Tuesday, February 21, 2006

Subjective Morality

Doctor Logic purports to write about morality, but I can't help but think that he's merely changing the subject. What his view really implies is that there is no such thing as (normative) morality. He writes about something called "subjective morality", but this appears to be synonymous with "arbitrary personal preferences", so we might as well just use the latter phrase and avoid any terminological confusion or unclarity. Anyway, I just want to quickly respond to some of his main points:

1) Even without objective morality, human societies will form "social contracts" and behave, well, just as they actually do.

True. Morality is not taken to be a causally efficacious part of the world. (Though it might make a difference if people believe in an objective morality, as it might deter them from defecting on the social contract in instances where it would be personally advantageous. Beliefs certainly are causally efficacious.)

2) "History teaches us that social morality changes. Slavery was once good."

No, slavery was once believed to be good. Compare: "History teaches us that social astronomy changes. The universe was once geocentric." This confuses belief with reality. If by "social morality" you simply mean the codes of conduct widely accepted in a society, then "social morality" has very little to do with normative morality, which is what we're interested in here. (See also my posts on society and morality, and especially moral diversity and skepticism.)

3) Subjective moral progress is possible.

In other words, it is possible for the world to become closer to how I want it to be. That's certainly true, but not very interesting. No genuine progress is possible on such a view. The world isn't made better by ridding it of racism and genocide, but merely "more to my taste". After all, racists might think the civil rights movement was positively bad. (Obviously they are mistaken, but we can only recognize this truth if we are moral objectivists.)

4) "How would you determine whether an action was universally moral?"

Well, there's this field called "moral philosophy", where people reason about such things. Some questions are harder than others. Consider the following:
You have the option of whether or not to press a button. If you press the button, then a roomful of innocent children will be subject to torturous agony. If you don't, then nothing happens. There are no other significant differences between the two outcomes. What is the morally best option?

That is not a difficult question. More generally, the moral facts are determined by facts about human welfare, which themselves supervene upon natural facts. (We don't always have access to all these facts, of course, but that doesn't mean they aren't there.)

5) "Suppose there were some way to know that a moral code was universal. Such a recipe would still rely upon subjective moral appeal!"

Yes, we've already granted that human behaviour is a result of psychological facts, not abstract metaphysical ones. See #1 above. Append "So what?"

6) "We have no reason to expect that universal moral laws would be appealing to humans"

On the contrary, I think there is a conceptual connection between morality and human welfare. DL asks us to imagine that moral experts discover that it's immoral to eat tasty food. But unless there is some reason behind the moral claim (e.g. that eating tasty food would somehow cause great harm to others), his scenario is conceptually impossible. You might as well ask us to imagine that baldness has nothing to do with hair.

7) "a universal moral law is nothing without a universal regime for compliance."

So practically-minded! We have social norms and criminal laws to cover such matters; it's not the job of the moral facts themselves to do the policing. Rather, they're there to make true statements (such as "slavery is wrong", and "genuine moral progress is possible") true. That is, we need to posit moral facts for theoretical, not practical, reasons.

8) "Advocates of universal moral law like to argue that we must accept the existence of universal moral truths, or else the things we regard as subjectively immoral would be acceptable. This claim is contradictory because the claim acknowledges the value of subjective morality..."

That's a terrible argument. (To illustrate, one could run an analogous argument with "physical" in place of "moral".) The point is not that whatever I believe must be true, therefore we should posit objective facts to match my arbitrary beliefs. What an absurd suggestion! The claim is instead that some particular beliefs themselves (e.g. "slavery is wrong", etc.) are true, and thus need a matching fact as their truthmaker. Note that it is the content of the belief that makes it true (if indeed it is), not the arbitrary fact of my believing it.

9) "Ironically, it is authoritarianism that poses the greatest threat to our subjective moral good"

I agree. But we shouldn't conflate dogmatism with objectivism. Relativism is not the way to avoid dogmatism. (Quite the opposite, in fact.) I've explained all this before.

10) There's a theistic moral objectivist commenting on DL's post. For the record, I disagree with much of what he says. My post on 'God given value' explains why.

Okay, that's all for now.


  1. Doctor Logic writes: 4) "How would you determine whether an action was universally moral?"

    That strikes me as a fair question.

    Richard responds: Well, there's this field called "moral philosophy", where people reason about such things.

    That strikes me as snide, at best. There are plenty of fields which we do not consider valid, and instead consider fraudulent, pseudoscientific, or whatever.

    Some questions are harder than others.

    Sententious platitudes: is that stock in trade for philosophers?

    [example omitted]: What is the morally best option? That is not a difficult question.

    Here Richard asks us to nod yes that his example has validity because we may have a gut feeling. Not because he has an argument. This may be traditional philosophical mummery, but this is the 21st century: we know now to ask how our gut feeling relates to UNIVERSAL morality.

    More generally, the moral facts are determined by facts about human welfare, which themselves supervene upon natural facts. (We don't always have access to all these facts, of course, but that doesn't mean they aren't there.)

    Whether or not moral facts exist and have the relationships you claim, Richard, you haven't established that they are universal rather than particular. And even if they were universal, you still haven't shown how to determine if an action was universally moral by those facts.

    In short, Richard, you haven't answered the question in any way.

  2. Erik, I haven't engaged in moral philosophy. I've been engaging in iconoclasm.

    It is totally fair to ask that Richard explain how he knows some truth such as the "objectivity of moral facts". Indeed, if they are objective, is there any reason why I shouldn't be able to perceive them also? Putting a burden of disproof on others is absurd. The person making the positive claim bears the burden.

    Sure there are objective truths about human beings, including our psychology. That doesn't establish that Richard is talking about them when he uses the term "moral facts". And it's far more distant from "universal morality".

    Don, I am reminded of a Saturday Night Live skit with Gilda Radner where the characters were children telling how to solve the worlds problems. The answers were essentially "grow up, then solve the problem (in some way they don't explain.)" And that's what I think you're saying with "Reason about it."

    Richard is making an extraordinary claim, which ought to have some backing, if not extraordinary proof. Perhaps such things are a dime a dozen in philosophy, but if so why not accept my theory that there are little demons inside your head that make you feel these things. If you need another 50 or so such theories, they're really easy to make up. Erik can take up the burden of disproving them all.

    I don't know why you accuse me of throwing out gut feelings: I wrote "we know now to ask how our gut feeling relates to UNIVERSAL morality." In other words, I'd expect a theory of universal morality to provide an explanation of gut feelings that matches our knowledge of them.

    It's obvious that intuitions conflict: there is no intuition you can name that is universal. Some percentage of people at some stage of their lifetime will disagree. There's a moral fact for you: there is no universality of moral intuition. So how do you proceed to discuss universal morality in spite of this fact?

  3. Richard,

    I'm thrilled that you took an interest in my blog post (I visit here often). Still, I was a bit surprised at your strongly-negative reaction. I think that in most contexts we would be in agreement. For example, I thought your post on "God-Given Value" was quite excellent. In fact, my post was directed at similar sorts of moral claims.

    I'll set aside those things that I think we agree upon, namely, the descriptive aspects of morality, e.g., the way moral codes and social contract work in practice, and the relevance of personal taste as a determinant of human welfare.

    I think our disagreement is about the nature of moral theories.

    We might think of morality as an algorithm that acts on the facts of circumstance and personal taste to produce a set of instructions for behavior.

    I would call such an algorithm objective when it is obtainable without reference to the data it is intended to act upon. That is, we should be able to determine what an objective morality is without considering the particular circumstances and personal sensibilities upon which it is supposed to act. This does not imply that we should know ahead of time what behaviors are expected of us (as simplistic religious doctrines often do).

    My claim is that the ideal function of this algorithm (that is, what the algorithm maximizes) is beyond objective determination as I define it. This is because what is "ideal" is itself subjectively determined. Your recent post on longevity illustrates this point well because it is not clear what metric should define an optimal life ("shape", quantitative joy, or some other metric).

    There's a long list of proposed algorithms or moral principles. We try to decide among the long list of candidate algorithms on the basis of our subjective revulsion to their respective consequences in thought experiments. Much like the button-pushing thought experiment you cited.

    To confuse the issue further, our very consideration of these algorithms changes our subjective sense of right and wrong. Personally, I think it's morally good to allow consideration of such algorithms to affect our personal, moral sensibilities, but I don't think it aids objectivity!

    You compared my statement about past morality with social astronomy. Physics and celestial mechanics are indeed objective.

    However, I think that morality is more like economics than physics. Economics has some objective elements, e.g., metrics of economic activity like housing starts and the inflation rate. We can also enumerate ideal economic algorithms like free market economies or controlled economies that maximize different things, e.g., growth vs. stability.

    However, we would not claim that there is one universal or objective system of economy. We may find that a free market economy objectively maximizes certain economic metrics, but who is to say that those economic metrics are the right ones to maximize?

    Likewise, we can study real life morality and social contract (individual and statistical moral norms), and we can enumerate models of ideal moralities (ethical systems), but I don't see on what grounds we could possibly determine that there is a single, objectively-best moral algorithm (i.e., we cannot objectively determine what metric to optimize).

    Also, I think that the moral status of slavery isn't as invariant as you suggest it might be. As I said, I think a lot of moral argumentation proceeds on the basis of case studies that have relatively unambiguous moral status for the reader. Today, an argument that demonstrated that a moral principle would lead to slavery would produce strong condemnation of the moral principle under test. But a couple of hundred years ago, the argument might be much less convincing, and not simply so because the reader would have an inferior understanding of the sciences.

    My question to you is this: If there are objective moral principles, shouldn't those principles be obtainable without reference to subjective moral opinion (e.g., emotive case studies)? If not, why should we regard morality a branch of philosophy as opposed to, say, anthropology or economics?

    P.S. Note that I'm not arguing against your thesis in "Society and Morality". What a society collectively thinks is moral does not translate into a "should" for any individual. A society in which slavery is acceptable does not make slavery good to an individual observer, no matter how large that society may be.

  4. Hi Doc, thanks for dropping by.

    I'm not sure I understand what you mean by an algorithm "obtainable without reference to the data it is intended to act upon". Could you unpack that a bit more?

    I think we need to distinguish moral intuitions from subjective emotional reactions. For example, one might find various forms of sexual "deviancy" to be personally repugnant, and yet recognize that there is no reasonable basis for considering such practices immoral. There seem two quite distinct forms of judgment going on here. (It's true that we assess moral theories on some intuitive grounds, so there is a "subjective" aspect to what we choose to believe; but the same is true of every subject whatsoever. I believe in an objective external reality because that seems more plausible to me. Same goes for moral reality. Perhaps one or other of my intuitions is misguided, and that's where philosophical debate comes in.)

    Intuitions can also be overriden by systematic theory. For example, mostly everyone in Western society grows up believing that (even consensual) incest and cannibalism are wrong. But once we recognize the plausibility of utilitarianism as the true moral theory, and the liberal principles that best implement it, then we might begin to question and even reject those prior intuitions. (I also agree that there are possible situations where slavery would be - objectively - right. But those are rare and I exclude them for simplicity's sake.)

    Regarding the diversity of values, and how this relates to moral principles, I would point you to my post on Objective Moral Relativism (which I think you might find congenial -- like the 'God-given value' post, I no longer entirely agree with it myself), and also of course the 'moral diversity' link in the main post above.

    So, to answer your core challenge: "If there are objective moral principles, shouldn't those principles be obtainable without reference to subjective moral opinion (e.g., emotive case studies)?"

    Any argument will have to appeal to foundational premises at some stage, so if that's all you mean by "subjective opinion" then we can't obtain any conclusions, in any field, without it. If you had a stronger notion of subjectivity in mind (e.g. 'emotional repugnance', etc.) then I think we can do just fine without that, as seen through my earlier distinction.

    Perhaps I would do better not merely to talk about how one might reason about objective morality, but to actually illustrate such reasoning in action. In that spirit, let me direct you to my post on Consistency and Utilitarianism, which proves that (some form of) utilitarianism is the correct moral theory :-)

  5. Richard,

    I'm not sure I understand what you mean by an algorithm "obtainable without reference to the data it is intended to act upon". Could you unpack that a bit more?

    We agree that that we are looking for a way to maximally fulfill the desires of the populace.

    But should we define fulfillment as joy minus suffering? Or the joy minus the suffering squared? Or just minimize suffering? Or do we consider the joy minus suffering of the final state? Or do we consider the average joy minus suffering per capita integrated over time?

    There are an infinite number of possible formulas (or algorithms) we might construct for evaluating how best to fulfill the desires of all persons.

    Unfortunately, there seems to be no objective basis for making this selection. We cannot even appeal to people's desire to use one formula over the others because there's no objective way to assess how to tally those desires for preference for formula.

    Let's differentiate this from a field like physics which is generally considered to be the study of something objective. Physical theories are rated according to their predictive power. There are many metrics of predictive power (availability of experiment, consistency with experimental results, ability to compute predictions, etc.), but we all agree on which way is up. For example, it's better for a theory to be consistent with experiment than not, to make more predictions than fewer, to be testable in an experiment tomorrow rather than in 1000 years, to have a simpler formulation than a more complex one, etc. The search for scientific explanations is like looking for high ground in multi-dimensional space: we may disagree on which direction will takes us higher faster (which is the steepest approach to the peak), but we all agree on which way is up.

    In contrast, I'm not sure we can say the same of moral optimization. We simply don't have a universal appreciation for what represents an optimum moral situation. As with physics, the question isn't whether we can presently identify the ideal solution, but whether we have the objective ability to know when we're moving in the right direction.

    My claim is that we cannot know objectively whether we are moving in the right moral direction (toward the peak) or not. Though we might all agree generally that more joy is good, and less suffering is good, we really need to know when to forego joy to reduce suffering, and when the ends justify the means. We don't seem to be able to agree on this, no matter how well informed we are and no matter how critical our thinking.

    A successful physical theory arrived at by mistake (e.g., in spite of initially flawed and contrary experimental results) is undiminished by the means of its discovery. Morality is not a simple search algorithm because making the wrong move can change a moral peak into an immoral valley.

  6. 1) in any group there is usually some agreement on what is moral and not moral.
    2) we are far from optimal utilitarianism (there are low hanging fruit that are good by almost all utilitarian framework - ignoring complex concequences).

    these things might fall apart when we talk about very large time/space scale objectives but at least there is a starting point.

    The good thing about utilitarianism with certain universal moral principles is you can get at least minimal "buy in" (in theory) from anyone. Religions are a problem in that regard and theories with no universal moral position just dont offer much anyway.

  7. Erik I am definitly not arguing against religion here.
    What I am saying is that I cant expect to ghet "buy in" from a muslim (lets say) via arguing using quotes from buddah.
    But I might get somewhere using a utilitarian argument (possibly baised on a sort of humanitarian grounding).

    In that regard I could personally have a religious position but see it as useless to argue it on that ground.


Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)

Note: only a member of this blog may post a comment.