Kan Fibration¶

A Kan fibration is a simplicial-set map that lifts every compatible horn over a specified simplex in its target.

Version: v1 · 2026-10-03 · History
Domain-specific #: 13358
Domain group: Formal Sciences
Origin domain: Mathematics
Subdomains: Simplicial Homotopy Theory, Algebraic Topology → Mathematics
Aliases: Kan fibration of simplicial sets

Core Idea¶

A Kan fibration is a map \(p:X\to S\) of simplicial sets with a particular relative filling property. For every dimension \(n>0\), every position \(0\leq i\leq n\), a horn \(\Lambda_i^n\) mapped into \(X\), and a full \(n\)-simplex of \(S\) whose boundary data agree with the horn's image, there is a full \(n\)-simplex in \(X\) that both extends the given horn and maps to the chosen simplex of \(S\). The property is a right lifting property against every horn inclusion, not a claim of unique fillers.^[1]^[2]

A horn is the union of all but one of a simplex's codimension-one faces, omitting its interior. The missing face and interior are what the lift must supply. Both inner and outer horns participate. If the base \(S\) is a point, its simplex is fixed; the same condition reduces to the ordinary horn-filling property of \(X\). Thus a Kan complex is a simplicial set whose map to a point is a Kan fibration, not another name for every Kan fibration.^[3]^[1]

Structural Signature¶

Simplicial map: \(p:X\to S\) identifies the total object and the base. The relation is about a map, not only about one simplicial set.
Partial simplex: \(\Lambda_i^n\to X\) prescribes every face except one, subject to the compatibility inherent in being a horn map.
Specified base filling: \(\Delta^n\to S\) extends the image of that horn; this is the particular target simplex over which lifting must occur.
Compatible lift: \(\Delta^n\to X\) extends the source horn and composes with \(p\) to the specified base simplex.
Universal test: the compatible lift must exist for all dimensions \(n>0\), all horn positions, and all commutative squares of this form.

Condensed: every horn over a chosen filled base simplex admits a matching full simplex upstairs.^[1]

Sig role-phrases: simplicial map; prescribed horn; specified base simplex; compatible lift; every dimension and horn position.

What It Is Not¶

Not just a Kan complex. That is the special case \(X\to *\); an arbitrary Kan fibration is a relative map between simplicial sets.^[1]
Not merely inner-horn lifting. Dropping the outer horns weakens the defining test and can yield an inner fibration without a Kan fibration.^[3]
Not an arbitrary filler. A simplex in \(X\) that extends the source horn but maps to a different base simplex does not solve the selected lifting square.
Not a unique-lift property. The definition guarantees existence, not uniqueness, a canonical selection, or an algorithm for constructing a filler.^[1]
Not Kan extension. A Kan extension is a separate categorical universal construction sharing Kan's name, not this horn-lifting condition.

Scope of Application¶

In simplicial homotopy theory, Kan fibrations single out maps for which every partial simplicial lift over a specified base simplex can be completed. This makes a map-level extension test available instead of only an object-level statement. Pullbacks and composites of Kan fibrations are again Kan fibrations, giving the class operational stability under two common constructions.^[2]

For simplicial groups, a theorem establishes that each such group is a Kan complex. Its projection to the point is therefore a Kan fibration. For simplicial abelian groups, a degreewise-surjective homomorphism supplies a different, genuinely relative example of a Kan fibration of underlying simplicial sets.^[2]

These are mathematical settings, not claims that every structure called a fibration in topology, graph theory or category theory satisfies this exact simplicial-horn test.

Clarity¶

Write the whole commuting square rather than saying “horns fill.” The phrase alone hides two questions: is the partial simplex in the total object, and which completed simplex in the base must its image match? Once both maps are given, failure of one compatible lift is a counterexample to the Kan-fibration property. This also prevents confusing a fibration of a map with fibrancy of one object.

The notation \(\Lambda_i^n\) matters because the condition covers every \(i\), including the outer positions. Restricting to \(0<i<n\) changes the claim.^[3]

Two tempting substitutions are logical mistakes, not structural tradeoffs. A filler of the source horn that maps to a different base simplex does not solve the given relative square. Conversely, the right lifting property says a solution exists; it does not specify a canonical algorithm or require uniqueness. These tests belong to the definition and to how one proves it, respectively.^[1]

Manages Complexity¶

An unbounded family of local extension questions is compressed into one reusable lifting predicate on a map. The compression does not make the verification trivial: a proof may exploit algebraic structure, known closure theorems or a construction of fillers. But it gives a disciplined interface. Once a map is known to satisfy the predicate, its pullbacks and composites can be recognized without re-deriving the definition from scratch.^[2]

Abstract Reasoning¶

To test a proposed \(p:X\to S\), choose a dimension and horn position, then specify a horn in \(X\) and a full simplex in \(S\) with matching boundary data. Seek a simplex in \(X\) that simultaneously extends the horn and lies over that exact base simplex. If one square has no lift, reject Kan-fibration status. If a theorem proves lifts for the entire universally quantified class, accept it. A proof in the special base-point case shows the domain is a Kan complex; it does not by itself prove an unrelated map from that domain is a Kan fibration.

The test can also run through structure rather than direct simplex search. The Stacks Project proves a degreewise-surjective morphism of simplicial abelian groups is a Kan fibration, and it proves pullback stability. Those are ways to warrant the universal property for a new case, not changes to the property's definition.^[2]

Knowledge Transfer¶

Within simplicial mathematics, the same lifting square applies to maps arising from algebraic objects and to maps studied in homotopy theory. What transfers literally is the horn inclusion, commuting-square condition and compatible lift; the proof of existence may differ. “Fibration” in another mathematical setting can be an analogy or a related formal class, but it does not inherit Kan status merely from the label.

Examples¶

An identity map: an outer horn lifts¶

Take \(p=\mathrm{id}_{\Delta^1}:\Delta^1\to\Delta^1\). Kerodon states that every simplicial-set isomorphism is a Kan fibration; the following \(n=1\) square shows the mechanism rather than replacing its universal theorem. Let \(\Lambda^1_0=\{0\}\) map to vertex \(0\) upstairs and let the specified base simplex be the nondegenerate edge \(e:0\to1\). The square commutes at vertex \(0\). Choose the lift to be that same edge \(e\) upstairs: its restriction is the selected vertex and \(p(e)=e\). For any horn and any dimension, the same identity-map argument makes the specified base simplex itself the unique possible lift.^[1]^[3]

Mapped back: identity simplicial map → initial-vertex horn → specified edge \(0\to1\) → that very edge upstairs → the same construction works for every horn.

A failed outer-horn square¶

Now let \(p:\Delta^0\hookrightarrow\Delta^1\) include only vertex \(0\). Prescribe the same horn \(\Lambda^1_0=\{0\}\to\Delta^0\) and the full base edge \(e:0\to1\). Compatibility still holds on the horn. But every \(1\)-simplex of \(\Delta^0\) is degenerate and maps to the constant edge at \(0\), not to \(e\). There is no diagonal \(\Delta^1\to\Delta^0\) over \(e\); this single square refutes Kan-fibration status. This calculation is author-derived from the formal definition and Kerodon's stated inclusion criterion; it is not presented as a quoted worked example.^[1]^[3]

Mapped back: vertex inclusion → selected initial vertex → nondegenerate base edge → no matching source edge → one failed outer horn is decisive.

A relative abelian-group quotient¶

Let \(A\) and \(B\) be the constant simplicial abelian groups on \(\mathbb Z\) and \(\mathbb Z/2\), with every face and degeneracy map the identity on its group, and let \(q:A\to B\) be reduction modulo \(2\) in every degree. For a visible \(n=1\) square, prescribe the horn \(\Lambda^1_0=\{0\}\) by the integer \(3\), and choose the base \(1\)-simplex \(1\pmod2\). It is a degenerate edge in the constant simplicial set, and its initial vertex equals \(q(3)\). The integer \(3\), considered as a (degenerate) \(1\)-simplex of \(A\), is a lift: its initial face is \(3\) and \(q(3)=1\pmod2\). The same check works at the other outer position. This elementary square does not by itself prove all-horn lifting; Stacks Lemma 14.31.7 does, because \(q\) is degreewise surjective.^[2]^[3]

Mapped back: quotient \(q:\mathbb Z\to\mathbb Z/2\) in each degree → horn value \(3\) → chosen base \(1\)-simplex \(1\pmod2\) → lift \(3\) → the source theorem covers every dimension.

Structural Tensions¶

No universal intrinsic two-sided cost is established for this formal map property. Relative compatibility, all-horn quantification and existence without uniqueness are definition boundaries, not competing objectives. A particular proof technique may carry computational costs, but those costs do not constitute a tension in the identity of a Kan fibration.^[1]

Structural–Framed Character¶

Kan fibration is strongly structural on the structural–framed spectrum. Its condition is a formal quantified relation, independent of evaluative weight or a human preference about good maps. Mathematicians choose why to study it, but that choice does not change whether a given square lifts. The name has a historical origin, not institutional authority that constitutes the property. Vocabulary can travel among mathematical texts while the exact horn and lift types stay fixed; applying “Kan-like” to an informal completion process imports an analogy rather than recognizing this same object. The portable skeleton is a relative extension or lifting relation, but its presence elsewhere does not grant this named all-horn simplicial condition cross-domain reach. Its character: a formally structural, domain-specific map property in simplicial mathematics.

Structural Core vs. Domain Accent¶

The abstract skeleton is a partial compatible specification + completed image → compatible completion above. Whether that compatible-lifting relation has an independently evidenced cross-domain prime identity is a future-prime question, not an existing parent asserted here. The actual Simplicial Set prerequisite supplies the typed carrier, not generic lifting. The domain accent is decisive: the partial specifications are simplicial horns, the completed images are simplices of a base simplicial set, and every dimension and inner or outer horn is quantified. Without these carriers and the right lifting property, the name Kan fibration no longer applies. Its possible broader analogy cannot promote this typed notion to prime status.

This entry presupposes Simplicial set.

The staged composition/presupposes parent is Simplicial Set: the fibration condition is stated for a map between simplicial sets and requires their face/degeneracy structure. It is not a strict subtype of a simplicial set, and not every simplicial set or map to a point has the Kan lifting property. The live Kan Extension entry is categorically distinct despite the shared name; Fibrations of Graphs use a different carrier.

Relationships to Other Abstractions¶

Current abstraction Kan Fibration Domain-specific

Parents (1) — more general patterns this builds on

Kan Fibration presupposes Simplicial set Domain-specific

Kan fibrations are lifting properties of maps between simplicial sets.

Hierarchy path (1) — routes to 1 parentless root

Kan Fibration → Simplicial set → Compositionality

Neighborhood in Abstraction Space¶

Kan Fibration sits in a sparse region of the domain-specific corpus (93^rd percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Unclustered & Miscellaneous (2551 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-10-08

Not to Be Confused With¶

Kan complex is the object satisfying horn filling, recovered as the map-to-point case. Inner fibration tests only inner horns. Kan extension extends functors by a universal property. Graph fibration is typed over graphs. None of these names alone proves that a given simplicial map lifts every horn over every compatible base simplex.

References¶

[1] Kerodon, “Kan Fibrations,” Definition 3.1.1.1 and examples. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i

[2] The Stacks Project, Section 14.31, “Kan fibrations,” definition and lemmas. registry ↩a ↩b ↩c ↩d ↩e ↩f

[3] Kerodon, “Horns,” Construction 1.2.4.1 and remarks. registry ↩a ↩b ↩c ↩d ↩e ↩f