- #1

- 91

- 0

You are using an out of date browser. It may not display this or other websites correctly.

You should upgrade or use an alternative browser.

You should upgrade or use an alternative browser.

- Thread starter sammycaps
- Start date

- #1

- 91

- 0

- #2

mathwonk

Science Advisor

Homework Helper

2020 Award

- 11,139

- 1,331

this isa fundamental construction in topology, the building up of fairly general spaces, like manifolds, from conjoining pieces which are simpler, like cells. Thus any discussion of cellular spaces, cellular homology, would discuss this, e.g. I presume the well known free algebraic topology book by A. Hatcher.

http://www.math.cornell.edu/~hatcher/AT/ATchapters.html

the appendix has a lot of properties of CW complexes, an important type of quotient or adjunction space.

if you want more elementary topics on those, a basic topology book might help, like kelley or maybe munkres?

here are some more free ones.

http://www.math.jhu.edu/~jmb/note/pushout.pdf

http://www.math.uchicago.edu/~may/CONCISE/ConciseRevised.pdf

here is maybe ,ore what you want: in introduction to topological manifolds by john lee:

pages 73-86? the problems start on page 81.

http://books.google.com/books?id=Me...epage&q=adjunction spaces in topology&f=false

by the way the point of the "push out" nonsense, is just that it is easy to define continuous maps OUT OF an adjunction space. I.e. if you have two maps, one on each piece X and B, and they agree where the two pieces are glued together i.e. along A in B, then they define one map out of the adjunction space (B joined to X along A).

http://www.math.cornell.edu/~hatcher/AT/ATchapters.html

the appendix has a lot of properties of CW complexes, an important type of quotient or adjunction space.

if you want more elementary topics on those, a basic topology book might help, like kelley or maybe munkres?

here are some more free ones.

http://www.math.jhu.edu/~jmb/note/pushout.pdf

http://www.math.uchicago.edu/~may/CONCISE/ConciseRevised.pdf

here is maybe ,ore what you want: in introduction to topological manifolds by john lee:

pages 73-86? the problems start on page 81.

http://books.google.com/books?id=Me...epage&q=adjunction spaces in topology&f=false

by the way the point of the "push out" nonsense, is just that it is easy to define continuous maps OUT OF an adjunction space. I.e. if you have two maps, one on each piece X and B, and they agree where the two pieces are glued together i.e. along A in B, then they define one map out of the adjunction space (B joined to X along A).

Last edited:

- #3

- 91

- 0

Great thanks! I don't much about cells yet, but it seems worthwhile to work ahead. The class will cover some algebraic topology, it just hasn't started yet.

I think my confusion stems from the fact that our professor didn't really explain why endowing a space (partitioned into subsets of X) with the quotient topology was particularly important, or I was too slow to get it. It seemed that even in defining adjunction spaces, it didn't come much into play except when he simply stated that we take the disjoint union of the two spaces we're trying to glue. I figured it had to do with the fact that the quotient topology made the quotient spaces homeomorphic to other spaces, and in fact, Munkres has a theorem about how a map g:X->Z and the set X*={inverses of the singletons in g} with the quotient topology induces a map f:X*->Z which is homeomorphic iff g is a quotient map. I'm not really sure though.

Is this the reason, or is there something more subtle I am missing?

I think my confusion stems from the fact that our professor didn't really explain why endowing a space (partitioned into subsets of X) with the quotient topology was particularly important, or I was too slow to get it. It seemed that even in defining adjunction spaces, it didn't come much into play except when he simply stated that we take the disjoint union of the two spaces we're trying to glue. I figured it had to do with the fact that the quotient topology made the quotient spaces homeomorphic to other spaces, and in fact, Munkres has a theorem about how a map g:X->Z and the set X*={inverses of the singletons in g} with the quotient topology induces a map f:X*->Z which is homeomorphic iff g is a quotient map. I'm not really sure though.

Is this the reason, or is there something more subtle I am missing?

Last edited:

- #4

mathwonk

Science Advisor

Homework Helper

2020 Award

- 11,139

- 1,331

i.e. an interval is simpler than a circle but a circle i q quotient o an interval.

'

'a torus is moire complicated than a rectangle but a torus is a quotient of a rectangle,....

almost any space is a successive union of quotients of rectangles of various dimensions.

- #5

- 91

- 0

i.e. an interval is simpler than a circle but a circle i q quotient o an interval.

'

'a torus is moire complicated than a rectangle but a torus is a quotient of a rectangle,....

almost any space is a successive union of quotients of rectangles of various dimensions.

I understand that, I'm just a bit unsure of why the quotient topology is the "right" one. Is it just because then these spaces are homeomorphic to each other (like you said, identifying the endpoints of an interval in the quotient space makes it homeomorphic to the circle in the quotient topology)? Or is there something deeper and more subtle I am missing.

Like, when I was learning about box vs. product topology in arbitrary products, there were all nice properties about the product topology that the box topology didn't have.

- #6

mathwonk

Science Advisor

Homework Helper

2020 Award

- 11,139

- 1,331

as i said before (push out), the idea is to define a topology so that a function on the numerator space that is constant on equivalent points, is continuous on the quotient space. think about it.

Share: