The Linnaean system and its 250-year persistence.
Biological nomenclature is a communication system. It provides a way for systematists and others to talk about taxa. As such, names have meaning only when those who use them jointly understand the concepts attached to them. Those concepts may refer to the attributes of individual taxa, such as species, as well as connote information about relationships among taxa. The effectiveness of a system of nomenclature, then, might best be judged on its ability to transmit such types of information. Whereas there is a substantial body of literature investigating the ability of hierarchical systems to transmit taxonomic information--much of it embodied in the debates among cladists, pheneticists, and evolutionary taxonomists--there seems to be little in the way of what I would view as objective analysis of the attributes of systems of nomenclature.
There is the widely pandered idea that biological nomenclature--especially as attributed to Linnaeus--is a mess. Those who promote this point of view argue that if we could only make nomenclature stable, not only would communication about taxa proceed in a more harmonious fashion but the taxonomic enterprise itself would be disencumbered of the Linnaean nomenclatural albatross that has long hung around its neck. They further argue that the existing nomenclatural approach--what I will call "Linnaean"--lacks an evolutionary foundation. According to this view of systematics, we have methods appropriate for discovering lineage relationships, but the methods of naming those lineages are not themselves properly evolutionary.
In the following pages I will describe what I believe to be some salient aspects that have allowed the Linnaean system to serve biologists well for nearly 250 years. I will then outline some of the criticisms of the Linnaean system and describe why I believe that most of those arguments are specious and misguided. Nixon and Carpenter (2000) published what appears to have been the most extensive rebuttal of arguments for "phylogenetic nomenclature" before the appearance of the articles in this issue of the Botanical Review For the most part I will not repeat subjects treated by them but, rather, offer what I believe to be new or extended criticisms.
III. The Linnaean System
The system of nomenclature currently used in biology can be traced back to the works of Linnaeus. Hierarchical relationships of taxa are conveyed--at least in part--by the names themselves. In the original works of Linnaeus himself, examination of the taxon names allowed only for direct recognition of taxa at the level of genus and species. It was nearly 100 years before a system of standardized endings was developed and consistently applied to names in multiple ranks above the level of genus. The incorporation of additional formal hierarchical levels allowed for the description of increasingly more complex classifications. The adoption of standardized endings for most of the ranks greatly eased communication.
What was lacking in the Linnaean system of nomenclature until early in the nineteenth century was a consistent system for determining how a name should be applied--the fixation of concept, what we now refer to as "typification." The method chosen in all branches of biology was that of nomenclatural types. With the exception of some of the most highly ranked taxa, under this system groupings above the species-group level derive their names--and consequently concepts--from genera. The types of genera are species, the latter being the only grouping level in which specimens serve as types. The practice of recognizing nomenclatural types was first introduced in the Stricklandian Code, prepared in the 1840s, and relies on the concept of priority. Typification was codified in its essentially modem form in the early twentieth century in what we recognize today as the International Codes of Botanical and Zoological Nomenclature. The system of nomenclatural types exists in contrast to what Moore (1998) and others have called "conceptual types." The conceptual system would allow the concept applied to names to become the personal choice of individual investigators. Whereas it might at times produce a system of names with more obvious concepts, it would undoubtedly do so at the expense of continuity in the use of names.
Under the Linnaean system, as currently applied, information on hierarchical relationships is conveyed directly. At the least inclusive level, species are nested within genera. Because species names comprise a combination of generic and species epithets, the placement--and therefore relationships--of species are self-evident from the use of their names. At more inclusive levels the endings of the taxon names connote relative positions in the hierarchical system, up to a given point, depending on whether we are using the botanical or zoological code. Thus, within the limits of the number of levels recognized in the hierarchy, the system directly conveys information on group membership and on group relationships.
Most taxon names in biology have little or no meaning in and of themselves. Although the hierarchical system of Linnaean nomenclature can be quickly learned--and therefore the rank order and inclusion/exclusion relationships of names comprehended--information on the meanings of the names is still dependent on knowledge of biology. Thus the method for applying names is largely disconnected from biological issues, excepting the issue of hierarchical relationship; the names themselves become meaningful only when the concepts attached to them are understood.
Taxon concepts exist in the form of diagnoses. This was as true in the works of Linnacus and his contemporaries as it is today. The justification for the concepts--i.e., whether the groupings they connote represent works of divine creation or the products of organic evolution--is what has changed overtime. Furthermore, the fidelity with which the concepts (diagnoses) recognized at lower levels nest within those used at higher levels has varied depending on the quality of the investigation and the grouping methods applied. The scientific utility of concepts is very closely tied to the latter point. Under the cladistic paradigm, the concepts for paraphyletic groups are flawed and must be revised. Therefore, the composition of such groups will of necessity change, and the application of names will sometimes change as well when only monophyletic groups are eventually recognized.
Taxonomic concepts are theories. Like all theories in science, theories about taxa are subject to testing and modification or rejection. Testing of taxonomic theories is a desirable practice. Each of us might wish to protect our own favored theories, especially those of our own authorship. Nonetheless, if those theories are not maximally informative and predictive, they should be replaced. This is a scientific goal for which we should strive. The issue of what constitutes a taxonomic theory has been greatly clarified under the application of cladistic methods. Prior to ascendancy and refinement of the methods articulated by Willi Hennig (1966), the concept of a taxonomic theory was vague at best, and the validity of a taxonomic theory often rested on power of assertion and strength of reputation of its author. Those "tests by authority" have now been replaced by the application of parsimony as a way of judging character distributions. The change in taxonomic methodology--whether viewed as incremental or revol utionary--does not, however, fit well with one of the major tenets of nomenclature: the concept of stability (see Gafffley, 1979; Dominguez and Wheeler, 1997).
A major tenet of nearly all systems of nomenclature ever proposed is stability. For example, the most recent International Code of Zoological Nomenclature (ICZN, 1999) indicates in its preamble that one of the objects of the code is "to promote stability and universality in the scientific names of animals." The International Code of Botanical Nomenclature (Greuter et al., 2000) notes that the code "aims at the provision of a stable method of naming taxonomic groups," a statement which on its face places emphasis on the stability of the method rather than of the names themselves.
It may seem odd that the codes speak of establishing a stable nomenclature (or at least having a stable method of applying names), whereas much of what is in them has to do with the management of a system in which synonymy, homonymy, and recombination--all mediated under the concept of priority--are paramount concerns. So, by their very nature, the codes are designed not so much to create a stable nomenclature but possibly better said, "an orderly one which minimizes changes in names as they are applied to taxonomic concepts."
The issue of whether nomenclature lacks stability because of the system of naming itself, or whether instability is the result of the taxonomy to which names are applied, can be explored by examining some case studies. First we might examine the issue of catalogs or indices. In botany the oldest and best known of these would be the Index Kewensis. Many comparable compilations exist in zoology for vertebrates and nonvertebrates alike, although none with the scope of the Index Kewens is, and few with the history. What these types of compilations offer is an organized presentation of taxon names in a hierarchical format, in the best cases providing a detailed history of their usage and where to find information about the taxa to which they refer. They may give the impression, especially when newly prepared, that taxonomy is a neat and tidy enterprise and that if we just had catalogs (checklists) for all groups, we would have an accurate assessment of numerical and hierarchical aspects of biological diversity. In a recently published article, Edwards et al. (2000: 2313) commented, with regard to the preparation of Web-based lists of world scope, that "many species have been named more than once by different scientists, and, as our knowledge of species has changed over the years, species have been merged together or split into several new species. Thus there are many (perhaps three times as many) more scientific names than there are valid species. The scientific community has begun several international projects to sort out the nomenclatural confusion that has resulted."
On one hand, the statement could be interpreted to mean that if taxonomy had been better done there would be no synonymy or change of species concepts. On the other hand, it strongly suggests that once the synonyms and varied combinations have been sorted out through the cataloging process, we would have only to add heretofore undescribed taxa to the classification, and an accurate map of biological diversity would be at hand--that taxonomy could be frozen in time.
Reality, in my view, could be no farther from the truth. My own experiences at cataloging and producing revisions, comparable experiences of my colleagues at the American Museum of Natural History, and my examination of the literature indicate that although catalogs serve as indispensable tools for improving the organization of our knowledge of biological diversity, in the end they provide only a summary of the state of the art at any given point in time. Whether the existing taxonomy is good or bad will be determined by examination of the original literature, subsequent revisionary work, and additional specimen materials, processes greatly facilitated by a well-prepared catalog but for which there is no substitute. Now for the examples.
1. I recently completed a revision of a Holarctic group of bugs, the genus Plagiognathus Fieber, 1858 (Schuh, 2001). The most recent catalog for the group had been published just six years earlier (Schuh, 1995). Table I provides information on taxonomic changes resulting from the revision. The existing taxonomy, although appearing orderly enough in the catalog, dealt with a group that at once was paraphyletic with numerous species placed in other genera, had a large number of undescribed species, had many named taxa that ultimately were treated as junior synonyms, and included many species that did not belong to the group, some of these being moved to other genera and the remainder placed as incertae sedis in lieu of additional revisionary work. All of these changes were the result of interpretation of taxonomic concepts. In only one case was a name changed for strictly nomenclatural reasons, because of a case of secondary homonymy.
2. As a second example, let us consider the taxonomy of birds at the species level. According to Haffer (1997) the number of bird species reached nearly 19,000 early in the twentieth century. By the 1940s the number of species had been reduced to 8616. The numbers have now begun to rise again, with somewhere between 9000 and 10,000 species being recognized. These changes have almost nothing to do with the vagaries of nomenclature but, rather, result from the application of varying concepts of species. The reduction in species numbers between 1909 and 1946 came about with the rise of population biology and the application of the methods of the "new systematics." More recent increases in numbers of species are not so much the result of the description of new species--because, according to Haffer's figures, only 153 taxa were described as new species over a period of about 70 years. Rather, the changes in numbers are largely the result of applying a phylogenetic species concept, wherein minimally diagnosable ent ities are recognized as valid species. Again, these changes are not the result of the system of nomenclature but altered views of what represents the most appropriate taxonomic practice.
3. Within botany, up until about 20 years ago, the prevailing view divided seed plants into Gymnospermae and Angiospermae, and within the latter into Dicotyledoneae and Monocotyledoneae. This view is no longer considered valid, with both the Gymnospermae and the Dicotyledoneae now being recognized as paraphyletic. These and other dramatic changes in seed-plant classification have resulted primarily from the application of cladistic methods, as for example in the work of Bremer (1985) and others; more recently, additional data from DNA sequences have substantially influenced plant classification, as reflected, for example, in the work of Chase et al. (1993). The methods of nomenclature are irrelevant to these changes; under any definition of the term, the higher-level taxonomy of green plants--and seed plants in particular--over the last 20 years would be interpreted as anything but stable.
From this limited sample, we must conclude that taxonomy is by its very nature not a stable enterprise, and consequently neither can be the nomenclature that is tied to it. One might venture that one day all taxa will be described and their diagnoses agreed upon. This eventuality is, however, a long time in the future, if it is ever to be achieved. Although a significant proportion of the species of macroscopic organisms have been described, a substantial number still remain undiscovered and/or undescribed. As the examples of bugs and birds given above clearly show, however, just because one person says the species in a group are now recognized and adequately diagnosed, another worker my have a quite different opinion. Possibly more important, studies of taxonomy at the supraspecific level often lag far behind what exists at the species level. A completely dichotomous hierarchical system would require n - 1 higher taxa, where n is the number of species. Nowhere near that many higher-taxon concepts exist. In e ven the best-known groups the goal of achieving such a system is decades away. Furthermore, many of the higher-taxon concepts now applied are flawed, as clearly evidenced by the example of seed plants given above. The classifications (taxonomy) of even those higher taxa for which the greater majority of species have been described--e.g., birds and angiosperms--is in most cases far from phylogenetic. This being the case, it does not take much thought to realize that classifications for the vast majority of taxa are obviously going to require substantial additional work before they can be labeled as phylogenetic and therefore containing maximum scientific value, let alone having achieved a modicum of completeness in terms of descriptive work. For these reasons it is not reasonable to think of nomenclature as stable in an absolute sense. It is much preferable to think of the methods of applying names as being stable. This conclusion presupposes, of course, that names cannot he perfectly stable if the taxonomy to which they are attached is not stable. Not everybody adheres to this point of view, a subject that I will return to later in the article.
D. SIMPLICITY AND CONTINUITY
The Linnaean system provides a relatively simple and effective system of communicating hierarchical information. It is not the only possible system capable of communicating such information, but it now has a nearly 250-year history of doing so. It is that continuity which provides one of the strongest arguments for the persistence of the Linnaean approach and also for maintaining it. With at least 2 million names for biological taxa now in existence, any efforts to replace Linnaean nomenclature should be very well founded indeed. Otherwise, massive confusion will result, and much of the existing literature might be rendered essentially useless.
IV. Criticisms of the Linnaean System of Nomenclature
A. PHYLOGENETIC NOMENCLATURE
The main detractors of the Linnaean system are those who advocate what I will refer to as "phylogenetic nomenclature." Under phylogenetic nomenclature, methods of diagnosis, typification, and indication of hierarchical placement differ from those in the Linnaean system. Although not all details of methods for "phylogenetic nomenclature" are well worked out and the system is still little applied, the general approach can nonetheless be articulated; an extensive review of the literature can be found in Bryant and Cantino (2002). A brief description will be of help for understanding the remainder of this article.
Definitions: Its proponents argue that in phylogenetic nomenclature it is taxa that are defined, not names. Accordingly, definitions might be of three types: node based, stem based, or apomorphy based.
The choice as to which form of definition is most desirable and will produce optimal results under the broader philosophy of phylogenetic nomenclature is not clear from the literature. Indeed, Sereno (1999: 334) has argued that apomorphy-based definitions should be avoided in favor of node-based or stem-based definitions because of drawbacks inherent in the first, whereas Lee (2001: 177) has advocated the use of apomorphy-based definitions because they will be more stable than the other two types. What is clear is that most of the literature is organized around the concept of node-based definitions; furthermore, Kojima (2003) has persuasively argued that even apomorphy-based definitions are node based. For these reasons, the remainder of this article will treat phylogenetic nomenclature as if all definitions within the system were node based. Furthermore, if the phylogenetic nomenclaturalists were to adopt apomorphy-based definitions, many of the arguments for their system would seemingly evaporate. Typical n ode-based definitions, then, might read as follows: "Lepidosauria is the most recent common ancestor of Sphenodon and squamates and all of its descendants" or "Mammalia is the most recent common ancestor of montremes and therians and all of their descendants."
Typification: As I noted earlier, in the Linnaean system the application of names is controlled by the designation of nomenclatural types. In phylogenetic nomenclature taxa are circumscribed by indicating in their definition two or more taxa--specifiers--that are descended from a node; in the case of apomorphy-based definitions one of those specifiers would be a character. For this reason Nixon and Carpenter (2000) have referred to phylogenetic nomenclature as the "node-pointing system."
Hierarchical position: In an effort to provide stability of names, indications of rank are removed from names. One of the more widely distributed proposals would have all supraspecific names terminating in "-ina," simply to indicate that the name was coined under the system of phylogenetic nomenclature.
B. DEFINITIONS OF NAMES, DEFINITIONS OF TAXA, AND ESSENTIALISM?
Some authors have referred to the methods of phylogenetic nomenclature as the "phylogenetic systematic" approach (e.g., Kron, 1997) or as "phylogenetic taxonomy" (de Queiroz and Gauthier, 1990; Lee, 2001). lam not interested in arguing the exact meanings of terms; nonetheless, as Nixon and Carpenter (2000) pointed out, the term "phylogenetic systematics" has long been used for the methods articulated by Willi Hennig; appropriation of the term for an activity with a totally different purpose does little but obscure the actual nature of the arguments being made. Furthermore, some authors, such as Lee, (2001:175) in his use of the term "phylogenetic taxonomy," seem to confuse the recognition of groups with the naming of them. The term "phylogenetic systematics" is best applied to the methods used to recognize monophyletic groups on the basis of synapomorphie characters; it should not be applied to conventions developed for assigning names to those groups. A similar comment could be made with regard to the use of the term "taxonomy" when referring to methods that deal solely with the process of applying names. Such usage will at best be confusing and at worst convey an impression that is just plain false. Worse yet is the frequently conveyed connotation that natural (monophyletie) groups can be recognized only in that system (of phylogenetic nomenclature) and that the existence of such groups will be obscured, or their recognition impossible, using any other approach.
All criticisms of Linnaean nomenclature by phylogenetic nomenclaturalists devolve from arguments about definitions; more specifically, the definitions of taxon names. What is being criticized is the use of the attributes of organisms to "define" taxon names. The phylogenetic nomenclaturalists believe this process to be essentialistic--to convey an immutable essence. This view, as espoused by de Queiroz and Gauthier (1990), de Queiroz (1994, 1997), and others, was adopted from works such as those of Popper (1964, 1966), Hull (1965), and Ghiselin (1984) rather than having been invented by the advocates of phylogenetic nomenclature; furthermore, the opinions of these authors are based on the application of classical ideas in the field of philosophy to a limited sampling of relevant literature, and in the case of Hull and Popper, by individuals with no personal experience in the field of taxonomy.
The alternative to the Linnaean system, as proposed by the advocates of phylogenetic nomenclature, is an approach in which taxa (lineages) are defined on the basis of their relationships to one another, under the model indicated above for Lepidosauria and Mammalia. Such definitions are said to achieve a level of acceptability because they are not attached to names, do not treat taxa as classes, and therefore do not connote an "essence." Thus, under this system of definitions, phylogenetic nomenclature purportedly turns taxonomy into a truly evolution-based enterprise. The substance of this line of reasoning may appear as little more than obfuscation, and I for one would contend that it is, a point of view that I will now explain in some detail.
De Queiroz's s writings are littered with disclaimers indicating that he is not labeling the majority of practicing taxonomists as essentialists. For example, "Most contemporary biologists no longer hold the view that eternal, immutable essences are the underlying bases of taxa, or even that all members of a taxon have certain invariant characters by which they can be distinguished" (de Queiroz, 1994: 501). Yet he makes it abundantly clear that he considers their thinking tainted with essentialism when he notes that "[t]hey have nonetheless continued to formulate taxonomic definitions as lists of organismal traits or characters" (p. 501), that "[t]reating lists of organismal traits, i.e., descriptions of individual organisms, as the defining formulas of taxon names implies that taxa are abstract categories (kinds), the members of which belong to those categories for the reason that they share certain (defining, essential) characters. This view is very much in keeping with--and probably has been inherited dire ctly from--the metaphysics of essentialism, in which individual organisms are viewed as the concrete manifestations of an abstract, ideal organism (essence)" (p. 501), and that "elements of the Aristotelean form of definition have persisted in modem biological taxonomy in that the names of taxa continue to be treated as if they are defined by lists of organismal traits (see the section titled 'Definition' in numerous recent taxonomic papers)" (de Queiroz and Gauthier, 1990: 308).
Nixon and Carpenter (2000) have labeled de Queiroz's presentations as condescending, a point of view with which I wholeheartedly agree. But then, de Queiroz is not the first person to have labeled working taxonomists as essentialists and not the first to use the term as one of derogation and wrongheaded thinking. Ernst Mayr (see 1988 for a summary) argued over a period of many years that taxonomy must be rid of essentialism, to be substituted in his case with "population thinking." Although Mayr was attacking working taxonomists as essentialists, he was also complaining about what he viewed as the defects of Hennig's then recently introduced methods and the fact that they would undoubtedly reduce the Ayes to a status lower than "class" (in a phylogenetic classification), the rank to which birds were truly deserving in his view.
Mayr thus seems to have proceeded de Queiroz in providing arguments that produce an obvious paradox: in this case that essentialism is a scientific scourge and at the same time that the appropriate ranks for given taxa are somehow preordained. The phylogenetic nomenclaturalists have simply avoided the paradox of their argument rather than addressing it. They are certainly correct that taxa in the Linnacan system taxa have character-based definitions (diagnoses). But then, taxon recognition in the cladistic system--the one from which a phylogenetic taxonomy is supposed to spring--is also character based. The issue here, then, is not the difference between character-based definitions and lineage-based definitions, as the proponents of phylogenetic nomenclature would have us believe. It is, rather, their particular construal--or misconstrual--of the term "definition" as widely used in the taxonomic literature.
De Queiroz (1997: 127), in a relatively recent form of his ongoing manifesto promoting phylogenetic nomenclature, has argued that there are two primary kinds of taxonomic and nomenclatural systems: Linnaean and phylogenetic. He observed that taxonomic systems can be both Linnaean and phylogenetic, whereas nomenclatural systems "must be one or the other." The statement suggests that the Linnaean system of nomenclature cannot be used in a phylogenetic context, an idea that will strike most practicing taxonomists as absurd.
De Queiroz offers no simple or direct statement about why the Liannean hierarchy cannot be interpreted in a phylogenetic context. His reasons, however, appear to be based almost totally on his concept of the definition of a taxon name. He stated that, although the codes of nomenclature are elaborate, "their method of definition is implicit rather than explicit" (1997: 133). De Queiroz ventured that the nature of the definition (of the term "definition") must be inferred from the codes. His inference is that the definition of the name Asteraceae is something like "the taxon including the genus Aster that is assigned to the category Family." He further states that the Zoological Code refers to definitions of names, whereas the Botanical Code more appropriately speaks of "descriptions." According to de Queiroz, the latter term (description) refers to taxa rather than names (p. 133) and, in the Linnaean system, definitions are dependent on the Linnaean hierarchy, so categorical assignments play a critical role in the application of taxon names (p. 134). On the other hand, according to de Queiroz, in a "phylogenetic" system of nomenclature, definitions are independent of the Linnaean hierarchy, so categorical assignments play no role in the application of taxon names (p. 135).
Stuessy (2000) recently stepped forward to counter de Queiroz's reasoning concerning definitions. He argued that taxon names are not defined--that lists of diagnostic attributes do not form a definition of a name--but rather that the names serve as labels for the taxa and that the lists of attributes are circumscriptions of the taxa. In line with the view of taxa as concepts, as expressed above, I believe that Stuessy has an accurate perception of what is actually taking place in taxonomic practice. Nonetheless, Stuessy's comments were summarily dismissed by de Queiroz (2000) as having been presented without any evidence to support his conclusions. To bolster this assertion, de Queiroz simply gave a citation of his own work and requoted his definition of Asteraceae as given above. This can hardly be viewed as anything but authoritarianism, because de Queiroz has no argument himself, as can be seen in this quotation: "Perhaps the reason Stuessy thinks that taxon names are not defined is that definitions under the traditional system are taken for granted. That is to say, the bacteriological, botanical, and zoological codes do not explain how taxon names are defined.... Nevertheless, taxon names have implicit definitions of the kind described above [for the Asteraceae]" (de Queiroz, 2000: 533-534). One wonders how it is that de Queiroz knows the true and correct method of defining taxon names, if the works that codify the use of those names themselves make no explicit mention of the nature of such definitions.
I agree with Stuessy that taxa are concepts and that character-based diagnoses are the definitions for the taxa to which those names apply. De Queiroz's perspective on essentialism aside, such a view makes the direct--and I would claim necessary--connection between the taxa, the means by which they are recognized, and the names by which we refer to them. Under de Queiroz's model, one wonders how we are to make the connection between taxa and names. The definitions of names, of necessity, make no mention of attributes. The taxa must remain nameless, otherwise their names would be character based, and we would have returned to what de Queiroz believes are entities with essences.
I have regarded (Schub, 2000), and still regard, nearly all of the arguments put forth by de Queiroz and those who adhere to his line of reasoning as specious, for the following reasons. First, taxonomists have long used the terms "definition," "description," and "diagnosis" as being virtually synonymous. Certainly the term "definition" is not restricted to the meanings of names in the modern literature--as suggested by Hull and endlessly reasserted by de Queiroz--but is broadly applied in the sense of "taxon concepts"; such usage is widespread in both botany and zoology and occurs in the phylogenetic, taxonomic, and nomenclatural literature. As such, "definition," "diagnosis," and "description" all have the intended meaning of listings or summaries of the attributes of taxa, the first two terms usually referring to a more restricted list of attributes than the last.
De Queiroz falsely assumes that definition is (always) used in the sense that he suggests, but he cites no explicit examples from the taxonomic or nomenclatural literature that would allow him to arrive at such a conclusion. Indeed, he says that the meaning of definition, as applied to taxa, must be inferred. One is forced to conclude that his inference is simply wrong. Names are used as connotations of those diagnostic attributes, which in sum are frequently referred to as definitions of taxa. Such an inference (fact?) would seem obvious, for de Queiroz himself has noted that the Botanical and Zoological Codes use "definition" and "diagnosis" in an interchangeable way, as do many works in the taxonomic literature; but he prefers to view this as an apparent lapsus, or the result of confusion, concerning what the "correct" term should actually be. The real issues, despite many pages of explication to the contrary on the part of de Queiroz, are his slavish adherence to a meaning of the term "definition," which few biologists but de Queiroz himself accept, and his belief that "defining" taxa by a list of attributes connotes for those taxa some essential and immutable properties.
The definition of Asteraceae (and similarly for any other taxon)--contrary to the view of de Queiroz--would certainly be treated by the bulk of practicing taxonomists as "the set of attributes shared by Aster and other plant taxa grouped together and placed at that categorical level." Although the name Asteraceae may indicate hierarchical level, the level of placement has no fixed reality but only indicates the relative hierarchical relationship of Asteraceac to other plant taxa at more and less inclusive levels. The continued insistence by de Queiroz, and others infatuated with the idea of phylogenetic nomenclature, that placement of a taxon at some level in the Linnaean hierarchy confers some "essential" quality to that taxon is simply misguided. Asteraceae can maintain its status as monophyletic, while retaining its position in the hierarchy of relationships at a level other than family. Nothing about the assignment to family rank is fixed. Indeed, by changing only the ending of the word, the relative hier archical placement of the clade within the phylogenetic system for flowering plants could be changed without altering the diagnosis for the taxon itself, or its composition. The assignment of the "-aceae" ending has everything to do with communication of information about relative rank and nothing to do with Aristotelian definitions or ascription of essential properties.
2. Karl Popper and Phylogenetic Nomenclature
The works of Karl Popper were cited by de Queiroz (1994) as a source of inspiration for his writings. Popper's works have also been accorded approval by many authors, including myself (Schuh, 2000), concerning the conception and testing of taxonomic theories in the cladistic paradigm. The reader may wonder how Popper's writings can figure as a positive influence in the writings of de Queiroz and me, when I have little good to say about the works of de Queiroz on the subject of phylogenetic nomenclature. I will explain.
In two of his works, Popper (1964, 1966) discussed at length what he believed to be the scientific status of the social sciences. These writings necessarily diverge to some degree from others of his that deal almost strictly with philosophy (or, if you will, epistemology) as it might be applied in the natural sciences, as for example The Logic of Scientific Discovery (1968). Nonetheless, de Queiroz (1994) has found in the aforementioned works statements that he believes support his cause; namely, Popper's arguments concerning "methodological nominalism" and "methodological essentialism." By way of paraphrase, methodological nominalism has everything to do with identifying concepts or ideas, describing them, and if need be, attaching terminology to them. Methodological essentialism is much more strongly oriented toward producing unbending definitions for terms representing favored concepts, irrespective of their scientific interest; the approach is strongly associated by Popper with the Aristotelian view of sc ientific thought, a view Popper treats most harshly.
It should not be thought unusual to find such distinctions in the works of Karl Popper. His disdain for formal definitions is well known, as is his advocacy (indeed, maybe even penchant) for conveying ideas through context rather than through formal definition, Indeed, The Poverty of Historicism (1964) includes no explicit definition of the term "historicism," and the reader is near the end of the work before the historicists are actually identified and labeled as such.
De Queiroz (1994) would have us believe that (most) practicing taxonomists are essentialists because he equates the methods of taxonomists with what Popper labels as "methodological essentialism." Even though de Queiroz allows that the distinction between nominalism and essentialism is a relative one, he nonetheless pursues the line of thought that the nominalistic perspective--his--represents an important step in the development of a more broadly scientific approach to taxonomy in the form of "phylogenetic nomenclature."
Under de Queiroz's construal of methodological essentialism, taxonomists are, for example, "associated with a seemingly endless and strictly verbal debate about the meaning of the name Mammalia" (de Queiroz, 1994: 499); one would suppose that just about any taxon name could be substituted for Mammalia if essentialistic beliefs are as widely held among taxonomists as de Queiroz would have us believe. Jade Queiroz's view, it would appear that taxonomists dote on the choice of the names and then spend all of their time devising concrete (essentialistic, character-based) definitions for them, rather than concerning themselves with more productive aspects of scientific inquiry, such as discovering and describing new species, recognizing monophyletic groups, and testing existing taxonomic theories. Practicing taxonomists would certainly question whether de Queiroz has ever taken a serious and broad-based look at the descriptive and revisionary literature.
There is no mistaking that Popper, in his writings, showed little tolerance for the Aristotelean approach, by which he meant positing what is true about the world (the essences), providing names for those truths, and producing definitions for those names. Popper does not speak of whether taxonomists--or biologists more broadly--are essentialists. But in the mind of de Queiroz, the masses of practicing systematists are guilty by association, because they are known to "define" taxon names, and those definitions consist of lists of traits. De Queiroz agrees with Popper that methodological essentialism stifles scientific progress and concludes that practicing taxonomists must therefore be stifling progress in the science of systematic biology because of his prior-erroneous--assessment of their methods as essentialistic.
The arguments of de Queiroz might be found worthy if his major premise had a leg to stand on. But it does not. Practicing taxonomists know full well what they are doing, and it has nothing to do with essentialism, in the sense used by de Queiroz, or any of those from whom he has adopted his ideas. Rather, taxonomic definitions (or diagnoses, if you prefer) are hypotheses about entities in nature. Those hypotheses are subject to test and rejection or revision. It is in this area that the philosophy of taxonomic practice has been substantially clarified since the last round of essentialist attacks by Mayr, a development that de Queiroz and his like-minded colleagues are so happy to overlook in their quest to provide what they believe will be a central role for evolutionary theory in taxonomic and nomenclatural practice.
C. CONCEPTS AND TILE EVOLUTIONIZATION OF TAXONOMY
Many authors have asserted that the publication by Charles Darwin of On the Origin of Species and the subsequent rise of the evolutionary paradigm revolutionized the methods of taxonomy. In the view of de Queiroz and his fellow phylogenetic nomenclaturalists, this apparently only leaves nomenclature to be revolutionized. In my view--and here I am not alone or expressing an original interpretation--nothing could be farther from the truth. Darwin may have offered a way of viewing classifications as reflective of genealogy and argued explicitly that such should be the case, but he certainly did not propose a method for constructing classifications that contain only monophyletic groups (e.g., Padian, 1999; Schub, 2000). This is not to say that many taxa recognized between the time of Linnaeus and the advent of cladistics were not monophyletic--we need only mention groups such as Angiospermae and Ayes. Nonetheless, explicit methods for recognizing such groups--those defined on the basis of synapomorphies--became c learly formulated and widely accepted only upon publication of the works of Willi Hennig (e.g., 1966) in English in the mid-1960s. No amount of argumentation over essentialism, definitions, or other subjects cloaked in philosophical jargon can change that fact.
De Queiroz (1997, and elsewhere) is fond of the phrase "evolutionization of taxonomy." The phrase includes, of course, the wishful adoption of his much-promoted system of phylogenetic nomenclature. What is rarely mentioned in his writings is the fact that the evolutionization of taxonomy (the acceptance of Darwinian theory by most taxonomists) had little effect on taxonomic practice (see Patterson, 1987; Schuh, 2000). The justification for much of taxonomic practice has no doubt changed markedly since the time of Linnaeus and his contemporaries and particularly since the publication of On the Origin of Species in 1859. However, the recovery of hierarchical structure in nature has been the driving force within taxonomy from Linnaeus to the present. Although Hennig may not have been the first--or the only--person to appreciate the idea that grouping by "synapomorphy" was the only way to construct maximally efficient and predictive diagnoses of taxa in a hierarchical classification, his writings were certainly t he ones that catalyzed the idea among biologists worldwide. The consistent and widespread application of those methods has revolutionized certain aspects of taxonomic practice in a way not seen since the time of Linnaeus. Nixon and Carpenter (2000: 300-301) have a slightly different take on this argument, but their comments--like mine--are not favorable with respect to the views of de Queiroz, as can be deduced from the fact that their discussion of this subject is placed under the heading "revisionist history."
De Queiroz, like some others, seems to have lost sight of (or possibly never read) the argument that the concept of evolutionary relationships among organisms gains strength from an independent taxonomy (e.g., Brady, 1985). Nor do he and his co-authors give much space to discussion of the source of information for phylogenetic relationships. If the practice of taxonomy has been strengthened in the past 35 years, it is not because of "evolutionization" but, rather, from "objectification." That objectification has to do largely with the way taxa are diagnosed; i.e., by apomorphies.
De Queiroz and his followers (e.g., Bryant and Cantino, 2002) would have us believe that their approach to producing taxon definitions--"node-pointing system" (Nixon and Carpenter, 2000)--will evolutionize nomenclature, because names in this system are defined on the basis of clade (lineage) relationships rather than by lists of diagnostic characters. One has to wonder about the relevance of the continued discussion of definitions, when in reality the clades that de Queiroz would wish to recognize receive their character-based diagnoses from cladistic analyses.
1. The PhyloCode
Stability under the codes of nomenclature has to a certain extent been engineered. As I pointed out above, the idea of stability will of necessity be transitory. We could, of course, devise a system of biological nomenclature that would be perfectly stable, but in the end it would be almost certainly be useless scientifically. The authors of the PhyloCode (Cantino and de Queiroz, 2000) argue that reassignment of a taxon to a higher or lower rank can cause an undesirable cascade of name changes, a situation that would be obviated within a system of phylogenetic nomenclature. Phylogenetic nomenclature would, of course, also remove much of the ease with which we are able to communicate concerning relationships among taxa. When dealing with typified taxa, shifts in categorical ranks in the Linnaean system will be no greater, and often much less, than would be the case under the node-pointing system, as shown by Nixon and Carpenter (2000). The proponents of phylogenetic nomenclature further argue that:
When the PhyloCode is extended to species, it will improve nomenclatural stability . . . by removing the linkage [of a species name] to a genus name. A major source of instability in species names under the preexisting codes[,] ... revision of generic limits, will thereby be eliminated. There will, of course, be a consequent absence of hierarchical information in species names governed by the PhyloCode; specifically, one will not be able to infer phylogenetic relationship from these names in the way that one can infer genus assignment from species names governed by the preexisting codes. However, an underlying principle of the PhyloCode is that the primary purpose of a taxon name is to provide a means of referring unambiguously to a taxon, not to indicate its relationships. From this perspective, the loss of nomenclatural stability of species names under the preexisting codes is too high a price to pay for incorporating taxonomic information (genus assignment) into the names. Moreover, although such informati on will not be built into species names under the PhyloCode, phylogenetic position can easily be indicated by associating the species name with the names of one or more clades to which it belongs. (Cantino & de Queiroz, 2000)
This statement offers a truly tortured piece of logic. Phylogenetic nomenclature would appear by definition to be designed to convey information on phylogenetic relationships. The name suggests as much, as does virtually all of the published discussion promoting such a system. Furthermore, in its role as part of the "evolutionization of taxonomy," phylogenetic nomenclature could hardly have a more important objective. Yet we are told that absolute stability of names is more important than being able to directly communicate relationships among species through the system of nomenclature, In the end, if producing stability is sublime in the PhyloCode, indicating phylogenetic affinity is ridiculous. According to the abovequoted statement, after we have discarded the association between generic and specific epithets in an attempt to promote stability in the application of scientific names, we can easily communicate this information by reassociating those names, presumably in the form of a uninomial. Thus the Phylo code proposes the reinvention an inferior version of the system that it just discarded. Furthermore, it presupposes a fixity for species concepts and lineage identities that has little to do with reality, as the examples described early in this article were designed to illustrate.
To date, few published works have addressed the issue of how to deal with species under the system of phylogenetic nomenclature. In one such work (Cantino et al., 1999) the authors offered 13 proposals on how species names might be dealt with. In addition to the argument that stability of species names is paramount, the paper is based on what I will call the "preemptive assertion" that the system of naming species must be changed because "Linnaean binomial nomenclature is logically incompatible with the phylogenetic nomenclature of de Queiroz and Gauthier." No doubt all working taxonomists have found aspects of species-group nomenclature under the codes tedious at times. I count myself as one of those. However, I am not encouraged by the arguments of Cantino et al., and those presented in the PhyloCode, for the following reasons. [Lee (2001) has also taken the "preemptive" approach to argumentation, suggesting that the work of Nixon and Carpenter (2000) actually advocated the methods of phylogenetic nomenclat ure (or phylogenetic taxonomy as he prefers to call it). Lee's argument is based on the following erroneous syllogism, whose major premise is clearly false: Any taxon definition involving the mention of apomorphies is apomorphy-based in the sense of phylogenetic nomenclature; the taxon definition advocated by Nixon and Carpenter (2000) involves the use of apomorphies; therefore, Nixon and Carpenter advocate the methods of phylogenetic nomenclature. In actuality, nothing could be farther from the truth concerning the conclusions of Nixon and Carpenter.]
First, assertions concerning the "logical incompatibility" of Linnaean nomenclature with phylogenetic nomenclature are irrelevant; this is the "preemptive assertion." What is relevant is whether the Linnaean system can accurately associate names with taxa and whether it can represent hierarchical relationships of monophyletic groups. The answer is that it unequivocally can.
Second, absolute stability is not necessarily a desirable goal. Species, like taxa at higher levels, are concepts. To imply that fixity of species names will somehow promote fixity of taxonomy at the species level (or any other level, for that matter) is a fatuous idea. It will more likely simply impede the discovery of concepts in need of test and revision (Gaffney, 1979; Dominguez and Wheeler, 1997).
Third, adding "some" stability at the expense of losing a relatively simple method of communicating information on relationships is a price not worth paying. Most of the proposals by Cantino et al. (1999) would adopt existing binomials for the names of species in the system of phylogenetic nomenclature. As such, nothing novel is offered by the move to phylogenetic nomenclature, except the claim by Cantino et al. that the names will not have ostensible definitions and will therefore have lost their "logical incompatibility" with phylogenetic nomenclature. Those names would remain the same in perpetuity. Thus, any revisions in concepts at the supraspecific level would complicate our ability to communicate directly information about relationships of species.
Fourth, with at least 2 million available species names already in existence, one wonders just how phylogenetic nomenclature will ultimately communicate about the relationships of species and distinguish among the taxa vulgarisl-vulgaris 1000. Homonymy, synonymy, and recombination may be messy issues for the taxonomist. Yet, because the classification of large groups is seldom the work of a single individual, and because different taxonomists with equally credible credentials and methods may reach different conclusions, taxonomy will never become an absolutely stable enterprise. The idea that species names can--let alone should--remain absolutely stable is simply misguided and unscientific.
2. Names, Ranks, and the Representation of Hierarchy
Nixon and Carpenter (2000) showed that the node-pointing system is no more stable than the Linnaean system. They based their observations on some of the examples favored for the promotion of phylogenetic nomenclature. I will offer a separate set of observations, also using examples from the work of de Queiroz (1997).
De Queiroz provided an example to justify his assertion that, under the Linnaean system, names can change under different ranking schemes and that the same names can refer to different taxa. His figure demonstrating these effects is here reproduced as Figure 1. Types are represented by asterisks. The differences between Scheme One and Scheme Two indeed show what de Queiroz asserts concerning the Linnacan system: that, in fact, the (endings of the) names are different because the ranks are different. What he views as irrelevant is that the hierarchical relationships among the taxa are identical in both diagrams. Under a system in which the "definition" of a taxon name is its diagnosis, exactly the same information on inclusion/exclusion relationships is transmitted by both schemes. De Queiroz would have us believe that the different schemes cause confusion about the information being transmitted, but this could only be the case if one were to accept his concept of definition for the taxa in question.
Figure 2 offers de Queiroz's justification of phylogenetic nomenclature from the point of view that taxon names are unaffected by rank (categorical assignment). Certainly the example offers evidence for his point. But information on inclusion/exclusion relationships is lost! Stability of names is the only criterion of interest to de Queiroz, but it can hardly be the only relevant criterion in evaluating a system of biological nomenclature. Thus this example cannot be viewed as anything but weak.
De Queiroz offers additional examples, in which the theories of phylogenetic relationships are changed and, therefore, the concepts applied to at least some of the included taxa are changed. He argues that because concept is tied to rank in the Linnaean system, change in rank produces change in meaning of taxon name. De Queiroz would have us believe in this case that phylogenetic nomenclature will be less confusing than the Linnaean approach because the names remain stable even though the taxon concepts have changed. This can hardly be the case. Nixon and Carpenter (2000) and Carpenter (2003) have shown that the Linnaean system will be more stable for typified taxa when taxon concepts change.
All of these arguments are essentially irrelevant for the evaluation of stability. The only relevant example offered by de Queiroz (1997) does not make a case for phylogenetic nomenclature on the basis of stability. Neither does it show well the difference between the two systems with regard to inclusion/exclusion relationships: namely, under the de Queiroz model no information on inclusion/exclusion relationships will be found in the names themselves, whereas the Linnaean system serves this function inherently, and well. On the basis of de Queiroz's own examples, I can only conclude that phylogenetic nomenclature will certainly be no more stable than the Linnaean system and that its methods for depicting even the most basic of inclusion/exclusion relationships will be substantially more cumbersome than those of the Linnaean system.
E. A PHILOSOPHICAL PERSPECTIVE
Elsewhere in the present issue Keller et al. (2003) argue for the grounding of phylogenetic nomenclature in the philosophical view of species as individuals. Although there are aspects of their presentation with which I do not agree, we are certainly in accord on the view that phylogenetic nomenclature is an ill-founded and illogical replacement for the Linnaean system.
There are also additional published arguments for why phylogenetic nomenclature should be preferred from a philosophical point of view. One of these, by Ereshefsky (2001), was published in book form. This might, in and of itself, be though to confer some special weight. The title, The Poverty of the Linnaean Hierarchy, would seemingly add credence to such thinking. Between the covers of Ereshefsky's work one finds many ideas that mainstream biologists might find perplexing, if not downright disturbing. For the purposes of the present discussion, I will restrict my comments primarily to points dealing with the issue of biological nomenclature.
Ereshefsky is what I would call relatively cagey in most of his presentation, weaving ideas together in such a way that it is difficult at times to determine his position. Our first clues undoubtedly come from his presumptuous title. Like those promoting phylogenetic nomenclature, and like his mentors David Hull and Eliot Sober, Ereshefsky argues that nomenclature must fit within an appropriate theoretical context. Not surprisingly, this context is "evolutionism." Ereshefsky's view, stated in his own words, sounds remarkably like what I quoted above from de Queiroz.
A defender of the Linnaean system would, of course agree that the vast majority of Linnaeus's theoretical [essentialist and creationist] assumptions are obsolete. Nevertheless, she could argue that we can continue to use various aspects of the Linnaean system without adopting his outdated assumptions. . . Still, the Linnaean system does carry unwanted theoretical baggage. There may be a consensus among biologists that taxa lack essences but the assumption that the Linnaean categories have essences is widely held. Taxa in a single category are often considered comparable, and taxa in different categories are considered importantly different. (Ereshefsky, 2001: 283)
Ereshefsky does not indicate who the offending parties might be or just what they have said and done that allows him to discern their essentialist leanings. As I argued above, such claims are largely empty and represent little more than futile efforts to justify a position based on philosophical preconception.
Ereshefsky further argues for what he describes as the practical advantages of phylogenetic nomenclature, as opposed to what he views as the practical limitations of the Linnaean system. Among these is stability. In this regard Ereshefsky notes that because "positions and ranks of taxa must be incorporated into their names" we are forced to accept "unwanted instability during revision because the names of the taxa must be altered." According to Ereshefsky, the post-Linnaean system avoids such instability because "the ranks and positions of taxa are not indicated by their names, so taxon names remain constant during revision" (2001: 285). Ereshefsky seems to admit to the necessity of studies in which taxa are reviewed and revised (taxonomically). He seems to overlook the fact that such studies will frequently force us to accept revised concepts. How nomenclature can remain entirely stable in the face of revised concepts has not been satisfactorily explained by de Queiroz, and it is certainly not explained by E reshefsky. The latter's assertion that stability of nomenclature is a desirable practical objective seems to rest on the assumption that "[t]wo biologists can disagree on the placement of a taxon yet refer to the taxon with the same name" (2001: 285). The absolute stability of names seems to be the only achievement realized under this kind of thinking, an empty objective from the scientific point of view, as Gaffney (1979) argued more than 20 years ago.
Ereshefsky further states that Linnaean ranks do not reflect true divisions in nature and that "biologists should be freed from the theoretically empty task of having to assign higher Linnaean ranks to taxa." With regard to species he argues that taxa typically thought of as species are incomparable entities. As he puts it, "It is doubtful that a feature exists that distinguishes species taxa from all other types of taxa" (2001: 285). After all of the disparaging remarks about essences, essentialists, and essentialism by Ereshefsky and others, these remarks come as close to describing an essentialist perspective as any one might find, no matter what the theoretical or philosophical persuasion of the author.
Ereshefsky's arguments, in my view, are carefully tuned to conform to his preconceived views on how evolutionary biology, and systematic biology more particularly, should function. Like many others, he seems happy enough to attempt to construct a new orthodoxy. His viewpoint rejects the continuity of observations from at least the time of Linnaeus to the present, that there does seem to he a hierarchical structure in nature and that it is the effective description of that hierarchy that has preoccupied systematic biologists. I can only conclude that Ereshefsky's claims concerning the "poverty" of the Linnaean hierarchy are empty, in large part because he misunderstands the very system he is trying to reform.
De Queiroz (1997) argued that the current codes of nomenclature promote "nomenclatural clarity, universality, and stability" (p. 137) but that they do so in an inappropriate theoretical context. That inappropriate context rests on de Queiroz's ideas concerning the definition of names and his belief that the only thing that is "clear, universal, and stable [in the Linnaean system] is the association between a name and one of the Linnacan taxonomic categories" (p. 137). He argued that the clear, universal, and stable association within the system of phylogenetic nomenclature will be the "association between a taxon name and a dade or monophyletic group of species" (p. 138). As I discussed above under "Definitions," I see little evidence in support of de Queiroz's point of view. Indeed, the characterization of the phylogenetic nomenclature by Nixon and Carpenter (2000) as one of "metaphysical correctness" would seem to be entirely accurate.
The context in which biological classifications must function is a hierarchical one. In this sense the Linnaean system has served biology effectively for nearly 250 years. Although some workers may have misconstrued the use of the "family" category to mean that placement of a taxon at that level conveys some particular (essential) properties, most investigators using the system in this day and age are well aware that "family" and all other categorical ranks are used to denote a scheme of nested hierarchical relationships, not to convey information on some additional aspect of reality over and above the characters by which the taxa are diagnosed.
In claiming that the "theoretical context" of hierarchical Linnaean nomenclature is flawed, de Queiroz, Ereshefsky, and others have produced arguments that are little more than distractions, The important issue is the portrayal of hierarchical relationships of taxa. The Linnaean system has done that for nearly 250 years and should not be abandoned simply to bring nomenclature into conformity with flawed metaphysics. If changes are to be made, they should be to repair deficiencies in the existing codes of nomenclature, along the lines suggested by Nixon (2003).
[FIGURE 1 OMITTED]
[FIGURE 2 OMITTED]
Table I Changes in taxonomy of Plagiognathus Fieber, 1858 (Insecta: Hemiptera: Miridae) Total Valid Valid valid Nearctic Palearctic Generic Species New New Year species Species species synonyms synonyms species genera 1923 57 39 18 0 - 20 - 1995 101 77 24 2 - - - 2001 110 86 24 3 30 (new) 23 1 Species incertae New Year sedis combinations 1923 0 - 1995 0 - 2001 16 26
I thank Jerry Davis for inviting me to contribute to this discussion and for his belief in the unity of the taxonomic enterprise, Dennis Stevenson for his efforts in helping me locate the botanical literature and for discussing botanical nomenclature, and Andrew Brower for a one-line comment that inspired the title of this article and a portion of its content. Pamela Beresford, James Carpenter, and Jun-ichi Kojima offered helpful comments on the manuscript. The article is dedicated to the late Ron Brady, in recognition of his arguments concerning the independence of systematics as a field of enquiry.
VII. Literature Cited
Brady, R. H. 1985. On the independence of systematics. Cladistics 1: 113-126.
Bremer, K. 1985. Summary of green plant phylogeny and classification. Cladistics 1: 369-385.
Bryant, H. N. & P. D. Cantino. 2002. A review of criticisms of phylogenetic nomenclature: Is taxonomic freedom the fundamental issue? Biol. Rev. Cambridge Philos. Soc. 77: 39-55.
Cantino, P. D. & K. de Quciroz. 2000. PhyloCode: A phylogenetic code of biological nomenclature. <http://www.ohiou.edu/phylocode/>.
-----,H. N. Bryant, K. de Queiroz, M. J. Donoghuc, T. Eriksson, D. M. Hillis & M. S. V. Lee. 1999. Species names in phylogenetic nomenclature. Syst. Biol. 48: 790-807.
Carpenter, J. M. 2003. Critique of pure folly. Bot. Rev. (Lancaster) 69: 79-92.
Chase, M. W., D. E. Snit is, R. G. Olmstead, D. Morgan, D. H. Les, B. D. Mishler, M. R. Duvall, R. A. Price, H. G. Hills, Y.-L. Qiu, K. A. Kron, J. H. Rettig, E. Conti, J. D. Palmer, J. R. Manhart, K. J. Sytsma, H. J. Michaels, W. J. Kress, K. G. Karol, W. D. Clark, M. Heden, B. S. Gaut, R. K. Jansen, K.-J. Kim, C. F. Wimpee, J. F. Smith, G. R. Fumier, S. H. Strauss, Q.Y. Xiang, G. M. Plunkett, P. S. Soltis, S. M. Swensen, S. E. Williams, P. A. Gadek, C. J. Quinn, L. E. Eguiarte, E. Golenberg, G. H. Learn Jr., S. W. Graham, S. C. H. Barrett, S. Dayanandan & V.A. Albert. 1993. Phylogenetics of seed plants: An analysis of nucleotide sequences from the plastid gene rbcL. Ann. Missouri Bot. Gard. 80: 528-580.
De Queiroz, K. 1994. Replacement of an essentialistic perspective on taxonomic definitions as exemplified by the definition of "Mammalia," Syst. Biol. 43: 497-510.
-----, 1997. The Linnaean hierarchy and the evolutionization of taxonomy, with emphasis on the problem of nomenclature. Aliso 15: 125-144.
-----, 2000. The definitions of taxon names: A reply to Stuessy. Taxon 49: 533-536.
----- & J. Gauthier. 1990. Phylogeny as a central principle in taxonomy: Phylogenetic definitions of taxon names. Syst. Zool. 39: 307-322.
Dominguez, E. & Q. D. Wheeler. 1997. Taxonomic stability is ignorance. Cladistics 13: 367-372.
Edwards, J. L., M. A. Lane & E. S. Nielsen. 2000. Science 289: 2312-2314.
Ereshefsky, M. 2001. The poverty of the Linnaean hierarchy: A philosophical study of biological taxonomy. Cambridge Univ. Press, Cambridge and New York.
Gaffney, E. S. 1979. An introduction to the logic of phylogeny reconstruction. Pp. 79-111 in J. Cracraft & N. Eldredge (eds.), Phylogenetic analysis and paleontology: Proceedings of a symposium entitled "Phylogenetic models," convened at the North American Paleontological Convention II, Lawrence, Kansas, August 8, 1977. Columbia Univ. Press, New York.
Ghiselin, M. T. 1984. "Definition," "character," and other equivocal terms. Syst. Zool. 33: 104-110.
Greuter, W., J. McNeill, F. R. Barrie, H. M. Burdet, V. Demoulin, T. S. Filgueiras, D. H. Nicolson, P. C. Silva, J. E. Skog, P. Trehane, N. J. Turland & D. L. Hawksworth. 2000. Intemational Code of Botanical Nomenclature (Saint Louis code), adopted by the Sixteenth International Botanical Congress, St. Louis, Missouri, July-August 1999. Regnum Vegetabile, 138. Koeltz Scientific Books, Konigstein, Germany.
Haffer J. 1997. Foreword: Species concepts and species limits in omithology. Pp. 11-24 in J. del Hoyo, A. Elliott & J. Sargatal (eds.), Handbook of the birds of the world. Vol. 4, Sandgrouse to Cuckoos. Lynx Edicions, Barcelona, Spain.
Hennig, W. 1966. Phylogenetic systematies. D. D. Davis & R. Zangeri, trans. Univ. of Illinois Press, Urbana.
Hull, D. L. 1965. The effect of essentialism on taxonomy: 2000 years of stasis (I). Brit. J. Philos. Sei. 15: 314-326.
ICZN [International Commission on Zoological Nomenclature]. 1999. International Code of Zoological Nomenclature. Ed. 4. International Trust for Zoological Nomenclature, London.
Keller, R A., R. N. Boyd, & Q. D. Wheeler. 2003. The illogical basis of phylogenetic nomenclature. Bot. Rev. (Lancaster) 69: 93-110.
Kojima, J-I. 2003. Apomorphy-based definition also pinpoints a node, and PhyloCode names prevent effective communication. Bot. Rev. (Lancaster) 69: 44-58.
Kron, K. A. 1997. Exploring alternative systems of classification. Aliso 15: 105-112.
Lee, M. S. Y. 2001. On recent arguments for phylogenetic nomenclature. Taxon 50: 175-180.
Mayr, E. 1988. Toward a new philosophy of biology: Observations of an evolutionist. Harvard Univ. Press, Cambridge.
Moore, G. 1998. A comparison of traditional and phylogenetic nomenclature. Taxon 47: 561-579.
Nixon, K. C. 2003. The PhyloCode is fatally flawed, and the "Linnaean" system can easily be fixed. Hot. Rev. (Lancaster) 69: 111-120.
----- & J. M. Carpenter. 2000. On the other "Phylogenetic Systematics." Cladistics 16: 298-318.
Padian, K. 1999. Charles Darwin's views of classification in theory and practice. Syst. Biol. 48: 352-364.
Patterson, C. 1987. Introduction. Pp. 1-22 in C. Patterson (ed.), Molecules and morphology in evolution: Conflict or compromise? Cambridge Univ. Press, Cambridge.
Popper, K. R. 1964. The poverty of historicism. Harper & Row, New York.
-----. 1966. The open society and its enemies. Ed. 5, rev. 2 vols. Princeton Univ. Press, Princeton, NJ.
-----. 1968. The logic of scientific discovery. Harper Torchbook ed. 2. Harper & Row, New York.
Schuh, R. T. 1995. Plant bugs of the world (Insects: Heteroptera: Miridae): Systematic catalog, distributions, host list, and bibliography. New York Entomol. Soc., New York.
-----. 2000. Biological systematics: Principles and applications. Cornell Univ. Press, Ithaca, NY.
-----. 2001. Revision of New World Plagiognathus Fieber, with comments on the Palearetic fauna and the description of a new genus (Heteroptera: Miridae: Phylinac). Bulletin 266. Amer. Mus. Nat. Hist., New York.
Sereno, P. C. 1999. Definitions in phylogenetic taxonomy: Critique and rationale. Syst. Biol. 48: 329-351.
Stuessy, T. F. 2000. Taxon names are not defined. Taxon 49: 231-233.
|Printer friendly Cite/link Email Feedback|
|Author:||Schuh, Randall T.|
|Publication:||The Botanical Review|
|Date:||Jan 1, 2003|
|Previous Article:||Apomorphy-based definition also pinpoints a node, and PhyloCode names prevent effective communication.|
|Next Article:||Critique of pure folly.|