COLLINGWOOD IN A NEW GARB
Bromberger’s modification of the deductive theory of explanation is, a deduction is an explanation if and only if the premiss answers the question, why is the conclusion true. For example, if we ask why is the sky blue, the theory of dispersion fluctuations answers our question and so explains why the sky is blue.
But to say that all skies with atmospheres are blue, our sky has an atmosphere therefore therefore our sky is blue, is proper deduction but it is an explanation to a smaller or narrower extent. And to say that Tom’s coin is golden because it responds properly to a touchstone will not do at all, because to say so is not to answer any why- question (it would answer a why-question if it talked about our belief that Tom’s coin is golden).This is very convincing. We cannot answer, why does a tuning fork vibrate? by, because it is made of a continuous medium. Here the continuum theory of elasticity is evidently non-explanatory. Even to say, the tuning fork vibrates in a series of harmonics each subject to force proportional to the width of the vibration, even this is recognized as not an explanation but a highly sophisticated restatement of the question. Perhaps, when we accept the Newtonian schema of explanation we deny that decomposing a vibration into series (Fourier analysis) is explanatory, and also that the acceleration of the parts of the vibrating fork is proportional to the width of the vibration; but we shall, then, accept the force causing this acceleration as explanatory. And here, when Bromberger says explanations are deductive answers to why questions, he enables us to distinguish between essential and accidental theories, as well as between philosophical and mathematical or fictional; when we philosophically accept Newtonian forces we accept a part of elasticity theory as explanatory, and the anti-atomic part of it as mathematical or fictitious; otherwise we declare the whole of it fictitious.
So far, then, so good. But Bromberger has only raised again our old question, whereas he seems to think that he has answered it.
Things are, really, very difficult and baffling. We do not even quite know what is a question. We seldom consider the question, what is the distinction between x and y? a question proper - except in exams. (Actually, it is ungrammatical because of the misuse of the definite article; discussed at length in Chapter 2 above). Popper declared all what is...? questions bad as they conform to methodological essentialism, tabooed since Galileo and Bacon (both of them epistemological, but anit-method- ological, essentialists). Yet what is...? questions can easily be converted into why-questions, and so not outlawed, unless we can answer the question, what is a question? But this question is taboo as well, and so we may convert all what is...? questions into why-questions. However, why- questions are very puzzling. Yes-or-no questions, as well as w-questions, to use Carnap’s term (i.e. questions of the kind, who, where, which, how), are treated by erotetic logicians as multiple-choice problems. These, though quite bothersome in so many aspects, are fundamentally less problematic than questions which are not multiple-choice questions. How, what, and why-questions can be multiple-choice but need not be. When we have a question which is not a multiple-choice question we may have a definite and clear ability to recognize an answer to it, yet be utterly unable to conceive or to discover any answer to it. This is what Bromberger calls a p-predicament. The existence of p-predicaments is a great puzzle: how is it that I feel so easily and clearly that x is or is not an answer to y the moment it is given to me, yet a moment earlier x was as remote from me as the mystery of the universe?
How is a p-predicament possible? I do not have an answer to this question, but I do have a hint of an answer, a matrix which I hope will be filled and thus provide an answer - true or false - to the question, which I have posed in Chapter 9 above.
My hint of an answer is not substantially different from that of Bromberger. It may be introduced by noticing the fact that Bromberger’s description of the p-predicament is not complete. There is little doubt that we can conceive of an historical instance of a person in a p-predicament concerning a given question, who died without having any answer to it, and who would have recognized, had he lived longer, an answer by one of his successors as an answer to his question; yet who would not have recognized an answer to the same question given by a much later successor. I shall not even bother to outline an example - it is too easy to do.And so, my correction or supplementation of Bromberger’s description of a p-predicament is as follows. A person in a p-predicament is puzzled over a question to which he has no answer whatever, who feels that presented with a putative answer he will immediately perceive that it is or is not an answer to his question; but whose feeling is justified by the facts up to a point and not necessarily to the full. I dare say my supplementation to Bromberger’s description is rather trivial because no one ever said that the list of answers to every question is finite; indeed, some questions, we know, avail themselves of infinitely many answers, and it has not been claimed by anyone that given an infinite list of answers anyone can decide, at once or at any given length of time, whether a given answer is included in that list or not.
Hence, the fact that a person is able to decide, in some cases, that a given answer is an answer to his question, must be relativized to a general matrix of answers that the person has, to some general presupposition, articulated or not, that a person has regarding any possible proposition which in the future may pose as an answer to his question. This general presupposition, when unarticulated, is a person’s point of view, and in any case it is his metaphysics regarding the question at hand.
And so, at last, I have managed to close the circle, and relate the father of the logic of questions and answers, R.
G. Collingwood, to his latest successors. It is not a mere sentimental point, but there is a logic of the situation which drives Bromberger to discussions akin to those of Collingwood. It is possible to show that articulating metaphysical viewpoints, i.e. matrices of answers to given questions, turn them from mere puzzles and p-predicaments to manageable w-questions, to sort of multiplechoice questions. I have argued my view on this point elsewhere at length, and so has Bromberger, whose view is not too different from mine. I shall not elaborate; I shall give a quick and simple instance.The problems solved within Newtonian metaphysics are all why questions turned into which questions, p-predicaments turned into multiplechoice. Newtonian metaphysics views the universe as a conglomerate of diverse kinds of particles each kind exerting a kind of central force on its own kind. So the question of celestial mechanics becomes, what is the central force acting between heavenly bodies, and the answer is, one proportional to the inverse square of the distance. The question what makes Boyle’s law true is answered, a central force proportional to the inverse of the distance. What makes a tuning fork vibrate? A force proportional to the distance. Electricity? The inverse square of the distance, again. And so, we have infinitely many functions of the distance, and each question is translatable to the question, which function? A why question is thus turned into a which-question.
This is an over-simplification. In some cases, though the above picture was deemed true, it was not deemed useful enough, and more ingenious methods of turning a why-question into a which were used. Laplace’s theory of the tides and his theory of capillarity are such instances. But I shall not dwell on this here.
What is more exciting about all this is the idea of not answering but of criticizing a question - which we have met with Harrah and Belnap. A why-question, a p-predicament, cannot be easily criticized.
Why is the sky blue? can only be criticized by the contention, the sky is not always blue. To which the response should be, why is the sky bleu under conditions x? This, no doubt, can be criticized again and lead to the response, why is the sky blue under conditions yl Such dialectic really cannot take us away from the level of observed phenomena, no matter how sophisticated it may otherwise become. Not so when we transform a why-question into a which-question.When we change why is the sky blue? into which central force renders the sky blue? we can criticize the question. Today we are all content that the sky is blue not thanks to any central force but thanks to fluctuations of distribution of atoms of atmosphere and thanks to non-central interactions between them and photons. We have transcended Newtonian mechanics by showing that it handles certain why-questions poorly. To use the idea of Aqvist in the language of Harrah and Belnap, we correct a false question here, and thereby, I should add, transcend one viewpoint, one set of presuppositions and go more deeply to a newer viewpoint, to a newer set of questions.
There is much more to say on this topic. Since the list of answers to a given question is, in Newtonianism, quite infinite, it is, in a sense, impossible to correct Newtonianism. Einstein was aware of this fact. He noticed that one can always correct Newton’s theory of gravity by choosing a function of the distance close to the inverse square. Yet he dismissed such attempts as too arbitrary. And so, my answer is but a matrix, merely an outline of an answer, not a full answer. It is under certain constraints which we must specify more fully, that we can correct a question and replace it by a deeper one, that we can transcend a viewpoint and attempt to construct a deeper one. I shall not discuss this issue any further. Let me merely conclude that to criticize a question is only a first step forwards replacing it with a better one.
When we translate a why-question into a which-question we do so from a given, often developed viewpoint.
When we transcend the question thus put, when we transcend the viewpoint from which it was put, the which- question fizzles out and we are turned back to our why-question, to our p-predicament. Or, alternatively, a b-question, one beyond our horizon, enters it as a new p-predicament. We may try to solve it directly, we may try to develop a new viewpoint in order to handel it; we may, more dialectically, try to do both with the aid of the partial success of each. This comes close to the theory of Bunge which I reviewed in Synthese, 1969: we try to integrate our general view of the world, he says, into a coherent picture. This, I say, we do by imposing a viewpoint on our background knowledge. Not successfully, of course, but with partial success.Hence, when we ask a why x? question, we want a theory from which x follows, but not necessarily any theory. We want one which conforms to present day metaphysics, or one which may help us develop a new metaphysics, as the case may be. (These two cases parallel somewhat Kuhn’s normal science and revolutionary science). Hence, what explanation is causal depends on current, or on tomorrow’s, metaphysics, depending on the situation at hand. And we choose questions which our metaphysics turns from why-questions to which-questions, sooner or later arriving at a stumbling block which may force us to alter our viewpoint or metaphysics altogether. Still better, and closer to reality, we may operate with more than one set of presuppositions, turn a why-question into two different and competing which-questions, and see which is preferable. This is a sort of crucial experiment between sets of presuppositions of metaphysical points of view - a crucial experiment not like the one in science which selects one answer from a set of answers to a given question, but which selects one translation to a which-question from a set of such translations to a given why-question. But all this is very far from being put into a formal language, or even into a fairly rigorous presentation, and it is still all up in the air.
This, then, is my view of the dialectic interaction between science and metaphysics, presented in the light of the latest developments in the logic of questions and answers.
It is in this context, or a similar one, that we can see the force of Robert S. Cohen’s introduction of why-not-questions. For, formally, there is no difference between why- and why-not-questions: why-not-x becomes why-7 when y is defined as not-x. But his questions, such as why there and not here (i.e. western but not Chinese science), why this and not that, make sense from a given specific viewpoint, from which one may wish, as Cohen says, to develop a logic of comparative analysis. For more details the reader is invited to examine R. S. Cohen’s ‘The Problem of 19(k)’ in the Journal of Chinese Philosophy 1 (1973), 103.
appendix: the anti-scientific metaphysician
The contemporary instrumentalist tradition stems from a frankly, and even initially, anti-scientific tradition.1 It is the traditional view not that the world is flat (Berkeley, Hume)2, but that the world of science is flat whereas the world of metaphysics is deep. This tradition, I suppose, goes back to Hegel. T. M. Knox, the English translator of Hegel’s Philosophy of Right, ascribes this view to Hegel, and contrasts “The categories of ‘essence’” with “mathematical and empirical science or... formal logic” and even “those philosophies which adhere to scientific method instead of abandoning it in favour of reason and the philosophic method...” I tend to agree, and find adumbration of this view in the Preface to the Phenomenology3, but I confess I find no clear statement of it anywhere in that work, in the Logic, the Encyclopedia, or in the Philosophy of Right. I suppose it is what Hegel would have liked to say; but it took about a century to learn to say it as crisply as contemporary Hegelians have learned to say it, judging from Bradley’s aptly named Appearance and Reality of 1893.
Sir William Hamilton commended logic as the science of essence and condemned mathematics as merely formal (i.e., descriptive, or purely instrumental). George Boole’s earliest publication is couched in the form of a retort to Hamilton: if logic is part of mathematics then his view is absurd.4
Subsequently the anti-scientific tradition jettisoned not only mathematics but also formal logic - as Hegel had recommended in the first place, according to T. M. Knox. Let me quote, though, some contemporary authors who are sufficiently influential and who hold this view, such as Croce and Sartre.5 Let me begin with Croce.
Science... cannot be anything but... philosoph. If natural sciences be spoken of, apart from philosophy, we must observe that they are not perfect... The so-called natural sciences... are surrounded by limitations... They calculate, measure... Even geometry... rests altogether on hypotheses... What of properly naturalistic they contain, is abstraction and caprice... The concepts of natural science are, without doubt more useful; but...
And here is what Sartre says:
... the further research of the scientist will reveal it as purely a thing - i.e. stripped of all instrumentality. But this is because the scientist is concerned only with establishing purely external relations. Moreover the result of this scientific research is that the thing itself, deprived of all instrumentality, finally disappears into absolute exteriority.
... the concept of objectivity, which aimed at replacing the in-itself of dogmatic truth by a pure relation of reciprocal agreement between representations, is selfdestructive if pushed to the limit.
But science has more extreme opponents. I shudder to quote Heidegger; even his interpreters are often too unclear. It seems to me that Paul Ricoeur claims that Heidegger’s Holzwege advocates the instrumentalist view of science;6 but I cannot say.
The following quotation from Maurice Merleau-Ponty7 is both conventionalist-instrumentalist and apriorist:
Husserl says in the first volume of the Logical Investigations that the physicists proceed by “idealisierende Fiktionen cum fundamento in re" - that is, by idealizing fictions which are nevertheless founded in the facts. Let it be, he says, the law of Newton. Basically it makes no assertion about the existence of gravitation masses. It is another one of those idealizing fictions by which one purely conceives of what a gravitating mass would be. Then one determines what properties it would have, on the supposition that it exists. According to Husserl, Newton’s law says nothing at all about existence. It refers only to what would belong to a gravitational mass as such.
The method actually used by physicists, therefore, is not the chimerical induction of Mill... It is rather the reading of the essence... That which gives it its probable value... is rather the intrinsic clarity... ideas shed on the phenomena we seek to understand...
I think the main question to ask about these people is what do they think sticks and stones and the cow and the moon are made of, or how and why their machines work. It seems obvious that they do not care, and so their views on these issues are vague, confused and outdated (see Sartre’s discussion of contemporary physics, loc. cit.). These views, however, not the deep metaphysical doctrines which their advocates propound, are what these advocates offer instead of sience- unlike Bellarmine or Duhemwho accept Aristotelian physics; and unlike philosophers who, with Hume and Mach, think the world is flat; and unlike Bohr and his followers whose discussions still reflect efforts to create newer and deeper pictures of the world. Their instrumentalism is simply the best excuse they have for their ignorance of and disregard for science.
NOTES
1 Sir Karl Popper, ‘Three Views Concerning Human Knowledge’, reprinted in his Conjectures and Refutations, London 1963.
2 Popper, ‘A Note on Berkeley as a Precursor to Mach’, op. cit. See also p. 211 above. 8 G. F. W. Hegel (transl. by T. M. Knox), Philosophy of Right, OUP, 1952,1967, pp. vii-viii and The Phenomenology of Mind, OUP, 1967, p. 88.
4 George Boole, Mathematical Analysis of Logic, Cambridge 1847, opening.
5 Benedetto Croce, Aesthetics as Science of Expression and General Linguistics, New York 1953, p. 30. J.-P. Sartre, Being and Nothingness, special abridged edition, 2nd paperback edition, New York, 1965, pp. 176, 283.
6 M. S. Frings (ed.), Heidegger and the Quest for Truth, Chicago 1968, p. 68.
7 Maurice Merleau-Ponty, The Priniacy of Perception and Other Essays, Evanston 1962, p. 69.