Horrocks' question for monomially graded modules (Q583289)

The paper under review contains a new approach to Horrock's question, which asks whether \(\dim_ kTor^ R_ i(k,M)\geq \binom{n}{i}\) for every regular local ring \((R,m,k)\) of dimension n and every finite length module \(M,\) \(i=1,...,n\). The main result gives a general inequality relating values of an additive function defined on a class of \(R[X_ 1,...,X_ d]\)-modules satisfying certain conditions. The author's techniques allow R to be a commutative ring with the price of imposing additional constraints on the modules themselves. This approach is fruitful indeed: there is an application where the additive function is defined via Serre's intersection multiplicity; one gets a general result on iterated Koszul homology; one arrives at a significant addition to the results known about Horrock's question. Moreover, the essential elements of the techniques used here seem to be of more importance than their utility in this paper.