About Rationally Speaking

Rationally Speaking is a blog maintained by Prof. Massimo Pigliucci, a philosopher at the City University of New York. The blog reflects the Enlightenment figure Marquis de Condorcet's idea of what a public intellectual (yes, we know, that's such a bad word) ought to be: someone who devotes himself to "the tracking down of prejudices in the hiding places where priests, the schools, the government, and all long-established institutions had gathered and protected them." You're welcome. Please notice that the contents of this blog can be reprinted under the standard Creative Commons license.

Monday, November 07, 2011

Ethical Pluralism: the ugly theory that could

by Ian Pollock

Warning: this post is long, and assumes basic knowledge of ethical philosophy 101 — you need to know and care what deontology, consequentialism and virtue ethics are, and have a basic grasp of the arguments for & against all three. In other words, this is a post for ethics geeks.

“In practice, virtually everyone seems to judge a large matter of principle to be more important than a small one of pragmatics, and vice versa — everyone except philosophers, that is.”
(Gary Drescher, Good and Real)

We like systems of thought, and we like them elegant and simple. This applies as much to ethics as to any other topic, and leads us to prefer simple ethical theories to complex ones. However, it may be that ethics (bound up in messy, evolved human preferences) is one place to expect complications in abundance. It may be that no simple principle can capture the whole of our ethical lives.

Let me see if I can make this more plausible. I will be talking in terms of the “big three” ethical theories — consequentialism, virtue ethics and deontology, because most ethical statements humans make seem to be grounded in thinking that comes from one of those three broad areas, though only a minority are aware of the vocabulary itself. (As an aside, I am also studiously ignoring a large portion of what some consider morality — what Jonathan Haidt labels “purity” concerns.)

First, I wish to establish a “bare minimum” — that consequences, virtues and justice do matter, at least whenever all other things are equal. In other words, for example, a consequentialist is also a deontologist in the sense that, given identical consequences, they’ll usually choose the same thing a deontologist would. Consider the following scenarios:

Scenario #1: You are considering donating with a limited budget to various charitable causes. Eventually the field is narrowed to two charities. According to an independent organization evaluating charities’ effectiveness, one saves approximately one life per $600, the other, one life per $2,000 dollars.

Solution: Choose the more effective charity.

Lesson: At least ceteris paribus, numbers matter in making moral decisions.

Scenario #2: A front-end car collision has occurred, and the two drivers involved are both bleeding profusely. You witnessed the accident and are the only one on the scene. The first driver caused the accident and appears to be drunk; the second was not at fault. You know you need to apply pressure to the wounds to stem the bleeding, but can only do this for one person at a time.

Solution: Prioritize the person who does not appear to have caused the accident due to their own negligence.

Lesson: At least ceteris paribus, an agent should consider questions of virtue and character in making moral decisions.

Scenario #3: Suppose (counterfactually!) that after decades of debate about capital punishment, it was finally determined by statisticians that, in terms of lives saved, the practice is a wash. In other words, without capital punishment, N innocent people per year per capita die (killed due to the lack of strong deterrent for murderers); while with capital punishment, N innocent people per year per capita die (killed due to wrongful convictions). It is in your power to either continue or abolish the practice. Knock-on effects on society are otherwise negligible.

Solution: Abolish capital punishment.

Lesson: At least ceteris paribus, an agent should personally respect the autonomy and rights of individuals in making moral decisions. It is better to allow an injustice (a murder by someone else) than to perpetrate one (a wrongful execution).

What I hope these scenarios establish is that, no matter how we may differ in which principles (utility, virtue, justice) we take to be overriding, the other principles have (and should have) some prima facie sway in our moral considerations if all other things are equal. If we can get ethical discourse to the point where we are talking not about whether, but about to what extent numbers should trump matters of principle and of character, and vice versa, we shall have made some strides toward a more plausible moral theory — plausible, that is, to people who don’t want to push fat people in front of buses, or tell an axe-murderer exactly where to find (and kill) their friend for fear of lying, or rubber-stamp whatever their society tells them the good life consists in.

Of course, if I am (for example) a consequentialist who uses virtue as a mere tie-breaker in unusual situations, that is not much of a rapprochement. However, let me see if I can carry the stronger claim that no matter which of the three approaches you consider most important, the other two have to have more than only a tie-breaker status. To carry this claim, I have to show that for each of the three approaches, the other two can sometimes override it in principle. This means I will need six counterexamples.

Let us start by undermining the exclusively consequentialist approach.

Counterexample #1 (utility vs character): Julia has mentioned, in a previous post that is one of my favorite on Rationally Speaking, the appalling phenomenon of “crush” videos — in which people derive sexual pleasure from videos of nonhuman animals being stomped to death by scantily clad women. Suppose that the practice of making these videos were now outlawed, but the videos themselves remained. Would it be ethical to watch them if one enjoyed it enough? At first glance, utilitarianism would seem to say it was mandatory — a straightforward net positive with no drawbacks. But most people would say no, it’s not ethical. You can call your enjoyment of such material “utility” if you like, but that enjoyment seems simply unworthy of being valued as a moral end. Damn you and damn your utility — you’re a nasty, twisted person.

Counterexample #2 (utility vs justice): Suppose you are the sheriff of a small town, in which a murder has been committed by a member of a hated minority. You have absolutely no leads, and if somebody is not punished, the townspeople might riot and do violence to random members of the minority. If you jail an innocent person, they will probably be placated. Should you do so? Utilitarianism again returns a straightforward answer of “Yes” here, at least if the risk of violence is high enough. Now in truth, situations like this are very difficult, but most of us feel that there is something appropriate and noble about a law enforcer having an absolute (or close to absolute) prohibition on committing injustice in the service of some good consequence — “fiat justitia ruat caelum.” Maybe the sheriff should scapegoat somebody if the situation is extreme enough and the numbers stack up disproportionately enough — but not at the drop of the hat, as soon as the consequentialist calculus returns even slightly higher utility for scapegoating. This holds even if we caveat that nobody will ever know what the sheriff did.

So much for utility as the sole principle. What about virtue? Counterexamples here are of a somewhat different sort owing to the rather different question virtue ethicists answer: not “what shall I do?” but “what kind of person shall I be?” However, it is worth noting that even a virtue ethicist will eventually have to make particular decisions, and we can still judge them on those decisions.

Counterexample #3 (character vs utility): Sally is a highly intelligent person with a college education in law, who displays gentleness and wisdom in her dealings with others. When she was a young person her father died of an extremely rare disease; she now spends much of her time at a non-profit, attempting to secure funding for the study of that disease by canvassing door to door. Her friends and family love her and find her optimism infectious.

Now, there is much to admire in Sally. She is a thoroughly decent person who is committed to philanthropy, in a world where most are not. However, it is a sad fact that her philanthropic efforts are mostly wasted. She is too concerned about a disease which does not affect many people — much better to focus on common causes of death and suffering. She is also ignoring the law of comparative advantage — her time is replaceable by people with less marketable skills, so it would be much more effective for her to land a well-paying law job and fund other people to perform charitable work. The sad thing is that a person of average income who gives maybe 0.1% of their disposable income to an effective charity and uses the other 99.9% to, I don’t know, buy caviar or something, may do more good in the world than someone who is totally committed to altruism as a way of life, but hasn’t bothered with the numbers.

So, the point here is that a person can be off the charts in all or most of Aristotle’s virtues, and yet still be outdone in improving the world by a grumpy rationalist willing to give a few highly leveraged bucks here and there. Stated more generally, good character is no guarantee that one’s life will have an especially good impact. Or if we wish to reverse it, a very consequentially bad life may not be traceable to bad character.

Counterexample #4 (character vs justice): Here counterexamples spring to mind so readily as to be practically truisms. Thomas Jefferson, an admirable man in so many ways, used slavery to support himself and his family. Most of the great people of bygone times, despite their high personal virtue, were out-and-out racists, homophobes and misogynists. The point is similar to the above counterexample #3 — good character is no guarantee of just conduct, for although the case for (say) women’s rights in terms of justice is a slam dunk, few of even the very best people of bygone times endorsed such a notion. People of otherwise good character seem only to be just to the extent that they have been taught to be, and rarely otherwise. In practice, thinking of morality solely in terms of a virtuous character ends up fetishizing personality, and even a good personality can countenance barbarous things.

What of deontology, then? What of justice?

Counterexample #5 (justice vs utility): All-out nuclear war threatens between Examplestan and the Gedanken Republic. The Republic demands amnesty for three terrorists who blew up an Examplestani post office several years ago. They say that this must occur as a gesture of good faith before they will come to the negotiating table at all. Should Examplestan consider this possibility? The answer seems to be yes — justice is extremely important, but we really don’t want the heavens to fall in that particular way. Sometimes you have to just set aside a matter of principle and take out your goddam calculator.

Counterexample #6 (justice vs character): Meet Fred. Fred has the nearly irresistible urge to torture people, but because he knows that this urge is not universalizable and that he is therefore duty-bound not to do so, he refrains with great difficulty from torturing people. Meanwhile, the idea of torturing people has never even occurred to Penelope, and it wouldn’t tempt her in the slightest if it did. Kant famously argued that the first case shows moral virtue, while the second shows a “mere” disposition. But this seems clearly wrong. For one thing, because a disposition does seem to be morally laudable — it’s just better not to want to torture people than to studiously refrain.

Now, these counterexamples are full of holes, complications and potential objections, but before you make an objection, please ask yourself seriously whether you’re rationalizing. For example, if I recall correctly, Kant was against cruelty to nonhuman animals (despite their not being “rational”) — but only because he bought into the empirical claim that cruelty to animals would eventually lead to violence against humans. Frankly, Immanuel, I call shenanigans. It’s true that violence against animals may lead to later human-on-human cruelty, but that is not what is wrong with torturing a rattlesnake. So when you come up with a clever objection to my counterexample to your-favored-approach, ask yourself if the objection really works in the least convenient possible world.

Now let me see if I can be more constructive.

I used to be a flat-out consequentialist, and in many ways I still consider myself one. I think an ethical theory that fails to relate to what, y’know, actually happens in the world, is essentially meaningless or worse. That said, I now consider consequentialism to be correct in principle, but also an incomplete practical methodology for dealing with ethical questions.

In this respect, it is a similar idea to reductionism. Yes, reductionism is entirely correct; yes, the basic units of physical reality (quarks and leptons?) are probably the only ontologically fundamental constituents of our universe; but no, particle physics is not good research methodology for studying the mating habits of penguins.

Likewise, there are some decisions (like not eating factory-farmed meat and eggs) that are slam-dunks on purely consequentialist grounds. But many others, such as how to train oneself to stand up to human evil instead of sliding into apathy, are much better conceived in terms of the conceptually higher-level ideas one finds in virtue ethics. Moreover, if consequentialist reasoning fails to take into account the undeniable... utility of certain close-to-absolute prohibitions, promises, commitments and concepts of honor and justice, what we end up with is a world where people make decisions that are locally consequentially optimal but globally hideous — consider the tragedy of the commons (I shall have much more to say about this particular issue in future posts).

Consequentialists have long argued that objections to their view end up confirming it — for example, the fact that you wouldn’t want to live in a world in which sheriffs were willing to scapegoat people, shows that that world has lower utility, so that the theory is saved. I think this is half-true, and rather hand-wavy. It shows that we have reasons for endorsing some deontological and virtue-ethical ideas on broadly consequentialist grounds, but it does not show that the consequentialist theory as stated really works as a guide to practice, unless we want to install a thousand ugly caveats into the theory. Sure, humans are made of quarks, but don’t throw out your psychology textbook in favour of physics alone!

So I think that in order to actually make your way in this world as an agent who is sincerely attempting to do the right thing, you’re going to have to use the vocabulary and concepts of consequentialism, deontology and virtue ethics, or your understanding will be impoverished. Exactly how they need to be traded off against each other is a subject for much further thought, assuming these ideas have merit.

In the comments to Julia’s article mentioned above, jcm christened this view as “Ethical Heterodoxy” or “Ethical Pluralism.” Let the critique begin!


  1. Great post, Massimo!

    But, then (as the commentator whose Google account formerly identified me as "jcm"), I'm quite biased towards your position here (and honored by your mention - thanks).

  2. mufi,

    except that the post is by Ian Pollock, not me... ;-)

  3. Woops.

    Same response, but to Ian.

  4. I have a similar view, but I'm usually a bit kinder to consequentialism. You can appeal to consequentialism as an arbiter between moral systems without it being as difficult as deriving psychology from physics. As a simple example, you can ask whether moralized disgust (or strict dedication to retributive justice, or a traditional belief in honor) does more harm than good. This may be an effective way of deciding between alternative virtues, duties, or principles.

    Maybe tolerance, fairness, and a nonjudgmental attitude are more important than purity, because in the end moralization of purity produces little good. Maybe our duty to maintain an orderly society is more important than our drive for retributive justice, because unrestrained righteous indignation leads to cycles of vengeance and bloodshed. Maybe developing the faculties of critical thinking and abstract reasoning is more important than adhering to the form of loyalty known as "faith". I think that these are judgments that we are forced to make on some level, and that are often made unconsciously when we decide which kinds of culture we would rather live in.

    I think that we'd be better off using consequentialism as an explicit arbiter (which at least encourages us to look at the facts of which values produce which actions) rather than using less exacting types of dialogue. This is even the case when faced with dangerous forms of consequentialism itself; if someone is already willing to use barbarous means to justify a glorious end, if they believe in an ideology that declares that duty and virtue lie in the courage to do what must be done, it may be that the only counterargument left is that the end that will be attained is not so glorious after all.

  5. re: counterexample #3

    Can't we branch out from Aristotle's list of virtues and state that "consideration for the consequences of your actions" is itself a virtue? The important difference, in my mind, between utilitarianism and virtue ethics isn't so much focused on different views about consequences, but rather that utilitarianism judges the act, while virtue ethics judges the agent and their reasons for a particular act.

  6. The post reminded me of getting all irritated with Sam Harris, as I thought consequentialism obviously inadequate, and then reading Bartley's Retreat to Commitment.

    One of the inadequacies of Harris' view, I thought, was his dismissal of Hume's Guillotine. Bartley (page 199 in the 2nd edition), however, pointed out that whilst an ethical statement could not be justified by a factual statement it could be rebutted by one.

    I don't think you have to be a lunatic Popperian to agree that he had a point.

    Reading the post made me think that may not be necessary if ethics were thoroughly liberal: that no act had to be justified but acts could be forbidden.

    There are swathes of behaviour that are not forbidden by one or two of the three theories. Here one is forced only by the one that stipluates a behaviour and only if it does so. Which particular one, of course, changes with the situation which makes it look as if you are moving allegence between the theories. In fact, though, you are following them all at the same time, they only impinge though where they forbid a behaviour.

    Of course, there is a problem where they say things that conflict. There one does have to choose. This would make your Consequentialism less of awarding more weight to Consequentialism in a decision and more that you resort to Consequentialism when there is a conflict. Sort of the difference between having more votes against having a veto.

  7. @mufi: I hadn't realized you & jcm were one & the same. Glad to see you're still sympathetic to the view you named!

    @Sean: I tend to agree that consequentialism still needs to be the "meta" view when we are considering which principles to adopt. As you very rightly put it, moralized disgust is not going to have a place in an ethical theory unless it can carry its weight by doing some actual good.

    @Greg Nirshberg: Certainly you can make such a move to save virtue ethics, but it seems rather strained, doesn't it? I mean, when it comes right down to it, making (explicit or implicit) expected utility calculations is not a trait like courage or generosity, it's an activity.

    And it just seems kind of trivial that you can take any good aspect X of a moral theory that is *not* virtue ethics (say, promise-keeping from deontology) and then say "and it is a virtue to do X." The bottom line, I think, is that VE looks at ethics from a completely different perspective than the other two, and rather than trying to awkwardly cram them together maybe we should just accept that they are different viewpoints useful at different times?

    @Tony Lloyd: "whilst an ethical statement could not be justified by a factual statement it could be rebutted by one." Nice quote!

    While the idea of letting ethical theories merely *prohibit* is an interesting one, I guess I don't see it as too fruitful, because I'm MOSTLY trying to use ethics as a decision theory to figure out, e.g., what to do with my life and how to use my time, rather than whether or not such-and-such is permissible.

    But as I was hinting to Sean, I do tend to see consequentialism as having "veto power." :)

  8. "While the idea of letting ethical theories merely *prohibit* is an interesting one, I guess I don't see it as too fruitful, because I'm MOSTLY trying to use ethics as a decision theory to figure out, e.g., what to do with my life and how to use my time, rather than whether or not such-and-such is permissible."

    But sometimes knowing what *not* to do is not only useful in decision making, but the best we can hope for in a given situation. It seems to me that knowledge of error precedes knowledge of how to correct for error. Sure this is oversimplistic, and runs into problems in the sense of if you use the argument by elimination of what not to do in decisions of how to live your life, you can never be sure you've entertained all the possible prescriptive options.

    "Certainly you can make such a move to save virtue ethics, but it seems rather strained, doesn't it? I mean, when it comes right down to it, making (explicit or implicit) expected utility calculations is not a trait like courage or generosity, it's an activity."

    I'm not sure Ian. Courage and generosity are labels we give to human character traits that are only possible through the observation of action. Are you saying that courage and generosity have real ontological status? Whether it's in courage or generosity or consideration for consequences, there's something more fundamental going on at root, right? Something about the motivations, dispositions, and beliefs of the epistemic agent that leads them to behave in certain ways. We've created labels to describe and differentiate different sorts of character traits that lead to different sorts of behaviors, and the ones we want to promote we've labeled "virues", so while it mean seam a bit ad hoc, I don't really see a problem in adding, deleting, or changing our lists of virtues whenever we find it prudent to do so. Embracing some sort of fallibilism and developmental process in our moral knowledge is a good thing I would say.

    So if you want to say there is a character trait that leads to courageous or generous behavior, I would say there is also a character trait that leads an agent to engage in expected utility calculations. If you're not comfortable with that, then it seems to me we should be discussing what is at root of all those disparate behaviors, and drop talk of virtues all together.

  9. I was sorry to disagree with most of the specific examples, but once the post shifted to its constructive portion you came out with an opinion I'm happy was very similar to a view I have even though I'm either a pure utilitarian or very close to it.

    So just for one of the examples, take the crush video one. I think it's ill-posed since you mix whether other people should object to someone who views and enjoys the crush videos with whether the viewers themselves can view the videos and produce net positive utility. Those two questions are different since they concern different action sets. The least-convenient viewer would be one who not only receives utility from viewing the vid, but also has perfect self-discipline so as not to pose any future risk of buying such videos on the black market or prompting (perhaps indirectly) others to make such videos to anticipate his demand. (Ordinarily, it'd be an endogenous preference so the viewer should just incur the cost of changing his tastes to avoid those harms-in-expectation.) (That said, I'm skeptical any real human can experience enough utility given the torture at stake, but the hypothetical objection basically goes back to Nozick's "utility monster" objection.) But it's b/c such a viewer is so unlikely that utilitarian observers would still condemn that viewer since s/he would be observationally indistinct from other, much more likely, viewers who are not so “inconvenient” to utilitarianism (at least observers in real-life situations). Really what I htink you want is to posit two unrealistic people - the viewer of inhuman self-discipline, and the perfectly-informed observer.

    The post's positive part though is something I agree with, which is the idea that some people, over the run of situations they face, really will do better by not looking at consequences but thinking about “virtues” or whatever. Some people are excruciatingly bad at associating expected consequences with actions, so they’d (ironically) be better consequentialists by thinking about ethics in terms of, say, virtue. Where we part ways is where you seem to believe such an approach should be uniform across people.

  10. This comment has been removed by the author.

  11. Wonderful post-
    I've been thinking along broadly similar lines, but I think there may be a way to present this kind of view in a less 'aesthetically messy'.

    It seems to me that there is a simple unifying concept beneath the big three moral systems, deontology, consequentialism, and virtue ethics (and almost certainly purtiy concerns and the rest). Each of these systems is trying to argue for a system of values, and seems to fail to the extent that it tries to universalize a single value (duty, utility, character.. although the last is more complicated).

    So ethical pluralism could be reframed as a system that acknowledges the multiplicity of important values in our moral lives. One advantage of this reframing besides its simplicity is that it gets close to makes a moral system *useful* in practice. When faced with a hard choice about what to do, or how to be, stopping to think about the things that you really value is very useful, in a way that calculating universal utility can never be. Ethics, as a philosophical endeavor, can then focus on balancing and interrelating our values, and in persuading others to share them

  12. This is good stuff. I’ll comment on the counterexamples.

    #1: The evaluation omits the effects that a person’s character has on future behavior and society’s legitimate interest in evaluating the same. A supposition that we can measure utility so precisely as to even out the differences is not reasonable or supported by any evidence. Therefore, social utility is a broader concern than the statement frames it to be. This illustrates the interplay between values and consequences.

    #2: The analysis seems to assume that honoring the worth and dignity of all persons is not a virtue. I maintain that it is best seen as a virtue. In a situation where many participants are prone to display an absence of virtue, in this case by harming innocent people, the sheriff may have to make a judgment about what is the best, i.e., the most life-affirming response. We can categorize that all day long but that categorization may say more about our inclination to want to categorize things than about anything else.

    #3: This statement of facts ignores the long-term effects of virtue. Sally may rightfully counter that her issue is being ignored and that by focusing on it, she may bring about good in the long run. I do not believe that her point of view can be assumed away by positing that her evaluation is contrary to the facts: we may never know. Many great things have happened, which “everyone” said were impossible but which advanced the human condition by light years. I am not convinced that a hypothetical case could be constructed that would allow us to evaluate the matter through reason and empiricism.

    #4: We might respond with other questions. Of what value is judging the life of one who has died? On what basis do we assume that there is any value whatsoever in making a wholesale judgment about a man who died 175 years ago, or yesterday for that matter? Why isn’t the self-indulgence implicit in such judgments a vice? In other words, why should we accept the assumptions of the hypothetical inquiry?

    #5: How does this make it “justice versus utility?” That assumes a definition of justice that I do not share.

    #6: As in #4, I don’t see the point of judging the parties wholesale. Fred may have built some internal strength overcoming his impulses but Penelope may be less likely to engage in torture or similar behaviors. Each may be more advanced in the dimension of their respective strengths. That’s more important than a wholesale judgment of either of them. If they were competing candidates for particular jobs or offices, I would want to know what jobs/offices those were, so I could evaluate which person was best suited to the position and whether either of them should be disqualified from it.

    In each case, I suggest that the philosophical argument does not look deeply enough into the question. To be useful, philosophy must relate to the world as it is. These counterexamples are challenging but only up to the point where you can see past their implicit assumptions. So the first assumption I question is how valuable these categories are; the second is how they are framed and what assumptions underlie them.

  13. Ian: Glad to see you're still sympathetic to the view you named!

    Yes, but I admit that it's only a default position, given my understanding of how moral reasoning actually works in the real world, which the normative theories each only partly describe (as your examples help to illustrate).

    On a slight tangent, I read and enjoyed that Non-Libertarian FAQ that you recommended here a while back, although I might quibble a bit with the morality section, in which he adopts a utilitarian/consequentialist position, defined in terms of "the set of principles which, when followed, will create on average the best world..." where people are "happy instead of unhappy" and there is "as little pain and suffering as possible..." I'm good with that...until I run into a counter-example (similar to one of yours) in which these criteria conflict with other criteria (e.g. virtue or justice) that (for whatever reason or non-reason) "win out."

  14. None of these counterexamples raise the faintest twinge of intuitive force in me that consequentialism isn't the only thing that matters to me - I DO want to live in a world where scapegoating achieves better results, and I question how reasonable a person is who denies this - as far as I'm concerned, they are simply allowing irrelevant, intruding emotions to cloud their judgments. I simply don't care about making decisions that have no impact on anyone's conscious states, and I have no idea why this strikes anyone else as compelling on reflection.

    If all deontology and virtue ethics amount to are withered stumps of jargon that we extract useful labels from for the heuristics we adopt, then I hardly see it worth the effort to argue for a "pluralist" account of ethics. And yet that's all that it seems to me we'd likely to get out of your proposal. We might simply be better off adopting new terminology entirely, terminology that acknowledges the underlying, implicit consequentialist rationale for our concepts and practices and doesn't trip people by retaining the the potentially confusing philosophical residuum of adapting virtue and deontic terms to consequentialist use.


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