The Fabric of Reality/Commentary

From Simulism

Jump to: navigation, search

The Fabric of Reality by David Deutsch is a hugely important book in relation to Simulism. There are a number of themes running throughout the book which are common to themes within this Wiki, and in addiiton, several chapters in the book discuss issues of direct relevance to the idea of a Simulated Reality. It is therefore useful to provide a commentary on the book, and a critique of some of the ideas and arguments presented.

Contents

[edit] Book Summary

In The Fabric of Reality Deutsch claims to have discovered common themes and similarities occurring within four major theories of reality:

  • Quantum Physics
  • Epistemology
  • Theory of Computation
  • Theory of Evolution
He further suggests that taken together, these theories can provide us with an enhanced understanding of what reality actually is, and how such understandings develop and mature via a process of evolutionary epistemology. Out of this analysis, Deutsch provides us with a criterion for reality:
" We should regard as real those complex entities which, if we did not regard as real, would complicate explanations"

Deutsch goes on to explore each of his main themes, and concludes:

  • What we take to be the physical universe is one of an infinite number of universes within the multiverse.
  • Knowledge is created via a process of problem solving, and is essentially an attempt to understand some phenomenon by creating a mental model of it.
  • Evolutionary processes are basically another form of knowledge development, in which information concerning the environment is transmitted down the generations via DNA
  • All physical processes can be viewed as computation, and as such are susceptible to the Turing Principle, that a universal computer could, in theory, be programmed to perform any computation which could be done by any other computer.

Taken together, these ideas form the following picture of reality:

All possible universes exist, as single universes within a multiverse extending through time: this is the only physical reality. What we take as reality is the journey of an individual (object or person) is through this multiverse. This occurs via choices made at each moment in time. Such choices may be conscious decisions or random chance. The path of an object or person up to a particular 'moment' in time is unique and fixed; after that moment, their journey potentially splits via a process of bifurcation into a myriad possiblities. When such a process occurs, resulting possible universes can be viewed as 'nearby', and information-rich large-scale structures such as galaxies are common across such nearby universes. In particular, life forms an important information-rich large scale structure which has the potential to affect the future course of the universe, and life forms across the multiverse constitute stable and pervasive elements. Conscious beings are integral to this process, and create knowledge within the universe by attempting to understand it through a process of modelling. This multiverse can be regarded as a physical computer, and the knowledge based structures (including life forms) can be interpreted as the results of a physical computation processes, one of which is evolution. All knowledge-based structures can be rendered as physical or virtual objects via computation as real objects or as virtual reality simulations, and in doing so, this creates further knowledge and understanding via the process of evolutionary epistemology. Each conscious individual renders their journey through the multiverse as a Virtual Reality representation in their own mind, created via a problem-solving process, akin to the Popperian version of the scientific method. The process of rendering is an active process, and structures, once having been rendered, cannot be un-rendered.


[edit] Criticisms

[edit] Deutsch's Criterion for Reality

We can first of all take issue with Deutsch's "criterion for reality". He derives this from the Gallilean argument that if the universe appears to behave "as if" a particular theory is true, and if in denying it, explanation is complicated, then we should take that theory as reality. The problem here is twofold. First of all, this is a disguised form of Occam's Razor, which is a good 'rule of thumb' for deciding between competing hypotheses, but does not in itself guarantee that one explanation is any better than any other at providing a valid description of reality. Indeed,it can be argued that in some respects, the simplest world-view is to believe that God created and runs everything, and all knowledge stems from Him. Many would take the view that the world appears "as if" someone had created it. Clearly the argument here is not straightforward, and such a view is open to the challenge that hypotheisisng the existence of such a God is to create a hugely complex and autonomous entity which in fact complicates, rather than simplifies, explanation. The point here is not to debate the existence of God, but to point up flaws in Deutsch's analysis. His criteria leads us into controversy: what is complicated to one person, may well be simple to another - it depends upon their world view and their own subjective model of reality.

[edit] Deutsch's Reading of Quantum Mechanics

There is a single view of Quantum Mechanics presented in the book, and noen of the other explanations are adequately explored. Deutsch admits this, but dismisses other explanations , saying "I have merely described some physical phenomena, and drawn inescapable conclusions". However, his conclusions are based on three premises:

  1. that photons in the 'slit' experiment are behaving "as if" they are interacting with imaginary photons
  2. that the only interpretation for this is that these imaginary photons exist in other, parallel universes
  3. that his criterion for reality (see above) is valid.

Deutsch does not present any actual evidence for (1), it is merely stated as a fact; however if we examine this, it does not appear to make sense. If photons were actually interacting with other photons as particles, we would expect them to behave as particles. In fact, this is not what is meant. Here what is actually being said is more like the following: Photons in the slit experiment behave "as if" they were waves interacting with other photons also behaving "as if" they were waves. This alters the interpretation entirely; if we are to use Deutsch's own criterion for reality, we might reject the fact that photons have any existence as particles, and look for a reality where they are viewed purely as waves.

There are clearly other interpretations of Quantum Mechanics, such as the classical 'Copenhagen Interpretation', which is dismissed by Deutsch as offering empirically verifiable predictions, but without any clear philosphical underpinning. The transactional interpretation of quantum mechanics does however offer an alternative to the 'many worlds' approach of Everett, whereby particles emit waves which interact with their surroundings in order to guide them to particular interactions. While we might take issue with this interpretation, it is clear that both this and the many worlds interpretation invoke highly complex and non-trivial entities within their explanations, and it is not a simple matter to select between these competing hypotheses.

Finally, as we have seen above, the use of the 'criterion for reality' in (3), may not be as simple as Deutsch suggests. He takes the multiverse to be real, because the experiment demonstrates that photons behave "as if" it is real; however, nowhere does Deutsch explain exactly how this process would work, if there were multiple universes, and what would be the consequences of this. For example, how, exactly is interaction supposed to take place between these universes, and why, if these occur at the quantum level, are such interactions prohibited at larger scale levels. There may well be good scientifc reasons for these, but these are not explored within the text.

[edit] Deutsch's Use of of Turing Principle

The Turing Principle, in essence asserts that anything which is Turing Computable, is computable on a Universal Turing Machine. All of these terms have very precise mathematical definitions, which only marginally coincide with the commonly understood notion of computability. Turing's statement is a very different matter from claiming that a machine can be built which can mimic the calculation of any physical process. There are many misunderstanding of what Turing's results actually claim see http://plato.stanford.edu/entries/church-turing/ for a discussion of these. Deutsch seems to be taking the widely held view that the "... Universal Turing machine can compute any function that any computer, with any architecture, can compute" (Dennett, D.C. 1991. Consciousness Explained.)

However, Deutsch takes this principle much further than most, claiming that as the physical universe is effectively a computer (following computationalists such as Seth Loyd), we can modify the Turing Principle to assert that that in theory, a universal computer could be constructed which would render any physical process in virtual reality. He further goes on to claim that "Since building a universal virtual-reality generator is physically possible, it must actually be built in some universes". There does not appear to be any justification for this claim, which has now taken us to the remakable position where Deutsch is asserting that there exists somewhere, in some universe, a computer capable of rendering the whole of creation. If this were true it would be a remarkable result, and one which would actually prove the simulation hypothesis.

[edit] Deutsch's View of Time

In essence, Deutsch's critique of the Newtonian view of time, rejecting the idea that the 'now' moves continuously along a timeline is correct. That model inevitably leads to the idea that we can slow down, speed up, stop, or even reverse the flow of time; however, in order to do this, we would need a 'meta-time' in which to operate. This eventually leads to an infinite regression of meta, meta-meta, meta-meta-meta etc., times, and does not advance the argument. On the other hand, the 4D model where all spacetime is fixed leads into difficulties of free will and determinism, but this is where Deutsch uses the Multiverse to escape from this conundrum. While it is true that we can contemplate an individual's journey through spacetime consisting of the mutiverse+time, in doing this, we are hooked once again by the notion of a meta-time. Deutsch tries to wriggle off this hook by claiming that if the time snapshots making up the timeline were shuffled, it would not matter to the individual, as we could still make sense of each here-now by constructing our history from past memories and experiences, which are present in our brains in the here-now.

This explanation seems to be unsatisfactory to say the least. If we were to apply Deutsch's own criterion for reality here, we might say that the universe appears to behave "as if" it were moving along a timeline; any notion that the individual times might be shuffled, and sense made out of them by reconstructing on a frame by frame basis seems convoluted in the extreme. We should therefore reject this as an explanation of reality.

Throughout the book there is an intriguing idea underpinning Deutsch's discussions that time 'unfolds' by rendering the universe in a VR-simulation, which does not seem to be made explicit. In his discussion of CantGoTu environments, Deutsch asks what it would be like to experience one of these environments. He comes to the conclusion that no matter how long you stay in such an environment, and no matter how like it has been so far to an environment that you have encountered previously, somewhere in the future, it will diverge and be different. This is clearly predicated on computation, and the iteration of the next steps in the algorithms which underpin the rendering. By this criterion time can be interpreted as the rendering of subsequent steps in the algorithm. What makes this problematic however, is that we appear to need our notion of time in which to process these steps.

All of these avenues lead us full circle back to the notion of time moving inexorably forward, step-by-step.

[edit] Deutsch's view of Mathematics

This is not so much a criticism of Deutsch's viewpoint here, but an accusation of irrelevance. At the end of the previous chapter he claims that the book will "provoke many mathematicians". His main conclusion is that mathematical truth depends on, and is no more reliable than, our knowledge of the physical world. However, I know no mathematician who would dispute that or at least not have heard this a million times. Since the start of the 20th Century, when Russell was writing Principia Mathematica, through Godel and the Incompleteness Theorems up to the present day, mathematicians have been struggling with the facts that:

  1. Mathematics appears to produce hard-and-fast provable , "true" statements
  2. Mathematical objects appear to have substance and can be regarded in some senses as 'real'
  3. Any mathematical system is merely a combination of its assumptions (axioms), and the logic used to derive results. Both of these can be changed, so that the results change.
  4. Godel has shown that in any mathematical system worthy of the name there will be unprovable results; the system will be incomplete.

Paul Ernest'a 1991 text called "The Philosophy of Mathematics Education" performed a critique on the Absolutist Philosophies of Mathematics, and argued the case for a Social Constructivist Theory of Mathematics, which claims that mathematics is simply a social convention, and that mathematical knowledge is agreed via shared understanding and negotiation of meaning. Nearly everything that Deutsch says is a re-iteration of Ernest's ideas; no doubt Deutsch has come independently to the same conclusions, but it has all been said before, much better, and is currently the prevailing paradigm, certainly amongst mathematics educationalists.

Views
Personal tools
Advertisements
Toolbox