Closure of good trees under disjoint union

Status partial high confidence

The general question of whether the disjoint union of two good trees is always good remains open as of 2026. Partial progress exists: arXiv:2202.10412 (submitted three days after the source paper) proves that if $H$ is a good forest then $H \cup P_4$ is also good, and the source paper itself notes that the disjoint union of two good forests is good whenever one is self-isolating. No paper resolving the full conjecture was found in the indexed literature.

Cited literature (1)

• partial Polynomial bounds for chromatic number VI. Adding a four-vertex path (2022)
  Maria Chudnovsky, Alex Scott, Paul Seymour, Sophie Spirkl · arXiv preprint · arXiv:2202.10412

  Proves that if $H$ is a good forest then the disjoint union of $H$ and $P_4$ is also good, establishing in particular that $2P_4$-free graphs are polynomially $\chi$-bounded, but falls short of proving the conjecture for arbitrary pairs of good trees.

Reviewer notes. arXiv:2311.05719 appears three times in the internal references with a contribution text that describes a completely different conjecture (clock-free graphs, treewidth) — likely a fuzzy-matching false positive. arXiv:2202.10412 is a companion paper in the same 'Polynomial bounds' series submitted three days after the source paper by the same four authors; its H ∪ P₄ result is the strongest known partial answer to the open problem.

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