This page is about the general concept of embeddings in category theory. For the special cases see embeddings of topological spaces, of smooth manifolds, or of types. For the notion in model theory see instead at elementary embedding.
homotopy hypothesis-theorem
delooping hypothesis-theorem
stabilization hypothesis-theorem
An embedding is a morphism which, in some sense, is an isomorphism onto its image.
For this to make sense in a given category , we not only need a good notion of images. Note that it is not enough to have the image of as a subobject of ; we also need to be able to interpret as a morphism from to , because it is this morphism that we are asking to be an isomorphism.
One general abstract way to define an embedding morphism is to say that this is equivalently a regular monomorphism.
If the ambient category has finite limits and finite colimits, then this is equivalently an effective monomorphism. In terms of this, we recover a formalization of the above idea, that an embedding is an isomorphism onto its image :
For a morphism in , the definition of an image as an equalizer says that the image of is
In particular, we have a factorization of as
where the morphism on the right is a monomorphism.
The morphism being an effective monomorphism means that is an isomorphism, and hence that is an “isomomorphism onto its image”.
A morphism of topological spaces is a regular monomorphism precisely if it is an injection such that the topology on is the induced topology. This is an embedding of topological spaces.
Last revised on January 4, 2026 at 19:31:50. See the history of this page for a list of all contributions to it.