logo

Isotropic Subgroups 📂Abstract Algebra

Isotropic Subgroups

Definition 1

Let’s consider GG as a group, and define XX as a GG-set. For xXx \in X and gGg \in G, let Xg:={xX  gx=x}X_{g} := \left\{ x \in X \ | \ gx = x \right\} and Gx:={gG  gx=x}G_{x} := \left\{ g \in G \ | \ gx = x \right\}. We define the Isotropy Subgroup of GG with respect to xx as GxG_{x}.

Explanation

To get a grasp of what an isotropy subgroup is, one needs to understand the action of a group.

20180720\_170542.png

For the set of points and lines on the left side of the above figure X:={1,2,3,4,C,p1,p2,p3,p4,s1,s2,s3,s4,d1,d2,m1,m2} X : = \left\{ 1,2,3,4 , C, p_{1}, p_{2} , p_{3} , p_{4} , s_{1}, s_{2} , s_{3} , s_{4} , d_{1}, d_{2} , m_{1} , m_{2} \right\} consider the icosahedral group G=D4G = D_{4}. XX becomes a D4D_{4}-set. In simple terms, XgX_{g} is a set unaffected by gg, while GxG_{x} becomes a subgroup that cannot influence xx. Here, CC is the center of XX, and ρ0\rho_{0} is the identity of GG, which for every x,gx,g, satisfies CXgC \in X_{g} and ρ0Gx\rho_{0} \in G_{x}.

Let’s verify whether the examples below actually align with the figure above: Xρ0=XXρ1=CXμ1={C,p1,p3,s1,s3,m1,m2}G1={ρ0,δ2}Gs3={ρ0,μ1}Gd1={ρ0,ρ2,δ1,δ2} \begin{align*} X_{\rho_{0}} =& X \\ X_{\rho_{1}} =& C \\ X_{\mu_{1}} =& \left\{ C, p_{1}, p_{3} , s_{1} , s_{3} , m_{1} , m_{2} \right\} \\ G_{1} =& \left\{ \rho_{0} , \delta_{2} \right\} \\ G_{s_{3}} =& \left\{ \rho_{0} , \mu_{1} \right\} \\ G_{d_{1} } =& \left\{ \rho_{0} , \rho_{2} , \delta_{1} , \delta_{2} \right\} \end{align*} Here, when there exists gGg \in G satisfying gx1=x2g x_{1} = x_{2} for x1,x2Xx_{1} , x_{2} \in X, let’s define the equivalence relation x1x2x_{1} \sim x_{2}. If xx belongs to the cell of the partition by \sim, that cell is called the orbit of xx and is denoted as GxGx. Though it might look complicated, it simply means that XX is split into parts like {C},{1,2,3,4},{p1,p2,p3,p4},{s1,s2,s3,s4},{d1,d2},{m1,m2}\left\{ C \right\} , \left\{ 1,2,3,4 \right\} , \left\{ p_{1}, p_{2} , p_{3} , p_{4} \right\} , \left\{ s_{1}, s_{2} , s_{3} , s_{4} \right\} , \left\{ d_{1}, d_{2} \right\} , \left\{ m_{1} , m_{2} \right\}.

It’s important to note that GxG_{x} and GxGx are distinguished by whether they are subscripts or not. For example, G1G_{1} acts as an isotropy subgroup that cannot influence 11, which is G1={ρ0,δ2}G_{1} = \left\{ \rho_{0} , \delta_{2} \right\}, and G1G1 is the orbit of 11, which is G1={1,2,3,4}G1 = \left\{ 1,2,3,4 \right\}.

Generally speaking, consider the following theorem. It looks like it reminds us of Lagrange’s theorem, but a closer look at the notation reveals it’s somewhat unrelated.

Theorem

If XX is a GG-set, then Gx=(G:Gx)|Gx| = ( G : G_{x}) holds. If GG is a finite group, then Gx|Gx| is a divisor of G|G|.


Proof

Assume x1Gxx_{1} \in Gx, then there exists g1Gg_{1} \in G satisfying g1x=x1g_{1} x = x_{1}. When ψ:GxG/Gx \psi : Gx \to G / G_{x} is defined as ψ(x1):=g1Gx\psi (x_{1}) := g_{1} G_{x}, showing that ψ\psi is a bijection completes the proof.


Part 1. ψ\psi is a function.

Assuming g1x=x1g_{1} x = x_{1}, suppose there exists g1Gxg_{1} ' \in G_{x} satisfying g1x=x1g_{1}’ x = x_{1}. Then, g1x=g1xg_{1} x = g_{1}’ x and x=(g11g1)xx = (g_{1}^{-1} g_{1} ’ ) x, thus (g11g1)Gx(g_{1}^{-1} g_{1} ’ ) \in G_{x} must hold. Therefore, g1g1Gxg_{1} ' \in g_{1} G_{x}, and since g1Gx=g1Gxg_{1} G_{x} = g_{1}’ G_{x}, ψ\psi is a function.


Part 2. ψ\psi is injective.

If x1,x2Gxx_{1} , x_{2} \in Gx and assuming ψ(x1)=ψ(x2)\psi (x_{1} ) =\psi (x_{2} ) , then there exists g1,g2Gg_{1} , g_{2} \in G satisfying x1=g1xx_{1} =g_{1} x, x2=g2xx_{2} = g_{2} x, and g2g1Gxg_{2} \in g_{1} G_{x}. For some gGxg \in G_{x}, g2=g1gg_{2} = g_{1} g, and x2=g2x=g1(gx)=g1x=x1x_{2} = g_{2} x = g_{1} (g x) = g_{1} x = x_{1} holds. Since ψ(x1)=ψ(x2)    x1=x2\psi (x_{1} ) =\psi (x_{2} ) \implies x_{1} = x_{2}, ψ\psi is injective.


Part 3. ψ\psi is surjective.

For a coset of GxG_{x}, g1Gxg_{1} G_{x}, if g1x=x1g_{1} x = x_{1}, then g1Gx=ψ(x1)g_{1} G_{x} = \psi ( x_{1} ) holds. Since x1Gxx_{1} \in Gx is valid for all ψ(x1)\psi ( x_{1} ), ψ\psi is surjective.

Parts 1~3 demonstrated that ψ\psi is a bijection, thereby proving the following: Gx=G/Gx=(G:Gx) | Gx | = |G / G_{x}| = ( G : G_{x} )


Part 4.

If G<|G| < \infty, then G=Gx(G:Gx)=GxGx|G| = |G_{x} | (G : G_{x}) = |G_{x} | |Gx| holds, making Gx|Gx| a divisor of G|G|.


  1. Fraleigh. (2003). A first course in abstract algebra(7th Edition): p157. ↩︎