Loan words with more specific meanings after than before the borrowing

[Rounin Ryuuji]

It's interesting to think how words we speak but don't often write will probably have descendants that are more commonly used (but also a fairly predictable outcome); I'm now amused with the thought of future people wondering the same things about future forms of English (Anglic?), which is sure to have a great degree of variation.

Yeah... It really disturbs me how in the entire pre-Classical and Classical Latin literary corpus (not counting inscriptions/tablets/graffiti), as collected in PHI Latin Texts (which has... close to all of it), the word panticēs 'belly' is only attested six times, all of them in vulgar texts:

...

And yet it's perfectly in use as a colloquial word all over the place today, in Portuguese/Spanish (pança/panza, as in Sancho Panza), French (panse), Italian (pancia, pancetta) and even freaking Dalmatian (panzaita, note: now extinct) and Romanian (pântece).

Meanwhile, from among the synonyms venter, alvus and the poetic īlia, which are so much more common in writing... only venter survives (Spanish vientre, Dalmatian viantro, Romanian vintre).

I don't think it's particularly disturbing, or even surprising. I can think of a whole litany of words I wouldn't use in that sort of writing (especially my own fiction), even if they are very common. It makes me think panticēs may have, in some registers, had vulgar connotations, or simply sounded unsuitable (I wouldn't use "tummy" outside of a book for children, and I have some dispreference for "belly", too, though I can't explain exactly why). If literacy had been more widespread, we probably would've seen writing in a far greater variety of registers, but literacy being, to my understanding, mostly an upper-class thing at the time, it was mostly that sort of language that was written down (there were, if I remember right, also authors who commented on the fact that written Latin was more archaic and "decorated" than the language people actually spoke).

And sometimes, words and expressions stay uncommon in the written language over multiple centuries. One of our oldest attestations of the modern sense of "I guess" is Chaucer, after all. Experience, though no authority were in this world, would tell us people stratify their language even with no central body trying (with whatever degree of success) to regulate its usage.

[Kuchigakatai]

I don't think it's particularly disturbing, or even surprising. I can think of a whole litany of words I wouldn't use in that sort of writing (especially my own fiction), even if they are very common. It makes me think panticēs may have, in some registers, had vulgar connotations, or simply sounded unsuitable (I wouldn't use "tummy" outside of a book for children, and I have some dispreference for "belly", too, though I can't explain exactly why). If literacy had been more widespread, we probably would've seen writing in a far greater variety of registers, but literacy being, to my understanding, mostly an upper-class thing at the time, it was mostly that sort of language that was written down (there were, if I remember right, also authors who commented on the fact that written Latin was more archaic and "decorated" than the language people actually spoke).

Maybe you could say I find it unsurprising rationally (or well, that it should be unsurprising, for the reasons you say), but very surprising viscerally. You rarely see panticēs in Latin at all, and yet it's so widespread in modern Romance, and has certainly always been.

And yeah, there's Cicero who at some point asks to his friend Atticus in a letter if he doesn't feel he (Cicero) is familiar and un-ornate in his personal letters. It's true he writes in rather short sentences in his letters (I once heard a Latin learner, who had read some formal Cicero stuff before, say: "oh, so Cicero could speak normal!!"). And I also remember once coming across a paper that claimed SVO is the most common word order in his letters too (something like 55% of sentences with a subject and object), while in his philosophical writings it's SOV (with something like 70%).

[Rounin Ryuuji]

I don't think it's particularly disturbing, or even surprising. I can think of a whole litany of words I wouldn't use in that sort of writing (especially my own fiction), even if they are very common. It makes me think panticēs may have, in some registers, had vulgar connotations, or simply sounded unsuitable (I wouldn't use "tummy" outside of a book for children, and I have some dispreference for "belly", too, though I can't explain exactly why). If literacy had been more widespread, we probably would've seen writing in a far greater variety of registers, but literacy being, to my understanding, mostly an upper-class thing at the time, it was mostly that sort of language that was written down (there were, if I remember right, also authors who commented on the fact that written Latin was more archaic and "decorated" than the language people actually spoke).

Maybe you could say I find unsurprising rationally (or well, that it should be unsurprising, for the reasons you say), but very surprising viscerally. You rarely see panticēs in Latin at all, and yet it's so widespread in modern Romance, and has certainly always been.

Its frequency probably also did vary with time and place; it might, in the idiolects of those specific authors, actually have been an uncommon term for all we know. I often hear people these days use "anymore" to mean "nowadays", which, while I understand it, registered as the guy from whom I first heard it trolling (I even asked him why he was using it that way, and he claimed it was "standard"; I thought he was making some sort of joke I wasn't getting); it wasn't until I heard it from other people that I understood it was pretty common, even though it is, to me, idiodialectally, ungrammtical. I could think of several uses like was sat and done gone that, while common, are not in any register I personally use.

And yeah, there's Cicero who at some point asks to his friend Atticus in a letter if he doesn't feel he (Cicero) is familiar and un-ornate in his personal letters. It's true he writes in rather short sentences in his letters (I once heard a Latin learner, who had read some formal Cicero stuff before, say: "oh, so Cicero could speak normal!!"). And I also remember once coming across a paper that claimed SVO is the most common word order in his letters too (something like 55% of sentences with a subject and object), while in his philosophical writings it's SOV (with something like 70%).

That would make some sense. The word order was probably already in the process of shifting, but educated and literary varieties of whatever language seem to tend to prefer older forms; some speakers also simply have more conservative dialects and idiolects than others. Compound that with an admiration of older writing (which seems a commonplace thing cross-culturally), and it isn't exactly surprising that most languages with a written tradition show some form of diglossia. It would have been very good of the ancient Romans to document it more.

[Talskubilos]

That's pretty interesting, but why is Coromines & Pascual's etymology pretty bad? It looks pretty possible: I notice that in my dialect (El Salvador), colcha primarily refers to some kind of simple makeshift bed (and never 'quilt', we only use edredón for that), that colchón is primarily a matress, and that culcita > Mozarabic [ˈkoltʃa] seems slightly problematic phonetically although it's by all means possible (via [ˈkoltʃeta] surely, and cf. Sanctium > Spanish Sancho while sanctum > santo).

Actually, Sancho would be the native Spanish output of Latin sanctu(m), while santo is a later loanword from Church Latin. Interestingly enough, Basque saindu shows the usual palatalization of the nasal in /n/+plosive groups.

There's also not much of a problem with colloc-am > colcha via French/Occitan, considering this kind of collusion of alveolar+velar produces alveolar+postalveolar quite often (monacum > Occitan monge, borrowed by Spanish as monje, and -āticum > French/Occitan -age, Old Spanish -adgo > modern -azgo, diem dominicum > French dimanche, and even corticem > corcho is an example of this).

IMHO, trying to explain ALL the cases of palatalization as Occitan/French loanwords doesn't work.

[Linguoboy]

There's also not much of a problem with colloc-am > colcha via French/Occitan, considering this kind of collusion of alveolar+velar produces alveolar+postalveolar quite often (monacum > Occitan monge, borrowed by Spanish as monje, and -āticum > French/Occitan -age, Old Spanish -adgo > modern -azgo, diem dominicum > French dimanche, and even corticem > corcho is an example of this).

IMHO, trying to explain ALL the cases of palatalization as Occitan/French loanwords doesn't work.

But he's not proposing we explain all cases by invoking Occitan/French, just this one. (And of all the terms to borrow from French/Occitan, this hardly seems like a farfetched one.)

[Talskubilos]

But he's not proposing we explain all cases by invoking Occitan/French, just this one.(And of all the terms to borrow from French/Occitan, this hardly seems like a farfetched one.)

In order to explain /ʧ/, linguists often recurr to a phonosymbolic origin to explain when a French loanword isn't possible. My point is French has only /ka, ga/ > /ʧa, ʤa/, whereas palatalizations were apparently more general in Mozarabic (cfr. the above mentioned corcho). In this particular case, the word is semantically more widespread in Basque than Spanish colcha, colchón:
koltxa (R) 'pincushion'
kolxa, kolxe, kolxoi(n) (L, LN) 'bed linen (matress, pillow, blanket); quilt (embroidered)'
kultxoin (L) 'matress'
kurtxoin (L) 'little quilt, blanket, carpet, matress'

[Zju]

My point is French has only /ka, ga/ > /ʧa, ʤa/

Isn't colchón just an augmentative of colcha, though? All the Basque words you listed and corcho can be derived from those two Spanish words.

[Kuchigakatai]

Actually, Sancho would be the native Spanish output of Latin sanctu(m), while santo is a later loanword from Church Latin. Interestingly enough, Basque saindu shows the usual palatalization of the nasal in /n/+plosive groups.

That doesn't seem right.

cincta, cinctūra > cinta, cintura
unctāre > untar
punctus > punto
iunctus, iunctāre > junto, juntar
plānctus > llanto
tinctus > tinto

But punctiāre > pinchar (apparently formerly punchar). I am unable to find any other inherited instance of -nct- > -nch-. Most native -nch- seem to come from -NCl- rather, possibly after syncope: amplius > ancho, īnflāre > hinchar, cingulum > cincho, macula > *mancula > mancha.

I think I'm still leaning towards Sanctius > Sancho, as it is "traditionally" done, more so seeing -ius is an existing affix in Latin in Roman nomina (e.g. Sallustius, Livius), and seeing that in Late Latin many of those became given names (bishop Aurelius Augustinus).

It's interesting you bring up sanctum as being something like [saɲto] though, at least at some point. I'd say the word isn't learned but probably "semi-learned" (i.e. inherited but with a pronunciation influenced by the Latin spelling) in Old Spanish; it is spelled <sancto> way more often than <santo>... I've wondered this exact question about a possible [-ɲt-] in French though. Latin sanctus > saint is a very strange development if I'm supposed to take <ai> as [ai]>[ɛi] (that only happens in open syllables otherwise, sānus > *sainu > sain), but it makes more sense if <saint> was [sãɲt] in Old French. Cuneus > coin and other words shows -in likely stands for final [ɲ], much like how -il stood for final [ʎ], e.g. travail [tɾaˈvaʎ] (which we know with certainty because this survived down to the 19th century!). So maybe saint is [sãɲt], developing later as [sãĩt] > [sɛ̃ĩt] > [sɛ̃(t)] anyway.

IMHO, trying to explain ALL the cases of palatalization as Occitan/French loanwords doesn't work.

Of course. I totally give credence to the idea Mozarabic or some other form of Romance south of the Cantabrian mountains had /tʃ/ in such cases, and even corresponding to Hispano-Romance /ts/ elsewhere. The other day I saw that the kharja #5 on this website even has an instance of "qrŷwn" (in the 2nd manuscript, a transliteration that I presume stands for قرجون) which has been interpreted as *koraˈtʃon or *koraˈdʒon, 'heart', Old Spanish coraçon.

García de Diego and you are making me at least definitely suspicious of the "traditional" colloc-a > Spanish colcha via French/Occitan. Not sure if French/Occitan would borrow culcita > Mozarabic colcha from Iberia, but maybe...? It is semantically more satisfying at least (as you said).

My point is French has only /ka, ga/ > /ʧa, ʤa/

Isn't colchón just an augmentative of colcha, though? All the Basque words you listed and corcho can be derived from those two Spanish words.

(Corcho from corticem seems to uncontroversially be from Mozarabic though.)

[Talskubilos]

Actually, Sancho would be the native Spanish output of Latin sanctu(m), while santo is a later loanword from Church Latin. Interestingly enough, Basque saindu shows the usual palatalization of the nasal in /n/+plosive groups.

That doesn't seem right.

cincta, cinctūra > cinta, cintura
unctāre > untar
punctus > punto
iunctus, iunctāre > junto, juntar
plānctus > llanto
tinctus > tinto

But punctiāre > pinchar (apparently formerly punchar). I am unable to find any other inherited instance of -nct- > -nch-. Most native -nch- seem to come from -NCl- rather, possibly after syncope: amplius > ancho, īnflāre > hinchar, cingulum > cincho, macula > *mancula > mancha.

That's right, but *punctiare is a Proto-Romance (Vulgar Latin) form, not a Classical one.

I think I'm still leaning towards Sanctius > Sancho, as it is "traditionally" done, more so seeing -ius is an existing affix in Latin in Roman nomina (e.g. Sallustius, Livius), and seeing that in Late Latin many of those became given names (bishop Aurelius Augustinus).

Actually, Sanctius is a retroformation (i.e. Latinization) from Sancho, not a genuine Latin word. Apparently. it was originated in Leonese sancho 'rabbit', a pre-Latin word corresponding to Basque untxi id. I'm a bit rusty these years.

García de Diego and you are making me at least definitely suspicious of the "traditional" colloc-a > Spanish colcha via French/Occitan. Not sure if French/Occitan would borrow culcita > Mozarabic colcha from Iberia, but maybe...? It is semantically more satisfying at least (as you said).

Alibert's dictionary of Lengadocian shows the form colca from the same etymon, whereas the Latin variant culcitra gave Spanish cólcedra, now in desuse. So I'm afraid Coromines was right and García de Diego was wrong.

[Otto Kretschmer]

Polish puszka which means a can was borrowed into Yiddish as pushke in a narrower sense of a can for collecting charity and the word was then borrowed into some Jewish English varieties

[Linguoboy]

Polish puszka which means a can was borrowed into Yiddish as pushke in a narrower sense of a can for collecting charity.

Depending on its range of meaning, puszka itself might represent a narrowing of Old High German buhsa (cognate to English "box").

[Otto Kretschmer]

In English the word cigarette means any kind of cigarette. In Polish a similar borrowed word, cygaretka, means a specific type of a cigarette

[Vijay]

In Arabic, كتاب kitāb just means 'book', but in Malayalam, Muslims apparently use [kiˈt̪aːbɯ] to specifically mean the Qur'an. Alkitabu in Indonesian (and Malay?) appears to mean 'Bible'.

[Creyeditor]

In Indonesian I mostly hear <alkitab> /alkitap/.

[Linguoboy]

In Arabic, كتاب kitāb just means 'book', but in Malayalam, Muslims apparently use [kiˈt̪aːbɯ] to specifically mean the Qur'an. Alkitabu in Indonesian (and Malay?) appears to mean 'Bible'.

Indonesian has three different borrowed words for "book": pustaka, kitab, and buku. Pustaka (< Sanskrit पुस्तक), the term of longest standing, is now found chiefly in higher registers and in derived terms like perpustakaan "library". It can also refer specifically to a primbon or divination text. Kitab is the preferred term for sacred works (usually Muslim, but even a primbon cab be called a kitab). Though it can also be used generically, the most common general term for "book" is contemporary Indonesian is buku (< Dutch boek).

[Otto Kretschmer]

Polish puszka which means a can has been borrowed as pushke into Yiddish with a narrowed meaning of a can for collecting charity and ha made it's way into some dialects of Jewish English

[Chuma]

I've also noticed on some international food packaging, the word Keks on what in English would be called "cookies" or "(sweet) biscuits", depending on where you learnt it.

Swedish kex and Finnish keksi do come from English "cakes", and mean cookie. In Swedish it can also mean wafer.

Further fun fact: The English word "cake" is itself an import from Scandinavia. So Scandinavian now has both kaka/kake/kage/kakku "cake, cookie", and kex/kjeks/kiks/keksi "cracker, biscuit, wafer, cookie".

[Qwynegold]

Huh, I never knew that about cake.

[Rounin Ryuuji]

That made me look up what the English cognate was. Apparently (according to Wiktionary, so keep your pinch of salt handy), it's related to cheek, and also cycel, which would've probably yielded a modern word kitchel.

[Ryusenshi]

Some examples where a generic word in language X was borrowed for "thing in the style of country X".

• EN chicken → FR chicken "chicken nugget" (colloquial only; the correct form is "nuggets de poulet". I blame McDonald's.)
• EN gospel → FR gospel "black gospel music, spirituals"
• EN cartoon "simple drawing style (that can be printed or animated)" → FR cartoon "American animated shorts" (think Looney Tunes, or Tom & Jerry)
• EN comics → FR comics "American-style comics", especially "superhero comics" (as opposed to the more general word "bande dessinée")
• FR Suède "Sweden" → EN suede "a type of deer leather invented in Sweden"