PREDICTIONS VERSUS EXPLANATIONS

Let us return to the original question proposed by Brush. Do predictions of novel facts provide stronger evidence than explanations of old ones, as Whewell and Popper claim? Or is the reverse true? My answer is this: Sometimes a prediction provides better evidence for a hypothesis, some­times an explanation does, and sometimes they are equally good.

To show this, let us begin with a case that violates the historical thesis of evidence. Here it should be easy to show that whether the putative evi­dence is known before or after the hypothesis is formulated is irrelevant for confirmation. Let the hypothesis be

h = This coin is fair, i.e., if tossed in random ways under normal conditions it will land on heads approximately half the time in the long run.

e = This coin is physically symmetrical, and in a series of 1,000 random tosses under normal conditions it landed on heads approximately 500 times.

We might reasonably take e to be empirically complete with respect to h. Accordingly, whether e supports h, and the extent to which it does, does not depend on empirical facts other than e. In particular, it does not de­pend on when, how, or even whether e comes to be known, or on whether e was known first and h then formulated, or on whether h was conceived first and e then stated as a prediction from it. Putative evidence e supports hypothesis h and does so (equally well) whether or not e is known before or after h was initially formulated, indeed whether or not h is ever formu­lated or e is ever known to be true.

So let us focus instead on cases that satisfy the historical thesis of evidence. We might suppose that at least in such cases explanations (or predictions) are always better for confirmation. Return once again to our drug hypothesis:

h = Drug D relieves symptoms S in approximately 95% of the cases.

Consider now two evidence claims, the first a prediction about an unknown future event, the second a report about something already known:

e₁ = In the next clinical trial of 1,000 patients who suffer from symptoms S and who take D, approximately 950 will get some relief.

e2 = In a trial that has already taken place involving 1,000 patients with S who took D (we know that) approximately 950 got some relief.

On the prediction view, e₁ is stronger evidence for h than is e₂. On the explanation view it is the reverse. And to sharpen the cases let us suppose that e₂, by contrast to e₁, was not only known to be true prior to the for­mulation of h, but that h was formulated with the intention of explaining e2. Which view is correct? Neither one.

Let us take the prediction case e₁ first. Whether, and to what extent, e₁ (if true) supports h depends on empirical facts in addition to e₁. In this case it depends on the selection procedure to be used in the next clinical trial. Suppose this selection procedure calls for choosing just 5-year-old girls with very mild symptoms who in addition to D are also taking drug D' which ameliorates symptoms S in 95% of the cases and potentially blocks D from doing so. Then e₁ would be very weak evidence for h, if it supports it at all. This is despite the fact that e₁ is a correct prediction from h, one not used in generating h in the first place. By contrast, suppose that the se­lection procedure used in the past trial mentioned in e₂ is much better with respect to h. For example, it calls for choosing humans of both sexes, of different ages, with symptoms of varying degrees, who are not also taking drug D'. Then e₂ would be quite strong evidence for h, much stronger than what is supplied by e₁. In such a case, a known fact explained by h would provide more support for h than a newly predicted fact would.

Obviously the situations here can be reversed. We might suppose that the selection procedure used to generate the prediction of e₁ is the one cited in the previous paragraph as being used to generate e₂ (and vice versa). In this situation a newly predicted fact would provide more sup­port for h than an already explained one.

In these cases what makes putative evidence have the strength it does has nothing to do with whether it is being explained or predicted. It has to do with the selection procedure used to generate that evidence.^[88] In one situation—whether it involves something that is explained or pre­dicted—we have a putative evidence statement generated by a selection procedure that is a good one relative to h; in the other case we have a flawed selection procedure. This is what matters for confirmation—not whether the putative evidence is being explained or predicted.

4.