The discussion at Bee's blog has broken the 200 comment mark.
My last comment there was #200, Aaron Bergman's was #201.
It's like a loud party.
http://backreaction.blogspot.com/2007/11/theoretically-simple-exception-of.html
the person who is really missing in that discussion is
John Baez
very good physics ideas can bring about the invention or recognition of new mathematics
(more often the recognition of already invented, and realization of its role in understanding nature)
G.L. E8 ToE in my humble opinion will eventually require the recognition of a new type of spacetime manifold and a new type of connection.
It could be a spacetime manifold that is locally deSitter instead of locally Lorentz.
Or where the local geometry is graded (i.e. energyscale dependent)
I guess in the best and free-est discussion people should be free to speak carelessly off-cuff and not suffer the chilling effect of being quoted, but I very much liked an off-the-cuff exchanged between G.L and Bee, and want to quote the essentials.
Garrett said the E8 theory under construction was neither top down or bottom up but, instead, might be described as
"top-down inspired, bottom-up". My punctuation.
that is it is being built up by hand from the ground of the standard models----to match GR and particle SM---but there is an overriding mathematical idea that inspires it.
Bee said "what is the top that the inspiration comes down from?" or words to that effect. It is a really good and persistent question and it points to where mathematical creativity could play a role.
I think the idea of naturalness at the top---or which is inspiring the construction of the theory---is that geometry and matter are the same thing and should be described by the same mathematical object.
however classical geometry dynamics (GR) the geomtry was described by the metric, the distancefunction played the role of geometry.
Garrett pointed out at the seminar that a CONNECTION is just as good a way to represent the geometry and in some ways more NATURAL. he mentioned that one can recover a metric from a connection and a connection is a more elegant or economical way---it describes the spacetime manifold's shape by how things roll and twist as you truck them around on it. With a metric you have to
figure out how to do transport, by studying distances. But the connection just tells you how, with less fuss and bother. That's its job.
So a connection is an inertial compass trucking dingus that covers the metric's job and the bonus is that it gives a natural way to describe FIELDS and their allowed interactions.
so the overriding math idea (from whence the inspiration for the bottom-up buiding work) is that geometry and matter are the same thing so let's try to describe them both with a connection dingus, and get a classical and eventually quantum dynamics of geometry and matter in terms of that.
And there is the question of WHAT KIND OF 4D SPACETIME MANIFOLD it should be built on (because there are various definitions of manifold available in differential geometry, and of course one can invent new ones) and then WHAT KIND OF CONNECTION on what kind of bundle. A bundle is where you plant a copy of E8 at each point of the manifold and then talk about connecting them up. E8 is the egg of the universe, it is what defines our world of interacting matter and geometry, so naturally you want a copy of it at each point because that describes each point of our world. The nontrivial part is connecting.
These are just my inexpert reactions as a spectator. What I am anticipating is that a real mathematician will show up and say something like----hmmm Garrett's E8 doesn't have Lorentz flat symmetry in it, it has deSitter, so we have to do something about the underlying manifold. It might have a curved tangent space. And also it looks like Lambda is energyscale-dependent, so the manifold might be scale-graded in some sense. It might need to be able to have a dimensionality that varies with scale---so that it becomes fractal-like and lower dimensiony at very small scale...
We have this odd thing that in nature space expands----but the flat Minkowski space of special relativity doesn't. To be fundamental it seems intuitive that a theory could not be built on a manifold that is locally Minkowski. More likely one that is locally deSitter...
but these are my hunches and they don't matter, I just want to indicate some of the room outside the box. If it turns out that the E8 theory has the potential to GROW mathematics, like feet that require a new size of shoes, so as not to mis-shapen their ToEs.