Recent arguments against phylogenetic nomenclature (PN) rest on the assumption that, generally, consensus about the meaning of taxon names arises spontaneously. A brief historical review shows that this is not the case. Comparisons with other fields as diverse as physics, geology and geopolitics show that precision in the meaning of terms is essential to produce consensus, which is precisely the opposite as the avowed aims of rank-based nomenclature (RN). The difficulty in reaching consensus increases with the weight of tradition and decreases with the number of objective tests to falsify competing theories. In both respects, biological nomenclature is handicapped because the weight of tradition is extreme and rules of nomenclature cannot be discovered in nature. These facts may explain the difficulty in reaching consensus on the most appropriate system of nomenclature for the 21st century. Therefore, comparisons between RN and PN should focus on minimal taxonomic stability, rather than realized or maximal stability. A four-taxon example shows that in this respect, PN vastly outperforms RN. Opponents of PN often predict that implementation of PN will cause considerable confusion. A comparison with computer science shows that confusion is often associated with progress, and may be unavoidable for nomenclature to prosper in the new millennium.