nLab cone type

Idea

The cone type is an axiomatization of the cone in the context of homotopy type theory.

Definition

A cone type on a type $A$ is a higher inductive type $Cone(A)$ inductively generated by

• A term $e:Cone(A)$

• A function $i: A \to Cone(A)$

• A term

$$p: \prod_{a:A} e = i(a)$$

In Coq pseudocode, the cone is given by

Inductive Cone (A : Type) : Type
  | vertex : Cone A
  | base :  A -> Cone A
  | edge : A -> Id Cone A vertex base(a)

It can equivalently be defined as the (homotopy) pushout of $A$ and the unit type $1$ under $A$. Similarly, it is the cofiber of the identity function of $A$. This definition makes it clear that the cone type is always a contractible type. As a result, cone types could be though of as a way of constructing free contractible types for any type $A$.

The cone type of $A$ can also be defined as the localization of $A$ at the empty type, $\mathrm{Cone}(A) \coloneqq L_\emptyset(A)$, and thus it is the (-2)-truncation of $A$.

Inference rules

Formation rules for cone types:

$$\frac{\Gamma \vdash A \; \mathrm{type}}{\Gamma \vdash \mathrm{cone}(A)}$$

Introduction rules for cone types:

$$\frac{\Gamma \vdash A \; \mathrm{type}}{\Gamma \vdash *:\mathrm{cone}(A)}$$

$$\frac{\Gamma \vdash A \; \mathrm{type}}{\Gamma \vdash i:A \to \mathrm{cone}(A)}$$

$$\frac{\Gamma \vdash A \; \mathrm{type}}{\Gamma \vdash \mathcal{g}:\prod_{x:A} * =_{\mathrm{cone}(A)} i(x)}$$

Elimination rules for cone types:

$$\frac{\begin{array}{l} \Gamma \vdash A \; \mathrm{type} \quad \Gamma, y:\mathrm{cone}(A) \vdash C(y) \; \mathrm{type} \\ \Gamma \vdash c_*:C(*) \quad \Gamma \vdash c_i:\prod_{x:A} C(i(x)) \\ \Gamma \vdash c_\mathcal{g}:\prod_{x:A} c_* =_{y:\mathrm{cone}(A).C(y)}^{\mathcal{g}(x)} c_i(x) \end{array}}{\Gamma \vdash \mathrm{ind}_{\mathrm{cone}(A)}(c_*, c_i, c_\mathcal{g}):\prod_{y:\mathrm{cone}(A)} C(y)}$$

Computation rules for cone types:

• Judgmental computation rules

$$\frac{\begin{array}{l} \Gamma \vdash A \; \mathrm{type} \quad \Gamma, y:\mathrm{cone}(A) \vdash C(y) \; \mathrm{type} \\ \Gamma \vdash c_*:C(*) \quad \Gamma \vdash c_i:\prod_{x:A} C(i(x)) \\ \Gamma \vdash c_\mathcal{g}:\prod_{x:A} c_* =_{y:\mathrm{cone}(A).C(y)}^{\mathcal{g}(x)} c_i(x) \end{array}}{\Gamma \vdash \mathrm{ind}_{\mathrm{cone}(A)}(c_*, c_i, c_\mathcal{g})(*) \equiv c_*:C(*)}$$

$$\frac{\begin{array}{l} \Gamma \vdash A \; \mathrm{type} \quad \Gamma, y:\mathrm{cone}(A) \vdash C(y) \; \mathrm{type} \\ \Gamma \vdash c_*:C(*) \quad \Gamma \vdash c_i:\prod_{x:A} C(i(x)) \\ \Gamma \vdash c_\mathcal{g}:\prod_{x:A} c_* =_{y:\mathrm{cone}(A).C(y)}^{\mathcal{g}(x)} c_i(x) \end{array}}{\Gamma, x:A \vdash \mathrm{ind}_{\mathrm{cone}(A)}(c_*, c_i, c_\mathcal{g})(i(x)) \equiv c_i(x):C(i(x))}$$

$$\frac{\begin{array}{l} \Gamma \vdash A \; \mathrm{type} \quad \Gamma, y:\mathrm{cone}(A) \vdash C(y) \; \mathrm{type} \\ \Gamma \vdash c_*:C(*) \quad \Gamma \vdash c_i:\prod_{x:A} C(i(x)) \\ \Gamma \vdash c_\mathcal{g}:\prod_{x:A} \end{array}}{\Gamma, x:A \vdash \mathrm{apd}_{y:\mathrm{cone}(A).C(y)}(\mathrm{ind}_{\mathrm{cone}(A)}(c_*, c_i, c_\mathcal{g}), *, i(x), \mathcal{g}(x)) \equiv c_\mathcal{g}(x):c_* =_{y:\mathrm{cone}(A).C(y)}^{\mathcal{g}(x)} c_i(x)}$$

• Propositional computation rules

$$\frac{\begin{array}{l} \Gamma \vdash A \; \mathrm{type} \quad \Gamma, y:\mathrm{cone}(A) \vdash C(y) \; \mathrm{type} \\ \Gamma \vdash c_*:C(*) \quad \Gamma \vdash c_i:\prod_{x:A} C(i(x)) \\ \Gamma \vdash c_\mathcal{g}:\prod_{x:A} c_* =_{y:\mathrm{cone}(A).C(y)}^{\mathcal{g}(x)} c_i(x) \end{array}}{\Gamma \vdash \beta_{\mathrm{cone}(A)}^{*}(c_*, c_i, c_\mathcal{g}):\mathrm{ind}_{\mathrm{cone}(A)}(c_*, c_i, c_\mathcal{g})(*) =_{C(*)} c_*}$$

$$\frac{\begin{array}{l} \Gamma \vdash A \; \mathrm{type} \quad \Gamma, y:\mathrm{cone}(A) \vdash C(y) \; \mathrm{type} \\ \Gamma \vdash c_*:C(*) \quad \Gamma \vdash c_i:\prod_{x:A} C(i(x)) \\ \Gamma \vdash c_\mathcal{g}:\prod_{x:A} c_* =_{y:\mathrm{cone}(A).C(y)}^{\mathcal{g}(x)} c_i(x) \end{array}}{\Gamma \vdash \beta_{\mathrm{cone}(A)}^{i}(c_*, c_i, c_\mathcal{g}):\prod_{x:A} \mathrm{ind}_{\mathrm{cone}(A)}(c_*, c_i, c_\mathcal{g})(i(x)) =_{C(i(x))} c_i(x)}$$

$$\frac{\begin{array}{l} \Gamma \vdash A \; \mathrm{type} \quad \Gamma, y:\mathrm{cone}(A) \vdash C(y) \; \mathrm{type} \\ \Gamma \vdash c_*:C(*) \quad \Gamma \vdash c_i:\prod_{x:A} C(i(x)) \\ \Gamma \vdash c_\mathcal{g}:\prod_{x:A} c_* =_{y:\mathrm{cone}(A).C(y)}^{\mathcal{g}(x)} c_i(x) \end{array}}{\Gamma \vdash \beta_{\mathrm{cone}(A)}^{\mathcal{g}}(c_*, c_i, c_\mathcal{g}):\prod_{x:A} \mathrm{apd}_{y:\mathrm{cone}(A).C(y)}(\mathrm{ind}_{\mathrm{cone}(A)}(c_*, c_i, c_\mathcal{g}), *, i(x), \mathcal{g}(x)) =_{y:\mathrm{cone}(A).C(y)}^{\mathcal{g}(x)} c_\mathcal{g}(x)}$$

Uniqueness rules for cone types:

$$\frac{\begin{array}{l} \Gamma \vdash A \; \mathrm{type} \quad \Gamma, y:\mathrm{cone}(A) \vdash C(y) \; \mathrm{type} \\ \Gamma \vdash c:\prod_{y:\mathrm{cone}(A)} C(y) \quad \Gamma \vdash p:\mathrm{cone}(A) \\ \Gamma \vdash l:\prod_{x:A} p =_{\mathrm{cone}(A)} i(x) \end{array}}{\Gamma \vdash \eta_{\mathrm{cone}(A)}(c, p, l):\mathrm{ind}_{\mathrm{cone}(A)}(c(p), \lambda x:A.c(i(x)), \lambda x:A.\mathrm{apd}_{y:\mathrm{cone}(A).C(y)}(c, p, i(x), l(x)))(p) =_{C(p)} c(p)}$$

Properties

Propositionally constant functions

In dependent type theory, a propositionally constant function from type $A$ to $B$ at an element $b:B$ is a function $f:A \to B$ with a dependent function $p:\prod_{x:A} f(x) =_B b$.

Given a propositionally constant function at $b:B$, $f:A \to B$, with $p:\prod_{x:A} f(x) =_B b$, by the recursion principle of the cone type, one has

$$\mathrm{rec}_{\mathrm{Cone}(A)}^B(b, f, p):\mathrm{Cone}(A) \to B$$

such that

$$\mathrm{rec}_{\mathrm{Cone}(A)}^B(b, f, p)(\mathrm{vertex}) \equiv b:B$$

$$\mathrm{rec}_{\mathrm{Cone}(A)}^B(b, f, p)(\mathrm{base}(x)) \equiv f(x):B$$

$$\mathrm{ap}_{\mathrm{rec}_{\mathrm{Cone}(A)}^B(b, f, p)}(\mathrm{edge}(x)) \equiv p:f(x) =_B b$$

In particular, by the judgmental congruence rule for the introduction rule of function types, the standard constant function $\lambda x:A.b:A \to B$ for $b:B$ with $\beta_{A, B}^{x:A.b}:\prod_{x:A} (\lambda x:A.b)(x) =_B b$ from the typal computation rules for function types factors through $\mathrm{Cone}(A)$ by

$$\lambda x:A.b \equiv \mathrm{rec}_{\mathrm{Cone}(A)}^B(b, \lambda x:A.b, \beta_{A, B}^{x:A.b}) \circ \mathrm{base}$$

where for strict function types we have $\beta_{A, B}^{x:A.b} \equiv \lambda x:A.\mathrm{refl}_B(b)$.

Propositionally constant functions can also be regarded directly as functions $\mathrm{Cone}(A) \to B$, similar to how paths in $A$ can be regarded as functions $\mathbb{I} \to A$ from the the interval type rather than the application of said function along the path generator of the interval type.

Since the cone type is a contractible type, it is equivalent to the unit type, with equivalences $e:\mathrm{Cone}(A) \simeq \mathbb{1}$, and every constant function $(b, f, p)$ factors the unit type through $e \circ \mathrm{base}:A \to \mathbb{1}$ and $\mathrm{rec}_{\mathrm{Cone}(A)}^B(b, f, p) \circ e^{-1}:\mathbb{1} \to B$.

Examples

• The unit type $1$ is the cone type of an empty type $0$.

• The interval type $I$ is the cone type of the unit type $1$

• A simplex type? $\Delta_n$ is the cone type of the simplex type $\Delta_{n-1}$. $\Delta_1 = I$ and $\Delta_0 = 1$.

See also

• cone

• constant function

• empty type, localization of a type

• cone type, bracket type, set truncation