You impose an equivalence relation ~ on morphisms of a category C where f ~ f' and g ~ g', but you find that g∘f and g'∘f' land in different equivalence classes. What has gone wrong?
ANothing — composed morphisms don't need to be equivalent just because their components are
BComposition in the quotient is ambiguous: the class [g∘f] depends on which representative you pick for [f] and [g], so the quotient structure is not a well-defined category
CThe identity morphisms are not preserved, which violates the axioms of the quotient
DYou need to also impose an equivalence on objects before the quotient on morphisms can be defined
This is the congruence failure. In the proposed quotient, [g] ∘ [f] should equal [g∘f] — but if different representatives g' and f' give a different composite, the composition law is ill-defined. The quotient category construction requires ~ to be a congruence: f ~ f' and g ~ g' must imply g∘f ~ g'∘f'. Without this, there is no canonical way to compose equivalence classes, and the putative quotient is not a category. This is the categorical analog of requiring a subgroup to be normal for the quotient group to have well-defined multiplication.
Question 2 Multiple Choice
In the homotopy category K(A), two chain maps f, g: A• → B• are identified if they are chain-homotopic. Why is this a valid quotient category construction?
AChain homotopy is a trivial relation — all chain maps between any two complexes are homotopic
BChain homotopy defines a congruence: if f ~ f' and g ~ g' (both chain-homotopic pairs), then the composites g∘f and g'∘f' are also chain-homotopic, so composition is well-defined on homotopy classes
CChain homotopy identifies objects (chain complexes), not morphisms, so the congruence condition does not apply
DThe construction is valid because chain complexes form an abelian category, which automatically makes any equivalence relation a congruence
The critical check for any quotient category construction is the congruence condition. For chain homotopy, if h: f → f' is a homotopy between f, f': A• → B• and k: g → g' is a homotopy between g, g': B• → C•, then one can explicitly construct a homotopy between g∘f and g'∘f'. This computation is the key step that validates K(A) as a category. The verification is not automatic — it is specific to the algebraic properties of chain homotopy — which is why checking congruence is always the first obligation in constructing a quotient category.
Question 3 True / False
Any equivalence relation on the morphism sets of a category can be used to form a valid quotient category.
TTrue
FFalse
Answer: False
This is the central misconception about quotient categories. An arbitrary equivalence relation on morphisms does not produce a category because composition may become ambiguous: if you represent [g∘f] as the class of g∘f, then using different representatives g' ~ g and f' ~ f might give g'∘f' in a different class. The relation must be a congruence — compatible with composition — to ensure that the composition of equivalence classes is well-defined. Without congruence, the putative quotient fails one of the basic axioms of a category.
Question 4 True / False
In a quotient category C/~, the objects are the same as in C, but morphisms are replaced by equivalence classes of morphisms under a congruence relation.
TTrue
FFalse
Answer: True
Quotient categories identify morphisms, not objects. The objects of C/~ are identical to those of C. For each pair of objects X, Y, the hom-set Hom_{C/~}(X, Y) = Hom_C(X, Y)/~ is the set of equivalence classes. Composition is defined on representatives: [g] ∘ [f] = [g∘f], which the congruence condition guarantees is independent of the choice. Identity morphisms descend directly: [id_X] is the identity in C/~. This construction is the categorical generalization of quotient groups, rings, or vector spaces — modding out by a compatible equivalence to coarsen the structure.
Question 5 Short Answer
Explain why the congruence condition (compatibility with composition) is necessary for forming a valid quotient category. What goes wrong if it fails?
Think about your answer, then reveal below.
Model answer: Composition in the quotient category C/~ is defined by choosing representatives: [g] ∘ [f] = [g∘f]. For this to be well-defined, the class of the composite must not depend on which representatives we choose from [f] and [g]. If f ~ f' and g ~ g' but g∘f is not equivalent to g'∘f', then [g] ∘ [f] would give different results depending on whether we compute g∘f or g'∘f' — composition is ambiguous, and we don't have a category. The congruence condition — f ~ f' and g ~ g' implies g∘f ~ g'∘f' — is exactly what prevents this ambiguity. It is the necessary and sufficient condition for the quotient to inherit a well-defined composition law.
An analogy: modding out a group G by a subgroup H gives a quotient group only if H is normal. The normality condition is the 'congruence' for groups: it ensures coset multiplication is well-defined. In categories, congruence generalizes normality to the setting of morphism composition. Without it, the quotient fails to be a group (in the algebraic case) or a category (in the categorical case).