Lee Smolin and his Failing Model of the Universe
No scientific theory has been more thoroughly tested, examined, expanded, applied, and verified than evolutionary theory, the claim that all organic life arose from a common origin millions of years ago. The driving force behind that process is natural selection: over thousands of generations, organisms better suited to the environment eventually replace less-fit organisms. While the theory applies to biological life, physicist Lee Smolin has argued that the universe is subject to a cosmological natural selection (“Did the universe evolve?”). According to Smolin, his quasi-evolutionary theory has the power to explain why the laws of physics permit the existence of life. However, when one carefully explains and analyzes the theory, several difficulties arise.
Cosmology (the branch of science that deals with the origin, structure, and fate of the universe) has a problem: in sheer terms of probability, the laws of physics should not allow the evolution of life anywhere in the universe. When physicists spell out the laws of nature in the form of mathematical equations, they find certain constants that take on arbitrary values. For example, Newton’s law of gravity expressed as an equation looks like this:
In this equation, G stands for a constant, a number that physicists discover through experiments and put into the equation by hand. Nothing from the law itself determines G’s value, so as far as we know, it could have any value. Physical laws have all sorts of constants like these. In many cases, if the values of these constants were slightly different, life would not be possible in the universe. One can thus claim that the laws of physics are finely-tuned for our existence; “in the set of possible physics, the subset that permit the evolution of life is very small” (Barnes 3).[1] Take the force of gravity: if it were stronger, stars would have collapsed into black holes long before life could have evolved, and if it were weaker, a diffuse gas would fill the universe instead of stars, planets, and life. Moreover, Roger Penrose has calculated that were the amount of entropy or disorder in the early universe different by one part in 10^10^123rd power, then life could not exist (220). In sum, given our incredibly improbable existence, something needs to explain why the universe does allow the evolution of life. [2]
It is beyond the scope of this paper to outline and defend all instances of fine-tuning, but considering the recent summary of contemporary physics by Luke Barnes (“Fine-tuning”), we can take fine-tuning as given. The question is whether Smolin’s scenario succeeds in explaining our finely-tuned universe.
Smolin’s “unique idea represents the first attempt to incorporate the principle of natural selection into cosmology” (Eillis and Rotham 201). He builds his theory from two assumptions.
A key point Smolin stresses is that his theory is testable and therefore scientific. He argues that given his theory and the assumption that our universe is a typical or average member of the multiverse, the following prediction follows:
In theory, then, natural selection can explain the large-scale structure of the universe and our own improbable existence. However, Smolin’s theory has been on the table for twenty-five years, and the physics community has expressed several concerns about Smolin’s application of a biological theory to a non-biological field, some of more weight than others.
For example, cosmologist Alexander Vilenkin argues that “black hole production can be enhanced by an increase in the value of the cosmological constant” (2), thereby falsifying prediction (A) of Smolin’s theory. The basic argument is that if the one slightly increases the value of the cosmological constant, quantum fluctuations would increase. Quantum fluctuations sometimes produce either black holes directly or massive stars that eventually lead to black holes. Thus, increasing the cosmological constant leads to a universe more efficient at generating black holes. Given that Smolin has argued that our universe is maximally fine-tuned for black holes creation, his theory is false.
In a published response to Vilenkin, Smolin points out that the objection assumes that the cosmological constant is the most plausible solution to General relativity. General relativity needs to be modified in key respects because the observed accelerating expansion of the universe is incompatible with relativity in its current form. One of the suggested modifications of the theory is placing a cosmological constant in the equations of general relativity. But other options on the table include “quintessence, a ghost condensate…a role for higher dimensions, or of new non-local effects” (Smolin “Status of Cosmological Natural Selection” 13). If one opts for these alternative suggestions, Vilenkin’s counter-example fails. Smolin combats the objection well. A single counter-example, predicated on a controversial theoretical entity, does not falsify cosmological natural selection.
However, more serious concerns threaten Smolin’s project. Physicists George Ellis and T. Rothman have pointed out that Smolin’s proposal assumes that most black holes in a universe are the result of star collapse. It is this assumption that allows Smolin to claim that universes finely-tuned for black holes will be tuned for life as well. This is a false assumption. As Ellis and Rotham have pointed out, “from the point of view of tunneling to new universes, primordial black holes would serve just as well as stellar remnants” (209). Primordial black holes are those that form near the beginning of the universe because of dramatic variations in mass density; too much matter packed into one spot generates a black hole. Indeed, it seems that universes with primordial black holes would soon dominate the multiverse on Smolin’s scenario because universes like those can produce “offspring” much faster than universes that must go through the process of star formation and death. Over time, then, primordial black hole universes will weed out late-formation black hole universes, and the multiverse will become increasingly devoid of life. Unlike evolutionary theory in the biological realm, Smolin’s theory makes life in the universe unlikely and improbable. Therefore, while his theory may be a successful application of natural selection to the universe, that application has no explanatory power.
Instead of giving an empirical argument, philosopher Gordon McNabe points to a conceptual difficulty in Smolin’s theory: universes that evolve by natural selection are not typical or average members of the multiverse. It is likely that given a random choice of a universe from the total collection, that universe will not be one capable of evolving by natural selection. The basic reason is that there are more ways to be a universe without natural selection than there are to be one with it (paraphrased from McNade 7-8). For a system to evolve by natural selection, that system must have four features: reproduction, inheritance, variation, and various levels of survivability among offspring. However, as McNab points out, universes could lack any of these four features:
It is open to Smolin to object that McNab has misunderstood his scientific process: he is assuming that the only physically workable mechanism is one that produces universes via natural selection. Looking to possibilities (1) through (3), at least, Smolin can claim that although these scenarios are logically possible, they are not physically possible. Taking this as a presupposition of the theory, Smolin then asks the physics community to test his theory. As the theory receives confirmation, so do the theory’s assumptions.[4]
Now, it is right that a theory need not prove its assumptions directly, but it is not clear that Smolin’s theory presupposes that possibilities (1) through (3) are physically impossible. Smolin could tack them on as auxiliary presuppositions, but then his theory becomes unacceptably ad hoc or artificially contrived. Take the following two statements,
In sum, we are left with two outstanding difficulties for the application of evolutionary theory to cosmology: (i) the problem of primordial black holes, and (ii) either Smolin’s theory is ad hoc or it does not prove that life-permitting universes are probable. Smolin, at least, does not successfully apply natural selection to the universe. Someone else may have better luck, but for now, the physics community must wait for the next Darwin.
I. Physics and the Improbability of Life
F = G x M(1) x M(2)/(r x r)
In this equation, G stands for a constant, a number that physicists discover through experiments and put into the equation by hand. Nothing from the law itself determines G’s value, so as far as we know, it could have any value. Physical laws have all sorts of constants like these. In many cases, if the values of these constants were slightly different, life would not be possible in the universe. One can thus claim that the laws of physics are finely-tuned for our existence; “in the set of possible physics, the subset that permit the evolution of life is very small” (Barnes 3).[1] Take the force of gravity: if it were stronger, stars would have collapsed into black holes long before life could have evolved, and if it were weaker, a diffuse gas would fill the universe instead of stars, planets, and life. Moreover, Roger Penrose has calculated that were the amount of entropy or disorder in the early universe different by one part in 10^10^123rd power, then life could not exist (220). In sum, given our incredibly improbable existence, something needs to explain why the universe does allow the evolution of life. [2]
It is beyond the scope of this paper to outline and defend all instances of fine-tuning, but considering the recent summary of contemporary physics by Luke Barnes (“Fine-tuning”), we can take fine-tuning as given. The question is whether Smolin’s scenario succeeds in explaining our finely-tuned universe.
II. Smolin’s Proposal: Cosmological Natural Selection
Smolin’s “unique idea represents the first attempt to incorporate the principle of natural selection into cosmology” (Eillis and Rotham 201). He builds his theory from two assumptions.
- Black holes do not end in points where space and time come to a stop, but rather “the interior of a black hole tunnels into a new spatially compact universe” (Smolin “Did the Universe evolve?” 175). Smolin suggests that black holes suck in matter and energy and, like the mouth of a balloon, swell into another universe, eventually branching off. Universes with many black holes will have many progenies of baby-universes. We may refer collectively to these universes as the multiverse. In terms of evolution, black holes are sources of reproduction.
- Universes resulting from black-hole production are not exact copies of their parents; they differ slightly in the physical constants and quantities that inform the laws of nature. This process is a source of both inheritance and variability, inheritance because baby-universes have more in common with parent universes than not, but nonetheless variable to a small degree.
A key point Smolin stresses is that his theory is testable and therefore scientific. He argues that given his theory and the assumption that our universe is a typical or average member of the multiverse, the following prediction follows:
A. Changes in the laws of nature will result in a universe less efficient at generating black holes.The reasoning behind (A) is simple enough: Smolin’s theory postulates that most universes in the multiverse are efficient at generating black holes (due to cosmological natural selection). Indeed, most of them are incredibly efficient, so long as the multiverse has been around for enough time. Just as in biology, there will be maverick outliers, but if we assume that our universe is an average member of the collection, it follows that our universe will be maximally finely-tuned to produce black holes. Therefore, if one could prove that our universe is not maximally tuned for black hole creation, then one would falsify Smolin’s theory.[3]Of course, one could abandon the assumption of typicality, but that move makes Smolin’s theory unfalsifiable and untestable. In that case, it would no longer be a scientific theory.
III. Is the analogy to natural selection valid?
In theory, then, natural selection can explain the large-scale structure of the universe and our own improbable existence. However, Smolin’s theory has been on the table for twenty-five years, and the physics community has expressed several concerns about Smolin’s application of a biological theory to a non-biological field, some of more weight than others.
For example, cosmologist Alexander Vilenkin argues that “black hole production can be enhanced by an increase in the value of the cosmological constant” (2), thereby falsifying prediction (A) of Smolin’s theory. The basic argument is that if the one slightly increases the value of the cosmological constant, quantum fluctuations would increase. Quantum fluctuations sometimes produce either black holes directly or massive stars that eventually lead to black holes. Thus, increasing the cosmological constant leads to a universe more efficient at generating black holes. Given that Smolin has argued that our universe is maximally fine-tuned for black holes creation, his theory is false.
In a published response to Vilenkin, Smolin points out that the objection assumes that the cosmological constant is the most plausible solution to General relativity. General relativity needs to be modified in key respects because the observed accelerating expansion of the universe is incompatible with relativity in its current form. One of the suggested modifications of the theory is placing a cosmological constant in the equations of general relativity. But other options on the table include “quintessence, a ghost condensate…a role for higher dimensions, or of new non-local effects” (Smolin “Status of Cosmological Natural Selection” 13). If one opts for these alternative suggestions, Vilenkin’s counter-example fails. Smolin combats the objection well. A single counter-example, predicated on a controversial theoretical entity, does not falsify cosmological natural selection.
However, more serious concerns threaten Smolin’s project. Physicists George Ellis and T. Rothman have pointed out that Smolin’s proposal assumes that most black holes in a universe are the result of star collapse. It is this assumption that allows Smolin to claim that universes finely-tuned for black holes will be tuned for life as well. This is a false assumption. As Ellis and Rotham have pointed out, “from the point of view of tunneling to new universes, primordial black holes would serve just as well as stellar remnants” (209). Primordial black holes are those that form near the beginning of the universe because of dramatic variations in mass density; too much matter packed into one spot generates a black hole. Indeed, it seems that universes with primordial black holes would soon dominate the multiverse on Smolin’s scenario because universes like those can produce “offspring” much faster than universes that must go through the process of star formation and death. Over time, then, primordial black hole universes will weed out late-formation black hole universes, and the multiverse will become increasingly devoid of life. Unlike evolutionary theory in the biological realm, Smolin’s theory makes life in the universe unlikely and improbable. Therefore, while his theory may be a successful application of natural selection to the universe, that application has no explanatory power.
Instead of giving an empirical argument, philosopher Gordon McNabe points to a conceptual difficulty in Smolin’s theory: universes that evolve by natural selection are not typical or average members of the multiverse. It is likely that given a random choice of a universe from the total collection, that universe will not be one capable of evolving by natural selection. The basic reason is that there are more ways to be a universe without natural selection than there are to be one with it (paraphrased from McNade 7-8). For a system to evolve by natural selection, that system must have four features: reproduction, inheritance, variation, and various levels of survivability among offspring. However, as McNab points out, universes could lack any of these four features:
- A universe might give rise to a universe of a different ontology (different physical properties, laws, spatial dimensions, and so forth) – In this case, there is no meaningful chain of inheritance from parent to child.
- The values of the constants and quantities might not vary from parent to child universe – Here, there is no variation among members of the multiverse. Universes come and go, but the overall composition stays the same.
- There may be no connection between the traits that a child universe has compared to its parent – On this scenario, there is no cumulative change over thousands of generations, a key reason inheritance is so important for natural selection.
- Whole collections of universes may be those that expand forever, without black holes and without progeny – Universes like (4) do not reproduce.
It is open to Smolin to object that McNab has misunderstood his scientific process: he is assuming that the only physically workable mechanism is one that produces universes via natural selection. Looking to possibilities (1) through (3), at least, Smolin can claim that although these scenarios are logically possible, they are not physically possible. Taking this as a presupposition of the theory, Smolin then asks the physics community to test his theory. As the theory receives confirmation, so do the theory’s assumptions.[4]
Now, it is right that a theory need not prove its assumptions directly, but it is not clear that Smolin’s theory presupposes that possibilities (1) through (3) are physically impossible. Smolin could tack them on as auxiliary presuppositions, but then his theory becomes unacceptably ad hoc or artificially contrived. Take the following two statements,
B. Given a collection of universes evolving by natural selection, a life-permitting universe is probable.
C. Universes described by (1) through (3) above are physically impossible.If Smolin’s theory asserts nothing more than (B), then he can still derive all the predictions outlined above that make his theory testable. And yet, he has merely proven a conditional probability. In all likely-hood, given the above discussion, a life-permitting universe is still improbable. Now, if Smolin’s theory asserts (C) in addition to (B), no new predictions follow. The theory looks the same. Therefore, any experiment or verification of a prediction could not inform us if (C) is true. In short, the original objection stays: Smolin has not shown that the multiverse truly behaves such that natural selection applies to it (or at least to most of it). His central project fails.[5]
IV. Conclusion
Notes
[1] With Barnes, we can define physics as the laws of nature, the values of the constants featured in those laws, and the conditions at the beginning of the universe (the “initial conditions” that natures laws act upon).
[2] Some physicists have proposed the anthropic principle as the answer: we should not be surprised that we see a finely-tuned universe because that is the only universe we can observe. Were the universe not finely-tuned, we would not be here to observe it. Thus, there is a selection-effect that deceives us into thinking the physics of the universe are special. However, the anthropic principle at most proves a conditional statement: given that life exists, a finely-tuned universe is probable. Of course, such a statement is true. That does nothing, however, to show that a finely-tuned universe is probable, period.
[3] Moreover, Smolin has filled out this general prediction with a specific one: given his theory, there will not be neutron stars larger than one and half times the mass of the Sun. Were scientists to observe a neutron star three times the mass of the Sun, Smolin’s theory would be laid to rest.
[4] Moreover, option (4) does not prove detrimental to Smolin’s project because universes without progeny cannot multiply their numbers. Universes that do have offspring will, given enough time, eventually outnumber universes of type (4).
[5] There are other objections that one could raise. For example, Rudiger Vaas has claimed that Smolin’s theory is unfalsifiable because it assumes that our universe is a typical or average member of the multiverse – the so-called principle of mediocrity – but this principle cannot be proven (14). This objection is unpersuasive because the principle of mediocrity builds from an intuitively plausible principle of probability: all things being equal, assume that a thing x is probable rather than improbable. We hit conceptual bedrock, to be sure, but it is a solid bedrock.
Barnes, Luke. “The Fine-tuning of the Universe for Intelligent Life.” June 7th, 2012. https://arxiv.org/pdf/1112.4647.pdf.
Ellis, George, and T. Rothman. “Smolin’s Natural Selection Hypothesis.” Quarterly Journal for the Royal Astronomical Society. 1993, vol. 34, 201-212.
McCabe, Gordon. “A Critique of Cosmological Natural Selection” [Preprint]. October 7th, 2010. http://philsci-archive.pitt.edu/id/eprint/1648.
Penrose, Roger. The Emperor's New Mind: Concerning Computers, Minds, and the Laws of Physics.
Smolin, Lee. “Did the Universe Evolve?” Classical and Quantum Gravity, vol. 9, no. 1, Jan. 1992, pp. 173–191., doi:10.1088/0264-9381/9/1/016.
ibid. “The Status of Cosmological Natural Selection.” ArXiv.org, Cornell University Library, 18 Dec. 2006, arxiv.org/abs/hep-th/0612185.
Vaas, Rudiger. “Is there a Darwinian Evolution of the Cosmos?: Some Comments on Lee Smolin’s Theory of the Origin of the Universe by Means of Natural Selection.” ArXiv.org, Cornell University Library, conference proceedings September 2-5, 1994. https://arxiv.org/ftp/gr-qc/papers/0205/0205119.pdf.
Vilenkin, Alexander. “On Cosmic Natural Selection.” ArXiv.org, Cornell University Library, November 27th, 2006. https://arxiv.org/abs/hep-th/0610051.
[2] Some physicists have proposed the anthropic principle as the answer: we should not be surprised that we see a finely-tuned universe because that is the only universe we can observe. Were the universe not finely-tuned, we would not be here to observe it. Thus, there is a selection-effect that deceives us into thinking the physics of the universe are special. However, the anthropic principle at most proves a conditional statement: given that life exists, a finely-tuned universe is probable. Of course, such a statement is true. That does nothing, however, to show that a finely-tuned universe is probable, period.
[3] Moreover, Smolin has filled out this general prediction with a specific one: given his theory, there will not be neutron stars larger than one and half times the mass of the Sun. Were scientists to observe a neutron star three times the mass of the Sun, Smolin’s theory would be laid to rest.
[4] Moreover, option (4) does not prove detrimental to Smolin’s project because universes without progeny cannot multiply their numbers. Universes that do have offspring will, given enough time, eventually outnumber universes of type (4).
[5] There are other objections that one could raise. For example, Rudiger Vaas has claimed that Smolin’s theory is unfalsifiable because it assumes that our universe is a typical or average member of the multiverse – the so-called principle of mediocrity – but this principle cannot be proven (14). This objection is unpersuasive because the principle of mediocrity builds from an intuitively plausible principle of probability: all things being equal, assume that a thing x is probable rather than improbable. We hit conceptual bedrock, to be sure, but it is a solid bedrock.
Works Cited
Barnes, Luke. “The Fine-tuning of the Universe for Intelligent Life.” June 7th, 2012. https://arxiv.org/pdf/1112.4647.pdf.
Ellis, George, and T. Rothman. “Smolin’s Natural Selection Hypothesis.” Quarterly Journal for the Royal Astronomical Society. 1993, vol. 34, 201-212.
McCabe, Gordon. “A Critique of Cosmological Natural Selection” [Preprint]. October 7th, 2010. http://philsci-archive.pitt.edu/id/eprint/1648.
Penrose, Roger. The Emperor's New Mind: Concerning Computers, Minds, and the Laws of Physics.
Smolin, Lee. “Did the Universe Evolve?” Classical and Quantum Gravity, vol. 9, no. 1, Jan. 1992, pp. 173–191., doi:10.1088/0264-9381/9/1/016.
ibid. “The Status of Cosmological Natural Selection.” ArXiv.org, Cornell University Library, 18 Dec. 2006, arxiv.org/abs/hep-th/0612185.
Vaas, Rudiger. “Is there a Darwinian Evolution of the Cosmos?: Some Comments on Lee Smolin’s Theory of the Origin of the Universe by Means of Natural Selection.” ArXiv.org, Cornell University Library, conference proceedings September 2-5, 1994. https://arxiv.org/ftp/gr-qc/papers/0205/0205119.pdf.
Vilenkin, Alexander. “On Cosmic Natural Selection.” ArXiv.org, Cornell University Library, November 27th, 2006. https://arxiv.org/abs/hep-th/0610051.
Comments
Post a Comment