Zompist Bboard Again

Starting your article "This Ancient Language..." makes you sound like an idiot. Hopefully the title was written by an intern and not the author themselves.

zompist wrote: ↑Sun May 21, 2023 4:01 pm I'd take the author's musings about early human grammar with a barrel of salt, though. No language is preserved in amber, and we are so far from the origins of language that we really know almost nothing about it.

Oh, yes. Disappointingly, I’m pretty sure I also spotted it in the reference grammar itself (though I can’t recall where).

Is unconditional l > r more common than unconditional r > l?

Original Algonquian *r (based on old attestations) has become *l in most cases

Am i right in assuming the pronounciation *ʃri: in sri lanka is influenced by *t and *d being palatal affricates before *r.

foxcatdog wrote: ↑Wed May 24, 2023 10:06 pm Am i right in assuming the pronounciation *ʃri: in sri lanka is influenced by *t and *d being palatal affricates before *r.

As far as I can see, it’s because the original Sinhalese is itself ශ්‍රී ලංකා Śrī Lankā. The first word is a standard Sanskrit honorific: https://en.wikipedia.org/wiki/Shri

Maybe sriracha was a better example

And now that i think about it we have shrink which also might have influenced it.

foxcatdog wrote: ↑Thu May 25, 2023 2:28 am Maybe sriracha was a better example

Sriracha is completely unrelated; I’ve never used the word myself, but Wikipedia has /sɪˈrætʃə/, so it’s pronounced differently too. (And incidentally, also similarly to its source: Thai ศรีราชา [sǐː.rāː.tɕʰāː].) I don’t know how the first ⟨r⟩ got into the spelling.

EDIT: According to Merriam–Webster, the city of Sriracha is alternately Sri Raja. Obviously this has been borrowed from Sanskrit through Thai, losing the consonant cluster in the process (as with most Sanskrit borrowings in Thai). However, the heavily redundant Thai orthography retains the ⟨sr⟩ in written form, leading to the romanisation Sriracha.

foxcatdog wrote: ↑Wed May 24, 2023 10:06 pm Am i right in assuming the pronounciation *ʃri: in sri lanka is influenced by *t and *d being palatal affricates before *r.

I should note that initial /sr/ is not found in native English words but /ʃr/ is.

Travis B. wrote: ↑Thu May 25, 2023 4:23 pm

foxcatdog wrote: ↑Wed May 24, 2023 10:06 pm Am i right in assuming the pronounciation *ʃri: in sri lanka is influenced by *t and *d being palatal affricates before *r.

I should note that initial /sr/ is not found in native English words but /ʃr/ is.

... and that goes back a lot further than the affrication of /t/ and /d/ before /r/.

An example of a non-English name which actually does begin [sr] is Srebrenica. I think that is also often pronounced with /ʃ/ by English speakers.

Is a in a priori and a posteriori a morpheme of English? Is the concept of 'morpheme' perhaps too fractured to have a yes/no answer? Does the answer depend on the user's educational history?

Richard W wrote: ↑Sat May 27, 2023 10:34 am Is a in a priori and a posteriori a morpheme of English? Is the concept of 'morpheme' perhaps too fractured to have a yes/no answer? Does the answer depend on the user's educational history?

I conceive of that as a set element of the phrase.

Richard W wrote: ↑Sat May 27, 2023 10:34 am Is a in a priori and a posteriori a morpheme of English?

I don't think so.

Richard W wrote: ↑Sat May 27, 2023 10:34 am Is a in a priori and a posteriori a morpheme of English?

Let's start with an easier example— how many morphemes are in hors d'oeuvre or laissez-faire? Probably just one, because unless you know French they are not decomposable or obviously related to particular English words. They're learned and used as a unit.

How about bon mot? Again, it's usually learned as a unit. But the meaning of bon is perhaps more widely known, and reinforced by bonjour, bon vivant, bon voyage, bon appétit. It's still heavily marked as foreign, though.

How about le? What's striking about this one is that it's been extended to English words, at least in brand names. E.g. there was a car named Le Car (that's not how you say 'the car' in French), there's a store called Le Shop, there's a glasses maker called Le Specs.

A priori should maybe be contrasted with a capella. Though capella has an English cognate, it doesn't automatically come to mind, especially given the meaning ('without instruments'). But a priori is obviously related to prior. So this might be more like film noir— it's obvious that this is a kind of film, so the phrase is partially analyzable, and in fact this leads to noir being borrowed on its own. Do people ever connect the a to phrases like à la carte, à propos?

In short, we should probably start with the assumption that these are learned and used as single units, but the more analyzable a phrase is, the more likely that the other parts get at least some marginal meaning. If you can extend it to new contexts it's definitely a morpheme. (But morphemes of course don't have to be extensible.)

zompist wrote: ↑Sat May 27, 2023 4:41 pm

Richard W wrote: ↑Sat May 27, 2023 10:34 am Is a in a priori and a posteriori a morpheme of English?

Let's start with an easier example— how many morphemes are in hors d'oeuvre or laissez-faire? Probably just one, because unless you know French they are not decomposable or obviously related to particular English words. They're learned and used as a unit.

How about two, twelve, twenty, twice and twin? Do they have a common morpheme? (Remember that most English speakers are literate.)

If sent is a morpheme in dissent, then hors d'oeuvre is relatable to hors de combat, de Whalley, and oeuvre. We also have laissez-passer and savoir-faire; the latter is incidentally backed up by savvy.

zompist wrote: ↑Sat May 27, 2023 4:41 pm How about bon mot? Again, it's usually learned as a unit. But the meaning of bon is perhaps more widely known, and reinforced by bonjour, bon vivant, bon voyage, bon appétit. It's still heavily marked as foreign, though.

How about le? What's striking about this one is that it's been extended to English words, at least in brand names. E.g. there was a car named Le Car (that's not how you say 'the car' in French), there's a store called Le Shop, there's a glasses maker called Le Specs.

And longer established place names, like Chapel-en-le-Frith, and obscurely in names like Barton-le-Clay, and also an allomorph in Ashby de la Zouche.

zompist wrote: ↑Sat May 27, 2023 4:41 pm A priori should maybe be contrasted with a capella. Though capella has an English cognate, it doesn't automatically come to mind, especially given the meaning ('without instruments'). But a priori is obviously related to prior. So this might be more like film noir— it's obvious that this is a kind of film, so the phrase is partially analyzable, and in fact this leads to noir being borrowed on its own. Do people ever connect the a to phrases like à la carte, à propos?

I think it's seen as a distinct morpheme. After all, they get tagged as being in the Latin and French subsystems. And à la is allegedly an English preposition, albeit unusual. It has restrictions on the form of the governed noun phrase, and weird phonetics, at least in England.

zompist wrote: ↑Sat May 27, 2023 4:41 pm In short, we should probably start with the assumption that these are learned and used as single units, but the more analyzable a phrase is, the more likely that the other parts get at least some marginal meaning. If you can extend it to new contexts it's definitely a morpheme. (But morphemes of course don't have to be extensible.)

Richard W wrote: ↑Sat May 27, 2023 11:01 pm How about two, twelve, twenty, twice and twin? Do they have a common morpheme? (Remember that most English speakers are literate.)

I think so, though that leaves -lve and -in unanalyzable. (Of course, not all morphemes are transparent!)

If sent is a morpheme in dissent, then hors d'oeuvre is relatable to hors de combat, de Whalley, and oeuvre. We also have laissez-passer and savoir-faire; the latter is incidentally backed up by savvy.

Some of that seems like a stretch... I mean, these things are evident if you know some French, but maybe it only means you know some French, not that they are English morphemes.

Even if you know oeuvre, and the meaning of hors de combat, that gives you "out of work", which is etymology, not semantics. Would looking at maneuver help or confuse...?

Linguoboy wrote: ↑Mon May 08, 2023 9:59 am

MacAnDàil wrote: ↑Sun May 07, 2023 11:18 amAnother possibility would be a loanword between the two Celtic subgroups.

Sure, okay, let's see the evidence. The diachronics of lexical loans between Irish and Welsh has been extensively studied; show us a plausible etymology which explains the existing attestations and history of variation in both languages.

I tried to give an example of how onomatopeic etymologies seem overused.

You gave one concrete example from one language pair on the literal other side of the world. Forgive me if I need a little more convincing than that.

You have appropriately underlined the importance of providing more relevant sources. I have therefore consulted such sources.

According to Alexander MacBain’s An Etymological Dictionary of the Gaelic Language, “<ach>, interjection of objection and impatience; founded on above
with leaning upon [och]”

‘Above’ here refers to the Gaelic word ‘ach’ “but”.

‘Och’ refers to the following word with the following etymology: “<och>, an interjection, alas! Ir. [och], [uch], O.Ir. [uch], vae, [ochfad],
sighing: [*uk]; Got. [aúhjôn], make a noise, Norse [ugla], Eng.
[owl]; Let. [auka], stormwind, Srb. [uka], a cry.”

We may therefore consider the Gaelic interjection ‘ach’, related to another Gaelic interjection, ‘och’, itself of Celtic, indeed Indo-European, stock.

For Welsh, I hereby refer to the following dictionary:
https://www.welsh-dictionary.ac.uk/

Here also, the reader of the entry for the interjection ‘ach’ is referred another interjection taking the form of ‘och’.

The origin of the Welsh interjection ‘och’ appears to also be Celtic, given that the reader is referred to several other Celtic languages at various periods: Medieval Cornish ‘ogh’, Old Irish ‘uch’, Medieval Irish ‘ach, och’, Medieval Breton ‘ach’ and Modern Breton ‘ac’h’.

I consider, therefore, that there is no coincidence that interjections with similar forms and meanings (‘ach’ and ‘och’) exist in closely related neighbouring languages (Gaelic and Welsh) while being absent from the vast majority of languages in the world.

I therefore thank you for having piqued my interest in this matter, notably in Welsh etymology which I hitherto know very little of.

MacAnDàil wrote: ↑Tue May 30, 2023 8:09 am According to Alexander MacBain’s An Etymological Dictionary of the Gaelic Language
For Welsh, I hereby refer to the following dictionary:
https://www.welsh-dictionary.ac.uk/

No serious linguist relies on McBain anymore; it's long outdated and contains a lot of speculative and nowadays rejected etymologies. The only reason it's still regularly referred to, mostly by amateurs, is that it's out of copyright and you can get reprints cheaply. I also have it at home, but would never rely on it. GPC is a much better source.

MacAnDàil wrote: ↑Tue May 30, 2023 8:09 am I consider, therefore, that there is no coincidence that interjections with similar forms and meanings (‘ach’ and ‘och’) exist in closely related neighbouring languages (Gaelic and Welsh) while being absent from the vast majority of languages in the world.

Have you checked which other languages have comparable interjections? FWIW, both German and Russian have each ach / ах and och / ох, with similar meanings. At least the Russian and the German interjections cannot be cognates with each other.
I'm not saying that the Celtic interjections can't go back to a common source, just that it's not excluded that this specific sound sequence may be an attractor for expressing a specific range in meanings in interjections.

MacAnDàil wrote: ↑Tue May 30, 2023 8:09 amFor Welsh, I hereby refer to the following dictionary:
https://www.welsh-dictionary.ac.uk/

Here also, the reader of the entry for the interjection ‘ach’ is referred another interjection taking the form of ‘och’.

What the etymological portion of the entry says is actually (in its entirety), "[cf. och]". No actual claims are being made here about the relationship between the two words. If you look at the entry for the word ych, it similarly reads "[cf. ach², och]". (It's also worth noticing that the entries for ach and ych specify pronunciations--something not normally done for native Welsh words because of how shallow the orthography is. This is a signal that these words are exceptional in some way.)

Notice further that the definitions do not line up: The primary English gloss for och is "woe!, alas!, O!, oh!" whereas for ach and ych it is "yuck, ugh"[*]. Ych is also used to gloss ach, so I think it's safe to consider these variants of each other. But "woe!" and "yuck!" are not interchangeable or even that closely related. Note that the second definition for och refers to its use "before the name of God (Christ, &c.) or a saint, in an oath", which is not where you would anticipate an expression of disgust.

MacAnDàil wrote:The origin of the Welsh interjection ‘och’ appears to also be Celtic, given that the reader is referred to several other Celtic languages at various periods: Medieval Cornish ‘ogh’, Old Irish ‘uch’, Medieval Irish ‘ach, och’, Medieval Breton ‘ach’ and Modern Breton ‘ac’h’.

Again, all the etymological entry really says is "cf.". The lexicographers are not taking any sort of stand on whether the similarity in form and meaning is due to coincidence, borrowing, or common inheritance. (I've already explained why the last of these is problematic due to the fact that Common Brythonic *x and Common Goidelic *x simply don't derive from the same proto-sources.)

MacAnDàil wrote:I consider, therefore, that there is no coincidence that interjections with similar forms and meanings (‘ach’ and ‘och’) exist in closely related neighbouring languages (Gaelic and Welsh) while being absent from the vast majority of languages in the world.

MAJOR question-begging unless you have actually spent hours searching lexica of the "vast majority of languages in the word" looking for " interjections with similar forms and meanings".

Looking closer at the data, I think what we actually have here for both languages is at least two different interjections: one of earlier origin with a meaning of "woe! alas!" with possible origin as an onomatopoeia for a sigh or groan (both additional definitions of the Welsh term och and the Irish/Gaelic och[**]) and a later one (not attested before the 18th century in Welsh) of definite onomatopoeic origin (due to the irregularities in pronunciation) which arose separately in each language. The long attestation of och in the literary forms of both languages[***] actually strengthens the argument for historical borrowing.

[*] The Welsh gloss reads "Interjection expressing disgust or strong loathing".
[**] See https://www.teanglann.ie/en/fgb/och for Irish and https://www.faclair.com/ViewEntry.aspx? ... 49E9195711 for Gaelic.
[***] Scottish Gaelic derives from Middle Irish, the most comprehensive dictionary for which is the Dictionary of the Irish Language.

Don't forget English agh /ɑːx/ and ugh /ʌx/, which are words with very similar meanings and are practically sole survivors of /x/ outside of Scottish English, Hiberno-English, and recent loanwords, and are almost certainly onomatopoeic in nature.