Lightnews — Scholar-powered news

Sāmapriyavasuḣ @avzaagzonunaada.bsky.social · 1d

Hansel & Gretel in Öömrang

1 11

Sāmapriyavasuḣ @avzaagzonunaada.bsky.social · 1d

The Grimm tales in Öömrang & Fering Insular North Frisian.

1 3 21

Sāmapriyavasuḣ @avzaagzonunaada.bsky.social · 4d

BlueSky, we got gray sky in North Frisia. Here’s the island church of Oomram — „sark“ if you will — under it.

27

Sāmapriyavasuḣ @avzaagzonunaada.bsky.social · 5d

Anglo-Frisian gin

1 1 9

Sāmapriyavasuḣ @avzaagzonunaada.bsky.social · 7d

Stay in Poznań over. Glad to meet several amazing colleagues (including Ronalld Kim & Henrik Liljegren!!) and hearing about lots of interesting research being done.

(Pic by Sakura Ishikawa of the title of my talk on Shina(ic) & Gawri historical accentology.)

7

Sāmapriyavasuḣ @avzaagzonunaada.bsky.social · 10d

The Narts immigrated from the Caucasus to Poland. Their lair.

10

Sāmapriyavasuḣ @avzaagzonunaada.bsky.social · 13d

... though problematic cases like ávi- ‘sheep’ and áhi- ‘snake’ do famously exist.

2 2

Sāmapriyavasuḣ @avzaagzonunaada.bsky.social · 13d

It doesn’t look like there is much consensus on the history of that word. Mayrhofer gives several options. I can sort of sea how ‘sea’ may change to ‘wilderness, desert’, but still feels not strongly supported enough. Skt. marú- is difficult to derive from *mor-u- because of Bruggmann’s Law, ...

1 3

Reposted by Sāmapriyavasuḣ

Gasper Begus @begus.bsky.social · 15d

is also more difficult for the models.

For methodology on how to analyze internal layers in tone learning, see the paper with the amazing Kai Schenck:

arxiv.org/abs/2509.17859

Unsupervised Learning and Representation of Mandarin Tonal Categories by a Generative CNN

This paper outlines the methodology for modeling tonal learning in fully unsupervised models of human language acquisition. Tonal patterns are among the computationally most complex learning objective...

arxiv.org

2 9

Sāmapriyavasuḣ @avzaagzonunaada.bsky.social · 16d

მადლობა

1

Sāmapriyavasuḣ @avzaagzonunaada.bsky.social · 16d

Hah, I had no idea!! Is the author a Japanese mathematician?

1

Sāmapriyavasuḣ @avzaagzonunaada.bsky.social · 16d

I kind of desperately want to take this survey, but I am not a native-speaker of English and don’t want to mess up their research. :(

Kirby Conrod @kirbyconrod.bsky.social · 16d

Please boost! #linguistics student here at Swarthmore is running a survey for English speakers (who don't know Mandarin / any other tone languages) about how we learn tones. Take a few minutes to help out with this student research!
swarthmore.az1.qualtrics.com/jfe/form/SV_...

Online Survey Software | Qualtrics Survey Solutions

Qualtrics sophisticated online survey software solutions make creating online surveys easy. Learn more about Research Suite and get a free account today.

swarthmore.az1.qualtrics.com

1 5

Sāmapriyavasuḣ @avzaagzonunaada.bsky.social · 16d

Is this open to statisticians in North America?

1

Reposted by Sāmapriyavasuḣ

Robin Ryder @robinryder.bsky.social · 16d

I'm super excited to announce that ISBA @isba-bayesian.bsky.social has voted to start a new section on Bayesian Social Sciences! It will be a great way to further collaborations with many disciplines in the Social Sciences and Humanities.

bss-isba.github.io

Home - BSS-ISBA

bss-isba.github.io

2 8 24

Sāmapriyavasuḣ @avzaagzonunaada.bsky.social · 16d

More seriously, it is interesting that English speakers prefer to not use their BUT vowel when pronouncing foreign words. To my ears, the “short a” of Indian languages is much closer to that than to any other English vowel.

2 5

Sāmapriyavasuḣ @avzaagzonunaada.bsky.social · 16d

The human ideal is to call the language Canarese like the English prioneers decided and the country yändátaʼ /çäˈⁿdäːtäʔ/ as in the original Huron-Wendat.

3

Sāmapriyavasuḣ @avzaagzonunaada.bsky.social · 16d

I say Canarese. :)

1 3

Sāmapriyavasuḣ @avzaagzonunaada.bsky.social · 16d

Thanks, I’ll read your thread in detail later in the day.

Reposted by Sāmapriyavasuḣ

Saulė / Saule @sankuperis.bsky.social · 16d

Så, today is the Baltic Unity Day, when the Balts (tribes which Lithuania and Latvia consist) united their forces and defeated Swordbrethren.

On that occasion, may I share my years of work dedicated to the Baltic languages with you? Please find the links below in this thread:

2 1 18

Sāmapriyavasuḣ @avzaagzonunaada.bsky.social · 16d

I would love to look at your slides (after the talk, of course) if you don’t mind sharing. (Unfortunately, can’t be at the talk that far away.)

1

Reposted by Sāmapriyavasuḣ

dialectmatt.bsky.social @dialectmatt.bsky.social · 16d

This Friday, I will be giving a talk at the Comparative Indo-European Linguistics (CIEL) Seminars, this time on how dialectometric methods can help us study historical linguistics, with case studies from Indo-European languages.
www.universiteitleiden.nl/en/events/20...

1 3 9

Reposted by Sāmapriyavasuḣ

Johan Schalin @schaljoh.bsky.social · 17d

For anyone’s information who is looking for my publications: I just deleted my Academia(dot)edu account. No explanations needed.
You still find my publications on Researchgate and I plan to start updating on Knowledge commons. My last one I uploaded yesterday:

www.researchgate.net/publication/...

(PDF) The non-front main-stressed vowels in Early Old Nordic – a reconstruction

PDF | The comparative data testifies to an Early Old Nordic three-way contrast between short labialised vowels of different origin. From these... | Find, read and cite all the research you need on Res...

www.researchgate.net

2 9

Sāmapriyavasuḣ @avzaagzonunaada.bsky.social · 17d

The consequence of intervocalic /-d-/ v. /-ḍ-/ is massive for later Indo-Aryan.

Medial /-d-/ disappears during the Middle Indic (early mediæval) period, while /-ḍ-/ lives on (usually tapped [-ɽ-], or [-ɭ̆-] in Insular, though secondarily merging with /-r-/ (or /-l-/) in scattered varieties).

Sāmapriyavasuḣ @avzaagzonunaada.bsky.social · 17d

Latin nīdus (< *nisdós) ‘nest’ and sīdit (< *sísdeti) ‘sits’ have obvious cognates in Sanskrit नी॒डः nīḍáḣ and सीद॑ति sī́dati.

However, only does nīḍáḣ show the expected retroflexion (Pre-Sanskrit *niẓḍáḣ), while in sī́dati we do not get **sī́ḍati (assuming Pre-Sanskrit *síẓḍati).

It could be that the ...

Yoïn van Spijk @yvanspijk.bsky.social · 17d

Historically, the word 'nest' consists of two parts.

The part 'ne-' is related to 'nether', while '-st' is related to 'sit'.

The distant ancestor of 'nest' meant "place to sit down".

'Nest' is also related to Spanish 'nido' and its Romance relatives.

Click my new graphic to learn more:

6

Sāmapriyavasuḣ @avzaagzonunaada.bsky.social · 17d

I love the background on this. :)

(By the way, I am not familiar with Hindi nīṛ. That is indeed the expected reflex of Skt. nīḍáḣ, but I have never seen it, nor does Turner report it. I imagine someone directly using the Skt. word in literature may use it.)

1 3

Sāmapriyavasuḣ @avzaagzonunaada.bsky.social · 17d

... verb ‘sit’ — more phonologically and semantically related to the base root than the lexicalized noun ‘nest’ — was restored to having /d/ (in place of regular /ḍ/) in transparent analogy with the still-productive root √s(a)d.

2