Monday, 2 March 2015

The Singular Universe and the Reality of Time – A review





Physicist and Philosopher Lee Smolin and Philosopher Roberto Mangabeira Unger have teamed up to produce their long awaited book The Singular Universe and the Reality of Time: A proposal in natural Philosophy
 
Throughout his career Smolin has injected novel ideas and principals into scientific descriptions of the universe. His 1992 paper (1), Does the universe evolve?, for the first time, offered a Darwinian process as an explanation in a fundamental physical theory. This paper was followed in 1998 by his book, Life of the Cosmos (2), which further develops and describes his Darwinian theory. He calls it Cosmological Natural Selection (CNS) and uses it to explain why our universe is a ‘Goldilocks universe’, just right for the existence of complex structures like galaxies, stars and chemistry. 

 Figure 1: Lee Smolin

He uses an analogy to natural selection in explaining that our universe is one in a long line of universes which evolved over cosmological time. What appears as the formation of a black hole in a parent universe is actually a disconnected child universe starting off in its own Big Bang.  With the bounce of a black hole into a child universe the laws and parameters of the parent universe are inherited by the child with slight variations. Over evolutionary time this Darwinian process has selected universes with ‘just right’ laws and parameters for fecundity in their production of black holes. Black hole fecundity relies on the existence of galaxies, stars and complex chemistry. Therefore our universe has evolved to its ‘Goldilocks’ state in order to achieve reproductive success in the production of black holes.

This theory does not provide much centrality for Homo sapiens. Our existence, and the existence of life, is merely made more plausible. It is not a full explanation of life, it only explains the existence of a cosmological context necessary for the evolution of life.  Smolin is content to leave life’s fuller explanation to Darwin’s theory of natural selection. 

It is perhaps of even greater importance that CNS marked a milestone in the largely unnoticed scientific revolution of universal Darwinism. Since Darwin’s time his theory had become ever more central to biology as its molecular details came into focus. Philosophers of science such as Donald Campbell and Karl Popper had since the 1970s described science itself as a Darwinian process which accumulates knowledge. In 1976 Richard Dawkins, in his book The Selfish Gene, speculated that there was a cultural replicator he coined the ‘meme’ which might be at the basis of all forms of cultural evolution. Today practically every field of the social science has a school bearing the ‘evolutionary’ prefix in its title: evolutionary psychology, evolutionary linguistics, evolutionary economics etc. In 1987 the Nobel Laureate Gerald Edelman introduced the Darwinian process to neuroscience where it rapidly spawned a large number of variations. 

Only the physical sciences remained unexplained by Darwinian Theory but with CNS this last gap was beginning to be filled. For those of us with our eye on this particular ball the next obvious question was why; why does nature repeatedly use this single mechanism to create and evolve so much of reality? Answering this question could possibly provide a unified view of reality which has so far eluded science.

Finding the answer would take a fine scientific mind with great mathematical expertise and a grand vision; it would take someone like Smolin. Surely, now that he had introduced the Darwinian processes to physics, he would take an interest in this larger question.
Since first developing CNS Smolin has spent much of his time at the forefront of the development of Loop Quantum Gravity, a leading contender in the race to find the theory of Quantum Gravity, the holy grail of modern physics.  However one part of Smolin’s creative mind has continued to think about the evolution of the Cosmos. In a number of follow-up papers (3; 4; 5; 6) he summarized the current scientific evidence pro and con his theory and argues for its many advantages over string theory.

Smolin, like Newton and Einstein before him, gives a lot of weight to philosophical guidance. He champions a relational view of physics as pioneered by Gottfried Wilhelm von Leibniz and Charles Sanders Peirce where the physical properties of entities can only be understood in relation to other entities. This view is also central to the epic research program conducted by Julian Barbour on the nature of time (7)

Much of Smolin’s recent writing features philosophical criticisms of widely held views within the physics community such as the Newtonian paradigm, string theory and the anthropic principle. He believes that physics and cosmology have lost their way and hopes to lead it back to the true path through development of guiding philosophical principles.
Smolin has accepted a faculty position in the philosophy department at the University of Toronto. Since 2006 he has collaborated with the Brazilian philosopher Roberto Mangabeira Unger on a project to develop philosophical principles capable of guiding research in physics and cosmology.

Unger is a polymath who has made important contributions to the law, philosophy, economics, natural philosophy and political science. His influence is practical as well as academic; he was a leader of the movement which brought progressive government to Brazil and was appointed minister of strategic affairs in 2007 and 2015.
His radical political and philosophical outlook are somewhat reminiscent of the counter-culture ideals many of us entertained at one time. He views the present tense of our lives as a creative interface between past and future where the individual’s potential has the capacity to transcend social roles and institutions. 

This notion of the present as the focus of reality also permeates his understanding of natural history where again it is the present which is the forum for the creative unfolding of reality. Philosophically he understands the universe to evolve through the inventive action of the present and thus finds greater explanatory value in becoming than in being.

It is perhaps his view of time which Smolin has found most stimulating. Unger’s view of the universe as an historical process resonates particularly well with Smolin’s CNS. In any case their relationship seems to have greatly energised Smolin.


Figure 2: Roberto Mangabeira Unger
Smolin’s February 2012 paper (5) A perspective on the landscape problem examines the current crisis in physics from a philosophical perspective. He considers Leibniz’s principle of sufficient reason, that there must be a reason that explains why each property of nature is what it is, as a bedrock principle for understanding physics. These reasons are often historical; everything is what it is because it got that way. Smolin’s theory of cosmological natural selection is proposed as a scientific theory that respects the principle of sufficient reason. Smolin also notes that this paper is partially based on his collaboration with Unger concerning the evolving laws of nature.

This joint project drives much of Smolin’s subsequent research. He describes his 2013 book Time Reborn as an introduction to the book he is writing with Unger. He describes his 2013 paper (8) Temporal Naturalism as outlining a project:
This essay is part of a larger project whose aim is to radically reconfigure the practice of science on a cosmological scale to admit three theses: 1) the reality of time, 2) the evolution of laws with respect to that time and 3) the uniqueness of the single causally closed universe that unfolds in time. 
This project was conceived with Unger and its main fruit is their book being reviewed here. For almost three years prior to its publication this project was central to Smolin’s creative output while little was heard from Unger on the subject.

The book is an unusual version of shared authorship as it is actually two books, the first by Unger and the second by Smolin. Each author gives a full account of their individual take on three central ideas: the singularity of the universe, the reality of time and the limited power of mathematics in cosmology. A long concluding chapter lists the many issues on which they disagree and records each of their divergent views on those issues.

The topic of most heartfelt agreement seems to be the reality of time. They share the belief that all of reality changes with the passage of time in a manner that is not fully predictable. For those familiar with Smolin’s body of work there is really very little new here. A natural philosophy which views the universe from an historical perspective has been central to Smolin’s research for over twenty years. CNS is only described in this book as an example of a theory which may fit within their natural philosophy.

Obviously his relationship with Unger has greatly stimulated Smolin and lends support to his view of a historical universe. However I cannot escape concluding that Unger’s philosophy is relatively sterile and serves to blunt Smolin’s great insight that the universe may evolve through Darwinian processes. Unger’s philosophy forms a defensive structure around his notion of the true nature of things; nothing is allowed admittance which does not conform. 

In a section of the book debating their differences Unger goes so far as to reject CNS on the grounds that time changes everything and that there can be no process which itself remains unchanged as he imagines is the case with CNS. 

It is clear, at least within the forum of the cosmos, that Unger is not a great fan of Darwinian explanations. One cannot help but imagine the outcome if Smolin had collaborated instead with a philosopher having greater understanding and sympathy for the Darwinian perspective. I can, for example, imagine Smolin teaming up with Daniel Dennett and the fruitful exploration which may have ensued.
 
Dennett understands the explanatory power of Darwinian processes. Probably his most quoted observation is that Darwin had the best idea ever (9), ‘ahead of Newton, ahead of Einstein, ahead of everybody else’. Certainly ahead of Roberto Mangabeira Unger. In his breakthrough 1995 book Darwin’s Dangerous Idea, Dennett outlines the notion of ‘design space’, a type of mathematical space containing all possible designs: physical, biological and cultural. For example we might conceive of the biological design space as composed of all possible sequences of DNA base pairs. He understands Darwinian processes to be active mechanisms which search through this design space and select those designs which are solutions to the challenges of existence.

Figure 3: Daniel Dennett

This great book was a milestone on the path to Universal Darwinism, the understanding that all complex entities may be created and evolve through the action of Darwinian processes. Dennett understood that both biology and culture could be put under this theoretical umbrella. He even discussed Smolin’s theory of cosmological natural selection as an example of how physical reality may also be included within this paradigm. Dennett has a deep understanding of Darwinian processes and runs with it. Here is someone who might have provided fruitful philosophical mentoring.

A scientific treatment of a design space might consider it in terms of a vast probability distribution where a probability is assigned to the existence of each possibility. The vast majority of designs must have a probability arbitrarily close to zero; most possible designs are not viable solutions to the challenges of existence. Darwinian processes work their way through the space of all possible designs trying out each in turn. Those that can have an existence do have an existence; knowledge of how existence may be achieved is selected and accumulates through an evidence based process.

Evidence based probabilistic models operate according to the mathematics of Bayesian inference which describes how probabilities are reassigned on the reception of new evidence or information. In this view Darwinian processes are physical implementations of the mathematics of Bayesian inference (10; 11; 12)

If Smolin had arrived at this paradigm via the philosophical stepping stone of Dennett’s insights he would be much better positioned to answer the many conundrums presented by the natural philosophy he has developed with Unger.  

First, their idea that all things change in a creative and unpredictable manner within the present moment of time, may solve the conundrum of the second law of thermodynamics at least when stated in terms of information entropy. In the universal Darwinism paradigm all of the many probabilistic models over design space, including physical reality, biology, culture and even science itself accumulate knowledge through Darwinian processes. This knowledge is an expectation of how to achieve an existence in the entity’s particular environment. New stuff popping into existence in the present moment is unforeseen by these probabilistic models and thus the passage of time tends to increase the ignorance or entropy of all models. 

In this context the second law is obvious: any model becomes less certain as time goes on because there is more new stuff to know. The continued existence of complex entities depends upon gaining knowledge of the current conditions which is the function of Darwinian processes.

In the last part of the book Unger lists his disagreements with Smolin’s published work and Smolin offers a rebuttal. Unger does not shy away from criticizing Smolin’s theory of CNS. He sees it as violating his principle that laws are not constant but change and evolve with their subject matter over time: 
the idea of cosmological natural selection would give a central and permanent role, in the history of the universe, to one mode of change: the one described by Darwin.
In his rebuttal Smolin points out that Darwinian processes are not laws but rather a form of logic and thus share in the timeless nature of mathematics. He is reduced to defending his most brilliant theoretical contribution to his collaborator. If his natural philosophy had evolved instead, under Dennett’s influence, to that of universal Darwinism he could have made much more powerful arguments defending CNS.

A litmus test for Smolin and Unger is that a good scientific theory must explain its laws. This is in contrast and forms a criticism of theories in the Newtonian paradigm which include just about all existing physical and cosmological theories. Theories in the Newtonian paradigm take as given a set of initial conditions and a set of dynamical laws that drive the evolution of those initial conditions through time. The question ‘why these laws?’ is not answerable within this paradigm which cannot therefore provide a complete explanation.

Amongst the most powerful possible explanations of ‘why these laws?’ is a demonstration that the laws are unique in their ability to explain the phenomena in question. Dennett describes such steps in reasoning as ‘forced moves’ such as in a chess game when there is only one move that will stave off defeat. For example, an early hope for string theory, was that a version would be found that was the only mathematically consistent theory of quantum gravity. Unfortunately this did not turn out to be the case with string theory as it has been shown that there are about 10500 consistent versions (5).

However it was proved by Richard Cox in 1946 that Bayesian inference is the uniquely consistent mathematics of probability (13). Any other theory of reasoning from uncertain knowledge must either be inconsistent or wrong. It has further been demonstrated that a reasonable definition of knowledge is contained within Bayesian inference (12). If we view Darwinian processes as physical implementations of Bayesian inference then it is clear that they are the unique method for accumulating this type of knowledge. If we take one further step along the path of universal Darwinism and accept that knowledge is required for the existence of complex entities and is in fact their essence then we are led to conclude that Darwinian processes are the unique method by which complex entities evolve. The last step in this argument is that all entities we know of, even down to quarks and gluons, are vast compared to the fundamental Planck scale and are therefore should also be considered complex entities.

Without the accumulation of knowledge to constrain local entropy production any complex entity will gain entropy and eventually lose its complexity and its integrity as an entity. Thus the question ‘why this law of Darwinian processes?’ can be answered: there is no other possible process by which complex entities can maintain their existence. Accepting this law is a forced move. 

Unger’s view of the historical universe includes the observation that ‘laws evolve with the systems they describe’. This minority view has a long history reaching back to Leibniz and Peirce. It is easily demonstrated by considering that the historical universe consisted, at an early stage, of only a quark/gluon plasma. At this stage there was no atomic physics, biology, cosmology or culture. All of these subject matters as well as the laws describing them came into being as the universe evolved.  Smolin shares this view with Unger’s but adds the insight that Darwinian processes are ‘the general logic making such evolution possible’.

Universal Darwinism might have provided him with a deeper explanation of the evolution of laws. In Dennett’s view Darwinian systems constantly probe design space searching for the good solutions. When good solutions are found they tend to spread and become essential to a wide range of phenomena. These commonly shared design details may be understood as laws (11). In this view scientific laws are merely the design details used in a wide range of solutions. Thus laws evolve with design.

For example with the evolution of sexual reproduction came the mechanism of dominant and recessive genes and the design details that form the law of Mendelian Inheritance. This law is not a timeless principle, it is a good solution discovered by natural selection and came to be used by a large class of organisms. The process is highly contingent and if life is found throughout the cosmos we may expect there will be many instances where this law does not apply. In these cases the Darwinian process which govern the life forms’ evolution may have found other, perhaps better, solutions to the challenges of existence.

Unger appears riled by Smolin’s seemingly innocuous observation that science is our most reliable guide to nature (8):
naturalists also hold that science is the most reliable route to knowledge about nature.
This quote is taken from the paper Temporal Naturalism which Smolin describes as part of his joint project with Unger although Smolin is the sole author of the paper. It may have slipped his mind that Unger is not a scientist and has an ego unlikely to let this kind of imagined slight go by. Unger does not let it go by but writes:
The claim that science is either the sole or even the most reliable source of insight into nature should exercise no influence on the temporal naturalism that we here espouse. There is no hierarchy of forms of inquiry that could entitle one to say that one such form trumps the others.
and
To assert, on behalf of science, a privilege over other sources of insight and experience does science no favor.
Is he having us on here? When he gets sick does he go to someone who has studied medical science for many years or does he go to someone who hasn’t?

These comments smack of post-modernist intellectual relativism where anyone’s views concerning nature are equally as correct as anyone else’s. Yet he presumes to be developing ‘methods and principles adequate to a science of cosmology that is not simply a scaled-up version of contemporary physics’. Why not develop methods and principles for a new astrology, after all cosmological science holds no privilege over astrology? Geez Lee, how could you have teamed up with this?

The superior knowledge of science is proven by its power. For instance scientific knowledge has been used to create the only known instance of atomic fusion outside of a star. From the beginning humans have dreamt of such power to deliver themselves from their enemies. For better or worse, after a history filled with sacrificial rituals, prayer and philosophizing, of all the schemes we humans have devised to bend nature to our will, only science has accumulated the knowledge to deliver on this and so many other dreams.

In my opinion Smolin makes only a lame rebuttal, to the effect that scientists tend to form dedicated communities of truth seekers. Universal Darwinism would provide him with a crushing counter.

Models can only accumulate knowledge if they follow the mathematical principles of Bayesian inference. This is an evidence based process where the model assigns probabilities to a family of competing hypothesis and these probabilities are updated to reflect the implications of any new evidence. In this manner the model gains knowledge and moves closer to certainty. All of human knowledge is based, at least loosely, on this process; we learn from experience. However some processes follow the mathematics more closely than others and are therefore more efficient. Science is clearly at the top of this hierarchy as Bayesian principles are the heart of the scientific method. Indeed E.T. Jaynes’ great textbook (14) on Bayesian inference is titled: Probability Theory: the logic of science.

As a demonstration of this claim I would point to moon rockets, cell phones and even steam engines as examples of the wondrous entities which may be produced from the knowledge which science has accumulated about nature. Please show me comparable entities which may be produced from the accumulated knowledge of, let’s say, astrology or even philosophy. I believe this proves that science has a privileged understanding of nature in that its knowledge of natural processes may be used to construct novel natural entities having great complexity and power. 

Unger also rejects the principle of sufficient reason which Smolin endorses so fervently. His argument seems to be that the universe must be free to do its own thing and cannot be hemmed in by reasons. In his view the principle of sufficient reason is rejected because:
It expresses a refusal to accept the factitiousness of the universe: that it just happens to be one way rather than another.
Sorry, but I cannot leave this quote without noting Unger’s fondness for impressive sounding words which hardly anyone has in their vocabulary. Actually, it seems, they are not in his vocabulary either but like Humpty Dumpty he assumes they can mean whatever he wants them to mean. Surely he does not mean factitiousness, a noun specifying a thing made from a human or artificial design rather than from a natural design. A Kindle search of his section of the book returns twenty-four instances of this word and along with many of those repetitions he insists, contrary to any dictionary I can find, that it means ‘a thing that is what it is’. Perhaps he conceives of himself as above the realm of dictionaries and like Shakespeare is free to give old words new meanings. ‘A thing that is what it is’; perhaps he meant ‘uncaused’ or ‘unexplainable’? Nope. He doesn’t like uncaused because everything is caused and he doesn’t like unexplainable because that paves the road to explanatory nihilism. Better to use factitiousness and then claim that it means ‘a thing that is what it is’. I dunno, there may be something very deep going on here which I do not have the wit to understand. 

Smolin makes an important observation concerning the equivalence of computation and physical processes:
There have been a number of suggestions that physical processes are computations.
He uses this observation to speculate that the universality of computation may resemble the universality of physical theory. Given the perspective of universal Darwinism he might understand the universality of computation as providing an explanation for his notion of creativity within the present moment of time.

Both MIT’s Seth Lloyd and the Nobel laureate Gerard ‘t Hooft have suggested that quantum interactions are quantum computations; what is computed is the outcome of the interaction (15; 16). In this view what takes place in the present is a vast array of quantum computations, computations whose outcome is the future.

As ‘t Hooft explains, the outcomes of these deterministic computations are creatively unpredictable:

'Deterministic' cannot imply that the outcome of the evolution process can be foreseen. No human, nor even any other imaginable intelligent being, will be able to compute faster than Nature itself. The reason for this is obvious: our intelligent being would also have to employ Nature's laws, and we have no reason to expect that Nature can duplicate its own actions more efficiently than itself.

We might wonder at the nature of quantum computation. In what sense are quantum interactions equivalent to quantum computations? It has been shown that quantum computation is homologous with quantum interactions as described within Wojciech Zurek’s theory of quantum Darwinism (17). Quantum Darwinism describe quantum interaction as the transfer of quantum information where only a very small subset of the information describing one entity can survive the transfer to another entity (18). The information that does survive is selected by a Darwinian process which Zurek has named quantum Darwinism. Thus the present moment is composed of a plethora of quantum interactions where the future is being instantiated through the probing, searching and testing of a Darwinian process.

Smolin’s greatest legacy to science may be his introduction of the Darwinian process as an explanatory mechanism within fundamental physics. At one point in his career he gave thought to intriguing questions such as: what do physics and biology have in common? His answers were very good, for instance (6):

  There is only one mode of explanation I know of, developed by science, to explain why a system has parameters that lead to much more complexity than typical values of those parameters. This is natural selection.

 This statement is tantalizingly close to a claim that Darwinian processes serve as an explanation for all complex systems. Disappointingly, at the current stage in his career, rather than evolving his brilliant insights this book stunts them by teaming up with a less fertile and at times antagonistic philosophy.
 
Note: One really great thing about Physics is the website www.arxiv.org. It is an electronic archive of physics papers hosted by Cornell University. Wikipedia reports that (19)In many fields of mathematics and physics, almost all scientific papers are self-archived on the arXiv’.  I have certainly found this to be true and marvel, for example, that most of Lee Smolin’s papers may be found here and read for free. His are surprisingly clearly written and I believe that any interested person will find them quite accessible.

Bibliography

1. Did the universe evolve? Smolin, Lee. s.l. : http://www.universaldarwinism.com/documents/smolin-did-the-universe-evolve.pdf, 1992, Classical and quantum gravity, pp. 173-191.

2. Smolin, Lee. The life of the cosmos. s.l. : Oxford University Press, 1998.

3. —. The status of cosmological natural selection. [book auth.] R. Vaas. Beyond the Big Bang: Competing Scenarios for an Eternal Universe. http://arxiv.org/abs/hep-th/0612185 : Springer, 2013.

4. —. Scientific Alternatives to the Anthropic Principle. [book auth.] Bernard Carr. Universe or Multiverse. s.l. : http://arxiv.org/abs/hep-th/0407213, 2007.

5. A perspective on the landscape problem. Smolin, Lee. s.l. : http://arxiv.org/abs/1202.3373, 2012, Preprint: Foundations of Physics.

6. The case for background independece. Smolin, Lee. s.l. : http://arxiv.org/abs/hep-th/0507235, 2005, arXiv:hep-th/0507235v1.

7. The Solution to the Problem of Time in Shape Dynamics. Barbour, Julian, Koslowski, Tim and Merc, Flavio . s.l. : http://arxiv.org/abs/1310.5167, 2013, arXiv preprint.

8. Temporal Naturalism. Smolin, Lee. s.l. : http://arxiv.org/abs/1310.8539, 2013, Preprint.

9. Dennett, Daniel C. Darwin's Dangerous Idea. New York : Touchstone Publishing, 1995.

10. Bayesian Methods and Universal Darwinism. Campbell, John O. s.l. : http://arxiv.org/abs/1001.0068, 2009. AIP Conf. Proc. 1193, 40 (2009), DOI:10.1063/1.3275642. pp. 40-47.

11. Campbell, John O. Universal Darwinism: The path of knowledge. s.l. : CreateSpace, 2011.

12. —. Darwin does physics. s.l. : CreateSpace, 2015.

13. Probability, Frequency, and Reasonable Expectation. Cox, R.T. s.l. : 
http://jimbeck.caltech.edu/summerlectures/references/ProbabilityFrequencyReasonableExpectation.pdf, 1946, Am. Jour. Phys. 14,.

14. Jaynes, Edwin T. Probability Theory: The Logic of Science. s.l. : University of Cambridge Press, 2003.

15. Lloyd, Seth. Programming the Universe. s.l. : Vintage; Reprint edition, 2007.

16. The Cellular Automaton Interpretation of Quantum Mechanics. 't Hooft, Garard. s.l. : Arxiv preprint, 2014, Vol. http://arxiv.org/abs/1405.1548.

17. Characterizing the Structure of Preserved Information in Quantum Processes. Blume-Kohout, Robin, et al. s.l. : http://arxiv.org/abs/1006.1358, 2008, Physical Review Letters, 100 (3). Art. No. 030501. ISSN 0031-9007.

18. Quantum Darwinism. Zurek, Wojciech H. s.l. : http://www.nature.com/nphys/journal/v5/n3/abs/nphys1202.html, 2009, Nature Physics, vol. 5, pp. 181-188.

19. Wikipedia. arXiv. Wikipedia. [Online] [Cited: 2 28, 2015.] http://en.wikipedia.org/wiki/ArXiv.