Chromatic graph homology for brace algebras (Q2407270)

Chromatic graph homology is a chain complex valued invariant. It takes two pieces of data as input: a finite graph \(G\) and a differential graded commutative associative algebra \(A\) (satisfying some further weak homological conditions). This invariant knows graph-theoretic information about \(G\) (its graded Euler characteristic is essentially the chromatic polynomial). There is also interesting geometric content: if \(A\) is the cohomology ring of a space \(X\) satisfying some technical conditions then the chromatic graph homology is the first page of a spectral sequence converging to the relative homology of the so-called \textit{graph configuration} space of \(X\). The paper under review gives a variation on chromatic graph homology where the algebra is more general: it need not be a commutative associative algebra, but a weaker kind of algebra, a \textit{brace algebra}. However, a price must be paid: the only graphs for which the invariant is defined are trees. In fact the trees must be equipped with additional combinatorial data: a cyclic order on the edges at each vertex (that is, the isotopy class of a planar embedding of a neighborhood of that vertex). To a reader from knot theory or commutative algebra, this price may seem rather steep: brace algebras are less familiar than commutative associative algebras, and local planarity is potentially a rather technical condition. But from the point of view of contemporary algebraic homotopy theory, this feels more or less natural. Brace algebras, as a model for \(E_2\)-algebras in characteristic zero, constitute the first step on a well-studied hierarchy of algebras interpolating between associative and commutative associative algebras. This step in the hierarchy is intimately related to the self-embeddings of the plane, and so (local) planarity as a condition for a construction involving brace algebras comes as no surprise. The necessity of the condition that the graph be a tree is less clear and the authors express hope that there should be a way to remove it.