On the weak Lefschetz property for Hilbert functions of almost complete intersections (Q1944721)

The last 10 or so years have seen a lot of investigation into the Hilbert functions of Artinian standard graded algebras: concretely, if \(R=k[x_1,x_2,\dots,x_n]\) is the polynomial ring over a field \(k\) (usually of characteristic zero) and \(I\) is a homogeneous ideal with quotient ring \(A=R/I\) of finite length, how are properties of \(A\) reflected in the possibilities for the Hilbert function \(H(t)=\dim_k A_t\), which can be viewed as a finite sequence of positive integers? One such property on \(A\) is the \textit{weak Lefschetz property} or WLP, meaning that there's a linear form \(L\) for which the multiplication maps \(L:A_t \to A_{t+1}\) have maximal rank for all \(t\). For WLP algebras it is known that the Hilbert function is a \textit{weak Lefschetz sequence}, meaning that (a) \(H\) is unimodal, i.e. \(H(0) \leq H(1) \leq \dots \leq H(m) \geq H(m-1) \geq \dots \) for some integer \(m\) and (b) the positive part of the first difference \(\Delta H\) is an \(O\)-sequence (the positive part satisfies Macaulay's growth bound). This paper deals with the special case \(n=3\) (WLP is automatic when \(n=2\)). Here \textit{T. Harima, J. Migliore, U. Nagel} and \textit{J. Watanabe} used the splitting type of the sheafified (rank 2) bundle of relations on \(\mathbb P^2\) to prove that if \(A\) is a complete intersection, then \(A\) has the WLP [J. Algebra 262, No. 1, 99--126 (2003; Zbl 1018.13001)]. Pushing this idea further, \textit{H. Brenner} and \textit{A. Kaid} gave examples of almost complete intersection algebras (meaning \(I\) is minimally 4-generated) which do not satisfy the WLP [Ill. J. Math. 51, No. 4, 1299--1308 (2007; Zbl 1148.13007)], answering a question of \textit{J. Migliore} and \textit{R. M. Miró-Roig} [J. Pure Appl. Algebra 182, No. 1, 79--107 (2003; Zbl 1041.13011)]. The main result in the paper under review states that the Hilbert function of an almost complete intersection nonetheless must be a weak Lefschetz sequence, generalizing the result of \textit{U. Nagel} and \textit{D. Cook II} [Ill. J. Math. 55, No. 1, 377--395 (2011; Zbl 1262.13032)] for monomial level algebras. The idea of the proof is to link \(A\) to a Gorenstein algebra \(G\) by a complete intersection and exploit the relations between the corresponding Hilbert functions. The proof is technical, but easy to follow.