Commutator relations and the clones of finite groups. (Q2510714)

The clone of a group \(G\) is the set of all finitary term operations on \(G\). Note that term operations clearly preserve all subgroups of powers of \(G\). Moreover by \textit{E. Aichinger, P. Mayr} and \textit{R. N. McKenzie} [J. Eur. Math. Soc. (JEMS) 16, No. 8, 1673-1686 (2014; Zbl 1432.08001)], for every finite \(G\) there exist \(k\) and a subgroup \(H\) of \(G^k\) (i.e., a \(k\)-ary relation of \(G\)) such that the term operations are exactly the \(H\)-preserving operations. However, in general, given a group \(G\) it is not known how to find this \(H\) or even its arity \(k\). In [Algebra Univers. 54, No. 1, 23-52 (2005; Zbl 1114.20014)] \textit{K. A. Kearnes} and \textit{Á. Szendrei} asked: Does there exist a uniform \(k\) such that the clone of every finite group is determined by \(k\)-ary relations? The author answers this in the negative by proving that for \(n\geq 3\) the \((n-1)\)-ary relations of the dihedral group \(D_{2^n}\) are not enough to determine its term operations. Using commutator theory he shows that the clone of \(D_{2^n}\) is determined by relations of arity \(2^{n-1}\).