[guided]The purpose of this step is to isolate exactly where the unknown cumulant $\kappa_n(x_1,\dots,x_n)$ appears. It appears only in the one-block partition $1_n$, because for that partition the unique block is $\{1,\dots,n\}$ and therefore $\kappa_{1_n}(x_1,\dots,x_n)=\kappa_n(x_1,\dots,x_n)$.
We now define the notation used to discuss the other terms. The set $NC(n)$ is the set of noncrossing partitions of the ordered set $\{1,\dots,n\}$. If $\pi\in NC(n)$ and $V=\{i_1<\cdots<i_r\}$ is a block of $\pi$, then the block contributes the scalar $\kappa_V(x_1,\dots,x_n):=\kappa_r(x_{i_1},\dots,x_{i_r})$. The whole partition contributes the product over all its blocks: $\kappa_\pi(x_1,\dots,x_n):=\prod_{V\in\pi}\kappa_V(x_1,\dots,x_n)$. This product is taken in $\mathbb{C}$, since cumulants are complex-valued multilinear functionals, so the order of the blocks does not affect the value.
The [moment-cumulant formula for free cumulants](/theorems/7107) says that $\varphi(x_1\cdots x_n)=\sum_{\pi\in NC(n)}\kappa_\pi(x_1,\dots,x_n)$.
Because $x_j=1_A$, the decisive feature of a partition $\pi$ is the block containing $j$. There are three possibilities. First, $\pi=1_n$, which gives the target term $\kappa_n(x_1,\dots,x_n)$. Second, $\{j\}$ is a singleton block; those terms will reproduce the shorter moment expansion after deleting $x_j$. Third, $j$ lies in a block of size at least $2$ but $\pi\neq 1_n$; those terms contain a lower-order cumulant with a unit entry and will vanish by the induction hypothesis.[/guided]