Intereting Posts

Orthogonality and linear independence
Flows of $f$-related vector fields
How to arrange functions in increasing order of growth rate , providing f(n)=O(g(n))
A determinant coming out from the computation of a volume form
Is a finite commutative ring with no zero-divisors always equal to the ideal generated by any of its nonzero elements
Solving the differential equation $\frac{dy}{dx}=\frac{3x+4y+7}{x-2y-11}$
adjunction relation
About linear bijection between Banach spaces!
Infinite Groups with Finitely many Conjugacy Classes
Expression for the highest power of 2 dividing $3^a\left(2b-1\right)-1$
How to calculate the number of possible connected simple graphs with $n$ labelled vertices
Convergence of a sequence implies convergence of the norm of the sequence
Characteristic polynomial equals minimal polynomial iff $x, Mx, \ldots M^{n-1} x$ are linearly independent
Show that the Sorgenfrey line does not have a countable basis.
Proving the series of partial sums of $\sin (in)$ is bounded?

Let $G$ be a finite group with $|G|>2$. Prove that Aut($G$) contains at

least two elements.

We know that Aut($G$) contains the identity function $f: G \to G: x \mapsto x$.

If $G$ is non-abelian, look at $g : G \to G: x \mapsto gxg^{-1}$, for $g\neq e$. This is an inner automorphism unequal to the identity function, so we have at least two elements in Aut($G$).

- A semigroup $X$ is a group iff for every $g\in X$, $\exists! x\in X$ such that $gxg = g$
- On $GL_2(\mathbb F_3)$
- homomorphism $f: \mathbb{C}^* \rightarrow \mathbb{R}^*$ with multiplicative groups, prove that kernel of $f$ is infinite.
- Pathologies in “rng”
- Why can't the Polynomial Ring be a Field?
- free subgroups of $SL(2,\mathbb{R})$

Now assume $G$ is abelian. Then the only inner automorphism is the identity function. Now look at the mapping $\varphi: G \to G : x \mapsto x^{-1}$. This is an homomorphism because $\varphi (xy) = (xy)^{-1} = y^{-1} x^{-1} = x^{-1} y^{-1} = \varphi (x) \varphi (y)$. Here we use the fact that $G$ is abelian. This mapping is clearly bijective, and thus an automorphism.

This automorphism is unequal to the identity function only if there exists an element $x \in G$ such that $x \neq x^{-1}$. In other words, there must be an element of order greater than $2$.

Now assume $G$ is abelian and every non-identity element has order $2$. By Cauchy’s theorem we know that the group must have order $2^n$.

I got stuck at this point. I’ve looked at this other post, $|G|>2$ implies $G$ has non trivial automorphism, but I don’t know what they do in the last part (when they start talking about vector spaces). How should this prove be finished, without resorting to vector spaces if possible?

Thanks in advance

- Are $C(\mathbb{R})$ and $D(\mathbb{R})$ isomorphic or not?
- Existence of Homomorphism so that Diagram Commutes
- Centralizer of involutions in simple groups.
- In a principal ideal domain, prove that every non trivial prime ideal is a maximal ideal. What could be wrong in this approach?
- Isomorphic quotient groups $\frac{G}{H} \cong \frac{G}{K}$ imply $H \cong K$?
- Are there broad or powerful theorems of rings that do not involve the familiar numerical operations (+) and (*) in some fundamental way?
- Existence of subgroup of order six in $A_4$
- $A_4$ has no subgroup of order $6$?
- Computing the Smith Normal Form
- Elementary approach to proving that a group of order 9 is Abelian

The reason why that last case is done separately is because in that case we cannot describe a non-trivial automorphism with operations intrinsic to the group. In the non-abelian case we could use an inner automorphism, and in other abelian cases we can use negation. Using vector space structure is the “next best thing” in the sense that then everybody will immediately accept multiplication by an invertible matrix as an automorphism.

That last case is about groups that are isomorphic to a direct product of finitely many copies of $C_2$

$$

G=C_2\times C_2\times\cdots C_2.

$$

Any permutations of components is then obviously an automorphism. But proving that such a group has this direct product structure, while not too tricky, is something the answerers did not want to do, because it is much more convenient to observe that a finite abelian group where all the non-trivial elements have order two is a vector space over $GF(2)$.

Here’s another way of doing this step. We inductively find elements $x_1,x_2,\ldots,x_\ell\in G$ such that at all steps $x_k$ is not in the subgroup generated by $x_1,x_2,\ldots,x_{k-1}$. Because $G$ is finite, at some point we will have exhausted all of $G$, and thus $G=\langle x_1,x_2,\ldots,x_\ell\rangle$. The claim is then that

$$

G=\langle x_1\rangle\times\langle x_2\rangle\times\cdots\langle x_\ell\rangle

$$

is a way of writing $G$ as a direct product of subgroups of order two. This is more or less obvious by construction.

But you should notice that this is essentially the same argument proving that a finitely generated vector space over a field has a finite basis.

By structure theorem for finite abelian groups, we have $$G\cong(\Bbb Z_2)^n$$ for some $n$. Consider the function on $(\Bbb Z_2)^n$ that swaps the $1$st and $n$th coordinates. Show that that is an automorphism.

Building off of OP’s work, we now consider Abelian groups, where each non-identity element has order 2.

Let $\{g_1, g_2, \ldots g_n \}$ be any minimal generating set of the group. Since $|G| > 2 $, thus $n\geq 2$.

Each element of the group $g$ can be uniquely written as $\prod_{i \in S_g} g_i$. (Exponents are no needed, since each element has order 2.)

Then, consider the the map that replaces $g_1$ with $g_2$ and vice versa.

Show that this is an automorphism.

- Primes between $n$ and $2n$
- Prob. 9, Sec. 4.3 in Kreyszig's Functional Analysis Book: Proof of the Hahn Banach Theorem without Zorn's Lemma
- Find the remainder when a number is divided by 999 given the following conditions
- Derivation of Pythagorean Triple General Solution Starting Point:
- Fermi – Walker coordinate system
- Intuitive proofs that $\lim\limits_{n\to\infty}\left(1+\frac xn\right)^n=e^x$
- $5$-adic expansion of $−2$
- So-called Artin-Schreier Extension
- If $2$ divides $p^2$, how does it imply $2$ divides $p$?
- Prove that $PSL(2,\mathbb{Z})$ is free product of $C_2$ and $C_3$
- Modules with projective dimension $n$ have not vanishing $\mathrm{Ext}^n$
- Partial derivative notation
- Difference between Modification and Indistinguishable
- Approximation of integrable function by continuous in norm of $L_2$
- Finding a closed form expression for $\sum_{i=1}^{n-1}\csc{\frac{i\pi}{n}}$