Gauthier (1986:35) proposed the initial definition for Maniraptora, a stem-based definition that used Ornithomimidae as an external specifier. This was not widely recognized because the definition was located in the conclusions of work rather than the discussion under the taxon. Padian (1997:411), for example, incorrectly cited Holtz (1996) for the initial stem-based definition of Maniraptora with ornithomimids as the external specifier. Three errors of interpretation involving Maniraptora have ensued:

First, Holtz (1994:figs. 4, 5) applied Maniraptora to a more inclusive clade that included ornithomimosaurs—a clade equivalent in taxonomic content to Gauthier’s Coelurosauria. He did so fully cognizant of Gauthier’s stem-based definition but unhappy that it would be limited on his cladogram to a poorly supported clade consisting only of dromaeosaurids and birds. With a curious twist of logic, Holtz argued that there was another definition implied by the etymology of Gauthier’s taxon (“hand snatchers”) and his list of synapomorphies. He then attempted to construct an “apomorphy-based” definition: “the first theropod possessing the derived fore limb [sic] structures described by Gauthier (including a semilunate carpal structure, a thin narrow metacarpal III, and bowed ulna) and its descendants” (Holtz 1994:1107). In this case, it is not clear if the remaining maniraptoran synapomorphies listed by Gauthier are part of the definition or if the definition should use Holtz’s, rather than Gauthier’s, version of the three cited apomorphies (because they differ in substantive details). More fundamentally, Holtz’s definition is incomplete. A complete stem-based definition with a form qualifier (i.e. an “apomorphy-based” definition) must include at least one specifier (Cantino and de Queiroz 2003, Sereno 2005). Given this background, there is little basis to argue that Holtz (1994) properly redefined Maniraptora as an “apomorphy-based” clade on the basis of wrist morphology (Holtz 1996:536, Padian 1997:411).

Second, Holtz and Padian (1995) redefined Maniraptora again in an abstract as a less inclusive node-based clade including only Dromaeosaurus, birds and descendants of their common ancestor.

Third, Sereno (1998) provided yet another definition for Maniraptora. Like Holtz and Padian (1995) it was node-based and less inclusive than Gauthier’s original definition. This definition, however, was more inclusive than that given by Holtz and Padian (1995), as it included Oviraptor, birds and descendants of their common ancestor.

The most sensible solution given the above is to abandon both Holtz and Padian’s (1995) and Sereno’s (1998) less inclusive, node-based definitions of Maniraptora and return to Gauthier’s original stem-based definition. Doing so would follow the path of many authors that have presented nuanced variations of Gauthier’s original definition, including Holtz (1996), Padian (1997), Chure (2001), and Maryanska et al. (2002). Also following Holtz (1996), Padian (1997) and Sereno (1998). Thus the active definition supported here is a first-order revision of that given by Gauthier (1986).

Maniraptoriformes is adopted in this compilation as a more inclusive group with a node-based definition that includes ornithomimids but excludes all more distantly related coelurosaurians, following Holtz (1996), Padian (1997) and Sereno (1998).

Taxa with less obvious use include:

(1) Tyrannoraptora (Sereno 1999b) is tentatively retained; it will be more useful when, and if, the position of Tyrannosauroidea is better established.

(2) Eumaniraptora, coined by Padian et al. (1997) and defined by Padian et al. (1999) differs from Sereno’s (1997, 1998) stem-based Paraves only by the type of definition; in most analyses, they include the same taxonomic content. Naming both a node- and stem-based taxon that have the same current taxonomic content is not a useful practice, because their difference is less obvious to practicing taxonomists, who them may use them in unintended ways.