principles of systematics and nomenclature general system and phylogeny of insects systematics of Ephemeroptera
last update 28.XII.2010





The Naming of Cats is a difficult matter,

It isn't just one of your holiday games;

You may think at first I'm as mad as a hatter

When I tell you, a cat must have three different names.

                            - T. S. ELIOT

see also: Dual Nomenclature System      

The division I.3. from the book by N. Kluge  


 translated by Vaso Marfus and the author.


Besides known data, this division contains new rules which allow to bring in order usage of all published Latin zoological names. The new system of nomenclatures suggested here, includes the International Code of Zoological Nomenclature as a part. As an example of usage of non-ranking nomenclature for taxa traditionally attributed to family- and genus-groups, see the "Revision of supra-species taxa of Ephemeroptera" in Phylogeny and Systematics of Ephemeroptera. See also list of publications on the principles of post-Linnaean nomenclature.




I.3 Principles of nomenclature of taxa
* Why do we need rules of nomenclature?
* Independence of taxonomy from nomenclature
I.3.1 General principles of nomenclatures of biological taxa
I.3.1.1 Availability and validity of names
I.3.1.2 Principle of priority
I.3.2 Different types of nomenclatures and names
* Why do different types of nomenclature coexist?
* Rank-based (ranking) nomenclatures
* Circumscription-based (circumscriptional) nomenclatures
* Description-based nomenclatures
* Phylogeny-based (phylogenetic) nomenclatures
* Hierarchy-based (hierarchical) nomenclatures
I.3.3 Rank-based (ranking) nomenclatures
I.3.3.1 Names regulated by the International Code of Zoological Nomencature (ICZN)
I.3.3.2 Rank-based names of higher taxa
I.3.4 Hierarchy-based (hierarchical) nomenclature
I.3.5 Circumscription-based (circumscriptional) nomenclature
I.3.5.1 Circumscription-related terms
I.3.5.2 Criteria of availability for circumscription-based names
I.3.5.3 Circumscription fit
I.3.5.4 Validity of circumscription-based names
I.3.6 Name-related misconceptions
I.3.6.1 Customary interpretation of non-typified names
I.3.6.2 Polyphyletic Hexapoda: How a myth was born
I.3.7 Combining circumscription-based and hierarchy-based nomenclature
I.3.7.1 Usage of different nomenclatures
I.3.7.2 Format of species name in non-rank-based nomenclature
I.3.7.3 Sliding binomina and polynomina
I.3.7.4 The layout of a taxonomic paper
I.3.8 Catalogues of zoological names


Why do we need rules of nomenclature? Scientists have to deal with millions of animal species and supraspecific taxa. Such abundance makes nomenclatural issues a true headache. No human language has a vocabulary to name all known taxa, and no one can master such number of names. Other fields of knowledge also deal with large numbers of objects far exceeding the capability of everyday language, such as celestial and geographic objects, chemicals, etc. However, in geography, we can unambiguously pinpoint any object - no matter how many times it has been renamed (as is often the case) - by referring to the object's coordinates. In chemistry we can refer to a substance by writing down its unique structural formula.

Unlike many other objects, living species cannot be given such ultimate definitions. Any pair of species differ in indefinitely large number of characters that can be described in any number of ways, so comparing descriptions we can never be sure whether the species they refer to are identical or not. Each species has infinitely large number of characters that can be used as identifiers; no description or illustration could reflect them all. (A picture can only show a single or a few specimens, while what we need is understanding the features of a species as a whole, so pictures without description may often be misleading.) Descriptions mention only each species' differences from already known ones, so when a previously unknown species is discovered those older descriptions turn to be insufficient since they do not provide information to match with the characters of such newly discovered species. Thus the need arises from time to time to refine and elaborate the descriptions of all species known up to date in order to accommodate additional forms.

For example, in XVIII-XIX centuries lepidopterists used to believe it was enough to describe in detail the wing patterns accompanying the description by an accurate colour illustration. Later it was found that some butterfly species may have virtually identical complex wings patterns being easily distinguishable in genital structure (whose examination requires cooking butterfly abdomens in alkaline solution); other butterfly species can be reliably identified only by chromosome number and/or shape. The old descriptions lack those essential details, so updated descriptions of species known for a long time had to be drafted. In the future we will likely face the need to supplement such descriptions with even more morphological, anatomical, biochemical or molecular-level details.

Updating a species' description we can never be sure that the author of the older one had not confused several species or described more than one species under the same name. And since neither original description nor amended ones can be entirely reliable, we need strict and coherent rules of naming organisms to help us untangling intricate cases.

The easiest way to come to an agreement on which names for which taxa should be used, is to have an official list of correct names published by a competent international body. However, such a solution is only possible if the set of taxa we need to name would remain unchanged. Yet every year dozens of thousands of previously unknown species are being described, dozens of thousands of already described species get redefined, thousands of new genera and subgenera and many new families are established, and new classifications suggested with new meanings assigned to old specific, generic, family and other names. No panel could issue competent decisions on the usage of every single name under such overwhelming pressure.

Even if all animals would ever be described (at least formally) - which is hardly foreseeable - the number of zoological taxa would not stabilize. The classification of animals is a work in progress (see I.2.3.2) and inherently transient, because we are to come to a natural, phylogeny-based classification - the only one that makes sense - while we can only move closer reconstructing animal phylogeny bit by bit as knowledge from all areas of biology keeps flowing.

Thus, to regulate the usage of taxonomic names we need universal rules of nomenclature in place, which would give any worker answers about what to do with this or that name without submitting every case to a commission.

The most elaborate rules include the International Code of Zoological Nomenclature (ICZN), International Code of Botanical Nomenclature, International Code of Nomenclature of Bacteria, and some other codes which govern ranking names of species and other biological taxa (about the concept of ranking name see I.3.2 and I.3.3 below). These codes are enacted by competent international bodies and are binding for all biologists. The codes (ICZN being the one relevant to zoologists) provide universal rules based on which anybody can decide how any particular name should be dealt with. Only in rare cases where these rules fail to help with a name-related issue, the Commission makes its special ruling. So regulating names by official lists also occurs, but this involves just a small fraction of names.

Some categories of names of taxa (see I.3.2 and I.3.5) are not regulated by codes in effect. Their usage is subject to established custom, and that practice is getting increasingly outdated. In this book we formulate universal rules of using such names. Unlike the provisions of ICZN (see I.3.3.1) and other international codes, the rules suggested here for rank-based names of higher taxa (I.3.3.2), hierarchy-based names (I.3.4) and circumscription-based names (I.3.5) are not official and thus not mandatory. Some principles apply to all rules herein discussed, including concepts of availability, validity and priority (I.3.1).

Independence of taxonomy from nomenclature. It is crucial not to confuse issues of taxonomy with those of nomenclature, just as objects are distinct from their names. Taxonomy is a science and hence can not be subject to outside dictate: researcher will decide which principles of taxonomy should be applied and what kind of classification created; he may choose to base his classification on phylogeny or something else, he may adopt or reject any phylogenetic hypotheses and follow any way of designing a classification - cladistic, numerical, gradistic, or any other. Since Renaissance times the science runs on the principle that no authority and no elected body can impose opinions on a scientist, the only thing capable to change his mind being logical argument. Unlike taxonomy, nomenclature is not a science, as the names are artificially produced. So a nomenclature can be subject to rules adopted by a competent body and binding for every worker in the field; such rules are indeed provided by applicable international codes. Each of these codes proclaims the non-interference of nomenclatural rules to taxonomy. As the preamble of the International Code of Zoological Nomenclature says, "The objects of the Code are to promote stability and universality in the scientific names of animals and to ensure that the name of each taxon is unique and distinct. All its provisions and recommendations are subservient to those ends and none restricts the freedom of taxonomic thought or actions".


In the paragraphs I.3.3-I.3.5 below various principles of nomenclature of taxa will be discussed - both officially set forth in the last version of the International Code of Zoological Nomenclature, and unofficial, offered to a reader's discretion. But first we need to take a look at the concepts shared by all types of nomenclature - these are concepts of availability and validity, and the principle of priority (I.3.1.2). 

I.3.1.1. Availability and validity of names

The taxonomic name which should be used is called valid (accepted, correct). The valid name is chosen, subject to the rules of nomenclature, among the names considered to be available for purposes of the same code. Thus, an available name can be valid or not, while an unavailable name is never valid. Availability criteria vary depending on nomenclature (see below), yet the following points are shared by all codes.

Availability of a name is established by a set of availability criteria. These include: (1) the name must be published after a certain date designated as starting date for this nomenclature (January 1st, 1758 for the names regulated by ICZN and other zoological names; botany and bacteriology use other dates); (2) the name must have certain format (e.g. Latin or latinized in taxonomy); and (3) the text of the name's first publication must meet some requirements (they vary depending on code and name category). Only names meeting all availability criteria are considered available for purposes of a given nomenclature, so within a nomenclature many different available names can be created for a taxon, while available names for different taxa may be identical.

Different available names given to the same taxon are called synonyms (synonyma). The concept of synonymy applies only to names subject to the same code, because the names unavailable under that code are not considered synonyms. Objective synonyms (synonyma objectiva) are different names deliberately given to the same taxon; they may appear, for example, when a new name is published to replace the existing one (about objective ranking synonyms - see also I. Subjective synonyms (synonyma subjectiva) are different names of such taxa which are differently defined but are identical from a worker's point of view. The concept of subjective synonymy is fundamentally different in rank-based and circumscription-based nomenclatures (see I. and I.3.5.3), while in hierarchy-based nomenclatures it simply doesn't exist.

Identical available names given to different taxa are called homonyms. Junior homonyms are also referred to as preoccupied names. Just like synonymy, homonymy may only exist within the scope of the same code, because names unavailable under this code are not considered homonyms. The term hemihomonyms was coined (Starobogatov, 1991) for identical names assigned to different taxa subject to different codes (i.e. where only one of these names is available under each code).

To assign every taxon one single name and make each name refer to a single taxon (i.e. to get rid of synonyms and homonyms), a single valid name is to be chosen among all available names of each taxon and among all identical names suggested for different taxa. Such choice is made based on the same code whose rules are used to establish name availability. So, within the scope of one nomenclature we come to situation where every taxon has a single valid name and each valid name refers to a single taxon. At the same time one taxon may have several different names considered valid under different codes, while different taxa can have identical valid names under different codes (i.e. hemihomonyms may coexist).

I.3.1.2. Principle of priority

Different codes provide different ways to choose a single valid name among available names, but some sort of priority principle always applies. According to this principle, the valid name is the one which was the first, among all competing available names, to meet the availability criteria - i.e. the one published in compliance with all applicable requirements. Priority must be subject to some restrictions: it applies only to names published after a certain date (01.01.1758 in ICZN); all names published prior to this date are considered unavailable (see above). Without such restriction no name could be deemed valid with any certainty, because the time before its first publication would be indefinitely long and it would be impossible to make sure that there is no earlier published name (which would be of higher priority).

The principle of priority implies that every name must have its authorship, since priority does not apply to anonymously published names. Authorship means not just the author's name, but his particular paper with the exact date of publication.

Priority principle is designed to prevent arbitrary renaming of taxa. Anyone can make up a new name for any known taxon and publish such name meeting all availability criteria, but priority principle precludes such names from becoming valid, since the only name of this taxon that may be valid is its oldest available name. Thus, priority principle discourages redundant name creation and allows zoologists to reach consensus on which one of the existent names should be used. Priority supersedes any other criteria ruling out any controversy by indicating which name among several available ones should be picked up as valid: if two names had been published with at least one day in between, there can be no doubt which one is older; among names published in the same paper, the one which appears first takes priority.

Here lurks an important shortcoming of priority: the valid name (i.e. the name which should be used) is the earliest one, which is usually also the most problematic in terms of which species does it refer to. When an author describes a new species, he only points out what distinguishes this species from those previously described. Unaware of the forms yet to be discovered, he is in no position to indicate all characters necessary to distinguish it from the forms not described by then. So, generally, the later a species is described, the more comprehensive its original description is (setting aside occasional substandard and inaccurate descriptions). To redefine previously described species the expert must reexamine specimens on which the species's original description was based; the earlier it was described the higher are the odds that these can be lost or damaged.

In cases where rigorous enforcement of priority principle may result in much inconvenience or is unfeasible, special ruling is to be made to override priority. For names subject to the International Code of Zoological Nomenclature (see below), such rulings are issued by the International Commission on Zoological Nomenclature.

Codes vary in how priority principle is applied and in defining sets of names among which the oldest one should be chosen. In ranking nomenclatures (see below) the oldest name is chosen only among names assigned to taxa of the same rank or rank group; for example, the name of a plant genus should be the oldest among generic names, while the name of an animal genus is selected from a combined pool of generic and subgeneric names. So the result of name choice using priority principle will depend upon applicable codes.

An alternative to the priority principle is the principle of the first reviser, i.e. the first author who believes some names to be synonymous has the right to choose valid names among available ones. On limited use of this principle in ICZN see I. below.


Why do different types of nomenclature coexist? As we discussed, classification cannot be permanent (I.2.3.2); instead, it is subject to incessant change, because it relies on phylogeny, and there is no direct way to have phylogeny established; as all methods of reconstructing phylogeny are indirect and rely on the entire body of biological knowledge, and the latter is continuously growing, the process of adjusting our idea of phylogeny, and hence changing classification, will be endless as well. So there is no hope that a perfect and final classification of living organisms would ever be built. Should a constant classification appear, rules of nomenclature would become redundant, as the names of all taxa in such classification will only need to be validated once and for all. It is the inability to create such classification that forces us to set universal rules of naming taxa.

All principles of nomenclature pursue the only aim: to have names firmly stick to taxa. Yet any taxon has many different attributes, which include its circumscription, diagnosis, rank, position in the classification, etc. It is impossible to make a name refer to all such attributes at once, because any change of the classification entails changes in these attributes. For example, in different classifications taxa of the same circumscription may have different ranks, different diagnoses or be assigned to different higher taxa; and vice versa, taxa of the same rank can have different circumscriptions, and so on. Nomenclature must support ever-changing classification, which implies that a name can only be associated with just one attribute of a taxon.

Based on the attribute with which a name is associated, several fundamentally different types of nomenclatures can be recognized, viz. rank-based, circumscription-based, description-based, phylogeny-based, hierarchy-based, etc. In following paragraphs we will explain why only rank-based, hierarchy-based, and circumscription-based ones are meaningful.

Rank-based (ranking) nomenclatures. In these nomenclatures a name is associated with a certain rank of taxon and is subject to change whenever the rank changes, but remains the same when other attributes (such as circumscription or position) change (Fig. 8). The ranking nomenclature still plays a major role in taxonomy, because all international codes, including ICZN, are based on this principle. However, the ranking nomenclature has significant shortcomings (I.3.6) and is not used in this book. For details about ranking nomenclatures see I.3.3 below.

Circumscription-based (circumscriptional) nomenclatures. Under this approach a name is associated with a certain circumscription of a taxon without regard of its rank or position (Fig. 8). The terms "volume" (Kluge, 1999), "circumscription" ar "membership", as applied to supraspecific taxa, all mean the particular set of members included, not the number of species. Circumscriptional naming is traditionally applied to higher taxa, but coherent rules of circumscriptional nomenclature were not formulated until recently. (For details see I.3.5 and I.3.7 below.)

Description-based nomenclatures. In this type of nomenclature a name is associated with the definition (diagnosis) of a taxon and changes accordingly whenever the taxon is redefined. Some authors tried to use this principle.


For example, different authors referred to the same taxon (springtails) as Collembola (derived from collophor or embolium, the sticky ventral tube), Podura (after the abdominal fork), Oligomerentoma (reflecting reduced number of abdominal segments), Protomorpha (from protomorphosis - a peculiar mode of embryogenesis), and so on (see V-


Following this path we can change names of taxa over and over, but none will fully reflect the diagnosis. That's why descriptive nomenclature is never stable and there is no point using it.

C. Linnaeus (1751) has developed strict rules of descriptive nomenclature for a specific name (nomen specificum). To form the full name of a species, such specific name was attached to a generic name (which was not descriptive) and could include from 1 to 12 words providing the diagnosis to distinguish this species from any other within the genus; as the classification changed, specific names were modified accordingly. Such specific names are no longer in use; in modern nomenclature the second - specific - part of each binomen consists of a word Linnaeus called "trivial name" (nomen triviale), which is not descriptive, as it may not refer to any diagnostic features whatsoever.

Phylogeny-based (phylogenetic) nomenclatures. Recently some authors suggested a phylogenetic (cladistic) nomenclature, where the name is associated with a common ancestor of the taxon, which means a name may only refer to a holophyletic taxon evolved from such ancestral species. This nomenclature is of no practical use, since no criteria of name availability have been formulated (about the importance of availability criteria see I.3.1.1), and it is hard to imagine what such criteria might be. A phylogenetic name may become available only if its publication refers to a common ancestor of the taxon. The ancestor, however, is never known, and only a limited number of its characters may be conjectured. Descriptions usually list general characters of a taxon and give a direct or indirect indication that this taxon is "natural". If we interpret "natural" as holophyletic, we have to accept that the common characters listed in the description are those of the common ancestor, i.e. that the name given to this taxon is available as a phylogenetic one; but should we interpret "natural" as, say, paraphyletic, we may not accept this name as an available phylogenetic one. Moreover, the format of phylogenetic names remains undefined. Phylogenetic nomenclature as suggested by some authors (not cited here because of their poor knowledge of nomenclatural principles) violates the ICZN: phylogenetic nomenclature takes family names generated according to the ICZN rules of ranking nomenclature (see I.3.3.1 below) are use them under entirely different set of rules. Using the same name in different - indeed, incompatible - nomenclatural systems is inadmissible, because this only leads to chaos.

Hierarchy-based (hierarchical) nomenclatures. In this book we use circumscription-based nomenclature and a hierarchy-based nomenclature developed by the author. In hierarchy-based nomenclature a name is associated with the taxon's placement within hierarchical classification and does not depend on rank (Fig. 8). This system is based on recently enacted International Code of Zoological Nomenclature, yet overcomes some important flaws of the ICZN's ranking principle. (For details see I.3.4 and I.3.7 below.)


Fig. 8. Principal difference between circumscription-based, rank-based and hierarchy-based nomenclatures

Hierarchy of taxa uniting 12 species:















Names of taxa in various nomenclatures:




Blue-shaded taxon

Green-shaded taxon

Yellow-shaded taxon

Classification 1

phylum C-us

classis B-us

ordo A-us

Classification 2

classis C-us

ordo B-us

familia A-us

Classification 3

ordo C-us

familia B-us

genus A-us

Classification 4

familia C-us

genus B-us


Classification 5

genus C-us






Blue-shaded taxon

Green-shaded taxon

Red-shaded taxon

Classification 1

phylum E-us

classis D-us

ordo C-us

Classification 2

classis D-us

ordo C-us

familia B-us

Classification 3

ordo C-us

familia B-us

genus A-us

Classification 4

familia B-us

genus A-us


Classification 5

genus A-us






nomenclature (here -us/1, -us/2, -us/3 and -us/4 are endings reflecting rank, such as -idae and others)

Blue-shaded taxon

Green-shaded taxon

Red-shaded taxon

Classification 1

phylum A-us/4

classis A-us/3

ordo A-us/2

Classification 2

classis A-us/3

ordo A-us/2

familia A-us/1

Classification 3

ordo A-us/2

familia A-us/1

genus A-us

Classification 4

familia A-us/1

genus A-us


Classification 5

genus A-us








Blue-shaded taxon

Green-shaded taxon

Red-shaded taxon






Principal differences between circumscription-based and rank-based, typified and non-typified nomenclatures.

A circumscription-based nomenclature (non-typified);   

B rank-based non-typified nomenclature;


C rank-based typified nomenclature;


D hierarchical nomanclature (typified).


Species are shown as black rings, supra-species taxa as ellipses. On the figures AC names of each taxon are given for five various variants of arrangement of ranks, numerated by numbers from 1 to 5. Instead of Latin names of taxa, everywhere arbitrary A-us, B-us, etc. are written. One of the ranks, genus, and one of the names, A-us, are everywhere marked by bold.




Name formats. Depending on usage mode, names may be rank-based, circumscription-based, hierarchy-based, etc., while by format they can be typified or non-typified, standardized or non-standardized.

Typified names are derived from the name of type genus (on type genus see I.3.3.1). Many rank-based names of suprageneric taxa and all hierarchy-based names are typified. Non-typified are rank-based names of species-group and genus-group, and supraspecific circumscription-based names. Occasional use of non-typified names as rank-based names should be discouraged (see I.3.6.1).

Standardized names are such rank-based names whose endings are identical in all taxa of the same rank but differ in names of different ranks.

Table 6 summarizes name format types.


Table 6. Examples of typifed, non-typified, standardized and non-standardized names; in each cell names of two superorders are given - dragonflies (with oldest genus Libellula) and stoneflies (with oldest genus Perla)
















Rank-based nomenclatures are those where a name is strictly associated with rank but remains unaffected by the circumscription or position of the taxon (Fig. 8). Significant shortcoming of ranking is that names are associated with a purely conventional taxon's attribute, i.e. its rank (see I.2.4.2 above). In different classifications the same ranking name can be assigned to taxa of different circumscription while taxa consisting of the same members (i.e. having identical circumscriptions) should be given different names within the same ranking nomenclature if such taxa have different ranks. As a result, ranking may cause confusion (see I.3.6 below).

An advantage of ranking principle of nomenclature is that it allows to set strict rules and thus assure a unique valid name for every taxon in any given classification. Until recently all such rules were invariably ranking-based, so in spite of its obvious shortcomings this approach is both generally recognized and widely used. (See, however, I.3.4 and I.3.5 below to learn about newly developed coherent rules based on non-ranking principles).

Ranking principle is used in all currently effective nomenclatural codes (ICZN, the International Code of Botanical Nomenclature, and the International Code of Nomenclature of Bacteria). ICZN governs names of taxa ranging from species to superfamily. Besides this, ranking are some names of higher zoological taxa (i.e. taxa with ranks higher than superfamily) not regulated by ICZN (see I.3.3.2 below).

I.3.3.1. Names regulated by the International Code of Zoological Nomenclature (ICZN)

I. Historical. First attempts to establish rules of nomenclature date back to the late 1700s. In 1842 Strickland's "Rules for Zoological Nomenclature" were published, in 1877 - the Dall's code, in 1905 - "Regles Internationales de la Nomenclature Zoologique". The first edition of the ICZN was published in 1961, the second (somewhat revised) in 1964, the third (amended) in 1985; currently the forth (amended) edition is in force, published in English and French in 1999.

The current Code contains 90 articles governing the status of names and a number of recommendations for authors and editors of scientific publications. Recommendations aim to promote stability of zoological nomenclature.


I. Scope of ICZN and criteria of name availability. ICZN regulates usage of names applied to zoological taxa of species-group (species and subspecies), genus-group (genera and subgenera), and family-group (superfamilies, families, subfamilies, tribes, etc.), but does not apply to taxa lower than subspecies or higher than superfamily. Zoological taxa are taxa consisting of organisms currently considered to be animals (since nomenclatural rules may not affect scientific freedom, they say nothing about which organisms should be considered animals.

Availability criteria somewhat vary depending on group (species-group, genus-group, and family-group) and in different cases (the concept of availability is discussed in more details in I.3.1.1 above); here we will only discuss some of these criteria.

The starting point of zoological nomenclature is January 1, 1758, considered to be publication date of the 10th edition of "Systema Naturae" (Linnaeus, 1758). The reason why the tenth edition has been chosen is that in that edition Linnaeus for the first time used binary nomenclature in a consistent fashion. All names published before 1758 are automatically considered to be unavailable (including names used by Linnaeus himself in the previous editions of "Systema Naturae"). Accordingly, all names used in 10th edition of "Systema Naturae" are considered published for the first time in 1758, and Linnaeus is regarded as their author (though many of these names had been in use well before Linnaeus).

To become available, the name has to be Latin or latinized, i.e. written in Latin letters only and treated in accordance with the rules of Latin grammar. Words of any derivation (Latin, Greek, or other languages) or any combinations of letters may be used as zoological names.

A name becomes available only after its publication in scientific literature; unpublished names are unavailable. So the author publishing a scientific name has to be fully aware whether this name has already been used or is a new one. For a name already used, one should make sure that it is available (in such case publishing this name will not affect the nomenclature in any way). In case the intent is to publish a new name, the author must analyze everything in the Code applicable to that case and meet all availability criteria to make sure the suggested name will become available. If by some reason it is impossible to publish a name of a new taxon in compliance with all the provisions of the Code, it is better to use some expressly unavailable designation (such as number) instead of Latin name, to show that no claim of availability is being made. In any case, suggested names should be made either unambiguously available or expressly unavailable in terms of ICZN, because names of dubious availability destroy nomenclatural stability.

Species-group name, i.e. the one of a species or subspecies (sometimes, mainly in botany, called epithet) becomes available only if originally published as a part of a binomen, i.e. in combination with a generic name (original combination). As systematic position of the species changes, the same epithet may appear in combination with another generic name(s), but neither its availability nor authorship and date get affected. We may change generic name in a binomen, but may not use the epithet without a generic name. This restriction is due to the fact that many epithets are repeatedly used in zoological nomenclature, and homonymy rules in species-group (see I. apply only to epithets associated with a generic name.


I. Name-bearing types. The most important principle of ranking nomenclature is the principle of fixation of type-taxa, or name-bearing types. For each taxonomic name a type is designated (type taxon); type taxon is one of subordinate taxa within the taxon to which the name applies.

For species-group (species or subspecies) the type is a specimen - one of specimens of this species preserved in collection. The type specimen is chosen arbitrarily, but once such specimen is designated, it cannot be changed. That's why type specimens are being marked with special labels and preserved with extreme care; even if damaged, destroyed, or lost, a type can not be replaced with another specimen. A single specimen designated as a type in the original description is called holotype (holotypus); if the type series includes other specimens, they are referred to as paratypes (paratypi). If in the original description compiled based on several specimens no holotype has been designated, all the specimens are called syntypes (syntypi); one of them can be designated (by the same or another author) as a lectotype (lectotypus) afterwards, in which case all the rest of syntypes become paralectotypes (paralectotypi). If the type series is lost, a neotype (neotypus) may be designated to solve nomenclatural issues that require a type specimen. The holotype, lectotype, or neotype is the actual type of a specific name.

The type of a genus-group name (generic or subgeneric) is a type species (in older papers perhaps referred to as "genotype", the term no longer used in that sense). The type species of a generic name should be designated in the original description of the genus, or - for older names, where this had not be done originally - in a later publication (in which case the type species designated the first is considered to be valid).

Each name of the family-group (superfamily, family, subfamily, tribe, subtribe, etc.) has its type genus and is derived from the latter's name by change of ending.

The types are often referred to as types of species, genera, families, etc., but in fact are types of the names, not the taxa. A name may have no more than one type, but a taxon includes as many types as many available names it has.

At the same time, the type itself is always either a specimen or a taxon (species or genus), i.e. a material object, not a name. Thus, if a species gets its name changed, this species (with a newly assigned name) remains the type of the same generic name as before.

The types are necessary to prevent transferring names from one taxon to another. As a taxon's circumscription changes, the former name applies to the taxon which includes the type. For example, when a group once regarded as a single genus is divided into smaller genera, the genus which includes the type species of the former generic name keeps that name.


For example, Linnaeus in his classification included all mayflies in a single genus and gave it the generic name Ephemera Linnaeus 1758 (the date ascribed based on ICZN rules of priority, although the name Ephemera had been in use for centuries). As Linnaeus did not designate a type species, Latreille (1810) designated Ephemera vulgata Linnaeus 1758 as the type species for this generic name. So ever since, as classification changed, the name Ephemera could only be applied to such genus including the species E. vulgata, while no genus which didn't include this species could be named Ephemera. In different periods various workers applied the name Ephemera to mayfly taxa of different content, but in every classification Ephemera was the only taxon which both had the rank of genus and included the species E. vulgata.


In the same way type specimens are used to deal with species-group names. If, for example, it is discovered that the same specific name is being applied to more than one species, the original name should be applied to animals conspecific with the type specimen. Sometimes an original description establishing a species's name mentions only characters shared by several species; in this case the type specimen should be reexamined to decide which of these species will keep the name.

We need types in nomenclature to decide which taxon should keep this or that name, rather than which species does this or that specimen belong to, or which higher taxon should include this or that species, i.e. to resolve issues of nomenclature, not taxonomy. So, designating name-bearing types (now a mandatory procedure) has nothing to do with typological approach in taxonomy (hardly used today by anybody, since no arbitrarily selected type can provide a basis for natural classification).


I. Principle of priority and principle of coordination. There are three groups of names subject to ICZN: (1) species-group names (names of species and subspecies); (2) genus-group names (names of genera and subgenera); and (3) family-group names (names of all suprageneric taxa up to superfamily, i.e. superfamilies, families, subfamilies, tribes, subtribes, etc.). ICZN rules for each group are different, but every group is subject to coordination. The principle of coordination works as follows:

Within the species-group (i.e. among species and subspecies names) a name available as a species name is automatically available, with its authorship and date, as a subspecies name, and vice versa (any available subspecies name is available, with its authorship and date, as a species name).

Within the genus-group (i.e. among the names of genera and subgenera) coordination works in the same way: a generic name is automatically available as a subgeneric one, and vice versa; in both cases names keep their original authorship/date. Thus, the oldest (valid) name of a genus or subgenus is chosen among both generic and subgeneric names, not only among names of the same rank. In other words, availability, authorship, and date of publication are effective throughout the genus-group, not just among the names of the same rank.


For example, the name Heptagenia Walsh 1863 was originally established for a genus; subsequently the genus Heptagenia (by that time its circumscription has changed, as it happens quite often in ranking nomenclature) was divided into subgenera. One of these subgenera included, of course, H. flavescens, the type species of the generic name Heptagenia. The correct name of a subgenus should be such available genus-group name (whether generic or subgeneric) which is the oldest among names whose type species are members of such subgenus. In this example such name is Heptagenia - the oldest one among all names whose type species are members of the entire genus, not just the subgenus. Thus, the subgenus containing H. flavescens should be named subgenus Heptagenia Walsh 1863 in spite of the fact that Walsh established no subgenera in 1863.


So the principle of priority, as applied to the genus-group, works in such a way that whenever a genus gets divided into subgenera one of the subgenera is always assigned the same name as the genus where it belongs; such subgenus is called nominal.

Coordination within the family-group, under current ICZN rules, works without regard to the genus-group names. Family-group names are typified (see I.3.2) and each is derived from a genus-group name by replacing its ending with a standardized rank suffix/ending (-oidea for superfamily, -idae for family, -inae for subfamily, -ini for tribe, -ina for subtribe, etc.). So, unlike genus-group names (i.e. generic and subgeneric), the names of family-group taxa feature distinct endings for each rank. Coordination within the family-group means that authorship/date of any name in the family-group (superfamily, family, tribe, subtribe or any other) automatically applies to all family-group names derived from the same generic name. Hence, the oldest (valid) name for a family-group taxon should be chosen not only among the names of taxa having the same rank, but among all family-group names whose type genera are members of that taxon. Changing rank endings does not affect the authorship/date. In other words, a valid name of a family, subfamily, or another family-group taxon is derived from such generic name from which the first family-group name had been derived. Coordination within the genus-group has no effect on coordination within the family-group; this means the genus-group name from which the family-group name is derived is not necessarily the oldest one.


For example, genera Heptagenia Walsh 1863, Ecdyonurus Eaton 1868, and others are considered members of the same family. To obtain the name for this family, we must chose one of these genera and change its ending to "-idae". The fact that the oldest generic name is Heptagenia is irrelevant since the priority within the genus-group does not affect the priority within the family-group. The first family name had been derived from the name Ecdyonurus - Ecdyonuridae Ulmer 1920. However, there is an even earlier subfamily name Heptageniinae Needham 1901 derived from the name Heptagenia. According to the principle of coordination, the correct family name with appropriate authorship will be Heptageniidae Needham 1901 (despite the fact that in 1901 Needham neither considered this taxon a family nor used the ending -idae). Superfamily Heptagenioidea, tribe Heptageniini, and any other family-group name derived from the generic name Heptagenia will keep the same authorship. As applied to the family-group, the principle of coordination means that within this group the name of one of directly subordinated taxa will always be derived from the same generic name as the name of the next higher taxon. For example, the superfamily Heptagenioidea Needham 1901 includes the family Heptageniidae Needham 1901 with the subfamily Heptageniinae Needham 1901; if any tribes are recognized, one of them will be named Heptageniini Needham 1901.


But since the Code requires no coordination between the genus-group and the family-group, a family may not include the genus on whose name the family name is based. For example, there is a valid family name Polymitarcyidae, yet no valid generic name Polymitarcys, from which it had been derived. Within the family-group, the name Polymitarcyini Banks 1900 (derived from Polymitarcys and originally given to a tribe) is older than the name Ephoronidae Traver 1935 (derived from Ephoron and originally assigned to a family), while within the genus-group the name Ephoron Williamson 1802 is older than Polymitarcys Eaton 1868, considered to refer to the same genus.

The provision on independent coordination within the family-group was introduced in 1961. Till then, only species-group and genus-group names were subject to coordination rules, so different authors applied different principles to choose type genera for family-group names. That's why so many family-group names still circulate defying current rules.

Under the current Code, the principle of priority strictly applies only to cases where a valid name is being chosen among names published on different dates; competing names published in the same paper are subject to another principle - the principle of the first reviser (see I.3.1.2). The first reviser is the first worker to establish synonymy and to choose the valid name among competing ones.


I. Synonymy. As any other typified rank-based nomenclature, the one subject to ICZN considers equally ranked names having the same type (i.e. type taxon - see I. to be objective synonyms (see I.3.1.1). Subjective synonyms are different equally ranked names whose types are not identical but regarded as referring to the same taxon by those who hold them to be synonyms.


For example, in the classification where the genera

Heptagenia, Ecdyonurus, Rhithrogena, Epeorus, and Arthroplea are all members of the same family, the family name will have the following synonyms:

Heptageniidae Needham 1901

= Ecdyonuridae Ulmer 1920

= Arthropleidae Balthasar 1937


Taxa circumscriptions have no effect here: e.g., the name Arthropleidae has never been applied to a family containing all these genera, being established for a small family including just Arthroplea.

As we can see, the names considered synonyms under ICZN may belong to taxa of disparate circumscription. So it is crucial to be aware that in rank-based nomenclature synonymy is not quite the same thing as in circumscription-based one (see I.3.5.3).

Under the principle of priority, only the oldest synonym may be the valid name of a taxon. All other (junior) synonyms are invalid and thus not used, but are kept in the nomenclature (i.e. remain available): they are considered for purposes of establishing homonymy (see below) and may also become valid names as classification changes.


I. Homonymy. Homonyms are identical names given to different taxa. The ICZN requires homonyms to be eliminated; here, as with synonyms, the principle of priority also applies.

In zoological nomenclature no species-group name (whether specific or subspecific) may be repeated within a genus, i.e. combined with the same generic name. However, if a subgeneric name is used, it does not affect the homonymy of the species-group names. Species-group homonymy may be primary (when original combinations of different species have the same generic name with identical epithets) and secondary (when original combinations of different species had different generic names with identical epithets but after being reclassified such species ended up in the same genus). According to ICZN, renaming junior secondary homonyms is reversible but renaming junior primary homonyms is not.

Genus-group name (generic or subgeneric) should be unique within the animal kingdom (i.e. within zoological nomenclature). In this respect zoological, botanical, and bacteriological nomenclatures are independent, i.e. each of these nomenclatures disregard the names used in the others. This means, for example, that the same word can be the name of both an animal genus and a plant genus, but cannot be applied to more than one animal genus.


I. Universal rules and powers of the Commission. Nomenclatural rules are designed to promote name stability, that's why they tend to perpetuate practice already in place. However, no single code, however carefully phrased, may reflect established custom for every individual case.

For example, while it is generally accepted that the valid name is always the oldest one among available names of a taxon, in some cases a well-known taxon may have generally accepted and long used name other than its oldest available name. Under ICZN rules, this widely circulated name is to be replaced with the technically correct one, but since in this case revitalizing a name hardly familiar to anybody would compromise nomenclatural stability, the International Commission for Zoological Nomenclature issues a special ruling to formally establish such generally accepted name as the correct one by suppressing the older unused name or otherwise declaring it unavailable.

Depending upon situation, the Commission may suppress a name either for purposes of priority but not for purposes of homonymy (so that such name becomes invalid but another identical name may not be created) or for purposes of both priority and homonymy. The Commission may overrule the Code to validate designation of a type taxon. Rulings of the Commission always supersede the Code.


I. Spelling. The Code regulates not only choice of names, but also their spelling. As we have already mentioned, the names should be Latin or latinized (i.e. spelled using Latin letters and treated in accordance with the rules of Latin grammar); in terms of etymology, any derivation is allowed (Latin, Greek, other languages, even random combination of letters).

All genus- and family-group names begin with a capital letter, while species-group names are always lowercase. In old papers some species names (such as those derived from personal names) used to be capitalized; now we must spell them in lowercase only. Diacritic marks, apostrophe and diaereses, once used in some names, should be omitted in zoological publications (e.g., the name originally spelled Baёtis is now Baetis). Hyphens, punctuation marks, or digits in compound names are also excluded from modern zoological nomenclature, and each name is spelled as a single word (e.g., decempunctata instead of 10-punctata, sanctijohannis instead of s-johannis or st. johannis), with the only exception of names including a Latin letter to refer to a physical character (such names are spelled with a hyphen, e.g., c-album). All of the above applies only to zoological names; plants names are governed by the International Code of Botanical Nomenclature which sanctions hyphens, diaereses, and caps in species epithets.

All species names are binary, i.e. consist of a generic name followed by a specific name written separately without punctuation in between (e.g., Heptagenia sulphurea). The name of an animal subspecies is a combination of three names, i.e. a binomen followed by a subspecific epithet, also without punctuation marks (e.g., Heptagenia sulphurea dalecarlica). If there are subgenera, the subgeneric name may be inserted in parentheses between generic and specific names (e.g., Heptagenia (Heptagenia) sulphurea). No other word may be inserted in parentheses between generic and specific names. If the second part (the specific name, or epithet) of a binomen is an adjective, it must grammatically agree with the generic name, so as the generic name changes, the ending of the epithet may also change to reflect gender (subgeneric names in parentheses do not affect the epithet's gender).

Citation of authorship is optional, yet usually recommendable. Authorship/date refer to the last word of the name. The name of the author who has made the name available - usually with the publication year after it - follows the name with no punctuation marks in between. Date can be separated from the authors name by comma (no other punctuation marks are allowed). In this book, to avoid confusion, we omit commas in authorship and use them in bibliographic references. If the generic name had changed since the first publication, the authorship is cited in parentheses, e.g., Heptagenia sulphurea (Muller 1776), but Ephemera sulphurea Muller 1776, because this species, originally described by Muller as a member of Ephemera, was subsequently transferred to the genus Heptagenia.

If the author cited after a name is not the one who made the name available (such as the author of a new combination or the one who reconsidered the circumscription of the taxon) such citation should be separated from the name in a way other than using parentheses or a comma (see I.3.5).

The International Code of Zoological Nomenclature, Appendix B6 contains the following recommendation: "The scientific names of genus- or species-group taxa should be printed in type-face (font) different from that used in the text; such names are usually printed in italics, which should not be used for names of higher taxa." Accordingly, in this book we italicize all rank-based names of genus- and species-group (no matter how we treat them), while the names of other taxa (rank-based family-group names, hierarchy-based or circumscription-based names) are formatted otherwise (regular, bold, italics bold), depending on context.

I.3.3.2. Rank-based names of higher taxa

The International Code of Zoological Nomenclature regulates names of taxa ranked not higher than superfamily. There is no generally accepted rules of naming higher taxa (orders, classes, phyla, etc.).

Some authors advocate introducing a mandatory standardized typified nomenclature of higher taxa. They suggest all names of higher taxa to be derived in the same manner as family-group names, i.e. by modifying names of type genera with endings to reflect the rank. There is no consensus on what such higher rank endings should be (see Table 7). A number of established practices exist as to the use of typified names of higher taxa, depending on animal group.

The issue of how the principle of coordination should be applied to typified names of higher taxa is also a matter of argument: we can use no coordination at all (as in botany), i.e. disregard other ranks' names while selecting the oldest name for a taxon of given rank; or several independently coordinated groups (e.g., order-group, class-group, phylum-group) may be established in the same way as the ICZN's independent family- and genus-groups; or else, we can coordinate typified names of higher taxa with the family-group. Actual names will depend upon which one of these alternatives is adopted.

Hence, to adopt standardized typified rank-based names, zoologists must agree on two points: (1) standard endings (reach consensus on which ending each rank will have), and (2) name bases (i.e. the way to apply the principle of coordination).

None of these problems exist in the non-ranking hierarchy-based nomenclature we suggest below, because no endings are used in typified names and because it entirely relies upon those rules of coordination already provided by the current ICZN for the genus-group and family-group.


An ICZN-based non-ranking hierarchy-based typified nomenclature has been recently developed (Kluge, 1998, 1999).

Generic names, that are rank-based names by definition (see above), can be used not only in rank-based nomenclature but to derive hierarchy-based names as well, in which case type species of generic names (see I. above), authorship, priority and coordination all work just as provided by the ICZN (see I. Unlike in the ICZN's nomenclature, in the hierarchy-based nomenclature no name is assigned absolute rank (such as genus, family, etc.), but refers rather to a relative rank indicated by the number of higher hierarchically subordinated taxa. That's why the hierarchy-based nomenclature can be used in non-ranked classification.

The hierarchy-based, or hierarchical name (nomen hierarchicum) consists of an available generic name (or, better, of an available name of genus-group in terms of the ICZN - see above) to which, after a slash, the letter(s) f and/or g are attached, and a number (1 or higher). Here is the procedure to create a hierarchy-based name: first we pick the oldest generic name within the taxon in question. Since in our classification we use no ranks and just disregard them, the priority of the generic name (for purposes of obtaining a hierarchy-based name of the taxon) can be established based on authorship of either genus- or family-group names (they may not be the same - see I. above). After the slash we insert either a "g" (as in "genus") if the priority gets established based on genus-group names, or an "f" (as in "familia" - family) if we establish it based on family-group names. In case we need to chose the oldest name among several genus-group names or among several family-group names published in the same paper and not formerly considered rank-based synonyms (i.e. failing the first reviser's decision - see I., we choose the one which comes first in such paper. Number 1 is attached to the taxon which in our hierarchical classification is the largest (highest) among taxa for which this generic name is the oldest within the chosen group of names (i.e. within the genus-group for names followed by "g" or within the family-group for names followed by "f"). Subordinated taxa with the same generic name are numbered according to their order of subordination in such a way that the smaller (lower) is the taxon, the higher is the number.

The taxa with the same oldest generic name are numbered from the highest to the lowest, not vice versa, because the highest one can be easily identified based on priority, while taxa splitting can be unlimited.


Fig. 9 provides an example of how hierarchy-based names are generated. Let's assume we have attributed generic rank to a supraspecific taxon (as there is no restrictions on ranking, we may assign such rank to any taxon, however large or small, provided it includes at least one formal species). Then, based on ICZN rules for the genus-group (see I., we shall decide which generic name is valid for this taxon; in the example on Fig. 9 this will be Habrophlebia. This generic name becomes the base for the hierarchy-based name. We add to it the slash and a "g", which means the name is chosen under genus-group coordination rules; it give us Habrophlebia/g. A classification may include several hierarchically subordinated taxa, each of which, if assigned generic rank, will have the same name (Habrophlebia, in our example); the names will differ in their numbers. To start numbering we identify the highest (i.e. the most inclusive) taxon to which such generic name may be applied; this will be our number one. In our example the highest taxon whose oldest generic name is Habrophlebia gets the hierarchy-based name Habrophlebia/g1. No taxon whose rank is higher than that of Habrophlebia/g1 can bear the name Habrophlebia, because such taxon would include not only the type species of the name Habrophlebia Eaton 1881 (which is Ephemera fusca Curtis 1834) but also the type species of an older generic name Leptophlebia Westwood 1840 (which is Ephemera vespertina Linnaeus 1768), so if we assign that taxon generic rank its name will be Leptophlebia, not Habrophlebia. All subordinated taxa having the same generic name get consecutive numbers, starting from the highest one, in such a way that the larger is the number, the lower is the rank. In our example Habrophlebia/g1 embraces Habrophlebia/g2 and Habroleptoides/g1; Habrophlebia/g2 contains Habrophlebia/g3 and Hesperaphlebia/g1, because the names Habroleptoides Schoenemund 1929 and Hesperaphlebia Peters 1979 are younger than the name Habrophlebia.



Example how hierarchical names are formed in one of groups of mayflies.

Species are shown as dots, supraspecies taxa - as rectangles.


So, based on genus-group priority rules alone, one can assign unique hierarchy-based names to all taxa within a classification. However, as the current Code provides for a separate priority for family-group names (see I., the hierarchy-based names generated under the rules for the genus-group give no idea what should taxa names be in a rank-based nomenclature if assigned family-group ranks (i.e. if we treat these taxa as tribes, subfamilies, families, or superfamilies). It would be practical to present hierarchy-based names in such a format that would allow converting them into familiar-looking rank-based names without recurring to any additional information (once hierarchy-based names will take hold, this procedure will become unnecessary). All we need to make a hierarchy-based name convertible to rank-based one is adding to it, with an "=" in between (no spaces), another hierarchy-based name, this time the one obtained based on family-group priority. In this case the name base is spelled out in its original form (i.e. as a generic name, without family-group endings) followed by a slash and an "f" instead of "g". Both the generic name and the number established based on genus- and family-group rules often match; in such cases we just write down one generic name with the letters "fg" and the number.


In our example such are the names Habrophlebia/fg1, Habrophlebia/fg2 and Habrophlebia/fg3. In this case the oldest family-group name derived from the generic name Habrophlebia is Habrophlebiinae Kluge 1994; this name is younger than the oldest name derived from the generic name Leptophlebia - Leptophlebini Banks 1900; there is no family-group names derived from Habroleptoides or Hesperaphlebia.


If the generic names are identical but the numbers are not, we insert into the hierarchy-based name both numbers with their respective letters, separated by an "=" without spaces; if the generic names are not identical we write down both generic names (separated by an "=" without spaces) with their respective letters and numbers.


For example, there is a taxon that includes type species of two generic names: Caenis Stephens 1835 and Brachycercus Curtis 1834, of which the latter is older; however, the oldest family-group name derived from the name Caenis - Caenidae Newman 1853 - is older than the oldest family-group name derived from Brachycercus - Brachycercidae Lestage 1924. The taxon including both type species will have the hierarchy-based name Caenis/f1=Brachycercus/g1. Such spelling means that under ICZN rules, if this taxon is assigned genus-group rank its name will be Brachycercus, while if assigned family-group rank its name should be derived from the generic name Caenis. One of subordinated taxa within Caenis/f1=Brachycercus/g1 also includes both type species, and its hierarchy-based name will be Caenis/f2=Brachycercus/g2. In rank-based nomenclature this taxon also can be either named Brachycercus or get a typified name derived from Caenis, depending on whether we consider it a genus-group or a family-group taxon. This taxon, in turn, contains two taxa: one including the type species of the generic name Caenis, and another - the type species of the generic name Brachycercus. The hierarchy-based names of these taxa will be Caenis/f3=g1 and Brachycercus/f1=g3, respectively. Hierarchy of these taxa looks as follows:


             /                  \
Neoephemera/fg1               Caenis/f2=Brachycercus/g2
                               /                  \
                     Caenis/f3=g1               Brachycercus/f1=g3

In a hierarchy-based name it is convenient to write first the part generated according to the family-group rules (the one with "f" at the end), and then the part generated according to the genus-group rules (and ends in "g"), not vice versa: if in the hierarchy-based name both generic names and their respective numbers, generated under genus- and family-group rules, are identical (as is usually the case), it is better to write "fg1", "fg2", "fg3", etc., than "gf1", "gf2", "gf3", etc., because digits look more legible after "fg". If the generic names or numbers are not identical, the reader, to avoid confusing numbers, may look only at the last number after the "g", since every hierarchy-based name has a "g" (with a number), while "f" may be absent.

As classification changes, numbering in hierarchy-based names also shifts, so depending on classification taxa of the same circumscription may have different names, while taxa of different circumscriptions may be named identically. Hierarchy-based and rank-based nomenclature share such disadvantage, only circumscription-based one (see below) is free of it. The important benefit of hierarchy-based nomenclature is that the names shift only if there are changes in the classification, i.e. if the subordination of taxa is modified, while in rank-based nomenclature names change with any rank shift as well. Unlike rank changes, always purely discretional, classification changes are always based on evidence and can be discussed.

If a hierarchy-based name is used, it may be helpful to provide details on how taxa are arranged in this classification, as a comment on the name's number. This can be done when the name is first mentioned, listing (in parentheses) generic names of closest excluded taxa (using "sine", Latin for "without") and those of directly subordinated taxa (using "incl." - incluso, including). For the names from the above example such comments may look as follows:


Leptophlebia/fg1 (incl. Calliarcys, Habrophlebia, Atalophlebia);

Habrophlebia/fg1 (incl. Habroleptoides);

Habrophlebia/fg2 (sine Habroleptoides; incl. Hesperaphlebia);

Habrophlebia/fg3 (sine Hesperaphlebia).


Hierarchy-based names of the taxa discussed in this book are listed in Table 8.

In this book we use both hierarchy-based and circumscription-based nomenclatures. Practical application of these to a non-ranking classification is discussed in I.3.7 below.


In addition to the names which are intrinsically circumscription-based (see below), any name (including rank-based one) may become circumscription-based if we cite the authorship of its circumscription. For example, "Libellula sensu Linnaeus, 1758" or "Libellula: Linnaeus, 1758" will mean the same thing as the circumscription-based name "Odonata" (see below), while the authorship cited without "sensu" or a colon (as in "Libellula Linnaeus 1758") refers to the rank-based generic name Libellula. Such widely used formats enable workers to provide circumscription-based names for any taxa they want, but the method is rather cumbersome and confusing: the same circumscription-based taxon may be named in many alternative ways if different names with meanings suggested by different authors are used.

Special circumscription-based names have a wider circulation. Such name can be called either circumscription-based, or circumscriptional (nomen circumscribens); in the previous papers this term was translated as "volumetric name" (Kluge, 1999a, 1999b). Each circumscription-based name is attached to a taxon of a given circumscription (i.e. a certain set of members included). Most generally accepted circumscription-based names are those of major taxa, but there were no rules governing their usage until recently, when such rules has been suggested (Kluge, 1999) based on long established and well-proven custom. Unlike ICZN rules for rank-based names, the rules proposed below for circumscription-based nomenclature are unofficial and thus not mandatory.

To make a decision on using a circumscription-based name, we assess the name based on three criteria: 1) availability, 2) circumscription match, and 3) validity (rank-based names are assessed by availability and validity only). The key concept of circumscription-based nomenclature is the "admissible membership", based on which one can decide whether several taxa are identical in terms of circumscription.

I.3.5.1. Circumscription-related terms

Comparing taxa mentioned and characterized in various publications may be difficult and generate conflicting opinions on which taxa treated in such papers are identical and which are not. To determine whether taxa match in circumscription, the following parameters had been introduced (Kluge, 1996):

Original includes membership: the set of species, explicitly or implicitly quoted in the original publication of thee as members of that taxon.

Original net included membership: original included membership less (1) the species whose position is considered uncertain in the original publication, and (2) the species erroneously listed as members of the taxon contrary to the diagnosis provided for that taxon in the original publication.

Original excluded membership: a set of species explicitly or implicitly quoted in the original publication of the name as non-members of that taxon.

Original net excluded membership: original excluded membership less (1) the species whose position is considered uncertain in the original publication, and (2) the species erroneously listed as members of another taxa contrary to the diagnoses provided in the original publication.

Original circumscription (or original admissible membership): any set containing all species of original net included membership and no species of original net excluded membership.

I.3.5.2. Criteria of availability for circumscription-based names

Since rank- and circumscription-based nomenclatures are inherently incompatible, it would be helpful to effectively separate rank-based names from circumscription-based ones. We propose to do so using such criteria of availability that would make names available as rank-related unavailable for purposes of circumscription-based nomenclature, and vice versa.

It would be appropriate to consider available for purposes of circumscription-based nomenclature all scientific names published since 1758 other than species-, genus-, and family-group names subject to ICZN, and other than typified names (those derived from type-genera names using only suffixes, endings and/or the stems "-form-" and "-morph-"). In that case all species-, genus-, and family-group names and all typified names (including both family-group names and typified names of higher taxa) would be available only as rank-based ones.


Examples: Odonata, Odonatoidea, Odonatoptera, and Odonatopterata are available circumscription-based names, as they are derived from "Odonata" which is not a generic name (though some of these names have been first introduced as rank-based). Ephemeroptera, Ephemeropteria, and Ephemeropteroidea are available circumscription-based names, as they are derived by adding the stem "-pter-" to the generic name "Ephemera". Oniscomorpha may be either an available circumscription-based name if assigned to a taxon within Diplopoda (where there is no genus named Oniscus), or a typified rank-based name if assigned to a crustacean group containing the genus Oniscus.

I.3.5.3. Circumscription fit

Depending on how well names fit taxa in terms of circumscription, names may be non-fitting, non-monosemantically fitting, or monosemantically fitting.

1. Non-fitting name is such available circumscription-based name whose admissible circumscription (see I.3.5.1) is inconsistent with the taxon's circumscription. In circumscription-based nomenclature such name cannot be valid for this taxon by definition.

2. Non-monosemantically fitting name (nomen circumscribens non monosemanticum) is such available circumscription-based name whose admissible circumscription is not inconsistent with the circumscriptions of both the taxon in question and another taxon (taxa) within the same classification.

3. Monosemantically fitting name (nomen circumscribens monosemanticum) is such available circumscription-based name whose admissible circumscription is not inconsistent with the circumscription of the taxon while being inconsistent with the circumscription of any other taxon within the classification.


Example: for the name Odonata Fabricius 1793, the admissible circumscription would be any set including its original listed membership (all 52 species Fabricius included into his class Odonata) and not including the original excluded membership (all species Fabricius assigned to other classes in the same paper). Adding thousands of dragonfly and damselfly species described thereafter into Odonata is not inconsistent with the original admissible circumscription of the name Odonata, because Fabricius, being unaware of such species, didn't refer to them as non-members of Odonata. Therefore, the name Odonata originally given to the taxon of 52 species, also fits the circumscription of the taxon currently known as Odonata and including over 5000 species. Likewise, Fabricius said nothing on whether Meganisoptera and other extinct groups remotely related to extant odonates are part of Odonata. This means that if we use a classification where a taxon includes all odonates and Meganisoptera while its subordinate taxon includes all extant odonates but no Meganisoptera, the name Odonata should be considered non-monosemantically fitting either taxon (see VI-1.2.1 Odonata).


Depending on classification, the same name may be non-monosemantically fitting or monosemantically fitting a taxon of the same circumscription.

Example. If we agree with those suggesting that different groups of odonates evolved from different Meganisoptera, and adopt a classification where extant odonates do not form a single taxon distinct from Meganisoptera, the name Odonata applied to the taxon including all odonates and Meganisoptera should be considered monosemantically fitting such taxon.

Circumscription-based synonymy. Different names fitting the same taxon are circumscription-based synonyms. Such synonyms may be monosemantic (synonyma circumscribentia monosemantica) or non-monosemantic (synonyma circumscribentia non-monosemantica). Circumscription-based synonymy fundamentally differs from the rank-based synonymy, including synonymy as defined by ICZN (see I. above); unlike rank-based synonyms, circumscription-based synonyms apply to taxa identical in circumscription, but not necessarily in rank. Therefore, whenever we mean circumscription-based synonyms we should always make it clear. The valid circumscription-based name should be chosen among circumscription-based synonyms, if any.

I.3.5.4. Validity of circumscription-based names

The valid circumscription-based name is the oldest one among monosemantically or non-monosemantically fitting available names of the taxon. If a monosemantically fitting name is also the oldest, it becomes the only valid name of the taxon; if the oldest name is only non-monosemantically fitting, then the oldest monosemantically fitting name becomes the second valid name of the taxon.

Thus, unlike the rules for rank-based names, those for circumscription-based names allow for more than one valid name for a taxon.

In this book we use both circumscription-based and hierarchy-based nomenclatures. Their practical application in non-ranked classification is discussed in I.3.7.


The rank-based and circumscription-based approaches to nomenclature are irreconcilable. Combined use of the two results in confusion about what does this or that name stand for.

I.3.6.1. Customary interpretation of non-typified names

Non-typified names other than those of the genus- or species-group constitute the majority of generally accepted names of taxa higher than family and some others; such names are not governed by ICZN. The rules we have proposed (see above) provide for their use as circumscription-based names only; however, they are still being used both ways (i.e. as circumscription- or as rank-based). Curiously, the lack of rules induces customary division of names into rank- and circumscription-based: some names are treated by majority of workers as rank-based (i.e. those attached to the taxon as far as the latter keeps its rank despite any changes in circumscription), while others as circumscription-based (i.e. attached to the taxon as far as the latter keeps its circumscription despite any rank changes). What causes such a division remains unclear.


For example, the name Chilopoda Latreille 1817 is traditionally used as circumscription-based: its rank had been promoted from family (Latreille, 1817) to class level (Pocock, 1893) with no changes in circumscription. The name Neuroptera is traditionally treated as rank-based and nearly always assigned to an order-level taxon, no matter how much its circumscription changed. In the 10th edition of "Systema Naturae" (Linnaeus, 1758) order Neuroptera included genera Libellula, Ephemera, Phryganea, Hemerobius, Panorpa, and Raphidia, and was obviously artificial. Later one group after another got moved out, so that presently the name Neuroptera is commonly used to refer to a holophyletic order including only members of the Linnaean genus Hemerobius (even not all of them). Aptera, another blatantly artificial order of the class Insecta in Linnaean classification, included all wingless arthropods. There is no currently recognized group fitting Linnaean Aptera (nor his Neuroptera, for that matter). Some authors tried to use the name Aptera as rank-based applying it to one of the orders into which Linnaean Aptera were split, viz. Aphaniptera, Anoplura, or Thysanura. However, most workers consider Aptera to be a circumscription-based name and thus reject it. Names like Apterygogenea Brauer 1885, Apterygota Lang 1888, Apterota Haeckel 1896, etc. had been proposed for wingless insects to avoid the name Aptera, considered preoccupied under the principles of circumscription-based nomenclature.


When one taxon gets a rank-based name while another - a circumscription-based name, this may look as if taxa subordination has been reversed.


For example, the name Insecta is rank-based: it always refers to a class-ranking taxon, whose circumscription may, however, change from Arthropoda (=Insecta sensu Linnaeus) to Pterygota (=Insecta sensu Lamarck) (see below). In contrast, the name Hexapoda Blainville 1816 is circumscription-based, attached to a particular group of animals. As rank-based names move from one group to another, the rank-based name Insecta may be given to either a taxon higher than Hexapoda (e.g., Kingsley, 1888) or to Hexapoda (e.g., Heymons, 1901), or else to a taxon within Hexapoda (e.g., Lameere, 1895):               

   Kingsley, 1888:        Heymons, 1901:         Lameere, 1895-1900:
   Insecta                Arthropoda               Hexapoda
      Malacopoda             Teleiocerata             Iapygiens
      Arachnida              Chelicerata              Campodes
      Myriapoda              Atelocerata              Collembola
      Hexapoda                  Myriapoda             Thysanura
                                Hexapoda=Insecta      Insecta

Using non-typified names as rank-based is obviously unpractical. Any demand for rank-based names can be satisfied by typified names derived from the name of a type genus (see I.3.3.2 above), while non-typified rank-based names are confusing since they have no designated type taxa. Rank-based nomenclature can be coherent and clear only if the ranks and type taxa are defined, as provided, for instance, by the ICZN (see I.3.3.1). Some authors attempted to designate type genera for non-typified names, but their effort met little support, probably because there is no explicit agreement on which non-typified names should be regarded as rank-based, and which as circumscription-based ones.


For example, in the case discussed above, McLeay (1819-1821) has designated Libellula as the type genus of Linnaean order Neuroptera, but many ignored this fact. After the Linnaean Neuroptera have been divided into smaller orders, some workers applied the name Neuroptera to the order which included Libellula, while others to orders that did not; after a long period of confusion, the latter trend prevailed.


The best practice would be treating all non-typified names, except for ICZN's genus- and species-group names, as circumscription-based (according to the rules of circumscription-based nomenclature proposed in I.3.5.2 above), and keeping rank-based usage only for names having generally accepted types (including genus- and species-group names).

I.3.6.2. Polyphyletic Hexapoda: How a myth was born

Myths in science are misconceptions generated by misinterpretation of texts rather than evidence. Unlike scientific theories and hypotheses, myths are not developed by anybody: they lurk in literature so people cannot but either trust or question them. The flaws of the existing nomenclature may cause readers misinterpret certain statements contrary to the author's will (Kluge, 1996). One such myth is that there is a theory about polyphyly of Hexapoda.

This myth stems from a change in the classification of Hexapoda in light of new data clarifying the phylogeny of this group. A classification with the taxon Hexapoda divided into two equally-ranking taxa, Apterygota and Pterygota (see below, V-1: Classifications of Hexapoda I), was common knowledge during a long period. Many authors considered Apterygota to be a paraphyletic taxon ancestral for Pterygota; it took a while to find out which group within Apterygota is the closest (ancestral or sister) to Pterygota, so Apterygota lingered as a taxon. After Diplura, Collembola and Protura were shown to share a unique synapomorphy, viz. pouched mouthparts [see below, V-1.1: Entognatha (1)], Apterygota was divided into Entognatha and Triplura. Later, Triplura and Pterygota were also found to have a synapomorphy: muscles missing in all but one (basal) antennal segments [see V-1.2: Amyocerata (1)].

Based on these apomorphies a new classification has been suggested, where the divisions of Hexapoda were not Apterygota and Pterygota, but two holophyletic taxa, viz. Entognatha and Amyocerata, the latter, in turn, divided into Triplura and Pterygota. These classification changes may be summarized as follows:

               Hexapoda                     Hexapoda 
               /      \                     /       \
       Apterygota    Pterygota      Entognatha     Amyocerata
        /      \                                    /      \
Entognatha    Triplura                        Triplura    Pterygota

So, taxa within Hexapoda got rearranged, but the integrity and holophyletic nature of the taxon Hexapoda remained unchallenged.

Until now we used in this discussion rank-independent names monosemantically fitting taxa circumscriptions. But a number of rank-dependent names have also been circulating in the meanwhile. Let us see how the same classification changes will look if instead of using the names strictly based on circumscription they are treated as rank-based. The familiar name Insecta is a good example.

The name "Insecta" seems to never have had an unambiguous meaning. Insecta is the Latin for the name Entomon used by Aristotle. Aristotle's Entomon embraced a range of terrestrial arthropods, but not crabs (he placed them into Malacostraca). Such conventional usage of the name Entomon has survived: the term "entomology" still means the study of insects, arachnids and myriapods, but not crustaceans. Linnaeus, contrary to established custom, included all crustaceans into the order Aptera of his class Insecta (Linnaeus, 1735-1796) coextensive with what is currently known as Arthropoda (while the class Entomon, or historically interpreted Insecta, has no match in modern taxonomy). Though some later workers (Fabricius, 1792-1798; Latreille, 1796, and others) did use the name Insecta in Linnaean sense, this name was also being used in its traditional meaning, i.e. without crabs (e.g., Latreille, 1802-1805). Later attempts to better define Insecta and make it natural caused this taxon to change its circumscription many times. Lamarck included into Insecta only winged insects; other authors, including some recent workers, used the name Insecta to refer to taxa of various circumscriptions in between Insecta sensu Linnaeus and Insecta sensu Lamarck. Since Linnaean times the word "Insecta" has had the following meanings:


Insecta: Linnaeus 1758 = Arthropoda Siebold et Stannius 1848;

Insecta: Lamarck 1801 = Pterodicera Latreille 1802;

Insecta: Latreille 1806 = Tracheata Haeckel 1866;

Insecta: Leach 1815 = Hexapoda Blainville 1816;

Insecta: Cuvier 1817 = Atelocerata Heymons 1901;

Insecta: Packard 1883 = Dimalata Sharov 1966;

Insecta: Kingsley 1894 = Opisthogoneata Pocock 1893;

Insecta: Handschin 1958 = Amyocerata Remington 1955;

Insecta: Chen 1962 = Pleomerentoma Krausse et Wolff 1919.


At no point during this whole period had the word "Insecta" have a universal meaning, but was always concurrently used by various authors in at least two or three different meanings. Before Linnaeus yet another meanings for Insecta existed: Reaumur, for one, applied this name to all animals but mammals, birds, and fish (i.e. treated both worms and reptiles as insects).

There is a tradition of applying the name "Insecta" to a class-ranking taxon, thus using it as a non-typified rank-based name. To follow the tradition, anyone inclined to use this name in a new classification must first identify the class-ranking taxon.

Whenever a new classification was created, the class rank (along with the name "Insecta") was ascribed, more often than not, to Hexapoda. In the old classification dividing the class Hexapoda (=Insecta auct.) into Apterygota and Pterygota, the taxon Pterygota got subclass rank, so the classification of Pterygota was designed down from this rank, i.e. next highest division of Pterygota were infraclasses, etc. Now that the class Hexapoda gets divided into Entognatha and Amyocerata, and the latter into Triplura and Pterygota, the rank of Pterygota is demoted to infraclass:


               class                       class
               Insecta(=Hexapoda)          Insecta(=Hexapoda)
               /        \                  /       \
        subclass       subclass       subclass    subclass
        Apterygota     Pterygota      Entognatha  Amyocerata
        /       \                                 /       \
infraclass     infraclass                 infraclass      infraclass
Entognatha     Triplura                   Triplura        Pterygota
Preserving the hierarchy within Pterygota would require one-rank demotion of all winged insect taxa, which would cause big inconvenience since Pterygota includes the majority of animal species. Keeping Pterygota in its former subclass status while upgrading taxa above Pterygota one rank seemed the least painful way out, so Amyocerata became a class, and Hexapoda - a superclass, while the name "Insecta", with its presumed class status, replaced the name Amyocerata, not Hexapoda.
               class                        superclass
               Insecta(=Hexapoda)           Hexapoda
               /        \                  /        \
        subclass       subclass        class         class
        Apterygota     Pterygota       Entognatha    Insecta(=Amyocerata)
        /        \                                   /          \
infraclass      infraclass                     subclass         subclass
Entognatha      Triplura                       Triplura         Pterygota

All this rank and name reshuffle had no theory behind it and was only done for the sake of convenience and tradition. Yet these technical changes brought to life a new myth about insect phylogeny.

Someone looking at the new classification may think of "class Insecta" as a well established concept needing no special comment, while actually the phrase "class Insecta" is used here in a new sense. The reader only notices that in the new classification the Entognatha, once a part of the class Insecta, has moved outside, and gets the impression that entognathous insects have by some reason been excluded from the class Insecta. The reader fails to notice that in fact all the taxa remain just where they used to be in the hierarchy, only the name "class Insecta" have moved from one taxon to another. He sees several renowned authors to adopt such classification change and concludes that it should be warranted by some solid theory (while the actual logic behind the consensus is anything but science - see above). As no explanations of the change is provided (the change being technical, there is nothing to explain), the reader believes that the underlying theory is commonplace. The theory that would make taxonomists exclude Entognatha from insects should be the one of polyphyletic Hexapoda, i.e. a claim that Entognatha and Amyocerata share no hexapodous ancestor. So the reader assumes that such theory had been developed, thoroughly debated, and endorsed by major authorities. In this case what works as a myth is not a theory about polyphyly of Hexapoda (there is no such thing), but the idea that such theory does exist. The myth is so pervasive that some workers are vigorously advocating this non-existent theory (see V-1: Status of Hexapoda).


I.3.7.1. Use of different nomenclatures

Among supraspecific taxa, the hierarchy-based nomenclature based on the natural hierarchy of the phylogenetic tree works better than the rank-based one relying on purely artificial absolute ranks, while on species level the rank-based nomenclature is adequate because it relies on evidence-based definition of species.

Therefore it would be only appropriate to switch to hierarchy-based nomenclature dealing with supraspecific taxa while retaining the rank-based nomenclature for species-group names. However, circumscription-based nomenclature is better equipped to satisfy taxonomic needs than hierarchy-based nomenclature, so if there is a choice between hierarchy-based and circumscription-based names, the latter should be preferred. Hierarchy-based nomenclature has only one, yet very important advantage over the circumscription-based one: to codify the circumscription-based nomenclature new rules are to be introduced (including homonymy rules - see Kluge, 1999), published names catalogued (see I.3.8 below) and many new names created, while to adopt the hierarchy-based nomenclature all we need is current ICZN rules and already available pool of genus-group and family-group names, so we may go ahead converting these rank-based names into hierarchy-based ones using a simple procedure.

I.3.7.2. Format of species name in non-rank-based nomenclature

Rejecting absolute ranks in supraspecific taxa would mean getting rid of the generic rank as well. But under ICZN any species name may only exist as a binomen, i.e. combined with a name of a genus (but not of a taxon of another rank). Specific epithet can not be used unless in such a combination since many epithets are used more than once throughout the nomenclature; it is also impossible to replace generic names in binomina with names of supraspecific taxa of another rank(s) because this would upset homonymy.

In a non-rank-based nomenclature, a species name might have the following format: first goes the specific epithet, then authorship and year (parenthesis is never used: secondary binomina do not exist since there is no binomina whatsoever), then the original generic name in square brackets (whether it reflects currently recognized position of the species or not); if the original combination included a subgeneric name, such name (in parentheses) may follow the generic name in square brackets. This or similar format is widely used in catalogues. The generic name works here as a surname: initially given based on kinship, it is not subject to change and is used for identification purposes no matter how accurately it describes actual relationship.

To indicate current position of the species we may add, before the epithet, a hierarchy-based name (see I.3.4 above) of any higher taxon where this species belongs; such names will not be mistaken for generic part of a binomen because of their distinct hierarchy-based format. For example, instead of Habrophlebia (Habrophlebia) fusca (Curtis 1834) we can write: Habrophlebia/fg3 fusca Curtis 1834 [Ephemera]; or, with less details on position: Habrophlebia/fg1 fusca [Ephemera]; or, with even less details: Leptophlebia/fg2 fusca [Ephemera]; or else: Leptophlebia/fg2 Habrophlebia/fg3 fusca [E.]. If our species is the type species of a generic name, we may insert an asterisk instead of the number after this name, e.g.: Habrophlebia/fg* fusca [Ephemera]. The asterisk means that in any classification the number here should be the highest. Unlike the original generic name (the one in square brackets) which can never be altered, the hierarchy-based name in front is subject to change to reflect progress in taxonomy.

Such format has obvious advantages over the one currently in use. The generic name within a binomen is charged with two conflicting functions: (1) ensuring the uniqueness of the species name, and (2) reflecting the taxonomist's opinion on the species position. Each of these tasks is difficult in itself, and trying to accomplish both at once it hardly practical. As species position changes, homonyms may emerge or vanish, thus creating the need to replace epithets; this may lead to extremely puzzling situations where the very identity of the species under discussion is not clear and downright confusing. On the other hand, there are cases requiring either more or less details about species position, so generic name alone may not be enough.


For example, in two different papers the names of two obviously different species look like this:


Species 1: Epeorus znojkoi (Tshernova 1938), and

Species 2: Epeorus znojkoi (Tshernova 1938).


Such name format sends confusing message on which species is referred to in each case. The following format carries much more information:


Species 1: znojkoi Tshernova 1938 [Ecdyonurus], and

Species 2: znojkoi Tshernova 1938 [Iron].


If our knowledge about these two species would be limited to original descriptions and type specimens (adults only), their respective positions could be presented as follows:


Species 1: Rhithrogena/fg2 znojkoi Tshernova 1938 [Ecdyonurus], and

Species 2: Epeorus/g2 znojkoi Tshernova 1938 [Iron]


since the taxa Rhithrogena/fg2 and Epeorus/g2 are well-defined based on adult stage. In current classification (see diagram below), Rhithrogena/fg2 is divided into Cinygmula/g1 and Rhithrogena/fg3; Cinygmula/g1 and Rhithrogena/g4 are defined based on larvae only. Epeorus/g2 is divided into Ironodes/g(1) and Epeorus/g3, and the latter into Epeorus/g4, Caucasiron/g(1) and other subordinated taxa; Ironodes, Epeorus/f4 and Caucasiron are defined based on larvae only. Now that the larvae of both species are known, we may elaborate:

Species 1: Rhithrogena/fg3 znojkoi [Ecdyonurus], and

Species 2: Caucasiron/g(1) znojkoi [Iron]

or in a more detailed way, Epeorus/g3 Caucasiron/g(1) znojkoi [I.].


All these alternative names showing the position of Species 2 are correct and differ only in the amount of details they provide. In contrast, traditional name format makes all but one binomina incorrect, while the only "correct" binomen turns to be different depending on publication.


The traditional binary format is very concise and convenient whenever species taxonomy is not an issue, e.g., in papers on ecology, physiology and other non-taxonomic texts. In this book we discuss the taxonomy of large supraspecific rather than species-level taxa, and therefore usually refer to species names is such short binary form. However, in papers dealing with species position and status it is advisable to use more elaborate non-rank-based name format.

I.3.7.3. Sliding binomina and polynomina

To show the hierarchical subordination concisely, names may be presented as binomina or polynomina. To do so, names of two or more taxa are to be arranged consecutively starting from the highest one; names of circumscription-based, hierarchy-based and/or rank-based nomenclatures may be used. There should be no punctuation marks between names, just like between elements of binomina in the current nomenclature.

For example, names of taxa may look like this:


Ephemeroptera Anteritorna Bidentiseta Branchitergaliae

Branchitergaliae Heptagennota Pseudiron

Heptagennota Pentamerotarsata

Pentamerotarsata Radulapalpata Rhithrogena/fg2

Rhithrogena/fg2 Cinygmula/g1 cavum Ulmer 1927 [Cinygma], and so on.


Polynominal names of supraspecific taxa in the above example refer to the underlined names in the following classification:


           Ephemeroptera s.str. (or Euplectoptera, or Ephemera/fg3)
           /    \
Posteritorna   Anteritorna (or Ephemera/fg4)
                /    \
      Tridentiseta  Bidentiseta (or Ephemera/fg5)
                     /    \
         Furcatergaliae  Branchitergaliae (or Heptagenia/f1=Oligoneuria/g1) 
                          /    \    
                  Eusetisura  Heptagennota (or Heptagenia/f2=g1)
                               /    \
                        Pseudiron  Pentamerotarsata (or Heptagenia/f3=g2)
                                    /     \
                            Arthroplea   Radulapalpata (or Heptagenia/f4=g3)
                                         /     |     \
                           Rhithrogena/fg1  Cinygma  Heptagenia/f5=g4
                             /           \
                Rhithrogena/fg2         Epeorus/g1
                 /     \                /      \
    Rhithrogena/fg3   Cinygmula/g1  Bleptus   Epeorus/g2
                                               /      \
                                         Ironodes    Epeorus/g3
                                                     /    |    \
                                            Caucasiron   Iron  Epeorus/g4

I.3.7.4. The layout of a taxonomic paper

Modern taxonomic papers of Linnaean tradition tend to give each taxon a separate diagnosis, a description, and a differential diagnosis (also called comparison or discussion); the diagnosis is supposed to contain a summary of diagnostic characters, the description to provide a detailed characterization of the taxon, and the differential diagnosis to pinpoint what distinguishes it from other individual taxa. Such reiteration of the same characters is unpractical, since description, diagnosis and differential diagnosis are but vaguely special. As a matter of fact, the description is just an elaborate diagnosis, while the latter is nothing but a concise description: both reflect only a part of the taxon's characters (the number of characters of any taxon being infinitely large), and only those of taxonomic importance. Differential diagnosis is believed to be special in that the characters are described in comparison with those of other taxa. However, any character is meaningful only in a comparative context, so the only thing which sets the differential diagnosis aside is that it presents the comparison expressly, while in both the description and diagnosis the comparison is implicit. Yet a scientific paper is no place for implicit statements: its very aim is to expose the author's findings and opinions. That's why it is essential for any taxonomic text (be it a description, a diagnosis, differential diagnosis or whatever you choose to call it) to provide comparison in an explicit form.

Within the Linnaean tradition taxonomic diagnoses usually follow a certain plan whose standard vary depending on rank. For example, if a diagnosis of an order says that forewings are transformed into elytra, the diagnoses of other orders within the same class will state that the forewings are not so transformed, while the diagnoses of superorders may not mention this character. Such approach makes it easier for the reader to find a character of interest while avoiding redundancy in diagnoses of consecutively subordinated taxa. But in papers following such a plan the diagnoses of taxa would depend on purely artificial absolute ranks, thus inconsistent with the goal of building a natural system where taxa are supposed to be natural entities.

In this book we use another approach to meet taxonomic requirements. Each supraspecific taxon is characterized following a universal rank-independent scheme allowing to do without assigning absolute ranks to taxa, i.e. to switch to non-rank-based post-Linnaean classification. First the autapomorphies are listed, then the characters of unclear phylogenetic status, then the plesiomorphies; finally, or perhaps among plesiomorphies, variable characters of the taxon may be mentioned (usually referring to lower taxa not discussed here). There is no special paragraph for synapomorphies: all the apomorphies shared with any other taxa are listed under "Autapomorphies" of a higher taxon rather than among characteristics of this one. All doubtful synapomorphies are listed as "Characters of unclear phylogenetic status"; this paragraph may be subdivided as necessary. Each character is not only described but also compared, whenever possible, to other taxa; its unique or otherwise status is indicated. If necessary, references are given [in square brackets] to the description of a more general character in a higher taxon.

All characters are numbered with the only purpose to facilitate looking for similar items in descriptions, not for counting characters. Character counting used in all recent cladistic computer programs deprives such programs of any scientific meaning (see I.1.3 above). One cannot assign numbers to characters unless for the sake of convenience, as we do when we arrange a text into sentences, paragraphs, chapters, etc.; the numbers may not be attributed any biological meaning.

A general "Index of characters" is attached to the main text (taxa characteristics). In the index, the characters are arranged by structural parts they belong; the arrangement of the latter follows the usual pattern. Under each character entry, taxa names are listed, each followed by a reference number under which the character is described. To facilitate the search, some characters are mentioned in the index more than once. From the outside, the index may look like character list and matrix of a cladistic paper, but it is just an index (like a table of contents or an alphabetic index), not a base for conclusions. The index lists all hierarchically subordinated taxa discussed in the text, not only "operational units" to be included into a matrix (the cladistic "operational unit" is rank-related, which makes it artificial and hardly meaningful). The numbers of structural parts used in the index are by no means a universal character numbering: each character is given a number to refer to the taxon description.


Nomenclatural principles cannot be consistently applied unless catalogues of published names are available.

The most comprehensive source on family-group, genus-group, and species-group names (and other published data on animal taxonomy) is the Zoological Record, published since 1864 by the Zoological Society of London, first alone, then (since 1981) in partnership with BIOSIS, a U.S. company. Each volume covers the literature published during a year. To find the information you need you must know (at least roughly) the year of publication, which should be not before 1864.

Information about earlier names or names whose date of publication you want to know, can be found in a number of catalogues, viz. Agassiz, 1848 (an early incomplete catalogue listing names of classes, orders, genera and species); Index animalium, 1902 (early generic and species names published during 1758-1800); Index animalium, 1922, 1931-1933 (early generic and species names published during 1758-1850); Nomenclator zoologicus, 1939-1996 and Nomenclator animalium generum et subgenerum, 1926-1940 (genus-group names published since 1758). None of these provides information about family-group names or types of genus-group names.

There are also more detailed catalogues dealing with individual taxa. Listed below are catalogues covering subject groups of this book, i.e. myriapods, wingless insects and palaeopterans.




The above catalogues list names subject to the ICZN, only Agassiz (1848) gives some non-typified names available for use in circumscription-based nomenclature, but fails to mention many names published by that time. As there is no other catalogue of circumscription-based names, they are often misinterpreted in terms of authorship and original meaning (for some the authorship was not established until recently). Recently N. Kluge and A. Lobanov started the TAXA_NOM project to create a data bank of supraspecific names other than ICZN's genus-group and family-group ones. The bank includes information on how various authors used each name in various classifications, and all authorships ascribed to each name. Unlike prior catalogues and data bases, based invariably on a presumed classification, TAXA_NOM treats all classifications impartially and has no such bias.


For catalogue of circumscriptional names of arthropod taxa - see NOMINA  CIRCUMSCRIBENTIA INSECTORUM

 to general index