IFLA's Functional Requirements for Bibliographic Records (FRBR) presents a compelling and influential model of the “bibliographic universe.” However there are interesting variations between FRBR's formal model and the narrative expositions of FRBR's authors and explicators – that is, between the formal model and the framework as more broadly understood by the FRBR community. In this paper we argue that despite a widespread belief to the contrary, attribute inheritance down the “hierarchy” of Group 1 entities is inconsistent both with the formal model and with the general spirit of the project. We believe these observations reveal an ongoing uncertainty about the nature of bibliographic entities as well as difficulties in maintaining a clear and exact understanding of the models we are using to represent those entities – even when those models are our own creation.