Robert L. Frazier ( www.rlfrazier.org )

Rationality and Science: Can Science Explain Everything?
By Roger Trigg
Blackwell, 1993. vii + 248 pp. £40.00 cloth, £12.99 paper


Science and the empirical method are great success stories. Not only have they given us a deeper understanding of our world, but they have also allowed us an extraordinary control over it. Indeed they are so successful that some believe that science can explain everything, or, at least, that if there is a rational explanation for something, then it will be a scientific one. According to this view, science and its methods are rationally privileged.

There are opponents to giving science this status. It has been argued that the norms underlying science are mere conventions, so science and the empirical method give us only one of many possible ways of understanding the world, and there is no independent standard for choosing amongst them.

Trigg, rightly I think, rejects both of these positions: that science is uniquely rational and that it is merely one method amongst many for understanding the world. He acknowledges the success of science and the rationality of its methods, but argues that there are some things apt for rational explanation that science cannot explain. For example, it cannot explain why the universe is comprehensible and why of all the possible worlds this one became actualized (ch. 6). Indeed, it cannot even explain why there is something rather than nothing (ch. 8). Nor can it explain the nature of consciousness (ch. 9). Also, the irreducibly chancy nature of the universe has made scientific explanation more difficult, as have problems of measurement associated with chaotic phenomena and the nature of fractals (ch. 8). (In passing Trigg mentions the problems that science has with morality (Introduction) and religion (ch. 10).)

Interesting as are these considerations of what science cannot explain, they are, it seems to me, tangential to Trigg's main project. The linchpin of his book, that which connects the claims that science is a rational activity and that it cannot explain everything, is Trigg's argument that science cannot explain why it is rational to engage in scientific activity. Yet if there is no explanation for this, then then engaging in it is pointless.

Roughly, the argument that science cannot explain its own rationality is that rationality involves a normative constraint on an activity, and such norms must relate the activity to a goal external to the activity. Trigg holds that truth or knowledge must be the goal of science. If it does not have this as the goal, then it is pointless; if it is not apt to achieve it, then it is not a rational activity. Concerning the goal of science he says that ``[a]ny account of human activity is liable to lapse into incoherence without such notions as reason, truth and reality. Certainly without them, all human belief, and not just religious belief, will lose its point'' (p. 34) and ``[t]he idea that our beliefs -- whether religious, scientific or whatever -- have to be measured against something beyond them is essential if they are to have any purpose'' (p. 75). About the rationality of science he says, ``...if rationality is linked to the ideas of truth and reality, science is only rational because it provides a reliable method for gaining knowledge'' (p. 62). His conclusion is that if we are to understand science as a rational activity, we must be committed to realism and the possibility of knowledge of the world. These, in turn, require metaphysics (realism and a robust theory of truth) and epistemology (normative, rather than naturalised), which are beyond the scope of scientific theory.

When Trigg considers the views of those who think that science can explain everything, (e.g., naturalists such as Quine, Patricia Churchland and Rorty), he argues that they have to give up the idea that science is rational. They cannot account for the normative nature of rationality. Anti-realists, internal realists and pragmatists (e.g., the later Wittgenstein, Lyotard, Dummett, Putnam, Dewey and Rescher), Trigg argues, have such a weak notion of truth (or the goal of science), that it either makes science pointless, or does not require us to distinguish between science and other possible methods of investigating the world.

There are a number of fairly minor problems with the arguments. One is that they are too sweeping. Trigg takes on so many different views that he papers over subtleties and differences that may be important. Another is that he sometimes simply points out consequences of his opponents' views without arguing that they are unacceptable. For example, when considering Rorty's views he says this: ``Reason must be able to transcend its immediate context. Otherwise it is true that the forging of consensus will be all we can achieve. Politics will replace metaphysics.'' (p. 79) To complete his argument, Trigg needs to show what is wrong with the conclusion that politics must replace metapysics.

However, there is another problem that I take to be even more significant. Even if we accept that, in order to be rational, science must have an external goal, Trigg never adequately establishes that this goal must be knowledge, or that the possibility of knowledge requires a commitment to realism. Without establishing this, there is no reason to think that pragmatism or anti-realism is incompatible with the rationality of science.

In his discussion of pragmatism, for instance, Trigg says that in doing science, ``we need to have some idea of what is the case. In order to underpin science we need to appeal to the actual structure of physical reality'' (p. 56). Otherwise it could be that ``we are harboring a major illusion about the success of science,'' (p. 56), ``[w]e can be driven to wonder whether science actually works as well as we thought'' (p. 56), and `[w]e are still left wondering why science works'' (p. 57). The idea seems to be that if science does not have a metaphysical grounding, then we we cannot explain why it works, which is inadequate. Why is this inadequate? It seems that he says that it is because if there is no explanation of why something works, then it does not work. He suggests that science would not be successful if ``[o]ur ability to manipulate the world in certain respects could be a lucky accident'' (p. 56). It is not clear how this could be an objection: lucky accidents can be useful. Furthermore, it is not even clear how it could be accidental that science works, given usefulness as the criterion for deciding between theories. More by the way of argument is needed.

This book is a disappointment. It deals with central issues in the philosophy of science and advances positions that are reasonable, but the arguments are so unpersuasive as to undermine its attractive features. I am afraid that this book has not advanced the debate. Nor do I think that it is really suitable as a textbook. The topics and arguments are not as well organised as one would like them to be in a book for students.

CHRIST CHURCH, OXFORD ROBERT L. FRAZIER



Robert L. Frazier