In our last discussion of morality and science, an interesting argument was raised in the comments (by rbd and then in more detail by Ben Finney), concerning an analogy between morality and health. Sam Harris has also brought it up. It's worth responding to because it (1) sounds convincing at first glance, and (2) has exactly the same flaw that the morality-as-science argument has. That's what a good analogy should do! If I can paraphrase, the argument is something like this: "You say that morality isn't part of science because you don't know what a `unit of well-being' is -- it's not something that could in principle be measured by doing an experiment. But one could just as easily say that you don't know what a `unit of health' is, and therefore medicine isn't part of science. The lack of some simple measurable quantity is a simplistic attack against a sophisticated problem." This gets right to the point. Because, in fact, I don't know what a "unit of health" is, which is why medicine is not -- solely -- part of science. Let me explain what I mean. Obviously we use science all the time when it comes to medicine. Similarly, we should be very ready to use science when it comes to morality -- it's an indispensable part of the endeavor. But in both cases there is a crucial component that lies outside the realm of science. Here's how we do medicine, in a cartoonishly simplified version that is nevertheless good enough for our present purposes. First, we decide what we mean by "healthy." Then, we use science to try to bring it about. That first step is not science, no matter how much science might be involved in the definition. Various measurable quantities certainly belong to the realm of science -- height, weight, pulse, blood pressure, lifespan, time in the 40-yard dash, etc. But what we decide to label "healthy" is irreducibly a human judgment, not an empirical measurable. Some people might think that extreme thinness is part of being healthy, while others might prefer a more robust physique. Some people might define health as the state that maximizes life expectancy, while others might put more emphasis on quality of life even at the expense of total years. It matters not a whit what people actually think, of course -- even if everyone in the world agreed on what "healthy" meant, it would still be a judgment rather than an empirical measurement. If one contrarian person came up with a different definition, they wouldn't be "right" or "wrong" in the conventional scientific sense. There is no experiment we could do to answer the question one way or another. In the real world, we more or less agree on what constitutes health, so the non-empirical status of this choice isn't treated as a crucially important philosophical problem. (At least, until you start reading the literature on disability studies, and you realize that what you thought was obvious maybe is not.) We agree on what health is, and we set out to achieve it, and that second part is very much science. Morality is exactly the same way, although with somewhat less unanimity in the first step. We agree (or not) on what morality is, and once we do the process of achieving it is very much a scientific issue, in the broad-but-perfectly-valid definition of "science" as "an understanding of how the world works based on empirical data." Once again, it doesn't matter whether we agree or not, because that first step is a decision we human beings make, not something we measure out there in the world. While both health and morality are human choices rather than empirically measurable quantities, they certainly aren't random choices. Human beings aren't blank slates; we have preferences. Most of us would prefer to live longer and be free of aches and pains; these preferences feed into how we choose to define "health." Likewise for morality. But "we broadly agree on X" is not, and never will be, the same statement as "X is a scientific truth." Understanding our preferences, turning vague impulses into precise statements, constructing logical frameworks based on them -- that's what the philosophy of medicine/morality is all about. The case of morality is actually much more difficult than the case of health, because most interesting moral questions involve tradeoffs between the interests of different people, not only the state of one individual. So even if we could do experiments to establish a unique map between mental states and human well-being, we wouldn't really be any closer to reducing morality to science. All very fun to think about, though.