Thanks to Irfan for his generous invitation to join the Policy of Truth blog, and thanks to all of you who are taking the time to read this. As Irfan has said, I’m a Visiting Assistant Professor of Philosophy at Providence College. If you’d like to know more about me or see some of my other philosophical writing or teaching materials, you can browse my website.
This paper is a conference-length version of a longer paper I’m currently preparing for submission to academic journals. Following the lead set by David Potts, I’ll be posting each section of the paper as separate posts, but you can also consult the full paper at any time. There are currently four sections, and if all goes well I may add a fifth post with material taken from the longer version of the paper. Today’s introductory section is quite short, but one thing we might talk about is why any of this matters. Why did Rawls – why might we – think it important to identify these “circumstances?” Or, in a more skeptical vein, does anyone think there is already something wrong in this way of talking about the role of principles of justice?
Without further ado:
I. The Idea of the ‘Circumstances of Justice’
Justice, in its classic formulation, is the virtue of rendering to each her due.[1] David Hume rejects this as a proper definition of justice fhttps://wordpress.com/post/irfankhawajaphilosopher.com/10033or the obvious reason that it gets its meaning only by a further specification of what each person is due – a specification which must be settled by appeal to principles of justice.[2] Nonetheless, the phrase is useful in locating the subject matter, if not the substance, of justice. As John Rawls says, the concept of justice is defined “by the role of its principles in assigning rights and duties and in defining the appropriate division of social advantages.”[3]
The circumstances of justice, then, are the conditions under which such principles apply. Rawls characterizes them as “the normal conditions under which human cooperation is both possible and necessary,”[4] and says that outside these conditions, “there would be no occasion for the virtue of justice.”[5] Without the need for cooperation, there would be no need to assign rights and duties and no social advantages to distribute. These circumstances generate the problem to which principles of justice prescribe a solution. What, then, are these circumstances? Rawls credits his account to Hume, claiming to have added “nothing essential” to Hume’s fuller treatment.[6]
As I will argue, however, Rawls’s claim is very much mistaken. His Kantian moral psychology and his more capacious concept of ‘justice’ lead him to adopt a list of circumstances that, despite their substantial overlap, cannot be justified, or even explained, in Humean terms. In order to resolve this problem with the Rawlsian-Humean account, I propose a revised account of the circumstances of justice. Thus, resolving this interpretive puzzle points the way to a deeper understanding of the nature of justice and its grounding in the human condition.
[Continue to Part 2]
[1] This formulation traces back at least to Plato’s Republic, where Polemarchus attributes it to the poet Simonides. See Republic 331c-332d.
[2] Hume Treatise, 3.2.6.2 [Book#.Part#.Section#.Paragraph#]
[3] Rawls 1999, p. 9
[4] Rawls 1999, p. 109
[5] Rawls 1999, p. 110
[6] Rawls 1999, p. 110
Works Cited
Hume, David. 2000. A Treatise of Human Nature. Edited by David Fate Norton and Mary J. Norton. Oxford: Oxford University Press.
Plato. 1992. Republic. Translated by G.M.A. Grube. Revised by C.D.C. Reeve. Indianapolis, Indiana: Hackett Publishing Company, Inc.
Rawls, John. 1999. A Theory of Justice, rev. ed. Cambridge, MA: Harvard University Press.
Policy of Truth 
Great to see you get started with PoT, Derek! Welcome!
I have several questions/comments, some of them no doubt pretty naive (I am not a political philosopher):
(1a) I take part of the set-up here to be that there is the set or system of claims and obligations (rights and duties) – “what we owe each other” – and then a subset of this set that concern or define the appropriate division of the advantages (benefits) and costs of living together in something that counts as a society. And it is these latter claims and obligations that are claims and obligations of justice. Is this right?
(1b) If so, I have the following concern: intuitively, many (most, the paradigmatic) claims and obligations of reciprocity are claims and obligations of justice. Yet it is not obvious how or in what sense these claims and obligations concern the division of social advantages (and costs). The worry, then, is that what we have here is not the essence of justice, but the essence of some branch of justice (perhaps social or political justice in some broad sense). Is this a good worry to have or am I missing something? (Honestly, this is more than a worry. Based primarily on this consideration, my provisional position is that these sorts of accounts are not general accounts of justice, but accounts of social or political justice in the broad sense.)
(2a) Another part of the Rawlsian (and perhaps Humean) framing of the question is that the relevant claims and obligations are “assigned” – constructed, institutional – and that we need an explanation for why they are to be defined and assigned in this way rather than that. This explanation (and justification) is constituted by what are called principles of justice or PJ (I’d prefer ‘basic principles’ to ‘principles’ to make clear that the topic is what explains or justifies the more-specific principles of claim and obligation, but hey I didn’t invent the approach!). Is this right? (Then, to round things out, (i) the conditions of justice or CJ are the conditions under which PJ apply and relatedly (ii) CJ generate a problem, relative to some normative background conditions, to which PJ are a solution. Correct?)
(2b) This part of the Rawlsian (and perhaps Humean) framing seems like just the right way to think of broadly social or institutional (and hence more narrowly political) justice. But if this is all that it does, then it does not concern all of justice. Nor does it say anything about why the state (or some other authority) might have the role of enforcing norms of justice (such as dyadic norms of reciprocity) that would be prior to the norms of social or institutional justice. (Though it would say something about how: presumably, the benefits and burdens of a scheme of prior-rules-of-justice enforcement would need to be distributed fairly.)
LikeLike
One more thing: you probably answer this later, but I wonder to what extent Hume’s framing shares (with Rawls’) any or all of the features that I comment on above.
LikeLike
Thanks, Derek, for posting here.
Just focusing on this section that you’ve excerpted, I was a little unclear about the scope of “…Rawls’s claim is very much mistaken.” Do you mean that (a) Rawls is substantively wrong about what he takes to be the circumstances of justice as such, or (b) that he’s basically right about the circumstances of justice but wrong to think that his account of it can be explained or justified in Humean terms?
I would also echo what Michael says in the first two sentences of his (2b): the same issue occurred to me, possibly in a slightly different way. If we suppose that justice is the virtue of rendering each her due, we can imagine circumstances of its application that involve no institutionalized division of social advantages. Two occasionally-interacting neighbors in a Lockean State of Nature could have duties of justice toward one another (or be required to act on the virtue of justice) but not live under any institutionalized assignment of rights/duties defining the appropriate division of social advantages. No one may have done any “assigning” or “defining.” They might just be living side by side in a Lockean SON and have a vaguer, non-codified, and more intuitive sense of the requirements of justice than is codified in any explicit “assignment.” It’s unclear whether you think of Rawls as giving an account of justice as such, or of political justice relativized to some very specific context (e.g., one that excludes, say, Lockean States of Nature).
I guess I’d also echo the question you pose in your introduction: Why is it important to narrow down the circumstances of justice at all? Why not think that justice has application wherever and whenever people interact–that social interaction is necessary and sufficient for the application of justice? Put another way, why think that there is any such thing as a context involving human interaction that affords “no occasion for the virtue of justice”?
LikeLike
I’m also a little concerned about Michael’s worry that what Rawls et al. are offering us is not an account of justice as such, but of one branch or kind of justice. Unlike (I think) Michael and Irfan, though, I don’t see much of a problem here. It would indeed be a problem if an account of the circumstances of justice turned out to give us only an account of social justice or distributive justice. But while I reserve judgment on whether Rawls himself faces this problem, nothing that Derek has said so far really suggests that we’re headed in that direction or that the attempt to give an account of justice by considering the circumstances of justice will tend to focus too narrowly on questions about the division of social advantages. Rawls wants an account of the circumstances of justice that will help us to understand the role that principles of justice play “in assigning rights and duties and in defining the appropriate division of social advantages.” I take it that both members of that conjunction are essential, not just the second (even if Rawls in fact focuses too narrowly on the second, though I am, as I say, agnostic on that point). There is, in any case, no reason why reflecting on the circumstances of justice would lead us to ignore the rights and duties that aren’t a matter of distribution. The crucial thing about the circumstances of justice is that they are the circumstances in which human beings can co-operate and must co-operate (presumably this is a conditional necessity of the practical variety, not a strict causal or metaphysical necessity). Provided that we (or Rawls, or Hume) do not make the silly error of supposing that the only kind of co-operation we need is primarily distributive, then we aren’t going to be led into the error of supposing that distributive justice is the whole of justice.
I’m also not too worried by the thought that what we’re talking about here is the institutional assignment of rights and duties. I’m not a big fan of State of Nature theorizing, but if a State of Nature is simply a condition in which there is no state, then it’s wrong to suppose that it involves no institutions (and indeed, it’s hard to read Locke as supposing that there aren’t, though maybe Hobbes does), so the worry has little bite so far as I can tell. If, instead, we stipulate that a State of Nature is a condition in which there are no institutions at all, this still seems to make no difference provided that we aren’t seriously supposing that co-operation would be neither possible nor necessary in such a condition. If we for some reason take that supposition seriously, then it would simply be irrelevant, as it would be a condition in which justice was impossible or had no application.
As for Irfan’s question, why think that there is any such thing as a context involving human interaction that affords “no occasion for the virtue of justice”? I take it that, whatever Hume or Rawls explicitly claims, we need not suppose that there is any such context in order to gain insight from thinking about the circumstances of justice. It may be that the circumstances of justice stem from necessary features of human nature. Talk about conditions in which those circumstances do not hold would then be counter-factual. But counter-factual discourse is frequently illuminating.
That may not be of much help, but I’d be curious to see whether Derek is thinking of these issues along the same lines I am.
LikeLike
[SUBSTANTIALLY REVISED 6/20] Thanks, David. That is very helpful. My worry in [1b] was grounded in the idea that, in the following characterization of what justice is from Rawls,
(A) “[the concept of justice is] defined by the role of its principles in [i] assigning rights and duties and in [ii] defining the appropriate division of social advantages,”
the second thing specifies the point or purpose of our coming to the understandings or codifications that constitute the first thing. You are right that [A] does not logically imply that. But I know that many contemporary Rawlsians would endorse the idea that the second thing is something like the point or purpose of the first (and perhaps Rawls had this in mind, too). Such folks would also, I now realize/remember, interpret divisions – or patterns – of benefits and costs to individuals very broadly. Broadly enough to – at least by stipulation – deflect the charge that their interpretation (or reworking) of [A] is in fact a characterization of distributive justice proper, not justice.
However, I think something like my charge stands, on this common reading/reworking of A (but not on your strictly logical, “mere conjunction” interpretation, David). For what is being implied, by having (ii) be the point or purpose (or standard for evaluating) (i), is that what is distinctive to the claims and obligations of justice is not simply that these, but not others, essentially concern cooperation, but that these, but not others, essentially concern achieving the proper patterns of benefits and costs to individuals that are cooperating. And this sort of characterization of justice seems to imply controversial explanatory information that might bias the inquiry (toward justice essentially concerning broadly distributive elements even if these are not the elements at stake in distributive justice proper). And it does seem to me that this information is controversial precisely because it makes the distributions or patterns of individual benefits and costs that result from cooperation – rather than the terms of cooperation conceived of more broadly – what makes the application of fundamental deontological moral norms to the rules of cooperation distinctive.
It does seem right that what makes the (conventional, institutional) claims and obligations of justice distinctive is that they serve some distinctive cooperation-related end. But what is this end? Often, the following answer goes along with the common interpretation/reworking of [A] that I just criticized: all you need is some kind of public understanding (of who owes whom what) – and a common understanding to leave each other alone or not initiate violence against each other suffices even if we basically have nothing to do with each other – to make almost any social arrangement count as cooperation in the relevant sense.
I want to disagree with this as well (and I think this matters in coming up with a good characterization of justice). I’m inclined to think that what separates (conventional, institutional) claims and obligations that shape social interactions generally from those that concern justice is that the latter specifically serve the end of achieving the more-intensive or more-mutually-dependent sort of cooperation that is exemplified by our living together in society (and by cooperating in a certain “involved with each other” enough way in cases of social interaction within a society). Moreover, this explanatory information is information that we should include in a good characterization of justice. For a good characterization of justice needs to distinguish the (conventional) rules of justice from the (conventional) rules of morality that are not rules of justice – and say something about the ways in which they are, in part, evaluated by different standards or relative to different ends. Note that, at least on its face, [A] does not do this – not even on your strictly logical “mere conjunction” interpretation of [A], David.
With this in mind, here is a characterization of justice that, I believe, succeeds in drawing our attention to the right target (by providing the right sort of explanatory information).
(B) Justice concerns (i) the common understandings or rules of claim and obligation that might determine the shape of fundamentally cooperative modes of social interaction and (ii) the moral evaluation of these rules in light of the circumstances relevant to the reality or prospects of fundamentally cooperative social arrangements. Among these rules are rules that would bind the cooperative system itself with respect to (a) costs imposed on individuals, (b) providing adequate access to importantly useful or good system-produced general conditions and (c) avoiding inappropriate distributions or patterns of benefits and costs to individuals (appropriateness/inappropriateness perhaps relative to individual contributions to the system of cooperation and hence ultimately a function of reciprocity).
If we omit the second sentence, we get a characterization of justice that is neutral with respect to the idea that justice is a strictly dyadic individual-to-individual normative affair (just applied to the prospects or reality of fundamentally cooperative social arrangements). I put the second sentence in because it reflects our common implicit belief and confidence in the idea that society can owe things to its participants, the idea that this sort of obligation is grounded in some sort of reciprocity relationship between society and individuals and the idea that the uniform application of reciprocity to all in society yields obligations to avoid certain patterns or distributions of benefit and cost. (At least in the arguably-fundamental case of society, we might want to include information about the ends of having a stable society and government and some of the all-purpose means to achieving these ends. This is just a first draft.)
Of course, the big “fudge” in this characterization of justice is not cashing out ‘fundamentally cooperative’. Though I think I’ve adequately gestured toward what this consists in, I don’t really have an account of what features would make cooperation not merely minimal but intensive-enough (or characterized by enough mutual dependence) to count as a fundamentally cooperative social arrangement. But you don’t need this in order to have a good characterization. For the purposes here, I think this fudgy bit is perfectly okay.
LikeLike
This responds to David Riesbeck’s June 14 post (1 pm):
Right, but I was taking issue with both conjuncts. In a Lockean State of Nature, you don’t necessarily have either assignments of rights or divisions of social advantages. A Lockean SON is compatible with no one’s assigning or dividing anything. Maybe “assigning” isn’t doing the kind of work in the first conjunct that “division” is doing in the second, but that’s not clear to me. But as I see it, an “assignment” seems to presuppose some entity centrally positioned to codify an assignment of rights and duties, and such assignments might not obtain in a Lockean SON.
One worry among others concerns the “must.” If the circumstances of justice were the circumstances in which people must cooperate, justice would seem not to obtain in those circumstances where people do not or need not cooperate. (I don’t mean “contexts in which there is literally no cooperation,” but contexts in which groups of cooperators live side by side with other, neither group cooperating with the other, and neither “having” to cooperate with the other.)
Two points. First, a Lockean State of Nature is a condition in which there is no state enforcing the rule of law, but so conceived a Lockean SON is perfectly compatible with the non-existence of institutions. I suppose the question turns on what counts as an “institution,” but there’s a straightforward sense in which you could have a Lockean SON that lacks both a state and institutions in the sense most Anglo-American political philosophers have in mind. It would just be a very primitive species of Lockean SON. (Would an illiterate society have institutions? How about a barely literate one?) But such things surely have obtained in history, and still do.
Even if a community in a Lockean SON had institutions, it’s compatible with a Lockean SON to have communities living side by side with one another without superordinate institutions. In other words, each community’s institutional norms would apply to it, but no institution would have codified norms applicable to everyone in the vicinity. As I said before, cooperation might obtain within each community but not between them, and it’s a question what implications that has for inter-community justice. Do non-cooperating communities in a L-State of Nature lacking superordinate institutions satisfy the conditions for the applicability of justice? I think so. But does Rawls? Not clear to me, at least from Derek’s account. I’m inclined to think that Hume doesn’t think so (and regard that as a strike against Hume).
Second, I don’t get the skepticism about State of Nature theorizing. Are you not a big fan of SON theorizing because you think they don’t exist, or because (granting their existence) it’s not theoretically useful to focus on them? I would deny both disjuncts.
Arguably, I’m writing from inside of a Lockean State of Nature right now. Area B in the West Bank is nominally under Israeli security control and Palestinian civil control, but in fact, it’s not under anyone’s control. “Israel security control” just means that the Israelis invade whenever they want to, and do so to protect the security of their own citizens, all of whom live (are obliged by law to live) outside of Area B.
Meanwhile, “Palestinian civil control” doesn’t exist, except on paper. There is effectively no criminal or civil law in Area B. There is no local government, no law enforcement (criminal or civil), and no functioning court system. There is no zoning, no planning, and no traffic control (or traffic laws), either. Disputes are settled informally (or by violence). People cooperate, but peaceful cooperation takes place on the basis of an internalized sense of justice/right conduct best known to insiders. Aside from the mosque and the university, it’s not clear to me that there are even any bona fide “institutions” here (within Abu Dis “city limits”). In any case, a mile or so from “city limits,” you’re in Bedouin encampments whose inhabitants regard themselves as subject to no government or institutions whatsoever, and don’t seem to have institutions of their own (in any sense of “institution” I would use). So Lockean SONs (or at the very least approximations to them) surely exist. But I think it’s an interesting question how one conceptualizes justice out here.
Fair enough, but I was just asking for the rationale. Put differently: if the circumstances never hold, why or in what way would counter-factual theorizing be illuminating?
LikeLike
Irfan, you seem to be resisting the idea, arguably implicit in the Rawlsian characterization of justice, that justice does not apply (that there are no obligations of justice) even when it is impossible or there is insufficient reason to create institutions that would secure cooperation. If we imagine a state-of-nature situation in which these conditions hold, though, your intuition seems to be that, in such a scenario, we are still obligated to treat each other in certain ways (and that these obligations count as obligations of justice). I can’t say that I don’t feel the pull of this intuition myself.
One way to handle this disagreement is to: (a) agree that the question of evaluating the relevant understandings or codifications of claim and obligation is an important part of justice (or question of justice) and (b) disagree about the nature of the prior normative elements relative to which such practices are to be evaluated. On your sort of view, some of these prior normative elements are basic obligations of justice that we owe to each other and these are not constituted by conventional obligations/claims that have the right sort of (more basic) normative backing. Does the Rawlsian characterization of justice rule this out? It seems to and I think that is potentially a problem with it. This is a respect in which the theory may not be adequately theory-neutral.
On the other hand, it is perfectly okay to restrict oneself to an account of some of the claims and obligations of justice (the one’s that are essentially conventional) that may or may not turn out to be an account of all of the claims and obligations of justice.
(On my view, all claims and obligations – not only the claims and obligations of justice – concern antecedent reasons, in a type of circumstance, for having relevant social conventions of claim and obligation and reasons for abiding by such conventions once they are instituted. The prior normative elements are (the relevant sorts of) normative reasons possessed by individuals – not anything like a claim/obligation. The conventions of claim and obligation are simply tools (to be used in conjunction with our capacity for reactive attitudes or moral emotion) – the question is what adopting these conventions (or abiding by existing ones) will get us or why we have reason to adopt or abide by them. Since this is a hypothesis about what claims/obligations are, it is hard to come up with a neutral characterization of claims/obligations that does not build in this hypothesis. Rawls seems to be operating with a somewhat similar framework, at least for the obligations of justice. The best way I can think of for not biasing an inquiry (or missing having any clear common target inquiry) is as just indicated: find the part of your account that answers a question any theorist would have and stipulate that it could be that your account X might turn out to be only a partial account of X.)
So: what makes a convention of claim/obligation a convention of claim/obligation that is in particular a claim/obligation of justice and what distinctive work might having such conventions do for us? Do they do this work in all circumstances or only some? And what are the prior normative elements in terms of which they are evaluated? Presumably, these prior normative elements apply in all contexts – in any states of nature that have normal humans in them – not just the circumstances of (conventional) justice (say, circumstances that are favorable-enough for cooperation). It is a substantive question whether any of these prior normative elements are claims/obligations of justice (or claims/obligations at all).
LikeLike
Responding to Michael:
I’m resisting the idea that justice doesn’t apply when there isn’t cooperation, full stop. I hadn’t meant to say that justice doesn’t apply “when [cooperation] is impossible or there is insufficient reason to create institutions that would secure cooperation.” I’m not sure about the conditions under which cooperation is impossible, but spelled out, I myself might be willing to say that justice doesn’t apply there. Cooperation is impossible in life-raft situations or in concentration camps, and maybe justice doesn’t apply there. (Obviously, I mean justice may not apply to the interactions of those imprisoned within a concentration camp, not that it fails to apply to those doing the imprisoning.) Anyway, that’s a more plausible thesis than saying that it doesn’t apply to cases where there simply is no cooperation.
Again, I wasn’t talking about cases where there is insufficient reason to cooperate. The cases I have in mind are cases where there is no cooperation, even if there may be reasons for cooperation that the relevant parties are ignoring or flouting.
What you say in the rest of the same paragraph does capture my intuition. My complaint with the Rawlsian conception of justice is that it fails to account for the application of justice to relatively primitive (Lockean) States of Nature–SON’s where there isn’t the degree of institutionalized cooperation that Rawls seems to envision as a necessary condition for the application of justice. (This complaint certainly applies to Hume’s theory of justice, and may well apply to Kant’s, though I’m less clear on that.)
I agree with what you say in the second paragraph as well, and regard the question you pose as exactly on target: It is not clear to me what Rawlsians want to say (or that they have anything to say) about basic obligations of justice that we owe others qua human, where these obligations aren’t constituted by conventional or institution-relative norms. Whatever else may be wrong with Nozick (including his deontology, property fetishism, and over-reliance on thought-experiments), he’s more alive to such issues than Rawls. He’s got a clear account of the moral norms that apply in a State of Nature, plus a relatively clear account of the State of Nature itself.
I take it as an adequacy condition on a theory of justice that it say something about the application of justice to Lockean States of Nature. I think it’s obvious that Lockean SON’s have existed and do exist, that justice applies to them, that gigantic injustices have taken place in them, and that we need an account of how to apply justice to them. A theory of justice that punts on all of that strikes me as radically incomplete. It has nothing to say about most of human history, and nothing to say about how to adjudicate or conceptualize conflicts between “primitive” peoples or between “civilized” and “primitive” peoples.
To be specific: Hume’s theory of property seems to apply that no one in a Lockean State of Nature has property rights worth respecting. Arguably, Kant’s theory has the same implication, though I’m less sure about that. I’m not sure (don’t remember) what Rawls want to say about this, but it’s worth noting that even if Derek is right about Rawls vis-a-vis Hume and Kant, this is a case in which Hume and Kant (arguably) share enough in common so that whether we attribute “Humeanism” or “Kantianism” to Rawls, we’re ultimately led to the same place.
It may be that Rawls deliberately wants to punt on how justice applies to Lockean SONs, which is fine (he has no obligation to tailor his theory of justice to my concerns), but in that case he’s not giving us an account of the circumstances of justice. He’s giving us an account of the circumstances of one part of justice without making sufficiently clear how much he’s omitting. (I am, by the way, less willing than either you or Derek to give up on the intuition that justice has a certain complex unity to it.)
Put another way: My main worry is that even if Derek is right that Rawls is less Humean than he (Rawls) realizes, his (Rawls’s) conception of justice may still share problematic features that Hume’s and Kant’s theories have in common. From that perspective, Locke and Nozick provide a useful foil to Rawls-Kant-Hume, whatever we ultimately want to say about the Humean component in Rawls.
LikeLike
Quick reply, Irfan. You seem to be assuming that, on a view like Hume’s, we get reasons of justice only after we have adopted public conventions of justice (say, specific rules of obligation/claim). I just asked Derek whether this is the right way to interpret Hume in making an initial comment on his Part II. In any case, one might take this view (and the further view that the antecedent normative backing for conventional norms of justice does not include reasons of justice, perhaps including only reasons of utility). I’m broadly sympathetic to your criticism of this sort of view, but I’m not sure that Rawls has a view like this. He seems to hold that basic norms of justice apply only to the circumstances of justice, whereas Hume seems to be doing something like merely asking what the conventions of justice get us in terms of outcomes.
LikeLike
Thanks all for the challenging feedback and commentary; I think it’s a good sign – for the paper and for the community at PoT – that such a short introduction has already generated such a deep discussion about the nature of justice. For this comment, I’ll start with a couple of easy points, then engage in more detail with the discussion of Michael’s (1a/b).
First easy point: On Irfan’s clarificatory question: When I say Rawls is mistaken, I mean the narrower claim: he’s mistaken in his own self-interpretation when he claims not to have added anything substantial to Hume. I ultimately argue (or at least assert!) that Rawls is more or less right in the things he identifies as circumstances of justice, but he is wrong in the explanation he gives (or often fails to give) for this list of circumstances.
Second easy point: David’s interpretation of me is basically spot on; I’ll try to elaborate on this agreement in a later comment.
(1a/b): What kind of justice are we talking about? Since I reference three philosophers in the opening section, perhaps it will be useful to say something briefly about how I think each of those philosophers would answer this question. For my part I remain somewhat undecided, but my general sympathies are toward a version of the Rawlsian answer, filtered through a Platonic lens.
Plato: Plato clearly thinks that justice itself is a single thing that manifests itself in a structurally isomorphic way in different appropriate subjects. So at the very least we get the justice of the city (“Civic Justice”) and the justice of the individual (“Individual Justice”). The initial idea is that Civic Justice will be easier to identify, and it can be used to discover what Individual Justice is (ultimately the relationship becomes more complicated (since, it is only in the just city that people can reliably receive the education and training needed to become just individuals). The shared structure is that both represent the harmony achieved by placing three conflicting parts into an appropriate ruling relationship. Presumably if there are any other specific kinds of justice, they will share these features (Is it necessary that there by 3 parts? Unclear).
If we force the idea of “circumstances of justice” backward to Plato, it seems that the circumstances are something like the need for coordination among multiple parts that are capable of being organized by an internal ruling/being-ruled relation. Put this way it seems to be pretty close to a generalization on Rawls’s language about the possibility/necessity of cooperation.
Hume: Here the answer is easy. Hume thinks he’s talking about “justice” as such, not some subset. I won’t go into too much detail, since I’ll give a more detailed interpretation of Hume in the next post. But for Hume “justice” is basically just an assignment of property rights, or the (artificial) individual virtue of being disposed to honor those rights. Hume would seem to agree that we have moral reasons (would he use the language of “duty”?) that aren’t matters of justice. But because these would have to be based in the (natural) motive of benevolence, it seems clear that they wouldn’t be best described in the language of what anyone is “owed.”
In connection to David’s comment, it’s wroth noting that for Hume (a) justice is entirely a matter of social convention – the need for a system of rights is a fundamental human need, but any such system will be an artificial construct; but (b) contra Hobbes, the existence of such a system does not depend on their imposition by a government. Indeed for Hume justice (as a system of property rights) is conceptually prior to government. I’ll resist the temptation to elaborate more.
Rawls: Here the answer is a bit more complicated. Rawls is explicit in most parts of Theory that he’s talking about a specific kind of justice “social justice.” Even at the outset he calls justice “the first virtue of social institutions.” And once he’s developing his principles, it’s clear that the justice he’s concerned with – regarding the basic structure of society – is not to be identified with the justice of other social institutions like the family or private associations.
I haven’t combed back through these pages recently, but I’m pretty sure that his discussion of the circumstances of justice has the former restriction but not the latter. That is, it’s the circumstances of “social justice” – the justice of social institutions or of social relations – but it need not be restricted to the fundamental “basic structure” of society.
So how restricted is this? As David points out, it need not preclude a discussion of the justice between neighbors in a Lockean state of nature. Such neighbors are already in a kind of cooperative relation to one another; even if they aren’t engaged in any joint productive ventures, they are cooperating in the sense of refraining from assault, theft, and other rights violations. Not only are they passively refraining from rights violations – they are mutually reliant on the expectation of such honoring of rights. If they weren’t, they would be in something much more like a Hobbesian state of war.
So for Rawls it is an account about the conditions of “social justice” – but social is understood broadly to include the kinds of interpersonal rights and relations that distinguish a Lockean from a Hobbesian state of nature. Although Rawls’s discussion of his own principles is ultimately focused on social institutions that involve a coercive state, I don’t think that narrow scope is meant to govern the general usage of “social” in social justice. On my reading this restriction comes in later as a consequence of the (mistaken) substantive view that only social institutions backed by the coercive apparatus of the state can adequately achieve the coordinating tasks required to address the circumstances of justice.
My take: How important is this restriction? In part it depends on whether (unlike Hume) we think there are other kinds of justice and how we think those other kinds of justice are related to social justice. The Platonic element of my thinking is that if there is another kind of justice, it must exhibit this same kind of structure, involving a need to coordinate multiple agencies through some shared principles of organization/decision-making. If our psychologies really did exhibit the kind of structure that Plato theorizes, perhaps a conceptually independent sense of justice would apply, but only because structurally similar conditions apply.
But I think the kinds of paradigmatic questions of justice we face are all social in the broad sense I’ve interpreted in Rawls. I think some kind of principle of reciprocity is at the heart of justice, and I think that principles of reciprocity are important primarily in virtue of the features of human life identified by “the circumstances of justice.” Creatures living in a Rousseavian state of nature (a condition which may well be impossible for human beings) are utterly unreliant on anyone else – even on the active forebearance of others – in a way that would preclude the need for principles of justice or reciprocity.
LikeLike
Just a very small historical point: it is by no means unduly anachronistic to see the notion of the circumstances of justice operative in Plato. In fact, Republic II seems to give two rival accounts: Glaucon’s contractualist account focusing on avoidance of injustice as a second best and Socrates’ more positive account emphasizing the need for co-operation in view of the lack of human self-sufficiency. One could also find another, slightly different account in Protagoras’ speech in the Protagoras. The notion isn’t formally elaborated as such, but it’s fairly plain that the basic strategy is the same: to think about justice by thinking about the conditions that make it possible and necessary.
LikeLike
Thanks, Derek, that was clarifying. Given the “first easy point,” do you think of our discussion here as focused on the narrow thesis you’re arguing for in the paper (Rawls’s Humeanism) or the broader claim the thesis presupposes (whether Rawls is substantively right about the cirumstances of justice)?
I’m not convinced Rawls is right, and not convinced that what he says applies easily to Lockean States of Nature. (I found Rawls’s account of Locke in his Lectures on the History of Political Philosophy problematically over-simplified) Let me zoom into my disagreement:
That strikes me as an implausibly broad and revisionary conception of “cooperation.” I don’t see how refraining from rights violations is a form of “cooperation.” The claim implies that when someone refuses to “cooperate” with a police investigation, they’re in fact cooperating with the police insofar as they refrain from violating the rights of the police. Or: If you ask for my help in a venture that I regard as immoral, and I refuse to participate or cooperate “in any way,” I am in fact cooperating with you just in case I don’t violate your rights. Or, closer to a Lockean State of Nature: two adjacent (or nearby) communities are “cooperating” with one another just in case neither attacks or raids the other, even if they absolutely refuse (on the basis of some ancient quarrel) to have anything to do with one another, and indeed, hold one another in utmost contempt. Indeed, on your view, they would be “cooperating” even if each community mandated a death sentence for its members’ having dealings with the other community, just as long as neither community attacked or invaded the other. All three of those implications strike me as reductios.
The use of “cooperation” to cover cases of non-interaction just seems to me to flout the accepted meaning of term. A cooperative venture is one in which we act together (relatively self-consciously or deliberately) to promote ends that we share. The term doesn’t apply to cases in which we merely refrain from interfering with the ventures of others, and certainly not to cases where the motivation for non-interference is compatible with intense hostility. I might de facto “respect your rights” simply because I’m indifferent to your existence (in the sense of literally forgetting that you’re there), or too afraid of the consequences of violating them, or believe myself incapable of violating them. More benignly, two communities could just live side by side and not cooperate with one another, not for malicious reasons, but simply because they regard themselves as fixed in their ways, have never cooperated with those people, and don’t feel like starting now.
I say “two communities” because I’m envisioning communities that cooperate internally but not externally, and don’t regard each community’s non-interference with the activities of the other as a form of “cooperation.” But it’s perfectly possible to imagine two individuals doing the same thing–two individuals might live side by side in a Lockean State of Nature, leaving one another alone, but not cooperating. They might be unreliant on one another (or think of themselves that way), hence not cooperating, but I don’t see how their active forbearance from rights-violation would add cooperation to the mix. That said, I don’t think their not cooperating puts them outside of the circumstances of justice, i.e., outside of the conditions within which justice applies.
One of my general worries about reciprocity- or cooperation-based conceptions of justice is that they have trouble dealing with cases in which justice would seem to have application in the absence of anything like mutual cooperation. I suppose one can handle the problem by stretching “cooperation” to handle such cases, but that seems to me to conflate cooperation with its complement–cooperation with non-cooperation–and to regard the conjunction of two incompatible ideas as a single concept.
LikeLike
This is a version of one of my criticisms of a common interpretation of the Rawlsian characterization of justice (in my revised reply to David, above): that it fails to distinguish the conventional claims/obligation of justice (and their point and proper evaluation) from conventional moral claims/obligations that are not claims/obligations of justice (and their point and proper evaluation). Here’s a proposal: at the heart of our moral, normative world, we have strong reason to bring about and play along with common understandings of who owes what to whom in order to organize our social life. Applying this (along with other antecedent normative elements or reasons) to circumstances of avoiding conflict and violence through mutual avoidance and forbearance suffices to justify conventional public understandings of claim/obligation that count as moral but need not count as a matter of justice. Apply this (along with other antecedent normative elements or reasons) to circumstances of intensive cooperation and mutual dependence – the kind of cooperation worthy of the name and characteristic of living together in society and society being functional – justifies conventional public understandings of claim/obligation that count as a matter of justice, not simply morality. To what extent does a picture like this satisfy your concern here? (I’d be surprised if it did not raise other concerns for you.)
LikeLike
You raise a lot of important points here, Irfan, and I kept putting off my reply in the intention of somehow covering them all.
Since I still haven’t done that, let me focus on one specific point. You suggest I’m offering an implausibly broad reading of cooperation, and you suggest that your clear examples of a refusal to cooperate (with the police and with an immoral scheme) suffice for a reductio of that analysis. I do worry that some version of your general claim might be right – that I might be operating with too broad a concept of ‘society’ and/or ‘cooperation’ here. But I don’t think your examples show this, and, to the contrary, and I think they provide some fodder for a defense of this broader interpretation.
Let’s start with the police case. Now we think about a wide range of suspects and witnesses who might be rightly described as “refusing to cooperate” with a police investigation. Let’s take the case of the well-heeled white collar criminal who is “refusing to cooperate” on the advice of his high-priced attorney. It is certainly true that the police are failing to cooperate *with the police investigation.* And, no doubt we can identify any number of specific actions and aims of the police investigators with which is is not cooperating, some of which he may be directly impeding.
Does this show that the white collar criminal is not part of any shared cooperative arrangement with the police? Not at all! He is in fact cooperating with many of their actions with regard to the investigation. For example, he appears in court when summoned. He doesn’t interrupt questions from the investigators, but instead waits patiently before replying that “On the advice of council, I decline to answer any questions pertaining to X, Y, and Z.” In the very act of appealing to the authority of his attorney he thereby acknowledges and acquiesces to the authority of the police investigators – recognizing that he needs to provide some counter-authority which they can be expected to recognize. Together, all these actors cooperate in the ongoing creation and maintenance of a shared system of criminal investigation, prosecution and defense, based on a complex network of shared understandings and mutual commitments.
The same, I think, is true of the neighbors in a Lockean SoN, even where one refuses the others’ offer of (let’s say) a labor-swapping arrangement between their two farms. A Lockean State of Nature already depends on a very rich set of shared understandings, practices, and expectations. Even this paradigmatic act of non-cooperation with a specific scheme of action only makes sense against a shared background framework of property rights. It’s only because I recognize some land as mine and some as yours and that you are free to choose when and where to use your labor that the proposed course of action even makes sense. I don’t think it’s crazy to say that we have to cooperate in creating and maintaining that basic practice of recognizing property rights, just as we do in the criminal investigation case. Even if the normative truth about property rights is in some sense prior to or independent of our practices, nonetheless the effective realization of those rights characteristic of a Lockean State of Nature is going to rely on a basic level of social cooperation.
LikeLike
That’s helpful, Derek. I had assumed that your Rawlsian characterization meant to capture something like “the first virtue of social institutions” and hence was a general account of justice, not an account of justice only as it applies to society. My main worries concerned the idea (not, as David pointed out, implied by the logical structure of the Rawlsian characterization that you give) that the point or purpose of coming to a common understanding (whether codified and coercively enforced or not) about “who owes what to whom” is to achieve some distribution or pattern of benefits and costs among the participants in a common, cooperative endeavor. Whether or not you (or Rawls) meant the characterization to capture this, it is a common idea among contemporary Rawlsians. As reflected in my above (revised) reply to David, my worry now is not the we have aimed at justice and hit distributive justice proper, but that we have characterized the point or purpose of our conventional rules “assigning rights and duties” too narrowly (and too “distributively”) in our characterization (or one leading interpretation or elaboration on it).
I agree with you that there is most likely a unity to justice (and therefore a unified account of what justice is) and that reciprocity is at least one of the main things at the heart of justice. I’m not sure I agree about the circumstances of justice being what makes reciprocity important (though I guess this depends on precisely what you mean, here). Perhaps reciprocity is important on its own but it is only worthwhile to realize regimes of reciprocity when relevant circumstances hold.
It is worth noting that, at the level of function and motivation, our being natural conditional reciprocators at the level of individual interaction does not suffice for the level and type of cooperation that humans typically achieve. Rather, things like having strong intrinsic motivations and emotions of group belonging seem to be necessary to achieve the level and kind of cooperation that humans typically achieve. To my knowledge, motivations and behavior (basic or otherwise) of conditional reciprocity toward society or the state are not a part of the received social scientific view of how our behaviors get oriented toward cooperation in the way that they are. Though it is of course quite controversial how, if at all, these motivational and functional facts about human nature are related to fundamental normativity – as a naturalist and a Humean of sorts, I want them to be quite relevant! – they would seem to be clearly relevant to the question of what the internal, subjective part of the circumstances of justice are (and the analogous circumstances for moral interaction that is not necessarily interaction governed by conventional claims/obligations of justice, if a similar explanatory and justificatory framework applies).
LikeLike
FYI, I’ve substantially revised my reply to David’s 14 June (1 pm) comment. Partly in light of David’s strictly logical interpretation of the Rawlsian characterization of justice as Derek presents it, I now not only withdraw my initial criticism of this characterization (as in fact being a characterization of distributive justice proper) but also
– develop a related “that still makes justice too damned distributive” type of criticism of the common interpretation of the Rawlsian characterization of justice that I was criticizing,
– formulate an important adequacy condition for a good characterization of justice (and argue that the Rawlsian characterization does not meet it – not even on David’s strictly logical interpretation) and
– propose a characterization of justice that meets this adequacy condition (and has some other virtues as well).
LikeLike
Pingback: The Circumstances of Justice: 2. The “The ‘Circumstances of Justice’ in Hume and Rawls” | Policy of Truth
Pingback: The Circumstances of Justice: 3. The Significance of What Rawls Added | Policy of Truth
Pingback: Shameless Plug: Free Range Philosophers | Policy of Truth
Pingback: The Circumstances of Justice: 4. A Revised Account of the Circumstances of Justice | Policy of Truth