A critique of "The End of Time", a book by Julian Barbour.

This is a book that claims to be about "The Next Revolution in Physics", a claim so bold, that it is very unlikely to be true. It is also a sign of a crackpot. The writer of the book, Julian Barbour, is a physicist by education, makes his living by translating from Russian, not as a physicist.

I am also a physicist, and I am certainly able to support myself from my education as a physicist. I am thus in a good position to criticize this book, which was given to me by Gry Gambert for that purpose.


First, the good things about this book:

I have never seen any other book explain the math behind modern physics as well as this book. If you want to get a feel, or impression of the mathematical framework of physics, this is the book for you. It may even clarify things before you start to read serious physics and math, and thus reduce the potential for confusion. It is very literate in its approach, but does contain some usable pictures as well.

Now the bad things:

First of all, Julian Barbours claim, is not a scientific theory, just a point of view, marked by excessive hubris.

What he claims, is that time does not exist, but is just an illusion.

Take a look on page 355, where Dowker points out that Barbours view is not a falsifiable theory, and thus not science, and Barbour admits he is right:
"...there is no way to make predictions about the results of our observations."
Falsification is the process of dropping theories that fail when tested. Theories that cannot fail are not scientific, and quite useless.

The main claim, that time does not exist, is simply a result of Barbour visualising a model of space-time, the Universe, the Multiverse, where he visualizes himself standing outside of it all, and then it looks as if time stands still, as if time is an illusion. But it will look like that no matter what the real nature of time is, so it is the lack of time that is an illusion.

To make an analogy: A video of a clock.
In the video, there is time, because it is a recording of a clock showing time. But when the video lies still on the table, it looks timeless. However, this does not change the fact that in the video, there is time, but of a different nature than the time outside of the video. There is recorded time, and there is real time, but both forms of time exists, just in different forms. Barbour does not understand this, and falsely thinks that time does not exist.


The word "time" represent 2 different physical phenomena, that points in mainly the same direction, but does not need to.

1. Direction of time.

We live in a 4 dimensional universe, with 3 dimensions of space, and 1 of time. However, these dimensions are not completely separable, as Einstein showed. Distance and time can to some degree be exchanged. However, by using Minkowskis methods, the directions of the universe can clearly be divided into spacelike directions, and timelike directions. This is no illusion, but something that can be measured pricesely with clocks and distance measurements.

This form of time can be measured with clocks.

2. Direction of increase of entropy

This is the effect that makes it possible to remember the past, but not the future, and forget the past, but not the future. It is this effect that makes a cold beverage warm, and a warm soup cold. It is this form of time that we directly experience when we remember things, and so for computers as well. This form of time is directly connected to the direction of computation, the direction it is possible to run computers in. It thus has deep roots in math as well, and is therefore the most fundamental aspect of time.

This form of time can be measured by forgetting, by thermometers, by computing, etc.

As long as these 2 forms of time can be measured, they exist.

Barbour makes a big mistake

Barbour makes a big mistake when it comes to quantum mechanics. He chooses the Schrödinger wave equation as basis for his models, page 231. The Schrödinger wave equation has 3 dimensions for each particle, and 1 dimension of time on top of that. This means that just 2 particles will give a model of 3+3+1=7 dimensions, which is an awful lot. This makes it impossible to calculate complex systems, as the number of dimensions get far too big, and cannout fit into a computer.

And then he proceeds to argue that in a model of thousand, or googoolplex dimensions, the single dimension of time can be removed.

However, there is one very serious flaw with this argument, and that is the fact that the Schrödinger wave equation is definitely not right. When used, it typically gives an accuracy of 0.01, and in some cases it is very inaccurate. It is simply wrong for a lot of cases. The correct equations to use, is Quantum Electro Dynamics, as presented by Richard Feynman, with its Feynman diagrams, built on the relativistic wave equations. Schrödingers wave equation is absolutely not relativistic, and is therefore wrong in this context.

Now, a really interesting thing about Q.E.D. is that each particle requires 4 dimensions, NOT 3. And which four dimensions are those? The 4 dimensions of space-time of course! This means that in a model of say 2 particles, there are 8 dimensions, of which 2 are time dimensions. In any model, 1/4 of the dimensions are time dimensions. This is exactly the opposite of what Barbour thinks: Instead of time being removed, time is immensely multiplied. One dimension of time each for all the particles in the universe. That is a LOT!!!!!!

I will now point out how time exists in his wrong examples of timelessness.

Page 1, picture "Snow storm"

This is a picture of a boat in a snow storm. It stands still, as all pictures do. It is thus also a picture of the time the boat was in the snow storm. It is only because we are outside the picture that we do not experience the snow storm or the time in the picture. If we were on that boat, we would have experienced both the snow storm and that time.

Page 21, Figure 1. A changing triangle.

The dimension of time is a diagonal on the page.

Page 29, Figure 2. 7 pictures mixed together.

Dimension and direction, from left to right, because of entropy of the mans movement.

Page 47, Figure 5. Picture of seed with parachute as analogy for histories in Hilbert/phase space.

If time is shaped as tendrils in some multidimensional space, it is still time, and it still exists.

Page 56, Figure 6. Simple diagram of Platonia

Again, if history is laid out as bifurcating lines in "Platonia"/Hilbert space, that does not remove time, but instead gives time a complex structure, along those lines.

Page 72,73, Figure 7,8. Just spaces of triangles

Does not contain time, or physics, just triangles.

Page 77, Figure 9. Path in shape space

Now, this is interesting, because he has put events into the space of triangles, events with a history, and thereby time. Time is along the curvy line, and starts at the upper left corner. This is possible to see because the entropy increases in this direction, mainly because C gets far away when the line reaches the bottom.

To rephrase: A planet getting thrown away by the gravity from 2 others is much more likely than a planet from far away getting captured instead of it just continuing to drift for away. That gives the direction of time

Page 78, Figure 10. Another possible path.

In this picture, time starts at the center of the triangle, because this is a low point of entropy. Again, time follows the line.

Page 80, Figure 11. Center of gravity.

This has nothing to do with time.

Page 82, Figure 12. Spaghetti strands.

These space-time paths has time in the z direction, starting at the bottom, because those equal sided triangles at the bottom have low entropy.

Page 84, Figure 13. 9 'spaghetti diagrams'

Same here. All 9 pictures has time as the upward direction, and earliest time at the bottom of the picture, again because of lower entropy there.

Page 85, Figure 14. Shape Space pictures of the same 9 systems.

In these pictures time is along the curvy lines, always starting at the middle of the triangle, as this is the point of lowest entropy. It is the same point as the bottom triangle in the previous pictures.

Page 95, figure 18. Galileos diagram.

Here, time is horisontal, as may be the direction of travel. The points e,d,c,I0,a are equally far apart in time. Direction is not apparant here, except it is most likely to start at I0.

Page 101, Figure 19. Taits clock.
Page 102, Figure 20. More Taits clocks.

This is a traveling clock, and direction of time is therefore same as direction of traveling.

Page 125, figure 22. Interference fringes.

Direction of time here is the same as the direction the waves travel, and it stops when they hit the wall, because they are observed at the wall, and observing is an irreversible process thermodynamically. Observing increases entropy, and therefore marks the direction of time.

Page 127, figure 24. Waves in pond.

An example of how it looks different for different spectators. A good starting point in explaining why the dimensions of time and space are related.

Page 132, 133, Figure 25,26. Space-time diagrams.

Illustrates even better how dimensions of time and space are related, and can partly be exchanged.

Page 148, 149, Figure 27,28. Light wave cones

These diagrams show very well how the directions of space and time dimensions are separated by the outside and inside of the light cone.

Page 174,175, Figure 29,30. Space/times

Examples of how space and time can partly be exchanged. But time still exists, as space does.

page 212,213 Figure 39,40. Phase space diagrams.

Snapshots of the phase space of 2 particles.

Page 214,215 Figure 41,42

Same, but after observation, an irreversible process, showing the direction of time.

Page 216, Figure 42. Entagled states.

Same, with entangled states.

Page 223, Figure 42. Entangling of states.

Entangling as a process requiring time.

page 321, Figure 55. History of universe.

Again, time is along a line.

You may email me if you have a comment.

Last modified: Sat Jun 25 14:17:50 CEST 2005