Thursday, April 8, 2010

What is 'Just-So' science?

We write a lot about the tendency toward ideology in the way some science is done, especially in relation to genetic determinism and Darwinian fundamentalism, pointing out how often and how far inferences in these areas go beyond the data, or scenarios are assumed and then treated as being proven. When applied to humans, the offenses are more numerous and more serious, especially since they can have social consequences and are often blatantly related to lobbying for careerist objectives like research funding. Misleading hypotheses masquerading as established truth can (are designed to) move money, and that takes it away from other areas that might be more fruitful.

When you construct a hypothesis about something you see in Nature, it almost always has some relevance to a currently accepted theory. But in the areas mentioned, the theory as well as the hypothesis tend to be accepted rather than subject to serious, objective testing. Thus, the idea that everything has to have a Darwinian explanation--that is, that natural selection is assumed to be the causative force--underlies much of what we object to in our posts. Yet proper evolutionary theory does not require the selective assumption, in the usual fine-tuning sense in which it is typically invoked. Even many biologists, not to mention hosts of pop-culture opiners, harbor what is largely a caricature of what we actually already know.

A Just-So story is one of Rudyard Kipling's children's tales that gives a fanciful, but of course seemingly air-tight explanation for some observed phenomenon. How did the leopard get its spots? An Ethiopian friend daubed the once all-tawny leopard with his blackened finger tips.

In science, a Just-So story is one that can hardly be refuted, is offered based on some sort of evidence, usually not very rigorous, and is treated as truth, almost defying anyone who criticizes to prove that the hypothesis is false.

In fact, in order not to be junk science, a Just-so hypothesis should be subject to stringent testing before being accepted, just like any other hypothesis. Instead, those offering such an hypothesis often act as if this is an insight that only a benighted heretic would challenge. They put a huge burden of proof on the challenger when, in fact, the Just-So hypothesis is often so air-tight with after-the-fact rationale as to be hardly an empirical suggestion at all--and not all that different from other kinds of statements of faith.

This is serious, because much of the life, social, and health sciences are victims of facile hypotheses of this kind. Some of them may be true, but the strength of evidence is usually wanting. In many cases collecting appropriate data is very difficult, if it's even possible. Many evolutionary scenarios are like this. But that is an argument against, not in favor of, asserting the hypothesis.

Questions that often tempt Just-So explanations include those related to our recent post on the widespread observation of 'gay' behavior, questions such as: Why are men like this and women like that? Why is gender behavior like it is? Or, in human genetics: Why do we get diseases like cancer or diabetes? What is the cause of mental illness? Can we explain these phenomena with 'Darwinian' scenarios?  Hypotheses invoking natural selection or genetic determinism are routine, yet we have tons of reason to realize that they are usually heavily over-stated.

The point is not that outcomes or traits have no cause, nor that the cause can't have a genetic component, nor that natural selection may not have played a role in the trait in the past. Instead, the point is that with pat hypotheses, often representing pre-formed worldviews, and/or vested interests, we don't know what level of truth applies since the assertions are being assumed rather than tested.

As scientists, we should stop telling Just-So stories, no matter how much they make us feel perceptive and intelligent, or lead us to be assertive when it comes to things that could be misused against people in society. Because it's awful science, it is often justifiably called 'junk' science, and detracts from good science.

Of course, those stories might be perfectly suitable for entertaining children, for whom contemplating fantasies may be an enjoyable stage of life.