4.1 Quasi-inverses


We have said that π—Šπ—‚π—‡π—β’(f) is unsatisfactory because it is not a mere proposition, whereas we would rather that a given function can β€œbe an equivalence” in at most one way. However, we have given no evidence that π—Šπ—‚π—‡π—β’(f) is not a mere proposition. In this sectionPlanetmathPlanetmathPlanetmathPlanetmathPlanetmathPlanetmathPlanetmathPlanetmath we exhibit a specific counterexample.

Lemma 4.1.1.

If f:Aβ†’B is such that qinv⁒(f) is inhabited, then

π—Šπ—‚π—‡π—(f)≃(∏x:A(x=x)).
Proof.

By assumptionPlanetmathPlanetmath, f is an equivalence; that is, we have e:π—‚π—Œπ–Ύπ—Šπ—Žπ—‚π—β’(f) and so (f,e):A≃B. By univalence, π—‚π–½π—π—ˆπ–Ύπ—Šπ—:(A=B)β†’(A≃B) is an equivalence, so we may assume that (f,e) is of the form π—‚π–½π—π—ˆπ–Ύπ—Šπ—β’(p) for some p:A=B. Then by path inductionMathworldPlanetmath, we may assume p is 𝗋𝖾𝖿𝗅A, in which case π—‚π–½π—π—ˆπ–Ύπ—Šπ—β’(p) is 𝗂𝖽A. Thus we are reduced to proving π—Šπ—‚π—‡π—(𝗂𝖽A)≃(∏(x:A)(x=x)). Now by definition we have

π—Šπ—‚π—‡π—(𝗂𝖽A)β‰‘βˆ‘g:Aβ†’A((gβˆΌπ—‚π–½A)Γ—(gβˆΌπ—‚π–½A)).

By function extensionality, this is equivalentMathworldPlanetmathPlanetmathPlanetmathPlanetmathPlanetmath to

βˆ‘g:Aβ†’A((g=𝗂𝖽A)Γ—(g=𝗂𝖽A)).

And by http://planetmath.org/node/87641Exercise 2.10, this is equivalent to

βˆ‘h:βˆ‘(g:Aβ†’A)(g=𝗂𝖽A)(𝗉𝗋1(h)=𝗂𝖽A)

However, by Lemma 3.11.8 (http://planetmath.org/311contractibility#Thmprelem5), βˆ‘(g:Aβ†’A)(g=𝗂𝖽A) is contractibleMathworldPlanetmath with center 𝗂𝖽A; therefore by Lemma 3.11.9 (http://planetmath.org/311contractibility#Thmprelem6) this type is equivalent to 𝗂𝖽A=𝗂𝖽A. And by function extensionality, 𝗂𝖽A=𝗂𝖽A is equivalent to ∏(x:A)x=x. ∎

We remark that http://planetmath.org/node/87864Exercise 4.3 asks for a proof of the above lemma which avoids univalence.

Thus, what we need is some A which admits a nontrivial element of ∏(x:A)(x=x). Thinking of A as a higher groupoidPlanetmathPlanetmathPlanetmathPlanetmathPlanetmathPlanetmath, an inhabitant of ∏(x:A)(x=x) is a natural transformation from the identity functor of A to itself. Such transformationsPlanetmathPlanetmath are said to form the center of a categoryMathworldPlanetmath, since the naturality axiom requires that they commute with all morphismsMathworldPlanetmath. Classically, if A is simply a group regarded as a one-object groupoid, then this yields precisely its center in the usual group-theoretic sense. This provides some motivation for the following.

Lemma 4.1.2.

Suppose we have a type A with a:A and q:a=a such that

  1. 1.

    The type a=a is a set.

  2. 2.

    For all x:A we have βˆ₯a=xβˆ₯.

  3. 3.

    For all p:a=a we have p⁒\centerdot⁒q=q⁒\centerdot⁒p.

Then there exists f:∏(x:A)(x=x) with f⁒(a)=q.

Proof.

Let g:∏(x:A)βˆ₯a=xβˆ₯ be as given byΒ 2. First we observe that each type x=Ay is a set. For since being a set is a mere proposition, we may apply the induction principle of propositional truncation, and assume that g(x)=|p| and g(y)=|q| for p:a=x and q:a=y. In this case, composing with p and q-1 yields an equivalence (x=y)≃(a=a). But (a=a) is a set byΒ 1, so (x=y) is also a set.

Now, we would like to define f by assigning to each x the path g⁒(x)-1⁒\centerdot⁒q⁒\centerdot⁒g⁒(x), but this does not work because g⁒(x) does not inhabit a=x but rather βˆ₯a=xβˆ₯, and the type (x=x) may not be a mere proposition, so we cannot use induction on propositional truncation. Instead we can apply the technique mentioned in Β§3.9 (http://planetmath.org/39theprincipleofuniquechoice): we characterize uniquely the object we wish to construct. Let us define, for each x:A, the type

B(x):β‰‘βˆ‘(r:x=x)∏(s:a=x)(r=s-1\centerdotq\centerdots).

We claim that B⁒(x) is a mere proposition for each x:A. Since this claim is itself a mere proposition, we may again apply induction on truncation and assume that g(x)=|p| for some p:a=x. Now suppose given (r,h) and (rβ€²,hβ€²) in B⁒(x); then we have

h⁒(p)⁒\centerdot⁒h′⁒(p)-1:r=rβ€².

It remains to show that h is identified with hβ€² when transported along this equality, which by transport in identity types and function types (Β§2.11 (http://planetmath.org/211identitytype),Β§2.9 (http://planetmath.org/29pitypesandthefunctionextensionalityaxiom)), reduces to showing

h⁒(s)=h⁒(p)⁒\centerdot⁒h′⁒(p)-1⁒\centerdot⁒h′⁒(s)

for any s:a=x. But each side of this is an equality between elements of (x=x), so it follows from our above observation that (x=x) is a set.

Thus, each B⁒(x) is a mere proposition; we claim that ∏(x:A)B⁒(x). Given x:A, we may now invoke the induction principle of propositional truncation to assume that g(x)=|p| for p:a=x. We define r:≑p-1\centerdotq\centerdotp; to inhabit B⁒(x) it remains to show that for any s:a=x we have r=s-1⁒\centerdot⁒q⁒\centerdot⁒s. Manipulating paths, this reduces to showing that q⁒\centerdot⁒(p⁒\centerdot⁒s-1)=(p⁒\centerdot⁒s-1)⁒\centerdot⁒q. But this is just an instance ofΒ 3. ∎

Theorem 4.1.3.

There exist types A and B and a function f:Aβ†’B such that qinv⁒(f) is not a mere proposition.

Proof.

It suffices to exhibit a type X such that ∏(x:X)(x=x) is not a mere proposition. Define X:β‰‘βˆ‘(A:𝒰)βˆ₯𝟐=Aβˆ₯, as in the proof of Lemma 3.8.5 (http://planetmath.org/38theaxiomofchoice#Thmprelem2). It will suffice to exhibit an f:∏(x:X)(x=x) which is unequal to λ⁒x.𝗋𝖾𝖿𝗅x.

Let a:≑(𝟐,|π—‹π–Ύπ–Ώπ—…πŸ|):X, and let q:a=a be the path corresponding to the nonidentity equivalence e:πŸβ‰ƒπŸ defined by e(0𝟐):≑1𝟐 and e(1𝟐):≑0𝟐. We would like to apply Lemma 4.1.2 (http://planetmath.org/41quasiinverses#Thmprelem2) to build an f. By definition of X, equalities in subset types (Β§3.5 (http://planetmath.org/35subsetsandpropositionalresizing)), and univalence, we have (a=a)≃(πŸβ‰ƒπŸ), which is a set, soΒ 1 holds. Similarly, by definition of X and equalities in subset types we haveΒ 2. Finally, http://planetmath.org/node/87644Exercise 2.13 implies that every equivalence πŸβ‰ƒπŸ is equal to either π—‚π–½πŸ or e, so we can showΒ 3 by a four-way case analysis.

Thus, we have f:∏(x:X)(x=x) such that f⁒(a)=q. Since e is not equal to π—‚π–½πŸ, q is not equal to 𝗋𝖾𝖿𝗅a, and thus f is not equal to λ⁒x.𝗋𝖾𝖿𝗅x. Therefore, ∏(x:X)(x=x) is not a mere proposition. ∎

More generally, Lemma 4.1.2 (http://planetmath.org/41quasiinverses#Thmprelem2) implies that any β€œEilenberg–Mac Lane space” K⁒(G,1), where G is a nontrivial abelian groupMathworldPlanetmath, will provide a counterexample; see http://planetmath.org/node/87582Chapter 8. The type X we used turns out to be equivalent to K⁒(β„€2,1). In http://planetmath.org/node/87579Chapter 6 we will see that the circle π•Š1=K⁒(β„€,1) is another easy-to-describe example.

We now move on to describing better notions of equivalence.

Title 4.1 Quasi-inverses
\metatable