Ultimate Explanations of the Universe (11 page)

The concept of an infinitely large number of universes, also referred to as the concept of the multiverse, grew out of the discussion concerning the anthropic principles. However, soon it started to appear in other contexts, and even developed a quantitative approach. This was true of the chaotic cosmology proposed by Linde, who constructed an inflationary cosmological model that led in a natural way to the “continual production of universes” (see Chap. 6). Although in his declarations he has often invoked philosophical motives (he wanted to secure “an eternal existence” for the family of universes), the chaotic inflation model was not prompted solely by philosophical inspiration.

One of the most radical and controversial concepts of the multiverse is the idea Lee Smolin presents in his book
The Life of the Cosmos
.
¹
Smolin introduces his concept as a falsifiable cosmological model, but in the heat of discussion and his impassioned journalistic approach he has not shied away from displaying his preferred worldview.

Smolin’s concept offers an excellent opportunity for a review of some of the typical problems besetting the idea of the multiverse as presented in a range of versions. This chapter will look at them.

The general picture of the universe proposed by Smolin is basically no different from the picture presented in Linde’s chaotic cosmology. In both models a parent universe produces descendant worlds in which the physics differs from that in the parent world. However, the mechanism by which universes are produced differs in the two models. In Linde’s model inflation and quantum fluctuations are responsible for the generation of universes, whereas Smolin’s version is based on two assumptions and an idiosyncratic understanding of the selection principle.

His first assumption concerns the problem of the singularities. Smolin assumes that the quantum effects of gravitation “prevent the formation of singularities, at which time starts or stops.”
²
Hence, when a collapsing object – a universe or a massive star – reaches its critical density, its shrinking transforms into expansion (following a “bounce”) and the whole process gives rise to a new universe. The Big Bang, too, according to Smolin, might have been the result of the collapse of another object in another world. Note that this is a very big assumption, which should be a programme for the construction of a cosmological model rather than an assumption.

For of course, on the grounds of the selection principle, the event with the highest probability is that the universe in which we live belongs to that most numerous subset in the multiverse which is most prolific of black holes. Hence the set of parameters characteristic of the physics of our universe should be the most favourable for the production of black holes. We don’t know whether this is indeed so, but that’s what Smolin has forecast.

There is yet another assumption that plays a significant role in Smolin’s line of reasoning: that the same set of initial conditions, values for the physical constants and other parameters characteristic of our universe favouring the generation of black holes also enables the onset of biological evolution and its continuation up to the appearance of conscious creatures. This assumption is completely independent of the former ones, but without it Smolin’s entire concept would founder. Thanks to it we obtain an explanation why the world we live in is as it is and not any otherwise. But isn’t this an explanation that begs the question? And it is certainly not an “ultimate” explanation. The problem of justifying the multiverse itself has not been touched on. In particular, according to Smolin, the laws of physics are the same in every universe and the question of where they have come from is still waiting to be answered. However, Smolin hopes that perhaps one day it will be possible to explain the existence of the laws of physics on the principles of “natural selection.”

It’s fairly easy to notice that probability theory plays a key role in Smolin’s concept. Basically the principle of selection boils down to a game of probabilities. Karl Popper had pointed out that in a large enough collection of component members competing with each other in some respect, on the strength of not much more than just probability theory, the selection principle will start to operate within the collection. In his
Intellectual Autobiography
Popper wrote, “Let there be a world, a framework of limited constancy, in which there are entities of limited variability. Then some of the entities produced by variation (those which ‘fit’ into the conditions of the framework) may ‘survive’, while others (those which clash with the conditions) may be eliminated.”
⁵
Popper called this mechanism “situational logic” and held that it created a situation “in which the idea of trial and error-elimination, … becomes not merely applicable, but almost logically necessary.”
⁶
Popper formulated this concept for the purposes of his analyses of the methodological status of Darwin’s theory, and hence could regard it as “almost logically necessary.”
⁷

We have no difficulty in seeing that Popper’s “situational logic” is also applicable to the population of universes in Smolin’s concept. But with one important reservation – in the case of universes there is no environment to which they could “adapt” (unlike the situation in biological evolution). However, it turns out that “situational logic” also works in the absence of an environment. Gordon McCabe has compiled a set of precise axioms and shown that if they are met in a collection, then “natural selection” will set in almost automatically in that collection
⁸
(he makes no reference to Popper and does not use the term “situational logic”). His axioms (in a simplified form) are as follows: 1. The objects belonging to the system must have certain characteristics which make them differ from each other. 2. These objects must have a finite lifetime. 3. The characteristics of these objects must include some which do not change throughout the object’s lifetime and which define the type to which the object belongs. 4. For every object there exists at least one object which generated it. 5. An object’s characteristics are at least partly inheritable (reproducible). 6. In the process of reproduction characteristics are not copied exactly, instead mutations occur. 7. For an object of a particular type the birth rate and mean lifetime depend on its type (viz. they may differ for different types). If all of these conditions are fulfilled in any system then a selection process ensues in that collection on the grounds of the laws of statistics. Note that none of these conditions relates to an environment. In biological evolution the environment determines the generally finite set of resources indispensable for survival, and competition for access to those resources becomes an important factor of selection. But in the general case the conditions enumerated above trigger the inception of the mechanism of evolution. This is the situation pertinent to the population of universes in Smolin’s concept.

Does Smolin’s concept give a satisfactory explanation why “we live in a universe which is as it is, and no otherwise”? Unfortunately not. Above all, we should remember that Smolin’s concept is based on strong assumptions and we have no guarantee – neither a theoretical one, nor an empirical one – that these assumptions hold true in reality. The arguments Smolin produces to support them effectively reduce to propaganda devices.

What’s more, the existence of the multiverse itself demands an explanation. We may imagine that its explanation will prove no easier than the explanation why conditions which are life-friendly have developed in one universe.

But the existence of a multiverse is not the only point that poses difficult questions. Some of its properties are also problematic. Notice that the conditions cited above necessary for a selection mechanism to be activated make up a restrictive set of requirements. There are “infinitely more” possible systems which do not meet these conditions than there are ones which do. This applies to possible families of universes as well. Thus the question arises why the population of universes considered by Smolin belongs to that special sub-family of all possible families of the multiverses, and we belong to the sub-family of families in which the sub-family of universes postulated by Smolin exists. For if our universe belonged to a different sub-family we could not explain why there are life-friendly conditions in our universe. The superfluity of this ladder of explanations is striking and, what’s more, entails a risk of regression ad infinitum.

By contrast let’s note that in the family of universes referred to by Brandon Carter to illustrate the strong anthropic principle (see Chap. 8) there is no need for a selection mechanism. It is simply an ensemble of all possible universes, in which by virtue of the definition (since it is the ensemble of all the
possible
universes) there must be at least one that is life-supporting. In Smolin’s concept it is not a loose ensemble or collection, but a
population
, of universes related to each other by the mechanism of selection. The anthropic principle does not impose conditions as demanding as those in Smolin’s concept on the family of universes.

These critical remarks may be supplemented with an observation made by Roger Penrose.
¹⁰
Biological evolution certainly requires highly specific conditions. One of them is the second law of thermodynamics, without which evolution would not work. Since, in line with this law, the universe’s entropy is growing, then retrospectively it must have been less and less the further back in time. In other words, however specific the universe might be at its present stage of evolution, with life on at least one planet, it must have been even more specific at the earlier stages of its evolution, when life had not yet developed (since entropy was lower). Penrose points out, and supports this observation with a detailed calculation, that in accordance with the philosophy of the anthropic principles the selection of a universe out of all the possible universes, in which life came into existence by sheer random chance with no preceding phase as required by the second law of thermodynamics, is far more probable that the selection of a universe like ours with the second law of thermodynamics applicable in it from the very beginning.

In Penrose’s opinion it would have been far “cheaper” to produce the entire Solar System along with its inhabitants by means of random collisions of particles than to explain why the universe had such a low entropy at the beginning.
¹¹
He thinks that this fact may be explained when we have a prospective theory of quantum gravitation. That is why resorting to anthropic arguments is simply premature.

Finally one more remark concerning both Smolin and many other adherents of the multiverse idea. Very often they claim that their concept is scientific because it is falsifiable. This criterion as a test of whether a hypothesis is scientific or not was put forward by Karl Popper, who said that if a hypothesis is not open to falsification by confrontation with the results of experiment or observation, then it does not qualify as scientific. Discussions are still going on in the philosophy of science as to how exactly the principle of falsification functions in science, and to what extent it discriminates between scientific and unscientific ideas. However, there is no doubt that Popper’s criterion gives an accurate description of certain aspects of the practice of science, and because of this it is often invoked by natural scientists, albeit sometimes in a not very critical manner. This often happens in disputes on the multiverse idea.

So, for example, Smolin maintains that his concept is falsifiable, ergo scientific, since it predicts that the universe in which we live contains many black holes (because it belongs to the family of life-supporting universes). I doubt whether any methodologist worth his salt would admit such a prediction as a genuine falsifier of the hypothesis proposed. His main criticism would be levied against its vagueness. It is not clear what is meant by “many,” and with respect to what. According to Smolin’s concept a life-supporting universe should contain many black holes in comparison with other universes. How are we to determine the quantitative rules for such a comparison and, above all, how are we to carry out the comparison?

Max Tegmark, another protagonist of the multiverse idea, was even more nonchalant with respect to the criterion of falsification. In his excitement to convince the reader of the scientific nature of the idea of “parallel universes” (as he called it) he cited the following as an instance of falsification: “For instance, a theory stating that there are 666 parallel universes, all of which are devoid of oxygen makes the testable prediction that we should observe no oxygen here, and is therefore ruled out by observation.”
¹²
So, although it is a false theory, nonetheless, on the grounds of falsification, it is still a scientific theory.

We should realise, however, that not every statement (theory, model, hypothesis) the consequences of which may in principle be compared with observational or experimental results may be regarded as falsifiable, in the sense normally ascribed this concept in the philosophy of science. Taking Tegmark’s style of comparison further, let’s imagine that someone before the age of space flight had claimed that the other side of the moon, which is not observable from Earth, is painted red and carries the inscription, “Coke is it!” in big white letters. It would definitely have been a falsified (and therefore falsifiable) prediction, but it could never be treated as a test of whether the hypothesis was scientific or not. It’s true that no theory or hypothesis which is not falsifiable even in principle may be regarded as scientific, but not all statements which are falsifiable (in the more colloquial sense of the word) may be regarded as scientific. The question of criteria distinguishing science from what is not science is a difficult methodological problem. Anyone who wants to write on this subject would do well to first look up the copious literature devoted to it.
¹³