Archive: The End of Time

First posted 15 January 2006. Barbour's picture is about as close to my worldview as you can get from a materialist perspective. I 'just' replace his infinite set of particle configurations with an infinite set of experiential moments.

I've just read Julian Barbour's The End of Time, a good history of the physics and philosophy of time that also puts forward a radical view of time itself.

Update to the Block View
Barbour's historical account helped me to realise that the view of time I put forward in All the Time in the World (or more specifically a couple of aspects of it) is Newtonian and doesn't recognise a couple of truths from the relativity revolution.

In that post I said, when distinguishing between the space dimensions and the time dimension in space time that although we can easily imagine orienting a three-axis grid however we like in space, we cannot think of 'rotating' the time axis at some angle.  In fact, special relativity and Minkowski space-time do away with absolute simultaneity and DO in fact allow rotating the time axis in 'trade-offs' with the space ones.  (It also gives us light cones and time-like, space-like and light-like relationships among events, but I'll save that for another post.)

General relativity goes even further by getting rid of the idea of 'clean' Euclidean planes of simultaneity altogether, at least in the vicinity of mass and the distorting effect it has on space-time.  Although I think I've got my head around the special relativistic implications for my view, I can't quite claim to be on top of what adjustments general relativity requires.

Getting rid of the vessel and defining a new space
Barbour doesn't adopt the 'block view' of time that appeals to me.  Instead he asserts that time does not exist.  Einstein's radical step was to dismiss the notion of absolute simultaneity and to take the very practical approach of asking what we could know via the use of rods and clocks.  Barbour thinks a further radical step is necessary, and his denies not only the existence of temporal becoming, which is denied by many others, but also the existence of any space-time 'vessel' for holding all the things that exist.  In the absence of that vessel, all that matters is the relative positions of the things themselves.  A scientist named Mach, whose influence on Einstein was great but fell away in Einstein's final analysis, is the father of this view.

Barbour believes that a Machian world is strongly suggested by Einstein's general relativity, but that Einstein didn't quite carry things through far enough to see it.  He goes further to say that such a view may be exactly what is necessary to reach an acceptable theory of quantum gravity, reconciling and uniting Einstein's cosmological theory with the weird microscopic world of quantum dynamics.

For Barbour the building blocks of reality are the infinite possible relative positional configurations of existing particles.  Each of these configurations might be thought of as an 'instant'.  (Certainly, in a layman's evolutionary view of time, each instant corresponds to some configuration of particles - a snapshot of the universe, although one might ask just how long an instant is.)  These snapshots are not layered neatly onto one another to create a book or block.  They are just in a jumble, like wooden shapes in a bag.  So Barbour's view might be called the 'timeless bag' view.

Now if we can imagine a mathematical space of huge (indeed, infinite) dimensionality, we could match each possible configuration of the universe to one point in that space.  Barbour calls this space Platonia, and it is central to his story.

The case for Structuralism
The term structuralism, as applied here, is mine, not Barbour's.  His theory is structural in that the 'shape' of configuration space (Platonia) determines how the wave function - that messy thing (or mathematical construct, depending on your view) central to quantum mechanics - hovers over it, making some configurations more likely than others.  So causality exists not 'vertically' (i.e. through time) but rather horizontally (i.e. sheerly through eternal relations and resonances).

Barbour reminds us that Schrodinger's first derivation of his wave equations was actually a time-independent one.  It predicted with great accuracy the energy levels in the Bohr model of the atom.  Schrodinger went on to develop time-dependent equations, which came to be viewed as the more fundamental relations.  For Barbour, the first is the more fundamental, and it is strong support for a timeless universe.

He also cites what is known as the Wheeler-DeWitt equation. This equation is controversial in that its derivation may be flawed, in that it is mathematically incomplete (new techniques must be developed to solve it), and in that its interpretation is not clear.  But Barbour thinks it could be an early version of the ONE equation that describes the whole universe, uniting the two great theories of relativity and the quantum.  And he thinks that it points to the 'timeless bag' view.

Philosophically, Barbour's interpretation of quantum mechanics has much in common with Everett's Many Worlds interpretation, in that it gives positive ontological status to all possible configurations of the universe.  But whereas Everett thought in terms of infinitely branching histories, all of which are 'real', Barbour sees, no 'real' histories, since histories require time, the existence of which he denies.  No, the infinite possibilities are simply configurations - those wooden shapes loose in the bag, those individual points in Platonia.  Every possible configuration exists, and the quantum wave function for the universe determines the 'number of copies' of each (which is a timeless way of thinking about the frequency of occurrence for each).  And because there is no time dimension, all of these instants exist simultaneously (or more properly, eternally).

Reconciling with our experience of temporal becoming
But why does it seem to us that only a tiny percentage of these possible states has been or will be realised?  And why is it that they seem to us to be strung together by a causal history (notwithstanding apparent quantum indeterminacy)?  Finally, why do we FEEL like we are moving through time or that events are flowing from the future to the present to the past?

The answer starts with this: all of the most probable states (again, as determined by the universe's wave function) are what Barbour calls 'time capsules' in that they contain 'records' referencing an apparent past.  These records include fossils, books, empirical test results and artefacts.  More important, they include human brain configurations that contain memories.

Barbour reminds us that any conscious access to 'the past' is through CURRENT brainstates.  In any instant a brain state is nothing but a small bit of the universe's overall configuration and but one of the many records potentially contained in that instant.  Barbour must also intend (although I don't recall him saying it explicitly) that 'time capsules' are specifically those configurations that contain overwhelmingly (although not necessarily perfectly) CONSISTENT records.  Otherwise, if we lacked a significant degree of inter-subjective agreement, or if physical records pointed randomly in different directions, life would be very difficult and confusing.

Another thing Barbour doesn't (as far as I could make out) make clear is whether the configurations to which records refer are in themselves time capsules.  If not, then the configurations (and the records' referents) still EXIST, because all possible configurations exist, but the records are in a sense false, because the apparent past to which they point is not a likely one.  If the  configurations to which time capsules point ARE time capsules themselves, then there is a linked path of time capsules through Platonia, and our memories are 'true' in a sense they would not otherwise be.

As for our feeling of temporal becoming, Barbour puts that too down to brain states and the notion of the specious present, familiar to phenomenologists.  I wrote about the specious present (retention, protention and the immediate now) in The Anatomy of Experience, so I won't go into any detail here.  The bottom line is that each instantaneous brain state contains not only the current input from the world but also a record of the last brain state, and within that of the one before, etc.  This recursive encoding of records in our brain states accounts for our subjective experience of the flow of time, of temporal becoming.  But, says Barbour, just like longer-term memories, this feeling is actually wholly contained in each single instant, so long as that instant is one of the probable ones he calls time capsules.

Do I buy this?
Now, although I have a lay interest in physics and read a fair bit about it, I obviously am not at its cutting edge, so I can't comment on the technical aspects of Barbour's theory.  But here is what I think of it at the level at which I can engage.  I am on record elsewhere as arguing for realism and determinism in the interpretation of quantum dynamics.  Barbour's view is a probabilistic one and therefore not deterministic, so my instincts lean away from it.  But his equating our experienced states with those that are most probable (maybe by a vast margin most probable) goes some way toward alleviating my discomfort.  If time capsules are, by definition, the states 'awarded' highest probability by the wave function (as it overlays the structure of the configuration space, Platonia), then our specific history IS given a special, privileged position, one that arises directly from the structure of Platonia.  But I'm not quite sure whether Barbour is saying this.

My biggest questions are those alluded to three paragraphs above.  Are the many instants that are referenced in the records at instant T, time capsules, as instant T is?  Are the records in instant T-3 the same as those in instant T, except that they exclude reference to instants T-2 and T-1?  If the answers to these questions are 'yes' then I think Barbour's proposal could be a sensible one, and I'm willing to throw away the dimension of time.  My comfort arises from the fact that Barbour would then be assigning to the wave function (as influenced by the shape of Platonia) the ordinal role we normally assume to be performed by time.  From a practical perspective, the upshot would be the same as in the block view of time that I have endorsed - all that IS, exists eternally.  The STRUCTURE of the universe is set, and it is a peculiarity of our experience that we seem to flow through a time dimension.  In either case, the god's eye view is of an eternal structure (one in Platonia, the other in 4-dimensional space-time).  I'm not saying that there are no differences between the two - there are very big ones, especially for physics.  But at the level of philosophical implication, I think that they are quite close.

If the answer to either question above is 'no', then I have a hard time swallowing the theory.  (This of course doesn't mean the theory is wrong, and it needn't bother Barbour at all!.)  If the instants, the configurations, of existence are not related to one another consistently - whether via a neatly stacked vertical causation or via consistent reference among configurations jumbled loosely 'in a bag', then there might as well be (and may indeed be!) only one instant.  This is certainly philosophically possible, but it makes me want to go to bed and not get up again.  (But this is what the determinism that I expound does to many other people, so there you go!).

For a more coherent and/but much more technical analysis of The End of Time by a philosopher of science, see Jeremy Butterfield, of Oxford University.