Sir

H. C. J. Godfray in his Commentary1 suggests that we throw out the past mechanisms of doing systematics and begin anew in a revolutionary 'brave new world' of unitary, web-based taxonomy, each group under the administration of an authoritarian body. We agree with Godfray that problems of synonymy can be frustrating and that the web is important in the dissemination of taxonomic information, but do we need a new mechanism to do taxonomy? Working within the current enabling conventions is more positive and practical than throwing them out and beginning again.

The international codes of zoological, botanical and bacteriological nomenclature are the rulebooks that have long governed how organisms are named, providing clear instructions on how to go about the process — a facility either lacking or in the process of being invented in other disciplines concerned with naming things.

At the core of the zoological and botanical codes is the type specimen. Types bring order and stability to taxonomy because they are the specimens upon which the original author based his or her descriptions, and which ultimately fix the name. Even DNA taxonomy needs archived voucher specimens to ensure that future generations can check and replicate findings. The codes operate for taxonomists as conventions, which, like UN conventions, are international in scope and have evolved during decades of intellectual debate.

Taxonomy provides a vocabulary to discuss the world: specifically, the names of organisms. We taxonomists therefore have a special role in communicating to the widest possible set of audiences. The Internet, being a distributed system of linked information, is ideal for this purpose — the various approaches under way2,3 will revolutionize taxonomical informatics in a positive, community-driven way.

Taxonomists have not yet made as much use of the web as they could, but first we need to identify with confidence core components of essential information, such as species lists, identification keys and illustrations. From our privileged position in the information-rich developed world, we run the risk of providing merely what we ourselves need and forgetting those in whose countries most of the diversity on Earth resides. Information is critical for global conservation, and much of it, particularly original descriptions of species and the specimens on which these are based, remains inaccessible to those who need it most.

Building a 'biodiversity commons' (ref. 4), where access to original taxonomic descriptions is part of the distributed web system, is critical. Much taxonomic information is already available online: FishBase (http://www.fishbase.org) and the Cycad pages (http://plantnet.rbgsyd.gov.au/PlantNet/cycad/index.html), to name just a couple, are excellent examples. At present, we need meta-databases to link dispersed information, which is where the Global Biodiversity Information Facility (GBIF) is concentrating its efforts.

Web-provided taxonomy is clearly the way for the future, but the technologies needed for this to operate successfully on the scale required are only starting to be available, and quality control is something that must also be addressed5. The Natural History Museum is putting on its website (http://www.nhm.ac.uk), as a priority, information about its type specimens, which will provide the kind of focus for scientifically rigorous and accessible taxonomy that Godfray advocates.

Informatics, however, is not a substitute for science. To ensure worldwide consistency and accuracy we also need open access to the foundations of our taxonomy, by making high-quality illustrations and descriptions of the type specimens available on the web. For an institution such as ours, housing nearly a million type specimens, this is a huge task. It will be achieved, but not soon, unless the world is ready to support more taxonomists doing taxonomy. This in turn requires widespread acceptance of taxonomy as a mature and stimulating science.