The Truth of Science: Physical Theories and Reality
Harvard University Press, 1997
£17.95 hardback (viii + 260 pages)
ISBN 0 674 91092 3
Review by Brian D. Josephson
Scientific knowledge is abstract (in the case of physics particularly), and linked only very indirectly to the world we perceive with the senses. To what extent is science concerned with a truth independent of the activities of scientists, and to what extent is it merely a 'social construction', consisting of the activities of a particular culture?
Such questions are the fundamental concern of this book, largely driven by disquiet at the claims of sociologists who have advocated the social construction point of view. Unfortunately (it seems to me), the author is neither enough of a philosopher nor enough of a sociologist to appreciate the issues fully, so that the book becomes rather of the nature of an apologia for science (but a interesting and instructive apologia nevertheless).
What is it about scientific 'truths', over and above social processes, that entitles us to consider them as true? Newton discards as too simple-minded the idea that scientific assertions are simply 'assertions of how things are', so an alternative is required. His alternative is the coherence or overall consistency of a large and complex body of scientific knowledge. As an illustration of this coherence (exemplified by the fact of the non-existence of an engine that runs without the input of energy) he asserts that 'when a well-corroborated theory implies that a phenomenon will never occur, it will, indeed, not happen'.
The problem with this particular line is that the history of science contains a number of examples of laws that appeared at one time to be universally obeyed but which later were found to have exceptions. All of chemistry appeared to confirm the law that elements could not be transmuted, while the physics of the past seemed to imply that space and time were absolute. The moral to be drawn is that no matter how coherent a scientific system may be, there is no guarantee that facts that are inconsistent with it will not emerge in the future. In the end Newton acknowledges this, and shifts to the idea that science, 'Kuhn notwithstanding', makes progress, and approaches the truth ever more closely over the course of time.
The reference to Kuhn is one to a statement of his that science does not move towards anything such as truth, but simply evolves. Attacking this position, the author quotes Shimony, who uses a supposed analogy between the deciphering of a coded text and the problem of understanding nature to argue that there could not be progress in science without there being some unambiguous truth that this progress was being made towards. But the analogy is unconvincing; it is like saying if a society continually makes progress then there must be some unique perfect society that is the limit of this progress (or more mathematically, that if we have an ascending sequence of integers there is some integer that is the limit of that sequence). The whole notion of scientific truth is so troublesome that it seems much better to avoid it, and to discuss the nature of scientific progress instead.
What of the idea, which Newton seems to detest, that science is an 'endeavour to construct reality'? One cannot argue with the idea itself; the objection is to the idea that that is all there is to science: Newton wants there to be something independent of the scientist that underlies the outcome of such endeavours. But it is not clear that relativism in itself excludes that possibility. If I stand on a mountain, I see things that have an existence independent of myself, and yet what it is possible for me to see depends on where I stand.
Some of Newton's criticisms of Pickering's glosses on relativism I can accept. Nevertheless Pickering arguably is right in contexts such as string theory, where different versions of theories that claim to be descriptions of reality come and go like fashions. Here objectivity fights against the human desire to profit from the latest ways of thinking, as (to quote Pickering) scientists attempt to do the thing that will bring them as individuals the greatest gain.
A better book could have been written if the central theme had been not scientific truth but scientific progress, and the question of what it is about science and its methods that allows it (sometimes) to make clear progress by resolving inconsistencies in a widely acceptable manner. What is important in this context, as is made clear, is that the methods of science (such as making experiment rather then theory the ultimate arbiter of the truth) leave less room for personal opinion to determine what is correct than may be the case in some other disciplines (though some scientists seem endemically given to underestimating the degree of rigour of other disciplines, particularly in the humanities, and to elevating science to a status of uniqueness that it perhaps does not deserve).
As a final point that I feel is instructive, Newton seems to think in the case of parapsychology that its 'incoherence with the rest of accepted science' is a good reason for ignoring the experimental evidence. As has already been explained it seems not, in the light of the history of science, to be a very good reason. But the fact that scientists are inclined nonetheless to see it as a good reason seems to demand a sociological explanation (e.g. in terms of taboo). The saga of parapsychology refutes the suggestion that 'science is objective to the extent it avoids bias ... because the public character of science produces a balance with that effect.' The public character of science can help foster a view consistent with revealed reality, but can sometimes instead produce blinkers that help inhibit such a state of affairs.
I do not think that Newton has made the case that he hoped to have made, but the book nevertheless makes very interesting reading for its analyses of how science works. It is also provides for the scientist particularly a useful introduction to relativist ideas.