Sen, ‘The Idea of Justice’ (Chapter 4, Voice and Social Choice)

This chapter tells us more about Sen’s understanding of the ‘transcendental’/comparative distinction. I’m not going to cover all (or even most) of the points he raises in this chapter. Instead, I want to raise a question that builds on a few comments about justification from the discussion of the Introduction (e.g., Cynthia #2, Colin #5/7, Charles #15, David W. #16, Blain #17, Aaron #18). Here is my question: Is Sen’s theory of justice ‘political in the wrong way’? I’m going to suggest that (i) Sen seems to be saying ‘yes’, (ii) he ought to say ‘no’, and (iii) if he says ‘no’, the difference between his approach and ‘transcendental’ ones is greatly diminished (or perhaps removed).

What does Rawls mean by ‘political in the wrong way’? In Part V of the Restatement, he says that political liberalism seeks a kind of consensus that is different from ‘consensus politics’. The latter aims to identify a particular policy that can gain sufficient political support in a particular time and place, without seeking agreement concerning the justification of the policy (and allowing the balance of power between various groups to influence the decision). For example, one might hope to reach agreement on the ‘diagnosis’ that X is unjust, without first (or ever) identifying why X is unjust.

In contrast, Rawls’s view is political not because it bypasses the question of justification, but because it deals with that question in a political way. On (my interpretation of) Rawls, he holds that a theory of justice must be based on a moral analysis of the problem that the theory is meant to solve. In Rawls’s case, the problem he addresses is how to reconcile individual freedom with the social necessity of using a coercive overarching authority to impose a cooperative scheme that yields unequal benefits and burdens. This, plus further considerations, gives us a moral analysis of ‘the political relationship’, which then leads to a conception of society (as a system of fair social cooperation) that is supposed to allow us to respond to the identified problem. (The coercion and inequality inherent in the political relationship can be reconciled with our freedom provided that the terms of social cooperation are fair.) Finally, that idea of society allows us to ‘work up’ a set of ‘political’ values that are appropriate for justifying the use of state authority, which enables us to answer why X is unjust.

With that in mind, in order to object to Rawls, one can either (a) argue that his understanding of the fundamental problem of political philosophy and/or his resulting conception of justice is mistaken, or (b) argue it is a mistake to think that a theory of justice needs to be based on a moral analysis of the problem(s) to which the theory is meant to respond.

Now, I think there are several plausible ways of pursing (a)-type challenges to Rawls. For example, one might question whether being a ‘fully cooperating participant’ in a system of social cooperation is a necessary condition for membership in the corresponding ‘justificatory community’ (to use Cohen’s phrase). However, it seems to me that Sen is proposing to pursue a (b)-type challenge. He wants to ‘diagnose’ particular instances of injustice in a way that allows for ‘plural grounding’ on the basis of multiple, conflicting ‘evaluative criteria’. Let’s consider how social choice theory is supposed to enable that.

I think (and I should note that I am not a social choice expert) that his point (6) on page 109 is the most important for the above discussion. It seems to me that the distinctive power of social choice theory is entirely clarificatory. It can do three things:

(i) Given a fixed set of individual rankings and priorities (‘inputs’), and given a fixed set of axioms meant to capture formal rational and/or moral requirements, social choice theory can identify the extent of agreement on what should be done (‘social conclusions’). In so doing, social choice theory can discover previously unrecognized areas of agreement.

(ii) Given a fixed social conclusion and a fixed set of axioms, social choice theory can identify how much disagreement in terms of inputs is compatible with agreement on the specified social conclusion.

(iii) Given a fixed social conclusion and a fixed set of inputs, social choice theory can identify which sets of axioms are compatible with agreement on the specified social conclusion.

Now, Sen’s view seems to (to me) be this: IF we improve the quality of the inputs (through [a] promoting public reasoning to encourage people to scrutinize their own rankings and priorities, [b] using informationally rich inputs – such as capabilities, and [c] battling parochialism by incorporating ‘enlightenment’ perspectives), THEN we can do (i) above to show that people’s actual rankings and priorities are compatible with more agreement about what is an instance of injustice than we intuitively assume, WITHOUT needing to engage in futile debates about why something that we agree is unjust is unjust. That is what Sen seems to be suggesting, and it fits the description of Rawls’s phrase ‘political in the wrong way’.

From a practical perspective, if this is treated as a recommendation of how to actually bring about social change, and if we assume that it is empirically plausible as such, then I have no problem with it. Indeed, it would quite exciting if Sen brings about more public reasoning and popular discussion of justice on a global level and then combines that with analytical efforts to discover hidden areas of agreements, followed by political campaigning to act on the findings.

However, from a philosophical perspective, I don’t think the view I’m attributing to Sen would achieve the ‘bypassing’ that it apparently sets out to do. As a result, I think Sen’s theory, despite appearances so far, is not ‘political in the wrong way’. Further, I think his view is not non-‘transcendental’. Consider the following questions:

Whose inputs matter? (Why does justice require us to take everyone’s rankings and priorities into consideration? Who is ‘everyone’? What counts as an ‘enlightened’ perspective? Why?)

Which kinds of inputs matter? (E.g., Should we ignore ‘external preferences’? If so, why? Why are individual rankings and priorities that survive reasoned scrutiny better than ones that do not? Are they more morally defensible? If so, according to what moral values, ideals or principles? Why are those moral values, ideals and principle significant for justice?)

How should the inputs be weighted? (If more people agree on one ranking than another, should the two rankings be weighted equally or according to the number of supporters? Why?)

Should the axioms aim to capture only rationality, or also some moral demands, like reasonableness? Why/why not?

Questions like these must be answered in order to defend the general claim that the outcome of a social choice procedure has anything to do with justice, as well as the particular claim that a particular social choice procedure has incorporated the best interpretations of inputs (page 108). And I think that all of those questions force us to address questions that are ‘transcendental’ (according to the way Sen uses the term). The questions above force us to explain what the connection is between agreement, justification and political legitimacy. They force us to ask what it is about instances of injustices that make them unjust. But if social choice theory needs to ask those kinds of questions in order to defend its favoured justice-generating social choice procedure, then it isn’t different in kind from ‘transcendental’ approaches.

That being said, Sen’s approach may still be significantly different from Rawls’s approach for (a)-type reasons. What I hope is that Sen does explicitly address questions like those listed above (I haven’t finished the book yet).

Sorry for going long,


Be Sociable, Share!

About Chris Lowry

I focus on liberal theory, esp. Rawls. My current work examines how the capability critique of primary goods has theoretical implications for the neutrality/perfectionism debate. In particular, I'm working on a defence of limited perfectionism that fits within a larger neutralist framework and is justified by familiar neutralist-type reasons.
This entry was posted in Posts, Reading Group and tagged , , , , . Bookmark the permalink.

9 Responses to Sen, ‘The Idea of Justice’ (Chapter 4, Voice and Social Choice)

  1. I have only caught up on reading this chapter today, so thanks Chris for this helpful summary.

    There are a couple of points I wanted to make here to (a) provide more support to the central arguments Sen advances in this chapter and (b) expand the analysis beyond Rawls (I don’t want to give the impression that I have it in for Rawls!) by critiquing Nozick along lines similar to what Sen argues in this chapter against the transcendental approach.

    Chapter 4 is a substantive chapter and Sen elaborates on many of the key arguments and points he outlined in the Introduction. In particular, he provides more support for the provocative claim that the question “What is a just society” is not a good starting point for a useful theory of justice.

    I believe this simple claim is actually among one of the most important theses that the discipline should address and debate. If our theories of justice have started off on the wrong path then investing more of our energies into waddling yet further along that mistaken path will be in vain. And that conclusion ought to trouble us as philosophers and as educators.

    [Just an aside: I think the fixation on the intellectual exercise of figuring out the demands of perfect justice is a much more modern phenomenon than Sen believes, and has more to do with the professionalization of the discipline than a common intellectual thread between thinkers as diverse as Hobbes and Rousseau]

    In this chapter Sen argues that the transcendental approach is neither sufficient nor necessary. On p. 98 he asks “does a transcendental approach produce, as a by-product, relational conclusions that are ready to be drawn out, so that transcendence may end up giving us a great deal more than its overt form articulates?” His answer is no. Transcendental approaches do not give us rankings of departures from justness in terms of comparative distances from perfection. And I think the reason why transcendental approaches do not, and cannot, provide this is that such theories bracket the non-ideal considerations that would bring such comparative concerns to the fore (e.g. non-compliance, scarcity, disease, etc.). And so asking what justice is in a closed, rich society full of compliant healthy persons is really the wrong question for us to ask. It doesn’t help us in our deliberations about what the demands of justice are in societies that have crime, disease, immigration, ballooning deficits, etc. Sen gives the example of comparing mountains (p. 102). Accepting that Mount Everest is the tallest mountain in the world isn’t necessary, nor particularly helpful, when comparing the peaks of Mount Kilimanjaro and Mount McKinley.

    To bring what I think is the power of Sen’s critique of the transcendental approach to the fore, permit me to apply it to Nozick’s theory (as I have already targeted Rawls and don’t want to give the impression that Rawls is the only culprit here). Sen himself doesn’t address Nozick’s theory in great detail in this chapter (there are one or two references to Nozick), but I think Sen’s general critique can be made more compelling by considering how it applies to a specific transcendental approach like Nozick’s.

    Nozick’s transcendental account of justice maintains that “whatever arises from a just situation by just steps is itself just (Nozick, 1974: 151). So a perfectly just society is one with a minimal state where the rights of self-ownership, including property rights, of all are protected. There is no taxation on earnings for things like satisfying the demands of equality or priority. The priority of liberty, for Nozick, means liberty cannot be sacrificed for these other societal aspirations or interests. While it might be morally laudable if citizens choose to donate a portion of their earnings to help the disadvantaged, for Nozick it is unjust to compel citizens to do so via the coercive use of state power (in the form of redistributive taxation).

    The great bulk of ASU is an exercise in ideal theory. However, there is one important “non-ideal” principle that Nozick invokes in his (scant) discussion of rectifying past injustice. Nozick’s entitlement theory of justice is premised on three principles- the principle of transfer, the principle of just initial acquisition and the principle of rectification. The first two principles help us to determine when one is entitled to a good or holding. But of course people do not always abide by the requirements of these two principles. Human history is not one of just initial acquisition nor just transfers. It is a history of slavery, conquest, theft and fraud. To remedy such injustices the entitlement theory must invoke a principle of rectification.

    Given the *actual* history of human acquisition and transfers, it is surprising that Nozick’s historical theory does not make the principle of rectification much more central to ASU. The topic ‘rectification’ appears only five times in the index and totals a meagre seven pages in a book that exceeds three hundred and fifty pages. The neglect of this topic no doubt reflects the fact that Nozick’s approach is a transcendental one, rather than comparative approach. The scant discussion of rectification gives some hints at how the entitlement theory might be applied in the real world. But what Nozick has to say about this is very unsatisfactory. Indeed, I think he struggles to try to make it cohere with his transcendental account. And I think this reinforces Sen’s critique that such theories can’t provide us with rankings we can use to tackle injustice in the real world.

    Here is what Nozick proposes as a way to tackling the concerns of rectification:

    Perhaps it is best to view some patterned principles of distributive justice as rough rules of thumb meant to approximate the general results of applying the principle of rectification of injustice. For example, lacking much historical information, and assuming (1) that victims of injustice generally do worse than they otherwise would and (2) that those from the least well-off group in the society have the highest probabilities of being the (descendants of) victims of the most serious injustice who are owed compensation by those who benefited from the injustices (assumed to be those better off, though sometimes the perpetrators will be others in the worst-off group), then a rough rule of thumb for rectifying injustices might seem to be the following: organize society so as to maximize the position of whatever group ends up least well-off in society. (Nozick, 1974, p. 231)

    Given the importance the issue of rectification has on Nozick’s entitlement theory one is bound to wonder why Nozick does not make this issue more central to ASU. Nozick now adds a very important qualification to his argument for the minimal state. To his original declaration that ‘taxation is on a moral par with forced labour’ we must add: if and only if no considerations of injustice could apply to justify such taxation. The minimal state is only justified provided all past injustices have been rectified. What society can claim to have satisfied such a requirement? Once we extend the issue to a global context the issue of rectification becomes even more complicated. Which of the developed countries can truthfully claim that they acquired their current level of prosperity from just transfers and just initial acquisitions?

    Nozick’s entitlement theory of justice is a good example of why it is a mistake to start with the question “What is the just society?”. Libertarians take the idealized story of the Lockean appropriation of initial property and apply its prescriptions to the real world without acknowledging the realities of how property was acquired.


  2. Chris Lowry says:


    I share your worries about Rawls’s idealizations. Insofar as his theory is designed for the ‘circumstances of justice’ he specifies, it’s not obvious what the theory can say about situations where those circumstances (either objective or subjective) are not satisfied. (Blain – does the ‘general conception of justice’ apply beyond Rawls’s circumstances of justice?)

    However, I don’t think that worry leads necessarily to a rejection of ‘transcendentalism’. I think one can object to an overly narrow account of the circumstances of justice, and yet still be committed to the view that we shouldn’t bypass questions of justification.

    One of the reasons I say that is that it seems to me that are three starting points, rather than Sen’s two:

    (a) We could aim to specify necessary and sufficient conditions for a perfectly just society. (What is a just society?)

    (b) We could aim to specify practically significant sufficient conditions for instances of injustice. (What is an injustice in the world as it is?) We could call this ‘transcendental minimalism’ or some such.

    (c) We could aim to identify instances of injustice without specifying sufficient, or necessary, conditions. (What should be done to improve justice?)

    I think the line between B and C is more significant than the line between A and B; and I think that Sen seems to want to situate himself at C (but I think he ought be at B). A and B both seek to understand why an instance of injustice counts as unjust; although B is less ambitious. (Pogge, World Poverty and Human Rights is probably an example of B). C, in contrast, seeks to bypass the question of what makes something unjust, and instead tries to jump forward to the diagnosis of instances of injustice.

    If that characterization is right, then one could favour B over A (say, for reasons related to objections to idealizations), and yet still be ‘transcendental’ – where that simply means being committed to the view that we need to tackle questions about justification (and legitimacy and so on) on the road to identifying instances of injustice.

    Is B compatible with your view, Colin?

    And to everyone – am I being fair to place Sen in C, rather than B?

  3. Chris,

    I suppose it depends on what (b) entails. As you state it, I would certainly agree with it, and I think Sen would as well. Specifying “*practically significant* sufficient conditions for instances of injustice” doesn’t, it seems to me, presuppose any commitment to a transcendental identification of justice.

    The “practically significant” condition (which refers to the world as it is) thus makes the sufficient condition for an instance of injustice (a) context dependent and (b) provisional (as the circumstances of the world change so too would the sufficient conditions of injustice).

    My own virtue-ethics approach to justice would be receptive to that. And I think Sen’s comparativist approach would be as well. Sen doesn’t think we need to specify sufficient or necessary conditions of transcendental justice. But I do think he believes the comparativist approach needs to do what you outline in (b). On p. 95, for example, when contrasting philosophical theories of justice with social choice theory, he claims:

    “The outcomes of the social choice procedure take the form of ranking different states of affair from a “social point of view”, in light of assessments of the people involved. This is very different from a search for the supreme alternative among all possible alternatives, with which theories of justice from Hobbes to Rawls and Nozick are concerned”

    So I think Sen would endorse (b) but reject the claim that it is “minimally transcendental” if by transcendental one means it possesses the two features Sen identifies in the Introduction- namely, it concentrates its attention on what it identifies as perfect justice and it concentrates primarily on getting the institutions right. Neither of those features seem to apply in (b).


  4. Derek Bowman says:

    1. Sen’s main achievement in this chapter is to show that accounts of transcendental justice are neither necessary nor sufficient for generating the comparative judgments that we face in making actual political choices. In human endeavors it is rarely the case that we find tools which are, by themselves, either necessary or sufficient for the task at hand. (See, e.g. Sen himself on the limits of reasoned scrutiny, p. 39-40 and p. 89). What Sen needs to show is that transcendental theories are not helpful or important for dealing with the actual choices we face, and the necessity and sufficiency arguments don’t even begin to do that. In light of this, it’s worth noting how little Sen has said so far about specific decisions faced by actual political actors and the role political theory can play in them.

    Let’s consider the example of Martin Luther King, Jr. He was part of a movement that fought for, and in no small part achieved, political change in the name of justice. Such success would not have been possible without careful strategic planning, including tough decisions about which incremental changes were worth struggling and even dying for.

    But King also had a Dream which represented an ideal of a perfectly just society. Lest we dismiss King’s moving vision as mere rhetoric, we must recognize the place of such ideals in real politics. In light of that recognition, it would be folly to abdicate our responsibility to critically examine such ideals.

    2. Colin,

    Far from illustrating the folly of ‘ideal theory,’ the example of Nozick demonstrates the absurdity of trying to derive any serious methodological conclusions from a crude division of political philosophy into a few narrow categories.

    First, Anarchy isn’t primarily a treatise on justice. Like Hobbes and Locke, Nozick is first and foremost concerned with the limits of justifiable coercion, especially state coercion. Second his entitlement theory as a theory of distributive justice, a topic confined to a single chapter.

    Once we turn our attention to the work as a whole, we see that the whole first part is driven by problems that arise from our need to protect ourselves from injustice as well as from the overzealous prosecution of injustice by others. This includes discussions of compensation and punishment for rights violations. Even within the account of distributive justice, much attention is paid to the problem of interpreting the Lockean Proviso in light of the fact that there is not an unlimited supply of unappropriated land available for acquisition.

    Does his attention to these problems mean these parts, like his discussion of rectification, are also ‘non-ideal theory’? Do these parts of the account still paint a picture of a perfectly just society? More importantly, would answering those questions, in those terms, tell us anything useful about how we ought to be doing political philosophy? Clearly not.

  5. Derek,

    I disagree that ASU isn’t primarily a treatise on justice. I take the concern for the limits of justifiable coercion to be a concern with justice (and I think Nozick does as well). The entitlement theory of justice is derived from a prior, and more foundational, normative premise of Nozick’s: the thesis of self-ownership. So the concern about coercion informs and determines (or limits) Nozick’s discussion of the redistribution of wealth. I don’t see how that characterization is a crude division of his work and the central aspirations of ASU.

    You claim “…Even within the account of distributive justice, much attention is paid to the problem of interpreting the Lockean Proviso in light of the fact that there is not an unlimited supply of unappropriated land available for acquisition.”

    I suppose one could quibble over what constitutes “much attention” in this context. Sure Nozick acknowledges the fact that there is no more unowned land to appropriate today. Thus he simply replaces key aspects of the Lockean proviso (e.g. “enough and as good proviso”) with the requirement that appropriation not worsen the (material) situation of others. And he then asserts that capitalism has, in fact, done this. But he provides little argument or empirical support to substantive such a claim.

    I think the Nozick example is an interesting one to consider because he offers us a theory of justice that, at times, does invoke non-ideal considerations (like rectification) and comparative judgments. But these fall to the background because Nozick is primarily interested in winning a philosophical debate about what constitutes the perfectly just society. Thus one comes away from ASU thinking that justice requires us to oppose redistributive taxation and implement the minimal state rather than attend to the rectification of past injustices.

    I think attending to the question of what the proper relationship is between facts (of our history, human nature and political life) and normative ideals does tell us something useful about how we ought to be doing political philosophy. It tells us that we should not approach the discipline as (to borrow David Miller’s distinction) “members of the Star Ship Enterprise” might (i.e. separating the project of defining concepts from applied political theory); but rather we should conceive of the discipline as something that is for us mere “earthlings”. And that means we must attend to the relationship between facts and normative ideals.


  6. Derek Bowman says:


    I very much agree that we ought to attend to the proper relations among facts about human lives and normative ideals. But dividing all of political philosophy into ‘ideal’/‘non-ideal’ theory or ‘trascendental institutionalism’/’social choice comparativism’ obscures rather than highlights those relations.

    The question of the justifiable limits of coercion is distinct from the question of how to characterize the perfectly just society. Had Nozick started with the second question, unconstrained by unpleasant facts of human history and human nature, the case for the minimal state would never have gotten off the ground. Far from “falling into the background,” such facts drive the entire inquiry.

    Why only these facts and not others? Why some facts here but other facts elsewhere? Why do Nozick and Rawls (and the many other philosophers of the past 4 decades and of centuries gone by) rely on different facts at different places? Which mistakes were predicated on the absence of which facts in what part of which inquiry? These are all interesting and vital questions, but we can’t ask them within a reductive framework that obscures those details.

  7. Sorry I’m coming late to this discussion – I’ve been catching up too. I want to focus on the early part of the chapter, where Sen gives us more argument in support of his claim that identifying perfect justice is neither necessary nor sufficient for making comparative judgments about the relative justice or injustice of particular proposals (96-106).

    Start with sufficiency. People such as Rawls allegedly offer us a “firm identification, with no indeterminacy, of the nature of just social institutions” (89). However, “the characterization of spotless justice… would not entail any delineation whatever of how diverse departures from spotlessness would be compared and ranked” (99). I think we are to imagine that we have been given a detailed institutional blueprint for the one 100% just society, call it society A, and that we know that society A is perfect in all of the dimensions relevant to justice; call them dimensions 1, 2, and 3. If that’s all we know, then we won’t be able to rank societies B and C, if B is higher than C in one of these dimensions, but lower in another. What we need to know is how to rank or weight dimensions 1, 2, and 3. Rawls’s ranking may not be right, and it certainly is not as detailed as one might wish, for the reasons Sen points out on p.99. But Sen is here (on p.99) simply asking for a more detailed ordering of priorities than the one Rawls gives.

    Of course (turning now to the issue of necessity) we don’t always need such a ranking; it may be that society B does worse than society C on all dimensions, or that it does much worse than C on some clearly important dimension, so much so that we feel confident ranking B below C even without a full ranking of our evaluative criteria. But this is just to say that easy cases don’t require much philosophical work.

    In summary, having a detailed ranking or weighting of conflicting evaluative criteria is not necessary for dealing with easy cases, while having a descriptive account of the one society that maximally satisfies all of these criteria is not by itself sufficient for making comparative judgments about hard cases – we need to be able to rank or weight these different considerations.

    None of this, as far as I can see, is in any tension with Rawls’s approach to questions of justice, at least if we set aside the issue of assuming full compliance, and the question of the utility of the original position. Those, I take it, are separate issues. I don’t want to belabour the Rawls vs. Sen issue, because this is Sen’s book, and what matters is what he’s arguing, ultimately, not whether he is fair to Rawls. But I do think it’s important that we be clear about what exactly it is in the Rawlsian approach that we want to reject.

  8. Great comment, Chris. I agree that Sen seems to be arguing for a ‘political’ account of justice that is ‘political in the wrong way.’ I also agree that (whatever we might think of the specific features of Rawls’s original position device, etc.) the question of the justification of principles of justice cannot simply be evaded by focusing on achieving ‘consensus’ on certain issues. Indeed, one of my main irritations with this chapter was the failure by Sen to explain why we should be trying to accommodate the various ‘inputs’ given by different persons in order to ascertain what justice requires us to do in particular cases. Surely there needs to be some kind of justification for doing so!

    Regarding your query: “does the ‘general conception of justice’ apply beyond Rawls’s circumstances of justice?” Yes, the ‘general conception of justice’ always applies. Recall that it simply asserts that any inequality that does not benefit society as a whole, and especially those who are the least-advantaged, is unjust. So it condemns those inequalities that benefit the few at the expense of the many. There is no reason why such condemnation should not apply to any society, no matter how far removed from ‘reasonably favourable circumstances’.

    Here are a few additional criticisms of this chapter:

    Sorry to harp yet again on Sen’s critical comments on Rawls (but, in my defence, Sen keeps criticizing him!), but I found Sen’s list of 6 ‘exclusions’ attributable to Rawls’s theory of justice on page 90 highly dubious. The case for alleged ‘exclusions’ (2), (4), and (6) seem especially implausible.

    (This seems to be a persistent characteristic of this book that I am finding increasingly disappointing. It is fine to reject Rawls’s views because one disagrees with them, obviously, but burning down a straw man is not going to convince anyone. Moreover, it seems entirely gratuitous. How is attacking, repeatedly, a straw man version of Rawls improving the case for Sen’s ‘positive’ approach to thinking about justice?)

    Sen argues that the ‘transcendental approach’ is neither sufficient nor necessary for thinking about all cases of justice. But does any philosopher whom Sen identifies as belonging to the ‘transcendental’ camp actually think this? Certainly not Rawls, for whom certain ‘considered convictions’ regarding justice and injustice serve as the starting points for his construction of ‘justice as fairness’. For instance, Rawls takes it as given that slavery is deeply unjust. We don’t need to know anything about the two principles of justice as fairness in order to condemn as unjust slavery. (This is why I found Sen’s comment at the bottom of page 21 and top of page 22 so bizarre.) Likewise for many other cases of severe injustice. Rawls’s theory of justice is intended to help us address ‘hard cases’ (e.g. “should freedom of expression take priority over equality of opportunity?”) not the ‘easy cases’ (e.g. Sen’s examples on p. 96 of “hunger, poverty, illiteracy, torture,” etc.). (I think that I’m simply restating one of the main points made by Andrew in his comment here.)

    I found Sen’s explanation of the ‘insufficiency’ of the ‘transcendental’ approach quite unpersuasive (pp. 98-101) (leaving aside the fact that it seems to be directed against a straw man target, in my opinion). Yes, in particular cases, it may be that comparisons of particular deviations from a conception of justice may be difficult to judge against each other, but does this really undermine the project of trying to develop a general conception of justice, or show that such conceptions will always be incapable of informing comparative judgements? All Sen’s argument seems to show, I think, is that in some cases judging how far 2+ societies deviate from the requirements of a conception of ‘transcendental’ conception of justice may be difficult, or even impossible. He certainly has not shown, as far as I can tell, that this will always, or even typically, be the case.

  9. Was no one else as flabbergasted by the Mt. Everest example as I was?

    To clarify:
    “The possibility of having an identifiably perfect alternative does not indicate that it is necessary, or indeed useful, to refer to it in judging the relative merits of two other alternatives; for example, we may indeed be willing to accept, with great certainty, that Mount Everest is the tallest mountain in the world…but that understanding is neither needed, nor particularly helpful, in comparing the peak heights of, say, Mount Kilimanjaro and Mount McKinley.” (102)

    To start, Mt. Everest is not perfectly tall. Mountains on Mars are much taller, for example. And it is quite easy to imagine a taller mountain here on Earth. Further, the idea of a “perfectly tall” mountain does not make sense at the outset. Height is a matter of measurement and thus is inherently relative–built into the concept is the possibility of the infinity of counting. Any tall mountain could always be one inch taller.

    Now, some might see justice in the same way–as an inherently relative matter, made up of individual units or packets of justice–but Sen seems to be taking that for granted when most would likely not agree.

    Even more, the ease with which geologists may compare heights of mountains is in large part due to total and universal agreement about units of measurement. No one in the world disagrees about what makes a centimeter a centimeter. Its universality is built into the unit of analysis.

    In order for this analogy to make sense, it seems to me, we would have to all agree on a pre-political measurement of justice against which situations may be judged. But that is precisely what is at stake for Rawls and company–the fact that measurement of justice requires first coming to some sense of what we will mean when we use the concept at all!

    Perhaps this is simply a re-statement in less useful terms of the arguments made by Chris about ‘political in the wrong way.’ Along those lines, I would say that the general conclusion that might be drawn from the sloppiness of this analogy is that Sen is imagining differences where they don’t exist and therefore reducing the effectiveness of an actually quite interesting argument that could be made about supplementing the ideal-driven focus with social choice theory in the evaluative process.

Leave a Reply