Exercise 26a.

A semigroup is a set with a product which associates to each ordered pair of elements of an element in such a way that for any elements . Show that the set of all mappings from a given set to itself forms a semigroup in which the product is composition of mappings. Show that the set of all one-to-one correspondences of with itself forms a group under composition.


Proof

We have two things to show:

  1. Show that the set of all mappings from a given set to itself forms a semigroup in which the product is composition of mappings.
  2. Show that the set of all one-to-one correspondences of with itself forms a group under composition.

Proof of 1.

Let be the set of all functions . We need to show that for all We know this is true because composition of function is associative, and we know we can compose the functions since share one the same domain and codomain, namely , since .

Proof of 2.

One-to-one correspondence (not to be confused with the one-to-one function, which is synonymous with an injective mapping), is a synonym for a bijection, that is, a function that is both surjective and injective.

Let be the set of all bijective functions (bijections) . We need to show that this set is a group under composition, that is, that the group axioms are true in this set, when we equip it with the composition operation.

  1. Closure: We know from set theory that a composition of bijections is a bijection, therefore
  2. Associativity: Since , is a subgroup as well and thus the associativity axiom is inherited.
  3. Existence of the identity element: The identity function, defined with is a bijection, and we know from set theory that is true for all functions, and particularly for bijections.
  4. Existence of an inverse for all elements of : Every bijective function has an inverse function, that is a bijection as well, and its inverse is the function of which it is an inverse. Formally: , which is what we needed to show.1

The proof is complete.

\(\square\)

Additional thoughts

This exercise made me think of Category Theory, and its connections with algebra and groups in particular. Let me explain some of these connections.

In this exercise we get introduced to the "semigroup", which is an algebraic structure "weaker" than the group, meaning that we are necessitating less2 rules which it has to obey, i.e. we don't care about identity elements, or inverse elements in a semigroup. So, what if we wanted to define an algebraic structure that's a little stronger than the semigroup, and add, say, the identity element axiom to its definition? This would give us the monoid, which is essentially a semigroup + identity element axiom. So if we order these structures by their "strength", we get the following ordering: semigroupmonoidgroup (or, every group is a monoid and a semigroup, every monoid is a semigroup, see image below). Here is where the connection with category theory comes in: A category with just one object is a monoid. A category with just one object, but where all the morphisms are invertible is a group. Zounds!

Here is a MSE thread with some nice answers that explains the connection with groups in particular, towards which this exercise hints at, if you are familiar with the definition of a category.

Alt text

image from wikipedia


  1. I'll often use results from set theory like this, I took a set theory course last year and I want to focus on algebra, because of that I won't be descending to the set theory realm, and will use those results as given. Additionally, since there is a Set Theory chapter in the book, with many of these results (most probably) proven, I think this is alright. 

  2. I say that "weaker" means "... less", but its tacitly meant that the set of rules for the weaker notion (semigroup) is a proper subset of the rules for the stronger notion (group). this is nicely illustrated by the attached image.