How Are Dinosaurs Named? Rules, Priority and a Great Deal of Latin

A scientific name has two parts, a genus and a species, and both are written in italics with only the genus capitalised. The genus groups closely related species and the species identifies one within it, so the full name identifies a single kind of animal unambiguously worldwide regardless of what language anyone speaks. Most dinosaurs are known popularly by their genus alone, which is why the familiar names are single words while the formal names have two. The words are constructed from Latin and Greek roots or latinised from other languages, and they must be treated as Latin grammatically even when the roots are not, which produces occasional awkwardness. The name is followed in formal usage by the person who described it and the year, which credits the work and makes the original description findable, and that citation is part of the name rather than an optional addition.

The choice of name is the describer's and the conventions are loose, which produces variety:

• •Anatomy, which is the most common, naming a distinctive feature such as a crest, a tooth shape or a limb proportion
• •Place, honouring where the specimen was found, from a country down to a specific formation or quarry
• •People, honouring a collector, a landowner, a funder or a colleague, which is permitted for the species name and discouraged for self-naming
• •Local languages, increasingly used for animals from regions with their own naming traditions, which has produced many recent names from indigenous words
• •Mythology, drawing on figures whose attributes suit the animal
• •Jokes and references, which are permitted, occasionally regretted and impossible to revoke once published

The governing principle is that the first validly published name wins, and that single rule explains most of the confusing cases. If an animal is named twice, the earlier name stands and the later one becomes a junior synonym, which is why some familiar names have been replaced by unfamiliar older ones. The most famous instance concerns a large sauropod named twice by the same researcher within two years, with the later and better-known name losing out to the earlier one for most of a century before the two were argued to be distinct after all. Changes of this kind are not fashion but the application of a rule, and they happen whenever earlier material is recognised as belonging to the same animal. A name can also be preserved against an older one by formal petition where the disruption would be severe, which exists precisely because strict priority occasionally produces absurd outcomes.

Above the genus the naming works differently and is governed less strictly, which causes its own confusion. Family, order and the other traditional ranks were designed for a system that assumed fixed categories, and modern classification instead names branches of the evolutionary tree, defining a group by its ancestry rather than by a list of features. A branch is generally defined as everything descended from the common ancestor of two named species, or as everything sharing a specified feature, and those definitions are stable in a way that rank-based ones are not. The practical consequence for dinosaurs is that birds fall inside the group rather than beside it, since the group is defined by descent and birds descend from it, which is why specialists refer to non-bird dinosaurs when they mean the familiar animals. The ranks persist in textbooks and are treated as conveniences rather than as claims.

Publishing a new name has requirements and the standards have risen considerably. A type specimen must be designated, which is the individual the name is permanently attached to, and it must be deposited in a public collection where other researchers can examine it, since a name based on material in private hands is regarded as unusable even where it is technically valid. The description must state what distinguishes the animal from everything already named, which is the substantive scientific claim and the part that gets challenged. Publication must be in a permanent and accessible form, which has become complicated with electronic journals and is now addressed by specific provisions. Naming from fragmentary material is legal and widely criticised, since a name based on a single bone frequently cannot be distinguished from later better material, and a substantial proportion of nineteenth-century names are unusable for exactly that reason.

Two parts, a genus and a species, both treated as Latin, followed by the describer and the year, which is part of the name rather than an addition. The first valid publication wins, which is why familiar names get replaced by older unfamiliar ones rather than by fashion. A type specimen must sit in a public collection, and names based on fragments frequently cannot be distinguished from later material.