In some languages, like German and French, every noun has a gender and each gender has its article. Whereas languages like English and Persian do not have genders. Why is that?

Even though these languages belong to the same family of languages. What is the philosophy behind gender in a language? Is it a universal concept?

  • 10
    Women Fire and Dangerous Things by George Lakoff is about the origin of noun classes (of which grammatical gender is a particular case).
    – Colin Fine
    Commented Nov 29, 2020 at 23:15
  • 5
    Note that grammatical gender is found in the oldest strata of Indo-European: English and Persian have lost the feature. In English this is largely because most grammatical endings were weakened or lost in the transition from Old English to Middle English. But this cannot be the whole story, since French has kept grammatical gender despite comparable reduction of nominal endings.
    – Colin Fine
    Commented Nov 29, 2020 at 23:19
  • 2
    @ColinFine your comments deserve being wrapped in a full-featured answer that could be upvoted and accepted by the OP. Commented Nov 30, 2020 at 0:00
  • 7
    English actually has genders. It's just that nearly anything that does not have an intrinsic gender is neutral. With the exception of ships, of course.
    – jcaron
    Commented Nov 30, 2020 at 13:53
  • 4
    @jcaron I'm curious how many speakers wouldn't accept "it" for ships. I certainly wouldn't care. Commented Nov 30, 2020 at 19:31

3 Answers 3


The origin of grammatical gender is not necessarily well understood, but presumably it originated like any other inflectional feature and then became associated with gender when it was noticed that some prominent things of one natural gender fell into one paradigm and things of another into another, upon which those paradigms might have been generalised to other things of the same natural gender and, eventually, things that couldn't reasonably be said to have a natural gender. As noted, grammatical gender doesn't need to have anything to do with natural gender—in plenty of languages the distinction is between animate and inanimate things, for example, or more complex than that—it just happens that in Indo-European languages there's an overlap between grammatical gender and natural gender, so the two became associated in our languages. (I initially wanted to blame the Greeks or the Romans, but the word their grammarians used, γένος/genus, just means "kind" and has no connotation of natural gender.)

The other part of your question, why various languages have different gender systems despite being in the same family, and specifically in the case of Indo-European languages, is more concrete and easier to answer: because sound changes and other processes of language change lead to loss of contrast between the genders.

Late Proto-Indo-European had three genders: masculine, feminine, and neuter. Roughly speaking, in Germanic languages, the free accent of PIE was replaced by an accent on (basically) the first syllable of words, and one that was so strong that it led to the progressive weakening of vowels in unstressed syllables (an ongoing process in e.g. English), and eventually the loss of some of them, especially at the end of words. This meant the loss of all or most of the case system in all Germanic languages, as well as most traces of the gender system: it mainly survives through articles. Proto-Germanic had a demonstrative pronoun *sa that became the definite article in the West Germanic languages. In Old English it was still inflected for case and gender, but as it was unstressed and English continued the process of vowel erosion more vigorously than some of its siblings, by Middle English it had just become þe in all contexts (with a þ by analogy with some of the other demonstrative pronouns), and English lost its genders. German, on the other hand, kept declining its definite article into modern times (der, die, das), and retained three genders as a result. (Why did it keep inflecting the article? Probably for the same reason it retained a bit of its case system.) Dutch is somewhere in the middle: its descendent of *sa eroded into undeclined de, but it innovated a new article for neuters along the way (het, cognate with English it), so today it's often said to have two genders: common (formerly masculine/feminine) and neuter (but N.B. that almost all dialects retain a three-way contrast in the indefinite article, e.g. ne, een, e, and thus three genders; Hollandic, and therefore Standard Dutch, does not).

In another branch of the family tree, Latin had three genders inherited from PIE, but even though Latin nouns were (relatively) highly inflected, the neuter gender was mostly not very distinct from the masculine anymore. Already in the Classical period we find graffiti in Italy confusing one for the other, and with the export of Latin across Europe and the simplification of language that close language contact often entails, the contrast was eventually deemed too trivial to maintain and lost almost entirely in Vulgar Latin. Today, the Romance languages (effectively) have two genders; because the merged gender stands in contrast with the feminine, we call it masculine instead of e.g. common. In Italian and Spanish you can often still tell the gender of nouns by their ending (not always, but about as often as in Latin), but in French the endings underwent an erosion nearly as severe as in Germanic, and the articles are again the only clue.

The history of the Indo-European languages is broadly one of loss of inflectional complexity, though the details vary from language to language. As a result, if an IE language's gender system differs from its parent it will be in a reduction in genders. That's not an iron law: genders can be innovated as well. Early Proto-Indo-European itself probably had two genders (animate and neuter), and what was originally a collective/abstract suffix *-h₂ was reinterpreted as a feminine ending at some point (becoming, for example, the -a of Latin first-declension nouns). That seems to be rarer, though.

  • 4
    I contest the notion that English has lost its genders: The noun actor is masculine and is referenced with the pronouns he, him, his, himself, an actress is feminine (she, her, her, herself) and the team is neuter (it, it, its, itself). Commented Nov 30, 2020 at 17:26
  • 15
    @jk-ReinstateMonica As far as the nouns are concerned, you're confusing grammatical and natural gender. The pronouns obviously do preserve inflection for gender, which makes clearly disentangling things hairier and is why I kind of ignored them—they're also inflected for a few cases, but few people would take issue with the general claim that English has lost its case system.
    – Cairnarvon
    Commented Nov 30, 2020 at 18:10
  • 3
    No, I am just talking about a gender system. In English, it is purely semantic, but it exists nevertheless, as the pronouns demonstrate. There are many languages that do not have a gender distinction at all. Commented Nov 30, 2020 at 18:14
  • 19
    @jk-ReinstateMonica The point is that there is a difference between saying a noun is grammatically gendered and a noun is referring to a natural gender. For example, "accepta" is a Latin noun with a feminine declension meaning "a portion of land granted by the state". The word is not trying to imply that the land in question is female. It is the word itself that is gendered, not the concept with which it is associated. As far as I am aware, English does not follow this pattern to any noteworthy degree.
    – Abion47
    Commented Nov 30, 2020 at 22:20
  • 6
    @jk-ReinstateMonica For a living example: the Italian word guardia (watchman) has feminine grammatical gender (and requires feminine adjectives and verbs, when verbs are marked for gender), but can refer to both men and women. I don't know of any similar example in English. Commented Dec 1, 2020 at 6:39

As you already noted, gender or more generally speaking noun class is not a universal linguistic concept. There are a lot of languages in the world with no traces of gender or noun class, not even in the pronouns where English retains the inherited three genders. Note also that a language can have gendered nouns and genderless articles. Arabic is an example of such a language.

For some high-level overview, see WALS, chapters 30–32. The chapters deal with the number of genders/noun classes, their relation to sex, and the assignment criteria of genders to words.

Main results from the three chapters are:

  • The majority of languages in the samples has no gender/noun class distinction at all.
  • Differing from the question, English is counted as a language with 3 genders because there are nouns that are referenced by three different pronouns he, she, and it.
  • Two genders is the most frequent number of genders apart from no gender at all
  • Out of the languages with gender system, three quarters have a system based on sex
  • There is an approximate 50:50 split between purely semantic gender systems and gender systems where formal assignment plays an important rôle.

Note that there is nothing said about the why, and the why-question is probably unanswerable with current linguistic knowledge. Gender systems are long lived and stable in their type historically. The "minority system" of Bantu noun classes exemplifies this point very well.

  • 6
    Hmmmm aren't answers supposed to contain the answer itself, instead of just a link to a page with the info? Your first paragraph "As you already [...] for such languages." just reflects on what OP said, without actually answering their question about why that happens. And your second paragraph falls outright into the "your answer is in another castle" category. A summary of what the linked chapters say on the topic of why is there such a difference between languages of the same family would be desired.
    – walen
    Commented Nov 30, 2020 at 8:29
  • @walen Since you put emphasis on the word why in your comment: This question is actually not answered and probably not answerable at all, at least with our current understanding and knowledge of human language.—I will expand my answer somewhat to give an impression what is in the WALS chapters. Commented Nov 30, 2020 at 17:10
  • 1
    English ... nouns that are referenced by three different pronouns he, she, and it. — “Referenced” is the keyword here, English nouns are just referenced by different pronouns, but nothing agrees with them in gender, while in Dutch attributive adjectives, articles, demonstr. pron.s and the possess. pron. ons/onze “our” agree in gender with the noun they refer to. The question is “Is purely referential gender a real gender?” If you say “yes”, then there follows that Japanese also has gender, since it also has gender-specific 3rd p. sg. pronouns, 彼 kare “he” and 彼女 kanojo “she”.
    – Yellow Sky
    Commented Dec 1, 2020 at 17:28
  • 2
    In their chapter on the number of genders, WALS write, “The defining characteristic of gender is agreement: a language has a gender system only if we find different agreements ultimately dependent on nouns of different types. ... Most scholars working on agreement include the control of anaphoric pronouns by their antecedent (the girl ... she) as part of agreement. If this is accepted, as we do here, then languages in which free pronouns present the only evidence for gender will be counted as having a gender system.”
    – Yellow Sky
    Commented Dec 1, 2020 at 18:22
  • 1
    This means it's just their choice to consider English a language with genders, there's no consensus among scholars as for whether the control of anaphoric pronouns by their antecedent should be considered as a kind of agreement. And you just quote WALS without giving at least a hint at that being far from the absolute truth, neither you present any proofs of them being right.
    – Yellow Sky
    Commented Dec 1, 2020 at 18:28

Anglo-Saxon, a Germanic language spoken by most residents of England, was much changed by the influence of our new Norman overlords after 1066 CE, who spoke a variety of French. The two gender systems were incompatible (e.g. la Lune (fr), das Mond (ger)).

So English more or less lost genders, and similarly lost most case endings and verb inflections.

  • 3
    1) you could have researched the genus of der Mond correctly … 2) Romance and Germanic languages live happily side by side in at least 4 European countries to a varying degree; most German-speaking residents of South Tyrol are also fluent in Italian (their ‘overlords’, to use your term) yet there is zero indication of of the South Tyrolean moon becoming feminine or losing its genus.
    – Jan
    Commented Dec 1, 2020 at 8:18
  • 2
    Several hundred years earlier, Anglo-Saxon was the language replacing Latin (which we can treat as having the same gender system as French). Why didn't Anglo-Saxon lose its gender then?
    – chepner
    Commented Dec 1, 2020 at 14:48
  • As Jan and Chepner point out, a bilingual community can speak two languages with conflicting genders. So here in England we could have been bilingual in Norman French and Anglo-Saxon. But actually we assimilated a lot of French words into the vernacular. I am suggesting that in that situation the contradictions of gender were intolerable. However, explanations like this are easier to make than to prove. Commented Dec 2, 2020 at 9:46

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