We are pleased to present our first Politics Philosophy & Economics discussion of AJ Julius's paper 'The Possibility of Exchange'.  Julius's paper is available on open access here.

Victor Tadros, professor at the School of Law, the University of Warwick, is kicking off the discussion with a critical précis. Here now is Victor Tadros:  

Comments on A J Julius: The Possibility of Exchange

AJ’s terrific paper is concerned with coercion and exchange. Here is the problem. Consider:

Inducement. I get you to x by indicating that if you x I will y.

Sometimes this is wrong: I get you to cut off your finger by threatening your life. Sometimes it is permissible. I get you to give me a newspaper by promising to pay for it. How to distinguish these cases? Sensible sounding proposals fail. For example, consider: an instance of Inducement is wrong iff it is wrong for you to x or for me to y. As is familiar from blackmail cases, this fails.

AJ draws on an intrapersonal principle

(2) I should not (do y, intend by ying to bring it about that I do x, and fail to believe with warrant that, for some reasons independent of me, my ying facilitates by [doing x because I take R as giving me sufficient reason to x].

From it, he endorses an interpersonal principle:

(3) The Independence Principle. I should not (do y, intend by ying to bring it about that you do x, and fail to believe with warrant that, for some reasons R independent of me, my ying facilitates [doing x because you take R as giving you sufficient reason to x].

AJ summarizes these principles together as follows: ‘they require of a person that she not aim to produce an action, whether her own or the action of another person, except by helping the person who performs the action to do it for a reason that favours the action independently of what is done to produce it.’ (363)

Amongst other things, AJ then considers, and rightly rejects, attempts to explain the difference between coercion and legitimate exchange on the basis of the obligation to reward others for services rendered. It is sometimes permissible for me to induce D to v by offering i even if there would be no obligation for me to give D i if he vs. Our reasons to form certain intentions or to make certain promises will also not do the work: the reasons that we have to form intentions and make promises in cases of exchange are normally the reasons that we have to execute our intentions and carry out our promises in a way that makes the exchange fair and valuable to both parties.

AJ condemns certain exchanges on the basis that they are inconsistent with The Independence Principle. However, he argues that there are exchanges that are consistent with this principle. These are exchanges that can be made by co-operators. He suggests that there are sometimes reasons for people to act in virtue of the fact that their actions would be part of a co-operating set of actions.

Consider his example:

Couch: You and I have sufficient reason to do (you lift the south end of the couch, I lift the north end).

In this case: it is valuable that both ends of the couch are elevated. It is valueless that my end is elevated unless yours is. It is valueless that your end is elevated unless mine is.

As I cannot elevate both ends on my own, I lack a reason to elevate my end unless you elevate yours and vice versa. We have a reason to cooperate.

Now compare:

Cooperation: You and I have sufficient reason to do (you fix my door, I make your shoes).

Though he does not spell this out, I think that AJ thinks that the value of this pair of actions is not only in the door-fixing and the shoe-making, but also in the fact that in performing this pair of actions we help each other. This seems plausible.

We are then in a position to determine which ‘inducements’ are permitted and which are not. Those that are permitted are those that contribute to cooperative patterns of activity that people have reason to participate in. Wrongful exploitation occurs when inducements cause people to act, but there is no such pattern of activity.

I was not completely clear about AJ’s principles to determine which patterns of activity can be jointly recommended to participants as suitably co-operative. His example involves a case where part of a person’s labour – the morning’s work – is used simply to benefit an already wealthy person. There is no good reason, he thinks, for a person to work the morning when the morning’s work will all but be wasted on the rich person. Hence, complete course of action cannot be recommended to that person consistent with The Independence Principle. It would help, though, if we had more complete criteria to determine when actions can be jointly recommended as suitably co-operative.

Two criticisms:


If I understand them, AJ’s intrapersonal and interpersonal principles seem invalid.

To see the invalidity of The Independence Principle consider:

Coercion: D threatens to shoot my child unless I induce P to v by offering i. If I offer i, D will not carry out his threat whether or not P vs.

Typically, I should induce P to v, even if there is no reason for P to v and no reason for P to want i (wanting i is irrational), either independently of, or after, I have offered i.

Perhaps The Independence Principle would be better if ‘I should not…’ was replaced by ‘It is pro tanto wrong to…’

 If this is right AJ could do more to explain the pro tanto wrongness of inducing a person to do what they lack a reason to do. We might think: there are many things that I do that I have no reason to do. When these things are not disadvantageous to me in any way, I have little reason to care that I do them. If other people benefit by inducing me to do these things, why should I object?

AJ might claim: because I use P for my sake. If this is right, AJ’s view might be a variation on the principle that it is wrong for one person to use another as a means to his good. This principle is normally understood as a principle concerned with coercion. AJ might suggest that he means to extend it to cases of non-cooperative inducement.


It seems problematic to distinguish wrongful from legitimate exchanges on the basis of what we have reason to do together. Here’s why. It can sometimes be coercive to induce a person to do what they have sufficient reason to do. And it can sometimes be unproblematic to transact with a person who does what they have no reason to do.

For the first case, consider:

Marriage: if you and I get married, we will live happily ever after. Irrationally, you don’t want to marry me. A kidney would significantly improve your quality of life. I have a ‘spare’ kidney.

Suppose that it is supererogatory for me to give you a kidney. If I offer you a kidney in exchange for you marrying me, we do what we have good reason to do together. There is good reason for me to give my kidney to you independently of us getting married. There is good reason for you to marry me independently of getting a kidney. And there is good reason for us to do (we get married and I give you a kidney). Yet it nevertheless seems wrong for me to offer you a kidney in return for marriage.

Similarly, consider:

Save: You can save my life at the cost of your leg. You refuse to do this. I have some medicine that belongs to you that you need for your survival. I refuse to give you the medicine unless you save my life. You do so.

Let us assume, as is plausible, that a person can sometimes have a decisive reason to v even if ving is supererogatory. This might be true of you. You might have a decisive reason to save my life at the cost of a leg. Let us suppose you do. If so:

You and I have reason to (You save my life at the cost of a leg and I give you your medicine).

Yet it is wrong for me to threaten you with withholding the medicine, even though by doing so I get you to do something that you have decisive reason to do anyway. AJ’s principle seems to permit this. Perhaps AJ might argue that these acts are not appropriately linked, and hence this is not a case of cooperation. That, though, needs explaining. It is not obvious what the difference is between this case and AJ’s cobbler/carpentry case.

Furthermore, if these acts are not appropriately linked, AJ’s principles seem too restrictive. It would be permissible for me to do threaten you if you were required to save my life. If saving my life would cost you only your little toe, you might be required to save my life, and your duty might be enforceable. In that case it is permissible for me to withhold your medicine. It is not clear from AJ’s principle how to distinguish these cases. The acts in this case seem no more linked than the acts in the previous case. The problem, I think, is that AJ’s principles do not distinguish between what we have reason to do and what we are required to do.

I also think that it is sometimes permissible to cooperate with a person to do something that has no benefit for them at all. Consider:

False God: You are rich. I am poor. You ask me to build a shrine to your false God and offer me a handsome reward in return.

You and I have no reason to (I build a shrine and you reward me handsomely).

Yet it does not seem wrong for me to build your shrine. Furthermore, AJ’s view seems to imply that the wrong that is done in False God is a wrong that you do to me – you get me to do something that I have no independent reason to do. But if anything, it is I who exploit your false belief. I think this permissible, but if there is wrongdoing, it is mine not yours.

6 Replies to “Politics Philosophy & Economics Discussions at PEA Soup: AJ Julius’s ‘The Possibility of Exchange’ with précis by Victor Tadros

  1. Comment on “The Possibility of Exchange”
    Richard Arneson
    “The Possibility of Exchange” is a marvelously tricky, subtle, original, and analytically careful contribution to the discussion of its topic. I admire the essay and its author. But I don’t buy any of it.
    My dissent starts at the beginning, with the independence principle that is the foundation of what follows. Recall that independence requires that “I should not (do y, intending by y’ing to bring it about that you do x, and fail to believe with warrant that, for some reasons R independent of me, my y’ing facilitates your [doing x because you take R as giving you sufficient reason to x]).”
    I reject independence, because I believe it can be morally permissible to induce you to do something that I foresee you will do for what you take to be an insufficient reason, provided that I correctly believe that in fact there is sufficient reason R independent of me for you to do that thing. For example, I offer to give you some ice cream if you give me a dollar, even though I correctly anticipate that you will be induced to give me a dollar because you are irrationally tempted by the ice cream offer. Overvaluing immediate gratification, you will act to get the ice cream even though, by your lights, you should not do so. But my ice cream is good stuff. It’s very tasty, so tasty that there is in fact good and sufficient reason for you to bring about the state of affairs in which you give up a dollar and get some of my ice cream. Since this is a permissible offer on my part and violates independence, independence is unacceptable and should be rejected.
    Suppose that we are both rational agents, in the sense that we both never act except with sufficient reason, and suppose further that the fact that we are rational agents is common knowledge between us. In these circumstances, is independence an acceptable constraint on permissible offers and exchanges? My response is a tentative No. My hesitation is due to my being unsure it means to say that reasons R are “independent of me.”
    I have more confidence that rational and morally permissible exchange is possible, whether or not independence is satisfied, and can take place without engaging in joint intention and joint action. I don’t say joint intention and joint action as characterized by Julius do not exist; I set that issue aside and take no stand on it. What I am sure about is that rational and moral exchange is possible without joint intention and joint action coming into play. At least, rational and moral exchange is possible provided that each person has a Scheffler prerogative sometimes to pursue her own personal projects and ends up to some point even though there are alternative actions she could instead perform that would produce greater good.
    Here’s an acceptable exchange, I submit. I promise to give you a back rub if you give me a dollar. Understanding the offer I have made, you give me a dollar. I give you a back rub. In this exchange one party performs first, the other second.
    In claiming this to be an acceptable exchange, I am importing some familiar background assumptions as context. These include: (1) Neither you nor I have sufficient reason to make a free gift to the other of the good that is given in exchange. (2) Each of us is better off, if the trade of a back rub for a dollar occurs, than if it does not. (3) Neither of us has a (substantially) better offer readily available, which our exchange would preempt. (4) If you unilaterally give me a dollar, in the absence of any mutual agreement to make an exchange, your giving me a dollar does not render it the case that I then have a reason to give you a back rub. (5) My promising you that if you give me a dollar I will give you a back rub provides you assurance that if you give me a dollar I will give you a back rub. (6) Promising in circumstances like ours can provide this assurance via the conventional understanding that if one person makes a conditional promise, and the condition is fulfilled, the person making the promise is then obligated to do what he has promised to do, and absent special intervening circumstances has sufficient reason to do what he has promised to do. (7) With the promise-given assurance that if you give me a dollar I will give you a back rub, you have sufficient reason to give me a dollar. Your giving me a dollar fulfills the condition stated and triggers my having an obligation to you to give you a back rub. (8) When you give me a dollar, all of background assumptions (1)-(7) are mutual knowledge between us. (8) My making the offer of exchange to you in these circumstances conveys to you the information that I believe that (1), (3), and (4) are true and that I believe that I would benefit from the exchange proposed and more tentatively believe that you would also benefit from this same exchange. In some circumstances you might have reason to doubt that these beliefs of mine are true. In these circumstances the exchange will not go forward if you are reasonable. But sometimes all things considered you do have reason to accept my beliefs as true and to trust that I will do what I promise. In this case my issuing my offer via a promise renders it the case that (5) is true and also conveys the information that I believe that (7) is true or might well be true. Your then giving me a dollar both indicates that you understand the promise I have made and are acting with the intention of triggering in me an obligation to give you a back rub, carry through my part of the deal. In addition to all this, we need to assume that the terms we offer in exchange are sufficiently reasonable on both sides so that a reasonable standing disposition on both our parts not to be unfairly treated in voluntary exchange is not triggered in either one of us.
    Are Julius’s conditions on rational exchange satisfied? I don’t think so. But anyway I am far more confident that under the stated assumptions and conditions, the exchange described here is rational on the part of both participants, and need involve no joint action or joint commitment, than I am confident that Julius’s proposed conditions on rational exchange really are necessary for rational exchange.
    In the simple example, I have sufficient reason to issue my promise, because there is a good chance that you will then have sufficient reason to accept my promise to our mutual benefit, and no chance that if I forego issuing the promise, rational exchange between us will occur. And once I have issued my promise, if all goes well, you then have sufficient reason to give me a dollar, and once that occurs, I have sufficient reason to give you a back rub. The promise I issue is not (or need not be) a morally defective promise: I have reason to believe that my issuing the promise improves your options and that if you accept my promise and I then fulfill it, you will thereby be better off. How does the independence principle judge this transaction? Maybe negatively. The reason you have to give me a dollar, when you act on the assurance of my promise, is that by giving me the dollar you will bring it about that I am obligated to do my part of the deal and that I will do my part of the deal. This reason you have is not independent of me. Recall, Julius stipulates, “a reason to x is independent of A if the fact that this reason counts in favor of x’ing does not depend for its being a fact on attitudes and actions in virtue of which some person A counts as trying to bring it about that x is done.” In the example the fact that your reason to give me a dollar counts in favor of that action depends for its being a fact on actions and attitudes of mine in virtue of which I count as trying to bring it about that you give me a dollar. The reason you have to give me a dollar depends on the assurance my promise to you has provided, and this being so depends on my having acted to try to bring it about that you give me a dollar. (At least, I think this is so. I am not sure I understand this notion of a reason’s being independent of a person.) But so what? My conviction that the exchange described here is morally acceptable survives my coming to believe that the exchange violates the independence principle, so we have reason to reject the independence principle.
    Not all versions of the simple exchange example do violate the independence principle. Consider a variant in which I am indifferent as to whether you accept my offer of exchange and indifferent as to whether the exchange occurs. My aim might be to increase your available options, or to see whether you value my giving you a back rub more than your retaining your dollar, or some mix of these two, or something else altogether. In the variant in which I am trying to increase your available options, when you on the basis of your understanding of my promise give me a dollar with the aim of triggering my obligation to you to give you a back rub and thus of bringing it about that I give you a back rub, the fact that your reason to give me a dollar counts in favor of doing that does not depend for its being a fact on attitudes and actions of mine in virtue of which I count as trying to bring it about that you give me a dollar. In this example I am not trying to bring it about that you do one thing rather than another, I am trying to expand your option set. But varying this detail of the example so that my participation in the exchange does not violate the independence principle does not strike me as marking the difference between my participation’s being morally acceptable or not. The independence principle looks to be an idle wheel whose spinning does not affect our reasonable moral judgments.
    Julius suggests that we do need to accept the independence principle or something very like it. We need to accept the independence principle in order to explain and justify our convictions regarding the wrongness of coercion. I doubt this is so. He considers the difference between two different ways of depriving you of the benefit of a seat on the bus: sitting down on the seat before you occupy it, and threatening to spit on you if you sit down on that seat. Both actions bring about the loss to you of the same benefit, but only the threatening action is coercive and wrong, so the wrongness of coercion cannot consist in depriving another person of benefit. In reply, there is a generally fair system in place of allocating scarce seats on buses. The person who gets to the seat first gets to occupy it. My threatening you in these circumstances violates this generally fair system of distribution of benefits and burdens, and is presumptively wrong. In a case in which the system would be generating unfair distribution of benefits and burdens, our moral judgments shift. If you are young and agile, and I am old and clumsy, the generally fair rule might be generating unfair results. Same goes if you have two good legs and I have only one leg, and can stand up during the bus ride only with difficulty and pain. We do not need to appeal to anything like the independence principle to explain the wrongness of coercion when coercion is wrong.
    No doubt there is always a pro tanto reason to try to bring it about that other people whose choices might be affected by our actions act with sufficient reason. It is always a good thing, so far as it goes, that a rational agent acts rationally. This mild generality should not be rigidified into a moral constraint on choice of action along the lines of the independence principle.
    In passing, I note that Julius puts further conditions on the type of exchange he is trying to characterize, such that my discussion might appear misdirected. This appearance is misleading.
    Julius say that in the kind of exchange he is trying to characterize, each person has decisive, not merely sufficient, reasons, to do what is done that brings about successful exchange. The fact that both agents involved always do what they have decisive reason to do is common knowledge between the parties involved. And the situation is symmetrical: both have decisive reasons to do their part in bringing about successful exchange. He then seeks to show that “if exchange partly consists in the one person’s promising to [do her task if the other does her task],” these assumptions could not all be true. But surely morally acceptable exchange must be possible when both agents act on decisive reasons and the situation is symmetrical, so we should conclude that exchange cannot take place among rational agents via one agent’s making such a promise.
    Let’s try to work this through with our simple example of acceptable exchange. Let’s stipulate I have decisive reason to promise that if you give me a dollar I will give you a back rub. In the circumstances, nothing else is a sensible course of action for me to take. And let’s stipulate that one you have the assurance my promise provides, you then have decisive reason to give me a dollar and thus bring about successful and rational exchange. I don’t see any irrationality or moral unacceptability lurking in the slightly revised example. So I don’t see that Julius has identified a problem for my very unoriginal suggestion that a promise can suffice to generate successful exchange—an interaction that is morally acceptable and rational on the part of all participants. There is asymmetry of a sort here. Once I have made my promise, you have no reason to issue any promise. But stepping back, there is symmetry in our situations. As I envisage the scenario, our roles could be reversed. Nothing dictates that I must issue the promise and you must be the one who accepts the promise and performs his part of the deal first. It could just as well have been that you issued the promise and I accepted it and performed my part of the deal first. The symmetry in our situations consists in the fact that we both have decisive reason to bring about successful exchange between us. Of course, there is then no decisive reason for me to issue a promise rather than you. But I don’t see that that fact matters so far as determining the possibility of rational and morally permissible exchange is concerned.

  2. I’d like to thank Avia Pasternak and Zofia Stemplowska for organizing this discussion and Victor Tadros for his summary and thoughtful criticisms.
    Victor’s first criticism seems sound to me. The second raises several hard questions.
    I would guess that the “offer” of a kidney in the circumstances set out by Victor violates Independence. Independence condemns action by which a person A would bring another person C to do x if C does not believe with justification that C’s activity facilitates A’s doing x for “independent” reasons that call for A to x. So it’s not enough that A have good independent reason for marrying C. C needs to be believing with justification that C’s offer of the kidney facilitates [A’s marrying C *for* the independent reason.] If C intends to bring it about that A marries C by bringing it about that A marries C for the kidney, she is presumably not believing that she’s helping A to act for the independent reason.
    Victor might mean to rest the counterexample on his assumption that A and C together have good, offer-independent reason for the pair of their individual actions (C gives up the kidney, A marries C). It seems strained. As Victor points out I haven’t said what makes it the case that two persons have good reason for a pair of their individual actions. If there’s anything to this idea, it presumably requires some interaction in the reason-giving features of the two actions. That seems to be missing in the scalpel wedding and life-saving holdups that Victor describes.
    This is bound to remind you that I owe you a better explanation of the interaction between carpentry and cobbling thanks to which a carpenter and cobbler might have good reason the one to carpenter and the other to cobble. I think the interaction might lie in the division of labor. By spending some of her own time meeting the other person’s need each person frees up the other person’s time for a productive specialization in the other kind of work. You can’t specialize all by yourself.
    I disagree with Victor about the one person’s withholding of the medicine as a way of bringing the other person to save the first one’s life at the cost of the other one’s toe. It seems wrong.
    Finally about contracts for statues to false gods. I would guess that, if the encounter is taken in isolation, and if the falseness of the God voids the statue building of any reason-giving value, this would-be exchange is no good. Each person makes a mistake in bringing the other person to act other than for sufficient independent reason. This tells against the statue-maker’s stance no less than against the believer’s ploy. The statue maker accepts the offer as a way of inducing the payment. She brings the believer to make the payment and not for any independent reason the believer has to pay.
    I emphasize the isolation of this isolated judgment. The matter might look different against the background of our actual societies. In our actual societies we claim and enforce private property in the means to variously valuable and worthless projects. We make people unfree to pursue the projects whose means they can’t buy from some other person who happens to be holding them. This is probably a mistake. But so long as that’s the rule of the game *liberals* might stand ready to make exchanges in the means to other people’s harmless but apparently evaluatively mistaken personal projects. Unless market-going people are local liberals the global reliance on markets in privately claimed goods might exclude evaluative minorities or evaluative have-nots from the means to controversial projects. Which might be a problem.

  3. Hi AJ
    Thanks for the nice response. I might pick up on one or two other points later, but let me focus first on the first longer para.
    In Marriage, I wasn’t sure why it should necessarily be true that the offer of the kidney should not facilitate A’s justified belief that he marries C for the independent reasons that he has to marry. Suppose A initially acts for the kidney, but as a result comes around to seeing the marriage as a good thing. It still seems wrong to offer the kidney.
    But also, the requirement seems very constraining – it seems insufficient to make a transaction permissible, on your account, that I induce you to do something that you already have decisive independent reason to do by offering you something that you have a decisive reason to want. That seems very demanding. Suppose that you have good reason to x, but irrationally do not believe that you have good reason to x. I induce you to x by offering y, and you have good independent reason both to x and y. You act only for y and not for x. Surely this is not always wrong. The marriage example seems a special case – certain kinds of inducement to get people to marry are unacceptable even when they have decisive reasons to marry. These inducements are acceptable in other cases.
    Consider: your house would look great with a repaint, but you are irrationally disinclined to repaint it. I offer you £1000 (or even a kidney) to repaint it. You act for the £1000 (or the kidney), not caring about the repaint – you don’t actively dislike it, you just couldn’t care less. This isn’t pro tanto wrong is it? But my offer of the £1000 (or the kidney) doesn’t induce you to believe with justification that your house needs repainting.

  4. Thanks Avia and Zofia for inviting me to comment.
    First, I want to know more about where sufficient independent reasons come from. In the last two sections of the paper, it’s implied that the independent reason A has to play his part in the group’s action is that this would significantly benefit the exchange partner. This seems to capture our intuitions about the consumption goods machine case described in Section 4. According to AJ, the reason the poor worker lacks sufficient independent reason to work the morning for the capitalist is that the capitalist is rich: “Other persons’ reasons for extending my already vast control over consumption goods are quite feeble”; “The worker’s poverty, the owner’s prosperity, and the owner’s enjoyment of a surplus are important elements of the story.” I think the idea here is that the capitalist wouldn’t benefit to a high enough degree from the worker’s labour.
    But if the reason the worker lacks sufficient independent reason to work the morning is that the capitalist wouldn’t significantly benefit from the goods thereby produced, then it starts to look like any two very well-off people simply cannot engage in exchange. Suppose A and B both have all the food they need, or could reasonably want. They’re both very greedy. A is greedy for apples while B is greedy for bananas, though they both already have plenty of each. I believe that on AJ’s account, A lacks sufficient independent reason to give B a banana and B lacks sufficient independent reason to give A an apple — the reasons they have to do these things are “quite feeble”. Nonetheless, it seems right and reasonable for them to perform the swap. So it seems that the sufficient independent reason parties have to engage in exchange can’t derive from each other’s significantly benefitting from the exchange objects.
    Perhaps I shouldn’t be thinking of sufficient independent reason to perform one’s part in a trade in terms of the other party’s “significantly benefitting” from the action. More often in the paper, AJ uses the word “help”. Exchangers are meant to perform exchanges as a way of helping on another. So we might think the two greedy fruit eaters can help one another by swapping the apple for the banana, and that is what gives each of them a sufficient independent reason to do their part in the swap. But if “A can help B by x-ing” gives A a sufficient independent reason to x, then it starts to look like the poor machine worker has a sufficient independent reason to work the morning for the rich capitalist. After all, it’d make the capitalist better off and that seems enough to count as helping him.
    Of course, earlier in the paper it’s stated that we are interested in cases where each party has a *decisive* reason to do their part. Perhaps the greedy fruit eaters have decisive reason to help each other, while the poor worker lacks decisive reason to help the capitalist. But it doesn’t seem true that the greedy fruit eaters have decisive reason to help one another. They might have decisive reason to engage in the overall trade – but then, so does the poor worker, given the crummy options the capitalist offers him. The point of the consumption goods machine example, I took it, was that the worker lacks a reason to benefit the capitalist – not that the worker shouldn’t accept the capitalist’s offer. And it seems that similarly the greedy fruit eaters lack a reason to benefit one another. So I think the above considerations still apply if we replace “sufficient” with “decisive”.
    For these reasons, I’d like to know more about the idea of “help” that’s at play in the paper, and the difference between “quite feeble” and “sufficient” reasons, to see how AJ makes sense of rich people having sufficient (or decisive) independent reason to give things to one another while not giving the poor machine worker sufficient independent reason to work all morning for the capitalist.
    Second, I’m not sure how we’re to deal with a variant on the consumption goods machine case. Suppose the machine produces cars. It makes them two at a time. It takes a day for the machine to make a pair of cars, which it builds simultaneously. It cannot be made to produce only one car. The capitalist offers the worker the opportunity to use the machine all day, with the capitalist taking one car and the worker taking one car. The key difference between this case and the one described in the paper is that we cannot easily divide the worker’s labour into that which is done for himself and that which is done for the capitalist. Both cars are products of the very same minutes of labour. So the two activities (producing a car for the capitalist, producing a car for the worker) have not been put into what AJ calls an “artificial relation” by the capitalist. Because the worker does have independent reason to make a car for himself, he seems to have independent reason to labour all day. So it seems AJ’s account condones the trade. But I can’t see why this trade should be acceptable while the version described in the paper is not acceptable. In both cases, the worker works all day, the capitalist doesn’t work at all, and they split the day’s product evenly. If we want to say that the simultaneous car production case is not genuine exchange, though, we will need something more than AJ’s account seems to give us.
    Third, I’m unsure about how groups’ reasons to act are meant to generate individuals’ reasons to play their part in the group’s action. I gather the idea is this. If the group made up of A and B has sufficient independent reason to do (cobbling and carpentry), then A has sufficient independent reason to play A’s part in that group action and B has sufficient independent reason to play B’s part in that group action. But, as AJ notes in footnote 8, the relationship between group reasons to act and individual reasons to play their part can’t be one of entailment. Suppose A knows that the two-person group has sufficient reason to do (cobbling and carpentry). But A knows that nothing could induce B to do B’s part in that group action—B is very lazy and/or malicious. It then seems that A lacks sufficient reason to play his part in the group action, even though the group has sufficient reason to perform that action. So in order to derive the individuals’ reasons from the group’s reason, we need to add in some supplementary principles. I suspect that in spelling out these supplementary principles, though, we might be able to do away with group reasons for group actions. I’d like to hear what these supplementary principles are and how they relate to the group’s reason to act. (This might be a request to hear more about the unpublished manuscript listed in the bibliography.)
    Fourth, I’d like to pick up on Victor’s comments about the independence principle’s requiring A to believe with warrant that A is facilitating B’s doing something that B takes himself to have sufficient reason to do, rather than just requiring A to facilitate B’s doing something B has sufficient reason to do. The former formulation of the principle is extremely demanding, since it seems A would have to have access all sorts of details about B’s psychology, history, etc. to have a warranted belief that B will act for reason 1 rather than reason 2. I’m not sure exactly what kind of facts are supposed to lead a cobbler to believe with warrant that a carpenter acts because she has good reason to do her part in the joint action, rather than because she wants the cobbler to cobble for her. Whatever these facts are, I’m not sure they hold in any of the trades ordinary people regularly engage in, such as trading money for a newspaper. Perhaps condemning many everyday trades as non-genuine exchanges is an intended upshot of the account, though.
    [My comments here are informed by a meeting earlier this year of the ‘Cooperation and Equality’ research group at the University of Manchester, where we discussed AJ’s paper. So a hat-tip is due to Joseph Kisolo-Ssonko, Holly Lawford-Smith, Joel Smith, and Hillel Steiner. Of course none of them are accountable for what I’ve said.]

  5. In this characteristically profound and intriguing paper AJ says that this principle of Independence regulates getting others to take some action.
    I should not (do y, intend by y’ing to bring it about that you do x, and fail to believe with warrant that, for some reasons R independent of me, my y’ing facilitates your [doing x because you take R as giving you sufficient reason to x]).
    Independence seems on its face to condemn exchanges (and, it would seem, other forms of cooperation), if exchanges are understood in terms of one’s getting others to do things, not for reasons independent of one’s action but rather for reasons that one makes by conditioning one’s action on theirs. Not so, says AJ. As it turns out, this model of exchange is untenable, for reasons unrelated to Independence. So we need a new conception of cooperation, which, among its other virtues, allows us to model exchanges in a way that reconciles them with Independence. On the new conception, there are reasons that apply to pairs or groups of agents and favour the pair or set of individual actions that defines the relevant cooperative project. This resolves the tension, for there can be, for each participant, independent reason that favours the project and thereby reason that favours the action that constitutes participation in the project. We, together, have reason to lift the couch. And I have reason to lift one end, and you have reason to lift the other, because each of these actions belongs in the pair of actions that we jointly have reason to take. So when you suggest that I lift one end and you lift the other, I am to understand you, not as purporting to make a new reason that wouldn’t exist but for your suggestion, but as calling my attention to the reason for the project of lifting that you and I already had, and hence the reason that I had for doing that which constitutes my part in the project.
    I can see how AJ’s model works in cases where we are considering offers or other actions designed to set in motion some cooperative project that is defined and available to be pursued independently of the actions; and the actions in question are taken by a participant in the project so defined. I’d like to hear more regarding some other cases. These include cases where the intervention designed to get cooperation going comes from outside the circle of those whose cooperation is sought. I suppose that in many such cases the outside intervention needn’t be seen other than as notifying insiders of the existence of some cooperative project, or otherwise helping them find their way to a project that was anyway open to them and in which they had antecedent reason to engage. In other cases however it seems that an intervention may open up possibilities of cooperation that wouldn’t exist but for the intervention, and it’s hard to see how we might reconceive those cases in terms of simply pointing the way to action one already had reason to take. Familiar examples include picking one among several eligible alternatives in pure coordination settings, or devising some eligible scheme in broader cooperation settings, where, in the standard way of looking at things, reasons independent of the intervention underdetermine the actions that would constitute coordination or cooperation. Of course, in those cases the intervention is permissible only in case it helps realize a project that was already supported by reason. But the action that participants get to have reason to take, given the intervention, was not itself so supported. I have reason to play my part in a sensible evacuation scheme, but that’s not yet reason to do anything determinate. It’s only once some such a scheme is devised and advertised that I get to have some particular role to play, for example to proceed to some specific assembly point or board some particular lifeboat. It would seem then that, as stated, the Independence condition would condemn interventions that set up the schemes and define such roles and thereby facilitate collective action.
    External intervention can help in the kind of situation described by Stephanie in her third point. Where A and B face a joint requirement but it is known that B would be disinclined, other things equal, to respond to such requirements, the requirement won’t resolve into sufficient reason for A to play the part that he would have reason to play if B were not so inclined. But if some influential player such as a legal institution, acting in the interest of securing the cooperation of A and B, announces that it will monitor and enforce individual contributions to the project, A may get assurance that B will be contribute and, because of that, get reason also to contribute. Insofar as the intervention tolerates B’s acting not in compliance with the joint requirement but rather to avoid the consequences attached to nonconformity, it would seem to be condemned by Independence (even as it might be ratified by Independence insofar as it facilitates A’s complying with the joint requirement and by virtue of doing so). In a related case, where the dispositions of parties to some cooperative project are unknown, e.g. because the cooperative project involves large numbers of strangers, institutional intervention can also work to secure cooperation by giving each reason to expect that others will play their parts, whether because they respond to the reasons for the project or just to stay out of trouble. And as noted above, in respect of both small-scale cooperation among intimates and massive cooperation among strangers, institutions often don’t merely enforce but also in part define the cooperative projects that they help realize. Again, it would seem that such interventions would be condemned by Independence.
    If I understand him correctly, AJ’s claim is that joint requirements help explain a possibility, one rather than the only way in which people can come to cooperate (or to exchange). Perhaps against that baseline the interventions that I mentioned are part of a further, perhaps deviant path to cooperation. If so, these cases needn’t compete with the claim that there exist joint requirements. Nor need they conflict with the background point that the constitutive explanation of reasons to play one’s part in a cooperative project must appeal to a principle of reason rather than to a pattern of behaviour and attitudes on the part of the agents whose project it is. For example, in the evacuation situation that I mentioned, those involved plausibly face a principle of reason that requires of them that in certain specified conditions they evacuate in an orderly fashion. So the action of designing and implementing an evacuation plan helps the relevant agents respond to that principle by moving them to a situation where they are in a position to do so, and because of that is itself supported by reason. But the action seems to violate Independence, and it is not obvious to me how it might be reinterpreted in a way that would remove the conflict.

  6. Many thanks to AJ for such an interesting and complex paper. I enjoyed reading it. I have two comments.
    I too was skeptical about accepting the Independence Principle. Like Arneson, it seems to me permissible to induce you to do something you do not take yourself to have sufficient reason to do, provided there actually *is* sufficient reason for you to do it.
    An example that occurred to me in this context: suppose my dad has a penchant for telling off-color jokes at social gatherings. It’s part of his personal code that one should feel free to express oneself, and he thinks that other people should stop be so easily upset. Nonetheless, people are seriously offended by these jokes. Tonight we are on the way to dinner with some friends of mine. I know that once we get there my dad will be tempted to tell one of his jokes, and that he will see no independent reason not to do so. I also know that if I make him promise me now, in the car, not to tell the jokes, he will refrain–but only because he does not want to break his promise to me. So I ask him to promise me not to tell any jokes. According to AJ’s independence principle, I have done something wrong: I have aimed to get my dad to act in a certain way, without enabling him to respond to the reasons that he sees as independently bearing on his action. But I think what I do is perfectly permissible here, indeed that it is probably the right thing to do in this situation. So AJ’s independence principle seems too rigoristic to me. I worry that it rules out lots of innocuous actions.
    Another worry I had about the paper is this: what would it mean for an economic exchange count as a joint project that we have independent moral reason to engage in? It seems to me that most everyday exchanges cannot readily be characterized, in this fashion, as morally recommended joint projects. Consider an average transaction: I walk down to the convenience store on the corner and buy a cup of coffee and a newspaper for $5. While it is nice to have the coffee and the paper, my need for these goods is not particularly urgent. The store owner would have had no compelling moral reason to make sure I was supplied with these things independently of my offer to buy them. Likewise, while the store owner is not a rich bastard, I doubt that he is a particularly morally eligible recipient of my $5 on grounds of need. He is a reasonably well-off middle class guy. Absent our exchange, probably I would have had more independent reason to give the money to the homeless man down the street. So it’s hard for me to see this exchange as a joint project that carries any particularly strong moral recommendation in favor of it. To confirm that, suppose I had simply woken up this morning and–contrary to my usual practice–decided not to walk down to the corner store to get my coffee and a newspaper, but rather to do without them today, or that I had gone somewhere else instead. There doesn’t seem to be any sense in which I’d be “letting down the team,” by failing to perform my part of a morally significant joint endeavor involving the convenience store owner. So I’m not sure where AJ’s views leave most everyday exchanges. These ordinary exchanges, I take it, are for the most part not ones in which the participants have significant independent moral reasons to engage.

Comments are closed.