The Uniqueness of the Universe - Ole Miss



The Uniqueness of the Universe

The claim that the universe is fine-tuned for life has prompted considerable metaphysical speculation recently. Some theists see in it fodder for a new design argument. Some atheists respond that fine-tuning is explicable in terms of the existence of a vast multitude of other universes.[i] Yet there is a long history of philosophers seeing such metaphysical speculation as inherently misguided. The universe, they say, is unique. As such, it makes no sense to talk of the cause of it, the explanation for it, or the probability of its existing. What are these skeptics getting at? Are their objections any good? I will examine three versions of the uniqueness objection and conclude that only one raises a serious problem for efforts to explain fine-tuning metaphysically.

THE CAUSAL VERSION OF THE UNIQUENESS OBJECTION

The causal version of the uniqueness objection to design explanations of the universe is well-known to those familiar with Hume.

When two species of objects have always been observed to be conjoined together, I can infer, by custom, the existence of one wherever I see the existence of the other; and this I call an argument from experience. But how this argument can have place where the objects, as in the present case, are single, individual, without parallel or specific resemblance, may be difficult to explain. And will any man tell me with a serious countenance that an orderly universe must arise from some thought and art like the human because we have experience with it? To ascertain this reasoning it were requisite that we had experience with the origin of worlds.[ii]

As J.C.A. Gaskin astutely observes of the passage above, "this is one of the few occasions in his critique of religion when Hume appeals to a thesis drawn from his own philosophy."[iii] In this case, Hume invoked his (positive) theory of causation. According to it, there are three general conditions that must be satisfied in order for us to believe that A causes B: (i) we must believe that A is temporally prior to B; (ii) we must believe that A is spatially contiguous with B; and (iii) we must expect B to follow A. What makes us expect B to follow A, Hume says, is that we have observed a "constant conjunction" between the types of things of which A and B are instances; out of custom or habit (rather than the recognition of a necessary connection), we come to expect B-type things to follow A-type things.

In the passage above, Hume's critique of the design argument relies exclusively on condition (iii).[iv] For us to think A caused B, we must have (a) observed a (b) constant conjunction between (c) the types of things of which A and B are instances. Yet to meet (a) the observational requirement, "it were requisite that we had experience with the origin of worlds" - presumably an impossibility for creatures like us. Even supposing we could have experience with the origin of this world, we would have to have experience with the origin of other worlds as well in order to satisfy (b) the constant conjunction requirement. And in any case, the universe is "without parallel or specific resemblance." Since it is not a token of an oft-instantiated type, it fails to meet condition (c).

To those who think the universe is clearly similar to an artifact, Hume argued in parts V-VIII of Dialogues that it is equally similar to a plant or an animal. Since the universe no more resembles the members of the class of artifacts than it does the members of these other classes – since the universe lacks "specific resemblance" – there is no more reason to think the cause of the universe was artisan-like than to think that the cause was seed-like or mother-like. In other words, there is no reason to prefer the design hypothesis to the hypothesis that the universe was the result of some sort of "generation and destruction."

Other philosophers have enthusiastically endorsed Hume's objection, with Anthony Flew saying Hume's point about the uniqueness of the universe delivers to the design argument a "killing blow."

However far back we may be able to trace the – so to speak – internal history of the Universe, there can be no question of arguing that this or that external origin is either probable or improbable. We do not have, and we necessarily could not have, experience of other Universes to show or suggest that Universes, or Universes with these particular features, are the work of Gods, or of Gods of this or that particular sort...[…]...the unique Universe is and must be itself all we have. How it is, is just how it is; and that's that.[v]

Notice Flew says "all we have" (by which he means, I take it, "all we have to go on") rather than "all there is." Had he said the latter, then it would follow analytically that there is only one universe. Contemporary physical cosmologists do not use 'universe' this way; they often speak of other, real universes, especially in connection with the anthropic principle. It would not affect Hume's point, however, if there were more than one universe that could exist. His complaint rests simply on the idea that we can only have experience of one universe.[vi]

Whether this causal version of the uniqueness objection works depends crucially on whether the universe is, indeed, "without parallel or specific resemblance." If not – if, in modern parlance, it can be placed in an appropriate "reference class" – then, contrary to Hume, we could observe a constant conjunction between the type of thing of which the universe is an instance and a type of cause. Yet if the universe does lack "specific resemblance," this is only a contingent fact. In his day Hume may have been warranted in insisting that the evidence did not justify selecting machine over animal or vegetable as that to which the universe bears specific resemblance. As our picture of the universe fills out, however, the evidence may tilt in favor of some specific analogue.

For example, recent mappings of the galaxies reveal the existence of "foamy" structures.[vii] Galaxies are stretched out in stringy filaments, arranged in huge bubbles, and clustered into gigantic walls with galaxy-free space in between. Some astronomers interpreting these maps take them as pointing to phase transitions in the early universe. Part of their reason is that we see the same sorts of features in other substances produced as a result of phase transitions (e.g. the cracks, streaks, and bubbles that develop as liquid water turns to solid ice).[viii] Of course it may turn out that a fuller picture of the universe leads cosmologists to abandon these analogies – but then again, it may not. The point is that there is no general reason for supposing that the only sort of thing to which a universe might properly be compared is another universe. Whether or not the universe bears a specific resemblance to some familiar kind of object is a thoroughly empirical matter.

THE EXPLANATORY VERSION OF THE UNIQUENESS OBJECTION

Under the influence of a deductive-nomological (DN) picture of scientific explanation, some philosophers think those who try to ascribe a cause to the universe are guilty of a category mistake. They say cosmologists are capable of offering us only a model of the universe, not a theory. W.H. McCrea illustrated the distinction between models and theories by contrasting ancient astronomy with modern. For the Greeks, the goal of astronomy was to develop a model that accurately described the motions of the planets and enabled future motions to be predicted. Newton, on the other hand, looked for a theory from which he could deduce, not just the motions of the planets, but the motions of any other physical system as well. The Newtonian approach is the ideal contemporary cosmologists follow as they study the universe, said McCrea, but they would be better off approaching the universe in the same way that the Greeks approached astronomy.

Now we maintain that Newton did better than the Greeks because his methods allowed him to deal with any number of gravitating systems as easily as with the solar system, whereas the Greek method entailed a repetition of the whole work for every system encountered. On the other hand, there is but one universe and so we should regard the Greek method here as the superior one. Of course, we expect some time to find a better method but not in this case one along the lines of Newtonian theory.[ix]

McCrea thought the only reason to theorize about the cause of the universe would be to uncover laws that would explain the existence and nature of other universes as well. Since there are no other universes (or, at least, none open to our observation), the search for causal laws is unnecessary. Milton Munitz adopted a similar position.

Insofar as inductive reasoning is the attempt to provide good reasons for inferring something about unobserved instances, on the basis of observed instances, it has to do with the inference to further instances of some regularity or law. However, the problem faced in cosmology is not that of finding a warrant in experience for establishing laws. This is the task of ordinary physics. The task of cosmology, rather, is to say something about [that to which the observed universe belongs]; for this purpose, a model of the universe needs to be employed. And a model is not a law.[x]

John Leslie says such people think "basic laws of nature cannot be explained and therefore cannot need to be explained."[xi] As he points out, this belief arises from a conception of explanation according to which "the right way of explaining any natural law is to reveal it as an instance of some more basic law."[xii] Assuming 'more basic' means 'applies to a greater range of phenomena', the most basic laws will be those which apply to all phenomena and hence are incapable of subsumption. The attitude toward these basic laws is (to paraphrase Flew) that they are what they are, and that's that.

This objection is compelling only if one thinks DN is the only correct model of scientific explanation. Yet clearly there are singular explanations – explanations in which the phenomena explained are not deducible from some conjunction of laws and initial conditions – that count as fully scientific. Such explanations are common in "historical" sciences such as archaeology and evolutionary biology. Furthermore, subsuming phenomena under a law is not the same thing as providing an explanation – at least if an explanation is supposed to increase our understanding.

Gregor Mendel identified a pattern regarding the heritability of pea-pod characteristics, yet arguably he did not explain pea-pod heredity. For one thing, Mendel's Law is silent about the mechanism of heritability. Providing a mechanism – answering the “How?” question – is an important aspect of scientific explanation. Primitive human seafarers were able to uncover the pattern behind the tides, but they had no explanation of the tides because they did not understand why or how they happened. A view of explanation according to which subsumption under a law automatically counts as providing an explanation thus appears to be entirely too easygoing. Non-DN explanations, on the other hand, are often the best examples of science leading to an increased understanding of the world. Consider, for example, the theory that the dinosaurs became extinct as the result of a massive asteroid impact 65 million years ago. Whatever a scientific explanation is, the asteroid impact theory should count as one. As with the long-ago extinction event, the origin of the universe, too, is a unique happening, but that fact alone does not rule out the possibility of explaining it.[xiii]

THE PROBABILISTIC VERSION OF THE UNIQUENESS OBJECTION

Contemporary arguments from fine-tuning are almost always couched in probabilistic terms, with the inference to a designer or to a multiverse cast as a Bayesian one. The central claim is that the probability on chance of the universe's being such as to permit life is extraordinarily low. Yet some allege that the uniqueness of the universe precludes meaningful talk of the probability of the universe's existing and being the way it is. If this probabilistic version of the uniqueness objection is right, then inferences from fine-tuning to a designer or to a multiverse will either have to be reformulated in a way that does not involve probabilities or will have to be abandoned altogether.

The doctrine of the meaninglessness of probability claims about unique events derives from frequentism. Richard von Mises, the 20th century's foremost proponent of frequentism, claimed that "the rational concept of probability, which is the only basis of probability calculus, applies only to problems in which either the same event repeats itself again and again, or a great number of uniform elements are involved at the same time."[xiv] A consequence of this view is that (as Peirce said) "in reference to a single case considered in itself, probability can have no meaning."[xv] If the universe is such a single case, it follows from frequentism that it is meaningless to talk about the probability of the universe's existing and having the features that it has. As previously noted, the universe may not be such a case. Perhaps there is some reference class (or "collective," to use von Mises's term) of observable phenomena into which the universe can naturally be put – say, the class of foamy physical systems. But what if such a reference class of observable phenomena cannot be identified?

A tempting view to take – indeed, the standard view in the fine-tuning literature – is that the reference class is the set of possible universes that follow the same basic laws as our universe but in which the values of the free parameters (e.g. the rate of expansion of the early universe) are different. These possible universes are unobservable, but knowledgeable physicists can grasp what they would be like (e.g. how conducive they would be to star formation).

This view suffers from a fundamental problem. Consider a particular free cosmic parameter P. Though there may be true counterfactuals of the form "If P were slightly different, then the eventual emergence of life in the universe would be physically impossible," this would not entail that P's taking a life-permitting value is improbable. A probability is a number between zero and one. For there to be a probability of a particular outcome there has to be a space of possible outcomes such that its area or volume adds up to one. That is, probabilities require normalizable possibility spaces. In an infinite space in which every value is equally probable, either each particular probability is zero (in which case they all add up to zero) or each particular probability is nonzero (in which case their sum is infinite).[1] The issue, then, is whether a basis has been provided (either in physical theory or on philosophical grounds) for restricting the possible values of the relevant cosmic parameter or for assigning higher probabilities to some of its possible values than others.[xvi] Without such a basis, there is no reason to think the space of possible values for that parameter is normalizable. And in that case, it is meaningless to talk about the probability of selecting a certain value (or one of a certain set of values) from that space.[xvii] Because of this normalizability issue, we cannot cash out claims about the low probability of the universe's being such as to permit life in terms of the relative frequency of life-permitting universes in some set of possible universes.

Things get no better if we work with one of the other standard philosophical accounts of probability.[xviii] Subjectivists analyze probability in terms of rational constraints on betting behavior. The problem, however, is that no sense can be made of the idea of our collecting on a bet regarding the universe’s having a given set of constants. [Here we pick up one of the main themes of the uniqueness objector. We are not angels, capable of peeking in on various universes. We are stuck inside this universe.] Propensity theory, meanwhile, is inapt because there seems to be nothing in which any propensity for the production of our universe could be grounded – unless we beg the question by assuming that there is some concrete reality beyond our universe in which such a propensity is grounded. That there is such a reality is supposed to be the conclusion of a design/multiverse argument from fine-tuning, not a premise.

To sum up, the causal and explanatory versions of the uniqueness objection are weak, whereas the probabilistic version is potentially quite strong. Whether there is an adequate reply to the latter is a matter for further discussion.[xix]

-----------------------

-----------------------

[i] For careful documentation of the claims of fine-tuning - claims that if various force strengths, mass ratios, and initial conditions had been slightly different, life could not possibly have arisen in the universe – see Robin Collins, “Evidence of Fine-tuning,” in Neil A. Manson (ed.), God and Design: The Teleological Argument and Modern Science, (London: Routledge, 2003). See also John D. Barrow and Frank J. Tipler, The Anthropic Cosmological Principle (New York: Oxford University Press, 1986) and John Leslie, Universes (New York: Routledge, 1989).

[ii] David Hume, Dialogues concerning Natural Religion, ed. Nelson Pike (Indianapolis: Bobbs-Merrill Educational Publishing, 1970) p. 30.

[iii] J.C.A. Gaskin, Hume's Philosophy of Religion (New York: Barnes and Noble, 1978), p. 20.

[iv] Interestingly, nowhere in Dialogues does Hume appeal to condition (ii) of his account of causation when criticizing the design argument. If the universe (or "world") is simply the totality of all physical things, then there could be no further physical thing spatially contiguous with it. Perhaps this absence indicates Hume did not define 'world' in this way. Peter van Inwagen pointed out (correspondence) evidence in favor of this interpretation. Hume discussed the possibility of a world's having arisen by "vegetation" from a comet-seed (see Hume, Dialogues, p. 62); this implies that a Humean "world" could be part of some larger universe that is open to experience.

[v] Anthony Flew, David Hume: Philosopher of Moral Science (London: Basil Blackwell, 1986), p. 66.

[vi] To say that we can only have experience of one universe is not to say the universe we inhabit is the only one of which we could have experience. It is logically possible that we exist in some other universe instead of this one, just as it is logically possible that Descartes was born in Scotland rather than France. The question of whether (due to the necessity of material origins) we are metaphysically tied to this particular universe turns out to be a sticking point with respect to the objection that multiverse explanations of fine-tuning fail to explain why this universe is fine-tuned. See Roger White, “Fine Tuning and Multiple Universes,” Nous vol. 34 (2000), pp. 260-76; and Neil A. Manson and Michael J. Thrush, "Fine-tuning, Multiple Universes, and the 'This Universe' Objection," Pacific Philosophical Quarterly vol. 84 (2003), pp. 67-83.

[vii] See Sidney Perkowitz, Universal Foam: From Cappuccino to the Cosmos (New York: Walker & Co., 2000) for an entertaining introduction to the physics of foam.

[viii] Ivars Peterson reports on these observations in "Seeding the Universe," Science News vol. 137 (March 24, 1990), pp. 184-7.

[ix] W.H. McCrea, "Information and Prediction in Cosmology," The Monist vol. 47, no. 1 (Fall 1962), p. 98.

[x] Milton K. Munitz, The Mystery of Existence (New York: New York University Press, 1974), p. 62.

[xi] John Leslie, Universes (New York: Routledge, 1989), p. 106.

[xii] ibid., p. 109.

[xiii] I owe this criticism of the DN account of explanation to Steven Meyer. See his “The Methodological Equivalence of Design and Descent,” in The Creation Hypothesis, ed. J.P. Moreland (Downers Grove, IL: InterVarsity Press, 1994), pp. 77-82.

[xiv] Richard von Mises, Probability, Statistics, and Truth, 2nd English edition (revised) (London: George Allen and Unwin, 1957), p. 11.

[xv] C.S. Peirce, "The Doctrine of Chances," in Writings of Charles S. Peirce: A Chronological Edition, vol. 3 (Bloomington, Indiana: Indiana University Press, 1982), p. 281. "The Doctrine of Chances" was published in 1878.

[xvi] Some attempts have been made to provide such a basis, but they seem to have failed. See John Earman and Jesus Mosterin, "A Critical Look at Inflationary Cosmology," Philosophy of Science vol. 66 (1999), pp. 31-4. Robin Collins, Alexander Pruss, and Jeffrey Koperski separately seek to rectify this problem in unpublished papers of theirs.

[xvii] For a much more detailed presentation of this argument, see Timothy McGrew, Lydia McGrew, and Eric Vestrup, "Probabilities and the Fine-tuning Argument: A Skeptical View," Mind vol. 110, no. 440 and Neil A. Manson, "There Is No Adequate Definition of 'Fine-tuned for Life'," Inquiry vol. 43, no. 3 (September 2000). A less formalized presentation of the same basic idea was given by David Shotwell in "Is the Universe Improbable?" The Skeptical Inquirer vol. 11 (1987), pp. 376-82.

[xviii] For an excellent survey of such accounts, see Donald Gillies, Philosophical Theories of Probability (New York: Routledge, 2000).

[xix] I’d like to thank the participants in the 2002 meeting of the Creighton Club for their valuable responses to an earlier version of this paper.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download