Personal pronouns genesis

[Zju]

Do you know of any examples of words becoming personal pronouns? I'm talking about content words here, so no this/that/that yonder → I/you/he, though this would be interesting to hear as well, since such semantic change is not that straightforward.
I know that some (south)east Asian languages use nouns instead of pronouns out of politeness, but have any actual cases of pp genesis been recorded, with the former meaning completely bleached?

[linguoboy]

Do you know of any examples of words becoming personal pronouns? I'm talking about content words here, so no this/that/that yonder → I/you/he, though this would be interesting to hear as well, since such semantic change is not that straightforward.
I know that some (south)east Asian languages use nouns instead of pronouns out of politeness, but have any actual cases of pp genesis been recorded, with the former meaning completely bleached?

Chinese 咱 zán "we [exclusive"] is a good example. In origin, it's a contraction of 自家們 zìjiāmen "own family COLL", but the etymology is so obscure that not only has a new character been created, but also it's generally used with the 們 suffix (i.e. 咱們).

[gach]

The local cases of the Nenets personal pronouns are based on the possessive forms of the postposition nya-, "at". The accusative and genitive pronouns in Tundra Nenets (in Forest Nenets only accusative) are similarly based on the stem syi- (šāʔ-/šič- in Forest Nenets) which might be related to the noun "face", compare FN šāʔ- (pron. stem) with šaʔ ("face"). Also the 2nd and 3rd person nominative pronouns use a common innovative stem pi- but I haven't read anything about it's origin.

[Ser]

Late Latin developed new third person pronouns from the intensive ipse '-self' (as in "he himself", "my sister did that herself"), which clearly survived as such in medieval Romance varieties of the Italian peninsula (they were likely the ones who brought it to the written language anyway, whereas Latin speakers from the Visigothic Kingdom probably didn't use ipse that way, as seen in modern Portuguese/Spanish/Catalan).

You can still find descendants in modern standard Italian pronouns: esso, essa, essi, esse (even though, yes, they've been largely replaced by lui, lei and loro, which come from demonstratives).

[zompist]

Portuguese você and Spanish Usted both derive from 'your mercy'.

On increasingly replaces nous in colloquial French; it derives from homo 'man'.

[Sacemd]

Dutch "jullie" (You-plural) is a compound of "je-lui" of "je-lieden" (both meaning you-people). Admittedly, it incorporates another personal pronoun, but I think a second person pronoun based on "people" alone is not inconceivable. (The disappearing Dutch pronoun "men" (one), seems to be a parallel construction of French "on" to me.)

[linguoboy]

Dutch "jullie" (You-plural) is a compound of "je-lui" of "je-lieden" (both meaning you-people). Admittedly, it incorporates another personal pronoun, but I think a second person pronoun based on "people" alone is not inconceivable.

In fact, a gente "the people" is used as a 1P pronoun in contemporary colloquial Brazilian Portuguese.

A fairly unusual example among the IE languages is colloquial Connacht Irish muid, which represents a detached verbal desinence. That is, synthetic forms like brisimid "we break" became reanalysed as briseann muid by analogy with briseann sibh "youse break" and briseann siad "they break". (Historically, the 1P pronoun was sinn, still found in Munster Irish and conservative literary forms.)

[Dewrad]

Dutch "jullie" (You-plural) is a compound of "je-lui" of "je-lieden" (both meaning you-people). Admittedly, it incorporates another personal pronoun, but I think a second person pronoun based on "people" alone is not inconceivable.

In fact, a gente "the people" is used as a 1P pronoun in contemporary colloquial Brazilian Portuguese.

A fairly unusual example among the IE languages is colloquial Connacht Irish muid, which represents a detached verbal desinence. That is, synthetic forms like brisimid "we break" became reanalysed as briseann muid by analogy with briseann sibh "youse break" and briseann siad "they break". (Historically, the 1P pronoun was sinn, still found in Munster Irish and conservative literary forms.)

Interestingly, Late Cornish does something not dissimilar: the 3pl pronoun an dzhei /ǝn dʒǝi/ is the result of improper segmentation of the 3pl desinence -ans followed by the original 3pl pronoun y.

[linguoboy]

Interestingly, Late Cornish does something not dissimilar: the 3pl pronoun an dzhei /ǝn dʒǝi/ is the result of improper segmentation of the 3pl desinence -ans followed by the original 3pl pronoun y.

There are parallels there to -nt hwy being resegmented as -n nhw in Modern Welsh, yn d'oes?

[marconatrix]

Dutch "jullie" (You-plural) is a compound of "je-lui" of "je-lieden" (both meaning you-people). Admittedly, it incorporates another personal pronoun, but I think a second person pronoun based on "people" alone is not inconceivable.

In fact, a gente "the people" is used as a 1P pronoun in contemporary colloquial Brazilian Portuguese.

A fairly unusual example among the IE languages is colloquial Connacht Irish muid, which represents a detached verbal desinence. That is, synthetic forms like brisimid "we break" became reanalysed as briseann muid by analogy with briseann sibh "youse break" and briseann siad "they break". (Historically, the 1P pronoun was sinn, still found in Munster Irish and conservative literary forms.)

Interestingly, Late Cornish does something not dissimilar: the 3pl pronoun an dzhei /ǝn dʒǝi/ is the result of improper segmentation of the 3pl desinence -ans followed by the original 3pl pronoun y.

AFAIK muid is used in Ulster Irish too. It's unknown in Scots and Manx which still use sinn. It's a rare case of an inflexion breaking free. The only other case I can think of is English possessive 's which is now more of phrase-final clitic. At least in my version of English you can say things like "[the man we saw yesterday]'s hat"

Late Cornish is often difficult to interpret as the personal ending of the verbs tend to be obscured and mashed together with the pronouns which generally follow. An jy however could at least in part be a reflex of the emphatic (reduplicated kind-of) equivalent of i 'they'. In Middle Cornish marginally attested as ensi. Equivalent to Welsh wyntwy. (I don't understand the e~wy correspondence, but cf. eneb~wyneb 'face, surface' and ebren~wybren 'sky', so there's some sort of minor rule at work here). Since the stress falls on second syllable, ens'i would be subject in Late Cornish to pretonic weakening, and voicing and palatalisation of /ns/ before /i/, and the change of final stressed /i/ > /əj/ (all independently attested), giving /ən'Zəj/ or /ən'dZəj/.

OTOH the development of -ns on the 3rd plural inflected prepositions in Late Middle Cornish seems to be copied from verbal endings, e.g. dhodhans replacing dhedhe 'to them'. Something similar happened in Welsh of course, but much earlier, while in Breton it never happened at all.

[linguoboy]

(I don't understand the e~wy correspondence, but cf. eneb~wyneb 'face, surface' and ebren~wybren 'sky', so there's some sort of minor rule at work here).

Nothing particularly "minor" about it. Late Common Brittonic stressed /eː/ yields Welsh wy (or, with a-umlaut, oe). Cf. Standard French where the same vowel diphthongised to [oɪ̯] whence it became [wɛ] or [wa] in lower-class Parisian, although still written oi. (E.g. VL *pēsu(m) > Fr. poids, W. pwys.)

[Zju]

That's intriguing. Has a language gone a step further and produced a full-fledged personal pronoun inventory out of content words, possibly replacing the old one?

[gach]

Well, Nenets appears to be nearly there as only the 1st person nominative pronouns are based on an old pronominal stem məny-. The pronoun sets in the rest of the cases are fully innovative. You see a similar development also in Enets but no further.

[So Haleza Grise]

That's intriguing. Has a language gone a step further and produced a full-fledged personal pronoun inventory out of content words, possibly replacing the old one?

I don't know Japanese well, but I understand this occurs there; boku for example is a somewhat informal, assertive "I", but it originally meant "servant."

I know even less Vietnamese, but I once heard in class that there is an argument that personal pronouns in the sense of independent lexemes are entirely missing in everyday spoken language; instead one simply uses terms such as "uncle," "older sister" etc. depending on social relations. As I understand it, common-nouns-as-pronouns is a feature of several East Asian languages. See also Chinese honorifics.

[linguoboy]

I know even less Vietnamese, but I once heard in class that there is an argument that personal pronouns in the sense of independent lexemes are entirely missing in everyday spoken language; instead one simply uses terms such as "uncle," "older sister" etc. depending on social relations.

That argument has been made for Japanese as well. Such "first-person pronouns" as (w)atashi could equally be analysed as meaning "self", and in any case it's not unusual to refer to oneself by name or by kinship relation.

This happens in Korean as well, but there's no real argument to be made there that 나 /na/ and 저 /ce/ aren't 1S pronouns comparable to those of other languages.

[----]

Even the 'generic' first person pronoun in Vietnamese (tôi) is just a noun that means 'servant'. There are some words in Vietnamese that do seem to have an exclusively pronominal meaning, but you wouldn't want to use them in polite conversation; they're all either pejorative or really casual/intimate. So the claim about Vietnamese is more or less true.

[Zhen Lin]

The inherited personal pronouns (1P wa-, 2P na-, 3P si-, Q ta-, REFL ono-) from Old Japanese are more-or-less extinct. Some of the personal pronouns that appeared in the interim have fairly transparent etymologies – as already mentioned, 1P boku ← 僕 servant (a Sinitic loanword!); watakushi ← 1P 私 private – and also 2P kimi ← 君 ruler, sovereign; 2P omae ← お前 in front; 2P anata ← 彼方 yonder (locational!); 1P ore ← 己 onore REFL.

[vec]

Icelandic like so many other Northern European languages uses maður 'man' as the unpersonal pronoun, but people use it a lot to refer to themselves. I could see a future Icelandic where it replaces the first person pronoun, at least in a casual register. It's often used when telling stories or when describing feelings. I think Icelanders like to detach themselves from their feelings as much as they can. A victim speaking out on the media describing their reaction will almost unvariably use maður instead of ég.

[hwhatting]

There's a similar thing happening with German man, and zompist mentioned a similar tendency for French on. I son't hink that it has progressed as far in German as you describe in Icelandic - it is mostly used to frame one's own experiences and judgments as something shared by the audience or by society as a whole.

[merijn]

In Dutch (and in English?) there is similarly a tendency amongst some people to use "je" to describe actions done by the speaker. "Je" means "you" but like in English it often has a more general meaning, akin to "one". For Dutch people who don't know what I am talking about, this is an example that I made up, but that could have been said by the coach of Feyenoord (the current coach of Feyenoord does this a lot, at least in my mind) after a game in which they were behind with 1-0 and he chose to substitute an extra striker:

"Je bent trainer van Feyenoord, je staat met 1-0 achter, dus dan breng je een extra spits in om toch de winst binnen te halen"
"You're coach of Feyenoord, you're behind with 1-0, so you substitute an extra striker in order to win the game"

I don't think "je" here has 1st person singular meaning, it feels to me more like a pragmatic trick to convince people that everybody would have done what the speaker had done in the same situation. Note that in this example, the actions are described in the present tense, even though they describe actions that happened in the past. I think you cannot use "je" this way in the past tense.

[gach]

I don't think "je" here has 1st person singular meaning, it feels to me more like a pragmatic trick to convince people that everybody would have done what the speaker had done in the same situation. Note that in this example, the actions are described in the present tense, even though they describe actions that happened in the past. I think you cannot use "je" this way in the past tense.

Yeah, that looks like the you-passive, or the Mika Häkkinen passive if you will.

More on the line of the French on and the Icelandic maður there's also the north Finnic use of the indefinite person for the marking more definite persons (PL1 in colloquial Finnish and PL3 in Karelian). However, there are no directly corresponding pronouns so the old pronouns stay in use.

[Melteor]

In Dutch (and in English?) there is similarly a tendency amongst some people to use "je" to describe actions done by the speaker. "Je" means "you" but like in English it often has a more general meaning, akin to "one".

As an aside, I'm pretty sure that in English at least, any plural pronoun can be used as an indefinite/impersonal pronoun.

[KathTheDragon]

I think 'you' is the most common. "You'd do this"

[Travis B.]

I think that is specific to you, actually.

Note that you is no longer plural in everyday speech in many English dialects today, which have innovated their own 2nd pl. pronouns to replace it outside of formal speech (even though they may still use you therefor after a dedicated 2nd pl. pronoun has been used).

[Richard W]

There's a similar thing happening with German man, and zompist mentioned a similar tendency for French on. I son't hink that it has progressed as far in German as you describe in Icelandic - it is mostly used to frame one's own experiences and judgments as something shared by the audience or by society as a whole.

One notices the same process in English.