Altanis

September 22nd, 2026

Updated 26 Sep 2026Notes (PDF)

Recall that subgroups can be pushed forward and pulled back by a homomorphism. Indeed, the image and preimage of a subgroup are subgroups. Moreover, the preimage of a normal subgroup is also a normal subgroup, and the image of a normal subgroup is normal as well, if the image is surjective.

[0.0.45]Theorem(Correspondence Theorem)#

Suppose φ:G→G′\phi: G \to G' is a surjective homomorphism. Let K=ker⁡(φ)K = \ker(\phi).

  1. There is a bijective correspondence

    {subgroups of G containing K}⟷{subgroups of G′}\Big\{ \text{subgroups of $G$ containing $K$} \Big\} \longleftrightarrow \Big\{ \text{subgroups of $G'$} \Big\}

    given by H↦φ(H)H \mapsto \phi(H) and φ−1(H′)← ⁣ ⁣∣H′\phi^{-1}(H') \mapsfrom H'.

  2. Suppose HH, a subgroup of GG containing KK, corresponds to H′H' under this mapping.

    • HH is a normal subgroup of GG if and only if H′H' is a normal subgroup of G′G'.

    • The restriction φ∣H:H→H′\phi|_H: H \to H' is a surjective homomorphism with kernel KK.

    • ∣H∣=∣K∣×∣H′∣|H| = |K| \times |H'|.

Proof.
  1. We first verify the mappings are well-defined and mutual inverses. We know the image and preimage of a subgroup under a homomorphism are always subgroups. Furthermore, for any subgroup H′⊆G′H' \subseteq G', note φ(K)={idG′}⊆H′\phi(K) = \{\id_{G'}\} \subseteq H', meaning K⊆φ−1(H′)K \subseteq \phi^{-1}(H'). Thus the domains and codomains of the bijection are correct.

    To show they are inverses, note that φ(φ−1(H′))=H′\phi(\phi^{-1}(H')) = H' follows immediately from the surjectivity of φ\phi. Conversely, we show φ−1(φ(H))=H\phi^{-1}(\phi(H)) = H for any subgroup HH containing KK. The inclusion H⊆φ−1(φ(H))H \subseteq \phi^{-1}(\phi(H)) is trivial. Let x∈φ−1(φ(H))x \in \phi^{-1}(\phi(H)). Then φ(x)=φ(h)\phi(x) = \phi(h) for some h∈Hh \in H. Rewriting yields φ(h−1x)=idG′\phi(h^{-1}x) = \id_{G'}, meaning h−1x∈Kh^{-1}x \in K. Because K⊆HK \subseteq H, we have h−1x∈Hh^{-1}x \in H, and since h∈Hh \in H, closure forces x∈Hx \in H. Thus φ−1(φ(H))⊆H\phi^{-1}(\phi(H)) \subseteq H, meaning φ−1(φ(H))=H\phi^{-1}(\phi(H)) = H, establishing the bijection.

  2. Suppose H=φ−1(H′)H = \phi^{-1}(H') and H′=φ(H)H' = \phi(H).

    • We previously proved that the preimage of a normal subgroup is always normal, and the image of a normal subgroup is normal provided the homomorphism is surjective. The result comes immediately.

    • The restriction φ∣H:H→H′\phi|_H: H \to H' is inherently a homomorphism. It is surjective because H′=φ(H)H' = \phi(H). The kernel of this restricted map is precisely {x∈H:φ(x)=idG′}=H∩K\{x \in H : \phi(x) = \id_{G'}\} = H \cap K. Because K⊆HK \subseteq H, this collapses to KK, and so ker⁡(φ∣H)=K\ker(\phi|_H) = K.

    • By applying the standard counting formula to the restricted map φ∣H\phi|_H, we immediately yield ∣H∣=∣ker⁡(φ∣H)∣×∣im⁡(φ∣H)∣=∣K∣×∣H′∣|H| = |\ker(\phi|_H)| \times |\im(\phi|_H)| = |K| \times |H'|.

[0.0.46]Remark#

This proof is mostly straightforward, but some remarks are in order for (1)(1). In general, the identity φ−1(φ(H))=H\phi^{-1}(\phi(H)) = H is false. If you push a set forward and then pull it back, the set can get bigger because you accidentally grab extra elements that map to the same place. By our previous theorem, however, the set of all elements mapping to φ(h)\phi(h) is exactly the coset hKhK. This is precisely why the assumption K⊆HK \subseteq H is so crucial. By the closure of HH, multiplying any h∈Hh \in H by elements of K⊆HK \subseteq H keeps you strictly inside HH. Thus, the entire pulled-back coset hKhK lies completely within HH. The “extra” elements we grabbed were already inside HH to begin with, ensuring the set does not actually grow.

Let GG be a group together with a subgroup NN. We already know that G/NG/N, as a set, is the set of all left cosets of NN. We ask if G/NG/N can be endowed with some natural operation that can make it into a group. The answer is yes, but we require NN to be a normal subgroup, for reasons we will see soon.

[0.0.47]Definition(Coset Multiplication, Quotient Group)#

Suppose GG is a group together with a normal subgroup NN. For any two cosets aN,bN∈G/NaN, bN \in G/N, define coset multiplication by

(aN)(bN)=(ab)N.(aN)(bN) = (ab)N.

Then G/NG/N forms a group under this operation.

Note that G/NG/N is not a subgroup of GG, which is simultaneously obvious and surprising. It is not immediately clear that this forms a well-defined group operation. Indeed, since aN=a′NaN = a'N if a′∈aNa' \in aN, we need to show that the choice of representative for the coset does not alter the result of multiplication.

[0.0.48]Theorem(Well-Definedness of Coset Multiplication)#

Let NN be a subgroup of GG. The operation (aN)(bN)=(ab)N(aN)(bN) = (ab)N is well-defined on the set of left cosets if and only if NN is a normal subgroup of GG.

Proof.

We show that if NN is normal, the operation is strictly independent of the chosen representatives. Suppose aN=a′NaN = a'N and bN=b′NbN = b'N. We wish to show that (a′b′)N=(ab)N(a'b')N = (ab)N.

By the equivalence class properties of cosets, we can write a′=an1a' = a n_1 and b′=bn2b' = b n_2 for some n1,n2∈Nn_1, n_2 \in N. Then note

a′b′=(an1)(bn2)=a(n1b)n2.a'b' = (an_1)(bn_2) = a(n_1b)n_2.

Because NN is normal, its left and right cosets coincide, meaning Nb=bNNb = bN. Thus, for our element n1∈Nn_1 \in N, there exists some n3∈Nn_3 \in N such that n1b=bn3n_1 b = b n_3. Substituting this yields

a′b′=a(bn3)n2=ab(n3n2).a'b' = a(bn_3)n_2 = ab(n_3n_2).

Since NN is a subgroup and is closed under multiplication, n3n2∈Nn_3n_2 \in N. Thus a′b′a'b' can be written as abab multiplied by an element of NN, which immediately implies a′b′∈abNa'b' \in abN. Consequently, (a′b′)N=(ab)N(a'b')N = (ab)N.

If NN were not normal, the element n1bn_1 b could permanently escape bNbN, shifting the product into a completely different coset and causing the entire operation to be ill-defined.

Thus the subgroup for which quotienting by yields a group is precisely a normal subgroup. Having established that the operation is completely independent of the chosen representatives when NN is normal, we may proceed to quickly verify the standard group axioms.

[0.0.49]Theorem(Quotient Group Structure)#

Let NN be a normal subgroup of GG. The set of left cosets G/NG/N forms a group under coset multiplication.

Proof.
  1. Closure. Since a,b∈Ga, b \in G, their product ab∈Gab \in G by closure of the parent group, and so (ab)N(ab)N is indeed a valid coset in G/NG/N.

  2. Associativity. For any aN,bN,cN∈G/NaN, bN, cN \in G/N, note that

    [(aN)(bN)](cN)=(ab)N⋅cN=((ab)c)N.[(aN)(bN)](cN) = (ab)N \cdot cN = ((ab)c)N.

    Because the operation in GG is inherently associative, ((ab)c)N=(a(bc))N((ab)c)N = (a(bc))N. Thus

    (a(bc))N=aN⋅(bc)N=(aN)[(bN)(cN)],(a(bc))N = aN \cdot (bc)N = (aN)[(bN)(cN)],

    satisfying associativity.

  3. Identity. The coset of the identity, idGN=N\id_G N = N, serves as the identity element. For any aN∈G/NaN \in G/N, we have

    (aN)(N)=(aidG)N=aN,and(N)(aN)=(idGa)N=aN.(aN)(N) = (a \id_G)N = aN, \quad \text{and} \quad (N)(aN) = (\id_G a)N = aN.
  4. Inverse. The inverse of aNaN is precisely a−1Na^{-1}N. Indeed,

    (aN)(a−1N)=(aa−1)N=idGN=N,(aN)(a^{-1}N) = (aa^{-1})N = \id_G N = N,

    and symmetrically (a−1N)(aN)=N(a^{-1}N)(aN) = N.

Earlier, we proved that the kernel of any homomorphism is inherently a normal subgroup. We now resolve the natural converse: is every normal subgroup the kernel of some homomorphism? Having constructed the quotient group G/NG/N, the answer is yes.

[0.0.50]Definition(Canonical Projection Map)#

Let GG be a group and NN a normal subgroup of GG. We define the canonical projection map π:G→G/N\pi: G \to G/N by

π(a)=aN.\pi(a) = aN.
[0.0.51]Theorem#

The canonical projection map π:G→G/N\pi: G \to G/N is a surjective homomorphism with ker⁡(π)=N\ker(\pi) = N.

Proof.

First, we show π\pi is a homomorphism. For any a,b∈Ga, b \in G, the definition of coset multiplication yields

π(ab)=(ab)N=(aN)(bN)=π(a)π(b),\pi(ab) = (ab)N = (aN)(bN) = \pi(a)\pi(b),

and so π\pi preserves the group structure. Surjectivity is trivial, as any coset aN∈G/NaN \in G/N is immediately mapped to by a∈Ga \in G. Finally, we determine the kernel. The identity element in G/NG/N is the coset idGN=N\id_G N = N. Thus, a∈ker⁡(π)a \in \ker(\pi) if and only if π(a)=N\pi(a) = N, meaning aN=NaN = N. This occurs if and only if a∈Na \in N. Therefore, ker⁡(π)=N\ker(\pi) = N.

Thus, a subgroup is normal if and only if it is the kernel of a homomorphism.

Having classically constructed the quotient group G/NG/N by defining operations on cosets, we now take a step back. The quotient construction is not something exclusive to groups; it is a fundamental, structural mechanism that works with several constructions (sets, rings, topological spaces, and more). To truly understand what a quotient is, we must examine it by working with abstract objects, not simply groups. We start with sets.

Let AA and BB be arbitrary sets, and let ∼\sim be an equivalence relation on AA. Recall that the equivalence classes of ∼\sim partition the set AA, and we denote the equivalence class of an element x∈Ax \in A by [x]∼[x]_\sim. We can define the projection map π:A→A/∼\pi: A \to A/{\sim} by π(x)=[x]∼\pi(x) = [x]_\sim.

Suppose we have a function f:A→Bf: A \to B. We ask if we can push this function down through the quotient to define a new map on A/∼A/{\sim}. For this to make sense, the original function must be blind to the differences between equivalent elements.

[0.0.52]Definition(Respecting an Equivalence Relation)#

Let f:A→Bf: A \to B be a function and ∼\sim be an equivalence relation on AA. We say that ff respects the equivalence relation ∼\sim if ff is constant on equivalence classes. That is, for all x,y∈Ax, y \in A:

x∼y  ⟹  f(x)=f(y).x \sim y \implies f(x) = f(y).

If ff respects the equivalence relation, it guarantees that ff acts on entire equivalence classes uniformly, allowing for a powerful factorization of the action of ff.

[0.0.53]Theorem(Universal Property of Quotients in Set\Set)#

Let ∼\sim be an equivalence relation on a set AA, and let π:A→A/∼\pi: A \to A/{\sim} be the canonical projection map. For any set BB and any function f:A→Bf: A \to B that respects the equivalence relation ∼\sim, there exists a unique function f~:A/∼→B\tilde{f}: A/{\sim} \to B such that f=f~∘πf = \tilde{f} \circ \pi.

Equivalently, the following diagram commutes:

Proof.

We must establish both the existence and the uniqueness of f~\tilde{f}.

  • Existence. We define f~\tilde{f} directly on the equivalence classes in A/∼A/{\sim} by setting f~([x]∼)=f(x)\tilde{f}([x]_\sim) = f(x). Because elements of A/∼A/{\sim} are sets (equivalence classes) rather than individual elements of AA, we must verify this function is well-defined. Suppose [x]∼=[y]∼[x]_\sim = [y]_\sim. This implies x∼yx \sim y. Because we assumed ff respects the equivalence relation, f(x)=f(y)f(x) = f(y), and therefore f~([x]∼)=f~([y]∼)\tilde{f}([x]_\sim) = \tilde{f}([y]_\sim). The definition is independent of the chosen representative, so f~\tilde{f} is a well-defined function. Furthermore, for any x∈Ax \in A, we have (f~∘π)(x)=f~([x]∼)=f(x)(\tilde{f} \circ \pi)(x) = \tilde{f}([x]_\sim) = f(x), ensuring the diagram commutes.

  • Uniqueness. Suppose there is another function g:A/∼→Bg: A/{\sim} \to B such that g∘π=fg \circ \pi = f. We wish to show g=f~g = \tilde{f}. Let [x]∼[x]_\sim be an arbitrary element of A/∼A/{\sim}. Because π\pi is inherently surjective, [x]∼=π(x)[x]_\sim = \pi(x) for some x∈Ax \in A. Then g([x]∼)=g(π(x))=f(x)g([x]_\sim) = g(\pi(x)) = f(x). But we already know f(x)=f~(π(x))=f~([x]∼)f(x) = \tilde{f}(\pi(x)) = \tilde{f}([x]_\sim). Thus g([x]∼)=f~([x]∼)g([x]_\sim) = \tilde{f}([x]_\sim) for all classes in the quotient, forcing g=f~g = \tilde{f}.

[0.0.54]Remark(The Canonical Decomposition in Set\Set)#

The universal property of quotients gives us a powerful tool to mechanically “repair” arbitrary functions. Let f:A→Bf: A \to B be any generic set function. Generally, ff fails to be a bijection (an isomorphism in the category of sets) for two reasons. First, it may fail surjectivity (missing elements in BB). Second, it may fail injectivity (collapsing distinct elements of AA to the same target).

We can resolve both of these flaws. First, we fix surjectivity by restricting our codomain to the image, im⁡(f)\im(f). The map from AA to im⁡(f)\im(f) is trivially surjective, and we can seamlessly include im⁡(f)\im(f) back into BB via an injective inclusion map ι:im⁡(f)→B\iota: \im(f) \to B.

Second, we fix injectivity by defining a natural equivalence relation on AA induced by the function itself. Define x∼fyx \sim_f y if and only if f(x)=f(y)f(x) = f(y). By definition, ff trivially respects the equivalence relation ∼f\sim_f. Quotienting AA by ∼f\sim_f “glues” all elements that map to the same target into a single equivalence class. Immediately, we see that the universal property of quotients yields a unique map f~:A/∼f→im⁡(f)\tilde{f}: A/{\sim_f} \to \im(f).

Because we restricted the codomain to im⁡(f)\im(f), f~\tilde{f} is surjective. Because we quotiented out the exact elements that caused collisions, f~\tilde{f} is strictly injective. Thus, f~\tilde{f} is a forced bijection.

[0.0.55]Theorem(Canonical Decomposition of Functions)#

Any function f:A→Bf: A \to B can be uniquely factored as the composition of a surjection, a bijection, and an injection. This factorization, f=ι∘f~∘πf = \iota \circ \tilde{f} \circ \pi, is captured by the following commutative diagram:

where π:A→A/∼f\pi: A \to A/{\sim_f} is the surjective projection map, f~:A/∼f→im⁡(f)\tilde{f}: A/{\sim_f} \to \im(f) is a natural bijection, and ι:im⁡(f)→B\iota: \im(f) \to B is the injective inclusion map.

Any function ff is secretly just a bijection wrapped in layers of redundancy (collapsing elements) and deficiency (missing the target). In our canonical decomposition diagram, the first part of the sequence (the projection π\pi) acts as the quotient fix to resolve injectivity issues, while the last part (the inclusion ι\iota) acts as the embedding fix to resolve surjectivity issues. By quotienting the domain and restricting the codomain, we strip away these flaws to isolate f~\tilde{f}, a faithful representation capturing the core mechanism as to how ff operates.

We now transport this idea into the category of groups. Previously, we classically constructed the quotient group G/NG/N by defining coset multiplication, proving that it forms a group if and only if NN is a normal subgroup of GG. Instead of building the quotient from the ground up and verifying it works, we can equivalently define the quotient entirely by demanding it satisfies a universal property analogous to the one in Set\Set, and then locate the object that fits the description.

[0.0.56]Definition(Categorical Quotient Group)#

Let GG be a group and NN a normal subgroup of GG. The quotient of GG by NN is a group QQ together with a surjective homomorphism π:G→Q\pi: G \to Q with ker⁡(π)=N\ker(\pi) = N, satisfying the following universal property.

For any group HH and any homomorphism φ:G→H\phi: G \to H where N⊆ker⁡(φ)N \subseteq \ker(\phi), there exists a unique homomorphism f~:Q→H\tilde{f}: Q \to H such that f~∘π=φ\tilde{f} \circ \pi = \phi.

One might wonder why we still constrain NN to be a normal subgroup in a supposedly pure, structural definition. Because we explicitly demand π\pi to be a homomorphism with ker⁡(π)=N\ker(\pi) = N, and the kernel of any homomorphism is inherently normal, the rigid structure of groups strictly prohibits non-normal subgroups from ever acting as kernels. Thus, in the categorical lens, a normal subgroup is not an arbitrary set-theoretic restriction, but simply the name we give to a subgroup capable of being a kernel for some homomorphism.

Furthermore, the condition N⊆ker⁡(φ)N \subseteq \ker(\phi) is exactly the group-theoretic translation of a set function respecting an equivalence relation. If elements in NN get crushed to the identity by φ\phi, then φ\phi is completely blind to differences between elements that differ by an element of NN.

Before we confirm our classical coset space fits this definition, we must show that this universal property is actually rigid. By a standard diagram chase, any two objects satisfying this property are uniquely isomorphic, meaning the quotient is structurally unique.

[0.0.57]Theorem(Uniqueness of the Categorical Quotient)#

Let (Q1,π1)(Q_1, \pi_1) and (Q2,π2)(Q_2, \pi_2) be two quotients of GG by NN satisfying the universal property. Then there exists a unique isomorphism α:Q1→Q2\alpha: Q_1 \to Q_2 such that α∘π1=π2\alpha \circ \pi_1 = \pi_2.

Proof.

Since (Q1,π1)(Q_1, \pi_1) satisfies the universal property and π2:G→Q2\pi_2: G \to Q_2 is a homomorphism with ker⁡(π2)=N⊆ker⁡(π2)\ker(\pi_2) = N \subseteq \ker(\pi_2) (trivially), there exists a unique homomorphism α:Q1→Q2\alpha: Q_1 \to Q_2 such that α∘π1=π2\alpha \circ \pi_1 = \pi_2.

Symmetrically, since (Q2,π2)(Q_2, \pi_2) satisfies the universal property and π1:G→Q1\pi_1: G \to Q_1 is a homomorphism with N⊆ker⁡(π1)N \subseteq \ker(\pi_1), there exists a unique homomorphism β:Q2→Q1\beta: Q_2 \to Q_1 such that β∘π2=π1\beta \circ \pi_2 = \pi_1.

Substituting the first identity into the second yields β∘(α∘π1)=π1\beta \circ (\alpha \circ \pi_1) = \pi_1, which we reassociate as (β∘α)∘π1=π1(\beta \circ \alpha) \circ \pi_1 = \pi_1.

Now, consider the universal property of Q1Q_1 applied to the homomorphism π1:G→Q1\pi_1: G \to Q_1. We seek a unique map u:Q1→Q1u: Q_1 \to Q_1 such that u∘π1=π1u \circ \pi_1 = \pi_1. The identity map idQ1\id_{Q_1} trivially satisfies this. However, we just showed that β∘α\beta \circ \alpha also satisfies this. By the strict uniqueness guaranteed by the universal property, we are forced to conclude β∘α=idQ1\beta \circ \alpha = \id_{Q_1}.

By an identical symmetric argument, α∘β=idQ2\alpha \circ \beta = \id_{Q_2}. Thus α\alpha is a bijection, and since it is inherently a homomorphism, it is an isomorphism. This establishes that Q1≅Q2Q_1 \cong Q_2.

We now confirm that our classically constructed set of cosets is indeed the exact object requested by this universal property.

[0.0.58]Theorem(Coset Space Satisfies the Universal Property)#

The classically constructed coset space G/NG/N equipped with coset multiplication (aN)(bN)=(ab)N(aN)(bN) = (ab)N, paired with the canonical projection π(a)=aN\pi(a) = aN, satisfies the universal property of the quotient group.

Proof.

Let φ:G→H\phi: G \to H be a homomorphism such that N⊆ker⁡(φ)N \subseteq \ker(\phi). We must show there exists a unique homomorphism f~:G/N→H\tilde{f}: G/N \to H making the diagram commute.

  • Well-definedness. We propose f~(aN)=φ(a)\tilde{f}(aN) = \phi(a). Because we are defining this on cosets (equivalence classes), we must check it is strictly independent of the representative. Suppose aN=bNaN = bN. This means b−1a∈Nb^{-1}a \in N. Because N⊆ker⁡(φ)N \subseteq \ker(\phi), this implies φ(b−1a)=idH\phi(b^{-1}a) = \id_H. Since φ\phi is a homomorphism, φ(b)−1φ(a)=idH\phi(b)^{-1}\phi(a) = \id_H, meaning φ(a)=φ(b)\phi(a) = \phi(b). Thus f~(aN)=f~(bN)\tilde{f}(aN) = \tilde{f}(bN), so f~\tilde{f} is well-defined.

  • Homomorphism. We verify that f~\tilde{f} respects the group operation. Indeed, note

    f~((aN)(bN))=f~((ab)N)=φ(ab)=φ(a)φ(b)=f~(aN)f~(bN).\tilde{f}((aN)(bN)) = \tilde{f}((ab)N) = \phi(ab) = \phi(a)\phi(b) = \tilde{f}(aN)\tilde{f}(bN).

    Thus f~\tilde{f} is a valid group homomorphism. By construction, f~(π(a))=f~(aN)=φ(a)\tilde{f}(\pi(a)) = \tilde{f}(aN) = \phi(a), ensuring the diagram commutes.

  • Uniqueness. Suppose g:G/N→Hg: G/N \to H is another homomorphism satisfying g∘π=φg \circ \pi = \phi. For any coset aN∈G/NaN \in G/N, we have g(aN)=g(π(a))=φ(a)g(aN) = g(\pi(a)) = \phi(a). But φ(a)=f~(aN)\phi(a) = \tilde{f}(aN). Since gg and f~\tilde{f} agree on all elements, g=f~g = \tilde{f}.

[0.0.59]Remark(The Canonical Decomposition in Grp\Grp)#

We conclude by applying our canonical decomposition setup from Set\Set to group homomorphisms. Let φ:G→H\phi: G \to H be an arbitrary homomorphism. Like any general function, it may fail to be an isomorphism due to a lack of injectivity or a lack of surjectivity.

We resolve surjectivity by mapping onto the image im⁡(φ)\im(\phi), which is a valid subgroup of HH. We resolve injectivity by defining the equivalence relation a∼φb  ⟺  φ(a)=φ(b)a \sim_\phi b \iff \phi(a) = \phi(b). In the language of groups, φ(a)=φ(b)\phi(a) = \phi(b) implies a−1b∈ker⁡(φ)a^{-1}b \in \ker(\phi). Thus, the equivalence classes [a]∼φ[a]_{\sim_\phi} are exactly the left cosets aker⁡(φ)a\ker(\phi).

Because the kernel of any homomorphism is inherently a normal subgroup of GG, the quotient space G/ker⁡(φ)G/\ker(\phi) naturally forms a group. By its very definition, φ\phi respects the equivalence relation generated by its own kernel. The universal property immediately gives us a unique, well-defined homomorphism f~:G/ker⁡(φ)→im⁡(φ)\tilde{f}: G/\ker(\phi) \to \im(\phi).

Because we quotiented by the exact subgroup causing injectivity collisions, f~\tilde{f} is injective. Because we restricted the codomain, f~\tilde{f} is surjective. Therefore, f~\tilde{f} is an isomorphism.

The First Isomorphism Theorem is not a standalone group-theoretic trick. It is the realization of the universal canonical decomposition of maps in an arbitrary category, simply applied to Grp\Grp. Injectivity is factored out by quotienting by the kernel, surjectivity is factored out by restricting to the image, and we are left with f~\tilde{f}, the underlying isomorphism that faithfully represents φ\phi.

[0.0.60]Theorem(First Isomorphism Theorem)#

Let φ:G→H\phi: G \to H be a group homomorphism. Then φ\phi induces a canonical isomorphism

G/ker⁡(φ)≅im⁡(φ)G/\ker(\phi) \cong \im(\phi)

given by the map f~(aker⁡(φ))=φ(a)\tilde{f}(a\ker(\phi)) = \phi(a).