Comment on the generation number in orbifold compactifications (Q1801507)

The authors have investigated in a systematic way orbifolds with (2,2) world sheet supersymmetry, which can be constructed by modding out symmetric \(\mathbb{Z}_ N\) actions from the six dimensional torus with vanishing discrete \(B\)-field. As a result they give a classification of these types of models. Preceding investigations in this direction have the drawback of being incomplete and more seriously of stating or using incorrect spectra. The source for the deviations is that properties of the twist, which are not directly related to the twist eigenvalues, were not taken into account properly. In the case of the \(\mathbb{Z}_ 4\), \(\mathbb{Z}_ 6\), \(\mathbb{Z}_ 6'\), \(\mathbb{Z}_ 8\), \(\mathbb{Z}_ 8'\), and \(\mathbb{Z}_{12}\) actions there exist, for the same twist eigenvalues, inequivalent automorphisms which are realized in different lattices. The complete reducibility of the twist over \(\mathbb{C}\), which is used to define the complex planes in the space-time bases, is often confused with the reducibility of the twist over \(\mathbb{Z}\), such that several authors erroneously imply that the twist can be made block diagonal in the lattice basis. Consequently their conclusions are only correct if this holds indeed, which is usually only for one of the above mentioned inequivalent automorphisms the case. Statements concerning the factorisation properties of the modular group, which have been made in the same spirit, are also wrong in the general cases.