Gender Agreement?
James Fetzner
Douglas Yanega
Hello Taxacomers,
I have a colleague that has sent me the following message. I wondered if I could get some opinions on his query and questions?
"Stenocephalemys albocaudata was named by Frick in 1914. In 2017, Denys et al. modified it to albocaudatus to have gender agreement. Since then, at least one author working on the genus, has accepted the changed name. There are other papers in the meantime, not taxonomic in scope, that continue to use the old name, likely not aware of the name change [proposed by Denys et al.]. I am writing up a paper on this taxon, as we have the holotype. Do I use the -us ending of the -a ending? Is it truly necessary for gender agreement to be the be all end all? I would guess that there are many taxa that are not in agreement. Does usage of more than 100 years have precedence?"
So, my feeling on this is that some taxonomists are sticklers for gender agreement, while others are not. However, my thoughts are that since the "-us" ending was so recently proposed and has been accepted by others, that he should use that ending in his paper. Thoughts?
This one is easy and non-controversial; "-caudata" is an adjectival termination, and must agree in gender with the genus name. Stenocephalemys is indisputably a masculine genus name, so the original spelling was in error, and must be albocaudatus under the Code. This is neither an emendation nor a change in spelling, technically, as the Code defines them; this is a mandatory change of termination for purposes of gender agreement.
It WOULD be controversial if the original name had been albocauda, because under the ICZN's Glossary definition of a noun phrase, albocauda is a noun phrase, and not an adjective (it consists of an adjectival modifer - alba - and a noun being modified - cauda). It would never change spelling regardless of the gender of the genus. Many, if not the majority, of practicing taxonomists would be inclined to treat albocauda as an adjective, and mistakenly think that it needs to be changed to albocaudus.
The Glossary of the Code is, however, considered part of the legislative text, and even if such a word happens to appear as an adjective in Greek or Latin, the Code's definition takes precedence. For reference, this includes species names ending in (or variants of) latinized Greek nouns like -brachium, -carpus, -cephala, -cercus, -chaeta (or -cheta), -coma (when used for "hair"), -derma, -discus, -(o)dontus, -gaster, -glottis, -gnatha, -lobus, -merus (when used for "thigh"), -morpha, -nema, -notum, -(o)nycha, -omma, -phyllum, -pilus, -pleura, -poda, -ptilum, -pyga, -r(h)achis, -r(h)inus, -r(h)ynchus, -soma, -spilus, -sternum, -tarsus, -(o)ura, etc. Most such terms are genuine Latin nouns, like -auris, -barba, -callum, -capillus, -cilium, -clavus, -clypeus, - collum, -corpus, -costa, -coxa, -crista, -dorsum, -folia, -forma, -gula, -gutta, -labium, -labrum, -linea, -macula, -manus, -nasus, -palpus, -pelvis, -penna (or -pinna), -pluma, -pterygium, -rostrum, -scapus, -scutum, -seta, -spina, -squama, -sulcus, -telum, -tergum, -tibia, -unguis, -valva, -velum, -venter, or (as we are discussing) the very commonly-used -cauda (Latin for "tail", where the Greek is "-ura" or "-oura" - so, yes, a name like "platyura" is not an adjective, and never changes to "platyurus").
These and nearly any other anatomical or structural terms that end in "-a", "-er", "-us", or "-um" are nouns, with VERY few exceptions, and species names using them as suffixes are noun phrases, as the Code explicitly defines them in the Glossary, a fact recognized by many authors (e.g., Langeneck & Strazzulla, 2025), but not all (e.g., David & Gosselin, 2002). That is why this is controversial. Sometimes it's easy to spot the difference between the noun version and the adjectival version (cauda/caudatus, crista/cristatus, dorsum/dorsatus, lobus/lobatus, macula/maculatus, rostrum/rostratus, spina/spinosus, tarsus/tarsatus, etc.), but often this is not intuitive, and it's easy to get this wrong.
Peace,
--
Doug Yanega Dept. of Entomology Entomology Research Museum Univ. of California, Riverside, CA 92521-0314 phone: 951-827-4315 FaceBook: Doug Yanega (disclaimer: opinions are mine, not UCR's) https://faculty.ucr.edu/~heraty/yanega.html "There are some enterprises in which a careful disorderliness is the true method" - Herman Melville, Moby Dick, Chap. 82
James Fetzner
Cheers,
Jim
laurent raty
Hi Doug,
I have long been (and still remains) highly uncomfortable with this reading of the Glossary definition of "noun phrase"...
In common parlance, a "noun phrase" is a group of words, formed from a "head noun", to which are added various possible types of modifiers, which as a whole denotes a specific entity or concept, and acts as a single noun in a sentence.
- The Glossary says :
- noun phrase, n.
- A compound word consisting of a noun combined with another noun or modifying adjective, the compound being treated as a noun in apposition; if the adjective is the final element in a species-group name, its ending is determined by the gender of the noun it modifies (and not by that of the generic name with which the species-group name is combined). For examples, see Article 31.2.1.
In Art. 31.2.1, we find : "the noun phrases in Melanoplus femurrubrum (Melanoplus masculine; but rubrum agreeing with femur, neuter) and Desmometopa m-nigrum (Desmometopa feminine; nigrum neuter, agreeing with m, because letters of the alphabet are neuter)".
A name which is a noun phrase as defined and exemplified in the above, is a name which, were it not for the Code mandating the opposite, would normally be written as two words (two nouns or a noun and an adjective, simply united into a single word, with their respective endings preserved). A name formed by combining, e.g., an adjectival stem (NB : this is not the same thing as "an adjective") and a noun (stem), perhaps with the addition of a connecting vowel and some kind of ending, into something that could only ever be written as a single word, is certainly not a "noun phrase" in this sense. (And to call it a "noun phrase" in the grammatical sense of the term would require accepting the notion of "single-noun noun phrases", i.e., that a single word alone can be a "group" of words, which not everybody agrees with.)
A few more examples just to illustrate the issues :
caputesocis = caput esocis, with caput = head (nominative singular, the head noun in the phrase) and esocis = esox, a pike (the fish ; in the genitive singular : "of a pike"), is a noun phrase combining two nouns, which I think we once discussed on the old Taxacom group.
albacauda = alba cauda, with alba = albus, -a, -um, white (in the nominative singular feminine, qualifying cauda) and cauda = tail (the head noun, nominative singular feminine), would be an acceptable noun phrase.
caudaalba = cauda alba, ditto but with the two words swapped, would be an acceptable noun phrase as well, and would not agree in gender with the generic name despite it ends in an adjective.
caudata = caudatus, -a, -um, tailed (in the nominative singular feminine) ; not in Latin dictionaries, but adjectival owing to cauda being a Latin noun, and -atus, -a, -um a Latin suffix turning nouns into adjectives.
albocaudata = albo caudata, arguably an adjective phrase (i.e., literally, "tailed in white", or "tailed with white"), although no such thing is defined in the Code.in which albo = album is the colour white (a neuter noun, in the ablative singular when it is used this way) and caudatus, -a, -um = tailed, Cannot be a noun phrase, because (1) caudata is adjectival, and (2) albo cannot be viewed as acting as a (feminine) head noun, which caudata would be modifying. One of these conditions is not sufficient.
"feminacaudata" = femina caudata, with femina = female (feminine head noun, in the nominative singular) and caudata = caudatus, -a, -um, tailed (in the nominative singular feminine, qualifying femina), although admittedly a rather improbable name, would in principle be an acceptable noun phrase if denoting a species characterized by females having a (long) tail.
albicauda / albocauda is certainly not a noun phrase in the sense of the Glossary -- even if, in nomenclature, such names are indeed nowadays generally assumed to "end in a noun", and are treated as compound nouns in apposition on this base. (Not because they would be "noun phrases" in the sense of the Glossary.) In actual Latin, relatively few words are formed this way : contrary to what we do in nomenclature, in their -i- version at least, they are primarily adjectives, although sometimes usage as a substantive (i.e., with some untold noun understood) can occur secondarily. Cauda is not very illustrative as an example here, as few actual Latin words are derived from it. Consider, instead, e.g. : coma, which is Latin for the hair of the head -- more of an "everyday-life concept" (in a transformed way, it can also mean the foliage of a tree), thus yielding more actual derivatives : coma, as such, is a feminine noun ; not a single noun formed by adding something ahead of -coma exists in dictionaries ; comatus, -a, -um is an adjective meaning haired, which will end in -comata if modifying a feminine noun ; not a single word formed by adding something ahead of -comatus exists in dictionaries ; albicomus, alticomus, anguicomus, auricomus, flammicomus, flavicomus, floricomus, frondicomus, ignicomus, lauricomus, mollicomus, multicomus, vernicomus, viticomus are all adjectives derived directly from coma, meaning [something]-haired, [something]-crowned, or [something]-leaved, which will end in -coma if they modify a noun which is feminine. None of these words is known to have been used as a noun. In fact, two words would normally always be used when the non-adjectival meaning is intended (e.g. white hair, alba coma) ; as a result, from the moment that you use a single compound word (white-haired, albicoma), it is automatically adjectival, and adding a suffix ("albicomata") would be redundant, hence is not normally done. For an example of such a word turned secondarily into a substantive : capillus is a masculine noun, and another word for the hair of the head ; atricapillus, -a, -um, black-haired, is an adjective found in Latin dictionaries ; atricapilla [with avis, bird, feminine, understood] is the blackcap, a kind of bird.
In the group I am familiar to, names that would be noun phrases in the sense of the Glossary are at best rare, and far between. To identify such a name meaningfully, I'm afraid that you must understand, and take into account, all the elements that compose it ; looking at the final part only will never be sufficient.
(NB - Many more names, such as sanctijohannis, sanctaecatharinae, novaehollandiae, terraesanctae, gulielmitertii, etc. are in fact noun phrases in the standard grammatical sense of the term. But these names are noun phrases in the genitive, and the Glossary appears to limit the concept to names that are "treated as a noun in apposition" : they do not seem to qualify as noun phrases in the sense of the Glossary. Note that the last two end in an adjective -- resp. sanctus, holy, and tertius, third.)
For what it's worth,
Laurent -
Douglas Yanega
Hi Doug,
I have long been (and still remains) highly uncomfortable with this reading of the Glossary definition of "noun phrase"...
In common parlance, a "noun phrase" is a group of words, formed from a "head noun", to which are added various possible types of modifiers, which as a whole denotes a specific entity or concept, and acts as a single noun in a sentence.
- The Glossary says :
- noun phrase, n.
- A compound word consisting of a noun combined with another noun or modifying adjective, the compound being treated as a noun in apposition; if the adjective is the final element in a species-group name, its ending is determined by the gender of the noun it modifies (and not by that of the generic name with which the species-group name is combined). For examples, see Article 31.2.1.
In Art. 31.2.1, we find : "the noun phrases in Melanoplus femurrubrum (Melanoplus masculine; but rubrum agreeing with femur, neuter) and Desmometopa m-nigrum (Desmometopa feminine; nigrum neuter, agreeing with m, because letters of the alphabet are neuter)".
I find it extremely disappointing and unhelpful that these are the only examples in the actual text of the Code, because in both examples, the adjective comes last. In 99% of all the noun phrases used in zoology, the adjective comes first.
You and I are both fully aware that the Code's definition is not what linguists and grammaticians would agree with.
That said:
albacauda = alba cauda, with alba = albus, -a, -um, white (in the nominative singular feminine, qualifying cauda) and cauda = tail (the head noun, nominative singular feminine), would be an acceptable noun phrase.
albicauda / albocauda is certainly not a noun phrase in the sense of the Glossary -- even if, in nomenclature, such names are indeed nowadays generally assumed to "end in a noun", and are treated as compound nouns in apposition on this base.
This is precisely why the Code's definition has to be taken at face value. To inform members of the taxonomic community that the connecting vowel (or a variant ending letter) determines whether a word is a noun or an adjective is far, far too esoteric to be of any value to a person who simply wants to know whether or not they need to change the termination of a species name to make it correct.
The "default" should, as often as possible, be to treat as a noun (e.g., ICZN Art. 31.2.2). As such, albacauda, albicauda, albocauda, albacaudus, albicaudus, albocaudus, albacaudum, albicaudum, and albocaudum are all noun phrases under the Glossary definition. It should not be possible (and I argue that it is NOT possible) to have multiple compound names with the same exact ending word (i.e., with the same definition, not similar by coincidence, and not requiring literally identical spelling), some of which are nouns and some of which are adjectives. To insist otherwise is to add an entire level of complexity that the Code neither requires, nor benefits from, and to no one in the community's benefit. It is (and should be) as simple as this: any name that ends in a noun (including a spelling variant of a noun, like "caudus") IS a noun, and any name that ends in an adjective IS an adjective (so, nigritarsum is a noun, but nigritarsis an adjective). Even that rule, as simple as it is, can be challenging to apply, and - despite some common misconceptions - the goal of the Code is not to make things MORE challenging.
Consider, instead, e.g. : coma, which is Latin for the hair of the head -- more of an "everyday-life concept" (in a transformed way, it can also mean the foliage of a tree), thus yielding more actual derivatives : coma, as such, is a feminine noun ; not a single noun formed by adding something ahead of -coma exists in dictionaries ; comatus, -a, -um is an adjective meaning haired, which will end in -comata if modifying a feminine noun ; not a single word formed by adding something ahead of -comatus exists in dictionaries ; albicomus, alticomus, anguicomus, auricomus, flammicomus, flavicomus, floricomus, frondicomus, ignicomus, lauricomus, mollicomus, multicomus, vernicomus, viticomus are all adjectives derived directly from coma, meaning [something]-haired, [something]-crowned, or [something]-leaved, which will end in -coma if they modify a noun which is feminine.
The rules and attested uses of Latin are good to know, but also not strictly relevant. Scientific names are not actual Latin words, and do not need to be. The rules governing them are NOT the rules of Latin grammar. Everything you say is correct in the context of Latin, but the discussion is how the Code insists these words are to be treated when used as scientific names. Under the Code, names ending in "-comata" are adjectives, and names ending in "-coma" (or comus or comum) are not. This is entirely sensible and practical, regardless of what Latin scholars might think about it.
None of these words is known to have been used as a noun.
Not in Latin, perhaps, but several taxonomists have used such names as nouns. As such, even in the narrower interpretation, all such names fall under 31.2.2 and can only be treated as adjectives if the original description explicitly stated they were intended to be adjectives.
NB: if you wish to argue that albicomus is an adjective but albicoma is not, then what happens when a species named albicomus migrates to a feminine genus? It changes to albicoma, right? But what happens if it is then moved to a different masculine genus? If your interpretation applies, it would be permanently stuck as albicoma. All transitions from a putatively adjectival ending to a putatively noun ending would be irreversible - "one way trips". This is very certainly NOT what the Code is intended to do.
NB: Among the most well-known species of bumblebees are Bombus campestris, Bombus terrestris, and Bombus sylvestris. However, each of these was originally named in the genus Apis, a feminine genus. According to Latin scholars, the correct masculine forms are campester (https://en.wiktionary.org/wiki/campester), terrester (https://en.wiktionary.org/wiki/terrester) and sylvester (https://en.wiktionary.org/wiki/sylvester). This rule of Latin grammar is not employed by the Code, either - it would be very, very confusing, because the spellings campestris, terrestris, and sylvestris are also allowed to be masculine in Latin. The more one learns about Latin grammatical rules, the more obvious it is that the Code should steer AWAY from them, rather than more stringently embracing them. You want to see taxonomists pulling their hair out? Just ask them what the neuter forms of the species names "major" and "minor" are, for another example. People already hate gender agreement, and we don't want that situation to get any worse than it already is.
The history of nomenclature is long, complex, and filled with a parade of taxonomists whose skill in the application of Latin grammar has ranged from stellar to abysmal. Faced with such a legacy, a policy that accepts only the "stellar" end of the spectrum is not going to work very well. It should not require a degree in linguistics to determine how to correctly spell a name under the Code (especially since the authors of names sometimes misspelled them). Again, the Code's rules are a deliberate oversimplification, to facilitate the process. It is MORE desirable to pare down those rules even more, and make them even simpler, than to do as you suggest, especially in the face of ever-increasing pressure from the community to abandon gender agreement entirely. I am fighting, tooth and nail, to salvage the practice by making it so easy to use that it can be automated; if we do NOT simplify it, then I'm pretty certain that it WILL be abandoned, and that will be disastrous. I have spent several years now researching this in excruciating detail, and have a file open on my computer desktop, every single day, displaying every zoological species name in the GBIF master file (over 3 million entries; the largest such resource available, as far as I am aware).
If we abandon gender agreement, something near or exceeding 15% of all names in zoology will have to change spelling, when they revert to the original spelling. The older the names, the higher the percentage, so Linnaean and Fabrician names (for some of the most common species on the planet) will disproportionately be affected; very, very few of these are still in their original genera. Note also that in order for this approach to work, we must have access to every single original description, and must compile a list of all the species names and their correct spellings. We do not presently have a complete list of every species name ever published.
If, instead, we simplify the rules of gender agreement to the point where automation is practical and universal, only around 2% of all names will have to change spelling. This approach does NOT require us to compile a list of every species name ever published, nor have access to every original description. We could implement this almost immediately, if people support the idea. Several disciplines already DO support it. I have already helped implement "proof of concept" examples, for collaborators associated with TaxonWorks, and Species Files. These early examples cover well over 30,000 species names (several orders and superfamilies of insects), and the protocol is not difficult to navigate. Nearly 95% of all genera and species can be immediately and unambiguously categorized (the former as to gender, the latter as to declinability) without needing to see the original descriptions at all. The difficulty lies entirely within that remaining 5%. That is quite manageable, and the best part is that no one will ever need to argue about Latin grammar again. ;-)
laurent raty
Doug,
On 6/26/26 01:21, Douglas Yanega wrote:
I find it extremely disappointing and unhelpful that these are the only examples in the actual text of the Code, because in both examples, the adjective comes last. In 99% of all the noun phrases used in zoology, the adjective comes first.
I disagree with the last sentence. (You evidently did not get the main point - what made me feel uncomfortable - in my previous post.)
In an adjective-noun "noun phrase" as defined in the Glossary, the adjective modifies a head noun in a phrase : both elements must agree in case and gender with one another. The main peculiarity of such noun phrases is, as indicated by the Glossary, that : "the compound [is] treated as a noun in apposition; if the adjective is the final element in a species-group name, its ending is determined by the gender of the noun it modifies (and not by that of the generic name with which the species-group name is combined)." These names are treated as nouns in apposition even when their final element is an adjective ; their true nature can only be identified correctly by considering them in their globality. This is fully explicit, and runs directly in the face of any claim that the nature of a name under the current Code can be determined just by looking at its final element.
My main point was that 99% of the names that you call "noun phrases" are, actually, names that start with an adjectival stem without any case ending (typically united to a second element by a connecting vowel) ; they are not names that start with an adjective agreeing in case and gender with a "head noun" that would follow it, both words together forming a "phrase" : these 99% are certainly compound words, but I do not believe that they can be called "noun phrases" in any meaningful sense of the term. I would not regard anything that the Code could tell us about "noun phrases" as potentially concerning them.
(The examples were evidently selected to illustrate the peculiarity of "real" noun phrases ending in an adjective.)
The "default" should, as often as possible, be to treat as a noun (e.g., ICZN Art. 31.2.2).
As such, albacauda, albicauda, albocauda, albacaudus, albicaudus, albocaudus, albacaudum, albicaudum, and albocaudum are all noun phrases under the Glossary definition. It should not be possible (and I argue that it is NOT possible) to have multiple compound names with the same exact ending word (i.e., with the same definition, not similar by coincidence, and not requiring literally identical spelling), some of which are nouns and some of which are adjectives.
Of these, I would regard only albacauda (i.e., alba cauda, in which alba is an adjective that agrees in case and gender with the noun cauda, which it modifies) as an uncontroversial noun phrase.
As I wrote above, the Glossary definition, in the current Code, is explicit that a noun phrase ending in an adjective is a noun. On the other hand, a name ending in the very same adjective, but which is not a noun phrase, will be an adjective. Perhaps this type of thing should ideally not be possible, but it pretty much seems unavoidable to me that, currently, it is possible.
To insist otherwise is to add an entire level of complexity that the Code neither requires, nor benefits from, and to no one in the community's benefit. It is (and should be) as simple as this: any name that ends in a noun (including a spelling variant of a noun, like "caudus") IS a noun, and any name that ends in an adjective IS an adjective (so, nigritarsum is a noun, but nigritarsis an adjective). Even that rule, as simple as it is, can be challenging to apply, and - despite some common misconceptions - the goal of the Code is not to make things MORE challenging.
Not in Latin, perhaps, but several taxonomists have used such names as nouns. As such, even in the narrower interpretation, all such names fall under 31.2.2 and can only be treated as adjectives if the original description explicitly stated they were intended to be adjectives.
NB: if you wish to argue that albicomus is an adjective but albicoma is not, then what happens when a species named albicomus migrates to a feminine genus? It changes to albicoma, right? But what happens if it is then moved to a different masculine genus? If your interpretation applies, it would be permanently stuck as albicoma. All transitions from a putatively adjectival ending to a putatively noun ending would be irreversible - "one way trips". This is very certainly NOT what the Code is intended to do.
To "argue" that something is, or not, an adjective is not what I'm trying to do here -- I'm just reporting what I find in Latin dictionaries, and exploring the possible application of the current Code to this. Here :
- In Latin dictionaries, albicoma occurs exclusively as an inflected form of albicomus, -a, -um, which is given exclusively as an adjective meaning white-haired. My inclination would be not to argue against congruent information provided by multiple dictionaries, unless forced to do so. (But I have little doubt some will succeed making up a convoluted interpretation that will turn this word into "a noun".)
- The word that can have both natures as per Latin dictionaries, is atricapilla : this occurs in dictionaries both as an inflected form of atricapillus, -a, -um, which is given as an adjective meaning black-haired, and as the name of a bird ; atricapillus and atricapillum, on the other hand, occur exclusively as inflections of the adjective. In some dictionaries, e.g. https://archive.org/details/ausfhrlichesun01sche/page/524/mode/1up , the use of the word for the bird is listed within the entry for the adjective.
NB: Among the most well-known species of bumblebees are Bombus campestris, Bombus terrestris, and Bombus sylvestris. However, each of these was originally named in the genus Apis, a feminine genus. According to Latin scholars, the correct masculine forms are campester (https://en.wiktionary.org/wiki/campester), terrester (https://en.wiktionary.org/wiki/terrester) and sylvester (https://en.wiktionary.org/wiki/sylvester). This rule of Latin grammar is not employed by the Code, either - it would be very, very confusing, because the spellings campestris, terrestris, and sylvestris are also allowed to be masculine in Latin. The more one learns about Latin grammatical rules, the more obvious it is that the Code should steer AWAY from them, rather than more stringently embracing them. You want to see taxonomists pulling their hair out? Just ask them what the neuter forms of the species names "major" and "minor" are, for another example.
The problem with issues like terrestris vs. terrester stems from the spelling of the name not being completely specified in the OD. I.e., a feminine OS in -is can imply an adjective in -er, -is, -e, or one in -is, -is, -e. Note that this type of issue is not limited to gender agreement -- it also exists with names introduced in a Latin text in an inflected form that does not correspond to a unique nominative. E.g., generic names introduced in the accusative case with an -um ending, which could become either -us or -um in the nominative. (The present Code offers no solution so far as I'm aware -- perhaps some kind of FR-like action could be envisioned ?)
BTW, can major not be a noun...? (The plural, "majores", in any case, has certainly been used as a substantive, meaning either "adults" (as opposed to "children"), or "ancestors".)
People already hate gender agreement, and we don't want that situation to get any worse than it already is.
The history of nomenclature is long, complex, and filled with a parade of taxonomists whose skill in the application of Latin grammar has ranged from stellar to abysmal. Faced with such a legacy, a policy that accepts only the "stellar" end of the spectrum is not going to work very well. It should not require a degree in linguistics to determine how to correctly spell a name under the Code (especially since the authors of names sometimes misspelled them). Again, the Code's rules are a deliberate oversimplification, to facilitate the process. It is MORE desirable to pare down those rules even more, and make them even simpler, than to do as you suggest, especially in the face of ever-increasing pressure from the community to abandon gender agreement entirely. I am fighting, tooth and nail, to salvage the practice by making it so easy to use that it can be automated; if we do NOT simplify it, then I'm pretty certain that it WILL be abandoned, and that will be disastrous. I have spent several years now researching this in excruciating detail, and have a file open on my computer desktop, every single day, displaying every zoological species name in the GBIF master file (over 3 million entries; the largest such resource available, as far as I am aware).
If we abandon gender agreement, something near or exceeding 15% of all names in zoology will have to change spelling, when they revert to the original spelling. The older the names, the higher the percentage, so Linnaean and Fabrician names (for some of the most common species on the planet) will disproportionately be affected; very, very few of these are still in their original genera. Note also that in order for this approach to work, we must have access to every single original description, and must compile a list of all the species names and their correct spellings. We do not presently have a complete list of every species name ever published.
If, instead, we simplify the rules of gender agreement to the point where automation is practical and universal, only around 2% of all names will have to change spelling. This approach does NOT require us to compile a list of every species name ever published, nor have access to every original description. We could implement this almost immediately, if people support the idea. Several disciplines already DO support it. I have already helped implement "proof of concept" examples, for collaborators associated with TaxonWorks, and Species Files. These early examples cover well over 30,000 species names (several orders and superfamilies of insects), and the protocol is not difficult to navigate. Nearly 95% of all genera and species can be immediately and unambiguously categorized (the former as to gender, the latter as to declinability) without needing to see the original descriptions at all. The difficulty lies entirely within that remaining 5%. That is quite manageable, and the best part is that no one will ever need to argue about Latin grammar again. ;-)
Again, I can understand the sentiment, but I have a very hard time reading the current Code as supporting the suggested "solutions". Beyond the definition of noun phrase making it explicit that a name ending in an adjective will be an adjective only if it is not a noun phrase, which such a name can be, other provisions of the current Code stand in the way of any automation, such as the possibility of opt-out which is offered in the current Art. 26. (I can remember a couple of cases of an author using a name that should have been an adjective, but deliberately declaring it an arbitrary combination of letters.)
Douglas Yanega
Doug,
On 6/26/26 01:21, Douglas Yanega wrote:
I find it extremely disappointing and unhelpful that these are the only examples in the actual text of the Code, because in both examples, the adjective comes last. In 99% of all the noun phrases used in zoology, the adjective comes first.
I disagree with the last sentence. (You evidently did not get the main point - what made me feel uncomfortable - in my previous post.)
In an adjective-noun "noun phrase" as defined in the Glossary, the adjective modifies a head noun in a phrase : both elements must agree in case and gender with one another.
Please note the bolded wording:
"A compound word consisting of a noun combined with another noun or modifying adjective, the compound being treated as a noun in apposition; if the adjective is the final element in a species-group name, its ending is determined by the gender of the noun it modifies (and not by that of the generic name with which the species-group name is combined)"
It says "IF the adjective is the final element in a species-group name" - meaning that it does not have to be. It can be the first element, and in 99% of the cases fitting the definition, it IS the first element. The definition is, strictly, "a noun combined with another noun or modifying adjective" - this does not require them to be in any particular order, which seems to be the core of your argument here. There are only around 200 names in all the history of zoology where the adjective is the final element of a compound.
This is fully explicit, and runs directly in the face of any claim that the nature of a name under the current Code can be determined just by looking at its final element.
This is only true for those incredibly rare cases where the adjective comes last, which I did not dispute previously (but I do actually think it's a bad rule). What I said is that it makes no sense, and only confuses taxonomists.
This confusion is demonstrable: an assay of all of the correctly-formed names like "x-nigrum" that have been used in zoology reveals that when such names have been moved from one genus to the next, over 50% of the time they have erroneously been treated as adjectival, and given changed spellings - in direct violation of the Code. Almost 20% of such names were formed incorrectly by the original author, as well - and, more problematic, in the majority of names of this particular form (a letter followed by a hyphen followed by a word), the final element is a passive participle, referring to the genus and not the letter of the alphabet, so they agree with the genus name and are not neuter. In many others the final element is itself a noun, and also not neuter. Examples of the former are are x-maculata, v-carinata, v-notatus, x-fasciatus, x-litterata, s-signata, x-ornatus, and of the latter, w-scripta, t-liturus, s-littera, v-macula.
Taxonomists generally do not appear to know, or care, what the rules are regarding these names. As this discussion demonstrates, it is inordinately difficult to explain exactly when (and when not) to treat the word after a hyphen as an indeclinable neuter adjective. It is vastly simpler if we accept the de facto treatment by the community that all of these names ending in adjectives are declinable, even "x-nigrum" and "femurrubrum". Is treating these as adjectives a grammatical abomination? Yes. Do taxonomists commit grammatical abominations like this? Yes, a LOT. Can we prevent them from doing this? No.
If we can't prevent it, then allowing it seems like a concession we will need to make for the sake of keeping things simple. In this one very special circumstance - noun phrases where the adjective comes last - I freely admit that what I advocate violates the Code. But that is NOT true for my saying that noun phrases ending in a noun are nouns - that IS what the Code says.
these 99% are certainly compound words, but I do not believe that they can be called "noun phrases" in any meaningful sense of the term. I would not regard anything that the Code could tell us about "noun phrases" as potentially concerning them.
And I regard the Code's definition of these as noun phrases to be binding, not just because it's the rule, but because it makes sense to treat a name ending in a noun as a noun.
To "argue" that something is, or not, an adjective is not what I'm trying to do here -- I'm just reporting what I find in Latin dictionaries, and exploring the possible application of the current Code to this. Here :
- In Latin dictionaries, albicoma occurs exclusively as an inflected form of albicomus, -a, -um, which is given exclusively as an adjective meaning white-haired. My inclination would be not to argue against congruent information provided by multiple dictionaries, unless forced to do so. (But I have little doubt some will succeed making up a convoluted interpretation that will turn this word into "a noun".)
Again, the Code does not give primacy to what the dictionary says. It says that a noun phrase is "a noun combined with another noun or modifying adjective" - and albicoma is precisely that.
Of course, "one-way trips" as you describe are not possible. If a species is named, say, atricapillus, and if this is accepted as adjectival as per dictionaries, this name is to be understood as being the adjective atricapillus, -a, -um, which "must agree in gender with the generic name with which it is at any time combined" (31.2) ; if the species gets moved to a feminine genus, its name changes to atricapilla, the feminine form of the adjective (this does not turn the adjective into a noun) ; if it is then returned to a masculine genus (be it the same as the original one or not), the name becomes atricapillus again.
You assume that all taxonomists, in all disciplines, will have access to the original descriptions of every name in every discipline. This is not so, and probably never will be so. If I give you a list of 200 names like this, you won't have any way of knowing which ones to treat as adjectives unless you track down all the original descriptions and CONTINUE to keep track of these, in perpetuity, so you know which ones change and which ones don't. It's a lot easier if you declare that NONE of them change. We live in an era when an awful lot of taxonomy and nomenclature is handled "in bulk" rather than in small does, and only by specialists. This puts a premium on having rules that can be applied in an automated fashion, not one name at a time, with rare scattered exceptions.
You want to see taxonomists pulling their hair out? Just ask them what the neuter forms of the species names "major" and "minor" are, for another example.
BTW, can major not be a noun...? (The plural, "majores", in any case, has certainly been used as a substantive, meaning either "adults" (as opposed to "children"), or "ancestors".)
My point was that few taxonomists know that the neuter forms of "major" and "minor" are "majus" and "minus". In the history of zoology, when a species with one of these epithets has been moved into a neuter genus, the name has not changed spelling in 2/3rds of those cases - 61 times incorrect versus 23 times correct. That's an appalling error rate, suggesting that taxonomists simply don't know the correct way to treat these names. Not surprisingly, the error rate on adjectives like "grandior" (which has "grandius" as its neuter form) is even HIGHER than this. Again, asking taxonomists to learn and apply the rules of Latin grammar is an increasingly difficult expectation to maintain - and the community of taxonomists is increasingly annoyed by it. We can maintain gender agreement without having to adhere to all of the esoteric rules of Latin grammar; ease of use is more important than linguistic accuracy, and a compromise approach is beneficial. That is especially true given how incredibly rare names like this actually are; a few hundred out of several million. It's like having a special exceptional rule that applies only to people born in an airplane over international waters; people like this do exist, but it would be rather troublesome to keep track of this, and treat them all differently from everyone else.
Again, I can understand the sentiment, but I have a very hard time reading the current Code as supporting the suggested "solutions".
I have not suggested much at all that deviates from what is presently in the Code. And, just to make sure this point is not being overlooked, those cases where I do advocate a deviation from the Code are the RAREST of all categories, such as hyphenated compounds. Out of over 3 million names in zoology, hyphenated compounds ending in adjectives account for 124 valid cases, and less than half of these are treated as noun phrases under the Code. I don't think many people are going to care if we change how we treat 50 or so valid noun phrase names, by treating them as adjectives. That is a trivial price to pay in order to achieve a general rule that everyone, including an AI algorithm, can understand and apply without error.
provisions of the current Code stand in the way of any automation, such as the possibility of opt-out which is offered in the current Art. 26. (I can remember a couple of cases of an author using a name that should have been an adjective, but deliberately declaring it an arbitrary combination of letters.)
It seems that you and I look at the same exact things, but perceive very different ways to address them: I think that we should simplify things that cause grief for taxonomists, regardless of the linguistic appropriateness, while you advocate that we adhere to linguistic principles, regardless of the effect on taxonomists. In an objective analysis, over the past 270 years, taxonomists have collectively gotten worse at adhering to linguistic principles, and less happy about it, and I don't think that trend is going to reverse. As such, a reassessment of our expectations and policies is long overdue, and I don't believe that we benefit from trying to compel taxonomists to learn and adhere to each and every linguistic principle; a compromise is best, in my opinion (and the Code is, to some extent, already a simplification, and a compromise), because abandoning gender agreement entirely would be a genuine disaster, and we are on the verge of having that happen.
We are finally in a position, with the emergence of taxonomic data management tools like TaxonWorks, to automate gender agreement. It would be a shame to fail to take advantage of this, and instead watch the proverbial mob, with pitchforks and torches, undo 270 years' worth of a demonstrably useful practice.
laurent raty
On 6/27/26 20:26, Douglas Yanega wrote:
- In Latin dictionaries, albicoma occurs exclusively as an inflected form of albicomus, -a, -um, which is given exclusively as an adjective meaning white-haired. My inclination would be not to argue against congruent information provided by multiple dictionaries, unless forced to do so. (But I have little doubt some will succeed making up a convoluted interpretation that will turn this word into "a noun".)
Again, the Code does not give primacy to what the dictionary says. It says that a noun phrase is "a noun combined with another noun or modifying adjective" - and albicoma is precisely that.
I guess I must sound like a broken record, and I apologize if I do. But no, it is not.
An "adjective" is a word. In "alb-i-coma", "alb-" is merely a stem (not a word), and "albi-", with the connecting vowel included, acts as a prefix (a so-called "learned prefix" ; not a word either) : albicoma does not meet the Glossary definition of "noun phrase" because it includes at most one noun, and no "modifying adjective" as such. (In fact, it could even be argued that albicoma also fails to match the Glossary definition of "compound word" - which "noun phrases" are a subclass of - because this definition expressly excludes prefixes.) This reading aligns the Glossary definition of "noun phrase" with standard grammatical terminology in which a "phrase" is indeed always a group of words, not a single word created by combining stems ; your reading accepts "noun phrases" which are evidently not "phrases" at all. In my reading, the main peculiarity which a nomenclatural context should bring to "noun phrases" is that, as is the case in any name formed from more than one word, "the component words" that make up a name that is a "noun phrase" "are to be united" (cf. Art. 32.5.2.) ; in standard written language, the component words that make up a noun phrase are always separated by word spaces.
The reason why the Code provides no example of noun phrase that match albicoma, its that albicoma is not at all formed in a way that could make it a noun phrase. Real noun phrases are very infrequent in scientific names, whatever the position of the head noun.
The Glossary defines "Latin" as ancient and mediaeval Latin, and this is binding, too. This makes that, e.g., "a Latin adjective", in the context of the Code, is by definition equal to what was treated as an adjective in ancient and mediaeval Latin. And the only authorities on this are Latin dictionaries and grammars.
Sorry again,
Laurent -
Douglas Yanega
The reason why the Code provides no example of noun phrase that match albicoma, its that albicoma is not at all formed in a way that could make it a noun phrase. Real noun phrases are very infrequent in scientific names, whatever the position of the head noun.
The Glossary defines "Latin" as ancient and mediaeval Latin, and this is binding, too. This makes that, e.g., "a Latin adjective", in the context of the Code, is by definition equal to what was treated as an adjective in ancient and mediaeval Latin. And the only authorities on this are Latin dictionaries and grammars.
Sorry again,
The Code does not indicate that the dictionary is the sole source of authority for the nature of a species epithet, and - if it did - then we would be facing an even more chaotic situation than we do under the status quo.
A quick extraction from GBIF regarding the suffix "-comus" gives over 100 such examples used in zoology.
Of these, only 11 are attested as Latin adjectives:
albicomus, anguicomus, auricomus, chrysocomus, flammicomus, flavicomus, floricomus, ignicomus, leucocomus, mollicomus, and vernicomus.
There are 91 more that are not attested at all in Latin dictionaries:
acanthocomus, acomus, aericomus, aetheocomus, albacomus, albocomus, altercomus, amphicomus, aphelocomus, aracomus, argenticomus, argyrocomus, atricomus, aurantilicomus, aureocomus, aurocomus, bathycomus, bicomus, bellicomus, brevicomus, callicomus, campycomus, chalcocomus, chalcomus, chlorocomus, compsocomus, crassicomus, curticomus, dasycomus, dicomus, erythrocomus, eucomus, flavacomus, flavocomus, fulvicomus, fusicomus, glaucomus, habrocomus, glauxicomus, heterocomus, hirticomus, horricomus, lenicomus, leptocomus, leucomus, longicomus, microcomus, microfloricomus, millicomus, minimicomus, nigracomus, nigricomus, nigrocomus, nitidicomus, nymphocomus, ochrocomus, octocomus, ophiseocomus, pachycomus, paraminimicomus, paramollicomus, paraschizocomus, parcicomus, parvaeucomus, pericomus, phyllocomus, plumicomus, pluricomus, polycomus, pseudauricomus, pseudocampycomus, pseudocomus, pseudoeucomus, pulchricomus, pyrrhocomus, quadricomus, quasimacomus, raricomus, rigidicomus, rubricomus, schizocomus, sericomus, sicilicomus, sparsicomus, tenericomus, tenuicomus, trichocomus, tricomus, ulocomus, viridicomus, and xanthocomus.
(There is also nosocomus, a noun for a nurse, which we can presumably agree to set aside).
I maintain that under the Code, none of these is declinable when used as a species name, even the 11 attested words. This is simple, even if you believe it is simple-minded. ;-)
Do I assume correctly that you would not recognize any names outside of those first 11 as being declinable, since none of the others appear as adjectives in a Latin dictionary? Or, to be more specific as to the question of importance, do you consider ALL of the 91 latter names to be declinable, NONE of them, or only SOME of them?
Several of the latter names have Greek prefixes (e.g. bathycomus, dasycomus, leucocomus, pachycomus, xanthocomus), and according to David & Gosselin (but not according to the Code), words composed of a mixture of Greek and Latin components must always be considered indeclinable because they are never attested, or in a dictionary.
I maintain that the Code avoids strict reliance on the dictionary for precisely the reason that is evident from this example:
Scientific names do not adhere strictly to the rules of Latin, or (especially) to the attested uses of Latin. Taxonomists create names of all types, forming tens of thousands of latinized epithets that never once appeared in Latin or Greek, and the Code must contend with all of them - good, bad, or unusual - in a coherent and consistent manner. I've never said that your research or conclusions are wrong regarding the linguistics; my argument is that this should not be taken as overriding the Code, especially given the consequences of doing so. Is it true that auricomus is an attested Latin adjective and leucocomus is not? Absolutely. But, the Code tells us to treat them both as indeclinable, regardless. Your insistence upon linguistic purity, and applying external definitions, incurs a very high price for thousands of taxonomists who, unlike you, do not enjoy linguistic research.
By considering all 102 of these epithets as indeclinable, the Code simplifies things to yield a coherent and consistent result - unless, as in your case, you challenge the principle that the text of the Code takes precedence, and insist on selectively treating some as declinable and others not, because this is linguistically correct. Under your interpretation, a taxonomist has no way of knowing, without digging into dictionaries, which of these names is declinable and which is not. That is not helpful, practical, or consistent. It is unfair to expect taxonomists to engage in linguistic research like this, especially when dictionaries themselves are not consistent or comprehensive - two different taxonomists could use two different dictionaries, and come to different conclusions (e.g., among three prominent Latin dictionaries - Lewis & Short, Georges, and Gaffiot - "pumilus" is treated solely as a noun in two of them, and as both a noun and an adjective in the other). Or are you proposing that the Code should specify WHICH dictionary to use?
One last note here: examination in GBIF of the original spellings and recombined spellings of all the species names published using one of the previous epithets or their variants reveals that between 10-20% were either treated as indeclinable by the coining authors, or by the taxonomists that recombined them when they changed genera (including names among the 11 attested). That indicates that there are a fair number of taxonomists who, as I do, consider these names to be indeclinable noun phrases. Neither you nor I are alone in our opinions, and that is the reason we are having this discussion. This also indicates the final issue: clearly, there are taxonomists (myself included) who consider names like these to be noun phrases under the Code's definition, while many others (like yourself) consider them to be adjectives. Under the strict interpretation of Article 31.2.2, any such disputed name is to be treated as indeclinable unless the coining author explicitly stated the name is adjectival. Such an outcome is even WORSE, because then not only would taxonomists need to check every name in a dictionary, but they would also need to have access to the original description of every name. We do not want to compel taxonomists to do MORE work, and we should take pity on those poor non-taxonomists who find two or three variants of a species name, and cannot tell which one is correct.
This is not an obscure, esoteric, or trivial distinction, between my view and yours - it affects tens of thousands of species names across zoology, and would have a profound impact on how easily taxonomists and non-taxonomists can determine the correct spellings of these names. I consider it unfortunate that lepidopterists do not comply with the Code's rules on gender agreement, and unfortunate that ornithologists have chosen to adopt their own set of rules (i.e., those of David & Gosselin) that are not in accordance with the Code.