Constructive Empiricism and Other Contemporary Surrealisms
Secular surrealism is the same as Bas van Fraassen's constructive empiricism. Van Fraassen's slogan is “The name of the scientific game is saving the phenomena”. A theory saves its phenomena if its predictions about observable states of affairs are true, if it is empirically adequate.
Strict or constructive empiricists should never assert that a theory is true, only that it is empirically adequate. But to say that a theory is empirically adequate is just to say that the phenomena are as if it were true. Yet another version of secular surrealism is involved in Kyle Stanford's abortive attempt to provide an antirealist explanation of the success of science. Stanford says that we need not invoke a theory's truth to explain its success, we can simply say that it makes the same predictions as the (a?) true theory would make. But to say that a theory T makes the same predictions as a true theory would make, is just a fancy way of saying that the observable phenomena are as if T were itself true.Sunday School precepts of scientific method help account for the popularity of surrealism. People say that what matters in science is predictive power. Well, T, TG and T* have exactly the same predictive power. People say that what matters in science is that its theories are falsifiable by observation and experiment. Well, T, TG and T* are equally falsifiable by observation and experiment. Vulgar American pragmatists say that the ‘cash value' of any theory is what it tells us about experience. Well, T, TG and T* have exactly the same ‘cash value'. Innocent victims of the verifiability theory of meaning say that theories that are verifiable by the same experiences are really the same theory expressed in different words. Well, on that view, T, TG and T* are really the same theory expressed in different words.
I think this is all wrong. Surrealism is not the same as realism. T, TG and T* are not the same theory, nor are they equally good theories. The difference between them has to do, not with predicting or saving the phenomena, but with explaining them. We may assume that the theory T with which we began explains its phenomena, shows how they are brought about, says what causes them. But T*, the secular surrealist transform of T, does not explain its phenomena. I might explain why the streets are wet by saying that it is raining. But I do not explain why the streets are wet by saying that they are as if it were raining. The ancients explained the (apparent) daily motions of the fixed stars by saying that they are all fixed on an invisible celestial sphere that rotates once a day on its axis about the earth located at its centre. But you do not explain the motions of the stars by saying that they move as if they were all fixed on an invisible sphere. So, if we are interested in explaining phenomena, not just in ‘saving' them, we should prefer the explanatory theory T, whatever it is, to its non-explanatory surrealist transform T*. Realists who value explanation will prefer T to T*. It follows that, for realists anyway, there are explanatory virtues that are not evidential virtues. It follows that realists are not strict empiricists, and do not accept that only empirical evidence should determine theory-choice.
Here, as my last example makes clear, a terminological point intrudes. Did the ancients really explain the motions of the stars, given that their theory is false and that the celestial sphere they postulated does not exist? Does it make sense to speak of a false explanation? As I use the words, this does make sense. This does not imply, of course, that explanation has nothing to do with truth. All it means is that truth is an adequacy condition upon explanation, not a defining condition of it. It makes sense to speak of a ‘false explanation', but it does not make sense to speak of a ‘false yet adequate explanation'.
Nothing hinges upon this terminological preference of mine. If you think that ‘false explanation' makes no sense, then speak instead of ‘false putative explanation'. My philosophical point still stands, but must be differently expressed. I must now say that we should prefer the putative explanatory theory T, whatever it is, to its non-explanatory surrealist transform T*.What about theological surrealism? This does explain phenomena, after a fashion, by saying that God produces or causes them. Better this supernatural explanation than no explanation at all. Many of Berkeley's readers thought God the weak link in his system, and removed Her from it. The result is phenomenalism, a secular surrealist metaphysic that has minds and their experiences but no explanation of why the experiences are as they are. If I had to choose between phenomenalism (secular surrealism) and Berkeley (theological surrealism), I would prefer the latter from an explanatory point of view. But we do not have to choose between surrealisms. There is a third alternative, realism about both common sense and science. Is realism better than theological surrealism, from an explanatory point of view? Are natural explanations better than an all-purpose supernatural one? And if so, why?
This is not an easy question to answer. Strict empiricists often invoke simplicity as a virtue as well as empirical adequacy. Duhem's slogan was that science aims to save the phenomena in the simplest possible way. But simplicity is not much help to realists. What is simplicity, for a start? Occam said that it required us not to multiply entities without necessity. But nothing is simpler than theological surrealism: only one entity is postulated as the cause of all observed phenomena. Occam's Razor seems to favour theological surrealism over realism. Besides, realists seek truth, and what has simplicity to do with truth? Even if naturalistic explanations could be made out to be simpler in some sense than a supernatural one, what does this tell us about their truth? How do realists know that Nature is simple?
Perhaps the difference has to do, not so much with the number of entities that we postulate, but with our ability to understand how these entities work.
This seems more promising. For it is admitted on all sides that we poor humans cannot comprehend how divine causality operates. “God said ‘Let there be light'—and there was light”. Gee, thanks! God puts tree-experiences into our minds as if there were trees. How does She do it? We have no idea, and are not meant to have any idea. We are just meant to comfort ourselves with the thought that She can do anything.The trouble with this response is that some naturalistic explanations are also difficult for us to understand. How do celestial spheres work and what are they made of? How does gravity work, or magnetism? How can light behave both as a particle and as a wave? Does anybody understand Quantum Mechanics, the best predictive theory we have ever had? Down the ages, strict empiricists have grown impatient with this demand for ‘understanding’ and claimed that saving the phenomena is all that really matters. Duhem famously made fun of Lord Kelvin, who sought mechanical models of electromagnetic action: “We thought we were entering the tranquil and neatly ordered abode of reason, but we find ourselves in a factory” [Duhem 1954, pp. 70-72]. Nowadays those who despair of understanding Quantum Mechanics opt for surrealism about it—it is called ‘the Copenhagen Interpretation of Quantum Mechanics’. But that brings us full circle. If the unintelligibility of electromagnetism or of Quantum Mechanics does not matter, neither does the unintelligibility of Divine causation. Indeed, better one big mystery than a host of little ones.
But is that really so? Reflection on the little mysteries that naturalistic explanations throw up has sometimes led to better and better explanations—even if ‘better’ is understood here in strict empiricist terms. Scientists seem to have found out more about how natural causes work, without perhaps ever understanding them fully. There is no finding out more about how divine causality works—that is beyond our ken. Such empirical progress as there is in theological surrealism is entirely parasitic on the progress of science.
Gosse said that God fixed things as if the geological theories of his day were true. But Gosse would not have minded if the geologists refuted those theories and replaced them with better ones. He would simply incorporate the improved geology into a new version of his surrealism. And neither the original theory nor its amended version tells us anything about how divine causality works.How are we to explain the fact that science has been empirically successful, that scientists have improved their naturalistic explanations of the phenomena? (Notice that what we seek to explain here is not a fact about the world, but rather a fact about our intellectual dealings with the world.) Suppose we have a theory that is completely successful, that saves all its phenomena, that is empirically adequate. How come? The obvious realist explanation is that the theory is true. The secular surrealist has no explanation or only a circular one—saying that the phenomena are as if the theory were true is just saying that the theory is empirically adequate. The theological surrealist avoids circularity by postulating miraculous divine intervention and saying that God fixes it that the phenomena are as if the theory were true. Which brings me to the so-called ‘Miracle Argument’ for realism. Hilary Putnam’s slogan was “Realism is the only philosophy that does not make the success of science a miracle”.
4