The Critical Rationalist Vol. 03 No. 01 ISSN: 1393-3809 17-Apr-1998
(35) Even if I am not right about this, I am convinced that the infinite content of our theories is of no great significance when we seek to explain the prodigious and perhaps unquenchable feats of technology. This may seem a blatant platitude, since most of the technical objections made above fizzle out if `very large but finite content' is substituted for `infinite content', whereas the attractiveness of the explanation itself is hardly affected. But I believe that the matter goes deeper than that.
(36) The principal consideration here is that theories are not exploited in practice by exploiting their consequences one by one, either routinely or imaginatively, but only in a negative way. What is important about scientific theories is not what they allow (which is a great deal, most of it useless), but what they forbid. The point was made by Popper (1957), §20, and has been well stated by Albert (1968), p 221: `The foremost function of the nomological science, under practical aspects, is to point out limits of realizability'. I have expatiated on the matter at some length in (Miller 1994, Chapter 2.2g, especially pp. 39-41), but Alain Boyer has made me realize that I slightly overstated my case there. I should accordingly like to present the case anew, though more briefly.
(37) The crucial point is indeed stated perfectly well by Percival himself (§52):
I must make clear at this point that I do not subscribe to the popular view that every technological decision and action (including inventions) is prescribed by one or more scientific theories; in fact none are. This would overlook the fact that scientific laws are universal and therefore can only proscribe; alone, they can tell us only what cannot happen, not what will happen, and therefore alone cannot tell us what we should do to achieve a given end. Building a bridge, car, space-ship and tube of toothpaste is a matter for engineers discovering sets of constructible [I should prefer to say realizable] initial conditions that typically lead efficiently to the desired result. This is a conjecture and refutation affair. Universal theories of science help the engineer insofar as they can be used to eliminate some of the hopeful candidates of efficient sets of initial conditions, namely the ones whose description contradicts the accepted scientific theories.
(38) Our theories themselves, that is to say, do not describe positively any event in the world. Only when they are supplemented by initial conditions, do they say anything categorical (to use old fashioned jargon) rather than something that is merely hypothetical. But that is exactly what inventions are: supplementary sets of (reproducible) initial conditions, without which our scientific theories are practically sterile. This is not to deny that scientific knowledge can provide inspiration. Electrodynamics suggests innumerable possibilities for using electric currents to generate motion; but it gives no hint as to how a workable electric razor, for example, may be constructed. Nor is it to deny that the historical sciences (such as cosmogony, geology, evolutionary biology) are concerned with singular statements as much as with universal theories. But these singular statements are not what technology is looking for.
(39) Once we have a proposed set of realizable conditions (invention) in front of us, we may be able to use our scientific knowledge to show that it will not do the job it is designed for. For this purpose the representability of our knowledge as an infinite collection of separate items of information is neither here nor there: the critical deployment of a component of a scientific theory can, objectively & logically speaking, be replaced by the critical deployment of the whole theory. Of course, that is not the only possibility: indeed, our knowledge may imply that the invention will be successful, and even explain why it will be successful. But in each case the invention must be proposed before the theory can be put to work. Inventions are not generated by a positive application of theoretical knowledge, though they may be eliminated by a negative application. It is in this sense that science can tell us only which inventions, which practical proposals, will not work, and cannot tell us which ones will work. As I say, a practical proposal may be explained by science, but it will never be a mere consequence of it.
(40) The Baconian doctrine that science provides the key to the mastery of nature is not rendered less false by being endorsed by Nietzsche in The Will to Power (`Science is the transformation of Nature into concepts for the purpose of mastering Nature') or by Habermas. Each confuses science with technology, thus masking the true relationship between the two. As several writers have recently observed (Grove 1989, p. 26; Wolpert 1992, p. 28; Stevenson & Byerly 1995, p. 2), for most of their histories science and technology proceeded independently of each other, and if there was any influence one way or another it was almost always from technology to science. Why should this have been so if science is, as so often thought, the inspiration of technology? The plain answer (not offered by any of the authors mentioned) is that science has nothing to contribute to technology except for criticism; and that this criticism can always in principle be replaced by empirical rather than theoretical criticism; that is, by practical testing. What is characteristic of modern technology is that many practical tests that on grounds of ethics, safety, or cost, simply cannot be performed are replaced by theoretical evaluation, evaluation that takes account of well tested scientific laws. This does not make technology a part of science, let alone an offshoot of science. And as I say, in principle the scientific evaluation is always expendable. Since the purpose of a technological innovation is to be practically successful, here at least the test of practice may be allowed to rule supreme.
(41) To be sure, inductivists and others who imagine that our scientific knowledge is open to justification or to empirical support usually imagine also that this justification seeps down to the inventions whose efficacy our knowledge has the ability to explain. Classical mechanics, inductivists will say, does not only explain the operation of, say, a barometer or a system of pulleys, it also justifies our confidence that such items of machinery will work properly. This surely is a service that science can offer technology beyond the purely eliminative one that I have outlined. But sadly, none of this is true. There exists no process of induction that allows scientific knowledge to attain any positive degree of justification. It is because of this that justificationist anti-inductivists such as Watkins (1984) regard the pragmatic problem of induction (the problem of how science is rationally applied) as the Achilles heel of falsificationism. They are quite wrong to do so. The application of scientific knowledge, like the growth of knowledge itself, consists only of conjectural forays into the unknown and eliminations of failures.
(42) Almost all these points may be made also about the contribution that mathematics makes to technology. It is transparent that a mathematical calculation on its own yields no positive information, and cannot help to solve any practical problem. A calculation--an interpreted one, of course--may in contrast make manifest the shortcomings of a proposed solution; though again a practical test or series of tests might serve as well. It is of some interest that mathematics seems to have been harnessed in this way in technology long before empirical science was.
(43) What is crucial to the growth of technology is that we maintain our level of inventiveness, our ability to conjure up ingeniously wrought assemblages of initial conditions with which to supplement our scientific theories. `In science we investigate ... reality; in technology we create a reality according to our design' writes Skolimowski (1966, p. 374; quoted by Grove 1989, p. 46). `Technology, unlike science, is not concerned with things as they are but with things as they might be' says Grove (loc. cit.). In other words, the growth of new invention requires a certain openness of the universe to our meddling, rather than the opposite. If the true theory of the cosmos were, as determinists pretend, in need only of a handful of initial conditions in order to have the power to predict all that would happen, then our scope for innovation would be sadly curtailed. Far from its being the case that technological progress is attributable to the infinite content of our scientific theories, such progress is entertainable only because our theories are not too strong. Animals with limited behavioural repertoires--that is, with endosomatically entrenched theories that go a long way to determining their interactions with the rest of the world--have less scope than we have for moulding the world to their desires. Nothing could show more clearly that it is not scientific knowledge that provides an inexhaustible economic resource, but extra-scientific (or perhaps para-scientific) invention.
(44) I cannot quite understand how Percival can have endorsed trenchantly the view that scientific theory has only negative impact on technology, and then at once have proceeded to ignore it. Dare I suggest that he may have moved from the perfectly correct claim that the greater informative content a theory has the more it excludes, through the equally correct claim that the stronger a theory is the more technologically useful it is, to the incorrect conclusion that it must be the inexhaustible content of our best scientific theories, if anything, that is the engine of unlimited technological growth? The argument is invalid because the principal service that science provides to the development of new inventions is a garbage-disposal service. (After the inventions have been invented, science may explain why they work.) Far from promoting technological growth science frustrates it (especially if our theories are false and encourage us to eliminate inventions that would actually function). The faster we clear away mistakes, no doubt, the better, and highly informative theories will be more efficacious in this regard than are weaker ones. But technological success cannot be built on the elimination of errors alone. Weeding a garden, though essential for growth, is not enough to make even a single flower bloom.
The Critical Rationalist Vol. 03 No. 01 ISSN: 1393-3809 17-Apr-1998
Copyright © 1998 All Rights Reserved.
TCR Issue Timestamp: Fri Apr 17 07:52:54 GMT 1998