# Stricter Notion of Crossing in a Partition

Let $k$ be an integer. Traditionally a partition $\pi=V_1\cup \dots \cup V_n$ of the set $[k]:=\{1,\dots, k\}$ is called crossing when there exist $a,c\in V_i$ and $b,d\in V_j\not= V_i$ such that $a<b<c<d$. For example, the partition $\{1,3\}\cup\{2,4,5\}$ (which, for convenience, I will just denote by $(12122)$) is a crossing partition of $[5]$ whereas $(21122)$ is a non-crossing partition.

For my needs I am interested in a different notion of crossing, namely only those crossings that can't be resolved in an even number of permutations. For example, in my stricter notion of crossing $(1223123)$ would not have a crossing between $1$ and $2$ since for resolving this crossing, we would have to move $1$ past an even number of $2$'s. It would, however, still have crossings between $1$,$3$ and $2$,$3$.

Is this notion known? Do there exist results on the combinatorics of this stricter notion of crossing?

• How does $(12122)$ denote $\{1,3\}\cup\{2,4,5\}$? Sep 25, 2014 at 23:36
• @GerryMyerson : Given a word $a_1a_2a_3a_4a_5$, put $i$ and $j$ in the same subset if and only if $a_i = a_j$. Sep 26, 2014 at 2:16
• Given $n$, compute the number of strongly non-crossing partitions (which can't be resolved), and plug into OEIS. You might find references there, and if not, it is most likely new. Sep 26, 2014 at 7:30
• Instead of "even number of permutations" I think you mean "even number of transpositions within each pair $i,j$", right? E.g. in your example, the total number of transpositions would be even+odd+odd=even, which is not what you mean. Sep 26, 2014 at 7:43
• I just noticed: in your example, if you move the last 1 to the right, you obtain 1223213, so the 1-2-crossing is removed. So I wonder how to come up with a well-defined parity of the "resolving number". Sep 26, 2014 at 7:49