On ethics, part II: Consequentialism

by Massimo Pigliucci

http://www.bbc.co.uk/ethics

[This post is part of an ongoing series on ethics in which Massimo is exploring and trying to clarify his own ideas about what is right and wrong, and why he thinks so. Part I was on meta-ethics]

Following up on my general discussion about what ethics is, I am now going to proceed with a series of brief commentaries on specific approaches to ethics, attempting to cover the major ways of philosophizing about morality that have been developed over time. Before we go on, I need to remind everyone that this is most certainly not a comprehensive, not to mention scholarly, survey (there is simply a huge amount of material on this out there), though I will provide links to resources for those who wish to pursue any of these topics in more depth. Rather, this is my way to try to sharpen my own thoughts about ethics, and to expose them to a discussion amongst friends, hoping for truth (or at least a better understanding) to spring from it.

So let's get started with consequentialism, a way of thinking about ethics that really subsumes a family of ethical theories, most importantly the various forms of utilitarianism (from Bentham to Mill to Peter Singer). I'm starting here because consequentialism is arguably the most popular form of ethical reasoning among philosophers, and perhaps even among the public at large (whether or not most people would actually recognize it). It is also both very appealing at first sight and turns into a highly problematic Pandora box once we look a bit closer.

Essentially, consequentialism is the idea that all that matters in morality are the overall consequences of our actions. As opposed to what, one might ask? Well, for people adopting other ways of looking at ethics consequences are not the overriding criterion, which may instead be duty (to law or to gods), personal integrity, respect of other people's rights, and so on.

The first obvious question is: what quantity should our concern for consequences try to maximize? The answer depends on what sub-type of consequentialist one is, but for modern utilitarians this amounts to maximizing both happiness (philosophically conceived as human well being, not hedonistically) and minimizing pain. One could, however, choose different criteria to be maximized and still be a consequentialist.

This in turns brings up the first obvious problem for basic consequentialism: it doesn't seem to provide a principled way to distinguish between different ways of maximizing, say, happiness. This is really a matter of math: if you want a particular number as the result of a given operation, there are many ways to obtain that number. Maximization could be achieved, for instance, by attempting to increase everyone's happiness, but also by increasing the happiness of a majority at the expense of a certain increase of pain in a minority. Heck, depending on how one measures happiness, one might even find that a higher total can be obtained by greatly favoring a small number of people while making things worse for the rest.

A similar point is that consequentialism is supposed to be agent-neutral, meaning that it doesn't matter whether a certain action has the consequence of increasing your own happiness, that of a friend, of a relative, or even that of your nemesis. This is something that strikes many people (and a good number of philosophers) as rather odd. For instance, we consider it moral for a parent to care first and foremost for his children, but there is no way to justify this within a consequentialist frame. One possible reaction could be: so much the worse for our sense of parental duty to children, but one might equally develop suspicions that there is something wrong with an ethical system that does not take the specialness of parent-child relationship into account. (Of course, the concept of “duty” itself lies outside of consequentialism's purview.)

Here is another obvious problem: in most cases we simply do not know what the consequences of a given course of action will be, certainly not the overall consequences. And what does “overall” mean here? Over what extent of space and time are we supposed to preoccupy ourselves? Surely, in the very long run, nothing will matter anyway, since the sun, and eventually the galaxy and the universe itself will not exist. But even over the much smaller scale of human existence, are we supposed to measure the lifetime consequences of our actions, for all people potentially involved?

Clearly this makes no sense, and no serious consequentialist pursues that line of reasoning. But here is the funny thing about the approach: every time consequentialists run into one of these daunting issues (there are many more, some of which I will mention below), they modify consequentialism itself. For instance, one can get around the problem posed by the impossibility of knowing all the consequences of an action by moving to what is called “reasonable consequentialism,” the idea that an action is morally right if it has the best reasonably expected consequences.

This move may seem strange, as it amounts to rescuing one’s theory from trouble by modifying the theory ad hoc so that it can absorb, rather than resolve, the trouble. As Popper argued in the case of scientific theories, however, this is the path that leads straight to pseudoscience (or pseudophilosophy). But in fact things aren't quite that bad: even scientific theories are often adjusted to account for new data, a practice that is both acceptable and in fact indispensable for scientific progress. Think of how the Copernican theory got “tweaked” by Kepler, who replaced circular planetary orbits with elliptical ones to better account for the actual movements of planets. No astronomer thinks that that move amounts to a step in the direction of pseudoscience.

And yet, as post-Popperian philosophers of science (e.g., Imre Lakatos, one of Popper's students) have argued, there must be a limit to these adjustments, beyond which we declare the research program associated with a given theory to be “degenerate” and we abandon the theory in favor of a better overall alternative. I think a major issue with consequentialism is that it is difficult to tell when so much tweaking has been done that the theory has become unrecognizable.

For instance, I mentioned above the contrast between consequentialism’s agent neutrality and the common moral thinking that we have special duties toward friends and relatives. One way that has been proposed to deal with this is to introduce so-called “friendly consequentialism,” the doctrine according to which an action is moral if it has the best consequences for one’s own friends, or family. Okay, but then we seem to lose a major appeal of consequentialism itself, precisely the fact that it doesn’t favor one person over another in terms of moral worth. Can friendly consequentialism still be meaningfully thought of as consequentialism at all?

Or consider another common issue raised in this context: if most of your neighbors engage in a behavior that pollutes a local resource (say, dumping garbage in a nearby river), then if you did the right thing instead and brought your garbage to the processing plant, consequentialism might lead you to conclude that doing the right thing is not moral in that case, since you are not making a dent in the situation (the river will still be polluted) and you are causing a problem for yourself (you have to spend time, energy and money to bring the garbage to the plant). The overall consequences of your actions are negative. All right, then, say some consequentialists: we are going to introduce “rule consequentialism,” the idea that something is morally right if it doesn't violate the kind of behavioral rules that would have the best consequences for the community. Neat, but, once again, is it still consequentialism?

Consequentialism also seems to have a hard time with the concept of rights, not just the positive ones considered necessary by progressive thinkers, but even the negative ones minimally accepted by conservatives. Suppose the best overall consequences could be achieved if we prohibited certain behaviors, like driving highly polluting cars; or, to take a more extreme counterfactual commonly invoked against utilitarianism, imagine that a doctor could save five people’s lives by harvesting the organs of a sixth, healthy person. Although it can be done, it is hard to see how a consequentialist can respond to these situations without further difficult mental gymnastics to save his conception of morality. Individual rights just don’t sit well with a doctrine established on maximizing overall happiness.

One more problem to think about: one of the premises of consequentialism is that it makes sense to think of an overall benefit resulting from our courses of action. But one worry is that “overall benefit” is as problematic a concept as “overall skill”: one can have this or that skill (painting, music, writing, plumbing), but it doesn’t make any sense to ask what is a person’s overall skill. Does it make sense to ask what is the overall benefit of a course of action?

As I mentioned in the beginning, this does not pretend to be either a comprehensive or an in-depth analysis of consequentialism. That’s not the sort of thing that can be accomplished in a blog post (or even a series of blog posts). Nonetheless, the reflections I forced myself into while writing this confirm my overall impression of utilitarianism: it is an initially appealing idea that is likely to die the death of a thousand small ethical cuts. This certainly doesn’t mean that consequences shouldn’t enter into moral discourse, but it does mean that I do not see an exclusive focus on consequences as viable to construct an entire ethical theory.

Next: deontological ethics.