An organization blog on math, physics and philosophy

Skip towards the Primary Content

Note: These pages make extensive utilisation of the latest XHTML and CSS Standards. They need to look wonderful in almost any standards-compliant modern browser. Regrettably, they’ll most likely look horrible in older browsers, like Netscape 4.x and IE 4.x. Furthermore, many posts use MathML, that is, presently only supported in Mozilla. My favorite suggestion (and you’ll thank me when surfing a constantly-growing quantity of sites on the internet that have been crafted to make use of the brand new standards) would be to upgrade towards the new edition of the browser. If that is difficult, consider relocating to the factors-compliant and open-source Mozilla browser.

September 11, 2006

Wirth and Stasheff on Homotopy Transition Cocycles

In 1965, James Wirth authored a PhD thesis around the description of fibrations

(1) p. E &#x2192 B p. E to B

when it comes to transition data

(2) p &#x2212 1 ( U &#x3b1 ) &#x2192 t &#x3b1 &#x2190 t &#xaf &#x3b1 U &#x3b1 &#xd7 F p &#x2193 &#x2193 p U &#x3b1 = U &#x3b1. array< p^(U_\alpha) \array \\ \stackrel> U_alpha times F p downarrow downarrow p U_alpha = U_alpha > .

Here U &#x3b1 U_alpha are aspects of a great covering of B B by open sets, F F may be the typical fiber. t &#x3b1 t_alpha is really a selected trivialization from the fibration over U &#x3b1 U_alpha. and t &#xaf &#x3b1 bar t_alpha its inverse, as much as homotopy.

A brand new arXiv entry now recalls the primary concept of this old operate in modern language:

James Wirth Jim Stasheff
Homotopy Transition Cocycles
math.AT/0609220 .

The problem looks nearly the same as that familiar in the local trivialization of principal fiber bundles. The important generalization, though, is incorporated in the final clause, which asserts that t &#x3b1 t_alpha and t &#xaf &#x3b1 bar t_alpha are just weak inverses of one another.

Because of that, the familiar cocycle equation

for that transition functions

(2) g &#x3b1 &#x3b2 = t &#xaf &#x3b1 t &#x3b2 g_ = bar t_alpha t_beta

will generally only support homotopy

If these f &#x22ef f_ satisfy a formula postulating a kind of associativity, we obtain a greater cocycle equation as known from gerbes. But, still more generally, this f &#x22ef f_cdots itself may be associative only as much as homotopy, and so forth. This yields cocycles from the form because they corespond to two-gerbes, 3-gerbes, etc.

You can nicely summary all of this information when it comes to things i would call a pseudofunctor

(4) U &#x2192 C. mathbf to C .

where U U is exactly what I understand because the &#x10cech groupoid from the good covering &#x2294 &#x3b1 U &#x3b1 sqcup_alpha U_alpha where C C is a few n n -category. Within the topological world this really is known as a functor as much as strong homotopy.

(For many images of how such functors seem like I’m able to point for example for this which. The overall concept of realizing general cocycles as functors from simplices to particular codomains seemed to be formulated by John E. Roberts.)

Wirth labored out how much fibrations are equal to their assortment of transition data (&#x201cdescent data&#x201d).

James F. Wirth
Fiber spaces and also the greater homotopy cocycle relations
PhD thesis, College of Notre Dame, 1965

The tough part would be to rebuild the fibration E E in the transitions between its local trivializations.

For any fiber bundle, we simply go ahead and take space &#x2294 &#x3b1 U &#x3b1 &#xd7 F sqcup_alpha U_alpha times F and identify points that are related under g &#x3b1 &#x3b2 g_. Clearly, for general fibrations this construction is much more involved, because the g &#x3b1 &#x3b2 g_ are not even close to inducing an equivalence relation.

The important tool to make progress may be the mapping cylinder theorem that was mentioned and proven in

James F. Wirth
The mapping cylinder axiom for WCHP fibrations
Pac. J. Math. 54 (2):263-279, 1974 .

I thank Jim Stasheff to make me conscious of the work within the comment portion of this entry. I would possess a comment along with a question, however i will publish those to the comment portion of this entry here.

Published at September 11, 2006 7:15 PM UTC

TrackBack URL with this Entry:&#xa0&#xa0 golem.ph.utexas.edu/cgi-bin/MT-3./dxy-tb.fcgi/933

Some Related Records

1 Comment 3 Trackbacks

Re: Wirth and Stasheff on Homotopy Transition Cocycles

This is the announced comment, in other words some hybrid from a comment along with a question:

John Baez and Candice Bartels have investigated the thought of studying such issues when it comes to n n -bundles .

An n n -bundle is really a (topological or smooth, based on the application) n n -category E E plus a appropriate functor

(1) p. E &#x2192 B. p. E to B .

where B B is really a discrete n n -category, the bottom space.

When it comes to such n n -bundles there are other of less apparent categorifications of most of the structures familar from ordinary gauge theory, like principal bundles, vector bundles, connections on bundles, etc.

Just like within the theory of homotopy transition cocycles, it’s relatively simple to pass through from the given n n -bundle structure E &#x2192 B E to B towards the corresponding cocycle data of transitions between local trivializations.

What is not apparent may be the inverse of the construction. I’m not conscious that anybody has attempted to deal with this poor n n -bundles. But possibly a number of Wirth’s work might be applied here.

Personally, I had been already puzzled for a long time in what ought to be the simplest nontrivial example.

Given a Deligne 3-cocycle on the base space B B. i.e. the cocycle describing a U ( 1 ) U(1) -gerbe with connection and curving on B B. I needed to understand which total 2-space E &#x2192 B E to B with connection

(2) tra. P 2 ( B ) &#x2192 Trans ( B ) mathrm. P_2(B) to mathrm(B)

has local trivialization data that brings about the required Deligne 3-cocycle.

My method of this was most likely hopelessly unsophisticated when compared with Wirth’s tools. However the solution which i finally created a minimum of includes a form that meets rather nicely the objective of this type of description within the context where I needed to put it on, namely the &#8220physics&#8221 of strings in Kalb-Ramond backgrounds.

The answer which i created is described here.

Roughly, the concept is that this:

Towards the U ( 1 ) U(1) -gerbe we might affiliate a PU ( H ) PU(H) principal bundle, and also to that people may affiliate a vector bundle whose fibers are algebras A x A_x of compact operators on some Hilbert space.

Then in the connection and curving data we discover a transport 2-functor, which transmits points x x to algebras A x A_x. pathways x &#x2192 &#x3b3 y x stackrel y to some x A_x – A y A_y bimdodules and cobordisms between &#8220parallel&#8221 pathways to bimodule homomorphisms.

Now, this yields a complete space which isn’t a category, but simply a collection (with extra structure).

However the 2-category (= bicategory, I usually say n n -category for that weakest possible situation) of bimodules naturally sits within the 2-category Mod Vect mathrm_mathrm of module groups for that monoidal category Vect mathrm .

(3) Bim &#x2192 Mod Vect mathrm to mathrm_mathrm

an algebra A x A_x is distributed towards the group of created a x A_x -modules (that is itself a module category for Vect mathrm. by tensoring in the right ), a bimodule is distributed to some functor between groups of left modules, and so forth.

By doing this we might consider all of our fibers A x A_x as really denoting a category, namely A x Mod <>_mathrm .

This perspective is great since it manifestly realizes a line bundle gerbe like a 2-vector bundle meaning that objects of Mod Vect mathrm_mathrm can be handled as 2-vector spaces # .

However, it’s not quite as nice, since the total &#82202-space&#8221

(4) E. = &#x2294 x A x Mod E := sqcup_x <>_mathrm

isn’t a topological (significantly less an even) category. A minimum of not without further work and additional assumptions.

Which means this modest observation of mine may be the comment I’ve. The issue obviously is that if maybe Wirth’s result might be useful for understanding such problems with 2-bundles.

Browse the publish Fahrenberg and Raussen on Continuous Pathways
Website: The n-Category Caf&#xE9
Excerpt: A theory of continuous n-pathways modulo continuous reparameterization.
Tracked: September 26, 2006 9:07 PM

Browse the publish On n-Transport: Universal Transition
Website: The n-Category Caf&#xE9
Excerpt: Pathways in groups from universal transitions.
Tracked: October 6, 2006 5:11 PM

Browse the publish Bakovic on 2-Groupoid 2-Torsors
Website: The n-Category Caf&#xE9
Excerpt: Igor Bakovic defines and classifies categorified groupoid bundles.
Tracked: October 31, 2006 10:35 PM

Access Keys:

Ease of access Statement 1 Primary Page 2 Skip to Content 3 Listing of Posts 4 Search p Previous (individual/monthly archive page) n Next (individual/monthly archive page)