I have two simple questions about the category theory.

  1. In any category, is $Hom(A, B)$ always nonempty? In some typical categories, it seems right but the definition of morphism does not give any information.

  2. In the category $\textbf{Set}$ of all sets, what is in $Hom(\varnothing, A)$ and $Hom(A, \varnothing)$?

Please anybody help me. Thanks in advance.

  • $\begingroup$ It can be empty, and the a relation on some empty set is empty, so there is an empty function. $\endgroup$ – user40276 Feb 2 '14 at 16:29
  • $\begingroup$ $\hom_{\mathsf{Ring}}(\mathbb{Q},\mathbb{Z})=\emptyset$. $\endgroup$ – Martin Brandenburg Feb 2 '14 at 21:19

Hom(A,B) can be empty. For example, the category with only two objects and only identity morphisms is a perfectly valid category. For a more "natural" example, the path groupoid of a space with two different path components will have no morphisms from any object in one component to the other component. (Actually the first example is a specialization of the second to the discrete space with two points)

In Set, Hom($\emptyset,A$) consist of a single morphism for each $A$, the empty function. $\emptyset$ is the initial object in the category of sets.

Hom($A,\emptyset$) is always empty unless $A = \emptyset$, in which case it has only the identity morphism.

  • 4
    $\begingroup$ A function $f:A \to B$ is by definition a subset of $f \subset A\times B$ such that for all $a \in A$ there is a unique $b \in B$ so that $(a,b) \in f$. There can be no function from $A \to \emptyset$ if $A$ is non empty, because there are no $b \in \emptyset$ to pair with a given $a \in A$. However, there is a function from $\emptyset \to A$ because the universal quantifier is vacuously satisfied: there are no $a \in \emptyset$ to place any conditions on the relation. $\endgroup$ – Steven Gubkin Feb 2 '14 at 17:56
  • 2
    $\begingroup$ Codomains and domains aren't symmetric in this way. Every value of the empty function is in A vacuously; that's enough for it to be the codomain. But a function must map every element in its domain somewhere. $\endgroup$ – Malice Vidrine Feb 2 '14 at 17:58
  • 1
    $\begingroup$ $A\times\emptyset =\emptyset$. $\endgroup$ – Malice Vidrine Feb 2 '14 at 18:11
  • 1
    $\begingroup$ Also, in any Cartesian closed category $A\times 0\cong 0$, so it's not just a quirk of $\mathbf{Set}$ either. $\endgroup$ – Malice Vidrine Feb 2 '14 at 18:19
  • 2
    $\begingroup$ @user40276 Perhaps this will help convince you. It is true for all finite sets that if $|A|=n$ and $|B|=m$ then $|A\times B|=nm$ where $n,m\in\mathbb{N}$. It follows that $|A\times\emptyset|=0$ for all finite sets $A$ and we know that the only set with zero elements is the empty set. Hence $A\times\emptyset=\emptyset$. $\endgroup$ – Dan Rust Feb 2 '14 at 18:48

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.