Zompist Bboard Again

Since I already started a Dravidian language varieties thread, I thought I might as well start an Indo-European language varieties thread! Any Dravidian language is fair game for the Dravidian languages thread. Similarly, discussion related to any Indo-European language is fair game for this one despite the size of the family!

I guess I'll just start off with a quick remark about Odia, since that's an Indo-European language I don't see much discussion of here and I'm interested in promoting such (i.e. rarely mentioned) languages more. Odia, for those who don't know, is an Eastern Indo-Aryan language closely related to Bengali and spoken in the eastern Indian state of Odisha to the southwest of West Bengal. Apparently, at least in standard Odia, words never end in a consonant, always in a vowel, and the default vowel is always [ɔ], so e.g. Bhuvaneswar (the capital of Odisha) is [bʱubɔneswɔɾɔ].

Talking about Indo-Aryan, are retreoflex stops ubiquitous in that family?

Raholeun wrote: ↑Wed May 12, 2021 1:18 am Talking about Indo-Aryan, are retreoflex stops ubiquitous in that family?

Wasn't "the Aryan language" once used to mean PIE? I have a very old Welsh grammar which keeps referring to "Aryan suffix this, X inherited from Aryan that" etc. PIE is the only logical ancestor I can think it must refer to.

Jonlang wrote: ↑Wed May 12, 2021 6:08 am

Raholeun wrote: ↑Wed May 12, 2021 1:18 am Talking about Indo-Aryan, are retreoflex stops ubiquitous in that family?

Wasn't "the Aryan language" once used to mean PIE? I have a very old Welsh grammar which keeps referring to "Aryan suffix this, X inherited from Aryan that" etc. PIE is the only logical ancestor I can think it must refer to.

"Aryan" was indeed once used for IE as a whole (hence also the "Aryan race" nonsense). This usage was abandoned for two reasons: 1. The self-designation Arya is attested only in Indo-Iranian. 2. The racist abuse of the term.

WeepingElf wrote: ↑Wed May 12, 2021 6:15 am

Jonlang wrote: ↑Wed May 12, 2021 6:08 am

Raholeun wrote: ↑Wed May 12, 2021 1:18 am Talking about Indo-Aryan, are retreoflex stops ubiquitous in that family?

Wasn't "the Aryan language" once used to mean PIE? I have a very old Welsh grammar which keeps referring to "Aryan suffix this, X inherited from Aryan that" etc. PIE is the only logical ancestor I can think it must refer to.

"Aryan" was indeed once used for IE as a whole (hence also the "Aryan race" nonsense). This usage was abandoned for two reasons: 1. The self-designation Arya is attested only in Indo-Iranian. 2. The racist abuse of the term.

The racist thing is kinda what I assumed to be the reason for it being abandoned, the book in question (if I recall correctly) is from the late 1800s or early 1900s.

Since abugidas are sometimes called "alphasyllabaries," could the Odia script be called an "omicronsyllabary?"

Raholeun wrote: ↑Wed May 12, 2021 1:18 am Talking about Indo-Aryan, are retreoflex stops ubiquitous in that family?

No. Assamese just has alveolar stops instead of the dental-retroflex distinction, perhaps due to Tai substrate influence.

EDIT: While retroflex stops are common to varying degrees all over India, deretroflexion is, too.

Moose-tache wrote: ↑Wed May 12, 2021 8:53 am Since abugidas are sometimes called "alphasyllabaries," could the Odia script be called an "omicronsyllabary?"

One of my professors once argued that Indic scripts in general are just alphabets with an inherent vowel.

Vijay wrote: ↑Wed May 12, 2021 10:09 am One of my professors once argued that Indic scripts in general are just alphabets with an inherent vowel.

Allowing vowels to appear before the consonants they phonetically follow causes no end of trouble.

One way of seeing it is that abugidas are abjads with obligatory vowel diacritics, more or less. At least, that is how I understand them to ahve emerged from Semitic abjads both in India and in Ethiopia.

the more i learn about different writing systems, the more i feel that the distinctions between alphabet, abjad, abugida, and syllabary are blurrier than they typically get presented, and the less useful i find these categories to be

Richard W wrote: ↑Wed May 12, 2021 4:16 pm

Vijay wrote: ↑Wed May 12, 2021 10:09 am One of my professors once argued that Indic scripts in general are just alphabets with an inherent vowel.

Allowing vowels to appear before the consonants they phonetically follow causes no end of trouble.

I have no idea what that's supposed to mean, but:

Linguoboy wrote: ↑Thu May 13, 2021 11:08 amI came across "Croesus" in Vonnegut yesterday and that reminded me how I used to transpose this to "Croseus", which I pronounced /'kroːsiːəs/.

Vijay wrote: ↑Sat May 15, 2021 12:07 pm

Richard W wrote: ↑Wed May 12, 2021 4:16 pm

Vijay wrote: ↑Wed May 12, 2021 10:09 am One of my professors once argued that Indic scripts in general are just alphabets with an inherent vowel.

Allowing vowels to appear before the consonants they phonetically follow causes no end of trouble.

I have no idea what that's supposed to mean, but:

A South Indic script will typically have about half a dozen vowels which are written, in part at least, to the left of the consonant they are associated with. Think of the difference between traditional and 'typewriter' Malayalam when dealing with a Sanskritic consonant cluster followed by the vowel /e/.

Richard W wrote: ↑Sat May 15, 2021 1:33 pm

Vijay wrote: ↑Sat May 15, 2021 12:07 pm

Richard W wrote: ↑Wed May 12, 2021 4:16 pm
Allowing vowels to appear before the consonants they phonetically follow causes no end of trouble.

I have no idea what that's supposed to mean, but:

A South Indic script will typically have about half a dozen vowels which are written, in part at least, to the left of the consonant they are associated with.

Half dozen? There are literally two.
EDIT: Oh, you said in part. Four, then.

You clearly aren't counting SIGN AI e.g. കൈ let alone the old SIGN AU e.g. കൌ. Thai has five purely standing on the left, and lots of multi-part vowels.

Okay, so what about all this "causes no end of trouble"?

Vijay wrote: ↑Wed May 12, 2021 10:09 am
EDIT: While retroflex stops are common to varying degrees all over India, deretroflexion is, too.

Oh, a question I've been meaning to ask.
I remember reading (though I can't remember where!) that Indo-Aryan retroflexes and Dravidian retroflexes are really not quite the same. (IA leaning towards apical-alveolar while Dravidian retroflexes are articulated at the hard palate.)
Is that correct?

Ares Land wrote: ↑Sat May 15, 2021 3:24 pm

Vijay wrote: ↑Wed May 12, 2021 10:09 am
EDIT: While retroflex stops are common to varying degrees all over India, deretroflexion is, too.

Oh, a question I've been meaning to act.
I remember reading (though I can't remember where!) that Indo-Aryan retroflexes and Dravidian retroflexes are really not quite the same. (IA leaning towards apical-alveolar while Dravidian retroflexes are articulated at the hard palate.)
Is that correct?

Sort of. Ladefoged says something like this in at least one of the earlier editions (fifth edition?) of A Course in Phonetics, but I've also had a native speaker of Hindi tell me that most native speakers produce them just like we Dravidian language-speakers do and the ones who use the tongue tip against the alveolar ridge are a minority.

Ares Land wrote: ↑Sat May 15, 2021 3:24 pm I remember reading (though I can't remember where!) that Indo-Aryan retroflexes and Dravidian retroflexes are really not quite the same. (IA leaning towards apical-alveolar while Dravidian retroflexes are articulated at the hard palate.)

The system was different. While Old Indic has dental v. retroflex, Dravidian had dental v. alveolar v. retroflex. I've read claims that contact with Dravidian lead to a muddle whereby some deretroflection occurred in the Indic language. Dental v. retroflex contrast was largely lost for non-plosives in at least some Middle Indo-Aryan, with retroflex v. dental being the reflex of /n/ v. /nn/.

Vijay wrote: ↑Sat May 15, 2021 1:58 pm Okay, so what about all this "causes no end of trouble"?

• One solution for truly complex conjuncts is to write two conjuncts. The question then arises of where the preposed vowel goes. This issue allegedly has different solutions (even in Devanagari!) for Marathi and Hindi. Anaptyxis can aggravate the issue even without conjunct formation. In Thai, the spelling of /tɕapʰɔʔ/ oscillated during the 20th century between เฉพาะ (implying a phonological consonant cluster) and ฉเพาะ (implying two syllables).
• When it comes to computers, the problems just multiply:
• If one stores the vowels on the left first, as with Thai, then there is a problem with doing automated sorting. For years the solution was to swap vowel and consonant for sorting. Then, with more complicated tables coming into vogue, this problem was subsumed in the general solution for complex look-ups used for sorting Welsh and the pre-computing sort order of Spanish. The downside is that the sorting rules are not very clever, and Lao needs far huger tables than packages like ICU can handle.
• If the consonants are stored before vowels, then to display text, consonants and vowels have to be swapped round. Unfortunately, on Windows, Windows software claimed the right to do this swapping round, and then was slow to implement this swapping. The consequence is hack fonts and character encodings so that displaying the text does not rely on support from Windows components.
• Sorting out what to do with preposed vowels when conjunct look-ups fail actually required a subtle redefinition of the text layout engines for Indian Indic scripts.

To change the subject completely: Wiktionary lists many different alternatives for various forms of the verb quethen (link). I assume the variation is at least partly scribal and partly dialectal. Does anyone know which set of forms would have been used in the East Midlands dialect? Especially regarding the past participle, which has like six different forms with and without prefixed y-.