John Rawls on Value and Justification

After reading the comments on a previous post, it has become apparent to me that some people who take seriously the subjectivity of value and the incoherence of free will are led to the position of moral nihilism, or at least something very close.  While I do believe that value is subjective and that free will is incoherent, I am not a moral nihilist.  Moreover, I do not think that the validity of either of the aforementioned positions commits one to moral nihilism.  In the comments on my last post I recommended that someone read John Rawls’ first published paper, "Outline of a Decision Procedure for Ethics" because I think Rawls’ theory of justification is neither contingent upon the objectivity of value nor upon the possibility of free will.   I do not expect that I will convince any reader of this blog to become a Rawlsian by reviewing the paper, but I do hope to make the case here that a coherent account of moral justification is not contingent upon the objectivity of value.  I plan to follow this post with a subsequent account of how contractualism is coherent even conceding the problem of free will.

Ethicists often cash out different moral theories in terms of the "right" and the "good."  The standard view is that in deontological systems (e.g. Kantian moral theory) the right is logically prior to the good because the consequences of an action can only be considered to have positive value (good) insofar as the action itself was right, meaning that the agent complied with a moral maxim (most commonly Kant’s categorical imperative), had the proper moral motivation, etc.  In consequentialist systems (e.g. J.S. Mill’s utililitarianism), the reverse is true:  The good is logically prior to the right, meaning that the rightness of an action can only be determined in light of the value (good), that the action is expected to promote.  The thesis that value is subjective- that nothing is intrinsically or objectively valuable- poses two problems for consequentialism.  First, if value is subjective, then any particular value is unlikely to be universally held.  If a value is not universally held, a rule to promote that value cannot be justified on the grounds that it promotes a "common good" because there is no such common good.  Second, even in those rare cases in which each individual values the same or some similar object, the universality of a value does not translate into the intrinsic value of an object.  For example, it may be universal that each of us values our own happiness, but this does not make happiness itself an intrinsic good- it is perfectly compatible with me valuing my own happiness and you valuing your own happiness that we wish each other extreme despair.

Most contemporary philosophers* consider Rawlsian contractualism to be a type of deontological moral theory,  not a form of consequentialism.  This would arguably insulate contractualism from the threat of value subjectivity, however it still begs the question of how to ground a system of moral right objectively.   In the Decision Procedure paper, Rawls concedes the problem of value subjectivity and sets his sights on finding an alternative route to grounding objective moral principles.  He explains his project as follows:

"the objectivity or the subjectivity of moral knowledge turns not on the question whether ideal value entities exist or whether moral judgments are caused by emotions or whether there is a variety of moral codes the world over, but simply on the question:  does there exist a reasonable method for validating and invalidating given or proposed moral rules and those decisions made on the basis of them?"

Rawls, of course, thinks the answer to this question is yes, and he spends most of the rest of the paper articulating the method for validating and invalidating moral claims.  It isn’t an elegant work, but the decision procedure he develops in the paper is the seed of arguably the greatest work in metaethics of the 20th century.  One fundamental feature of the decision procedure is Rawls’ description of moral reason and moral reasonableness because, in the contractualist view, the principles of justice are those rules which a group of reasonable judges would consistently determine under specified conditions* of deliberation and reflection.  What is striking about this view, much more than Rawls’ specific conclusions, is the argument that Rawls uses for the validity of appealing to the "reasonableness" of a hypothetical judge.  The argument turns on an analogy drawn between the empirical sciences and moral inquiry.  Rawls says:

"the present method of evidencing the reasonableness of ethical principles is analogous to the method used to establish the reasonableness of the criteria of inductive logic.  In the latter study what we attempt to do is to explicate the full variety of our intuitive judgments of credibility which we make in daily life and in science in connection with a proposition, or theory, given the evidence for it.  In this way we hope to discover the principles of weighing evidence which are actually used and which seem to be capable of winning the assent of competent investigators."

Rawls’ notion of  "moral reasonableness" in ethics is analogous to epistemologists’ notion of epistemic virtue.  In both cases, the substance of the view- actual moral values or views about what constitutes knowledge- is much less important than identifying those who are committed to justifying their beliefs and actions, moral or otherwise.  A commitment to reasonableness may, of course, be seen as a value itself, but contractualism does not hinge upon the objectivity of this value any more than epistemology hinges upon the objective value of truth.  This is because the moral rules themselves do not derive from a foundation that specificies that reasonablenss is objectively valuable, only from a foundation that specifies that reasonableness is a prerequisite for for determining moral principles.  This is analogous to the observation that certain sensory perceptions may be a prerequisite for having knowledge about the external world but having these sensory perceptions does not, by itself, constitute knowledge.  This opens the possibility of grounding objective moral principles (rules) without postulating the existence of any objective moral value.

*I feel compelled to note here that some critics of Rawls have argued that his Theory of Justice amounts to a consequentialist political theory because certain political (property) rights are justified on the grounds that they are likely to promote a certain outcome.  I think the substantive conclusions of Rawls’ theory are much less important than the metaethical project of contractualism, but I leave it to the reader to research right-wing criticisms of the Difference Principle.

**Over the course of Rawls’ work, these conditions eventually evolve into what is known as "the original position."

Add to FacebookAdd to DiggAdd to Del.icio.usAdd to StumbleuponAdd to RedditAdd to BlinklistAdd to TwitterAdd to TechnoratiAdd to FurlAdd to Newsvine

Advertisements

18 Responses to “John Rawls on Value and Justification”

  1. Michael Dunne Says:

    ‘Michael,

    You are certainly entitled to your own view of what is “philosophically interesting” but I am inclined to agree with Jim that it is inappropriate to discount the work of many, many brilliant contemporary philosophers on the grounds that it is not philosophically relevant. Personally, I cannot think of any philosophical question that is more relevant (or interesting) than the question of what I should do, and the fact that many philosophers leave open the possibility that they may discover a satisfactory answer to this question is not proof of their incompetence or lack of philosophical depth. I have made a post about the Rawlsian view of justification and how it corresponds to value subjectivity. I plan to make a second post about the relationship between Rawlsian justification and free will. If you are not personally committed to justification (epistemic or otherwise), I expect that you will not find the Rawlsian account compelling, but I hope that you will at least understand what I mean when I say that there are coherent moral theories which are not contingent upon the existence of objective moral values. I welcome your comments on the Rawls posts.

    -Liza’

    Liza,

    I am not entitled to my view of what is ‘philosophically interesting’ unless I can support my claim with arguments when challenged. It just so happens that I can and have. You are not entitled to agree with Jim unless you present better arguments than him i.e. better arguments than no arguments. I’m pleased to see that you have done so and I now have no pressing objection to reading the Rawls article, even though at present I don’t find your arguments particularly persuasive. I have read it and am currently working on the presentation of criticisms.

    ‘Personally, I cannot think of any philosophical question that is more relevant (or interesting) than the question of what I should do’

    Sorry but personally I cannot think of any philosophical question that is more incoherent.

    ‘and the fact that many philosophers leave open the possibility that they may discover a satisfactory answer to this question is not proof of their incompetence or lack of philosophical depth.’

    Actually it seems to me exactly that, since what else could explain how they can believe that a concept shown to be incoherent might one day be shown coherent after all. Logic doesn’t work like that, as you surely know. However I will concede that a lot of philosophers who are surprisingly wrong about free-will are right about other things. I think in any case that this debate about ‘philosophical interest’ has dragged on long enough. If it satisfies everyone I’d like to forget about it for the moment.

    I appreciate your invitation to comment in this thread but there are still unanswered objections in the previous one – the objections to your claim that preference-talk is as incoherent as moral-talk for instance. Please address these when time permits. Sincere thanks for your presentation of Rawls and contractualism.

    Michael

    • Michael Dunne Says:

      On a recommendation from Jack I will elaborate on this –

      ”‘Personally, I cannot think of any philosophical question that is more relevant (or interesting) than the question of what I should do’

      Sorry but personally I cannot think of any philosophical question that is more incoherent.’

      It will be easier to just post the following extract from a conversation we had, which I do with permission.

      Jack Angstreich: there might be one point that you want to clarify

      michael.dunne_: which one?

      Jack Angstreich: you should explain that “what should I do?” is a serious question if it refers to a hypothetical imperative

      Jack Angstreich: but that is not a philosophical question

      michael.dunne_: i was playing with her equivocation – i’ll spell it out but i’ve already done so several times

      Jack Angstreich: yes

      Jack Angstreich: I think it’s worth doing

      Jack Angstreich: how could one answer the question “what should I do?” independently of prior commitments?

      Jack Angstreich: that is the dream of liberalism

      Jack Angstreich: and the dream of rawls

      Jack Angstreich: but there cannot be any such answer

      Jack Angstreich: logically

    • Jim Says:

      I doubt Liza will be happy with this response, but it needs to be done.
      Listen up, you dense moron. I made an argument for the only positions I asserted. I explained why your position about what is and isn’t philosophically interesting is wrong. I also explained why your attempted deal was just stupid. Your desire for me to make an argument for Liza’s position was ridiculous, and I refused. It was Liza’s position, and it’s not my job to make her arguments for her.
      If you have something to say to me, address it to me. Some members of Jack’s crew, including you, seem to share a passive-aggressive tendency to take shots at others without directly addressing them. It’s small, and it’s pathetic. Do it again, and you’re banned.

    • Liza Says:

      I am amused that a person who doubts the possibility of evidential objectivity, not to mention value objectivity, could say something like “I’m not entitled to my view of what is ‘philosophically interesting’ unless I can support my claim with arguments.” This seems absurd to me. Do you not think that “philosophical interest” is a subjective preference? If it is, then it is obviously not the sort of claim that can be supported with arguments. That is like saying ‘I am not entitled to be in love with so-and-so…’ or ‘I’m not entitled to have my taste for chocolate ice cream, unless I can support my claim with arguments.’ If ‘philosophical interest’ is not a preference, then what is it? You can’t mean ‘what professional philosophers find interesting’ because, as I have already observed, that question IS interesting to them. So, I’m stumped. If “philosophically interesting” is an objective label, please give it a definition for me, otherwise, I think you’ll have to concede that what you said is mistaken.

      Also, while we’re on the subject. What do you think it is that philosophers do? Or, to put it another way, what dispositions, actions, or values make a person into a philosopher?

      As to your other question, my point was only that, given that we do not have free will, our intuitive understanding of what it is to have a preference is as problematic as our understanding of what it is to have a moral obligation. In both cases, there is no counterfactual possibility that makes terms like “would choose” or “should choose” meaningful. Preference talk does no more explanatory work than value talk and both could easily be replaced by a more predictively accurate, refined neuroscience. It is as meaningless to say “I grabbed the chocolate bar because I love chocolate” as it is to say “I answered honestly because I think lying is wrong.” In both cases, my own awareness of my preferences/values seems causally relevant to what I did, but it isn’t causally relevant- at least not in the special sense that we intuitively feel that is- because I couldn’t have done otherwise. That being said, of course, I still do think these terms are quite coherent. In the big picture, I have no choice about what I will pick up to eat at the grocery store or whether I will be killing or saving someone tomorrow because I have no choice about what my values and preferences are, but my awareness of my values and preferences gives me an understanding of what I am likely to do.

      I hope this clarifies my point.

  2. Michael Dunne Says:

    You obviously prefer a more direct form of aggression. Frankly I’m starting to get tired of it. No, I have nothing to say to you at the moment. If you want to ban me there’s not much I can do.

  3. Michael Dunne Says:

    The comment in my original post was a cheap-shot. I apologise.

  4. Michael Dunne Says:

    Liza I have just noticed your post now. I’m going back to bed and will respond tomorrow. Thanks.

  5. Michael Dunne Says:

    Liza,

    It seems to me that hypothetical imperatives don’t have anything to them that would necessitate them being considered part of philosophy rather than, say, politics or economics or sociology. My claim was that since hypothetical imperatives cannot have any ground independent of potential interests one has in promoting this one or that, they are thus philosophically irrelevant, but may indeed be of great political interest. If you have a broader definition of philosophy than me that’s no major worry – maybe we can argue about it sometime.

    The issue I thought pressing was simply whether one could have a moral system ‘grounded’ on hypothetical imperatives. I took this to entail the irrelevance of the article you recommended to anything I had claimed. My demand that one justify what they consider ‘philosophically interesting’ in this case is just a demand to justify the assertion that an article on contractarianism, being premised on hypothetical imperatives, is relevant to the claim that it is impossible for a moral system to be normative or ‘grounded’ in any way. So I think this issue is clearly a red herring, which is why I wanted to forget about it. I will reply to the second part of your post soon.

  6. James Gray Says:

    Why can’t Rawls’s epistemology be consistent with Nihilism? You seem to think of Nihilism as being an uneducated view by definition, but I would think an error theorist could agree with what you said and I would consider the theorist to be a nihilist.

    I would consider that subjective value could be intrinsic. I don’t see subjective value and intrinsic value to be at odds. By “subjective value,” do you mean that value has to be an illusion?

    • Liza Says:

      James,
      I think that you may have misunderstood me or the context of this blog. I do not think moral nihilism is an uneducated view at all. On the contrary, I take it quite seriously. I wrote the post on Rawls in response to comments that had been made on one of my previous posts by Michael Dunne. I thought that Michael was conflating two issues which pose separate threats to moral realism: the subjectivity of value, and the problem of free will. I believe that Michael considers himself to be a moral nihilist, which is a perfectly defensible view. However, he suggested that the problem of free will makes moral realism incoherent. I don’t think this is true. I offered up Rawlsian contractualism as an example of a moral theory which can make sense of moral prescriptions without presupposing free will or some universal objective value. I decided to do two posts to address each issue separately, but I have only completed the first, so far. The follow-up will be about contractualism and free will.
      As to your question about whether Rawls’ epistemology is compatible with nihilism, my answer is a qualified ‘yes’. An error theorist might agree with Rawls’ descriptive account of what we take to be a ‘reasonable’ person. He might also agree that a contractualist account of moral language makes the claims truth-apt, at least within the parameters of contractualism. That being said, error theorists are rather tame, as nihilists go. If you are a nihilist about Truth or Knowledge (as some readers of this blog appear to be), then I think you would have to reject Rawls’ epistemology. After all, you can’t say that you know what a reasonable person is if you don’t think we can know anything.
      It is certainly the case that something which is intrinsically valuable could (and probably would) also be subjectively valuable, but that still begs the question about whether anything is intrinsically valuable. It’s easy to make the case for intrinsic value if all you mean is that some object is intrinsically valuable when some person values it as an end rather than as an instrument for some other end. In that case, yes, there are lots of intrinsically valuable things and their value is also subjective. Some people don’t mean that, however. Utilitarians, for example, believe that the intrinsic value of pleasure means that we all have reason to promote pleasure whether it is our own or that of someone else. So, for them, when something is intrinsically valuable it’s not just an end in itself but also universal. I am skeptical about anything being intrinsically valuable in that sense of the word, but feel free to offer up a persuasive counter-example.
      I don’t think that value is illusory at all. I am puzzled that you would infer that from my posts and also curious. Could you point me to the bit in the post that made you think that?
      Also, thank you for your thoughtful questions. Please let me know if I haven’t answered satisfactorily.

      -Liza

  7. Michael Dunne Says:

    Sorry for the long delay – I haven’t forgotten about this thread. James, are you talking to Liza or me? I’ll get round to posting sooner or later.

  8. James Gray Says:

    I was posting to Liza concerning the original anti-nihilism post.

  9. James Gray Says:

    I’m still not sure what you think nihilism is. I would consider any anti-realist to be a nihilist to some extent. I would also be tempted to think of anyone who rejects intrinsic value to be an anti-realist.

    How can we determine whether or not a “subjectively valuable” object is intrinsically valuable? If we take value to mean “instrumentally valuable,” then nothing truly ethical is being said. How can we talk about noninstrumental value that lacks intrinsic value? This is a third category that you seem to agree with.

    Utilitarians of the sort you mention that see pleasure as intrinsically valuable will say something like the following: My pleasure is good no matter what anyone else believes about my pleasure. That fact, if accepted, would prove that it has intrinsic value. Other people’s pleasure also has intrinsic value for the same reason.

    We know what has intrinsic value because of subjective experience. Subjective experiences can have intrinsic value. At what point do we want to say, “This subjective experience has subjective value,” and what exactly will that mean?

    I think part of the problem is that people tend to think subjective experiences are some kind of illusion. “Subjective value” sounds less real than “objective value.” Subjective experiences can only have “subjective value” because they are less real than other objects. If we see subjectivity as an equal part of reality and reject a subjective-objective metaphysical dualism, then the thought that subjective experiences have intrinsic value is not strange, and the notion of “subjective value” becomes a puzzling one.

    • Liza Says:

      James,

      If nihilism comes on a sliding scale then I would imagine that a great many of us are “nihilists to some extent,” maybe even you. Normally, I take nihilists to be those who reject the possibility that there is any legitimate foundation for belief. So, moral nihilists reject the possibility that there could be any legitimate foundation for moral prescription. I’m not sure that all anti-realists reject this possibility, but it’s futile to argue about it without defining anti-realism first, as many people who use the term seem to mean different things. In my view, a moral anti-realist is anyone who holds that there is no entity (e.g. God or Forms) or fact about the world (e.g. a fact about intrinsic value) that makes any moral principle true by necessity, and under that definition I would consider myself a moral anti-realist. That being said, I think that the process of justification itself provides a legitimate foundation for moral prescription. By “process of justification” I mean the way we make claims and give reasons to others and also to our recognition of the fact that others will make claims upon us and have their own reasons for doing so. This is an anthropological phenomenon. If we had evolved differently we might not have a justification process or values at all. So there’s nothing necessary about it- it’s an accident.

      I tried to make my view on intrinsic value quite clear, so I’m not sure if you misunderstood what I said or if you just think that I’m wrong about the meaning of the term. My view is that plenty of things are valuable as “ends in themselves” to some person, so they are non-instrumental, but this does not mean that that those things are valuable to anyone else. I don’t think that there is any object that every person must value and promote, and I think that’s what you mean to describe when you postulate something that has “intrinsic” value. The usual candidates for this object would be God, the Good, the Good Will, self-perfection, or human happiness. I deny that the first three exist, and I don’t think anyone must value and promote the latter two, though it is a happy accident that most of us do value and promote at least one of the five. So, it’s fine with me if you want to say that that I deny intrinsic value, just as long as we are clear that I don’t deny that individuals desire some objects as ends in themselves.

      I think all value is contingent upon us. Objects are only valuable insofar as we value them. Now, human beings are remarkably similar creatures, and so we tend to share values, and it may be the case that certain accidents of our biology make us so similar that we happen to share some values universally. I am fine with saying that if something is subjectively valuable universally then it is also objectively valuable, so long as it is understood that this objective value is contingent upon an accident. As I am sure you are aware, Kant says something similar about our understanding of the empirical world, though I don’t think he called it an accident.

      Also, since I already mentioned Kant: I am curious, do you think that the truth of deontic rules about write and wrong is contingent upon intrinsic value? If you think that the Categorical Imperative is prescription based upon some prediction about bad consequences, then I would guess that you do think that, but many Kantians would disagree.

      • James Gray Says:

        Liza,

        I don’t understand your idea of “valuable for x.” This is why I called “subjective value” a third category and discussed it at some length. Does it mean x believes it is valuable? Does it mean that it benefits x? If it means it benefits x, then don’t we need an objective (intrinsic value) meaning of “benefit?” For example, life or pleasure could be taken to have intrinsic value. If life or pleasure has no intrinsic value, then it might benefit me just as much to die and experience no pleasure as the opposite.

        I think many people would take “subjective value” to be like an intrinsic value in a relativist sense. Everyone lives in a different reality, so “valuable to x” would mean “valuable in x’s reality.” This sounds absurd. A more plausible answer was already given: “subjective value” is taken by many philosophers to mean a kind of value less real than an objective value. What is subjective isn’t factual to people like David Hume.

        I don’t see Kant as necessarily requiring moral realism. We can just accept the categorical imperative if it we can rationally accept it, perhaps as an anthropological fact as you mentioned. He seemed to justify ethics on pragmatic grounds: It seems that we have to be ethical to enjoy our lives. However, the categorical imperative doesn’t make sense to me. Why does everyone having a behavior make it specifically morally relevant? The categorical imperative is self-promoting. You shouldn’t murder because then we couldn’t all follow the categorical imperative anymore. But, so what? Morality or the categorical imperative itself might not matter at all.

        The strong intuition of moral realism (for me) is that somethings seem to really matter. They are really important. What I subjectively care about is what I believe has objective intrinsic value. If I found out that human beings have intrinsic disvalue, then it would be a lot easier for me to stop caring about them. Emotions and a pro-attitude seem to be tied to moral realist beliefs. Having a pro-attitude “just cuz” doesn’t make sense to me.

  10. Three Loan Wolves » Blog Archive » Non-aggression principle Says:

    […] John Rawls on Value and Justification « Apple Eaters […]


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: