Zompist Bboard Again

Pabappa wrote: ↑Thu Jun 18, 2020 8:20 pmedit: oh by the way, i love correveidile. there dont seem to be too many Spanish words like that.

There are a few others:

bienmesabe ("[it] tastes good to me") a kind of dessert
hazmereír ("make me laugh") laughingstock
metomentodo ("[I] stick myself in everything") meddler
vaivén ("[it] goes and comes") swaying

All of these can be pluralised as well, e.g. vaivenes "ups-and-downs". One name used for them in the literature is "syntactic freezes" and they've been compared to English pre-noun inserts like "a you've-got-to-be-kidding-me look on her face".

bradrn wrote: ↑Thu Jun 18, 2020 8:42 pmAh, right. So most non-IE languages keep these constructions separate.

But actually, now that I think about it: how many non-IE languages have a embedded questions in the first place?

I'm pretty sure it's fairly common. Both Standard Arabic and Mandarin have indirect/embedded questions at least, and yeah, both differentiate them fairly clearly from "headless relative clauses". (Or "fused relative noun phrases" if you prefer... I'm not much of a fan of the common term.)

For what it's worth, Spanish and French also differentiate them some of the time, just not all of the time. Quiero lo que me mostraste 'I want what you showed to me' ("headless relative clause"), No sé qué me mostraste 'I don't know what you showed to me' (embedded question). On the other hand, subclauses with cuando/cuándo are used both ways though: Recuerdo cuando lo viste 'I remember when you saw it' ~ No sé cuándo lo viste 'I don't know when you saw it' (although even here there is arguably some tendency to differentiate them with sentence-level stress; the written acute accent on the embedded question pronoun cuándo doesn't actualy feel unnatural).

This reminds me of Walmatjari, which I was reading about yesterday. Walmatjari marks personal agreement and mood on an obligatory auxiliary rather than the verb itself:

yani marna I went
yani pajarra We 2 went
yani pa He went
yani pila They two went
etc.

(Note that the verb here is yani; the auxilliary is stated as usually being the second word in the sentence.)

This also happened in the evolution of Ancient Egyptian into Coptic as the old conjugation of Old and Middle Egyptian was abandoned in favour of inflected auxiliaries in Late Egyptian, and then these auxiliaries show up merged into the verb stem in Coptic. It has had the fun result that, although linguists have been able to reconstruct the pronunciations a fair number of Middle or Late Egyptian nouns and adjectives, even if only in partial form (by having a close look at variation in Coptic / Egyptian spelling variants / Egyptian names and words in neighbour languages), the old verbal system was so heavily remodelled by the time Coptic shows up that hardly anything is reconstructible.

Linguoboy wrote: ↑Thu Jun 18, 2020 9:53 pmbienmesabe ("[it] tastes good to me") a kind of dessert
hazmereír ("make me laugh") laughingstock
metomentodo ("[I] stick myself in everything") meddler
vaivén ("[it] goes and comes") swaying

el dimequetediré ("tell me 'cuz I'll tell you") 'gossiping, sharing rumours; quarrel, angry argument' (plural los dimequetedirés)
el quedirán ("what will they say?") 'idea/opinion people in general have of something, usual opinion' (los quediranes)
el pordiosero / la pordiosera ("by-God-ist") 'street beggar' (los pordioseros)
el/la mileurista ("1000-euro-ist") 'person from/in Spain holding a bachelor's degree but earning around €1000 per month, person in underemployment' (los mileuristas)
el/la sabelotodo 'know-it-all' (los sabelotodos)
el nomeolvides 'forget-me-not' (los nomeolvides)

See also: il tiramisù ("pick me up") 'tiramisu' (sadly, it doesn't have a distinctive plural in Italian: i tiramisù).

Ser wrote: ↑Thu Jun 18, 2020 11:04 pm el pordiosero / la pordiosera ("by-God-ist") 'street beggar' (los pordioseros)
el/la mileurista ("1000-euro-ist") 'person from/in Spain holding a bachelor's degree but earning around €1000 per month, person in underemployment' (los mileuristas)

I wouldn't class these with the others, which are all full sentences on their own. By contrast, pordiosero is just agentive -ero affixed to a phrase rather than a simple noun and mileurista looks like an ordinary compound comparable to manicurista or tenismesista. Parallel coinages would be *dicepordios and *ganomileuros.

bradrn wrote: ↑Thu Jun 18, 2020 8:42 pm This reminds me of Walmatjari, which I was reading about yesterday. Walmatjari marks personal agreement and mood on an obligatory auxiliary rather than the verb itself:

yani marna I went
yani pajarra We 2 went
yani pa He went
yani pila They two went
etc.

(Note that the verb here is yani; the auxilliary is stated as usually being the second word in the sentence.)

That doesn't look so different from English, even though it does retain some direct marking, which it might very easily have lost.

Richard W wrote: ↑Fri Jun 19, 2020 12:53 pmThat doesn't look so different from English, even though it does retain some direct marking, which it might very easily have lost.

I easily envision Future English going the way of Late Egyptian. It would be really funny if it ended up like Coptic too.

I use, you use, we use.
I used, you used, we used.
I am used, you are used, we are used.

Then by the 26th century:
(From I am using) [ɑm ˈjoʊzi, jɔ: ˈjoʊzi, vja ˈjoʊzi] (colloquially spelled: am usy, your usy, vyah usy)
(From I did use) [ˌɑdi ˈjoʊz, ˌidi ˈjoʊz, ˌvidi ˈjoʊz] (ady use, edy use, vedy use)
(From I got used) [ˌɑgə ˈjoʊz, ˌigə ˈjoʊz, ˌvigə ˈjoʊz] (aga use, ega use, vega use)

Then by the 32nd century:
[aˈmo:z, joˈo:z, vaˈo:z] (colloquial spellings: am use, yo use, vah use)
[aˈdo:z, iˈdo:z, viˈdo:z] (ad use, ed use, vid use)
[aʁeˈo:z, iʁeˈo:z, viʁeˈo:z] (aga use, ega use, viga use)

Richard W wrote: ↑Fri Jun 19, 2020 12:53 pm

bradrn wrote: ↑Thu Jun 18, 2020 8:42 pm This reminds me of Walmatjari, which I was reading about yesterday. Walmatjari marks personal agreement and mood on an obligatory auxiliary rather than the verb itself:

yani marna I went
yani pajarra We 2 went
yani pa He went
yani pila They two went
etc.

(Note that the verb here is yani; the auxilliary is stated as usually being the second word in the sentence.)

That doesn't look so different from English, even though it does retain some direct marking, which it might very easily have lost.

Not really — it just looks similar since it’s pro-drop. That’s an auxiliary being inflected, rather than a pronoun.

Ser wrote: ↑Fri Jun 19, 2020 7:02 pm

Richard W wrote: ↑Fri Jun 19, 2020 12:53 pmThat doesn't look so different from English, even though it does retain some direct marking, which it might very easily have lost.

I easily envision Future English going the way of Late Egyptian. It would be really funny if it ended up like Coptic too.

I use, you use, we use.
I used, you used, we used.
I am used, you are used, we are used.

Then by the 26th century:
(From I am using) [ɑm ˈjoʊzi, jɔ: ˈjoʊzi, vja ˈjoʊzi] (colloquially spelled: am usy, your usy, vyah usy)
(From I did use) [ˌɑdi ˈjoʊz, ˌidi ˈjoʊz, ˌvidi ˈjoʊz] (ady use, edy use, vedy use)
(From I got used) [ˌɑgə ˈjoʊz, ˌigə ˈjoʊz, ˌvigə ˈjoʊz] (aga use, ega use, vega use)

Then by the 32nd century:
[aˈmo:z, joˈo:z, vaˈo:z] (colloquial spellings: am use, yo use, vah use)
[aˈdo:z, iˈdo:z, viˈdo:z] (ad use, ed use, vid use)
[aʁeˈo:z, iʁeˈo:z, viʁeˈo:z] (aga use, ega use, viga use)

My speech is actually approaching this — the perfective aspect and copula are already fusing with the subject pronouns:

	… have	… am/are/is
I	ɨ̆v-	ŋ̍-/ə̆m-
we	wɨ̆v-	wɘ-
you	jɨ̆v-	jɘ-
he	(h)iz-	(h)iz-
she	ʃiz-	ʃiz-
it	s-	s-
they	ðɨv-	ðɘ-

bradrn wrote: ↑Fri Jun 19, 2020 8:23 pm

Richard W wrote: ↑Fri Jun 19, 2020 12:53 pm [That doesn't look so different from English, even though it does retain some direct marking, which it might very easily have lost.

Not really — it just looks similar since it’s pro-drop. That’s an auxiliary being inflected, rather than a pronoun.

My point was that in English, negativity and most mood have to be marked via auxiliaries. Some dialects seem to have replaced the simple present and simple past with constructions using 'do' and 'did'.

bradrn wrote: ↑Fri Jun 19, 2020 8:23 pm My speech is actually approaching this — the perfective aspect and copula are already fusing with the subject pronouns:

	… have	… am/are/is
I	ɨ̆v-	ŋ̍-/ə̆m-
we	wɨ̆v-	wɘ-
you	jɨ̆v-	jɘ-
he	(h)iz-	(h)iz-
she	ʃiz-	ʃiz-
it	s-	s-
they	ðɨv-	ðɘ-

This is true of most Americans, I think. I'd be similar except I'm rhotic, and would have different vowels in some of them.

I'm not convinced that this is fusion, however, rather than sandhi. I suspect you make some pretty complicated adjustments with any verb that begins with a vowel, but I don't think every such verb has a suppletive fusional set of forms.

At least one test of word status is whether other material can come between, and it certainly can in English: "I absolutely have done it", "I myself am...", "I in fact am..." (I'm open to saying that we often instead move such insertions after the subject+auxiliary combo ("I've absolutely done it"), but my examples are far from being ungrammatical.) One reason I like to analyze the French verbal complex as a single spoken word is that you can't insert such material there.

I think we'd have a much better appreciation of sandhi if we had to write it all over, as in Sanskrit or even Masoretic Hebrew. What we really say diverges quite spectacularly from what we write, or even from a phonemic representation of the word spoken slowly in isolation.

zompist wrote: ↑Fri Jun 19, 2020 9:49 pm

bradrn wrote: ↑Fri Jun 19, 2020 8:23 pm My speech is actually approaching this — the perfective aspect and copula are already fusing with the subject pronouns:

	… have	… am/are/is
I	ɨ̆v-	ŋ̍-/ə̆m-
we	wɨ̆v-	wɘ-
you	jɨ̆v-	jɘ-
he	(h)iz-	(h)iz-
she	ʃiz-	ʃiz-
it	s-	s-
they	ðɨv-	ðɘ-

This is true of most Americans, I think. I'd be similar except I'm rhotic, and would have different vowels in some of them.

Well, I’m not American, but I’m not surprised to hear that others have this as well. (I speak either Australian or South African English, although I haven’t yet figured out which.)

I'm not convinced that this is fusion, however, rather than sandhi. I suspect you make some pretty complicated adjustments with any verb that begins with a vowel, but I don't think every such verb has a suppletive fusional set of forms.

I don’t think so… If I try this exercise with another vowel-initial word such as allow, I get:

I allow [ˈɐ͡i.ə̆ˈɫæ͡ʉ]
we allow [ˈwi.ə̆ˈɫæ͡ʉ]
you allow [ˈjʉ.ə̆ˈɫæ͡ʉ]
he allows [ˈ(h)i.ə̆ˈɫæ͡ʉz]
she allows [ˈʃi.ə̆ˈɫæ͡ʉz]
it allows [ˈɪd.ɮ̍ˈæ͡ʉz]
they allow [ˈðej.ə̆ˈɫæ͡ʉ]

There’s certainly some sandhi — due to the /tə̆l/ sequence in it allows — but it’s nowhere near the amount there is with … have and … am/are/is.

At least one test of word status is whether other material can come between, and it certainly can in English: "I absolutely have done it", "I myself am...", "I in fact am..." (I'm open to saying that we often instead move such insertions after the subject+auxiliary combo ("I've absolutely done it"), but my examples are far from being ungrammatical.) One reason I like to analyze the French verbal complex as a single spoken word is that you can't insert such material there.

I think this is one of those cases where you need to distinguish grammatical and phonological words. Grammatically, those are definitely two words, as you say. However, phonologically, those are clearly not even one word, as shown by the reduced vowels — or, in some cases, the lack of vowels entirely. So it’s basically a clitic, but containing two grammatical words rather than one. (Is there even a name for this sort of thing? A ‘hyperclitic’, perhaps?)

For me, the pattern is: huge reduction before stressed initial vowels (I'm always running. > 'malways running). But before unstressed vowels (like in your "allow" example) or before consonants other than h (I'm ten gerbils > *mten gerbils) the pronouns are normal. I think part of the illusion comes from the fact that initial pronouns that are not being emphasized are almost always unstressed in a language that's already notorious for grinding unstressed vowels into jelly. It really seems that, at least for me, there is no grammatically-specific aspect to this.

bradrn wrote: ↑Sat Jun 20, 2020 6:44 am I don’t think so… If I try this exercise with another vowel-initial word such as allow, I get:

I allow [ˈɐ͡i.ə̆ˈɫæ͡ʉ]
we allow [ˈwi.ə̆ˈɫæ͡ʉ]
you allow [ˈjʉ.ə̆ˈɫæ͡ʉ]
he allows [ˈ(h)i.ə̆ˈɫæ͡ʉz]
she allows [ˈʃi.ə̆ˈɫæ͡ʉz]
it allows [ˈɪd.ɮ̍ˈæ͡ʉz]
they allow [ˈðej.ə̆ˈɫæ͡ʉ]

There’s certainly some sandhi — due to the /tə̆l/ sequence in it allows — but it’s nowhere near the amount there is with … have and … am/are/is.

I don't want to suggest you're wrong about your own speech ... but, analyzing one's own speech can be tricky. If you say these word pairs in isolation, I'm sure you get exactly what you wrote above. But then if someone didn't hear you saying "I'm", you might repeat it as "I am".

How does it sound in a regular-tempo longer sentence, e.g. "You wouldn't believe the chaos in her classroom, she allows the kids to do anything"? I think you'd get closer and closer to [ˈʃi.ˈɫæ͡ʉz].

On the other hand, I'm not opposed to a special handling of the auxiliaries, since after all that's actually reflected in the writing system— we write "I'm", "she's", etc. I just think it's a difference in degree rather than in kind from other sandhi phenomena.

Apparently, crack pots are still going strong. This guy, who is a real linguist at the University of Paris, is trying to convince the world that Muskogean is related to Turkish, because... well, I really don't want to spoil it for you.

there is good evidence that Yeniseian is related to the Na-Dene family (Navajo etc), so if he were able to prove that Yeniseian is also related to Turkish, we'd have a genuine connection between Turkish and the languages of the New World. However his theory is different .... he's not connecting it to Na-Dene, but to Muskogean,Hokan, etc which are not visibly related to Na-Dene. While in theory both connections could be true, one would expect the proto-languages to resemble each other. I wouldnt quite call this a crackpot idea, but i agree with you that the provided wordlist ... especially given that only *some* of the languages on each side are represented for each word .... is very weak evidence for such a bold claim.

zompist wrote: ↑Sat Jun 20, 2020 2:43 pm

bradrn wrote: ↑Sat Jun 20, 2020 6:44 am I don’t think so… If I try this exercise with another vowel-initial word such as allow, I get:

I allow [ˈɐ͡i.ə̆ˈɫæ͡ʉ]
we allow [ˈwi.ə̆ˈɫæ͡ʉ]
you allow [ˈjʉ.ə̆ˈɫæ͡ʉ]
he allows [ˈ(h)i.ə̆ˈɫæ͡ʉz]
she allows [ˈʃi.ə̆ˈɫæ͡ʉz]
it allows [ˈɪd.ɮ̍ˈæ͡ʉz]
they allow [ˈðej.ə̆ˈɫæ͡ʉ]

There’s certainly some sandhi — due to the /tə̆l/ sequence in it allows — but it’s nowhere near the amount there is with … have and … am/are/is.

I don't want to suggest you're wrong about your own speech ... but, analyzing one's own speech can be tricky. If you say these word pairs in isolation, I'm sure you get exactly what you wrote above. But then if someone didn't hear you saying "I'm", you might repeat it as "I am".

How does it sound in a regular-tempo longer sentence, e.g. "You wouldn't believe the chaos in her classroom, she allows the kids to do anything"? I think you'd get closer and closer to [ˈʃi.ˈɫæ͡ʉz].

On the other hand, I'm not opposed to a special handling of the auxiliaries, since after all that's actually reflected in the writing system— we write "I'm", "she's", etc. I just think it's a difference in degree rather than in kind from other sandhi phenomena.

Oh, yes, I am indeed aware about the difficulty of analysing one’s own speech. But I’m pretty sure here that they are indeed as I said — when I say your example sentence, it’s still [ˈʃi.ə̆ˈɫæ͡ʉz]. (Although I’m not ruling out that that [ə] could be deleted — that’s why I marked it extra-short. But I’m not sure that deletion of reduced vowels counts as full-on sandhi.)

bradrn wrote: ↑Sat Jun 20, 2020 8:20 pm(Although I’m not ruling out that that [ə] could be deleted — that’s why I marked it extra-short. But I’m not sure that deletion of reduced vowels counts as full-on sandhi.)

That is exactly what sandhi does in Sanskrit. (Among other things.) And Hebrew, for that matter.

(I'm not trying to be especially argumentative. I think sandhi is pretty neat, and I think it gets ignored or forgotten a lot in languages that don't mark it.)

zompist wrote: ↑Sat Jun 20, 2020 9:37 pm

bradrn wrote: ↑Sat Jun 20, 2020 8:20 pm(Although I’m not ruling out that that [ə] could be deleted — that’s why I marked it extra-short. But I’m not sure that deletion of reduced vowels counts as full-on sandhi.)

That is exactly what sandhi does in Sanskrit. (Among other things.) And Hebrew, for that matter.

(I'm not trying to be especially argumentative. I think sandhi is pretty neat, and I think it gets ignored or forgotten a lot in languages that don't mark it.)

Ah, right. I think it’s clear now that I have absolutely no idea what sandhi is. (I thought I knew, but clearly I didn’t.) Do you know of any resources which I can use to learn about it?

Two sets of examples of what probably is sandhi in the dialect here are:

I don't [ˈaːõʔ], or optionally [ˈaːõn] before a vowel, semivowel, or nasal
you don't [jõːʔ], or optionally [jõːn] before a vowel, semivowel, or nasal
he doesn't [ˈ(h)iːʌːzn̩ʔ], or optionally [ˈ(h)iːʌːzn̩] before a vowel, semivowel, or nasal
she doesn't [ˈʃiːʌːzn̩ʔ], or optionally [ˈʃiʌːzn̩] before a vowel, semivowel, or nasal
we don't [ˈwiːõʔ], or optionally [ˈwiːõn] before a vowel, semivowel, or nasal
they don't [ˈðeːõʔ], or optionally [ˈðeːõn] before a vowel, semivowel, or nasal

I didn't [ˈaːɘːnː],]; [ˈaːɘːːʔ], or optionally [ˈaːɘːːn] before a vowel, semivowel, or nasal
you didn't [ˈjʉːɘːnː]; [ˈjʉːɘːːʔ], or optionally [ˈjʉːɘːːn] before a vowel, semivowel, or nasal
he didn't [ˈhiːɘːnː]; [ˈhiːɘːːʔ], or optionally [ˈhiːɘːːn] before a vowel, semivowel, or nasal
she didn't [ˈʃiːɘːnː]; [ˈʃiːɘːːʔ], or optionally [ˈʃiːɘːːn] before a vowel, semivowel, or nasal
we didn't [ˈwiːɘːnː]; [ˈwiːɘːːʔ], or optionally [ˈwiːɘːːn] before a vowel, semivowel, or nasal
they didn't [ˈðeːɘːnː]; [ˈðeːɘːːʔ], or optionally [ˈðeːɘːːn] before a vowel, semivowel, or nasal

The reason for treating these as sandhi is that they're simple flap elision followed by vowel sequence reduction. The only reason to not treat these as sandhi is that the initial flapping that permits it is not regular, and only occurs (to my knowledge) in these places along with the words to, today, and tomorrow.

Travis B. wrote: ↑Sat Jun 20, 2020 11:21 pm Two sets of examples of what probably is sandhi in the dialect here are:

I don't [ˈaːõʔ], [ˈaːõn] before a vowel, semivowel, or nasal
you don't [jõːʔ], [jõːn] before a vowel, semivowel, or nasal
he doesn't [ˈ(h)iːʌːzn̩ʔ], [ˈ(h)iːʌːzn̩] before a vowel, semivowel, or nasal
she doesn't [ˈʃiːʌːzn̩ʔ], [ˈʃiʌːzn̩] before a vowel, semivowel, or nasal
we don't [ˈwiːõʔ], [ˈwiːõn] before a vowel, semivowel, or nasal
they don't [ˈðeːõʔ], [ˈðeːõn] before a vowel, semivowel, or nasal

I didn't [ˈaːɘːnː],]; [ˈaːɘːːʔ], [ˈaːɘːːn] before a vowel, semivowel, or nasal
you didn't [ˈjʉːɘːnː]; [ˈjʉːɘːːʔ], [ˈjʉːɘːːn] before a vowel, semivowel, or nasal
he didn't [ˈhiːɘːnː]; [ˈhiːɘːːʔ], [ˈhiːɘːːn] before a vowel, semivowel, or nasal
she didn't [ˈʃiːɘːnː]; [ˈʃiːɘːːʔ], [ˈʃiːɘːːn] before a vowel, semivowel, or nasal
we didn't [ˈwiːɘːnː]; [ˈwiːɘːːʔ], [ˈwiːɘːːn] before a vowel, semivowel, or nasal
they didn't [ˈðeːɘːnː]; [ˈðeːɘːːʔ], [ˈðeːɘːːn] before a vowel, semivowel, or nasal

The reason for treating these as sandhi is that they're simple flap elision followed by vowel sequence reduction. The only reason to not treat these as sandhi is that the initial flapping that permits it is not regular, and only occurs (to my knowledge) in these places along with the words to, today, and tomorrow.

That’s some pretty extreme elision and reduction there! I think it’s fair to say that if I were to hear someone saying something like [ˈaːõʔ], I wouldn’t guess that they were saying I don't.

bradrn wrote: ↑Sat Jun 20, 2020 11:31 pm That’s some pretty extreme elision and reduction there! I think it’s fair to say that if I were to hear someone saying something like [ˈaːõʔ], I wouldn’t guess that they were saying I don't.

The standard stressed pronunciation in the dialect here of I don't is [aːe̯ dõʔ] or, (optionally) before a vowel, semivowel, or nasal [aːe̯ dõn]. However, do when unstressed has the tendency to flap the initial /d/, resulting in [aːe̯ ɾõʔ] or [aːe̯ ɾõn]. Then, there is a very common phonological process in the dialect here that I, for lack of a better term, call flap elision, which is exactly that; here it results in [aːe̯õʔ] or [aːe̯õn]. However, /aɪ/ has a tendency to monophthongize before a back vowel, which here results in [aː], giving us [aːõʔ] or [aːõn].

Travis B. wrote: ↑Sat Jun 20, 2020 11:40 pm

bradrn wrote: ↑Sat Jun 20, 2020 11:31 pm That’s some pretty extreme elision and reduction there! I think it’s fair to say that if I were to hear someone saying something like [ˈaːõʔ], I wouldn’t guess that they were saying I don't.

The standard stressed pronunciation in the dialect here of I don't [aːe̯ dõʔ] or, (optionally) before a vowel, semivowel, or nasal [aːe̯ dõn]. However, do when unstressed has the tendency to flap the initial /d/, resulting in [aːe̯ ɾõʔ] or [aːe̯ ɾõn]. Then, there is a very common phonological process in the dialect here that I, for lack of a better term, call flap elision, which is exactly that; here it results in [aːe̯õʔ] or [aːe̯õn]. However, /aɪ/ has a tendency to monophthongize before a back vowel, which here results in [aː], giving us [aːõʔ] or [aːõn].

No, I do realise how you got [ˈaːõʔ] from I don't — I was just commenting that the result ends up seeming reduced in a rather extreme way.