deepdream of
          a sidewalk
Show Question
Math and science::Algebra::Aluffi

Groups. Group action.

Consider a group and a target object, such as a set. A group actions associate each group element with an automorphism of a target object, such as a set.

Group action

Let G be a group with operation mG:G×GG, and let A and object. A function σ:GAuto(A) is a group action iff:

  1. the identity of G maps to the identity of A.
  2. either:
    1. for any g,hG, σ(mG(g,h))= σ(g)σ(h) [left-action]
    2. for any g,hG, σ(mG(g,h))= σ(h)σ(h) [right-action]

Why are there two types of actions?


Composition order

The need for two action types arises from the two ways of mapping the parameters that go into mG(_,_) to the "inputs" of function composition __.

Consider σ as a group homomorphism from (G,+) to a group (FA,@) where the elements of FA are automorphisms of A. The operation @ can take one of two forms:

  1. @ is standard function composition: f@g(x)=f(g(x))
  2. @ is the opposite of standard function composition: f@g(x)=g(f(x))

The 1st operation corresponds to a left-action and the 2nd operation to a right-action. If, for the 2nd operation, we write functions on the right of their arguments:

(x)f@g=((x)f)g

Then we can see how the terms left and right arise. This is covered more below.

Compact notation

Let g,hG and xA. For a left-action (second parameter goes first), typical notation will express the second requirement in the definition above as:

σ(mG(g,h))(x)=σ(g)σ(h)(x)

to:

(ah).x=a.(h.x)

The dot is often dropped to arrive at:

(ah)x=a(hx)

And because these two terms are equal, the brackets are often dropped.

The requirement for the right-action is similarly shortened to:

x.(ah)=(x.a).h

and

x(ah)=(xa)h

Comparing the two versions, we see that left-actions act like functions that are written on the left of their arguments, while right-actions are like functions that are written on the right of their arguments.

Datum of a group action

A datum of a group action can be considered a table with a single column:

Table for σ:GAuto(A)
eG iA
a ϕj
b ϕk
c ϕl
d ϕx
f ϕy
g ϕz
h
ϕw

Alternatively, if the data of the automorphisms is unwrapped, the datum can be considered a 2D table. For the case where the target object is a set, it might look like the table below.

Table for ρ:GA
x1 x2 x3 x4 x5 x6 x7 x8 x9 x10 x11 x12 ... xn
eG x1 x2 x3 x4 x5 x6 x7 x8 x9 x10 x11 x12 ... xn
a x17 x41 x12 x4 x10 x14 x2 x8 x21 x19 x15 x14 ... x?
b
c
d
f
g
h

Example

Target is a set

When the target is a set, A, the set of automorphisms is the set of permutations of the set—the symmetry group, SA (or is it an isomorphism to it?).