Question about languages that use gendered nouns: what actually determines whether a noun is masc, fem or neut? When a new noun enters the language how does it get assigned a gender? Who decides and on what basis?
My recollection from spanish classes is that, at least in spanish, pretty much all new words take the masculine gender unless something about their formation suggests they would normally be feminine or the word is coming from another gendered language where it is feminine. Sometimes words will differ in gender regionally where one geographic area decides it is masculine and another decides it is feminine.
The final decision is made by what you might call the community of speakers who over time gravitate toward one particular formation and eventually that becomes the standard and the other versions either die off or are relegated to dialectical status.
Guy Deutscher relates a story in one of his books about how an aboriginal language marked airplanes as belong to the noun class for edible vegetables. Why? Because at some point, the class was widened to include other plants such as trees. The natives made their canoes out of trees so it seemed reasonable to include canoes in there as well. When they learned about airplanes, well, what is an airplane but a canoe in the sky? So clearly it belongs in there as well. But you can imagine if this language sticks around for another half a century, people could lose the connection betwene all of them and just have a noun class that includes seemingly at random vegetables, trees, canoes and airplanes.
Custom decides. If the term is linked to previous ones, or if it had an assigned gender in the original language, then that is one of the factors to consider.
Spanish calls computers computadora, short for máquina computadora (computing machine); máquina is fem so the adjective gets the feminine ending and once the noun is ellided, the new noun stays feminine. We also call them ordenador, which comes from the French ordinateur: masculine in both languages.
It depends on the language. You might take a look at the “Gender assignment” section in the Wikipedia article. In some languages, the gender almost exactly matches the actual gender of the referent – that is, words for masculine people and animals have the masculine gender, female people and animals have the feminine gender and everything else is neuter. But wiki says this is actually rare. In most languages the gender of a word doesn’t exactly correlate to the sex of the referent, although there is usually some degree of correlation. For example, in Latin, the word for “farmer” is agricola which is feminine even though most Roman farmers were men.
Also note that linguistically, “gender” does not always equate to “sex”. Some languages have genders that are based on, for example, animacy (animate objects regardless of sex are one gender, while inanimate objects are another gender).
You are misremembering your Latin. Agricola is a noun of the first declension, of which the vast majority are feminine, but it is one of the exceptions along with a number of other traditionally male occupations (such as poeta (poet) and nauta (sailor)).
For the main question, one factor in addition to the ones already mentioned is euphony. Depending on the rules of the language, one gender may simply lead to less awkward combinations of sounds than another when applied to a loanword.
Doh! You’re quite right, thanks for the correction.
I believe that in French new words are almost always masculine. This is the default. One exception I am aware of was because the word had a feminine cognate. So why was the word borrowed at all? Because it was a proper noun and thus invariable. When I used to go to hockey game in the early 70s the tickets were issued by “La Canadian Arena Company”. The cognate “compagne” is feminine. Note that the hotel downtown is called “Le Reine Elizabeth” because the implied “hotel” is masculine. Similarly for the French boat “Le France”.
You mean compagnie (also feminine) or société (ditto).
I would like a reference for this all new words in French are masculine phenomenon. Why should this be so, rather than follow the natural or customary gender, as in Spanish?
The distinction is relevant in, for instance, adjective-noun agreement. So a “good farmer” would be “bonus agricola” (agreeing in gender), not “bona agricola” (agreeing in declension). But a “good lady” would be “bona femina”, because “femina” is (unsurprisingly) feminine.
What if a new word is determined to be more feminine by Spain, but not so much by Mexico? With all of the many Spanish speaking countries there are in the world, has such a situation ever occurred?
IME, but IANALinguist, what’s more likely to happen is that we end up with different words by dialect. Ordenador is more common in Spain, computadora in Latin America.
I’ve encountered people who insisted in using what everybody else including DRAE agreed was the wrong gender, generally by treating the word as something that became a noun by ellision of something else, rather than having always being a noun. Example: a few days ago I was watching a movie in which someone said “gracias por el margarita”… ¿EL margarita? ¡AAAG! Margarita is feminine damnyou! The translator had treated “margarita” as short for “cóctel margarita”, and since cóctel is (m), he’d treated margarita as (m). FTR, DRAE agrees with me that margarita is always (f), definition 7.
Another example: maquila is (f) and means “subcontracted industrial work” or “factory which subcontracts for another” (DRAE really needs to update those particular definitions). It is mainly used in Northern Spain (from Cantabria to Aragon both included) and in those parts of Latin America which end diminutives in -ico (a characteristic of the Spanish of Northern Spain). I’ve had coworkers who had not encountered the word before and who insisted in saying el maquila because they claimed it was short for el proceso maquila, proceso being (m). The word is found by the first time in a medieval document in Arabic - you guys think them 10th-century Arab-speaking Riojanos used the word “proceso” in a subcontracting contract?
But wouldn’t a female farmer be bona agricola? Sometimes a word can be treated as either gender without changing its own forms.
In Hebrew, I think the rule of thumb is that a word is male - unless it sounds like a female noun, usually by ending with an open ‘a’ or with an ‘it’. Thus, for instance, “Telephone” (טלפון) is male, while “Encyclopedia” (אנציקלופדיה) is female.
There are Latin pairs like servus (m) / serva (f), but I think agricola is always masculine. Since all classical Latin texts have been digitized, perhaps someone can do a search to confirm this?
If (this is a big if) ordenador is cognate to Latin ordinator (m), does anyone use the analogue of Latin computator (m)?
Hmm; one could argue that those two specific words were simply borrowed into Hebrew as-is, along with their pre-existing genders.
ETA: on the other hand, maybe not, as “telephone” is more of a neuter, which is not an option in Hebrew.
Ordenador and ordinator are related in that both mean “it who puts things in order” and derive from related verbs (think alphabetizing, or ordering a series of numbers, rather than any other kind of ordering).
But computadora comes from English and I’ve never encountered its male version to refer to a machine; the verb computar exists but it’s rarely used, its synonyms calcular(1) and contabilizar(2) are much more frequent. The verb computar is mainly encountered in Latin America, in IT contexts in which people are trying to avoid repetition (in Spanish and except when used for emphasis, repetition is considered bad to the point of apologizing for it), or in very formal documents such as regulations.
1: leads to calculador(a) to refer to people, and calculadora to refer to a calculator
2: leads to contable for accountant, both genders are identical
…and the root φωνή itself is feminine, so it is not so straightforward.
Some Norwegian dialects only have common gender and neuter, where common corresponds to the masculine in the other dialects, and Norwegian in general seems to correspond to the only rule mentioned so far, new loan words these days are masculine. At least I can’t think of any recent counterexample.
It’s not historically correct though. “Helikopter” is for instance obviously not originally a Norwegian word, and today at least it is neuter.
How this is decided? By “feeling” to the first users and then by whichever form gets the wider usage.
I’m not a linguist. But I’ve often wondered if this is a common situation. A lot of words do not have an obvious gender; is a spoon, for example, masculine or feminine?
So my theory is that in many cases, words are assigned a gender because they sound similar to another word that has an established gender. Say you have a language where the word for father is vader and the word for water is vater. A vader is obviously masculine so you’ll say el vader. Water does not have an inherent gender. But if you’re used to saying el vader then el vater is going to sound more natural than la vater.
Only el váter happens to be the toilet… you seem to be mixing German and Spanish bigtime there.
water is neuter in German 
So far we have listed in this thread
-
Assignment by obvious noun class (airplanes are vehicles so have the same gender as canoes; plutonium takes the metals gender)
-
Constructions like máquina computadora
-
Loanwords (keep the original gender where possible)
-
Sounds like an existing word/declension
-
Lumped into a default miscellaneous gender (eg masculine)
Any data on the relative frequency of these?