you use that normative power on the goal, fulfillment of it becomes valuable (it becomes, perhaps, a constitutive part of the basic human good of achievement), in something like the way that when you use your normative power to make promises, actions incompatible with what is promised become disavaluable.

Nonetheless, however, the focus with which you strive to fulfill the goal may legitimately be greater than would be explained by the good of achievement alone. For instance, suppose I need to park a car correctly in order to save a life (I am parking the dictator's car, and any scratch means death), and I have the parking system I described at the beginning of the post. To park well, I need to avoid distractions and focus on moving the little on-screen car into its on-screen parking spot. But the focus with which I should pursue this may exceed the focus that is legitimate simply for an instance of the good of achievement. I'm doing this to save a life after all, and so I should reject distractions from more minor goods which normally I should take into account. But I think when I reject distractions, I need to include among my reasons for rejecting the distractions not just the good of achievement, but the good of saving a life by parking correctly.