The imminent introduction of the draft PhyloCode (Cantino & de Queiroz, 2000) governing application of phylogenetic nomenclature (de Queiroz & Gauthier, 1990, 1992) has generated considerable discussion in this journal and elsewhere. Some contentious issues include the information content of Linnaean ranks, stability in meaning and content of taxon names, relative merits of node-, stem- and apomorphy-based definitions, restric- tion of widely-used names to crown-clades, and format of species names (e.g., see Forey, 2001; Lee, 2001; de Queiroz & Cantino, 2001; Brochu & Sumrall, 2001; Ereshefsky, 2001; Bryant & Cantino, 2002). However, one subject has not been sufficiently discussed: whether the Phylocode should govern species names. This is a most important issue, since it will determine whether the PhyloCode will eventually affect the definitions and for- mat of species binomials, which form the vast majority of existing taxon names. Although the draft Phylocode currently contains rules for naming only clades, there are plans to eventual- ly extend it to also cover species (Cantino & de Queiroz, 2000), and various possible options for naming species in phylogenetic nomenclature have been presented (Cantino & al., 1999). These options all involve convert- ing binomial species names defined under the current Linnaean codes into a different (either totally or func- tionally uninomial) format under Phylocode. There are compelling reasons for such a change (see de Queiroz & Gauthier, 1990, 1992; Cantino & al., 1999; Ereshefsky, 2001; Pleijel & Rouse, in press; and references therein). The mandatory incorporation of the genus into a binomi- al species name creates instability in not just the genus name (e.g., new generic assignments due to splitting and lumping, or newly discovered phylogenetic relation - ships), but also destabilises the species epithet (e.g. sec- ondary homonymy, or changes in gender of species epi- thet to match that of a new generic assignment). New species cannot be described until their generic assign- ments (i.e., phylogenetic relationships) are known pre- cisely. Mandatory generic assignments for species pro- duces redundant higher taxa (e.g., monotypic genera are identical conceptually to their sole species). Stem species of two sister genera cannot be assigned to either descen- dant genus, or placed in their own genus, without creat- ing a paraphyletic genus. Finally, the genus rank does not exhibit any unique features distinguishing it from other Linnaean higher taxon ranks and is thus biologically meaningless. All these problems can be avoided by drop- ping the genus name from the names of species taxa. Species names would then be unaffected by generic reas- signments, new species could be described without knowing their phylogenetic relationships, and monotyp- ic (redundant) genera would never be created. In an ideal world, switching from binomial to uninomial species names would be easy and eliminate all the above prob- lems. However, the real world contains substantial phy- logenetic inertia. Converting the million-plus existing binomials into a different and (at least functionally) uni- nomial format represents a mammoth task. However, an even bigger obstacle is that binomial names are deeply entrenched throughout society, through their use in, for instance, popular books and field guides, faunal surveys, legislation (endangered species, animal ethics), and museum labels and catalogues. To expect all these insti- tutions to smoothly adopt a new format for species names is overly optimistic, especially given the substan- tial effort required on their part, and the lack of appreci- ation by most non-scientists of the biological reasons for such a change. Thus, the question of whether the PhyloCode should be extended to species names requires careful consideration, and it is hoped this commentary will help generate relevant debate. I first discuss how PhyloCode can easily function without any reference to species at all, and then mention some problems that need to be addressed if species are to be included. One possible reason why species names should be included in the PhyloCode is that species are commonly employed as reference (= specifier, “type”) taxa employed in phylogenetic definitions (e.g., Cantino & al., 1997; Baum & al., 1998; Härlin, 1999; Rouse, 2001). For instance, the definition “the least inclusive clade con- taining Species X and Species Y” is affected by the sta- bility of the names of the two reference species. If those species names are defined under the Linnaean system, 1 Lee • Species and phylogenetic nomenclature 51 • August 2002: XX–XY POINTS OF VIEW Species and phylogenetic nomenclature Michael S. Y. Lee Department of Palaeontology, The South Australian Museum and Department of Environmental Biology, University of Adelaide, North Terrace, Adelaide SA5000, Australia. E-mail: lee.mike@saugov.sa.gov.au.