nLab equivalence induction

Context

Type theory

natural deduction metalanguage, practical foundations

  1. type formation rule
  2. term introduction rule
  3. term elimination rule
  4. computation rule

type theory (dependent, intensional, observational type theory, homotopy type theory)

syntax object language

computational trinitarianism =
propositions as types +programs as proofs +relation type theory/category theory

logicset theory (internal logic of)category theorytype theory
propositionsetobjecttype
predicatefamily of setsdisplay morphismdependent type
proofelementgeneralized elementterm/program
cut rulecomposition of classifying morphisms / pullback of display mapssubstitution
introduction rule for implicationcounit for hom-tensor adjunctionlambda
elimination rule for implicationunit for hom-tensor adjunctionapplication
cut elimination for implicationone of the zigzag identities for hom-tensor adjunctionbeta reduction
identity elimination for implicationthe other zigzag identity for hom-tensor adjunctioneta conversion
truesingletonterminal object/(-2)-truncated objecth-level 0-type/unit type
falseempty setinitial objectempty type
proposition, truth valuesubsingletonsubterminal object/(-1)-truncated objecth-proposition, mere proposition
logical conjunctioncartesian productproductproduct type
disjunctiondisjoint union (support of)coproduct ((-1)-truncation of)sum type (bracket type of)
implicationfunction set (into subsingleton)internal hom (into subterminal object)function type (into h-proposition)
negationfunction set into empty setinternal hom into initial objectfunction type into empty type
universal quantificationindexed cartesian product (of family of subsingletons)dependent product (of family of subterminal objects)dependent product type (of family of h-propositions)
existential quantificationindexed disjoint union (support of)dependent sum ((-1)-truncation of)dependent sum type (bracket type of)
logical equivalencebijection setobject of isomorphismsequivalence type
support setsupport object/(-1)-truncationpropositional truncation/bracket type
n-image of morphism into terminal object/n-truncationn-truncation modality
propositional equalitydiagonal function/diagonal subset/diagonal relationpath space objectidentity type/path type
completely presented setsetdiscrete object/0-truncated objecth-level 2-type/set/h-set
setset with equivalence relationinternal 0-groupoidBishop set/setoid with its pseudo-equivalence relation an actual equivalence relation
equivalence class/quotient setquotientquotient type
inductioncolimitinductive type, W-type, M-type
higher inductionhigher colimithigher inductive type
-0-truncated higher colimitquotient inductive type
coinductionlimitcoinductive type
presettype without identity types
set of truth valuessubobject classifiertype of propositions
domain of discourseuniverseobject classifiertype universe
modalityclosure operator, (idempotent) monadmodal type theory, monad (in computer science)
linear logic(symmetric, closed) monoidal categorylinear type theory/quantum computation
proof netstring diagramquantum circuit
(absence of) contraction rule(absence of) diagonalno-cloning theorem
synthetic mathematicsdomain specific embedded programming language

homotopy levels

semantics

Contents

Idea

In dependent type theory, equivalence induction says that equivalence types are inductively generated by the identity equivalence, in the same way that identity types are inductively generated by reflexivity. Like the induction principle of identity types, there are two versions of equvialence induction, the standard (Martin-Lof style) equivalence induction and the based (Paulin-Mohrung style) equivalence induction.

Equivalence induction is equivalent in strength to the univalence axiom, because equivalence induction makes the equivalence type behave like an identity system, which then makes the equivalence type equivalent to the identity type between types, which is precisely the univalence axiom. Conversely, corollary 5.8.5 of the HoTT Book derives equivalence induction from the univalence axiom.

Definition

Using universes

Equivalence induction is usually formulated using universes UU, such as in Corollary 5.8.5 of the HoTT Book.

Definition

The (based Paulin-Mohrung style) equivalence induction states that given UU-small types A:UA:U and a UU-small type family P(X,e)P(X, e) indexed by UU-small type X:UX:U and equivalence e:A≃Xe:A \simeq X, and an element T:P(A,idequiv A)T:P(A, \mathrm{idequiv}_A), for all X:UX:U and e:A≃Xe:A \simeq X one can construct an element ind ≃ P(T,X,e):P(X,e)\mathrm{ind}_\simeq^P(T, X, e):P(X, e) such that T=ind ≃ P(T,A,idequiv A)T = \mathrm{ind}_\simeq^P(T, A, \mathrm{idequiv}_A).

The inference rules for this version of equivalence induction are given by:

Γ⊢A:UΓ,X:U,e:A≃X⊢P(X,e):UΓ,T:P(A,idequiv A),X:U,e:A≃X⊢P(X,e)⊢ind ≃ P(T,X,e):P(X,e)\frac{\Gamma \vdash A:U \quad \Gamma, X:U, e:A \simeq X \vdash P(X, e):U}{\Gamma, T:P(A, \mathrm{idequiv}_A), X:U, e:A \simeq X \vdash P(X, e) \vdash \mathrm{ind}_\simeq^P(T, X, e):P(X, e)}
Γ⊢A:UΓ,X:U,e:A≃X⊢P(X,e):UΓ,T:P(A,idequiv A)⊢ind ≃ P(T,A,idequiv A)≡T:P(A,idequiv A)\frac{\Gamma \vdash A:U \quad \Gamma, X:U, e:A \simeq X \vdash P(X, e):U}{\Gamma, T:P(A, \mathrm{idequiv}_A) \vdash \mathrm{ind}_\simeq^P(T, A, \mathrm{idequiv}_A) \equiv T:P(A, \mathrm{idequiv}_A)}

Definition

The (standard Martin-Lof style) equivalence induction states that given a UU-small type family P(X,Y,e)P(X, Y, e) indexed by UU-small types X:UX:U and Y:UY:U and equivalence e:X≃Ye:X \simeq Y, and a dependent function T:∏ A:UP(A,A,idequiv A)T:\prod_{A:U} P(A, A, \mathrm{idequiv}_A), for all X:UX:U, Y:UY:U, and e:X≃Ye:X \simeq Y one can construct an element ind ≃ P(T,X,Y,e):P(X,Y,e)\mathrm{ind}_\simeq^P(T, X, Y, e):P(X, Y, e) such that T(X)=ind ≃ P(T,X,X,idequiv X)T(X) = \mathrm{ind}_\simeq^P(T, X, X, \mathrm{idequiv}_X).

The inference rules for this version of equivalence induction are given by:

Γ,X:U,Y:U,e:X≃Y⊢P(X,Y,e):UΓ,T:∏ A:UP(A,A,idequiv A),X:U,Y:U,e:X≃Y⊢P(X,Y,e)⊢ind ≃ P(T,X,Y,e):P(X,Y,e)\frac{\Gamma, X:U, Y:U, e:X \simeq Y \vdash P(X, Y, e):U}{\Gamma, T:\prod_{A:U} P(A, A, \mathrm{idequiv}_A), X:U, Y:U, e:X \simeq Y \vdash P(X, Y, e) \vdash \mathrm{ind}_\simeq^P(T, X, Y, e):P(X, Y, e)}
Γ,X:U,Y:U,e:X≃Y⊢P(X,Y,e):UΓ,T:∏ A:UP(A,A,idequiv A),X:U⊢ind ≃ P(T,X,X,idequiv X)≡T(X):P(X,X,idequiv X)\frac{\Gamma, X:U, Y:U, e:X \simeq Y \vdash P(X, Y, e):U}{\Gamma, T:\prod_{A:U} P(A, A, \mathrm{idequiv}_A), X:U \vdash \mathrm{ind}_\simeq^P(T, X, X, \mathrm{idequiv}_X) \equiv T(X):P(X, X, \mathrm{idequiv}_X)}

Using type variables

Traditionally, equivalence induction is formulated using universes, such as in Corollary 5.8.5 of the HoTT Book. However, with type variables, one can express equivalence induction without using universes.

Definition

The (based Paulin-Mohrung style) equivalence induction states that given type AA and a type family P(X,e)P(X, e) indexed by type XX and equivalence e:A≃Xe:A \simeq X, and an element T:P(A,idequiv A)T:P(A, \mathrm{idequiv}_A), for all types XX and e:A≃Xe:A \simeq X one can construct an element ind ≃ P(T,X,e):P(X,e)\mathrm{ind}_\simeq^P(T, X, e):P(X, e) such that T=ind ≃ P(T,A,idequiv A)T = \mathrm{ind}_\simeq^P(T, A, \mathrm{idequiv}_A).

The inference rules for this version of equivalence induction are given by:

Γ⊢AtypeΓ,Xtype,e:A≃X⊢P(X,e)typeΓ,T:P(A,idequiv A),Xtype,e:A≃X⊢P(X,e)⊢ind ≃ P(T,X,e):P(X,e)\frac{\Gamma \vdash A \; \mathrm{type} \quad \Gamma, X \; \mathrm{type}, e:A \simeq X \vdash P(X, e) \; \mathrm{type}}{\Gamma, T:P(A, \mathrm{idequiv}_A), X \; \mathrm{type}, e:A \simeq X \vdash P(X, e) \vdash \mathrm{ind}_\simeq^P(T, X, e):P(X, e)}
Γ⊢AtypeΓ,Xtype,e:A≃X⊢P(X,e)typeΓ,T:P(A,idequiv A)⊢ind ≃ P(T,A,idequiv A)≡T:P(A,idequiv A)\frac{\Gamma \vdash A \; \mathrm{type} \quad \Gamma, X \; \mathrm{type}, e:A \simeq X \vdash P(X, e) \; \mathrm{type}}{\Gamma, T:P(A, \mathrm{idequiv}_A) \vdash \mathrm{ind}_\simeq^P(T, A, \mathrm{idequiv}_A) \equiv T:P(A, \mathrm{idequiv}_A)}

With type variables and impredicative polymorphism, one can also express the standard equivalence induction without using universes.

Definition

The (standard Martin-Lof style) equivalence induction states that given a type family P(X,Y,e)P(X, Y, e) indexed by types XX and YY and equivalence e:X≃Ye:X \simeq Y, and an element T:ΠA.P(A,A,idequiv A)T:\Pi A.P(A, A, \mathrm{idequiv}_A), for all types XX, YY, and equivalences e:X≃Ye:X \simeq Y one can construct an element ind ≃ P(T,X,Y,e):P(X,Y,e)\mathrm{ind}_\simeq^P(T, X, Y, e):P(X, Y, e) such that T(X)=ind ≃ P(T,X,X,idequiv X)T(X) = \mathrm{ind}_\simeq^P(T, X, X, \mathrm{idequiv}_X).

The inference rules for this version of equivalence induction are given by:

Γ,Xtype,Ytype,e:X≃Y⊢P(X,Y,e)typeΓ,T:∀A.P(A,A,idequiv A),Xtype,Ytype,e:X≃Y⊢P(X,Y,e)⊢ind ≃ P(T,X,Y,e):P(X,Y,e)\frac{\Gamma, X \; \mathrm{type}, Y \; \mathrm{type}, e:X \simeq Y \vdash P(X, Y, e) \; \mathrm{type}}{\Gamma, T:\forall A.P(A, A, \mathrm{idequiv}_A), X \; \mathrm{type}, Y \; \mathrm{type}, e:X \simeq Y \vdash P(X, Y, e) \vdash \mathrm{ind}_\simeq^P(T, X, Y, e):P(X, Y, e)}
Γ,Xtype,Ytype,e:X≃Y⊢P(X,Y,e)typeΓ,T:∀A.P(A,A,idequiv A),Xtype⊢ind ≃ P(T,X,X,idequiv X)≡T(X):P(X,X,idequiv X)\frac{\Gamma, X \; \mathrm{type}, Y \; \mathrm{type}, e:X \simeq Y \vdash P(X, Y, e) \; \mathrm{type}}{\Gamma, T:\forall A.P(A, A, \mathrm{idequiv}_A), X \; \mathrm{type} \vdash \mathrm{ind}_\simeq^P(T, X, X, \mathrm{idequiv}_X) \equiv T(X):P(X, X, \mathrm{idequiv}_X)}

In cubical type theory

In cubical type theory with glue types and either type universes or type variables, equivalence induction is provable from the glue types. This is in contrast with the usual Martin-Lof type theory where equivalence induction is a set of two unjustified inference rules equivalent in strength to the univalence axiom.

References

Equivalence induction is defined using universes in corollary 5.8.5 of:

Equivalence induction is defined using type variables in:

  • Dependent Type Theory vs Polymorphic Type Theory, Category Theory Zulip (web)

Last revised on September 20, 2026 at 02:04:17. See the history of this page for a list of all contributions to it.