On the International Code of Area Nomenclature (ICAN)


*E-mail: rene.zaragueta_bagils@upmc.fr


A recent work proposed a new naming system for biogeographical areas: the International Code of Area Nomenclature (ICAN). This system aims to standardize area names used in biogeography and constitutes a key step towards maturity of this discipline. We point out a number of issues in this first version of ICAN. These include the lack of distinction between classification and nomenclature, the use of ranks, the criteria of area definition, the conflation of diagnosis and description, the ambiguous use of valid/available, and the system of amendments. We suggest postponing the initial date of application of this code until consistent rules are provided.

Ebach et al. (2008) recently proposed a new naming system for biogeographical areas: the International Code of Area Nomenclature (ICAN). This system aims to standardize area names used in biogeography and constitutes a key step towards maturity of this discipline. However, if ICAN’s foundations lack precision, ICAN will actually complicate the naming of areas. We would like to raise some critical observations that will contribute to the development of an internally consistent ICAN.

Classification vs. nomenclature

First, taxonomic nomenclature is ruled by five codes, not three (as ICAN stipulates). In addition to codes cited by Ebach et al. (2008), one can mention the International Code of Nomenclature for Cultivated Plants (Brickell et al., 2004) and the International Code of Virus Classification and Nomenclature (ICVCN) (in Fauquet et al., 2005).

Like the Phylocode, ICAN does not distinguish between nomenclature and classification. In their introduction, Ebach et al. (2008) state that ‘[n]omenclatural rules have been constantly revised during the 250 years since Linnaeus first proposed his classification system’. Despite the fact that nomenclature has been revised since Linnaeus (1736) proposed his nomenclatural rules, current nomenclatural codes date mainly from the 19th century (Strickland et al., 1843; Anonymous, 1847; de Candolle, 1867). A nomenclatural code is not intended to construct classifications, but is used to name entities. In other words, nomenclatural rules concern names, not taxa or areas. ICAN Sec. A Art. 3.2 states that ‘ICAN does not endorse any particular method among those in use for validating or classifying areas’. Such a statement is trivial. Likewise, in Sec. C Art. 1.1, it is stated that ‘[a]rea names may be grouped under more inclusive area names in order to represent a biogeographical taxonomic hierarchy’. The name ‘North America’, however, does not contain the name ‘New York’. The concept of hierarchy applies to areas, not to names.

A nomenclatural code aims to create objective links between names and concepts (taxa, taxonomic assemblages, temporal intervals or areas) through an object, i.e. a particular specimen called a holotype, whose properties are constant through time. In contrast, the concept associated with a name is subject to scientific discussion. To the greatest extent possible, the link between a scientific concept and the name that refers to that concept must be kept devoid of subjectivity. Similarly, a nomenclatural code must also be normative: decisions taken following its rules must be subject to as little interpretation as possible. Nomenclature should deal with: (1) the objective link between a name and the real world through an object, i.e. the holotype – or, more properly, the onomatophore (the name bearer); (2) synonymy, homonymy and the principle of priority; and (3) definition of publication, validity and availability of names. In its current form, the ICAN draft fails to specify some of these points.


In ICAN, the use of ranks is quite arbitrary, because there may be many more areas (nodes) than ranks, at least in some methods such as area cladistics and Brooks parsimony analysis.

It is not stated in ICAN how higher-rank areas inherit the type-locality of names given to included areas. In a ranked taxonomy, the onomatophore of a species is a holotype, the onomatophore of a genus is its type species and that of a family is its type genus. This system guarantees the link between the name of any supraspecific taxon and the holotype without multiplying onomatophores. In ICAN, the type-locality of a higher-rank area can be different from those of included areas. This renders decisions about synonymy/homonymy virtually impossible.

The statement that ‘[a] rank can only have one valid name’ (Sec. C Art. 2.7) is meaningless. ICAN’s authors probably meant ‘an area can have only one valid name at a given rank’.

Area definition

Ebach et al. (2008) state that areas of endemism ‘have been recognized as the principal units of modern biogeographical analyses’. Each one is viewed as ‘the area delimited by the distribution range of a taxon’. ICAN, however, specifies that only a type-locality and diagnosis or description is needed to define a biogeographical area (Sec. C Art. 2.1). Examples given in the introduction only include geographical descriptions and do not take taxa into account. This contradicts the statement that ‘[t]he Northern Adriatic would be synonymized with the Central Adriatic, for example, if it were concluded that the presumed endemics live in both the Central and Northern Adriatic’. In this case, taxonomic distributions are used to describe the Adriatic, not geography. In addition, it is uncertain whether the northern Adriatic and central Adriatic would be synonymized if only some endemics were found in both areas. ICAN’s authors should clarify their criteria, because the consistency of ICAN depends on the rigour of area definition.

Further, we did not find in ICAN any reference to the temporal significance of named areas. For instance, Ebach et al. (2008) describe the Adriatic by the presence of the Foci del Po. Yet we do not know where the Foci del Po was placed in the Eocene. We suggest that an area name should be associated with an explicit temporal interval (e.g. Recent/Pleistocene Foci del Po) because current biogeographical studies cannot deal with geographical variations through time, and limit themselves to a single geological age. This will prevent synonymization of area names based on studies concerning organisms of different geological ages.

Diagnosis vs. description

Ebach et al. (2008) conflate diagnosis and description throughout the ICAN draft (including Sec. C Art 2). For instance, they state that ‘the name and type-locality must be linked to a diagnosis (ICAN Sec. C Art. 2.3), which may consist of a detailed description of the area’. However, ICAN (Sec. C Art 4.2) states that ‘[a] name can be rejected if it has the same diagnosis, description, geographical coordinate or distribution as an existing name’. The latter article distinguishes these two concepts without means of comparing them, i.e. without explaining in which case a description may be identical to a diagnosis. According to the Botanical (McNeill et al., 2006; art. 32.2) and the Zoological (International Commission on Zoological Nomenclature, 1999) Codes, a diagnosis is a statement of the characters that distinguish a taxon from other taxa. In contrast, a description is a list of properties of the taxon itself (Vignes, 1991). In the context of ICAN, a diagnosis should compare the extent of the described area with those of published areas, while the ‘description’ paragraph should include a list of localities and endemic taxa. In reference to ICAN Sec. C Art 4.2 we suggest that decisions about rejection be taken only for diagnosis vs. diagnosis and description vs. description.


ICAN fails to define the date of publication of a name. Ebach et al. (2008) state that names have to ‘be published in a refereed journal’, which implies that monographs or technical/edited books are excluded from the naming system. Sec. C Art. 2.6 states that ‘[n]ames that are published with a diagnosis and/or descriptions linked to a previous discussion of reference are valid’, whereas Sec. C Art. 2.8 states that ‘[t]he principles of homonymy, synonymy and priority apply to the names within any rank that have valid diagnoses’. Accordingly, it is unclear whether validity applies to names and/or diagnoses. Moreover, Sec. C Art. 2.1 specifies that ‘[a]n available area name has a type-locality and either a published diagnosis or a description in a refereed publication’. Following this ambiguous statement, an unpublished area name can be available just because a type-locality and a description are published. Likewise, ICAN leaves uncertain which distinction, if any, is made between ‘available’ and ‘valid’ throughout the ICAN draft. The two distinct uses of the term valid in the Botanical (valid publication of a name) and Zoological codes (valid name) generate the aforementioned confusion.


Modifications of ICAN rules are made through a series of amendments (Sec. B Art. 2), instead of being directly incorporated into the code. Thus, independent modifications of an article and its subsequent amendments will eventually be impossible to track. In the future, ICAN will be unusable if this system of amendment is maintained.


In this paper, we point out a number of inconsistencies and ambiguous statements in the first version of ICAN. We suggest postponing the initial date of application of this code until internally consistent rules are provided. Members of the Systematic and Evolutionary Biogeographical Association are warmly encouraged to provide a completed version of ICAN. Such a universal naming system will significantly increase the efficiency of communication in biogeography and other related disciplines.

Editor: David Bellwood