the shape of space
Apr. 20th, 2004 11:47 amJeff Weeks' MSRI-Evans talk yesterday was quite different than the usual MSRI talks "for grad students and Mathematicians" — much of the talk was illustrated by a fly-through ("first person shooter" for you silly gamers) of the universe given many different possibly topologies. It's easy to think of a spherical surface or a hyperbolic surface or a planar surface when you're talking about embedding a two dimensional surface in a three dimensional world — but when you start talking about three dimensional manifolds with various "shapes" (spherical, etc), things get a little weird.
For instance, if you look straight ahead on a spherical surface, you will see yourself from behind. But the weirder thing is that perspective is nonlinear — if you're standing at the pole, then things get smaller as you head towards the equator, but things further than the equator get bigger and bigger! Something directly opposite you on the sphere will occupy your whole field of view — in the background! The universe might really have such topology, but the reason we don't see ourselves in the distance has to do with the horizon, which is set by the speed of light (?).
But the key thing that was sort of an "ah-ha!" moment was when he made the connection between topologies of space and lattice symmetries. Actually, in the talk it was one of those things that was assumed to be obvious. And it is obvious, but I hadn't thought of it before! Say you live in a surface with toroidal symmetry — then you can meet yourself again if you translate a certain distance either horizontally or vertically. In other words, the toroidal space corresponds to a rectangular lattice! (But there's got to be a difference.. after all, you have curvature to worry about, as you "go around the donut.") Likewise, a klein bottle surface is like a lattice with one line of mirror symmetry — go off in one direction and you'll encounter a reversed copy of yourself, then keep going and you'll find an identical copy.
In the talk he also managed to explain the microwave background radiation and how it is used to determine Ω (which is related to the shape of the universe — "open," "flat," or "closed") in an understandable and accessible way. Anyway, it was an amusing talk — not particularly deep, but a good introduction to concepts I think will be quite useful to me.
Another awesome thing about Jeff Weeks is that he is a freelance mathematician. His web site includes the program that was rendering curved spaces.
For instance, if you look straight ahead on a spherical surface, you will see yourself from behind. But the weirder thing is that perspective is nonlinear — if you're standing at the pole, then things get smaller as you head towards the equator, but things further than the equator get bigger and bigger! Something directly opposite you on the sphere will occupy your whole field of view — in the background! The universe might really have such topology, but the reason we don't see ourselves in the distance has to do with the horizon, which is set by the speed of light (?).
But the key thing that was sort of an "ah-ha!" moment was when he made the connection between topologies of space and lattice symmetries. Actually, in the talk it was one of those things that was assumed to be obvious. And it is obvious, but I hadn't thought of it before! Say you live in a surface with toroidal symmetry — then you can meet yourself again if you translate a certain distance either horizontally or vertically. In other words, the toroidal space corresponds to a rectangular lattice! (But there's got to be a difference.. after all, you have curvature to worry about, as you "go around the donut.") Likewise, a klein bottle surface is like a lattice with one line of mirror symmetry — go off in one direction and you'll encounter a reversed copy of yourself, then keep going and you'll find an identical copy.
In the talk he also managed to explain the microwave background radiation and how it is used to determine Ω (which is related to the shape of the universe — "open," "flat," or "closed") in an understandable and accessible way. Anyway, it was an amusing talk — not particularly deep, but a good introduction to concepts I think will be quite useful to me.
Another awesome thing about Jeff Weeks is that he is a freelance mathematician. His web site includes the program that was rendering curved spaces.