Order of Join and Intersection of Subgroups

In summary: Your solution looks good. The key result you used, [ST : S] = [T : (S \cap T)] is a standard one which most books either prove or assign as an exercise (so at least the statement appears in the book). I guess having that result makes your problem quite easy, so perhaps MacLane and Birkhoff expected you to discover it for yourself, or perhaps to solve the problem in some other way. Are you using their elementary text (A Survey of Modern Algebra) or the more advanced one (Algebra)? In summary, Lagrange's theorem states that for some subgroup S of a finite group G, there are the same number of cosets of B in A as there are
  • #1
wnorman27
14
0

Homework Statement


If S and T are subgroups of a finite group G, prove that
[S:1][T:1] ≤ [S[itex]\cap[/itex]T:1][S[itex]\vee[/itex]T:1]

Homework Equations


Notation: [A:B] is the number of cosets of B in A for some subgroup B of A.
note that [A:1] is the order of A.

Lagrange's Theorem:
For some subgroup S of G.
[G:1]=[G:S][S:1]

The Attempt at a Solution


[S:1] ≤[S[itex]\vee[/itex]T:1]
and many other statements like this... none of them seem to help much.
 
Physics news on Phys.org
  • #2
What can you say about the order of the set

[tex]ST = \{st | s \in S, t \in T\}[/tex]?

Can you relate this to the orders that appear in the desired inequality?
 
  • #3
|ST| = Sum of orders of distinct cosets of S of the form St for some element t.

For two cosets St and St', we have that St=St' iff St*t'[itex]^{-1}[/itex]=S, which means [itex]t*t'^{-1}\in S[/itex]. But [itex]t*t'^{-1}\in T[/itex] also, so we must have [itex]t*t'^{-1}\in (S\cap T)[/itex]. Then [itex](S\cap T)t=(S\cap T)t'[/itex] so that there are the same number of cosets St in ST ([ST:S]) as there are cosets [itex](S\cap T)t\ of\ (S\cap T)\ in\ T \ \ [T:(S\cap T)][/itex]. Applying Lagrange's theorem twice:

[itex][ST:S][/itex]=[itex][T:(S\cap T)][/itex] implies:

[itex]\frac{|ST|}{|S|}=\frac{|T|}{|S\cap T|}[/itex], so that

[itex]|ST|=\frac{|S||T|}{|S\cap T|}[/itex] and since clearly [itex]ST\subset(S\vee T)[/itex]

[itex]\frac{|S||T|}{|S\cap T|}=|ST|≤|S\vee T|[/itex], so

[itex]|S||T|≤|S\vee T||S\cap T|[/itex]

I got the proof that [itex][ST:S][/itex]=[itex][T:(S\cap T)][/itex] from Dummit and Foote - this is not in the book I'm working through (MacLane and Birkhoff), which is where this problem is from. Is this supposed to be obvious? I don't know how I would have ever done this problem without just happening upon this result. Thanks!
 
  • #4
Your solution looks good. The key result you used, [itex][ST : S] = [T : (S \cap T)][/itex] is a standard one which most books either prove or assign as an exercise (so at least the statement appears in the book). I guess having that result makes your problem quite easy, so perhaps MacLane and Birkhoff expected you to discover it for yourself, or perhaps to solve the problem in some other way. Are you using their elementary text (A Survey of Modern Algebra) or the more advanced one (Algebra)?

Do they prove the 2nd ismorphism theorem? Namely, if N is a normal subgroup and H is any subgroup, then [itex]NH/N[/itex] is isomorphic to [itex]H/(N \cap N)[/itex]. This implies the order result you used in the case where one subgroup is normal. From this, it's not a major stretch to conjecture that the order result might be true even if neither subgroup is normal.
 
  • #5
Hi Jbunniii,

Thanks for all your help! I'm using Algebra (the advanced one) - as my primary text as I work through abstract algebra for the first time, but I've been reading parts of a few other algebra textbooks (Dummit+Foote, Fraleigh, Artin, and some course notes from a prof at uc berkeley). I'm studying the subject on my own, which is proving to be a bit insane, though this forum helps.

They don't prove any isomorphism theorems until 300 pages after this problem (in a chapter called "The Structure of Groups"). Due to your hint about ST, I was reminded that I had seen reference to such a set in D+F, which turned out to be right next to the section on the isomorphism theorems. In MacLane and Birkhoff, the definitions of kernel, image, and normal subgroup don't even appear until the next section!

The only thing I can think of is that they expect you to contemplate the nature of |S∨T| until you get the idea to consider |ST:S|?

Thanks again for your help, I feel I'd have been stuck on this forever.
 

FAQ: Order of Join and Intersection of Subgroups

What is the order of join of subgroups?

The order of join of subgroups is the number of elements in the subgroup formed by taking the union of two subgroups. It is denoted by the symbol $\displaystyle |H\lor K|$, where $H$ and $K$ are subgroups of a group $G$.

How is the order of join of subgroups related to the orders of the individual subgroups?

The order of join of subgroups is related to the orders of the individual subgroups by the formula $\displaystyle |H\lor K| = \frac{|H||K|}{|H\cap K|}$, where $|H|$ and $|K|$ are the orders of subgroups $H$ and $K$ respectively, and $|H\cap K|$ is the order of their intersection.

What is the intersection of subgroups?

The intersection of subgroups is the set of elements that are common to both subgroups. In other words, it is the subgroup formed by taking the intersection of two subgroups $H$ and $K$, denoted by the symbol $\displaystyle H\cap K$.

How does the intersection of subgroups affect the order of the join of subgroups?

The order of the join of subgroups is affected by the intersection of subgroups in the formula $\displaystyle |H\lor K| = \frac{|H||K|}{|H\cap K|}$. If the intersection of subgroups is non-trivial (i.e. has more than just the identity element), then the order of the join will be smaller than the product of the individual subgroup orders.

Can the order of join of subgroups be larger than the order of the group?

Yes, the order of join of subgroups can be larger than the order of the group. This can happen when the subgroups are not normal in the larger group, as the order of the join will then include all the elements of the larger group that are not in the subgroups. However, if the subgroups are normal, then the order of join will always be smaller or equal to the order of the group.

Similar threads

Back
Top