Exercise 14.3.21

Generalize (14.15) by showing that we have inclusions

𝔽 p mn = 𝔽 p m n AGL ( n , 𝔽 p m ) AΓL ( n , 𝔽 p m ) AGL ( nm , 𝔽 p ) S p nm .

Answers

Proof. We write q = p m . Let B = ( e 1 , , e m ) be a base of 𝔽 q over 𝔽 p . As 𝔽 p -vector spaces, 𝔽 q n and 𝔽 p nm are isomorphic, where an isomorphism φ is given by

φ { 𝔽 q n 𝔽 p nm ( α 1 , , α n ) ( x 1 1 , , x m 1 ; ; x 1 n , , x m n ) ,

where α i = j = 1 m x j i e j , i = 1 , , n . (this isomorphism depends of the choice of the base B ). So we can write

𝔽 p mn = 𝔽 p m n .

𝔽 p m n = 𝔽 q n is identified with { γ I n , v v 𝔽 q n } AGL ( n , 𝔽 q ) , via the isomorphism

{ 𝔽 q n { γ I n , v v 𝔽 q n } v γ I n , v .

This shows that

𝔽 p m n AGL ( n , 𝔽 p m ) .

If f = γ A , v is any element of AGL ( n , 𝔽 q ) , then f = γ A , e , v AΓL ( n , 𝔽 q ) (where e is the identity is S q ). Therefore

AGL ( n , 𝔽 p m ) AΓL ( n , 𝔽 p m ) .

By Exercise 3(c), we know that elements of AΓL ( n , 𝔽 q ) give maps 𝔽 q n = 𝔽 p nm 𝔽 q n = 𝔽 p nm that are affine linear over 𝔽 p , so that these elements are in AGL ( nm , 𝔽 p ) .

AΓL ( n , 𝔽 p m ) AGL ( nm , 𝔽 p ) .

The elements of AGL ( nm , 𝔽 p ) are bijective maps from 𝔽 p nm to 𝔽 p nm , thus AGL ( nm , 𝔽 p ) S ( 𝔽 p nm ) .

Via an arbitrary numbering of 𝔽 p nm , say γ : { 1 , , p nm } 𝔽 p nm , we obtain the isomorphism ψ : S ( 𝔽 p nm ) S nm defined by ψ ( σ ) = γ 1 σ γ , which allows us to identify S ( 𝔽 p nm ) S nm .

AGL ( nm , 𝔽 p ) S p nm .

To conclude, using several improper non-canonical identifications, we write

𝔽 p mn = 𝔽 p m n AGL ( n , 𝔽 p m ) AΓL ( n , 𝔽 p m ) AGL ( nm , 𝔽 p ) S p nm .

User profile picture
2022-07-19 00:00
Comments