A new fabric of reality?

June 5, 1998

Are today's scientists no better than the priests of the Inquisition, who forced Galileo to declare that his scientific theory was not a description of reality but merely an observation of appearances? David Deutsch reports.

Science in the modern sense began with Galileo's conception of a law of nature: a universal statement about reality tested by what he aptly called "ordeals" (we would call them crucial experiments). Ever since then, a recurrent theme in the history of science has been the tension between two great purposes that are implicit in Galileo's conception: science as a means of making predictions and giving us control of the world; and science as a means of understanding what the world is really like.

No one claims that prediction and understanding are incompatible. Yet they sometimes pull in different directions, and questions about how they are related have been the subject of bitter controversy. This is not just a storm in an academic teacup. Scientific progress itself has always depended on answering such "metaphysical" questions correctly.

Galileo's own confrontation with the Inquisition was a case in point. Following his brilliant advocacy of Copernicus' theory that the earth moves round the sun, Galileo was convicted in 1633 of heresy, sentenced to house arrest for life, and forced, under the threat of torture, to declare that he "abjured, cursed and detested" the Copernican theory. As Jacob Bronowski put it in The Ascent of Man: "The result was silence among Catholic scientists everywhere from then on ..."

But it is important to understand what Galileo was not forced to recant. The Inquisition did not object to his claim that Copernicus's theory provided a useful, pragmatic model of the solar system and gave accurate predictions of the effects of the workings of the solar system on the earth. Predictions, you see, refer only to observations and are therefore merely a matter of appearances, while what the Church cared about was reality. They were happy to cede utility and observational accuracy to the domain of science, so long as science was prepared to leave objective reality to them. And that is exactly what Galileo was not prepared to do. He knew that if his symbols were understood as referring only to appearances and not to objective reality then they, and the whole edifice of scientific reasoning, made no sense.

By a scandalous quirk of intellectual history, many of the most influential philosophers and scientists this century have in effect sided with the Inquisition in this matter, and against Galileo. Going variously under such names as "instrumentalism", "logical positivism" and "I'm just a simple scientist, I don't hold with metaphysics", the doctrine has been that the entire content of a scientific theory lies in its observable predictions. Explanations of why the predicted events come about, if couched in terms of entities that cannot be directly observed, are regarded as matters of taste: optional extras, not really part of science at all. Any consistent explanation that a theory may give for its predictions is deemed to be as good as any other so long as the predictions are borne out by experiment.

So was Galileo arguing about nothing when he insisted that the earth really is moving and does not just appear so to astronomers? Of course not. Certainly, performing crucial experimental tests is the defining method of science. But to think that surviving experimental tests is the purpose of our theories is like thinking that the purpose of an aeroplane is to pass airworthiness tests. Worse, actually, because one can fly an aeroplane into new territory without altering its design or knowing how it works. But in science, progress means better explanation, and all the other functions, methods and uses of science are dependent on explanation. Take medicine for example, one of the most pragmatic of the sciences. We may be able to recognise a disease without understanding its origin. We may be able to predict its symptoms and give a prognosis. If we are extraordinarily lucky, we may even be able to cure it with a medicine of unknown mechanism. But to make any further progress - say, to find a way of wiping the disease out altogether - only explanations will do. Predictions by themselves are useless.

From time to time, science comes up with a radical new theory which is not only successful at providing predictions and explanations in its own sphere, but forces a shattering change in our overall world view. Such was Copernicus's theory. Darwin's theory of evolution is another well-known example, and in this century, the theory of relativity and then that of quantum physics are among several shocks that have yet to be fully assimilated. It is at such moments that it is most tempting for people to fall into the instrumentalist error. "Let's not reject what can be observationally verified," they concede, "but why should we accept any metaphysical baggage about what lies behind our perceptions? Reality? That's just a word. What does it mean?" And thus some of them persuade themselves that the dinosaurs never evolved, but came into being already fossilised, together with the rest of the Earth in 4004bc. That's logically consistent with all known observations, isn't it? Thus, also, the Inquisition argued that observations made on earth were not the proper arbiters of disputes about the true nature of the sky.

And the same argument is still made today, against the most important implication of quantum theory: that the universe we see around us is only a small facet of physical reality; and that the whole of reality - the multiverse - contains many such universes affecting each other through the phenomenon of quantum interference. "Why should we believe that those other universes exist?" some people argue. "Why can't we just say that the universe we observe behaves as if they were out there, affecting it?" I wish I could pretend that those people were themselves not out there; or at least, that they did not really believe such nonsense, but were merely behaving as if they did. I wish it were unnecessary to explain that taking that view of quantum mechanics - the fundamental theory that underlies the whole of physics - impedes further progress and makes it well-nigh impossible to understand that theory or its astonishing connections with other fields of knowledge.

But it is necessary, because, this time, it is not only the die-hard opponents of the new theory who have retreated into instrumentalism, it is the founders themselves. In most branches of physics the prevailing attitude towards quantum mechanics has been pragmatic acceptance, combined with every sort of refusal to take it seriously as a description of reality. Moreover, this sad state of affairs has been repeated, with variations, in three other fields where important new explanations have been discovered: the theory of evolution (as updated by the Oxford scientist Richard Dawkins); Karl Popper's theory of the growth of knowledge in science; and Alan Turing's theory of the universal computer.

I believe that all of these theories should be taken seriously, not just in their own right, but jointly. For it turns out that they are so closely interrelated that none can be properly understood without reference to the other three. Recent research on quantum computers - machines that perform computations of unprecedented complexity by sharing sub-tasks among vast number of universes - provides one example of this.

Considered together, these four strands of explanation reveal a unified fabric of reality that is objective and comprehensible and in which human actions and ideas play essential roles. This is a rational world view that ought to be the prevailing one - not the final "theory of everything", but the starting point for the discovery of even better explanations, and the tentative standard against which all new theories should be judged.

David Deutsch is a researcher at the Centre for Quantum Computation, the Clarendon Laboratory, Oxford. His is one of six shortlisted books for the Rhone-Poulenc prize for science writing. The winner is announced next week.

This year's Rhone-Poulenc prize

The record number of entries for the Rhone-Poulenc prize this year - 109 - exceeded those for the Booker (104) and Whitbread (84) prizes, reflecting the continued growth of the public's interest in popular science books. The strength of this interest, which surprises many scientists, has confirmed my belief in people's desire to understand reality and their place in it.

Paradoxically, this trend goes along with the increasing alienation of young people from science, and the increasing popularity of every kind of pseudo-science and anti-science, as well as what has perhaps become the most dangerous of these trends, scientism (the specious use of scientific authority). But I think that all these trends have the same origin: the innate human sense of wonder, and the desire for things to make sense.

I hope that readers of the shortlisted books will find that science has what they want - not only in the sense of Winston Churchill's austere reflection that "facts are better than dreams", but quite literally: science has led us to facts beyond our wildest dreams, and ideas that make sense at levels that were literally unthinkable before. David Deutsch The Rhone-Poulenc shortlist

The Fabric of Reality David Deutsch Allen Lane

Guns, Germs and Steel Jared Diamond Jonathan Cape

Life: An Unauthorised Biography Richard Fortey HarperCollins

This is Biology Ernst Mayr Bellknap/Harvard

Fermat's Last Theorem Simon Singh Fourth Estate

Twins Lawrence Wright Weidenfeld and Nicolson

Please login or register to read this article

Register to continue

Get a month's unlimited access to THE content online. Just register and complete your career summary.

Registration is free and only takes a moment. Once registered you can read a total of 3 articles each month, plus:

  • Sign up for the editor's highlights
  • Receive World University Rankings news first
  • Get job alerts, shortlist jobs and save job searches
  • Participate in reader discussions and post comments

Have your say

Log in or register to post comments