Classes of words in languages tend to be either "open" (accepting new members readily) or "closed" (rejecting new members). This distinction is fairly easy to see: compare how readily English accepted the verb "to google", versus the hundreds of years of resistance to singular "they". So descriptively, verbs in English are an open class, while pronouns are closed.

But have we ever observed a class of words changing from open to closed, or from closed to open? Or an intermediate "half-open" stage, where new words are accepted but not commonly used or always seen as foreign?

  • 1
    Related: Languages with different open and closed word classes and the discussion there and in linked pages about Japanese verbs and pronouns acting as "closed" or "open" classes; the situation is a bit complicated and seems to have changed somewhat over time Commented Nov 26, 2017 at 21:39
  • ...in Old Japanese, pronouns were apparently more of a distinct "morphological class" Commented Nov 26, 2017 at 21:45
  • Japanese counters seem to have been opener in the past. They're still not completely closed, but the way they're open seems to have changed. Commented Dec 28, 2017 at 23:24
  • 1
    Bahasa Melayu (Indonesian) is claimed to be a language with an open personal pronoun class; any noun that can refer to a person (especially kinship and relational terms) can be used as a referential personal pronoun. Of course, the most commonly-used personal pronoun in all 3 persons is Zero, so the system is considerably more complex than usual, too.
    – jlawler
    Commented Jan 26, 2018 at 17:51
  • 4
    To state the obvious, all classes must have started as open. Going from closed to open is less self-evident. Commented Apr 22, 2019 at 7:46

6 Answers 6


Maybe we can see a (semi-)opening of the personal pronoun class in English just now: There are not only a lot of suggestions of an epicene pronoun floating around (this would extend the closed class by one member, but is in itself not an opening), but there are also a bunch of new pronouns on the verge for a "Third gender", going as far as allowing every individual to chose Eir own pronoun.

EDIT: Right now here on stack overflow pronouns are an open class, see I5 in this official staff post here

  • 7
    Yeah, but the native speaker has already chosen they, them, their, themself. Long ago. Other epicene pronouns may come in certain other contexts, but not in speech.
    – jlawler
    Commented Jan 26, 2018 at 17:46
  • I wonder if there are previous examples of pronouns becoming an open class through enforcement by an authority and expulsion of community members who objected to this language modification...
    – LjL
    Commented Oct 12, 2019 at 15:10
  • @LJL I am not aware of pronoun examples, but language planning and language enforcement happen all the time in the real word, I can think of the prescribed vocabulary (with alternatives forbidden and punished by law) in France, the changes to make Croatian intentionally different from Serbian, and more- Commented Oct 12, 2019 at 15:13

Well, this is no defenitive answer, but I think it ia possible, for instance: many brazilian dialects have changed the first person plural pronoun "nós" ['nɔ(j)s] for the article + noun "a gente" [ɐˈʒẽj̃.t͡ʃɪ], which literally means "the people".

The current status of "a gente" is the one of a full pronoun, always followed by a verb conjugated like the third person singular, but without adjectives treating it as a singular feminine noun as it is "in nature".

  • 2
    That doesn't make the class open. You can't just add new pronouns willy-nilly because of that.
    – CJ Dennis
    Commented Apr 4, 2020 at 0:05
  • Yeah, you're right... Commented Apr 4, 2020 at 16:45

But have we ever observed a class of words changing from open to closed, or from closed to open? Or an intermediate "half-open" stage, where new words are accepted but not commonly used or always seen as foreign?

In this example a full verb acts like a modal:

brauchen = need (full verb)

du brauchst nicht (zu) kommen
you need not (to) come
you need not come

cf. sollen (clearly a modal)

du sollst nicht (*zu) kommen

This is (Austrian) German.

  • 3
    Isn't that just grammaticalisation?
    – curiousdannii
    Commented Nov 27, 2017 at 3:33
  • 3
    yes. i think grammaticalisation is the term op is looking for
    – purlupar
    Commented Nov 27, 2017 at 21:42
  • 1
    Need is a (semi)modal in English, though it is certainly not cognate with brauchen historically. Need and dare can function as modals only in negative environments; semi-modals are negative polarity items. Members need not reply vs *Members need reply. And then there are the English periphrastic modal idioms: hafta, gotta, gonna, wanna, oughta, able to, willing to, etc; they have their own interactions with negation: He must not leave vs He doesn't have to leave and their own syntax.
    – jlawler
    Commented Feb 25, 2018 at 18:40
  • in either case, this might just be in part due to phonetics, if (the ancestor of) not-to and nichtzu respectively had contracted and finally erroded, or vice versa geminated (as the -t in not is kinda odd, as well as the grammaticalization of to). There's at least one instance of positive polarity need: "if need be" (sure, see it as a noun, as if that made any sense).
    – vectory
    Commented Jun 13, 2020 at 22:59

Yes we have, in English. Our prepositions used to be an open class (its why we have so many). New prepositions were derived using the suffixes a- be- de- and a fourth one I can't seem to recall nor find right now.

So yes, it can happen. I admit though, I have no idea why it happened in English. All I know is that it did.

Though come to think of it, the same thing may be happening to Chinese's pronouns. Thanks to communism, they've done away with their old system of pronouns, which relied heavily on their honorific system. Now there's only one word for I, one for we, etc... The older system is only still remembered because its used to stories taking place in the past, just for the sake or historical accuracy.

  • 3
    Are English prepositions not still a somewhat open class? It seems to me like there is still a tendency in English for certain words or phrases to come to be lexicalized as expressions that act like "prepositions": e.g. "despite", "regardless of". I found on Google a paper proposing a cline of more vs. less lexicalized prepositions, with only the prepositions "closer to the grammatical end of the cline" constituting a closed class: cercles.com/occasional/ops2010/bordetjamet.pdf Commented Nov 27, 2017 at 7:04
  • 1
    Also, a relevant ELU question: What are the historical processes of preposition coining in English? Commented Nov 27, 2017 at 7:07
  • Re: 'The older system is only still remembered because its used to stories taking place in the past, just for the sake or historical accuracy.': It's also used in formal letter-writing. Most people write emails in English here, but when we do write the odd letter in Chinese we stick somewhat to the old style (I do, anyway). Commented Nov 29, 2017 at 8:50
  • However, I'm not sure it used to be an open class: You can create new honorific terms using existing morphemes (e.g. sticking gui, ling etc. to nouns) but a) the words that can go through these processes seem to be more like nouns than pronouns and b) it's unclear that these morphemes are bound rather than free (wordhood is notoriously difficult to determine in Chinese after all). Commented Nov 29, 2017 at 8:53
  • Personal pronouns have various places on the referential hierarchy; first and second person (deictic) pronouns, for instance, are often deleted and indicated by syntax; e.g, Ever been to Vegas? vs Never been to Vegas. Third person pronouns can be referential or not, specific or not, definite or not, and so on. Epithets, for instance, can function pronominally: Bill caught the ball, but then the bastard ran the wrong way. This doesn't make it an open class.
    – jlawler
    Commented May 27, 2018 at 16:22

Yes, but my answer point in two directions:

  • Czech language may have lost (not so fat for the oldest generation) short predicative adjectives, only few remain as part of idioms, hence the whole class disappears.

  • answering functionally I have to point, as in the case of English prepositions to numerous and growing number (i.e. in open class) of multi-word prepositions as against the single word ones.


First of all, we have to clarify, that word classes do not tend but are closed or open.

The closed classes are determiners, pronouns, prepositions and conjunctions. All these classes have in common that they are either function or structural words. They can be seen as concepts that structure the language which surrounds these words, but on their own they are meaningless.

If we just look at prepositions this becomes pretty obvious. In for example carries no real meaning. It is a word that shows the position of something else but without any further context it makes no sense. We can now add for example the derivational unbound suffix -side in order to create Inside. Now we might understand what the other person wants to say. Inside is not not anymore a normal preposition. It can be also used as a adverb or noun because it has a meaning now. When talking about the closed class of preposition we are talking about the core, the so called prototype, prepositions and those are clearly closed and not open. Some now might argue that despite and all these other forms may be "new" prepositions but if you break them down to the core they refer to one of the prototype prepositions. In the case of despite it would be "in spite of". In order to create a new preposition you have to create a word that shows a new connection between things and unless someone can do this they should be seen as a closed class.

Thus I don´t think that any of us ever recognized a change of open to closed or closed to open classes. When you look at the dimension of language (Halliday) it would also make no sense that they change.

  • 3
    Those are the closed classes in English, but not universally/across languages: Japanese doesn't really have a closed class of pronouns (and arguably doesn't have a class of pronouns at all), but does have closed classes of verbs and verb-like adjectives, while Swahili has a closed class of noun-like adjectives.
    – Draconis
    Commented Nov 27, 2019 at 3:50
  • This is true but in the question he was talking about English. Furthermore although the classes may shift the same distinction can be found in every single language. Every language has closed and open classes. Where the closed one act as functional words or structural that hold the language together and open classes that give content. If you look at the real basis how language actually might have developed this also makes sense.
    – Jannik
    Commented Nov 27, 2019 at 10:11

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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