χ-boundedness of P₄-orientation forbiddance classes

Conjecture 5 · arXiv:1605.07411

arXiv Conjecture high confidence— first stated 2016-05-24

Status solved medium confidence

Conjecture 5 asserts that $\mathrm{Forb}(\mathcal{P})$ is $\chi$-bounded for every non-empty $\mathcal{P} \subseteq \mathrm{Or}(P_4)$ with $\mathcal{P} \neq \{P^+(3)\}$ and $\mathcal{P} \neq \{P^+(1,1,1)\}$. The source paper itself proved the case $P^+(3) \in \mathcal{P}$ (Corollary 35). Cook, Masařík, Pilipczuk, Reinald, and Souza (EJC 2023) subsequently proved the directed analogue of the Gyárfás–Sumner conjecture for all orientations of $P_4$, establishing $\chi$-boundedness of $\mathrm{Forb}(F)$ for each non-exceptional orientation $F$ of $P_4$; together with Corollary 35 this resolves all remaining cases of Conjecture 5.

Cited literature (1)

  • Linda Cook, Tomáš Masařík, Marcin Pilipczuk, Amadeus Reinald, Uéverton S. Souza · Electronic Journal of Combinatorics

    Proves the directed analogue of the Gyárfás–Sumner conjecture for all orientations of $P_4$, establishing $\chi$-boundedness of $\mathrm{Forb}(F)$ for each orientation $F$ of $P_4$ other than the two known exceptions $P^+(3)$ and $P^+(1,1,1)$, thereby resolving the open cases of Conjecture 5.

Reviewer notes. Cook et al. (EJC 2023, v30i3p36) proves the directed Gyárfás–Sumner analogue for all orientations of P4; combined with Corollary 35 of the source paper, this fully resolves Conjecture 5. Confidence is medium rather than high because the WebFetch of the Cook et al. article page did not return enough text to confirm an explicit citation of Conjecture 5 by number, though the mathematical content squarely aligns.

Auto-reviewed 2026-05-15 with claude-sonnet-4-6 (web search enabled).

Conjecture. Let $\mathcal{P}$ be a non-empty subset of $\mathrm{Or}(P_4)$. If $\mathcal{P} \neq \{P^+(3)\}$ and $\mathcal{P} \neq \{P^+(1,1,1)\}$, then $\mathrm{Forb}(\mathcal{P})$ is $\chi$-bounded.

Context

Among the four non-isomorphic orientations of $P_4$, the paper shows $\mathrm{Forb}(P^+(3))$ and $\mathrm{Forb}(P^+(1,1,1))$ are not $\chi$-bounded. The conjecture asserts these are the only obstructions; the paper proves it when $P^+(3) \in \mathcal{P}$ (Corollary 35).

Notes. PDF source — math may be garbled. Preceded by the informal statement 'We believe that $\{P^+(3)\}$ and $\{P^+(1,1,1)\}$ are the only non-empty subsets $\mathcal{P}$ of $\mathrm{Or}(P_4)$ such that $\mathrm{Forb}(\mathcal{P})$ is not $\chi$-bounded.'

Source paper

$χ$-bounded families of oriented graphs
Pierre Aboulker, Jørgen Bang-Jensen, Nicolas Bousquet, Pierre Charbit, Frédéric Havet, Frédéric Maffray, Jose Zamora · 2016-05-24
https://arxiv.org/abs/1605.07411 PDF source