17

I get the concept of (Western) grammatical gender and why it is used in the languages I know. However, I do not understand how native speakers casually avoid mistakes in grammatical gender.

For example, I have heard in Spanish someone saying "Hay una [pause] falta de agua". How did she figure out the gender of the noun (not "un", but "una") before figuring out what the noun even is?

The pause lasted about a second by the way.

If it were me, as a person whose strongest languages are genderless, I would say "Hay un [pause] fa- una falta de agua".

The process gets harder with languages where the gender of nouns are harder to distinct, like in Dutch and (oh no) German.

I am also asking for the linguistic perspective of this. Do native speakers naturally "feel" the need to articulate feminine instead of masculine?

21
  • 18
    They don’t, necessarily. The ‘workaround’ you describe as what you would probably do is also quite common among native speakers when the right word won’t come to them. Sometimes, even though the exact word escapes them momentarily in a ‘right on the tip of my tongue’ type of situation, its gender may be part of the information they do have available. In other cases, it’s not, and they may well get the gender wrong. Just as in English, when you blank on a word, you may use a or /ðə/ initially and then have to correct yourself to an or /ði/ when you find the word, if it begins with a vowel. Commented Dec 19, 2022 at 18:48
  • 3
    Minor nitpick about the examples: Dutch and German both have very straightforwards grammatical gender, the issue there is simply that it doesn’t map cleanly to how you spell the words like is generally the case in languages like Swedish or Italian (though both do have exceptions to this). If you want examples of difficult grammatical gender systems, I encourage you to instead look at Russian (which further subdivides animate from inanimate) or Polish (which has three different but partially overlapping masculine genders). Commented Dec 20, 2022 at 2:47
  • 2
    @AustinHemmelgarn because the F and M articles merged in Dutch there is actually major confusion among native speakers, we usually don’t have a clue of the original gender of a word. And lately the N seems to be losing ground to this new common gender as well.
    – Keelan
    Commented Dec 20, 2022 at 7:25
  • 2
    As @JanusBahsJacquet said, "hay un- [pause] ... una falta..." happens constantly. But in the example you picked, almost all of the alternatives for falta in that phrase are also feminine: "hay [una falta/una escasez/una ausencia/una carencia/una insuficiencia] de agua". (In fact, I can't find a masculine synonym w/o resorting to a thesaurus.) But if the phrase allowed for equally-common f. and m. alternatives, e.g. "ha cometido [un error/una equivocación/un fallo/una falta/...]" then it'd be normal to go with un by default and then change to una if needed (and viceversa).
    – walen
    Commented Dec 20, 2022 at 8:58
  • 2
    One thing to think about as well is that the speaker might guess wrong on other occasions. Once she said una she had a roughly 50-50 chance of being right. On other occasions she might have had to correct herself.
    – adam.baker
    Commented Dec 20, 2022 at 10:03

7 Answers 7

21

Do native speakers naturally "feel" the need to articulate feminine instead of masculine?

Your assumption regarding how speech works is not quite right regardless of whether a language is gendered or not. Hesitation, double-backs, repetitions, omissions, ellipses, mistaken terms not immediately realized and all sorts of things can happen anywhere along an utterance chain.

For example, in English, a speaker might mean one determiner (a or the) and say another. I'm using the a/the thing as a comparable to the Spanish un/una thing.

  • Determiner: "Oh, but I saw a...the possum in the driveway".
  • Verb: "Oh, we went...have gone recently".

So, in Spanish, maybe it could have "una escasez de agua". One cannot assume that the person was trying to figure out the gender. as other words that are also feminine could fit here perfectly well.

It is more likely they they had another term in mind such as escasez or ausencia, both feminine, and paused until the word falta came to mind.

3
  • 4
    +1. I once had an assignment transcribing some speech by a French professor. Some excerpts: "nous n’avons—nous ne sommes pas encore exposés" ; "les sciences de—médicales—sont très soutenues" ; "des cours d’internship, des cours de … qui permettent" ; "Écoutez : le ... la moyenne des étudiants" ; and once, "de ........ du programme". I was stunned at that last one because I thought he would say "le", that the contraction would be interrupted by such a long pause — but that rule is stronger than a pause and necessited restating the whole preposition! Commented Dec 20, 2022 at 4:04
  • You seem to be answering the opposite question. The example in the question is showing that they got the article gender correct even though they didn't know the word when they uttered it.
    – Barmar
    Commented Dec 21, 2022 at 15:34
  • We can't know if it was "correct" until the utterance is completed. They might have come up with another term, and one that is masculine (they might have meant a masculine term). Who knows?. Except for two comments (adam and Luke), I'm hard pressed to believe that people can't get that the features of spoken language do not depend on supposed correctness of a determiner. Hesitation in speech doesn't prove anything at all and correctness is not a linguist's term, as it were.
    – Lambie
    Commented Dec 21, 2022 at 17:40
8

As a German native speaker, I find the gendering of nouns something that happens absolutely automatic and with no thought whatsoever, like most other aspects of talking in a natural, un-forced manner. I daresay I cannot remember much of my Deutsch classes in school, but I cannot recall ever hearing of any rule, regularity, rhyme or reason for the gender of most words - it just is what it is. (I do not mean rules like "-er" or "-ling" usually being male, but a generic explanation for why a certain noun ended up with "-er" or "-ling" in the first place instead of ending up with a typically female suffix.)

From experience with my children, I cannot recall them mixing up the gender frequently when they were very small; presumably gender is learned together or inseparable with the meaning of the words.

As a side note, there are a very small amount of nouns whose gender seems variable based on dialect or region (e.g., "das Radio", "der Radio"), but that's not at all common. Also the gender for the same word might be different in different (but related) languages, like the moon being female in French, and male in German.

As to why a speaker may utter the correct combination of words/word endings even if it seems like they're searching for words mid-sentence, I can think of one reason: by some interpretations/theories, the thinking brain can be thought of to work in "modules" or "components".

One basic example would be one module which is a pure generator of thoughts - it just spews out random thoughts all day long with high frequency; and another module then acts as a filter and filters out thoughts which make no sense right now, and a third component which is conscious of thoughts. A setup like this would explain many aspects of our brain which we can often witness ourselves, for example, that when we're in a "flow state" with just the right intensity of problem solving, we are not aware of random thoughts (the filter filters out all thoughts not relevant to the problem at hand before they reach the "awareness" component). But if we are trying to sleep, and have no problem to solve, then the filter has nothing to filter by, and passes along random thoughts to our awareness.

The same could well be the reason for your phenomenon: there could be some component of the brain which wants to communicate something to the world, with a representation of objects in an abstract manner (i.e., not in the form of words), and a different component which translates the abstract thing into words. The abstract representation of the object might as well already contain the gender; so the "verbalization" stage already has the information that what's coming will be feminine even though it's still "looking up the word" in its "dictionary".

Note that I am using a lot of quotes here - treat this as a casual, naive, non-scientific description please, I am not a brain scientist. I am not sure if the above component-based model of the brain is still en vogue or not so much; I am pretty sure that nobody really knows for sure how the brain works mechanically in this aspect.

11
  • This is irrelevant because the speaker might have meant something altogether different. We simply do not know.
    – Lambie
    Commented Dec 21, 2022 at 17:45
  • 1
    "the moon being female in English" nouns are not gendered in English.
    – nasch
    Commented Dec 21, 2022 at 23:52
  • 3
    @AnoE Referring to the Moon as ‘mistress’ does not make it a feminine noun. Anthropomorphising a noun and assigning biological gender to it is not the same as grammatical gender; compare how ships are often referred to as sie in German (just as in English), even though das Schiff is neuter. Commented Dec 22, 2022 at 10:07
  • 1
    @ScottishTapWater, that is what my intended meaning was - I hope it's clear now with the elaboration I've added.
    – AnoE
    Commented Dec 22, 2022 at 10:38
  • 1
    @AnoE - Aye that's better Commented Dec 22, 2022 at 10:52
7

The linguistic answer is that speakers learn the gender of words from experience, they devise rules, and apply those rules. You have to know the grammatical rules of a language to speak the language.

That doesn't address the more interesting question of how knowing grammar figures into actual production of utterances, and really we haven't made a lot of progress in understanding how that happens. We know of a lot of things that explain why people may produce things that don't conform to what is said by the idealization that is grammatical theory. "Predicting the future" is a classical puzzle. In a simpler and more dramatic form, we encounter this in details of pronunciation in many languages, where the pronunciation of one word is determined by things that follow, sometimes by quite a chunk. Some things are in principle computable "before the fact", so you can apply a rule "x→y when z follows in the sentence" to the sentence /a b x q w r s y m/. You first string the elements together, then you compute the changes, and then you spew out the resulting words. You can do this kind of advance planning with some things, but not others. For instance, when you produce consonants, your lips will protrude somewhat just before you produce the vowel [u]. "Just before" is in terms of actual time, not abstract "precedence", so one doesn't say "I might useʷ... Uber" with a big pause but anticipatory labialization.

I would start with the premise that the speaker had at least computed "Hay una falta", and the pertinent question is "why and where do people pause". For example, the speaker is fishing for the right complement "de agua", but "falta" is already figured out.

I think the most important point to bear in mind is that the grammatical mechanism abstractly says what a speaker is striving for in creating an utterance, and it isn't a description of how people decide what things to talk about or what words to use when talking. It's not a real-time behavior generating mechanism.

4
  • 2
    Re 'the premise that the speaker had at least computed "Hay una falta"': could it also be that the lexical entry for falta is retrieved in parts, so that the speaker first has access to its gender, and then to its full form?
    – Keelan
    Commented Dec 19, 2022 at 17:36
  • 4
    I remember a situation in which a speaker of an African tone language with downstep was recording a text, turned the page, and found that the sentence continued farther than he had expected, so the downstep was out of his range. He felt he had to go back and start over with a higher base tone.
    – jlawler
    Commented Dec 19, 2022 at 18:17
  • Indeed, downstep computation is a piece of low-hanging fruit that is fairly easily replicable, for exactly the reasons you note. My own observations are that speakers vary in how they cope – "try harder" vs. "try again". As I recall, Rialland and/or Clements ran an experiment along these lines to see if there is a preferred strategy – I can't remember what the bottom line result was, though.
    – user6726
    Commented Dec 19, 2022 at 18:27
  • Learning genders is irrelevant to how speech flows.
    – Lambie
    Commented Dec 21, 2022 at 17:46
4

In German, there are two rules-of thumb for grammatical gender:

  1. Nouns "sound" male or female. There are some hard rules for this, e.g. words with -ung or -keit or -keit are almost always feminine. Then there are softer rules, like words that end with -e are more often feminine than not. But even if you make up a word that does not contain any obvious suffixes, people will still often have a tendency towards certain genders.

  2. Nouns that describe similar things tend to have similar genders. E.g. alcoholic beverages (except beer) are usually masculine. Der Chianti, der Ouzo, der Curacao usw. Exotic fruit are usually feminine. Die Khaki, die Durian, die Kiwi, die Litschi etc.

As others have pointed out, Germans have usually no problem at all with the articles in their language. However, it can become difficult for words they have never seen before ("Hornung"?). And the existence of (relatively new) words with several possible articles (der/die Paprika, der/das Ketchup, der/die/das Nutella) shows that the choice of an appropriate article is not always straightforward.

With nouns for similar concepts often having similar genders, it is indeed sometimes possible to guess the correct article before you remember the exact word you are looking for. But not sure how relevant that is in real-life settings.

2

I've read many descriptions of psychological studies that show that there are many automatic, unconscious processes going on in the brain before we become consciously aware of them. This is how we do so many complex things fluidly (once we've become practiced in them), and speaking is one of the most complex things we do on a regular basis.

A good book that explains much of this for lay people is Thinking, Fast and Slow by Daniel Kahneman.

So it's not unlikely that the unconscious "fast" system has already chosen the word, and thus triggered the corresponding article. Then maybe something interrupted their train of thought so the word didn't come out smoothly. The pause may seem like they're using the "slow" system to come up with the word, but it's being nudged by the fast system, so they'll usually utter the word that was originally intended.

This doesn't always happen, and the other answers present a number of examples of speakers going back and replacing a word when they realize the original utterance didn't match the word they eventually use.

1

In English, some nouns are countable. Some are not. Some are countable, but only in certain contexts, otherwise they aren't. Often, one item will be countable and another very similar item will not be, seemingly arbitrarily. And then there's 'data'...

The point is that, in making use of those words, you will generally not pause to specifically think of whether they are countable or not, or to apply the vague and not very consistent rules. You simply know by experience and usage, in just the same way that native speakers of a language with grammatical gender simply know how the words should be used.

10
  • 3
    The mass/count distinction is much more grounded in denotational semantics than is typically assumed of grammatical gender.
    – Keelan
    Commented Dec 20, 2022 at 20:50
  • True as that might be, people don't distinguish between the two by performing denotional semantics. They're not identical, of course, but similar enough to give, I think, an understanding to the asker of what gendered grammar feels like to someone for whom it is natural.
    – Kinro
    Commented Dec 21, 2022 at 14:08
  • Your claim that "often" the mass/count of a noun is "seemingly arbitrary" is simply not true, because of the denotational semantics. And this does, in fact, make it very different from grammatical gender, in a way that is crucial for this question.
    – Keelan
    Commented Dec 21, 2022 at 15:37
  • You can have a beer, but you can't have a wine. Lentils is countable, while rice isn't, though they're about the same size. Sometimes you can count hair, sometimes you can't. Data has been going from a countable plural to an uncountable. But that's just details anyway. Please think of what happens when you use those words. Do you think about these various rules each time?
    – Kinro
    Commented Dec 21, 2022 at 18:42
  • Ad beer / wine, (1) those are universal packager contexts, you should exclude them, and (2) just google for "I'll have a wine". Ad hair, countability depends on whether you conceive of the hair as a mass or individual units, it isn't arbitrary. Same to some extent for data. Yes, some things are lexically specified, but the countability of the vast majority of nouns is not, and that is the crucial difference with grammatical gender.
    – Keelan
    Commented Dec 21, 2022 at 19:44
0

My mother tongue is Spanish, it is simple, normally in Spanish we know the gender of the word depending of the last letter. If ends in a, it is feminine, if it ends in o it is masculine, off course there are a few exceptions.

In this case falta ends with a, so it is feminine la falta or una falta.

Regarding German language, it is more difficult, because there are too many exceptions to the existing rules. So mostly you need to learn by memory all the articles. Some fixed rules might be if a word ends in -ung or -keit the sustantive is feminine, if the word is a foreign word, normally is neuter.

6
  • der Ayran, der Ouzo, die Khaki, der Subotnik, der Jihad, der Hijab: lots and lots of non-neutral foreign words in German.
    – Jan
    Commented Dec 20, 2022 at 16:39
  • If there is any default for the gender for a foreign word in German, it is probably feminine (e.g., die Firewall despite the existence of a German masculine word der Wall), but there are too many exceptions to this to be practical. Commented Dec 20, 2022 at 17:24
  • 3
    This doesn’t answer the question. Rules for which words are what gender isn’t what the question is about. Commented Dec 20, 2022 at 19:28
  • 1
    The whole point of the question was to understand why the speaker appeared to know the gender of the word they wished to use, before knowing what the word itself was. This answer does not address that. Commented Dec 21, 2022 at 1:52
  • @SirCornflakes "Der Wall" is old-fashioned. People usually use die Mauer. E.g. in architecture the translation of firewall is die Brandmauer. I am not totally sure the gender here is derived from die Mauer, but one could make that case IMO.
    – Jan
    Commented Dec 21, 2022 at 9:20

Not the answer you're looking for? Browse other questions tagged or ask your own question.