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, January 23, 2012

Considering the consequences

by Michael De Dora

I have never thought much of consequentialism, the moral theory which asserts that determining “the good” or “the moral” is a matter of measuring outcomes. Decisions about what is moral, consequentialists say, should depend on the potential or realized costs and benefits of a moral belief or action. There are myriad problems with this line of thought, and while I have already discussed several on this blog, I would like to use this post to examine in more depth what I think are the four strongest objections to consequentialism.

First, consequentialism says nothing about the substance of one’s ethic. While most consequentialists are utilitarians — a position I also consider vague and tenuous — one obviously needs only value consequences to qualify as a consequentialist. Yet, since everyone has different moral goals, everyone will have different views about potential outcomes. For reasons discussed below, consequentialism does not help us decide which are better or worse. Rather, one’s moral values come prior to consequential calculation, and help determine what one thinks about the consequences.

Second, consequences are often not at all predictable or in line with the actions that caused them. For example, does the fact that certain Muslims riot over the printing of anti-religious cartoons suggest that printing said cartoons is immoral or wrong in some other way? Not in the slightest. It only suggests people have some twisted ideas regarding free expression. Or, consider an exchange I witnessed at a recent Intelligence Squared debate. At the event, two sides of two speakers each debated the motion “The U.N. should admit Palestine as a full member state.” The side taking position against the motion argued that the audience ought to stand with them because of the potential military situation — probably started by Israel — that could be brought on as a result of the U.N.’s recognition of Palestine. Unfortunately, there was no discussion about whether such military action itself would be reasonable.

This gets at a third problem with consequentialism: it often ignores foundational questions of right and wrong for questions of expediency. Or, it ignores concerns about intent for pragmatic concerns. The question of whether a war might start due to the U.N. admitting Palestine as a full member state is an important and interesting one, but it does not answer the distinct question of whether it is right to admit Palestine to the U.N. as a full member state. Those are two different questions that must be considered separately.

Lastly, consequences must be weighed alongside other factors and possibilities. Let us examine a recent exchange on this blog. It occurred in the comments to the recent post, “Massimo’s Picks, special Hitchens edition.”

In the comment thread, Massimo wrote about his skepticism toward the effectiveness of New Atheists like Christopher Hitchens to better the public acceptance of atheism. I replied that:

“Hitchens might not have been the person best fit to sway the majority to our side, but he was part of a movement (the so-called "new atheists") that I think did do two things to help us get to the point where that’s even feasible. First, their out-front writings and speaking engagements put atheism on the forefront of the Western world's consciousness, and created the space for more widespread conversations on religion (like this one!) that were not happening here beforehand. Second, their public work encouraged many apathetic secularists and fence sitters to be more assertive and engage with the problem of religious dogmatism. I think both of these were productive first steps toward getting a majority to embrace secular thinking. And I think these two points can be accepted whether or not you agree with their arguments, or how they stated their arguments.”

Massimo replied:

“… for an allegedly evidence-driven community I hear a lot of claims about all the good that the New Atheists have done, with precious little backing up in terms of data. Are we seriously arguing that atheism wasn’t widely discussed before the Hitchens-Dawkins-Harris-Dennett books? And on what evidential grounds are you asserting that more fence sitters have been drawn inside the movement rather than repelled by the NA’s rhetoric?”

I replied by asking Massimo: “certainly atheism was being discussed long before the arrival of the New Atheists, but on such a widespread and popular scale? The NA all had best-selling books, major TV and magazine appearances, and auditoriums packed with sometimes thousands of people.” His reply: “Nobody doubts that the NA have had an impact. The question is whether it was an overall positive one.”

Massimo’s legitimate empirical question aside (any takers?), I think his last comment is most relevant to our discussion on consequentialism. Whether or not the New Atheists were effective in broadening public acceptance of secular thinking, Massimo raises the following questions: Were the New Atheists necessary to raise such recognition? Couldn’t atheism have been put on the map in some other form or fashion? Indeed, hadn’t atheists previously in human history tried other effective methods? If not, why?

The point here is that there are certainly other possibilities for fostering the kind of space atheists wanted, or an even better space. None of those possibilities were enacted, so we should be thankful for where we are right now. But that does not make what happened desirable.

Regardless, consequentialists could reply that ignoring what may be terrible consequences is unethical. Would they have a point? Consider this common thought experiment: you are a German hiding a Jewish family during World War II, and Nazi guards are at your door asking if you have seen any Jewish people lately. Do you lie to potentially save their lives? Or do you tell the truth and essentially kill the Jewish family? The point here is not that there is an easy answer between lying and not lying. The point is that the consequences — a dead family — are so compelling that they warrant consideration. And this is just one of numerous examples.

What are the implications of all this? That consequences are important, you might conclude? Not necessarily. Instead, I think we have realized only that we have a range of different values, some of which are or can be in tension among themselves. For example, in the case we just considered, we might be stuck between, on one hand, the value of honesty, and on the other, the value of human lives.

As such, perhaps consequentialism should not be looked at as an ethical system in itself — again, it is bare of ethical content — but as a way to figure out if our different ethical systems — based on duties, obligations, virtues, rights, etc. — are working properly or as intended. In other words, consequentialism might help us to see if we are securing the kind of consequences we want. And if we aren’t, it’s time to adjust our aim and try for better consequences.


  1. To me, consequentialism seems more of a meta-ethical assertion than an ethical one. I think of consequentialism as simply stating that values are either a) part of the basic set of values that one uses to evaluate consequences (e.g. arguments to a utility function) or b) can be explained (reduced) in terms of such values and consequences of actions. I actually have yet to see a very convincing counterexample; most criticisms of consequentialism either focus on particular variants (e.g. act utilitarianism) or assert a value from a different ethical framework, but without really showing that it is a counterexample.

    In order for the value of honesty to be truly alien to consequentialism, we need to

  2. Oops. Browser glitch

    I was saying, in order o show that the value of honesty is alien to consequentialism, you'd have to show that there was something about honesty that could never be encapsulated using consequences and values that were metrics on consequences, which is actually somewhat of a tall order. Since we're talking about New Atheists, I can point out that Sam Harris's recent essay "Lying" gives a brief defense of honesty from a consequentialist point of view.

    Of course consequentialism is still a family of ethical theories and not a theory itself. But that's where you have to get more specific about what you're trying to refute. I wouldn't have high hopes for refuting it on the meta-ethical level it stands in.

  3. Consequentialism is a philosophical doctrine, but the empirical question of what people regard as moral does include a consideration of consequences (among other considerations). This does not make the people consequentialist, nor it proves or disproves consequentialism as a philosophical doctrine. The actual morality of people does not necessarily conform (even for a single individual) to a simple organizing principle such as consequentialism. It may even include contradictory choices at different times.
    This is to underline an idea I have expressed before in this blog: trying to figure out a logically coherent philosophical doctrine about something is not equivalent to accounting for the actual way in which that "something" exists or behaves (in this case, the actual way in which people make moral judgments). People's actions are not the faithful application of some principles and their logical implications. Moreover, faced with reasoned evidence of logical contradiction, people would often insist in their original judgment.
    A philosopher could, of course, wave away such "illogical" behavior in the same way that Michael dismisses Muslim protests about Danish cartoons as a "twisted" idea of freedom of expression. But such dismissal is liable to the opposite rejection: a Muslim morality scholar may dismiss preoccupations about freedom of speech as a twisted conception of the real significance of blasphemy, and since moral philosophy depends on a set of axioms (as we have learned in this blog before), we are none the wiser by looking alternately to the Western and the Muslim morality scholars. A sociologist or anthropologist may learn valuable lessons about cultural differences from such an exchange of views about the Danish cartoons, but no honest philosopher can dismiss either view as "twisted" or not valid without assuming some set of (essentially arbitrary) axioms. At that point we are back at square one: why and on which principles (or meta principles) should one accept or dismiss a moral judgment?
    We humans, imperfect creatures of a lesser god, do not invariably conform to some simple and rational set of principles in all our actions. Making sense of our behavior is not a task for logical deduction, but for analysis of the messy evidence found in reality. You cannot do it from your armchair.

  4. "For reasons discussed below, consequentialism does not help us decide which [outcomes] are better or worse." No kidding, here's why.

    Consequentialism is a theory of right action; it states that an action is right iff it produces the best consequences.

    It is not a theory of value, or of what makes for a good action. That is why it is non-committal about moral goodness. This is not a bug of the theory, but a feature.

    In order to have a complete normative theory that is consequentialist, one must supplement it with a theory of the good, e.g., hedonism, or preference satisfaction, or ideal list (G.E. Moore).

  5. "Consequentialism is a theory of right action; it states that an action is right iff it produces the best consequences."

    But the word "right" here is just that, a word, a synonym of "producing the best consequences". This formulation leaves undefined the concept of BEST (or BETTER) consequences, especially in the interesting cases in which the consequences are multiple and perhaps contradictory. Thus, taking a certain decision, a health planner may cause some deaths to be avoided and other deaths to not avoided; or some deaths avoided but more people bedridden with chronic illness; or some deaths avoided but less children receiving education or nutritional support; or some more deaths avoided, but costing a zillion dollars also needed for other goals. This leaves outside the theory any consideration of the reason why one of these actions, to which these sets of (joint) consequences correspond, is "better" or "more right" than the other available courses of action. Except if one introduces a set of (essentially arbitrary) values (one death is equivalent to X malnourished or uneducated children, avoiding one death should not cost more than X dollars, and so on). Philosophers, at this point, may of course introduce their own preferred values, based on whatever axioms of their choosing, but that is not to say that those values are "the right ones", not even "the ones preferred by most people in the world" or even "by most people in the cultures I value most". These latter criteria may be ascertained empirically, but even so they have no particular privilege to be enthroned as "the right values", nor justifies their use to tell which are "the best consequences".
    All in all, this sort of armchair exercise in morality seems pretty vacuous to me, but that is surely a sign of my crass ignorance.

    1. "This formulation leaves undefined the concept of BEST (or BETTER) consequences"

      That's exactly my point. Consequentialism, as a theory, does not specify any particular theory of value. To criticize it for failing to specify what makes consequences best or better is like criticizing evolutionary theory for failing to specify the origin of life. That is not an aim of the theory.

      A normative ethical theory consists of two other theories: a theory of the good, and a theory of the right. A theory of the good specifies what things have moral value, and a theory of the right specifies what makes an action morally permissible or obligatory. The two are not identical. For instance, Kantians place the theory of the right in a logically prior position to the theory of the good. Consequentialists, on the other hand, tend to place the theory of value in a logically prior position. As you say, in order to be a good consequentialist, you must know what makes a consequence best or better. For a Kantian, you need only know what makes an action right (C.I.).

    2. And what do ordinary people do? How do they determine what is good and/or what is right? Are they consistent in these judgments? Do they agree across cultures? Do they agree within cultures and subcultures? To all the precedent questions: why?
      What a sophisticated philosopher may lucubrate about these matters is relatively unimportant for these vast sociological processes determining how people behave and how they evaluate behaviors. And furthermore, the philosopher's armchair musings are of little help unless such thinker starts by looking at reality through scientific glasses.
      People's morality is not the implementation of a prior theory of morality. On the contrary, theories of morality are attempts to rationalize people's actual morality, but usually without a scientific analysis of actual morality: philosophers only compare their conclusions with their own subjective "intuition", mostly in an informal manner. My grandly capitalized Conjecture on this matter (see above) is that no such attempt would be ultimately fruitful.

    3. In my previous comment I alluded to "My grandly capitalized Conjecture on this matter (see above)" but in fact it should be "see below": that Conjecture is in a comment below this one in the thread.

  6. morals = right/wrong = values = evolved emotional response
    ethics = utilitarianism

    is this right?

    any discussion of morals leads to dilemmas and inconsistencies. if a set of beliefs leads to a contradiction there is something wrong. i think the concept of morality is simply not needed - it is a language construct of emotion that is logically inconsistent.

    1. Dear Res Cogitans,
      the notion of morality is needed to deal with the empirical fact that people judge some things as morally right and wrong (the application of these concepts varying from person to person, from culture to culture, from time to time). It is also needed by people in their ordinary life, to discuss what they think is morally right or wrong.
      What you probably mean is that the philosophical discussion of morality is fatally flawed, i.e. you refer to morality as an object of abstract armchair philosophical thinking. I do not know: perhaps it is possible for somebody to construct a logical framework to discuss issues that people regard as "moral problems". What I tend to believe is that nothing substantial can come from such purely logical endeavors.
      More to the point, I dare to launch a Conjecture, namely:
      That for every set of apparently rational and reasonable moral axioms, there would be some logical implications of those axioms that would not feel "right" for the majority of people, even if those same people initially agreed with the "reasonableness" of the values embedded in the axioms.
      For the economically minded: This conjecture is an analogue of Kenneth Arrow's "impossibility theorem" (except that Arrow did actually prove his theorem, unlike my Conjecture which for the moment remains purely conjectural). For the non initiate: Arrow's theorem states that given a plurality of people with individual preference orderings over a certain set of outcomes, there is no aggregate or "social" preference ordering satisfying a small number of perfectly reasonable requirements (such as: that if A is preferred to B by every individual, then A should be also socially preferred to B). A disturbing proposition, Arrow's is. And so, I mischievously hope, is my Conjecture.

    2. the notion of morality is needed... by people... to discuss what they think is moral...
      well yes indeed. wittgenstein would love that one.

      what we cannot speak of, we must pass over in silence.

      your conjecture boils down to "people's emotions are not rational".
      game theory and evolution together explain all our emotional responses to decisions - including what is called morality.

    3. No, Res, no "not rational". Only not deducible from a set of philosophical axioms. By the same token, mathematics are rational in spite of Gödel incompleteness theorem.

      Evolution (understandable with the help of, among other instruments, game theory) tends to ensure that the aggregate behavior of people (and other living things) is "rational" in the particular sense of evolution, just as physics ensure that water runs down a hill following the Hamiltonian line of least effort. But neither the physics of running water nor the theory of evolution would predict the behavior of every water molecule or every individual plant or animal at every particular occasion. Even if evolution generates a viable and "rational" morality among humans, mistakes are (often) made, by all kinds of people, even by entire groups of people (e.g. tribes or groups that lead themselves to extinction through perpetual warfare or an excess of human sacrifice). Human morality is not irrational but imperfect and messy, just as everything else in Nature.

  7. "For example, does the fact that certain Muslims riot over the printing of anti-religious cartoons suggest that printing said cartoons is immoral or wrong in some other way? Not in the slightest. "

    Um, circular reasoning here. If I am a consequentialist, then printing such cartoons might well be immoral...by definition. You can't show that I'm wrong merely by asserting it or by appealing to some other moral theory. The best you can hope for is to show that I am inconsistent.

    I have little respect for moral theories, though, since everyone tosses them out once they contradict intuition.

    1. I think that printing cartoons while aware that doing so will lead to riots where people will get injured is, prima facie, an immoral act. However, there are second string reasons to think that in the larger picture (censorship and freedom, for example) it should be a permissible act.

      I see no reason why a consequentalist theory can't include second string consequences. In fact the over-demanding objection to utilitariamism seems to assume that it must include these.

  8. First Objection: Consequences are adjudged against personal values - what is good for me and what is not. That is the only viable substance to one's ethics. Even the law of reciprocity is subject to this individual judgment. There are no moral absolutes. The only real morals/ethics are indeed extremely personal.

    Second Objection: You've mistaken poor judgment for unpredictable consequences, social responsibility for ethics. You yourself have pointed out how these examples are poor execution rather than poor premise, yet even you did not delve into the personal ethics which drives such poor decision making. Complicate it with one bad choice on top of another and it only gets worse. You've let a chain of events cloud your vision or the base or root of the chain.

    Third Objection: There are no fundamental questions of right and wrong, only people who cannot understand ethics or refuse to understand that human ethics are derived from a time when we still lived in trees. Your ilk complicate the discussion then claim there is no easy answer and so yours must be correct. It's the kind of argument that wants to examine the tires while sitting inside the vehicle.

    The statement: "...consequences must be weighed alongside other factors and possibilities." is patently not true. It may seem common sense to yourself, but that does not make it true. When you boil human decisions down to what is good for me or bad for me and the law of reciprocity all the complex matters simplify nicely. In your example, one might choose to risk possible death to avoid doing to another what you would not want done to yourself. It's a personal choice of what you alone believe to be true for yourself. There it is, what is good for me, what is bad for me, and the law of reciprocity. That is all that is needed to explain morality or ethics. Consequentialism is nothing more than looking at these during the decision making process.

    Your conclusion is in error. You have again assumed that ethics is more than weighing a thing as good for me or bad for me and how it fits into my version of the law of reciprocity. There are no absolutes, no objective moral values, no guiding definition of morality, no objective moral content. Any assumption of such cautiously skips over the part where we humans arrived at a consensus about general social behaviors. It was a brutal, bloody, violent affair that did not end quickly - it is in fact still going on.

    Hitchens and others do bring positive value to the world and to other atheists. They have turned the discussion from a lopsided discourse where believers were assumed to be the good guys into a discussion about who is right and why. As with our race's discourse on social behaviors, this one too will have victims and collateral damage. For us apes there seems, as yet, no way to avoid it. That said, the sooner we start, the sooner we finish.

  9. Suppose you hold a set of values, and wish to evaluate your possible course of action. Should you take consequences into account? Naturally. Should you consider both intended and unintended consequences? Of course. Only direct consequences, or also indirect consequences? Even indirectly indirect consequences to materialize from the possible actions of others as a response to yours, or as a response to responses to yours? Where do you stop?
    And then, besides consequences, should you also consider other possible criteria, such as respect for some other value? (for instance, consider J. Swift's modest proposal for the [Irish] poor, to eat their own children. The consequences were good in two ways: it would reduce overpopulation while increasing the food supply (even increasing its dietary quality by including meat in an otherwise dull diet of oats and potatoes); should people also incorporate the deeply ingrained respect for the sacredness of human life, or rather consign it to the realm of superstition and go ahead with Swift's proposal for the, well, swift elimination of hunger? Is there any objective basis for this special respect for human life (compared to the life of pigs and lambs)? Even unwanted human life, as was the case with many children of the poor?
    The answer, generally, is yes. One should also consider other values, besides the practical consequences of one's actions. At least those values that are incorporated into one's moral framework, arbitrary as they might be (for instance, Indians abstain from eating not only their own children but also cows: is cow life sacredness equivalent to baby life sacredness?
    Why not?
    We can continue in this vein forever, I suspect, going precisely nowhere, running in circles around our own armchair, and advancing not a iota in our understanding of human (actual) morality.

    And, of course, if one chooses to look first to actual morality through the glasses of science, and only then start theorizing about it, not in order to regulate but to understand it, one would become a (social) scientist, and thereby cease to be regarded as a philosopher in the ordinary sense.

    On the other hand, if one insists on regulating human behavior in a normative way, just from one's armchair, one may be seen as the victim of an extreme condition of conceit and presumption, since the endeavor of armchair moralizing could be undertaken with equal right by many people, with presumably different results, and none would be the wiser.

  10. Hector, I'm not sure I follow you. In the J. Swift example, why do we distinguish the death of children as an "other value" -- different from a practical

    Isn't the fact that innocent children will die a consequence that is contrary to values held a priori? That is, we never said we must eliminate hunger by any means necessary. Rather, there is an ordering of moral preferences: we would like to end hunger, but we
    would like to preserve life even more. These are both moral values that are specified a priori.

    In other words, the example doesn't show a case where we should consider something *other* than consequences, only something other than *one* consequence, given an "uber alles" value.

    P.S. Thanks for your great comments -- particularly the one about the Danish cartoon.

  11. It helps to recall that the whole point of the label is to set varieties of classical utilitarianism against varieties of deontology. The latter focuses on agent intentions, the latter on consequences. But it would be similarly absurd to criticize classical deontology on the grounds that it tells us to value intentions, but doesn't tell us what sort of intentions are valuable. The consequentialist label only specifies the (contended) morally relavant features of some action. Typically, the claim is that an action counts as "good" insofar as it tends to maximize certain outcomes. So far, it's merely a topographical term; the theorist has then to *specify* those features among possible outcomes which count as valuable.

    Your confusion is nicely captured when you write that "one obviously needs only value consequences to qualify as a consequentialist." This would be patently absurd, for the reason you suggest. The whole point of the label is to focus attention on one feature of action, and then to go on and account for which features, among different possibilities, are to be maximized.

  12. Your second criticism is plausible, but very poorly illustrated. The line between the initial publication of cartoons and the street protests was extremely indirect. The episode does point toward a more general problem with many types of consequentialism, namely, that they tend to over-generalize lines of agent responsibility, but your casing of the episode simply ignores all the relevant issues of consequentialist calculations. Similarly with the I2 debate. The fact that people in a debate expressed an awful argument about possible consequences is simply beside the point, simply a straw man. It's rather akin to criticizing the modern evolutionary synthesis on the grounds that Hitler used eugenics. I mean, it's done, but it's surely not model behavior.