Science in Christian Perspective
Clockwork Image Controversy
DONALD M. MacKAY
Department of Communication
University of Keele
From: JASA 28 (September 1976):
I am afraid that Cramer has sadly misunderstood each of the arguments he criticizes. Because I too am anxious that our "front-line troops?' should not be sold useless weapons, I have accepted the Editor's invitation to reply.
Let me begin with three points on which there is no dispute whatsoever:
1. Occam's methodological principle, that "entities
should not be multiplied beyond necessity", is
crucial to rational thinking in science and elsewhere,
and must be retained and respected. (Cramer, for
reasons to be examined below, accuses me of a
"direct denial" of it.)
2. There are no biblical grounds for expecting special events such as the resurrection of Christ to be explicable on the basis of scientific precedent. (See The Clockwork Image, p. 64). In that sense, neither Cramer nor I believe that a scientific explanation of the universe could in fact "account completely for all events", even at its own level (see below).
3. Scientific determinateness is not to be equated with predictability. Our inability to predict an event (as I expressly point out on p. 15 of The Clockwork Image) does not mean that it is indeterminate. "To describe a system as indeterminate is to imply not only that it is unpredictable, but (much more
Far from "denying" Occam's razor,
The Clockwork Image
strongly) that there exists no determining formula whereby future state-descriptions are completely and necessarily implied by present data."1
What, then, has misled Cramer, whose whole attack is designed (no doubt with the best of goodwill). to give the opposite impression of my position? Basically, it seems to be his failure to grasp the central contention of the book, that (p. 90) "different levels of description of reality are logically necessary in order to express all that truthfully needs to be said about it ... Each reveals an aspect which is there to be reckoned with, but is unmentioned in the other". "One and the same situation may need two or more accounts, each complete at its own lever'. So far from "denying" Occam's principle, these statements presuppose it! They claim that in certain circumstances there is necessity not just for "more complicated terms", but for a whole new level of conceptualization, if we are to do justice to all that is there to be reckoned with. There is here no question of multiplying entities without necessity. The necessity is there to be discovered empirically, by finding that there is indeed something (or someone) to be reckoned with at the higher level.
My various examples (such as the advertising sign) that Cramer mistakes for "arguments by analogy" are simply proofs of the converse thesis, that the necessity for a given level of concepts cannot be disproved by demonstrating the completeness of another account at its own level.Completeness
Evidently, despite all the illustrations, this notion of "completeness at its own level" has misled Cramer. To be sure, if an explanation of a situation in given terms were complete in the sense of answering all possible questions about the situation, then Occam would say that further concepts are unnecessary. But if, as in the case of an electronic computer (p. 72) the explanation in electronic terms does not even mention the mathematical significance of what is going on, the fact that it accounts completely for every electronic event does not in the least "force us to reject as superfluous" the mathematician's account in other terms. Anyone (with or without a "secular mind") who insists that the electronic explanation (because complete-in-its-own-terms) is the only valid or necessary one, would be self-evidently missing the point. Occam's razor offers him no defence, for the mathematician is amply prepared to show the necessity and relevance of his categories as well as those of the electronic engineer. There is no circularity, no violation of parsimony here. All the blunders would be on the part of the imaginary "nothing-butter".
My attack on anti-religious reductionism here, in other words, is a reductio ad absurdum. In order to show that the form of an argument is invalid, it is enough to find one good counter-example in which that form of argument would lead to an absurd conclusion. This is the purpose of my illustrations of nothing buttery"-to show that the logical form of the ontological reductionist's argument is unsound, and not at all to argue by analogy from man's purposes to God's purposes, or anything of the kind, In my experience of "front-line combat" this reductio ad absurdum has proved quite a useful, if modest, weapon.
It is a pity in this connection that Cramer seeks
support from the unfair remark by the agnostic Beloff.
The context2 was a discussion of mechanistic theories of
brain function, which I claimed were not ruled out by
the biblical doctrine of man. I believe that this is
strictly true and defensible; but i an case, it offers
no logical justification for the charge t at on this view
Christianity is compatible with Materialism, still less
with "anything at all". To test the coherence of Cramer's
logic at this point, consider again the computer. The
electronic engineer has no need of the categories of
the programmer in order to explain the events he observes in the computer; both his explanation and that
of the programmer, like the physiological and the biblical accounts of man, are true and necessary for their
respective purposes. But does this mean that the two
explanations "cannot conflict because they cannot contradict"? Is the mathematical significance of what is
going on therefore "useless baggage" to the engineer,
"compatible with anything at all"? Such talk makes
good rhetoric but bad logic. Front-line troops, beware!
Limits of Science
I made three main points (mainly pp. 36-39) about the "limits" of science.
1. They are methodological, not territorial. They arise wherever data are accessible only at the cost of participation, so that detachment is ruled out. (This covers far more than the "human emotional experiences" mentioned by Cramer, see ref. 3.)
2. They are self-imposed. "No part of the world of observable events is outside the boundary of scientific study . . . The limitations (of science) will show up rather in the restricted kinds of description his language allows (the scientist) to make of the events he studies, and the kinds of point be will be obliged to miss (theoretically) in consequence". (It is in elaboration of this point that I go on in the same section to expose the fallacy of "nothing-buttery".) Where the data in question are accessible only through becoming oneself involved as a participant (as in our experience of God), then (p. 39) scientific detachment is ipso facto impossible, and science is not arbitrarily excluded but "bows itself out". As I made clear (p. 36) this does not in the least preclude the behavioural manifestations of religious or any other experience from being scientifically studied by an outside observer for what they are worth.
3. Nevertheless, to refuse to expose oneself to knowledge available only through participation (such as is the knowledge of God), on grounds of "scientific conscience", would be irrational, and contrary to the scientific spirit of openness to evidence.
It is a pity that Cramer fails to specify the "materialistic and positivistic arguments that so impress him. This makes comment impossible.Determinism
On this Cramer has quite missed the point of the argument in The Clockwork Image. Briefly, if all you believe were rigorously represented by the detailed state of your brain, so that any change in your belief required a correlated. change in brain-state, then no completely detailed specification of your immediately future brain-state could exist, with an unconditional claim to your assent (i.e., such that you would be correct to believe it and in error to disbelieve it). This has nothing whatever to do with counter-suggestibility on your part; it is simply a logical consequence of the assumption of complete correlation. It proves not just unpredictability (the impossibility of discovering a complete specification) but (logical) indeterminacy (the non-existence of a complete specification with an unconditional claim to your assent), even if your brain were scientifically-determinate. So no complete specification of your future can exist, unknown to you, that you would be correct to accept as inevitable now if only you knew it. In that sense, even if future brain states of yours were scientifically determinate, and predictable by others, they would not be inevitable (in every detail) for you.
To the extent that your future actions depended causally on these logically-indeterminate details of your brain-state, they too would not be inevitable-for-you. As I point out, all this requires us to recognize a logical "relativity principle". according to which the agent-view and the observer-view of a future action must differ systematically if both are to be correct.4,5.Finally, I pointed out that this whole argument
Cramer has sadly misunderstood each of the arguments he criticizes.
relates to cognitive agents who have an I-story" to tell. Unless we credit computing machines with conscious agency, it would be not so much false as nonsensical to call them "morally free", even if programmed as Cramer suggests. This point like several others he raises, is explicitly mentioned in the Appendix to the book, and in references given in Chapter 8.
I am sure that with Cramer's help I could have made The Clockwork Image a better book. As it stands it is evidently open to even greater misunderstandings than I had thought possible. I hope that this exchange may help fend off at least a few of these.
2J.R. Smythies, ed.: Brain and Mind, 163-200, and Comments on pages 129-131, Routledge and Kegan Paul, 1965.
3MacKay, D.M.: Information
and Prediction in Science (S. Dockx and P. Bernays, eds.), Symp. of
Int. Acad. for Phil. of Sci., 1962, Academic Press, New York, 255-269,
4MacKay, D.M.: On the Logical Indeterminacy of a Free Choice, Mind, 69, 31-40, 1960.
5MacKay, D.M.- Freedom of Action in a Mechanistic Universe, (Eddington Lecture), Cambridge Univ. Press, London and New York, 1967. Reprinted in Good Readings in Psychology (M.S. Gazzaniga and E.P. Lovejoy, eds.), Prentice Hall, 121-138, 1971.