Chiara Lepora and Robert Goodin. On Complicity and Compromise (2013)
Compromise as a Template
Compromising is something we do all the time. Maybe ‘compromise’ is itself too grand a term to apply to all of the mundane adjustments that we make to one another in everyday life. But it is not only doctors and lawyers, politicians and humanitarians who find themselves having to compromise on matters of real moral importance.3 Ordinary people, too, know that feeling all too well.
The act of compromising involves, first, coming to an agreement with someone else as to what will be done and, second, implementing that agreement. Theactof reaching that agreement is clearly a joint action. Parties agree the compromise with one another. In the terminology of Chapter1 (and which will loom large in our subsequent analysis of complicity), each party to a compromise is a ‘co-principal’ in that agreement. Each shares responsi bility for formulating the plan of action and for what can be expected to occur as a result of it.4 That act of ‘agreeing a plan’ constitutes what might be called the ‘first moment’ of compromise
The ‘second moment’ comes when the plan of action, agreed through the compromise, is actually implemented. Different agents contribute in differing waystotheimplementationof thatplan. Under the terms of the compromise, they ‘do’, ‘omit’, ‘permit’, ‘induce’, and ‘enable’ different things. Some of those actions count as direct wrongdoing by the agent herself (from that agent’s own perspective). Other of those actions count as contributions to the wrongdoing of others, making the question of ‘what is wrong with the compromise?’ in large part a question of ‘what is my part in the wrongs being done by anyone involved, as part of the compromise?’ The latter is of course the classical focus of discussions of complicity—the question of one’s blame worthiness for contributing to the wrongdoing of others.
2.1 A Conflict of Principled Concerns
Our first observation is that compromise has an intra-personal [개인 내의] aspect, as well as an inter-personal one. The former is a matter of ‘being compromised’: the latter is a matter of ‘compromising with’. Although the inter-personal form (the transitive version of ‘compromising with’ someone else) tends to com mandthe most attention, it is the intransitive intra-personal form that in our view actually lies at the core of the notion.
Compromise arises in situations of conflict. Sometime that conflict manifests itself in physical violence, and a compromise is required to put an end to the fighting. The physical violence, however, is only the surface manifestation of something that is more fundamental, which is a conflict among the agents’ underlying aims and values. The conflict is not only between people: it is also within them. Each has to give up something of importance to her, in order to get something else that is of even more importance to her. That intra-personal calculation must perforce [부득이] go on inside the head of each, in the process of deciding whether or not she will enter into a compromise with the other.That is the reason we say the intra-personal form of compromise is the more fundamental of the two.
An agent would not feel herself to ‘be compromised’ by the compromise unless what she forsakes is something that is of principled concern to her. True, something is lost when you strike a bargain with someone else over the price of a used car. But assuming it was ‘just money’, rather than anything of genuinely principled concern from your point of view, then the outcome would more naturally be described as ‘a deal’ or ‘a bargain’ rather than as ‘a compromise’. Something has surely been relinquished [포기하다, 내주다], but nothing has been truly ‘compromised’ in a situation like that.
It is only when the intra-personal conflict forces an agent to choose among items of principled concern from to her that a compromise is genuinely involved. That explains, in turn, the phenomenology of compromise—why one feels ‘compromised’ when engaging in a compromise, even when on balance she thinks it was the right thing to do. Something of principled concern to her had to be sacrificed, and she rightly regards that fact as a source of regret.
A gratuitous sacrifice of something of principled concern, when there is no need to sacrifice it, would clearly be mistaken from the agent’s own point of view. A person would naturally regret such a mistake, as well; but there is no surprise in that, requiring any deep analysis.
Here we will instead confine our attention to cases in which the compromise was strictly necessary, in the sense that not all of the agent’s (or in the inter-personal case, all agents’) principled concerns could be simultaneously realized. Maybe they are merely ‘incompatible’ in the sense that it is pragmatically impossible for all of them to be satisfied, given resource and other constraints. Or maybe they are truly ‘in opposition’ to each other, in the sense that if one of the principals involved is true then other(s) cannot be.
2.2 Types of Compromise
2.2.1 Substitution Compromise
Suppose two physicians share responsibility for the treatment of a terminally ill patient who is unaware of his diagnosis. Suppose that one doctor thinks, as a matter of principle, that a patient in such a situation should be told his diagnosis so he can start considering end-of-life decisions; and suppose the other doctor thinks, as a matter of principle, that a patient such as this one ought not be told anything that will further impair his already fragile psychological state.11 Suppose the physicians resolve that conflict by each setting aside his or her own original principles and adopting instead a new principle, viz., ‘in such cases, at least the patient’s family must be informed’. Suppose that each physician thinks that it is better to do that than not.
This type of compromise can be called a substitution compromise. In it, the original principles of each of the parties to the compromise are set aside [한쪽으로 치워두다], and some altogether different principle is substituted for them.
Suppose Agent 1 holds principles {A, B, C, D} and Agent 2 holds principles {E, F, G, H}. A substitution compromise occurs when Agent1 and Agent2 agree to acton some different principle {X} that was not originally in the set of principles of either of the agents. Each agent regards {X} as good, but {X} as less good than her original principles. Still, each is prepared to agree to joint action to attain {X}, on the grounds that that each regards {X} as better than what will occur in the absence of any agreement between them.
2.2.2 Intersection Compromise
Imagine now a different pair of agents. Agent 3 holds principles {I, J, K, L} while Agent 4holds principles {K, L, M, N}. There is some conflict between the principles of Agents 3 and 4, but there is also some overlap.The two agents can agree to compromise on that area of overlap. Specifically, the two agents can agree that they will jointly pursue {K, L} alone, with Agent 3 agreeing to forsake pursuit of {I, J} and Agent 4 agreeing to forsake pursuit of {M, N}. Neither agent thinks that subset is actually better than his or her own full set of original principles, but each thinks that achieving that subset is better than what may occur in the absence of an agreement between them.
doctor 1 wants to prioritize on the basis of need (so the worst-off patient comes first); doctor 2 wants to prioritize on the basis of benefit (so the patient who would benefit the most from the intervention comes first). Further suppose that the specialisms of the doctors are complementary, in a way that neither can treat patients successfully without the assistance of the other; and suppose that they are the only two doctors with those specialisms in the vicinity. Each of the doctors can agree that they prioritize those patients who fall in the overlap of the two doctors’ lists of priority patients (those patients who are both very needy and stand to benefit a lot from treatment), with each doctor regretting simultaneously that some other patients on his or her own preferred priority list will be denied priority treatment because the other doctor disagrees.
2.2.3 Conjunction Compromise
A third type of compromise occurs when agents do not share any principle whatsoever in common, either from the outset or after further discussion. Their principles are completely in conflict, and fundamentally (i.e., logically rather than merely pragmatically) so. In set theoretic terms, this can be represented by saying that Agent 5 holds principles {O, P, Q, R} and Agent 6 holds principles {not-O, not-P, not-Q, not-R}.
Nonetheless, the two agents might still agree to a conjunction compromise consisting of some elements of each of the conflicting sets.They might compromise, for example, on {O, P, not-Q, not-R}... each agent agrees to the conjunction compromise (if she does), because each nonetheless thinks that implementing that compromise is better than what would occur in the absence of any agreement.
2.3 Responsibility for What?
In compromising, agents acquire responsibilities of two different sorts.
① they acquire responsibility for what they do directly, as co-principals to the compromise agreement and through the joint agency that is created by it.
② they acquire responsibility for what they do indirectly, insofar as the compromise has them contributing to what they see as the wrongdoing of others.
2.3.1 Direct Wrongs as Co-principals
There are, as we have said above, ‘two moments’ of compromise. First is the moment of agreeing the terms of the compromise. Parties to the compromise are co-principals in that act of agreement, at the very least; and depending on the details of the compromise agreement, they might also be co-principals in the exercise of joint agency following from it. Furthermore, insofar as the compromise involves forsaking things of principled concern to her (as it must in some way or another to qualify as a compromise at all), each party is thereby responsible for what are ‘direct wrongs’ from her own point of view [???]
Here, however, we will be principally concerned with two more distinct kinds of direct co-principal responsibility that might be involved in a compromise, depending upon its type. One is responsibility for what one under takes to perform together with the other as a part of the compromise. Call that committing responsibility. As part of a conjunction compromise, for example, one agrees to commit what one oneself regards as a wrong, as part of a plan of joint action with another agent with values diametrically [전혀 다른] opposed to her own.
Parties to a compromise can also acquire direct responsibility as co-princi pals for what they agree to omit doing as part of the compromise. Call that omitting responsibility. If one is morally required to do something and one fails to do it, one commits a wrong by that omission. In intersection compromise, for example, one omits to act as required by some of one’s own prin ciples—specifically, those that do not overlap the other party’s principles. In a substitution compromise, even more dramatically, one omits to act on literally all of the principles one initially harboured. In both cases, one does so as part of a plan of joint action with another agent with values at least partly opposed to one’s one.
2.3.2 Indirect Wrongs as Contributors
In addition to the wrongs that the parties commit directly themselves as co-principals in the compromise, they can also commit indirect wrongs by contributing causally to the wrongdoing of others in implementing the compromise agreement.
These contributions are causal contributions of a particularly strong sort in the limiting cases we are here discussing. Ex hypothesi, the other party would not have been able to do any of those wrongful things had it not been for what the first party is now doing or refraining from doing in consequence of the compromise. Providing a sine qua non of that sort is a major causal contribution. Indirect contributions of this causal sort might involve either ① permitting or ② inducing or ③ enabling.
③ In a conjunction compromise both parties pursue some of their principles which are wrong in the eyes of the other. There, each party enables the wrongdoing of the other, insofar as she enables him to pursue some of his own principles which she herself regards as wrong, in exchange for being enabled to pursue some of her own that he regards as wrong.
② In a substitution compromise, both parties are prompted to abandon the wholeset of principles they initially wanted to pursue. They are induced by the compromise to pursue a new set of principles.
① In an intersection compromise, one party does not so much enable or induce the other to pursue some of his principles as merely permit him to do so. An intersection compromise permits to both parties to pursue some common principles they held from the beginning, while abandoning some others
2.3.3 Compromise Increases Individuals’ Responsibilities
He may dosoreluctantly, even regretfully, but nonetheless for good reason: because doing what is required by the compromise is better on balance than anything else he can do in the circumstances. But the fact that there were good reasons against the compromise as well as for it should not obscure the fact that it was a matter of choice.
The second thought concerns dilution of responsibility. In shared actions one might think that if a wrong is committed by a dozen people rather than by two, then the responsibility falling to each of them for doing so is correspondingly diminished. Psychologically, the ‘diffusion of responsibility’ phenomenon is undeniable. But morally it is suspect. Certainly it is so in the case of a compromise.
For a start, remember that a crucial part of compromise (agreeing the compromise) is a joint action, not a shared one. In a business partnership, all partners are fully liable for all the actions of all the other partners under taken pursuant to the partnership.So too are all co-principals in a compromise morally liable for the existence of the compromise and all actions pursuant to it.
Responsibility is multiplied rather than divided as a result of the compromise. In a compromise, each party not only retains responsibility for what he himself commits and omits; he also acquires contributory responsibility for what the other does as a result of his permitting, enabling, or inducing. The greater the wrongs that you facilitate, the more blame that you share. Multiplying the numbers in that sort of case increases rather than reduces the blame that you bear.
2.4 Moral Discomfort in Compromise, and Why it is Warranted
Moral discomfort inevitably, and rightly, arises from sharing responsibility for wrongdoing in those ways through the compromises that one has entered into.
In part, that is responsibility for direct wrongs that one has committed oneself: through agreeing a compromise in the first place, and through wrongful acts that oneself commits (or right acts that one wrongfully omits to commit) pursuant to the compromise.
In part, that is responsibility for indirect wrongs one has done in contributing causally to the wrongdoing of others, permitting or inducing or enabling it through the compromise.
In all of those ways, compromise involves an agent in sacrificing something that is of prin cipled concerntoher. Hence the moral discomfort that is invariably associated with compromise. In our subsequent discussion of complicity, that is not a feature that will necessarily arise in all cases of complicity more generally (it may or maynot).
The crucial point is this. A moral bottom line of +2 can arise in many different ways. It might arise from a balancing of +4 goods against 2 bads. Alternatively, it might arise from a balancing of +1,002 goods against 1,000 bads. The bottom line is the same +2 either way. But anyone who sees no difference—anyone who says, ‘same bottom line, no difference’—is clearly missing something of very great moral importance. Call that the ‘moral gravity’ of the situation.