Topos Theory — Elementary Toposes and Subobject Classifiers

⚡ Key Takeaways

An elementary topos is a cartesian closed category with finite limits that possesses a subobject classifier $\Omega$ — a single object that internally represents the notion of "subset" or "truth value." Key result: every subobject of an object $X$ corresponds bijectively to a morphism $X \to \Omega$ (its characteristic map), and $\Omega$ carries the structure of an internal Heyting algebra, making the internal logic of a topos intuitionistic. Applies to: categorical logic, sheaf theory, algebraic geometry, set theory (forcing), and constructive mathematics. Who/when: Lawvere and Tierney, c. 1969–1970, building on Grothendieck's sheaf-theoretic toposes.

💡 The Core Insight

The defining move of topos theory is to internalize the membership relation. In $\mathbf{Set}$ , a subset $S \subseteq X$ is the same data as its indicator function $\chi_S : X \to \{0,1\}$ . Topos theory takes this representability of subobjects as an axiom: there exists a universal object $\Omega$ and a universal "element" $\mathsf{true}: 1 \to \Omega$ such that every monomorphism is a pullback of $\mathsf{true}$ along a unique characteristic map. Once subobjects are represented by morphisms into $\Omega$ , logical operations on subobjects become algebraic operations on $\Omega$ , and the entire apparatus of logic becomes categorical.

DefinitionSubobject

For an object $X$ in a category $\mathcal{C}$ , a subobject of $X$ is an equivalence class of monomorphisms $m: S \rightarrowtail X$ , where $m \sim m'$ iff there is an isomorphism $\varphi$ with $m' \varphi = m$ . The collection of subobjects is denoted $\mathrm{Sub}(X)$ ; it is partially ordered by $[m] \le [m']$ iff $m$ factors through $m'$ .

DefinitionSubobject Classifier

A subobject classifier in a category $\mathcal{C}$ with finite limits is an object $\Omega$ together with a monomorphism $\mathsf{true}: 1 \rightarrowtail \Omega$ (from the terminal object $1$ ) such that for every monomorphism $m: S \rightarrowtail X$ there is a unique morphism $\chi_m : X \to \Omega$ , the characteristic map, making the following square a pullback:

1

\quad \begin{array}{ccc} S & \xrightarrow{\;!\;} & 1 \\ \;\downarrow{m} & & \downarrow{\mathsf{true}} \\ X & \xrightarrow{\;\chi_m\;} & \Omega \end{array}

Derivation hint: The pullback condition says $S$ is exactly the preimage $\chi_m^{-1}(\mathsf{true})$ ; uniqueness of $\chi_m$ encodes that subobjects are faithfully represented.

DefinitionElementary Topos

An elementary topos is a category $\mathcal{E}$ that 1. has all finite limits (equivalently a terminal object and pullbacks), 2. is cartesian closed: for every object $X$ the functor $(-) \times X$ has a right adjoint $(-)^X$ (exponentials exist), and 3. has a subobject classifier $\Omega$ . The cartesian closure is expressed by the natural bijection

2

\quad \mathrm{Hom}_{\mathcal{E}}(Z \times X,\, Y) \;\cong\; \mathrm{Hom}_{\mathcal{E}}(Z,\, Y^X).

Derivation hint: This is the currying isomorphism; it is natural in $Z$ , $X$ , $Y$ and exhibits $(-)^X \dashv (-)\times X$ .

TheoremPower Objects and Internal Hom into

\Omega

In any elementary topos, every object $X$ has a power object $P(X) = \Omega^X$ representing subobjects of $X \times (-)$ :

4

\quad \mathrm{Sub}(X \times Z) \;\cong\; \mathrm{Hom}_{\mathcal{E}}(Z,\, \Omega^X).

There is a universal membership subobject $\in_X \rightarrowtail X \times \Omega^X$ such that every relation arises by pullback along a unique classifying map $Z \to \Omega^X$ . Derivation hint: Combine [2] with $X \leftarrow Y$ replaced by $\Omega$ , then apply [3]; the membership relation is $\in_X = (\mathrm{ev})^{-1}(\mathsf{true})$ where $\mathrm{ev}: X \times \Omega^X \to \Omega$ is evaluation.

Theorem

\Omega

is an Internal Heyting Algebra

The object $\Omega$ carries internal operations $\wedge, \vee, \Rightarrow : \Omega \times \Omega \to \Omega$ and constants $\mathsf{true}, \mathsf{false}: 1 \to \Omega$ making $(\mathrm{Sub}(X), \le)$ a Heyting algebra for every $X$ , naturally in $X$ . The Heyting implication satisfies the adjunction

5

\quad a \wedge b \le c \iff a \le (b \Rightarrow c)

internally in each $\mathrm{Sub}(X)$ . Derivation hint: $\wedge$ classifies the subobject $\langle \mathsf{true}, \mathsf{true}\rangle: 1 \rightarrowtail \Omega\times\Omega$ ; $\Rightarrow$ classifies $\{(p,q): p\wedge q = p\}$ ; the adjunction [5] is verified by transposing pullbacks.

ProofConstruction of

\wedge

on

\Omega

and the Heyting structure

We build the conjunction operation and show it induces meets on subobjects.

Step 1 — Construct $\wedge$ . Form the morphism $\langle \mathsf{true}, \mathsf{true}\rangle : 1 \to \Omega \times \Omega$ . Because $1$ is terminal and the composite $1 \to \Omega \times \Omega$ is a monomorphism (it is split by either projection composed appropriately), it is a subobject of $\Omega \times \Omega$ . By the classifier axiom [1] there is a unique map

\wedge \;=\; \chi_{\langle\mathsf{true},\mathsf{true}\rangle} : \Omega \times \Omega \to \Omega.

Step 2 — Meets of subobjects. Given $[m], [n] \in \mathrm{Sub}(X)$ with characteristic maps $\chi_m, \chi_n : X \to \Omega$ , form $\langle \chi_m, \chi_n\rangle : X \to \Omega \times \Omega$ and compose with $\wedge$ . The pullback of $\mathsf{true}$ along $\wedge \circ \langle \chi_m, \chi_n\rangle$ is the simultaneous pullback of $\mathsf{true}$ along $\chi_m$ and along $\chi_n$ , i.e. the intersection $S \cap T$ as a subobject of $X$ . Hence

\chi_{m \wedge n} = \wedge \circ \langle \chi_m, \chi_n \rangle, \qquad m \wedge n = m \cap n.

Step 3 — Order. Define $\le$ on $\Omega$ as the subobject $\{(p,q) : p \wedge q = p\} \rightarrowtail \Omega\times\Omega$ , classified by some $\le\, : \Omega\times\Omega \to \Omega$ . Under [3] this recovers the partial order on $\mathrm{Sub}(X)$ .

Step 4 — Implication via the adjoint. The operation $(-) \wedge b : \mathrm{Sub}(X) \to \mathrm{Sub}(X)$ preserves all meets and (using cartesian closure and that $\mathrm{Sub}(X)$ is a complete lattice — a consequence of finite limits plus $\Omega^X$ ) has a right adjoint, the Heyting implication $b \Rightarrow (-)$ , giving [5]. The internal implication $\Rightarrow : \Omega\times\Omega \to \Omega$ classifies the subobject $\{(p,q) : p \le q\}$ pulled into the diagonal.

Step 5 — Distributivity and bottom. $\mathsf{false}: 1 \to \Omega$ is $\chi_{0 \rightarrowtail 1}$ , the characteristic map of the initial object's unique mono into $1$ . Distributivity of $\wedge$ over $\vee$ (where $\vee$ is built from the image of $\langle \mathrm{id},\mathsf{true}\rangle \sqcup \langle \mathsf{true},\mathrm{id}\rangle$ ) follows from the adjunction in Step 4. Thus each $\mathrm{Sub}(X)$ is a Heyting algebra. $\qquad\blacksquare$

∎

⚠️ Why Heyting and not Boolean

The proof never produces a complemented lattice. In general $a \vee \neg a \ne \top$ where $\neg a := (a \Rightarrow \bot)$ . A topos is Boolean exactly when $\Omega \cong 1 + 1$ (equivalently the law of excluded middle holds internally). Generic toposes — e.g. sheaves on a non-discrete space — are not Boolean.

mathematician · F. William Lawvere & Myles Tierney | 1937– / 1937–

Lawvere, following his 1963 thesis on functorial semantics, sought a purely categorical, elementary (first-order, finitary) axiomatization of "the category of sets" and of Grothendieck toposes. Working with Tierney around 1969–1970 at Dalhousie University, he isolated the subobject classifier as the crucial new ingredient, replacing the infinitary descent data of Grothendieck topologies with a single object $\Omega$ and the operator $j: \Omega \to \Omega$ (a Lawvere–Tierney topology). This recast sheafification, forcing, and intuitionistic higher-order logic in a uniform categorical language, and turned topos theory into a foundation for mathematics.

🧠 Intuition

Think of $\Omega$ as the object of truth values internal to the world $\mathcal{E}$ . In $\mathbf{Set}$ , $\Omega = \{0,1\}$ and $\chi_S$ is the indicator function — asking "is $x \in S$ ?" returns a value in $\{$ yes, no $\}$ . In a sheaf topos $\mathbf{Sh}(X)$ , the truth values are the open sets of $X$ : the "truth value of $x \in S$ " is the largest open set on which the property holds. So truth becomes local and stagewise, not global. This is why excluded middle fails: a property can hold on a dense open set without holding everywhere, so " $P$ or not $P$ " need not be globally true. Power objects $\Omega^X$ are the internal "power sets," and exponentials $Y^X$ are internal function sets, so a topos genuinely is a self-contained universe of (constructive) sets.

Figure

The subobject classifier pullback square: a subobject as the preimage of

\mathsf{true}

ExampleThe classifier in

\mathbf{Set}

Given: $\mathcal{E} = \mathbf{Set}$ , take $X = \{1,2,3,4\}$ and the subset $S = \{2,4\}$ with inclusion $m: S \rightarrowtail X$ .

Setup: $1 = \{\ast\}$ , and we claim $\Omega = \{0,1\}$ with $\mathsf{true}: \ast \mapsto 1$ .

Steps: 1. Define $\chi_m : X \to \{0,1\}$ by $\chi_m(x) = 1$ if $x \in S$ , else $0$ . Explicitly $\chi_m(1)=0,\ \chi_m(2)=1,\ \chi_m(3)=0,\ \chi_m(4)=1$ . 2. Compute the pullback of $\mathsf{true}$ along $\chi_m$ : it is $\{x : \chi_m(x) = 1\} = \{2,4\} = S$ . ✓ 3. Uniqueness: any $\chi'$ with this pullback must send $S \to 1$ and $X\setminus S \to 0$ ; this is forced, so $\chi' = \chi_m$ . ✓ 4. Heyting structure: $\wedge$ is multiplication on $\{0,1\}$ , $\vee = \max$ , $\neg 0 = 1, \neg 1 = 0$ . Here $\Omega = 1+1$ , so $\mathbf{Set}$ is Boolean and excluded middle holds.

\boxed{\;\Omega_{\mathbf{Set}} = \{0,1\},\quad |\mathrm{Sub}(X)| = 2^{|X|} = 2^4 = 16\;}

ExampleThe classifier in the topos of

G

-sets and in presheaves

Given: Let $G = \mathbb{Z}/2 = \{e, g\}$ and let $\mathcal{E} = \mathbf{Set}^{G}$ be the topos of left $G$ -sets (functors $BG \to \mathbf{Set}$ ).

Question: What is $\Omega$ , and how many subobjects does the regular representation $G$ (acting on itself) have?

Steps: 1. A subobject of a $G$ -set $X$ is a $G$ -invariant subset. The classifier must record, for each point, "the set of $h \in G$ for which $h\cdot x$ lies in the subset" — i.e. truth values are subobjects of $G$ in the category, namely the set of subgroups-with-coset structure. For a general topos $\widehat{\mathcal{C}} = \mathbf{Set}^{\mathcal{C}^{\mathrm{op}}}$ , $\Omega(c) = \{ \text{sieves on } c \}$ . 2. For $BG$ (one object, morphisms $= G$ ), a sieve is a right-ideal $\subseteq G$ ; the right ideals of a group are $\varnothing$ and $G$ . So $\Omega = \{0,1\}$ as a set, but with $G$ acting trivially — $\Omega \cong 1 + 1$ and $\mathbf{Set}^G$ is in fact Boolean (groups are groupoids). 3. Subobjects of the regular $G$ -set $X = G$ : invariant subsets are unions of orbits. $G$ acting on itself by left multiplication is transitive (one orbit), so $\mathrm{Sub}(G) = \{\varnothing, G\}$ .

\boxed{\;\Omega_{\mathbf{Set}^{G}} \cong 1+1 \ (\text{trivial action}),\qquad |\mathrm{Sub}(G)| = 2\;}

Contrast (non-Boolean): in presheaves on the poset $\bullet \to \bullet$ (the "walking arrow"), $\Omega$ has three global truth values $\{0, \tfrac12, 1\}$ — false, "true later", true — and the topos is not Boolean.

⚠️

\Omega

need not be

1+1

A frequent error is to assume $\Omega \cong \{ \text{true},\text{false}\}$ in every topos. This holds only in Boolean toposes. In $\mathbf{Sh}(X)$ , $\Omega$ is the sheaf of open sets ( $\Omega(U) = \{$ open $V \subseteq U\}$ ), which can have uncountably many sections. Treating $\Omega$ as a two-element set destroys the intuitionistic content.

⚠️ Monos vs. injections, epis vs. surjections

In a general topos monomorphisms and epimorphisms are defined by cancellation, not by underlying-set behavior. They do coincide with "pointwise injective/surjective" in $\mathbf{Set}$ and many functor categories, but the classifier axiom is about monos, and one must verify a map is a mono categorically (left-cancellable) before forming $\chi_m$ .

CounterexampleA cartesian closed category that is NOT a topos

The category $\mathbf{Cat}$ of small categories is cartesian closed (the exponential $\mathcal{D}^{\mathcal{C}}$ is the functor category) and has finite limits. But $\mathbf{Cat}$ has no subobject classifier: subcategories of a category $\mathcal{C}$ cannot be uniformly classified by maps into a single object, because a "sub-thing" must specify both objects and arrows with closure conditions that are not representable. Hence $\mathbf{Cat}$ is cartesian closed but not a topos — showing axioms (1),(2) do not imply (3).

CounterexampleA category with a "classifier-like" object failing uniqueness

In the category of topological spaces $\mathbf{Top}$ , one might hope the Sierpiński space $\mathbb{S}$ classifies open subobjects. Indeed open inclusions $U \hookrightarrow X$ correspond to continuous maps $X \to \mathbb{S}$ . But not every monomorphism in $\mathbf{Top}$ is an open inclusion (e.g. inclusion of a non-open subspace), and those are not classified. So $\mathbf{Top}$ fails the classifier axiom for general monos: $\mathbb{S}$ classifies only the open subobjects, not all subobjects. $\mathbf{Top}$ is not a topos.

connection

Elementary toposes sit at the crossroads of several fields. The Grothendieck toposes — categories of sheaves $\mathbf{Sh}(\mathcal{C}, J)$ on a site — are exactly the elementary toposes that are also cocomplete and have a generating set (Giraud's theorem), linking to Sheaf Theory and Grothendieck Topology. The internal Heyting algebra of $\Omega$ ties the subject to Heyting Algebra and Intuitionistic Logic: the Mitchell–Bénabou language makes any topos a model of higher-order intuitionistic type theory, connecting to Categorical Logic and the Curry–Howard Correspondence. A Lawvere–Tierney topology $j: \Omega \to \Omega$ generalizes Grothendieck topologies and yields sheaf subtoposes, the categorical face of Forcing in set theory. Cartesian closure [2] relates $\Omega^X$ to the Cartesian Closed Category notion underlying the Lambda Calculus.

🔬 Where this appears

⏳ History

- 1957–1963 — Grothendieck. In the Tôhoku paper and SGA, Grothendieck develops sheaves on sites; a topos is born as "a category of sheaves," the natural home for cohomology in algebraic geometry.

- 1963 — Lawvere's thesis. Functorial Semantics of Algebraic Theories recasts model theory categorically, seeding the program of axiomatizing mathematics in categorical terms.

- 1964 — Lawvere, ETCS. An Elementary Theory of the Category of Sets axiomatizes

\mathbf{Set}

categorically, a precursor to the topos axioms.

- 1969–1970 — Lawvere & Tierney. Identification of the subobject classifier

\Omega

and the operator

j

; the elementary (finitary, first-order) definition of a topos is given, divorcing it from Grothendieck's site data.

- 1971 — Lawvere–Tierney sheaves. Sheafification and forcing reconstructed via topologies

j:\Omega\to\Omega

; Cohen forcing seen topos-theoretically.

- 1972 — Freyd, Kock, Wraith, Mikkelsen. Proofs that elementary toposes have finite colimits and are exact; Mikkelsen shows power objects suffice.

- 1977 — Johnstone, Topos Theory. First comprehensive monograph systematizing the field.

- 1980s — Hyland, the effective topos. Realizability toposes connect topos theory to recursion theory and constructive mathematics.

- 1992 / 2002 — Mac Lane–Moerdijk; Johnstone's Elephant. Sheaves in Geometry and Logic and the two-volume Sketches of an Elephant become standard references, consolidating the subject.

open-problem

When is a topos Boolean, and how "non-classical" can it be? The internal logic is intuitionistic; classifying exactly which categorical conditions force the law of excluded middle, the axiom of choice, or De Morgan's laws remains a delicate study (De Morgan toposes, Boolean toposes, $\neg\neg$ -sheaves). Higher topos theory: the passage to $(\infty,1)$ -toposes (Lurie) reframes $\Omega$ as an object classifier and is central to homotopy type theory; the precise elementary axioms for an elementary $\infty$ -topos are still being worked out (Rasekh, Shulman). Independence and large cardinals: how topos-theoretic forcing relates to the full strength of class forcing and large-cardinal hypotheses is an active interface. Predicativity: whether key topos constructions can be carried out predicatively (without impredicative power objects) motivates predicative alternatives like Σ-frames and arithmetic universes, an ongoing program of Joyal and Maietti.

Topos Theory — Elementary Toposes and Subobject Classifiers

Formal Statement

Why It Holds / Derivation

Building Intuition

Worked Examples

Common Misconceptions

Connections

Real-World Applications

Historical Development

Subtleties & Limits

See Also