Spurious states in boson calculations - spectre of reality?

We discuss some prevailing misconceptions about the possibility that spurious states may in general contaminate boson calculations of fermion systems on either the phenomenological or microscopic level. Amongst other things we point out that the possible appearance of spurious states is not inherently a mapping problem, but rather linked to a choice of basis in the boson Fock space. This choice is mostly dictated by convenience or the aim to make direct contact with phenomenology. Furthermore, neither well established collectivity, nor the construction of boson operators in the Marumori or OAI fashion can as such serve as a guarantee against the appearance of spurious boson states. Within an SO(12) generalisation of the Ginocchio model where collective decoupling is complete, we illustrate how spurious states may appear in an IBM-type-sdg-boson analysis. We also show how these states may be identified on the boson level. This enables us to present an example of an sdg-spectrum which, although it may be reasonably correlated with experimental data, nevertheless has the first few low lying states all spurious when interpreted from the microscopic point of view. We briefly speculate about the possibility that certain types of truncation may in fact automatically circumvent the appearance of spurious states. (orig.)