Absolute (Linnaean) ranks are essential to rank-based nomenclature (RN), which has been used by the vast majority of systematists for the last 150 years. They are widely recognized as being subjective among taxonomists, but not necessarily in other fields. For this reason, phylogenetic nomenclature (PN) and other alternative nomenclatural systems have been developed. However, reluctance to accept alternative nomenclatural systems and continued use of higher taxa of a given Linnaean category in comparative analyses presumably reflect a lack of appreciation of the deleterious effects of the subjective nature of Linnaean categories in other biological fields, such as conservation and evolutionary biology. To make that point clearer, evolutionary models under which such categories would be natural are presented and are shown to be highly unrealistic and to lack empirical support. Under all realistic evolutionary models, ranking of taxa into Linnaean categories is highly subjective. Solutions that could make taxonomic ranks objective are surveyed. A review of the literature illustrates two problems created by the use of Linnaean categories in comparative or evolutionary studies, namely suboptimal taxonomic sampling schemes in studies of character evolution, and unreliable biodiversity assessment drawn on the basis of counting higher taxa (taxon surrogacy).