4

I've figured out that if I know $G$ is not cyclic, then it for any $a \in G, o(a) \neq 4$ (or the order of any element in group $G$ is not 4).

I know ahead of time that the elements in the group ($\forall x \in G$) must have order $o(x)=k$ where $0 < k \leq 4$ where $k \in \mathbb{Z}$, so $k = 1, 2, 3, 4$.

If $G$ not cyclic, then we know $k \neq 4$ so we have $k = 1, 2, 3$ left.

I know how to show that $k \neq 3$ but I am not sure if I am supposed to do it for all possible $k$ because obviously if our order was higher (let's say $n$) then it would get messy. It ends up being in this case that $k = 2$ is good and $k = 1$ is trivial. I have looked up online and it says using Lagrange's Thm we know it has to be $k = 1, 2$ since $3 \nmid 4$, but I cannot use that theorem as we have not have learned it in class. How can I show that $k = 1$ or $2$ in another way?

However working from using the result that if we have $x \in G$ and $H = \langle a \rangle$ and knowing that $\left\vert{H}\right\vert=1$ or $2$, then $x^2=1$ for either case. Then I am done right?

What is another, cleaner, better way of trying to answer the question in the title?

I apologize if my formatting is poor since I am new at LaTeX and extraordinarily bad at algebra it seems. Thank you in advance.

  • Are you allowed to use the fact that $x^{|G|} = e$? – Jim Mar 06 '14 at 23:48
  • 2
    Why not just write out all the times tables for groups of order 4? – Emily Mar 06 '14 at 23:53
  • @Jim. I am not familiar with that notation from the book and my class but I assume you mean $x^4 = e$ right? Isn't that given knowing that the order of $G$ is 4? – user133636 Mar 06 '14 at 23:57
  • 1
    suppose $k = 3$. Then $(e,x,x^2)$ are different, then $G = {e,x,x^2, y}$. Now, what is the inverse of $y$? or $y^2$? – mookid Mar 06 '14 at 23:58
  • @Arkamis. I am not sure in how the times tables would differ. Depending on if $x^2 = e$ or not? A cyclic group's times table would result in a sudoku type result where each element appears once in every row and column since its order is 4, right? This is just a speculation off the top of my head. – user133636 Mar 07 '14 at 00:00
  • @user133636 Actually, every group's times table has that sort of sudoku structure. It's a very useful result that gets used in Lagrange's theorem, so you'll definitely come across it soon. – Gyu Eun Lee Mar 07 '14 at 00:02
  • @mookid. I already know how to show that $k \neq 3$. If $k = 3$, that means $a^3 = e$. But $G$ has order 4 meaning $a^4 = e$.

    Then I get $a^3a=ea$ and $a^4 = e$ which leads to $a=e$ which cannot happen.

    Is my reasoning correct?

    – user133636 Mar 07 '14 at 00:02
  • this is correct, but my justification was to avoid using the theorem $x^{|G| }= e$ ;) – mookid Mar 07 '14 at 00:04
  • @mookid. Ok thank you. What would be wrong about using the fact you stated? What is the reasoning? Is it not applicable to my specific problem or is it wrong, because as far I see the fact that $x^4 = e$ where 4 is the order of group G seems to be true. – user133636 Mar 07 '14 at 00:06
  • Your reasoning is true, but it is a quite sophisticated tool for the kind of problem you try to solve here. – mookid Mar 07 '14 at 00:09
  • @user133636: Yes, I mean $x^4 = e$ because $G$ has order $4$. If $x^4 = e$ because $G$ has order $4$ and $x^3 = e$ because $o(x) = 3$ then $x^4 = x^3$. Multiply by $x^{-1}$ three times and you get $x = e$, but $o(e) = 1$ so this contradicts our assumption that $o(x) = 3$. Thus $o(x) \neq 3$. – Jim Mar 07 '14 at 05:48
  • You can have a look at some answers in these posts: http://math.stackexchange.com/questions/165341/any-group-of-order-four-is-either-cyclic-or-isomorphic-to-v or http://math.stackexchange.com/questions/443642/prove-that-every-group-of-order-4-is-abelian (And you can probably find several other posts on this site which describe all groups of order 4.) – Martin Sleziak Nov 12 '15 at 06:50

3 Answers3

3

First of all, by definition the order of an element $x\in G$ is the smallest positive number $n$ such that $x^n=e$. So we might as well rule out $O(x)=0$. ($O(x)=0$ doesn't even make sense anyway...what is $x^0$ supposed to mean?)

If $O(x)=1$, this by definition means that $x^1 = x = e$. In this case $x^2=e$ is trivial.

If $O(x)=2$, then $x^2=e$; this is already what we want.

$O(x)$ cannot be $4$, for then $x$ generates a cyclic group of $4$ elements, but we know $G$ is not cyclic.

The last case is $O(x)=3$. You say you know how to show this cannot be the case; now you're done!

Gyu Eun Lee
  • 18,793
  • Thank you very much @kigen. I understand what I need to explicitly show now.

    On the top of my head, what if $G$ was of order $n$ instead of specifically 4. The $x^2=1$ only "works" for this specific case because $O(x) = 1$ or $2$ which results in $x^2 = e$ for both cases right? This should fail for a different order for $G$ right?

    – user133636 Mar 07 '14 at 00:05
  • That is absolutely correct. For a trivial example, think of the cyclic group of order $3$. More deeply, $x^2=e$ for all $x\in G$ implies that $G$ is a commutative group (can you prove this?), so the moment you see a noncommutative group this proposition fails. As it happens, the smallest noncommutative group is order $6$ (the symmetric group on $3$ letters), so order $6$ is where you'd find the first interesting counterexample. – Gyu Eun Lee Mar 07 '14 at 00:10
  • Thank you for the clarification. I appreciate your help. – user133636 Mar 07 '14 at 00:14
1

Here is another way to look at this question:

$G$ has $4$ elements; call the identity $e$ and the others $a,b,c$. In writing the multiplication table for $G$ (which is feasible since $|G|$ is small:), we must have

\begin{array}{|c|c|c|c|c|} \hline \star& e & a & b & c \\ \hline e &e &a &b & c\\ \hline a &a & &\\ \hline b &b & &\\ \hline c &c & & &\\\hline \end{array}

Now, each row and column of the multiplication table of a group must contain each element of the group exactly once (this is a good exercise to prove, using uniqueness of identity and so on). If $a^2=b^2=c^2=e$, then this is a valid multiplication table. If not, then there must be an element which is not its own inverse; WLOG assume $a\ne a^{-1}$. Then either $a^2=b$, which implies $ab=c$ and $ac = e$. In this case, $G=\{e,a,a^2,a^3\}$. Otherwise, $a^2=c$, which implies that $ba=e$ and $ca=b$. In this case again $G=\{e,a,a^2,a^3\}$. Quod erat demonstrandum.

(The purpose of this tedious approach is to motivate why we use theorems and do not attempt to enumerate algebraic objects by brute force ;)

Math1000
  • 36,983
1

This is essentially showing that if a group is of order 4 and not cyclic, it is the Klein 4 group isomorphic to $C_2 \times C_2$ - all groups of order 4 are isomorphic to one of these groups.

So to prove this without Lagrange's Theorem, we can suppose $o(x)=3$, else $o(x)=1$ if $o(x) \neq$, so x=e and this is trivial. So if $o(x)=3$, the group can be written as $G={e,x,x^2,y}$. But y has no inverse, so this is a contradiction. Therefore $o(x)=2$

  • I do not fully understand the Klein-4 group being isomorphic to what you said. We haven't learned isomorphism or even used the word Abelian yet in my textbook (but obviously we have learned commutativity). I loathe the book and my teacher but I always find better resources online that I try to read in relation to the class text + notes.

    The second part of your answer I understand and I am able to show that $o(x) \neq 3$. Thank you.

    – user133636 Mar 07 '14 at 00:10
  • The klein Group is $V = {e=(0,0),\alpha=(1,0),\beta=(1,1),\gamma=(0,1}$ where $\alpha^2 = \beta^2=\gamma^2 = e$ and $\alpha\gamma=\beta$ – mjb4 Mar 07 '14 at 00:18
  • We were exposed to what a Klein-4 group was in class in a kind of four square relationship where the identity was staying put on the same square, one relation was to move vertically, the other horizontally, and last one was diagonally, which is a combination of vertical + horizontal movement.

    I just don't immediately see how this question or the characteristic of having an order of 4 relates to Klein-4 right now. I am quite bad at algebra/group theory so it will take me a while for it to click in my head eventually.

    – user133636 Mar 07 '14 at 00:26
  • If it's not clicking right now, it's not your fault. Later on in your course, after isomorphisms are introduced, it may be proved that in fact there are only two "types" of groups of order $4$ - the cyclic group of order $4$, and the Klein $4$ group. If you know this then your original question is extremely simple. If you don't, you need to do this the hard way, the way you've already done it. – Gyu Eun Lee Mar 07 '14 at 00:34
  • Thank you for the clarification @kigen. I appreciate your tremendous help. – user133636 Mar 07 '14 at 00:36