Good and Bad Names ∞
One of the most popular use cases is finding whether or not a name string corresponds to a current and valid scientific name. And most users want just one “correct” answer. Currently Global Names Index/Resolver contains about 23 million name strings coming from many sources. To figure out which of these name strings we should return to users we started a Google Summer of Code project in 2015 and @wencanluo tried to classify them into “good” and “bad” ones.
What is name-string?
A name-string is a collection of characters, spaces, and punctuation that makes up a name: “Homo sapiens Linnaeus 1758” and “H. sapiens L.” are two name strings representing the same name.
What are ranks?
|Ranks - in the hierarchy of taxa||Example|
|Kingdom||Metazoa or Plantae|
|Division or Phyla||Vertebrata|
What is a good name?
Most people will NOT know what they mean by “good”. They probably expect that there is a correct name, and do not want to use a name that is wrong. There is in fact no single correct name, because one element of applying a name to a taxon (a related group of organisms) is subjective. There is usually broad, frequently unanimous, consensus, but there will always be a minority who think that they should use different names for the same thing.
What was considered the right name also changes over time.
I understand that our task is only to find out which name should be considered “good” today. That makes things easier.
So, it is wiser to think about “good” and “bad” names, and then to identify factors that help to distinguish the better names from the less good names.
For a name to be “good” three factors need to be considered:
- Does the name-string represent a scientific name
- Does the name comply with a relevant code of nomenclature
- Is the name being used now by one or more taxonomists to refer to a taxon (a group of organisms with a common taxonomic history)
What are we trying to establish?
Is this a scientific name, is it nomenclaturally correct, and is endorsed by a taxonomist
Ideally, a good name is one that is nomenclaturally good (complies with a code) and also is deemed to be the best name for a taxon (taxonomically endorsed) according to one or more good taxonomists.
The long term vision
In the long term, we would expect that GN is dynamically connected to other on-line specialist sources that keep GN informed as to whether the name is scientific, and whether it is taxonomically endorsed. One parts of those resources are available, and sometimes not freely and openly; such that the infrastructural service that we dream of, cannot be created today. Because of this, we need to find some way to estimate the likelihood of a name being good.
What makes a name “scientific”?
Scientific names are latinized. Alternative names are common names (also referred to as vernacular or colloquial or familiar or informal - such as cat, dog, crow, maple); there are surrogates for names (being name-strings that may refer to a culture, or some term by which an organism is widely known in, for example, research settings).
Names have ranks, that indicate where in the hierarchy of classification they belong: Kingdom, Phylum, Class, Order, Family, Genus, Species) - see the vocabulary document as these are just the most common ranks. The most numerous names are those for species (there are about 2.3 million species, and most will have had more than one name). A species name is distinctive. It is made of two parts, a genus and a species part (Homo sapiens, Drosophila melanogaster). These are binomial names. Not all latinized binomial names that have a capital letter at the front are organisms: ‘Anorexia nervosa is an eating disorder, and Habeas corpus is a legal term’).
Most other names are uninomial: the Hominidae is the family that includes Homo - us.
But, you will find trinomials and polynomials, when an author wants to indicate, as an example, which subgenus a species belongs in: Drosophila (Sophophora) melanogaster tells us that the author places the species ‘melanogaster’ in the subgenus Sophophora.
Namestrings and names
A name-string is the collection of characters, spaces, and punctuation that can be used for the name. There can be very many name-strings for the same name - as seen below. In addition there are abbreviations, mis-spellings, and who knows what. In trying to match name-strings together, we try to identify the ‘canonical’ version, being just the latinised words without annotations, abbreviations, or names or authors or dates. Dima’s parser seeks to do this.
The following illustrates a variety of name-strings (not all!) that are used for the same scientific name, a subspecies called Carex scirpoidea convoluta. Names may also be mis-spelled or annotated in various ways.
Carex scirpoidea convoluta Carex scirpoidea var. convoluta Carex scirpoidea subsp. convoluta Carex scirpoidea var. convoluta Kuk. Carex scirpoidea convoluta Gardner Carex scirpoidea convoluta Kükenth. Carex scirpoidea var. convoluta Kük. Carex scirpoidea var. convoluta Kükenth. Carex scirpoidea var. convoluta Kükenthal Carex scirpoidea Michx. var. convoluta Kük. Carex scirpoidea ssp. convoluta (Kük.) Dunlop Carex scirpoidea Michx. var. convoluta Kükenth. Carex scirpoidea subsp. convoluta (Kük.) Dunlop Carex scirpoidea ssp. convoluta (Kukenth.) Dunlop Carex scirpoidea Michaux var. convoluta Kükenthal Carex scirpoidea subsp. convoluta (Kük.) D.A.Dunlop Carex scirpoidea subsp. convoluta (Kük.) D.A. Dunlop Carex scirpoidea Michx. ssp. convoluta (Kük.) Dunlop Carex scirpoidea subsp. convoluta (Kuk.) D. A. Dunlop Carex scirpoidea Michx. subsp. convoluta (Kük.) Dunlop Carex scirpoidea Michx. ssp. convoluta (Kükenth.) Dunlop Carex scirpoidea subsp. convoluta (Kükenthal) D.A. Dunlop Carex scirpoidea Michx. subsp. convoluta (Kük.) D.A.Dunlop Carex scirpoidea Michx. subsp. convoluta (Kük.) D.A. Dunlop Carex scirpoidea subsp. convoluta (Kükenthal 1909) D.A. Dunlop 1998
Where do we find information about whether a name is nomenclaturally correct, next …
Names which are good nomenclaturally
Theses are names that comply with relevant code(s) of nomenclature. This category is more or less the same as we call ‘scientific names’. Code compliancy usually stops at the level of Family.
Codes of nomenclature
The codes of nomenclature provide a set of rules as to how names of organisms should be formed. There are a number of codes, but most follow the same basic rules
- Names should be formed of latinized names
- Names should be introduced (published) in a scientific document. The part of the document that includes the introduction of a name (or a change that complies with a code) is referred to as a nomenclatural act.
- Names of species have two words, the first, the genus, is capitalized; and they are usually presented in Italics
- There has to be a single specimen placed in a Museum or similar place that is an example and the ultimate reference of the species to which the name is used.
- In addition, the first species described in a genus is the type species for that genus, and the first genus described in a Family is the type for the family.
- If a name is used, the same name cannot be used for another kind of organism (but see homonyms and ambiregnal issues below)
- If two or more groups come up with different names for the same species, the first one to apply a name has priority
Different codes of nomenclature apply to:
- Animals (International Code of Zoological Nomenclature, or ICZN also referred to as the zoological code),
- Plants (The International Code of Nomenclature of Algae, Plants and Fungi - ICN or ICNAFP, the ‘botanical code’
- For bacteria, viruses, cultivated plants, and more
- The ‘Biocode’ an effort to create a unified framework for all of these ‘typological codes’ - i.e. ones that require a type specimen.
- Finally, there is the ‘Phylocode’ which is not typological, and is used to refer to clades of life.
In our case, the majority of names will be those that relate to the Botanical and Zoological codes.
Nomenclatural criteria are ‘objective’ because scientific names are created in a published nomenclatural act.
The names follow a standard format: Homo sapiens Linnaeus 1758, but may be represented by variations on this theme - see the Carex example above.
The name may not be considered the appropriate name for a taxon. Even if these names are considered ‘wrong’ now – they still represent a nomenclatural act which really happened and therefore “good” nomenclaturally.
Code compliant names may be:
- Currently accepted names, i.e. a taxonomist says that this is the best name for a taxon.
- Valid synonyms of any sort – homotypic synonyms (old combinations), heterotypic synonyms (two names for what was found be one species) - see below
- They may be annotated to, for example, to distinguish different taxon concepts (sensu, sec. … “Stenometope laevissimus sec. Eschmeyer 2004” or “nec Stenometope laevissimus sensu Patterson, 2001”). This is used to tell someone that the taxon I am talking about is as defined or used by Eschmeyer, but it is not the same as what Patterson meant when he used that name.
- Other annotations may refer to names described after the fact (ex … like “Osmundaria obtusiloba (Mertens ex C. Agardh) R.E. Norris” which means Agardh added a name to an item, but did not describe it in a way that complies with a code; later Mertens described it in proper compliance with the code, and later Norris created a new combination)
Such names do not include:
- Misspellings except inasmuch as permitted or enforced by the codes, as they may require changes to names after publication
- Chresonyms (names containing authors who had nothing to do with the corresponding nomenclatural act … like “Monochamus galloprovincialis De Fluiter, 1950”). Typically a chresonym is used to indicate the use of a name by a particular author. In such a case, it should have a colon – Homo sapiens:Mozzherin, 2015; but many taxonomic sources will include chresonyms among synonymies, and not distinguish the two.
- A name not properly described or published
Who keeps track of nomenclaturally correct names?
A subset of particularly diligent taxonomists will keep track of all of the Code compliant names in their area of interest - Bill Eschmeyer, for example, is the principal source for fish names. There have been a major number of books that have sought to keep track of nomenclaturally compliant names - Airey Neave’s Nomenclator Zoologicus, Sherborne’s Index Animalium, or the Index Nominum Genericorum (for plant genera).
Taxonomically authoritative sources, such as the World Spider Catalog are good sources of nomenclatural information. There is no complete list of such sources.
This responsibility is now passing to ‘registries’, one line environments that keep track of new names. Examples are Index Fungorum, ZooBank, and International Plant Name Index (IPNI). Increasingly, these will replace publications as the places where names officially appear.
The Global Name Usage Bank, based at Bishop Museum in Hawaii, was set up to manage nomenclatural acts, and is the basis of ZooBank.
Ideally, GN could find code-compliant names by calling on services from registries. Nice idea, BUT
- Even a good nomenclator is not complete
- Many taxonomic areas are not covered by nomenclators of any sort
- Digital nomenclators may not provide web services
- Some nomenclators do not freely release their information.
Can we be certain if a name is code-compliant?
Yes, when a name occurs in a nomenclator, or in registry, or in an exemplary taxonomic source - you should trust that source. BUT, many (most) code compliant names will not have made their way to on-line sources, so in the digital age, you must exercise caution.
Special problem of ambiregnal taxa
The jurisdiction of the codes assume that all organisms fall comfortably into plants, animals, bacteria and so on. This is not the case. Some single-celled protists are considered to be animals by some researchers, and the zoological code has been applied to them. But other researchers think of them as plants, and have treated their nomenclature as those of plants. In the case of the cyanobacteria or blue green algae, they have been subject to both the botanical and bacterial codes.
As the codes are independent of each other, the ‘correct’ name in one code may be different to the correct name under another code. That is, there are two objectively correct names for the taxon.
Special problems of chresonyms
Chresonyms are not things that many people are familiar with, but are a major problem. A chresonym is a reference to the use of a name. For example, a review might state
Xysticus cristatus (Clerck, 1757) Koch, 1843 Araneus cristatus Clerck, 1757 Xysticus promiscuus O. Pickard-Cambridge, 1876 Xysticus cristatus Levy, 1976 Xysticus cristatus Levy 1985 Xysticus cristatus: Azarkina & Logunov, 2001 Xysticus caspicus Azarkina & Logunov, 2001
The indentation suggests the names following the first entry are synonyms. The list may even have the term ‘synonym’. While the first name is the basionym, first use of cristatus for a spider, the second is the author’s view that promiscuus, named as a different species refers to the same organism (i.e. is a heterotypic synonym), but the other four entries are chresonyms. The author of the list is trying to say, I think Xysticus cristatus (Clerck, 1757) is the same species as the thing described by Levy in 1976 and 1985, and by Azarkina and Logunov under the name Xysticus cristatus, and also the organisms described by Azarkina & Logunov under the name Xysticus caspicus.
Chresonyms are NOT code-compliant, because the person referred to is not the author of the scientific name, rather they are users of the name. The formulation of the name-string of a chresonym can be indistinguishable from a code-compliant name. In some cases, a reference is indicated by a colon, and that is helpful.
How can you tell if a name string that parses to a canonical form is a code-compliant scientific name (i.e. is good) or is a chresonym (is bad).
- It is possible that a name string with the same genus and species but different authors are homonyms. The number of species level homonyms are very low, and most have been compiled by Tony Rees (see IRMNG).
- If there is a colon in the name before author, then it is a chresonym
- Xysticus caspicus sensu Azarkina & Logunov, 2001 and Xysticus caspicus sec Azarkina & Logunov, 2001 are other ways that references to names may be written out
- If the genus name and species names match perfectly, but there are different authors and / or dates (that can’t be explained by a typographic error or different transliterations), then the later dated name will be the chresonym.
Special problems of Homonyms
Homonyms are names that are spelled in exactly the same way, but point to different taxa. Homonyms may occur because a later taxonomist may not be aware that the name has already been used. When this is discovered, the later homonym has to be changed. Another source of homonyms is that the codes are independent, so a name might have been used for an animal, but a botanist is still free to use the same name for a plant
A major source of names is IRMNG. It is compiled by Tony Rees. He believes he has compiled over 95% of all generic names, and therefore knows most homonyms. Perhaps 15% of generic names are homonyms, but the number of special names that are homonyms is around a hundred or so. Names currently accepted by taxonomists (subjective) Taxonomists publish their views about the species, and the most appropriate name for them in papers and books - some of which may take many years to be published. For each species, they will use the name that they think is nomenclaturally correct, and based on their concept of the species, decide which of the names that have ever been applied to this group, is the best one.
The scientific names used by taxonomists to refer to species and other taxa are fewer than all of the code compliant names. Such names are subjective because they represent an opinion of taxonomists, and different taxonomists may slice and dice the process of evolution (see heterotypic synonymy below).
The scientific names for taxa is a ‘perfect’ subset of names from the first group. It can only be formed from nomenclaturally valid names.
Which names are “good” taxonomically?
What all taxonomists must do is to use the single nomenclaturally valid name for the group that contains the type material.
Taxonomic subjectivity → judgement
Different taxonomists may differ in their views. This is because some are lumpers and some are splitters, and when a taxonomist casts a ‘wider net’ to draw in a larger sample or set, they may find that one name in the broader set has priority over the name that is widely used for the narrower set preferred by splitters. Also names may change because taxonomy is a dynamic discipline and closely linked to evolutionary studies. As relationships are revealed, taxa may move from one place to another in a classification, and names have to change so that the earliest name for the genus applies.
They will list names that have been used for a taxon, but are no longer the right name, as ‘synonyms’. Typically, there will be list with the correct name (in their view) being listed without comment or stated to be the senior synonym, and with incorrect names, sometimes prefaced with the word ‘synonyms’ listed below.
Synonyms may be homotypic or heterotypic. A homotypic synonym is ‘objective’ and comes from the priority rule in the codes of nomenclature. If I describe a new species of Paramecium (let us say Paramecium irelandicum), but someone later shows that the type material has a feature that is distinctive for and unique to an adjacent genus, Frontonia, then the name will be changed to Frontonia irelandica. We have simply moved the species into a different genus, and with that act, we must, according to the codes, change the generic vehicle to reflect the genus that contains the earliest described type species. As Paramecium and Frontonia must have had different types, when we create a new combination, at lkeast one name has changed. The original name is the basionym. In cases where the genus has been changed, taxonomists may or not include the authors of the names. When this change is referred to, some pedants may well include the name of the author of the basionym in parentheses, and the name of the author of the combination at the end. Paramecium irelandicum Me, 2014 Frontonia irelandica (Me, 2014) Her, 2015 or Frontonia irelandica (Me) Her
When a taxonomist revises a taxon, they consider all of the nomenclatural and taxonomic acts that have created species names, identified new species, and so on. Sometimes, they find that one later author (Patterson) has described as a new species (Volvox irelandicum) something that has already been described by Linnaeus (Volvox globator). Perhaps Patterson is a splitter and the reviewer is a lumper, or perhaps Patterson described an unusual form of Volvox globator, or a special stage in the life history, or perhaps Patterson just hadn’t read all of the literature. The reviewer will conclude that Volvox irelandicum is a synonym of Volvox globator. As Patterson, to be code compliant, had to place a type in a Museum, the two names are ‘heterotypic synonyms’, and the rule of priority demands that the first created name takes precedence.
How do we get information from taxonomists
Taxonomists publish their views in scientific papers and books.
Some have on-line web-sites that carry information about species, one of the best of which is FishBase. Many who have web sites do not have any services to allow machine-to-machine dialogue. For many other taxa, there are no on-line sources.
The largest compilation of taxonomies is the Catalogue of Life. It however collects data on each taxonomic area from a ‘Global Species Database’ (GSD), which means that the diversity of taxonomic opinion is not represented, and some GSDs may be very slow in responding to requests for updates. As a result, CoL is not consistent in quality.
For the moment, we should accept the assumption, if the name is in Catalogue of Life, it is taxonomically endorsed.
But, if a name is not in catalogue of life does not mean that is is NOT taxonomically endorsed.
If a name is not in Catalogue of Life, but is in another taxonomic source, then it is taxonomically endorsed.
Some compilations are less good and some more good.
Global Names collects lists and classifications, and we know where all names in GNI come from. That is, Global Names can indicate if it thinks a source is taxonomically reliable and therefore is a source of ‘Good’ names.