Alon–Tarsi Latin Square Conjecture
Canonical statement
View source LaTeX
Let \(n\) be even. Regard every row and every column of a
Latin square \(L\) of order \(n\) as a permutation of its symbol set, and
define
\[
\operatorname{sgn}(L)=
\prod_{i=1}^{n}\operatorname{sgn}(\text{row}_i)
\prod_{j=1}^{n}\operatorname{sgn}(\text{column}_j).
\]
The number of Latin squares of order \(n\) with sign \(+1\) is different
from the number with sign \(-1\).Notes
Assign to a Latin square of even order the sign , the product of the signs of its row and column permutations. The Alon–Tarsi conjecture asserts that the number of Latin squares of order with sign differs from the number with sign . It was posed by Alon and Tarsi around 1990 in connection with their polynomial method for colorings and orientations of graphs [AlonTarsi1992].
The restriction to even is essential, since for odd orders the two counts coincide and the difference vanishes identically. Much of the conjecture's interest lies in its consequences: by the Alon–Tarsi machinery, nonvanishing of the signed count yields strong list-coloring conclusions [AlonTarsi1992]. On the positive side, Drisko established the conjecture for orders of the form with an odd prime [Drisko1998], and it is now known for several infinite families of even orders.
No argument covers all even , however, and the conjecture, together with the list-coloring consequences that would flow from it, remains open.
References (2)
- [AlonTarsi1992]
Colorings and orientations of graphs
Open ↗Noga Alon and Michael Tarsi · 1992 · misc
- [Drisko1998]
On the number of even and odd Latin squares of order p+1
Open ↗Arthur A. Drisko · 1997 · misc
The boxed statement is the canonical open formulation — not a stronger variant or a related research program. The status reflects the catalog's last review; do your own literature search before investing serious effort.