---

General Articles and Nomenclatural Notes with the following titles were published on 19 December 2001 in Volume 58, Part 4 of the Bulletin of Zoological Nomenclature

Copies of these General articles and Nomenclatural Notes can be obtained free of charge from the Executive Secretary, The International Commission on Zoological Nomenclature, c/o The Natural History Museum, Cromwell Road, London SW7 5BD, U.K. (e-mail: iczn@nhm.ac.uk).

Nomenclatural Notes

The authorship and date of the specific name of Ursus or Thalarctos maritimus, the polar bear, is Phipps (1774) and not Linnaeus (1758)

Anthea Gentry
ICZN Secretariat, c/o The Natural History Museum, Cromwell Road, London S W7 SBD, U. K.

  In 1934 C.W. Stiles (the first Secretary to the Commission) sought to place several generic names for carnivores on the Official List. The name Thalarctos Gray, 1825 (p. 62, published as a subgenus of Ursus Linnaeus, 1758, type species by monotypy T. polaris Gray, 1825) was one of those considered. A number of mammalogists, including Dr Angel Cabrera (a Commissioner from Argentina), were invited to comment. Dr Cabrera stated that the specific name of Ursus maritimus, already known as a senior synonym of T. polaris and usually cited from Phipps (1774) or Erxleben (1777), dated from Linnaeus (1758). This authorship and date (Linnaeus, 1758) for maritimus was incorporated into the eventual ruling on the case (Opinion 384, April 1956; see Opinions and Declarations 12: 71-190) and recorded in subsequent compilations of Commission rulings: The Official List of Specific Names in Zoology (1958), Official Lists and Indexes of Names and Works in Zoology (1987), and Official Lists and Indexes of Names and Works in Zoology, Supplement 1986-2000 (2001).
  Linnaeus (1758, p. 47), however, recorded `Ursus maritimus albus major, arcticus' under Ursus arctos (the brown bear), and he did not adopt `maritimus' as a valid, or even binominal, name. In these circumstances `maritimus' is not available (Articles 5.1 and 11.5 of the Code). Linnaeus referred to Martens's (1675, p. 73, pl. O, fig. C) description and illustration in Spitzbergische oder Groenlandische Reise Beschreibung and noted `forte distincta species est, nobis non visa' [perhaps a different species, I have not seen it], indicating that he doubted that it was a true species distinct from the brown bear.
  The entry for `Ursus maritimus albus major' in Linnaeus's Systema Naturae, Ed. 12 (1766) is a repeat of that in Ed. 10 (1758) with the added words `capite longiore, collo angustiore'.
  It seems that the first author to make available a name for the polar bear was Phipps (1774, p. 185). His description of Ursus maritimus was very brief (`This animal is much larger than the black bear'), but measurements were given and there was an unambiguous reference to Pennant's (1771) Synopsis of quadrupeds, which included a detailed description and illustration (p. 192, pl. 20, fig. 1) of the `Polar Bear' (but no latin name). The reference to Pennant's text and plate renders Phipps's name available by indication (Articles 12.2.5 and 12.2.7) even if Phipps's own description is considered to be insufficient for availability.
  Subsequently the name Ursus maritimus was adopted in Schreber (pl. 141, 1776 and p. 513, 1777), Erxleben (1777, p. 160) and Gmelin (1788, p. 101). Nineteenth century authors cited the name maritimus from a variety of sources, including those just mentioned, but rarely from Linnaeus (1758). Palmer (1904) was probably the first to cite the name from Phipps (1774) and this has been followed by nearly all subsequent authors.
  It is clear that attribution of the name maritimus to Linnaeus (1758) in Opinion 384 was an error. Authors both before (see, for example, Ellerman & Morrison-Scott, 1951) and after (for example, Corbet, 1978 and Wilson & Reeder, 1993) the 1956 ruling have cited Phipps (1774) as the author and date of the name; this practice should be continued and the entry on the Official List should be corrected.
  The name Ursus marinus, independently proposed by Pallas (1776, p. 691), is a junior synonym of Ursus maritimus Phipps, 1774.

References

Corbet, G.B. 1978. The mammals of the Palaearctic Region: a taxonomic review. 314 pp. British Museum (Natural History), London.
Ellerman, J.R. & Morrison-Scott, T.C.S. 1951. Checklist of Pataearctic mammals 1758 to 1946. 810 pp. British Museum, London.
Erxleben, J.C.P. 1777. Systema Regni Animalis ... Classis 1, Mammalia. xlviii, 636, [64] pp. Lipsiae.
Gmelin, J.F. 1788. Caroli a Linne, Systema Naturae, Ed. 13, vol. l, part 1 (Mammalia). xii, 232 pp. Lipsiae.
Gray, J.E. 1825. On the genus Ursus of Cuvier, with its divisions into subgenera. Annals of Philosophy, (N.S.)10: 59-62.
Linnaeus, C. 1758. Systema Naturae, Ed. 10, vol. 1. 824 pp. Salvii, Holmiae.
Linnaeus, C. 1766. Systema Naturae, Ed. 12, part 1. 530 pp. Salvii, Holmiae.
Martens, F. 1675. Spitzbergische oder Groenlandische Reise Beschreibung ... 1671. [6], 232, 3 pp., [14] pls. Hamburg.
Pallas, P.S. 1776. Reise durch verschiedene Provinzen des Russischen Reichs, vol. 3. [18], 760, 23 pp. St Petersburg.
Palmer, T.S. 1904. Index Generum Mammalium: a list of the genera and families of mammals. 984 pp. North American Fauna, No. 23. U.S. Department of Agriculture, Washington.
Pennant, T. 1771. Synopsis of quadrupeds. xxv, 382 pp. Chester.
Phipps, C.J. 1774. A voyage towards the North Pole undertaken ... 1773. viii, 253 pp., 14 pls. Nourse, London.
Schreber, J.C.D. von. 1776, 1777. Die Skugthiere, vol. 3. Pls. 92B, 139-145 (1776), pp. 457-584 (1777). Leipzig.
Wilson, D.F. & Reeder, D.M. (Eds.). 1993. Mammal species of the world. A taxonomic and geographic reference, Ed. 2. xviii, 1206 pp. Smithsonian Institution Press, Washington.

 

Availability of zoological names published in theses

P.K. Tubbs
Executive Secretary, International Commission on Zoological Nomenclature

  The Commission Secretariat is frequently asked about the availability of names (and nomenclatural acts, such as the designation of type species) from their publication in theses, and it may be helpful to state the position.
  There has never been a provision in the Code to the effect that a name or act cannot be made available from its appearance in a thesis. It follows that if a thesis is `published' in the sense of the Code (Articles 8 and 9) names and acts in it will be available if the other necessary conditions are met.
  However, extremely few theses count as published works, because nearly all fail to meet all the requirements of those Articles. Even if numerous copies are printed these are usually only deposited in prescribed libraries or distributed to colleagues of the author - they are not `obtainable, when first issued, free of charge or by purchase' by the zoological public, and therefore they do not satisfy Article 8.1.3; the subsequent supply of copies in response to individual requests would not satisfy Article 9.7.
  Abstracts of theses often appear in works which clearly are published in the sense of the Code; a name could be available from such an abstract but only if qualifying information (e.g. description and typification of the taxon) also appeared in it. This is not usually the case, however, and after 1999 is particularly unlikely in the case of a species since under Article 16.4 a holotype or syntypes must be explicitly fixed to establish the name.
  Many theses do contain proposed new names and nomenclatural acts, since these are indispensable for treatment of the subject matter. The author of such a thesis should include in it a disclaimer (Article 8.2) to the effect that the thesis is not to be taken as published for the purposes of zoological nomenclature or within the meaning of the Code. Disclaimers should also be provided by editors of all works which include abstracts of theses so that names and acts are not made available unintentionally. As a corollary of this, people who are aware of new names in theses should take great care not to cite those names in their own publications before the author has made them available.
  The recomendations in Appendixes A and B of the Code and those attached to Articles 8 and 9 give further guidance on the publication of new taxonomic names.

 

General Article

Phylogenetic Nomenclature and the PhyloCode

Kevin de Queiroz
Department of Systematic Biology, National Museum of Natural History, Smithsonian Institution, Washington, DC 20560-0162, U.S.A.

Philip D. Cantino
Department of Environmental and Plant Biology, Ohio University, Athens, OH 45701-2979, U. S. A.

  In a recent paper Forey (2001; BZN 58: 81-96) provided a description of the draft Phylogenetic Code of Biological Nomenclature (PhyloCode; http://www.ohiou.edu/phylocode/), followed by a largely negative commentary. Several of Forey's criticisms of the system of phylogenetic nomenclature embodied in the PhyloCode stem from misunderstandings about that system, and several confuse taxonomic with nomenclatural issues. In fact, the PhyloCode would regulate the naming of taxa and the subsequent application of taxon names in ways that are thoroughly consistent with the taxonomic approach that he advocates. In this essay, we comment upon some aspects of Forey's description of the draft PhyloCode, and we explain why none of his criticisms represent serious problems.

Forey's Part 1 (Goals and Mechanics of the PhyloCode)
  Part 1 of Forey's paper was intended to provide readers with an impartial description of the goals and mechanics of the PhyloCode. This section is largely accurate but omits some important issues, which we would like to describe, and contains some misleading statements, which we would like to clarify.

Motivation for the PhyloCode
  One important topic omitted by Forey is a discussion of the pragmatic issues that motivated development of the PhyloCode. The PhyloCode is designed to make explicit the reference of taxon names to clades, and thereby bring the subsequent application of taxon names into line with contemporary (i.e. evolutionary) conceptualizations of taxa (de Queiroz & Gauthier, 1994; de Queiroz, 1997). In so doing, it simplifies the process of naming clades and thereby facilitates communication about phylogeny. The need for an effective and efficient system for naming clades is particularly urgent now, as the unprecedented progress in phylogenetics in the past decade is likely to accelerate even further in the coming years, and the current system of nomenclature, as embodied in the International Code of Zoological Nomenclature (Zoological Code) and its botanical and bacteriological counterparts, is poorly suited to govern clade names. Under the current system, authors use the same names for different clades, and different names for the same clade, even when there is no disagreement about relationships and composition (de Queiroz & Gauthier, 1994; de Queiroz, 1997). Moreover, many newly discovered clades, even well-supported ones, are currently left unnamed, at least in part because it is often difficult: (1) to name clades one at a time (in the way that species are named as they are discovered) without having to develop an entire new classification and thus change the names of (Kron, 1997; Hibbett & Donoghue, 1998), and (2) to name those clades that one wants to name without having to recognize groups that one does not want to recognize (Cantino, 2000).
  The feature of the traditional system that underlies all of these problems is the link between names and ranks. Because of this link, authors who agree about the relationships and composition of clades but disagree about ranks will use different names for the same clade and the same name for different clades. Moreover, because a clade must be given a rank in order to name it, naming a newly discovered clade under the Zoological Code may require developing a new classification, which authors may be reluctant to do. The ranks of all taxa in a classification are interdependent. Therefore, depending on the availability of unoccupied ranks, naming a new clade may cause a cascade of name changes at higher or lower levels in the hierarchy when clades that include or are included within the newly discovered clade shift in rank and must therefore be renamed (Kron, 1997; Hibbett & Donoghue, 1998). Finally, because the genus rank is mandatory, and others (e.g. family) are treated by convention as though they were mandatory, naming a new clade may necessitate naming other taxa at the same rank even though one does not accept those taxa because they are paraphyletic, redundant (monotypic), or poorly supported (Cantino et al., 1999; Cantino, 2000). Under the PhyloCode, these problems do not exist because taxonomic rank has no bearing on the spelling or application of names. Instead, names are linked directly and explicitly to clades through phylogenetic definitions.

Similarities and differences between traditional (rank-based) and phylogenetic nomenclature
 Another important issue not discussed by Forey concerns the fundamental similarities and differences between the PhyloCode and the Zoological Code (and the other codes of rank-based nomenclature). Regarding similarities, the PhyloCode has the same general goals as the Zoological Code, namely, the provision of rules for naming taxa and applying existing names in new taxonomic contexts so that the names of taxa, and the application of names, will be unambiguous within a given taxonomic context. In addition, the PhyloCode is like the Zoological Code in attempting to promote stability and universality in the names of taxa and the application of names, so far as that is possible given that both codes permit disagreements concerning taxonomic hypotheses. Moreover, the PhyloCode accom¬plishes these goals using the same general mechanisms as in the Zoological Code, that is, by establishing precedence (an order of preference) among synonyms or homo¬nyms, which is normally based on priority of publication (seniority) but which allows for exceptions (usually through rulings by a commission or committee) in cases when using priority to determine precedence would compromise nomenclatural stability or universality.
  The main difference between the PhyloCode and the Zoological Code concerns the manner in which names are linked to taxa. In both cases, names are linked to taxa using definitions, but differences between the types of definitions used under the two codes result in differences in how names are applied in new taxonomic contexts and thus which names are regarded as synonyms. (It should be noted that the definitions referred to here are statements specifying how names are to be applied, as opposed to statements describing the characters of the taxa to which the names refer.) Forey described three categories of phylogenetic definitions (the type of definitions used in the PhyloCode), and illustrated how a particular name (`Aves') might be defined using definitions in each of the three categories (i.e. node-based, stem-based, and apomorphy-based). He did not, however, describe the rank-based definitions used in traditional nomenclature. This omission is important both because rank-based definitions, though they are the foundation of the Zoological Code and other traditional codes, are not described explicitly in those codes (instead, their use is implied by the way traditional nomenclature works), and because the difference between rank-based and phylogenetic definitions is the most fundamental difference between traditional and phylogenetic nomenclature.
  In contrast with phylogenetic definitions, which are based on the phylogenetic relationships of designated specifiers (e.g. `Aves' is the name of the least inclusive clade containing (say) Struthio camelus and Corvus corax), traditional definitions are based on the ranks of taxa containing designated types. Thus, to use the same name used by Forey in his examples, `Aves' is the name of the class containing (say) Corvus corax. This example is, of course, hypothetical, since the Zoological Code does not extend its principle of typification (and thus its method of definition) to names above the level of the family group. To use a real example, `Corvidae' is the name of the family containing Corvus. The fundamental difference between phylogenetic and traditional definitions results in an important difference regarding the associations between names and clades. Phylogenetic definitions tie names directly to clades; in contrast, traditional definitions tie names to clades only indirectly through the ranks to which the clades are assigned. The most important consequence of this difference is that names in phylogenetic nomenclature are more strongly tied to clades than to ranks (i.e. in the face of changing taxonomic proposals), while in traditional nomenclature the reverse is true - names are more strongly tied to ranks than to clades (de Queiroz, 1997). This difference underlies both the problems with traditional nomenclature and the advantages of phylogenetic nomenclature described in the previous section.

Phylogenetic definitions and specifiers
  Regarding definitions, a few statements in Forey's Part 1 are potentially misleading. On p. 84, Forey stated (para. 3) that specifiers (species, specimens, or apomorphies cited in a phylogenetic definition to specify the clade to which the name applies) `serve exactly the same function as Linnaean types except their characters do not define the clade'. There are two ways in which this statement may be misleading. First, while it is true that the specifiers of phylogenetic nomenclature and the name-bearing types of traditional nomenclature both serve as reference points for the application of names, there are also differences in their functions. The most fundamental difference is that specifiers are so called because they specify the taxon to which a name refers. Thus, the specifiers of phylogenetic nomenclature are used, as parts of phylogenetic definitions, to specify particular clades. In contrast, in traditional nomenclature types do not, by themselves, specify particular taxa (clades or otherwise) because several nested taxa may contain a given type. A rank is needed to restrict the reference of the name to one of the several nested taxa containing that type, and thus, in one sense, the specifiers of traditional nomenclature are both typesand ranks. However, in another sense, traditional definitions do not really specify particular taxa (i.e. `taxonomic taxa' in the sense of the Zoological Code-that is, taxa that are conceptualized in terms of composition, characters, or relationships, rather than solely in terms of a rank and a type), because a given name can be applied to any one of several taxa in a nested series, depending on which one is assigned the specified rank. In this sense, types are not really specifiers at all. Therefore, regardless of whether taxa are conceptualized solely in terms of ranks and types, types are not functionally equivalent to specifiers.
  Another difference between types and specifiers (related to the fundamental difference described above) is that single types are used in traditional definitions while, in contrast, multiple specifiers are required in phylogenetic definitions. Furthermore, under the traditional codes, the type used to define a name in the family group provides the stem of the name of the taxon of which it is the type (e.g. Zoological Code, Article 29). In contrast, under the PhyloCode, (1) the specifiers used to define clade names need not provide the stem of the name of the specified clade (e.g. neither Struthio camelus nor Corvus corax provide the stem of the name `Aves' in the above example), (2) one or more of the specifiers can serve this function (e.g. Corvus corax for `Corvidae'; Gallus gallus and Anser anser for `Galloanserae'), and (3) when a specifier provides the stem of a clade name, it does so regardless of rank.
  An additional problem is that Forey's statement could be interpreted as implying that the characters of types define clade names while those of specifiers do not. In fact, the characters of neither types nor specifier species or specimens define the names of clades. In the case of types, a clade name is defined in terms of the rank of the group that contains the type, rather than the characters of the type (e.g. Corvidae = the family containing Corvus). In the case of specifier species or specimens, a clade name is defined in terms of the relationships of the specifiers, rather than their characters (e.g. Corvidae = the least inclusive clade containing (say) Corvus corax and Platylophus galericulatus). The only characters that are used to define clade names are specifier apomorphies, which are used in apomorphy-based phylogenetic definitions (e.g. Diapsida = the clade stemming from (say) the first amniote to evolve two temporal fenestrae homologous with those in Sphenodon punctatus) but not in the rank-based definitions of traditional nomenclature. Of course, regardless of whether one adopts traditional or phylogenetic nomenclature (and regardless of the type of phylogenetic definition used), the relationships and composition of taxa are inferred using characters. This, however, is a taxonomic rather than a nomenclatural issue. Compositional changes and nomenclatural stability
  Although Forey's Part 1 was intended to be impartial in its portrayal of the PhyloCode, a subtle bias was introduced through his choice of examples (see particularly his Fig. 2), all of which concern taxa ranked above the level of the family group. The names of such taxa are not defined (i.e. according to rank and type) and, for the most part, are not regulated by the Zoological Code. Consequently, Forey implicitly contrasted the ramifications of the PhyloCode not with those of the Zoological Code but with the more or less total nomenclatural freedom that would exist in the absence of any code. Most of the rank-based problems cited above, which come into play when taxon names are defined according to rank and type, do not apply to the names of zoological taxa at ranks above the family group, where the principle of typification does not extend. As a result, readers of Forey's paper who might not like the changes in taxon composition that occur when names are applied in the context of different phylogenetic hypotheses under the PhyloCode (Forey's Fig. 2) might overlook the fact that similar changes in taxon composition occur under the traditional codes (de Queiroz, 1997). Changing ideas about phylogeny cause changes in the hypothesized composition of taxa under both systems, but under the Zoological Code, unlike the PhyloCode, additional instability in the names of clades and the membership of taxa results from changes in rank (i.e. through `lumping' and `splitting') even when ideas about phylogeny are stable (de Queiroz, 1996, 1997; Bryant & Cantino, in press). Furthermore, with regard to zoological names above the level of the family group, the PhyloCode will increase nomenclaturai stability. Currently, no code governs the definition and application of these names, and thus, there is nothing to prevent the capricious renaming of clades - that is, the replacement of existing names with unnecessary substitute names.

The primary function of taxon names
  Forey's discussion (p. 85) of PhyloCode Principle 1 suggests a misunderstanding of its intent. Principle 1 states that `the primary purpose of taxon names is to provide a means of referring to taxa, as opposed to indicating their characters, relationships, or membership'. This statement is adapted from item 1 in the Preamble of the International Code of Botanical Nomenclature. Its purpose is to describe the principle that although taxon names often describe the characters (e.g. Gnathostomata = jaw mouth), relationships (e.g. Paradipsosaurus = near Dipsosaurus), or membership (e.g. Galloanserae = Galliformes plus Anseriformes) of the taxa to which they refer, conveying such information is a secondary function of taxon names; the primary function is to supply a means of referring to taxa. Consequently, the PhyloCode does not permit rejection of a name simply because the name does not accurately describe the characters, relationships, or composition of the taxon to which it refers. The same is true under the Zoological Code (see Article 18). Thus, Paradipsosaurus is still the valid name of a taxon, even though that taxon is no longer thought to be closely related to Dipsosaurus (Estes, 1983).
  It appears that Forey misinterpreted Principle 1 by confusing taxonomic and nomenclatural issues. He quoted that principle in three successive paragraphs (p. 85) to point out three different properties of phylogenetic nomenclature: (1) that `a shift in taxon membership with changing ideas of phylogeny is perfectly acceptable to the PhyloCode'; (2) that `ideas of relationships can vary substantially. . . but ... there will always be some position ... on a phylogeny where [a name] will apply'; and (3) `a name is applied to a phylogeny without reference to why that phylogeny should have been chosen'. Forey described these properties as if they were undesirable, but all three are also properties of traditional nomenclature (or at least have analogs therein). Thus, in traditional nomenclature: (1) changes in taxon membership often result from changing ideas about phylogeny; (2) ideas about relationships can vary substantially, but certain names will always apply to some taxon, and (3) names can be applied in the context of a taxonomic proposal without reference to the justification for adopting that proposal. All of these properties, which are common to both traditional and phylogenetic nomenclature, are related to the basic principle that nomenclatural codes do not infringe upon taxonomic judgement but only regulate the application of names (Zoological Code, Principles 1 and 2; PhyloCode, Principle 6). These properties are neither unique to phylogenetic nomenclature nor problematical.

Synonymy
  Another point of confusion in Forey's Part 1 concerns synonymy. Forey stated (p. 87) that under the PhyloCode `With regard to synonymy there is the possibility of two names specifying the same taxon but since they may be defined in different ways (e.g. stem- and node based) they may both be valid'. To support this conclusion, Forey cited PhyloCode Note 14.1.2, which reads: 'Node-based, apomorphy-based, and stem-based definitions (Note 9.4.1) usually designate different clades, although they may be nested clades that differ only slightly in inclusiveness. Therefore names based on two or more of these different kinds of definitions are usually not synonyms'. The qualifier `usually' was included to cover the rare possibility that names defined using different kinds of phylogenetic definitions might refer to the same clade (e.g. if the apomorphy specified in an apomorphy-based definition originated (or became fixed) at precisely the same moment as the divergence (from its sister lineage) of the stem lineage specified by a stem-based definition). However, in this rare event, the names in question would be synonyms despite their being based on different types of definitions (such `hetero¬definitional synonyms' are analogous to names that the Zoological Code terms `subjective synonyms' in that the conclusion that they refer to the same taxon depends on a taxonomic judgement). According to the PhyloCode (Principle 3 and Article 14.2), if two names denote the same taxon, then they are synonyms and cannot both be valid (in the terminology of the Zoological Code ='accepted' in the terminology of the PhyloCode). Thus, although Forey is correct in pointing out that names defined using different types of phylogenetic definitions can sometimes refer to the same taxon, he is incorrect in stating that more than one such name can be valid.

Forey's Part 2 (Commentary)
  The second part of Forey's essay is explicitly critical of phylogenetic nomenclature and the PhyloCode. Forey's criticisms, however, either misrepresent the PhyloCode or are no more problematical for phylogenetic nomenclature than for its traditional counterpart. In this section, we address each of Forey's criticisms and show that phylogenetic nomenclature stands up to every one.

Taxonomic ranks
  In the introduction to his commentary (pp. 88-89), Forey incorrectly implied that phylogenetic nomenclature and the PhyloCode require the abolition of taxonomic ranks. Although it is true that some advocates of phylogenetic nomenclature favor the abolition of ranks and that the nomenclatural system described by the PhyloCode is rankless (Article 3.1), adoption of phylogenetic nomenclature and the PhyloCode does not require the elimination of ranks. The statement that the system of nomenclature is rankless does not mean that taxa cannot be assigned to ranks (de Queiroz, 1997); instead, it means only that `assignment of a categorical rank (e.g. genus, family, etc. ) is not part of the formal naming process and has no bearing on the spelling or application of taxon names' (Article 3.1). In other words, if a name refers to a clade, then changing the rank of that clade does not cause a change in its name (de Queiroz, 1997). Under the traditional system, changing the rank of a taxon from family to subfamily, for example, requires a change in the name of that taxon (e.g. from `Corvidae' to `Corvinae'). Under the PhyloCode, the same change in rank would not result in a name change. In any case, the PhyloCode does not prohibit the use of ranks, and therefore, Forey's concerns about its effect on the assessment of biodiversity are unfounded. Biologists will still be able to rank taxa, if they so desire, and thus to count numbers of taxa at particular ranks.
On the other hand, there are problems with these simple counts of equally ranked taxa. For one thing, such counts generally do not distinguish between monophyletic and paraphyletic taxa (Smith & Patterson, 1988; Smith, 1994). Moreover, it is widely acknowledged that taxa of the same rank generally are not comparable with respect to any biologically significant property, such as age, number of species, or disparity (Hennig, 1966; Mayr, 1969; Mayr & Ashlock, 1991), and that rank assignment is largely subjective, varying from one taxonomist to another (Simpson, 1961; Mayr & Ashlock, 1991). The PhyloCode's de-emphasis on ranks permits (without requiring) the abandonment of ranks and thus encourages biologists to develop more meaningful ways of assessing diversity. One obvious possibility is to count numbers of species (i.e. separately evolving lineages). Another possibility is to count the number of mutually exclusive clades possessing properties that are relevant to the question being addressed. For example, one might count the (minimum or maximum) number of non-nested clades that originated or became extinct in a particular time period, or the number that are characterized by organisms exhibiting different natural history strategies with regard to reproduction (e.g. oviparous, viviparous), feeding (e.g. carnivorous, herbivorous), metabolism (e.g. ectothermic, endothermic), etc. To assess overall similarity or disparity, multivariate measures can be used (e.g. Foote, 1995) rather than using subjectively assigned ranks. And, in biodiversity inventories, organisms that cannot be assigned to a species can still be assigned to more inclusive clades, regardless of whether those clades are ranked. In short, the PhyloCode's de-emphasis on ranks, rather than hindering studies of biodiversity, might actually contribute to the development of improved methods for such studies.

Annotated Linnaean systems
  Because ranking is often associated with the recognition of paraphyletic taxa, Forey himself has `some sympathy' (p. 89) for the development of rank-free approaches. On the other hand, he believes that `there are ways around the problem which do not involve the adoption of a PhyloCode' (p. 89, para. 2), specifically `the annotated Linnaean system' (p. 89) developed by authors such as Nelson (1973), Patterson & Rosen (1977), and Wiley (1979). Forey's statements are misleading on several counts, which (in addition to resting on the incorrect premise that phylogenetic nomenclature prohibits the use of ranks) result from his not distinguishing consistently between taxonomy and nomenclature. First, although it is true that paraphyletic taxa can be eliminated and the relationships of monophyletic taxa can be conveyed using annotated Linnaean systems, these are taxonomic solutions that are logically and pragmatically separate from the nomenclatural problems that the PhyloCode is designed to solve. Rather than being designed to convey the relationships of monophyletic taxa (clades), the PhyloCode is designed to prevent unnecessary changes in the associations between taxon names and clades that result under the Zoological Code from changes in taxonomic ranks. This nomenclatural problem is not addressed by the annotated Linnaean system advocated by Forey, which consists of conventions - such as phyletic sequencing (Nelson, 1973) and the plesion category (Patterson & Rosen, 1977) - designed to reduce the proliferation of taxonomic ranks, as well as other conventions for representing polytomies, uncertain placement within a larger clade, non-monophyletic groups, ancestors, taxa of hybrid origin, and distinctiveness (Wiley, 1979, 1981). Most of these conventions are taxonomic rather than nomenclatural in nature and are compatible with both traditional and phylogenetic nomenclature. In any case, they do not solve the problem of rank changes causing name changes.
  It is worth pointing out that several of the conventions of the annotated Linnaean system advocated by Forey de-emphasize the use and importance of ranks and might therefore be considered to anticipate the development of phylogenetic nomenclature in this regard (de Queiroz, 1997). For example, the sequencing convention (Nelson, 1974) uses the sequence of taxon names in a list, rather than ranks, to convey information about relationships. Similarly, the plesion, a category used for extinct taxa regardless of their position in the taxonomic hierarchy, is basically a rankless category. It might even be argued that the plesion category is incompatible with traditional nomenclature, given that it is rankless and that ranks are necessary for traditional nomenclature. In short, the conventions advocated by Forey do not constitute an alternative to the PhyloCode; instead, most are taxonomic conventions the use of which is entirely compatible with phylogenetic nomenclature.

Types and specifiers
  Forey argued (p. 89) that there is no fundamental difference between the specifiers of the PhyloCode and the name-bearing types of traditional nomencla¬ture and that the replacement of types by specifiers in the PhyloCode is therefore unnecessary. As explained above (see Phylogenetic definitions and specifiers), types and specifiers have both similarities and differences, though the concept of a specifier is more general than the concept of a type. Thus, specifiers include not only specimens and taxa, but also apomorphies in phylogenetic nomenclature and ranks in traditional nomenclature. Some other differences are as follows. (1) Although both specifiers and types serve as reference points for the application of names, the use of multiple reference points (specifiers) is necessary in phylogenetic nomenclature because a single specimen or subordinate taxon cannot unambigu¬ously specify a clade in the way that a single type can unambiguously specify a ranked taxon. (2) Types are necessarily included within the taxon whose name they are used to define, while in stem-based phylogenetic definitions, some specifiers (called `external specifiers' in the PhyloCode) are necessarily excluded from the specified clade (as noted by Forey on p. 84). (3) In contrast with the rule of the Zoological Code that the name of a taxon in the family group must be formed from the stem of the name of the type genus, the PhyloCode does not require that the name of a clade be formed from the stem of the name of one of the specifiers used to define that name. Given these differences between types and specifiers, introduction of the new term `specifier(s)' in the PhyloCode is appropriate.

When to name
  Forey made much of the statement in the PhyloCode Preface that `Criteria that influence the decision whether to name a clade include level of support, phenotypic distinctiveness, economic importance, etc'. He referred (p. 90) to this as a `recommendation' of the PhyloCode and concluded that `advocates of phylogenetic taxonomy really do not have any more precise reasons for naming a group than do followers of Linnaean Taxonomy and to include advice in the PhyloCode registers a precision which is both unnecessary and undesirable'. This criticism is misdirected. For one thing, advocates of phylogenetic nomenclature do not claim to have more precise or objective reasons for naming taxa than do practitioners of traditional nomenclature. Such decisions are taxonomic, not nomenclatural, and therefore are beyond the scope of both the PhyloCode and the Zoological Code. Moreover, contrary to Forey's assertion, the PhyloCode does not include advice about when to name a clade. The statement that he quoted is in the Preface, and although there are many formal recommendations in the PhyloCode itself, this is not one of them. It was included in the Preface simply to elaborate on the preceding statement that not all clades need be named. Furthermore, it is difficult to see how the statement itself, which lists only very general criteria and ends in `etc'., conveys an unwarranted level of precision. It should be apparent from both the context and the wording that none of the cited criteria is definitive, and that the list is not exhaustive. The listed criteria are simply examples of criteria that would generally be considered when one is deciding whether to name a clade.
  In this context, Forey's criticisms of the specific criteria lose their force. The PhyloCode is entirely neutral regarding the various measures of support that he lists (number of synapomorphies, Bremer support, bootstrap proportions, etc.); what is considered an adequate level of support is a taxonomic issue that is to be decided by the individual systematist. The same holds for levels of phenotypic distinctiveness and economic importance. Incidentally, Forey's point that the criterion of pheno¬typic distinctiveness implicitly advocates use of apomorphy-based definitions but that `apomorphy-based naming is less favoured than the other two [kinds of] definitions' (p. 90) is both questionable and irrelevant. For one thing, at least some PhyloCode proponents have argued for the use of apomorphy-based definitions (e.g. Pleijel, 1999; Lee, 2001; see also Gauthier & de Queiroz, 2001). Moreover, regardless of the types of phylogenetic definitions favored by individual systematists; there is nothing in the PhyloCode indicating that one kind of definition is preferred over others.

Compositional stability
  In his section entitled `How to name', Forey first argued (p. 91) that phylogenetic nomenclature is `curiously illogical' in attempting to choose definitions that will promote stability in the composition of taxa given that `taxonomic content is not the primary purpose of Phylogenetic Nomenclature (PhyloCode, Division 1. Principles)'. His argument, however, is based on his misinterpretation of PhyloCode Principle 1 (see The primary function of taxon names), which does not state that compositional stability is unimportant but only that the primary purpose of taxon names is to refer to taxa rather than to describe (i.e. through the meanings of the words from which the name is formed) their composition (or characters or relationships). Forey then correctly noted that stability in taxon composition will depend on the stability of the phylogenetic hypothesis, but then he reiterated his irrelevant complaint that `Phylogenetic Nomenclature is mute in offering guidelines since there are no agreed criteria [for assessing support]', concluding (again correctly) that although the name itself may remain stable, the composition of the taxon to which it refers `may be decidedly unstable'. As we argued above (see When to name), the issue of support is a taxonomic rather than a nomenclatural issue. In addition, neither traditional nor phylogenetic nomenclature can guarantee compositional stability. On the other hand, under phylogenetic nomenclature, changes in taxon composition result only from changes in hypotheses about phylogenetic relationships, while under traditional nomenclature, such changes can result both from changes in phylogenetic hypotheses and from changes in rank assignments, and the latter can occur even when ideas about phylogenetic relationships remain unchanged (de Queiroz, 1997). Moreover, phylogenetic definitions can be worded so as to limit potential changes in taxon composition (see PhyloCode Article 11.9), an option that is unavailable under the Zoological Code. Thus, far from highlighting shortcomings of phylogenetic nomenclature, the issue of compositional stability reveals significant advantages of that approach.

Nomenclatural stability
  Later in his section titled `How to name', Forey argued (p. 91) that PhyloCode rules regarding conservation can lead to instability in names (as opposed to taxon composition). In his hypothetical example, identical definitions are given to the names `Sarcopterygii' and `Gnathostomata', followed by conservation of `Sarcopterygii' and redefinition of `Gnathostomata', so that the application of the name `Gnathostomata' is unstable. This example is flawed in several ways. First, under the PhyloCode, the establishment of different names with identical definitions will be very unlikely to occur because all names and their definitions will be registered (see Article 8). The implementation of the PhyloCode will coincide with the establishment of a registration database, which will be accessible through the Internet. In addition to providing a useful entry to the literature relevant to particular names, this database will make it very easy for authors to avoid accidentally publishing homodefinitional synonyms (i.e. the sort in Forey's example) and homonyms. An author who proposed to give the name `Sarcopterygii' the same definition that had previously been published for `Gnathostomata' would have to register the name and definition, and the registration number would have to be included in the publication, in order for the name to be established under the PhyloCode (i.e. be `available' in the terminology of the Zoological Code). If a definition submitted for registration were identical to one that had previously been registered, the submitting author would be notified (see PhyloCode Appendix A). It is very unlikely that the author would then proceed to publish that definition, knowing that it could never be accepted (i.e. be `valid' in the terminology of the Zoological Code) unless it were conserved by the International Committee on Phylogenetic Nomenclature (ICPN).
  On the other hand, suppose that the earliest phylogenetic definition of the name `Gnathostomata' (e.g. the least inclusive clade containing the specifiers coelacanth and frog, symbolized `clade (coelacanth + frog)' though under the PhyloCode one would use scientific names of species for the specifiers) were highly inconsistent with prevailing use and ended up referring to a taxon that had traditionally been called `Sarcopterygii' (as in Forey's hypothetical example). Under these circumstances, an author might purposely publish the same definition (i.e. clade (coelacanth + frog)) for the name `Sarcopterygii' and then apply for conservation. If the ICPN agreed that stability would be promoted by conserving Sarcopterygii = clade (coelacanth + frog) over Gnathostomata = clade (coelacanth + frog), it would formally suppress the latter name-definition combination, and, as Forey stated, `Gnathostomata' could then be redefined (e.g. as clade (shark + frog)). In Forey's view, `this is hardly stability' (p. 91). On the contrary, permitting redefinition of taxon names following suppression enhances stability in that it permits their continued use in a manner consistent with prevailing use. Otherwise, a well known name such as `Gnathostomata' might have to be abandoned simply because the first definition published for it was inappropriate.