The Failure of Swinburne’s Simplicity Recipe for Verisimilitudinous Theories
In his books Simplicity as Evidence of Truth (1997) and Epistemic Justification (2001), Richard Swinburne argued strenuously that simplicity provides probabilistic evidence of truth by being a tie-breaker among conflicting theories as follows: greater simplicity is a criterion for “choosing among [competing] scientific theories of equal scope [or content] fitting equally well with background evidence and yielding the same data. . . .” This choice of the simpler of two theories purportedly yields the theory that is more likely to be true precisely by virtue of being simpler.
However, in a lecture that I delivered in March 2006 at All Souls College in Oxford when Swinburne was in the audience, and also at the Royal Institute of Philosophy in London, I demonstrated at least two results that fundamentally subvert Swinburne’s thesis, as he, in effect, conceded in the ensuing public discussion. To state my results, let me speak of a theory B that is more likely to be true than a theory A as “having greater verisimilitude” than A. Then two of my damaging contentions against Swinburne can be stated as follows:
- His comparative simplicity ratings, which are to yield a verdict of greater verisimilitude, avowedly pertain to rival hypotheses of equal content or scope. Yet Swinburne, like Karl Popper before him, has left the implementation of this crucial content-parity requirement glaringly unfulfilled.
- In The Existence of God, Swinburne wrote: “. . . if there is to exist something, it seems impossible to conceive of anything simpler (and therefore a priori more probable) than the existence of God.” Furthermore, he told us that “. . . the [explanatory] choice is between the universe as [explanatory] stopping point [i.e., as existing qua brute fact] and God as [explanatory] stopping point [i.e., existing as a matter of brute fact].”
Thus, the God hypothesis is supposedly the conceptually and ontologically simpler option. And the avowed cardinal thesis of Swinburne’s 2004 book, The Existence of God, is “an argument for God being the cause [ex nihilo] of the existence of the universe.”
Just for argument’s sake, posit with Swinburne that in the class of all existential hypotheses, the theistic one is the simplest and therefore simpler than its atheistic competitor. It is then a patent corollary of the conclusion of the argument I presented in my 2008 paper, “Is Simplicity Evidence of Truth?,” that this supposed greater simplicity does not show theism to be inductively more likely to be true than atheism, absent some demonstration of the requisite content-equality of these two rival hypotheses.
Indeed, even if such content-equality were to be demonstrated, the supposed greater a priori simplicity of the God hypothesis would not confer any greater verisimilitude upon it: as I have argued elsewhere, a priori simplicity and a priori probability are not at all ontologically legislative, and thus Swinburne cannot milk any theological capital out of the purported a priori simplicity and a priori probability of the Deity, even if true! Yet his theism is explanatorily omnivorous, avowing very dubiously that theism explains “everything we observe.”
Moreover, there are further difficulties: theory B might be simpler than theory A in one respect while being more complicated in another. But inter-theory comparisons of simplicity for assessing relative verisimilitude call for criteria of greater overall simplicity. Yet, previously I have used the comparison of Einstein’s general theory of relativity with Newton’s theory of gravitation to impugn the feasibility of ratings of comparative overall simplicity for rival theories. However, differing verdicts from just such ratings are indispensable to both Swinburne’s prescription for the greater verisimilitude of one of these theories and to philosopher J. J. C. Smart’s aspiration of simplicity as an avenue to truth, an aspiration set forth in my Part 1 (FI, June/July 2008, p. 35).
Indeed, if Einstein’s theory of gravitation were held to be more complex overall than Newton’s in virtue of the nonlinearity of its partial differential field equations, then its presumed greater verisimilitude is the death knell of both Swinburne’s prescription and Smart’s aspiration.
It has been said that scientists think they know when one theory is simpler than another overall as a matter of its greater beauty and elegance. But beauty is in the eye of the beholder, and, as Einstein aptly remarked, elegance had best be left to tailors.
The Demise of Leibniz’s 1714 Justification for the Primoridal Existential Question (PEQ)
Now, let us come to grips with the specific, 1714 context in which Leibniz formulated his PEQ and tried to justify it at once by relying carefully on both of the following two premises: (1) His well-known Principle of Sufficient Reason, to which I shall refer by the acronym PSR, and (2) his a priori argument from simplicity for the presupposition SoN, which is inherent in his PEQ, an argument I articulated in Part 1 (pp. 34–35).
Thus, Leibniz declared: “the great principle [of sufficient reason, PSR] . . . holds that nothing takes place without sufficient reason, that is . . . a reason sufficient to determine why it is thus and not otherwise. This principle having been laid down, the first question we are entitled to ask will be: Why is there something rather than nothing? For ‘nothing’ [i.e., the Null World] is simpler and easier than ‘something.’ Further supposing that things must exist, it must be possible to give a reason why they must exist just as they do and not otherwise” [italics in original].
These very ambitious avowals by Leibniz invite my clarifying comments:
- Right after enunciating his PSR, he poses PEQ, “Why is there something rather than nothing?,” as “the first question we are entitled to ask.” However, immediately after raising that question, he relies on the supposed simplicity of the Null World to justify the presupposition SoN of PEQ, claiming, in effect, that the Null World would be spontaneously realized ontologically in the absence of an overriding external cause. As we recall, he puts it concisely: “For ‘nothing’ [i.e., the Null World] is simpler and easier than ‘something.’” And clearly, there is either something or nothing.
- Evidently, Leibniz is not content to rely on his PSR alone to ask the truncated question “Why is there something contingent?” without the accompanying contrasting clause “rather than nothing.” Instead, he uses SoN as presupposed in this contrasting clause to assert a dual thesis: (a) the existence of something contingent is not to be expected at all, and (b) its actual existence therefore cries out for explanation! As will be recalled, just this dual thesis was philosopher Derek Parfit’s implicit rationale for embracing Leibniz’s PEQ.
Thus, the soundness of Leibniz’s justification of his PEQ evidently turns on the cogency of his PSR as well as of his a priori argument from simplicity for SoN. But we have already discounted his a priori argument for SoN in Part 1 (p. 35). Thus, we can now concentrate on appraising his PSR.
Consider the grounds in twentieth-century quantum theory for the demise within our universe of the universal causation familiar from Newton’s physics as codified by Laplace’s “determinism.” This empirically well-founded quantum theory features merely probabilistic, rather than universal, causal laws governing such phenomena as the spontaneous radioactive disintegration of atomic nuclei, yielding emissions of alpha or beta particles and/or gamma rays.
In this domain of phenomena, there are physically possible particular events that could but do not actually occur at given times under specified initial conditions. Yet it is impermissibly legislative to insist, via Leibniz’s PSR, that merely because these unrealized events are thus physically possible, there must be an explanation entailing their specific nonoccurrence and, similarly, a deductive explanation of probabilistically governed actually occurring events.
This admonition against PSR was not heeded by Swinburne, who avowed entitlement to universal-explainability, declaring in The Existence of God: “We expect all things to have explanations.” In just this vein, we recall, Leibniz had demanded for every event an explanatory “reason [cause] sufficient to determine why it is thus and not otherwise.” Hence, the history of modern quantum physics teaches that PSR, which Leibniz avowedly saw as metaphysical, cannot be warranted a priori and indeed is untenable on ultimately empirical grounds.
Thus, to discover that the universe does not accommodate rigid prescriptions for deterministic explanatory understanding is not tantamount to scientific failure; instead, it is to discover positive reasons for identifying certain coveted explanations as phantom.
As we saw, Leibniz had generated PEQ by conjoining his untenable PSR with SoN. Yet since his a priori defense of SoN via simplicity has also failed, it remains to inquire whether his avowed ontological spontaneity of the Null World might possibly be warranted empirically. My answer will be emphatically negative for the following reason: it turns out, as an induction from various episodes in the history of science, that SoN is altogether ill-founded empirically, as we are about to see.
To examine the empirical status of SoN, it will be useful to reformulate it in Richard Swinburne’s aforecited words as follows: “Surely the most natural state of affairs is simply nothing: no universe, no God, nothing.” But since our empirical evidence comes, of course, from our own universe U, consider the corollary of SoN that pertains to U. This corollary asserts that it is natural or spontaneous for U not to exist rather than to exist. Contrary to any a priori dictum on what is the “natural” ontological behavior of U, the verdict on that behavior will now be seen to depend crucially on empirical evidence, and indeed to provide no support for SoN.
Two specific cosmological examples spell out this empirical moral, leaving aside some pertinent technicalities that I do not have space to develop here. Yet the point of my examples will be clear.
- The first example is furnished by the natural evolution of one of the Big Bang models of the universe countenanced by general relativistic cosmology, the dust-filled, so-called Friedmann universe described by R.M. Wald. This universe has the following features relevant to my concerns:
- It is a spatially closed, three-dimensional spherical universe (a “3-sphere”), which expands from a pointlike Big Bang to a maximum finite size and then contracts into a pointlike crunch.
- That universe exists altogether for only a finite span of time, such that no instants of time existed prior to its finite duration or exist afterward.
- As a matter of natural law, its total rest-mass is conserved for the entire time period of its existence. Thus, during that entire time, there is no need for a supernatural agency to generate that mass out of nothing and/or to prevent it from lapsing into nothingness, contrary to both SoN and to Thomas Aquinas and René Descartes.
Evidently, the “natural” dynamical evolution of the Friedmann Big Bang universe as a whole is specified by Einstein’s empirically supported cosmology. Thus, the “natural” or spontaneous ontological behavior of Big Bang worlds is not vouchsafed a priori.
- The same epistemic moral concerning the empirical status of cosmological naturalness is spelled out by the illuminating case of the now largely defunct Bondi and Gold so-called steady-state cosmology of 1948.
Assuming the mutual recession of the galaxies discovered by Hubble, the average density of matter ought to decrease with time. But in their 1948 theory, Bondi and Gold boldly postulated that, nevertheless, in a spatially and temporally infinite universe, as a matter of natural law there is large-scale, temporal constancy of the matter-density. Note that this conservation or constancy is not of matter but of the density of matter over time.
The conjunction of this constancy of the density with Hubble’s mutual recession of the galaxies from one another then entails the following quite counterintuitive consequence: throughout space-time, and without any matter-generating agency, new matter (in the form of hydrogen) literally pops into existence, completely naturally, in violation of matter-energy conservation.
Hence, the Bondi and Gold world features the accretion or formation of new matter as its natural, normal, spontaneous behavior, yet terrestrially at a very slow rate. And although this accretive formation is indeed out of nothing, its complete spontaneity clearly precludes its being “created” by an external agency. Apparently, if the steady-state world were actual, it would discredit the doctrine of the medieval Latin epigram “Ex nihilo, nihil fit,” which means “from nothing, you cannot get anything,” or more familiarly, “you cannot get blood out of a stone.” But Bondi and Gold could.
The steady-state theory owes its demise to the failure of its predictions and retrodictions to pass observational muster in its competition with the Big Bang cosmology. This episode again teaches us that empirically based scientific theories are our sole epistemic avenue to the “natural” behavior of the universe at large, though of course only fallibly so.
What then is the empirical cosmological verdict on the corollary of SoN that asserts that “It is natural for our universe not to exist, rather than to exist”? Apparently, there is no empirical evidence for this corollary from cosmology, let alone for SoN itself. Its proponents surely have not even tried to offer any such evidence for SoN, believing mistakenly, as we saw, that it can be vouchsafed a priori á la Leibniz.
PEQ as a Failed Springboard for Creationist Theism: The Collapse of Leibniz’s and Swinburne’s Theistic Cosmological Arguments
Probably any one of us who was reared in the Occident has wondered at some time: “Where did everything come from?” As we know, typically this question is not a demand for a statement of the earlier physical history of our existing universe. Instead, the question is driven by the largely unconscious assumption of SoN and is thus simply another version of Leibniz’s query PEQ. Yet as I have argued painstakingly, PEQ rests on the ill-founded premise SoN, as well as on Leibniz’s very questionable PSR. Therefore, PEQ is an ill-conceived nonstarter that poses a pseudo-issue and does not warrant the quest for any sort of first cause as an answer!
But, as we know, both Leibniz and Swinburne raised PEQ as an imperative question, and thence they concluded misguidedly that the answer to it mandates divine creation. Indeed, in the May 26, 2007, issue of the German magazine Der Spiegel, journalist Alexander Smoltczyk tells us incoherently: “For most believers, God is neither a person nor a principle, nor yet an existing entity, but rather an answer to the question why there is something rather than nothing” [my English translation]. Thus, PEQ is invoked to generate incoherent theological capital in the culture at large: how, one needs to ask, can God not be an “existing entity” and yet be “an answer” to PEQ? Such is the desperation to give God a job to do.
However, PEQ evidently cannot serve as a viable springboard for creationist theism, because it is a pseudo-issue based on quicksand! By the same token, Leibniz’s and Swinburne’s cosmological arguments for divine creation are fundamentally unsuccessful.
Hence, I say to you: whatever philosophical problems you may have, it is my plea here that answering Leibniz’s PEQ should not engage your curiosity, because his question is just a will-o’-the wisp.
Coda on Richard Dawkins and Sam Harris vis-á-vis PEQ
Despite all the telling arguments for atheism offered by both Richard Dawkins in The God Delusion and Sam Harris in The End of Faith and Letter to a Christian Nation, they both caved in altogether unwarrantedly by countenancing PEQ, very misguidedly, as a searching question that rightly calls for an explanatory answer.
Thus Dawkins wrote: “Time and again, my theologian friends returned to the point that there had to be a reason why there is something rather than nothing. There must have been a first cause of everything, and we might as well give it the name God.”
Alas, entirely unaware that PEQ is an abortive nonstarter posing a pseudo-issue, Dawkins replied unavailingly by actually allowing a “first cause, the great unknown which is responsible for something existing rather than nothing” but objecting most feebly that this notion of a first cause as “a being capable of designing the universe and of talking to a million people simultaneously, is a total abdication of the responsibility to find an explanation.” However, this purported explanatory responsibility is a mere phantom, an ill-conceived contrivance of the ill-fated PEQ, as I showed in great detail in my 2004 “The Poverty of Theistic Cosmology.”
Relatedly, Sam Harris embraces PEQ to his detriment, declaring in Letter to a Christian Nation with very misplaced intellectual humility: “Any intellectually honest person will admit that he does not know why the universe exists. Scientists, of course, readily admit their ignorance on this point” [italics in original]. But it is a direct corollary of my argument above that there is no ignorance at issue here of the kind that is in need of being admitted!
To boot, Harris writes: “The truth is that no one knows how or why the universe came into being. It is not clear that we can even speak coherently about the creation of the universe, given that such an event can be conceived only with reference to time, and here we are talking about the birth of space-time itself.” At this point Harris offers a footnote pointing out that the physicist Stephen Hawking “pictures space-time as a four-dimensional, closed manifold without beginning or end. . . .” But, in the very next sentence, Harris misguidedly capitulates to PEQ by making the aforecited ill-fated categorical declaration: “Any intellectually honest person will admit that he does not know why the universe exists.”
Clearly, Harris applies this unqualified admission of ignorance to every instant at which the universe exists, even if it has existed forever. And evidently, he thinks that even Hawking’s model of a universe whose time is unbounded though finite (like a circle) cannot purchase absolution from this challenge by PEQ, although that model undercuts the particular question of “how or why the universe came into being” at a moment in the finite past. A fortiori, Harris again implicitly capitulates to PEQ in the case of other models of the universe, featuring other time-structures. Interested readers may consult my anticreationist account of the Big Bang universe in the essay “Theological Misinterpretations of Current Physical Cosmology.”
- Bondi, H. Cosmology (2nd ed.). Cambridge: Cambridge University Press, 1960.
- Dawkins, R. The God Delusion. New York: Houghton Mifflin, 2006.
- Grünbaum, A. “Can a Theory Answer More Questions than One of Its Rivals?” The British Journal for the Philosophy of Science 27(1)1976.
- ———. “Theological Misinterpretations of Current Physical Cosmology.” Philo 1(1), 1998.
- ———. “The Poverty of Theistic Cosmology.” The British Journal for the Philosophy of Science 55(4), 2004.
- ———. “Is Simplicity Evidence of Truth?” American Philosophical Quarterly 45(2), 2008.
- Harris, S. Letter to a Christian Nation. New York: A.A. Knopf, 2006.
- Leibniz, G.W. “Principles of Nature and of Grace Founded on Reason” Trans. G.H.R. Parkinson and M. Morris. In Leibniz: Philosophical Writings, edited by G.H.R. Parkinson. London: J.M. Dent & Sons, 1973.
- Smoltczyk, A. Der Kreuzzug der Gottlosen (The Crusade of the Godless). Der Spiegel 56–69, May 26, 2007.
- Swinburne, R. The Existence of God. Rev. ed. Oxford/New York: Clarendon Press/Oxford University Press, 1991.
- ———. Is There a God? Oxford/New York: Oxford University Press, 1996.
- ———. Simplicity as Evidence of Truth. Milwaukee, Wisconsin: Marquette University Press, 1997.
- ———. Epistemic Justification. Oxford: Clarendon Press, 2001.
- ———. The Existence of God. 2nd ed. Oxford: Clarendon Press, 2004.
- Wald, R. M. General Relativity. Chicago: University of Chicago Press, 1984.