The unappreciated, fundamentally analytical nature of taxonomy and the implications for the inventory of biodiversity.

Several ways have been suggested to improve the poor status of taxonomy as well as to overcome the taxonomic impediment and thus to speed up species discovery and documentation, such as: DNA barcoding, creation of databases of taxa and identification tools, online quantum contributions, standardization of morphological features, training of a new generation of taxonomists. The paper comments on the desirability and effectiveness of these presumed remedies. It is argued that the analytical, hypothesis-testing nature of taxonomic research is not well understood or appreciated and forms a major constraint on the rate of taxonomic descriptions. The various components of the taxonomic method, such as exploration, data, analysis, and results, interact in a complex manner that resembles the erratic, bouncing behaviour of a pinball machine. Species hypotheses probably are the most tested hypotheses in the natural sciences. The introduction of cybertaxonomy instrumentation and infrastructure will not alleviate the time-consuming, intrinsically analytical and hypothesis-testing process underlying the description and/or identification of each taxon. Long-term survival of the discipline of taxonomy, thus guaranteeing the future cumulative taxonomic output of amateurs and professionals, is endangered by a diminishing workforce of adequately trained professional systematists. The only way to increase the pace of a well-documented and scientifically useful taxic inventory of the world's biodiversity is to increase the number of professional taxonomists. [ABSTRACT FROM AUTHOR]