How statement and proof provenance work
The first chip identifies the source of the statement or construction; the second identifies the source of its local proof or verification.
- Literature-sourced: the exact statement appears in a cited source; only wording and notation differ.
- AI-adapted: a semantically identical restatement of literature-sourced material, modulo indexing, notation, and boundary cases adopted by the library.
- AI-generated: a genuinely novel statement formulated by AI, with no source for the claim itself.
These labels describe origin, not correctness: citations and verification chips remain separate evidence.
For every , the class forbidding and has the strong Erdős–Hajnal property
Statement
For every integer , there exists a real constant such that every finite graph with no induced and no induced and with contains disjoint sets with
and such that is a pure pair. Equivalently, the hereditary class forbidding and has the strong Erdős–Hajnal property.
Facts & Assumptions
Given: An integer .
For every graph and every there exists such that every nonempty -free graph has a linearly large vertex set whose self-density is at most or at least (The edge-density form of Rödl's theorem: every nonempty -free graph has a linearly large set of self-density at most or at least ).
If a nonempty set has self-density at most , then it has a subset of at least half its size that is -sparse (A set of self-density at most has a subset of at least half its size that is -sparse).
Connected components partition the vertex set, and distinct connected components are anticomplete (The connected components of a graph partition its vertex set and are its maximal connected subgraphs, Distinct connected components are anticomplete, and distinct anticonnected components are complete).
A graph class has the strong Erdős–Hajnal property exactly when some linear constant works for every nontrivial graph in the class (The strong Erdős–Hajnal property for a hereditary graph class).
A set is -sparse exactly when every vertex of its induced subgraph has degree at most times the set size (-sparse, -dense and -restricted vertex sets).
Proof
We first prove the connected-case claim: for each there are constants and such that every connected graph on vertices has a vertex of degree greater than , or contains an induced starting at every vertex, or has a biclique of size at least in . We prove this by induction on .
For , choose and . Every vertex of a connected graph on at least two vertices is incident with an edge, so every vertex starts an induced . Thus the claim holds for .
Fix , assume the claim for , and let . Since , choose any constant . Now let be a connected graph on vertices for which the first outcome fails, so every vertex has degree at most . Fix a vertex and put . Then . Since is connected, has degree at least , so and therefore .
Suppose every connected component of has size at most . Choose components greedily until their union has size in the interval : if the running union first reaches before it exceeds , stop there; otherwise the next component itself has size in and we take that one alone. Let . Then , and [L3] makes anticomplete to in . Hence and form a biclique of size at least in .
Suppose instead that has a connected component with . Because is connected, some vertex has a neighbour in . Let , which is connected. Every vertex of still has degree at most , and because . So the first outcome of the induction hypothesis is false for .
Apply the induction hypothesis to with parameter . If outcome 2 holds there, then contains an induced starting at , and prefixing this path with gives an induced in starting at because is adjacent to and has no neighbours in . If outcome 3 holds there, then contains a biclique of size at least , and the same biclique lies in . Thus, whenever outcome 1 fails for , either outcome 2 or outcome 3 follows. This completes the induction and proves the connected-case claim.
Let . Because is the standard -vertex path, [L1] applied to yields a constant such that every nonempty -free graph has a vertex set of size at least whose self-density is at most or at least .
Let be a graph with no induced and no induced , and let . If has a set of size at least and self-density at least , apply the same argument to : an induced in would be an induced in , and an induced in would be an induced in . So belongs to the same forbidden class. Replacing by if necessary, we may assume that has a set with and self-density at most .
By [L2], the set contains a subset with such that is -sparse. By [F1], every vertex of therefore has degree at most .
If every connected component of has size at most , then the same greedy argument as in step 4.1 partitions those components into anticomplete sets with . This is already a pure pair in .
Otherwise has a connected component with . Every vertex of has degree at most , and is -free because induced subgraphs preserve forbidden induced paths. Applying the connected-case claim from step 5.1 to the connected graph , outcome 1 is false by the degree bound and outcome 2 is false because contains no induced at all. Hence outcome 3 holds, so has a biclique with both sides of size at least . Equivalently, has an anticomplete pair of that size.
Let . Steps 9.1 and 9.2 show that every graph with no induced or and at least two vertices contains a pure pair with both sides of size at least . By [L4], the class forbidding and has the strong Erdős–Hajnal property.
Depends on
- The strong Erdős–Hajnal property for a hereditary graph class
- $H$-free and $\mathcal F$-free graphs under the induced-subgraph convention
- Empty and complete graphs, complete bipartite graphs, and the convention that $P_n$ and $C_n$ have $n$ vertices
- The edge-density form of Rödl's theorem: every nonempty $H$-free graph has a linearly large set of self-density at most $\epsilon$ or at least $1-\epsilon$
- A set of self-density at most $c$ has a subset of at least half its size that is $4c$-sparse
- The connected components of a graph partition its vertex set and are its maximal connected subgraphs
- Distinct connected components are anticomplete, and distinct anticonnected components are complete
- Connected graphs and connected components defined by the existence of vertex paths
- Adjacency, incidence, open and closed neighbourhoods, vertex degree, minimum degree and maximum degree
- Graph isomorphisms, automorphisms and graph complements
- Edges between disjoint vertex sets; complete, anticomplete, pure and mixed pairs
- $c$-sparse, $c$-dense and $c$-restricted vertex sets
Used by
Dependency tree · two levels
37 results within two dependency steps of this one, each drawn at its shortest distance from it. An arrow runs from a result to what uses it, so the chart reads left to right and ends at this result, which carries a heavier outline. Every node is a link to that result. Click elsewhere on the chart to enlarge it.
Sources
- Nicolas Bousquet, Aurélie Lagoutte, and Stéphan Thomassé, The Erdős-Hajnal Conjecture for Paths and Antipaths, Lemma 3 and Theorem 4 (standard reference, not scraped)