A homomorphism is a map between two algebraic structures (like groups or rings) that preserves their operations โ combining elements first and then mapping gives the same result as mapping first and then combining. An isomorphism is a homomorphism that is also perfectly reversible, revealing that two seemingly different structures are, for all algebraic purposes, exactly "the same" โ just relabelled.
Imagine one clock labelled with the numbers 0 through 11, and another clock labelled with the letters A through L (A=0, B=1, C=2, and so on). These look completely different โ numbers versus letters โ but they behave identically under "clock addition" (adding hours and wrapping around). Adding 3 and 5 to get 8 on the number clock corresponds exactly to adding D-hours and F-hours to get I-hours on the letter clock. The two clocks are secretly the same mathematical structure, just wearing different labels.
Once you know two structures are isomorphic, you never need to study them separately again โ any true fact about one is automatically true about the other. This lets mathematicians classify entire families of structures by finding which ones are secretly identical to each other, dramatically simplifying what would otherwise be an overwhelming zoo of seemingly distinct objects.
Converting between Celsius and Fahrenheit temperature scales is a kind of structure-preserving map (though not quite a group isomorphism in the strict technical sense, since it involves an added constant) โ the relationships between temperatures are preserved even though the actual numbers used look completely different.
For groups (G,โ ) and (H,โ), a function ฯ:GโH is a homomorphism if ฯ(aโ b) = ฯ(a)โฯ(b) for all a,bโG. This single condition automatically forces ฯ to send the identity of G to the identity of H, and inverses to inverses โ the entire group structure is respected by just this one requirement.
The kernel ker(ฯ) = {gโG : ฯ(g)=eH} is the set of elements mapped to the identity โ it is always a normal subgroup of G. The image im(ฯ) = {ฯ(g) : gโG} is always a subgroup of H. A homomorphism is injective (one-to-one) exactly when its kernel contains only the identity element.
A homomorphism that is also a bijection (both injective and surjective) is an isomorphism. If an isomorphism exists between G and H, we write Gโ H and say the groups are isomorphic โ structurally identical, differing only in the labels or names given to their elements.
For any group homomorphism ฯ:GโH:
This landmark theorem says that "collapsing" G by its kernel produces something perfectly isomorphic to the image โ a foundational structural result used constantly throughout group theory, ring theory, and beyond.
An automorphism is an isomorphism from a structure to itself โ a way of relabelling the elements that perfectly preserves the entire structure. The set of all automorphisms of a group G forms its own group, Aut(G), under composition โ a "group of symmetries of the group itself," which plays an important role in Galois theory (where the Galois group is precisely the automorphism group of a field extension โ see that entry) and throughout the classification of algebraic structures.
The same core concepts extend naturally to rings (a ring homomorphism must preserve both addition and multiplication) and fields. A crucial and perhaps surprising fact: any homomorphism between fields is automatically injective โ because a field's only ideals are the zero ideal and the whole field itself, so the kernel of a field homomorphism (always an ideal) must be trivial, or else the homomorphism would collapse the entire field to a single point.
In category theory, homomorphisms are the prototypical example of morphisms โ structure-preserving maps between objects of any kind. This vastly more general and abstract framework (developed by Eilenberg and Mac Lane, 1945) treats the specific algebraic structure (group, ring, vector space, topological space) as almost secondary, focusing instead on the network of relationships (morphisms) between objects, and often reveals deep, unexpected structural parallels across seemingly unrelated branches of mathematics.
Many of the deepest results across all of algebra are fundamentally statements about isomorphism. The Classification of Finite Simple Groups (see the Group entry) is precisely a statement that every finite simple group is isomorphic to one of a specific, fully enumerated list of families. Similarly, the classification of finite-dimensional vector spaces (see the Vector Space entry) reduces entirely to a single number โ the dimension โ because any two vector spaces of the same dimension over the same field are automatically isomorphic to each other.
Despite being technically indistinguishable from a purely algebraic standpoint, isomorphic structures can still differ importantly in other, non-algebraic respects โ for instance, in computational efficiency of representation, or in which specific isomorphic copy naturally arises in a given real-world application. A significant portion of computational algebra is concerned precisely with efficiently identifying and constructing explicit isomorphisms between structures that are known, in principle, to be isomorphic.