18

Imagine if every French speaker suddenly agreed that nouns were one of 'animate' and 'inanimate', or 'chocolate' and 'strawberry', or 'A' and 'B' instead of 'masculine' and 'feminine'. The language could go on being used identically to how it was before.

Given this it's not obvious why their grammatical genders are 'masculine' and 'feminine' in the first place. Are these just arbitrary labels applied by ancient linguists to the different noun forms after they came into use, or were the words we use for different grammatical genders generated naturally (and perhaps obviously, to them) by people who spoke the language?

AML
  • 291
  • 2
  • 5
  • 4
    Note that in the Northeast Caucasian (Nakh-Dagestanian) languages a similar system of genders (usually 3–6 of them) is called ‘classes’ and they are named by the consonant used as an agreement affix: “y-class, b-class, d-class”, etc. Also, in the Bantu languages where there are up to a dozen classes, they are named simply by numbers, “class 1, class 7”, etc. – Yellow Sky Jul 10 '22 at 06:01
  • 11
    It is in fact entirely obvious why it's masculine and feminine: it's because grammatically it behaves exactly like a man or a woman subject would in its place. It requires the same conjucations and the same pronouns and shares the same morphology (eg word endings). – Roman Starkov Jul 10 '22 at 13:06
  • The etymology of mās is ultimately uncertain, far as I know. In this view, it seems the question pertains to the Latin stack (incl. Greek), inasmuch as there may be history behind the grammatical nomenclature. – vectory Jul 11 '22 at 08:06
  • @Vectory What did I miss, there? Where did "mās" come in, and is that "mās" as in various central European tongues, or "mas" from Japanese, or what? – Robbie Goodwin Jul 11 '22 at 13:47
  • @RobbieGoodwin I assume vectory means the Latin word which is the root of the word "masculine." – DLosc Jul 11 '22 at 18:42
  • @DLosc How is your assumption helpful, please? My first concern was that Vectory's Comment on "mās" popped up with neither introduction nor apparent reason, although it seemed to be a reference to something previous that I - and apparently you, too - had no knowledge of. Doesn't the accent over "mās" signify that this is not about Latin? How could "mās" be part of Latin? – Robbie Goodwin Jul 11 '22 at 21:48
  • 2
    @RobbieGoodwin (1/2) I think it's a very safe assumption. For one thing, the comment "it seems the question pertains to the Latin stack (incl. Greek)" indicates that vectory is talking about something from Latin or Greek. I did not previously know that mās was a word, but I knew that the word "masculine" comes from Latin masculus, which could plausibly be derived from mās plus the diminutive suffix -culus (and indeed, that is the etymology that Wiktionary gives). – DLosc Jul 11 '22 at 22:49
  • 1
    @RobbieGoodwin (2/2) The accent in mās is a macron, which is sometimes used in Latin spelling to distinguish long vowels from short vowels. You will notice that Wiktionary spells the word as mās in the article text. – DLosc Jul 11 '22 at 22:49
  • @RobbieGoodwin it seems you received a notification about my comment, which can happen if you have the question on follow, or sometimes automatically if you are part of the comment chain. If so, note that I had not strictly meant to addressed you directly. – vectory Jul 12 '22 at 07:27
  • A very similar question exists already: Who was the first to call noun classes "genders"?. TL;DR: the greeks used genos for both natural categories (ie. "kind", which is actually cognate) and sex (hence "gender" through French, from Latin, also cognate) since before Aristotle – vectory Jul 12 '22 at 07:30
  • @vectory I apologise for that, reading through these answers I have realised I hadn't thought out the question fully. Asking "Do the pronunciation differences in typically masculine and typically feminine French word endings ultimately descend from ancient words, and if so were those words related to masculinity and femininity?" would have better expressed what I was thinking. – AML Jul 12 '22 at 17:50
  • @DLosc Thanks and how could you not see, my Comment sought to discover how or why “mās” came into this? Vectory told us “The etymology of mās is ultimately uncertain, far as I know.” I know no better - quite likely less - and so what? I’m not challenging what Vectory said. I’m simply asking why it was helpful to say that? How did it fit here? – Robbie Goodwin Jul 13 '22 at 01:29
  • @DLosc I will not notice how Wiktionary spells the word, unless you - or perhaps Vectory - first explains its relevance. How hard could that be? – Robbie Goodwin Jul 13 '22 at 01:32
  • @Vectory Thanks and how could it matter whether you meant to address me directly? My queries remain first, how and why "mās" suddenly popped into this thread with no introduction and further, whether the reference was to modern European "mās", or Japanese "mas" or what? How hard could that be? – Robbie Goodwin Jul 13 '22 at 01:37
  • @Vectory Can you explain "cognate", please? I take "cognate" to be, very broadly, associated with; stemming from the same root; having a similar meaning… and I expect to see not "cognate" alone, but "cognate with." You seem to be using the term quite differently. What's going on, please? – Robbie Goodwin Jul 13 '22 at 01:44
  • I was tacitly implying the following: if masculin as a grammatic term goes back to Latin, and is related to mās, while the etymology of it is unknown, it could as well have had a more generic meaning than "manly" when the grammarian first used it. This may be unlikely for all that I know very little. Comments are not intended for extended discussions that should be moved to the chat, I know that much. – vectory Jul 13 '22 at 10:49
  • Arrg, someone claimed (modern) English has grammatical gender (in comments?), but now I can't find it. This is contradicted at 08 mins 34 secs in How the Vikings changed the English Language: ""In modern English there is no grammatical gender... only for nouns referring to a particular sex ... grammatical case ... exception of the Saxon genitive..."" – Peter Mortensen Jul 13 '22 at 21:57

4 Answers4

21

The names currently used for French are inherited from Latin, which had three genders: masculine, feminine, and neuter. (Some ancient grammarians added "common" and "epicene" to this list, for both Latin and Ancient Greek.)

Since the vast majority of words for male humans were masculine, and the vast majority of words for female humans were feminine, these were the most obvious labels to choose. There are a few other rules of thumb for what gender a noun will be—types of trees, for example, are almost always feminine in Latin—but none of them were as consistent and useful as "words specifically for men are masculine and words specifically for women are feminine". So those names stuck.

Draconis
  • 65,972
  • 3
  • 141
  • 215
  • 1
    I think French in OP is an example, but I do think it's worth mentioning that some non-IE languages have a very similar breakdown, so masc/fem/neut seem to have developed independently in many language families (I'm no expert, but I believe e.g. Khoisan languages have that tripartite gender division). – cmw Jul 10 '22 at 02:57
  • 2
    @cmw Correct — according to WALS feature 31A, 75% of languages with noun classes have a sex-based gender system. (WALS isn’t always reliable for individual languages, but this aggregate value seems reasonable.) – bradrn Jul 10 '22 at 07:18
  • There’s variability though in the degree to which nouns without biological sex can be assigned to masc/fem classes. Dravidian languages for instance tend to have very strict semantic systems where only humans and deities are in masc/fem and everything else is in a residue class, unlike Romance languages where mas/fem have absorbed many nouns without biological sex. – Keelan Jul 10 '22 at 17:38
  • 1
    @Draconis How is "the vast majority of words for male humans were masculine, and the vast majority of words for female humans were feminine…" not pure tautology, explicable only in terms of itself? – Robbie Goodwin Jul 10 '22 at 20:59
  • 5
    @RobbieGoodwin Well, whether it's a tautology depends "masculine" is understood to mean "the gender that a word for a male human has", or whether it's understood to have some pre-existing meaning that came to mean "referring to men" because of its association. Even if the former, it can be made non-tautological by rewording it as "There is a gender such that the vast majority of words for male humans have that gender, and so that gender is referred to as 'masculine' ". – Acccumulation Jul 10 '22 at 21:43
  • 6
    @RobbieGoodwin Yes, as Acccumulation put it, the vast majority of words for male humans fall into "category 1", and the vast majority of words for female humans fall into "category 2", so we decided to name those categories masculine and feminine. The language's structure predates the labels we put on it. – Draconis Jul 10 '22 at 21:56
  • @cmw Would it be fair to say "if the m/f/n system developed multiple times independently its origin is an anthropological matter"? – AML Jul 10 '22 at 22:04
  • @AML Anthropological as opposed to what, though? I'd say it's still a matter of linguistics (since we're talking about patterns that appear in human languages), but linguistics can easily overlap with other fields. – Draconis Jul 10 '22 at 22:05
  • @AML I'm not sure we can know. It's impossible to guess at a "primordial" language. Hittite, for example, seems to have lacked a "feminine" gender, and some think (and I agree) that PIE did as well: you have masc/fem as a "common" gender and then a neuter gender. Even in Latin, 3rd declension nouns are either common or neuter: it's only with 1-2 adjectives that we can tell there's a masc/fem gender. Algonquian languages have an animate/inanimate gender, too, although there is some controversy in that (I'm no expert in those languages). – cmw Jul 10 '22 at 22:23
  • That said, Draconis is right that this is primarily a grammatical concern. I doubt there's a single language where the dividing principles of nouns are clear and consistent. – cmw Jul 10 '22 at 22:25
  • 1
    @draconis I'm not an academic, and not confident my question even makes sense. I apologise if I am asking a nonsense question and think it would be reasonable that you tell me if I am: A German man starts talking to me about 'der boden' and he tells me that the word is masculine. He says this because he has learnt it from someone else, who has learnt it from someone else and so on and so on. Go back far enough and you reach a contemporary of the first person to ever consider the ground (or the word 'ground'?) to be masculine. Why did that person say it was masculine? Is it possible to know? – AML Jul 10 '22 at 23:07
  • @AML Ah, so your question is why languages have gender systems in the first place? – Draconis Jul 10 '22 at 23:10
  • 1
    @Draconis Not really. Consider an alternative history: My distant ancestor meets an owl and lets it live in his woods. Later he wants to tell his friend about 'the owl in the woods I forage', but his language is ambiguous and he knows his friend will laugh at him for foraging an owl. So: instead of just saying 'owl, woods, forage' he says 'owl, woods, forage(sun)'. If he somehow was foraging that owl he might have said 'owl, woods, forage(moon)'. This catches on. Much time passes. '(sun)' and '(moon)' have now degenerated into verb endings. 1/ – AML Jul 11 '22 at 02:12
  • 1
    Modern speakers of the language don't recognise those verb endings as meaning 'sun' or 'moon' per se, any more, but they still say one ending is used for 'day' nouns, and the other for 'night' nouns. Some nouns are obviously one type - 'owl' is a night noun - but for many the reason for being day or night is lost to history, or perhaps entirely arbitrary.

    I'm trying to ask is it an accident of history that something like the above didn't become predominant, or is there something innate in us that predisposes us towards developing the gender systems we have? 2/2

     – AML Jul 11 '22 at 02:13
  • 2
    @AML I see; I'd recommend asking that as another question ("why do sex-based gender systems arise?") but in this case it's theorized to have to do with Proto-Indo-European animal husbandry. – Draconis Jul 11 '22 at 02:16
  • 2
    @AML On that "German" example: it's really just a matter of linguistics at that point. There's no non-linguistic "reason why" Mädchen is neuter and Frau is feminine aside from form. Likewise, in Latin, why is baculum (stick) neuter but truncus (trunk) masculine and arbor (tree) feminine? You can come up with ad hoc explanations, but the answer is much more dependent on analogies and word-form than anything else. – cmw Jul 11 '22 at 03:36
  • @Draconis, re: your comment, "The language's structure predates the labels we put on it." Is it notable that the ancestral language is thought to have had an animate / inanimate distinction at some point, with Hittite having only a feminine marker for persons (which may itself be significant or not), just as the surrounding language, in particular those written in Sumerian cuniform made liberal use of in at least one casr very grafic determiner - think fore-man and the like. I mean, the innovation upto IE's three classes was perhaps advanced by talking about it, same as is the case today??? – vectory Jul 11 '22 at 06:00
  • 1
    @AML w.r.t. der Boden I feel qualified to respond as a German speaker (xD). 1. The assignment of gender to new loanwords seems to be arbitrary (cf. Joghurt for e.g.) unless conditioning factors work out, begging the question. Semantics can be relevant (m. Computer - a profession prior to electronics; early n. Laptop, but m. Laptop-Computer, now m. Laptop, but still n. Notebook, beside n. Buch "book"; vis-a-vis morphonology see also m. Topf ~ regional Topp "pot, cauldron", etc.), maybe soundsymbolic (productive diminutive n. -chen, cp. Hündchen "puppy", Kind "child"; deverbal f. -ung; etc.) – vectory Jul 11 '22 at 06:16
  • [cont.] I'm not even sure what the origin of Boden is. I commented to note that you are probably asking about the Humus <> Humanus dichotomy, but that's a proper question about semantic fields that is too broad for this. Cf. husband, buddy, bottom, bunda, bantu, and of curia Adamus, etc. This is either sheer coincidence or complete madness. See also my Ger.SE question about Busen (bossom?), cp. Schoß, lit. "cradle, origin" – vectory Jul 11 '22 at 06:30
  • 1
    My favorite irregular word in Latin was agricola, agricolae - farmer. This word is feminine, despite the fact that farmers were men. The connection to mother earth apparently dominated over the biology of the farmers themselves. – Cort Ammon Jul 12 '22 at 04:34
  • @CortAmmon Not quite, I'm afraid: it's in the first declension, but generally masculine. For example, Cato refers to a bonum agricolam various times. Similarly nauta, poeta. – Draconis Jul 12 '22 at 05:06
8

Imagine if every French speaker suddenly agreed that nouns were one of 'animate' and 'inanimate', or 'chocolate' and 'strawberry', or 'A' and 'B' instead of 'masculine' and 'feminine'. The language could go on being used identically to how it was before.

All words are more or less arbitrary labels. Regardless of what you call the French morphological genders, if the language goes on being used identically to as it is currently, we would see that:

  • When the subject of a sentence is the first-person singular pronoun referring to a female speaker, or the name of a single female person, a predicative adjective takes the same form as when the subject is a singular noun belonging to the same class as eau or bière.

  • When the subject of a sentence is the first-person singular pronoun referring to a male speaker, or the name of a single male person, a predicative adjective takes the same form as when the subject is a singular noun belonging to the same class as arbre or thé.

  • There are many pairs of words where the one belonging to the eau, bière class denotes a female person of some kind, while the one belonging to the arbre, thé class denotes a male person of that kind (or a generic person of that kind). For example, voleuse belongs to the eau, bière class and means "female thief". Voleur belongs to the arbre, thé class and means "male thief" or just "thief". Aside from derivationally related pairs like these, there are words with unrelated forms but coordinate meanings that form such pairs, such as sœur "sister, female sibling" (belonging to the eau, bière class) and frère "brother, male sibling" (belonging to the arbre, thé class). We don't find any such pairs where the noun denoting a female kind of person belongs to the arbre, thé class and the noun denoting a male kind of person belongs to the eau, bière class. There are a small number of words belonging to the eau, bière class that can denote a male or female person equally, such as personne "person", but most words denoting types of persons are not like personne.

  • Nouns for animate beings (living creatures that move around and seem to have some agency) and inanimate objects (non-living things that don't seem to have any agency) are frequently found in either noun class. Given this, it seems it would be purely arbitrary to call the French noun classes "animate" and "inanimate".

  • In contrast, as illustrated by the first three bullet points, there are many contexts where the noun classes have a strong correlation with whether a person is male or female. Given this, it is not purely arbitrary to refer to the French noun classes as "masculine" and "feminine". This does not necessarily mean that every noun in the masculine noun class refers to or is associated with male beings, or that every noun in the feminine noun class refers to or is associated with female beings.

brass tacks
  • 18,116
  • 1
  • 45
  • 91
  • 1
    Now I need to know which genders chocolate and strawberry are – No Name Jul 10 '22 at 16:51
  • "All words are more or less arbitrary labels." Technically true, but when we reuse or words or derive new words from old ones, there's usually a metaphorical reason, and that's not arbitrary. – Barmar Jul 10 '22 at 21:46
  • However, there are exceptions, too. When physicists came up with the flavors up/down/strange for quarks, these were just arbitrary choices. – Barmar Jul 10 '22 at 21:49
  • @Barmar I'd say that most metaphorical applications are arbitrary, but not random. – cmw Jul 10 '22 at 22:27
  • @cmw Yeah, but it's not arbitrary like the original words. Deciding which metaphor to use is arbitrary, but given that choice the words are somewhat forced. – Barmar Jul 11 '22 at 03:30
  • "most words denoting types of persons are not like personne." This may be correct in a frequentist view. Other than that, there's no weight to meassure this claim accurately. Arguably, their existence is of some importance for the system where most nouns have no biological gender at all. At least in phonology it has been argued that rare, complex phonemes can contribute to the stability of a system (IIIRC). – vectory Jul 11 '22 at 07:53
  • @Barmar the names of the quarks aren't arbitrary. Up and down refer to the isospin, strange refers to being the odd one out (the others hadn't been discovered yet), charm refers to the longer life charmed hadrons had, and top and bottom are analogies with up and down, coupled with the (yes, I'll admit it) arbitrarily chosen original names of truth and beauty. You have a point though, I have long bemoned the continued use of strange and charm, and wish they'd be renamed to high and low. Or would it be low and high? I can never remember... – No Name Jul 12 '22 at 03:49
  • @NoName There's still more to it: b,c,d and s,t,u form two triples of consecutive letters. – Hagen von Eitzen Jul 12 '22 at 05:03
  • @HagenvonEitzen huh... interesting. Although I just checked, and strange is in fact down type – No Name Jul 12 '22 at 05:05
2

The defining property of grammatical gender/noun class is agreement.

With that in mind, unlike the other answers I would say that instead of looking at nouns to explain their names, we ought to look at the behaviour of words that agree with nouns - that is, adjectives and determiners.

In particular, adjectives can be used to describe pronouns that themselves lack any gender or class (in Indo-European, this is generally the first and second person in either the singular, dual, or plural).

So, we can now consider someone addressing a variety of people or objects and assigning them adjectives e.g. "you are large". If the speaker is speaking French we quickly see that male humans receive one set of adjective forms, whilst female humans receive the other, whilst for all other categories of objects some take one and some take the other. Gender is the only one that reliably predicts the adjective form, and so it is natural to name those adjective forms (and by extension, the class of nouns that take that adjective form) masculine and feminine.

Z4-tier
  • 103
  • 2
Tristan
  • 8,262
  • 25
  • 46
  • 1
    I don't see what any of this has to do with language acquisition. You are right that determiners are relevant and they cover the pronouns that agree so I've edited that in. The reason I discuss adjectives specifically is that, unlike determiners, they can be made to agree with things that lack grammatical gender themselves (e.g. second person pronouns). As such they give us a way to diagnose how grammatical gender relates to objects in the real world without relying on how it relates to nouns – Tristan Jul 11 '22 at 11:48
  • 2
    also petit/petite is marked in Parisian French. The final -e may be silent, but the -t is pronounced in the feminine and silent in the masculine. Regardless, provided there exist some adjectives with distinct masculine and feminine forms the proposed test can be applied, no matter if there are plenty of adjectives that don't distinguish gender – Tristan Jul 11 '22 at 11:49
  • lastly, questions that ask "why" are generally going to be difficult or impossible to source so I'm not sure it's a reasonable objection to make in general, but if there are some specific claims you feel ought to be supported by a citation I'll do my best to dig one out – Tristan Jul 11 '22 at 11:51
  • Well, French was a long time ago for me. On the other hand, adj. jeune, n. le/la jeune might be a better example in my view, except that German der Junge for comparison is literally the boy and petite is tendentially associated with girly characteristics at least in English. Most interestingly, ancien(ne), /ɑ̃.sjɛ̃/, /ɑ̃.sjɛ.n/ sounds like ablaut to me. Whereas, given le beau, la belle, f. /-t/ is considerable, cp. elle est moins bête*, Biatch. There's clearly circular reasoning, a bit of Whorf with respect to language awareness. – vectory Jul 11 '22 at 20:45
-4

The Question denies itself.

Choose a well-recognised language and list 50 common nouns, with their genders.

If that list makes the reasons for gender assignment obvious, extend it to 100…

If on the other hand, a list or 50 - or 500 - nouns fails to make the reasons for gender assignment obvious, does it not necessarily reveal the assignment as broadly random?

Robbie Goodwin
  • 93
  • 2
  • I'm trying to ask how the names for the different grammatical genders originated rather than how nouns were assigned to one gender or another. I suppose I'm also asking in a roundabout way if this is even a meaningful question in linguistics, or if it is a question for another field. – AML Jul 10 '22 at 21:39
  • 1
    Rhetorical questions are best avoided here. This is a Q&A site, not a forum. Better state it directly. You can change you answer. (But *without* "Edit:", "Update:", or similar - the answer should appear as if it was written today.) – Peter Mortensen Jul 11 '22 at 22:29