Zompist Bboard Again

https://indo-european.info/pokorny-etym ... ictionary/

Not sure if this has been posted, but it's an interesting PIE dictionary, some of the secondary links are quite useful as well.

Sure, if you regard "willful ignorance of laryngeals" as interesting. I can't think of any reason to use Pokorny over the LIV.

dhok wrote: ↑Tue Jan 21, 2020 6:47 am Sure, if you regard "willful ignorance of laryngeals" as interesting. I can't think of any reason to use Pokorny over the LIV.

For those roots and for the material LIV leaves out? LIV features only root for which the IE languages show primary verbs, i.e. no purely nominal roots. And LIV lists nouns or adjectives derived from the roots it features only rarely. The sad truth is that no full dictionary of PIE exists that has all currently reconstructed roots with laryngeals plus the attested continuants in the PIE languages, or that includes up-to-date material from Anatolian and Tocharian. Therefore, Pokorny with its breadth of material is still useful.

William Annis has reposted his Conlanger's Thesaurus, with some updates, the biggest one of which is that the PDF is now searchable.

Oooh. Excellent news.

The best overview of shared Balto-Slavic and Germanic subtrate vocabulary I've seen so far. It breaks down over a hundred roots by category, root shape, etc.

Most of you may already know about this, but Harold Orton's mid-century survey of English dialects is available online, and searchable by keyword. I am going county-by-county from Cornwall to Durham.

https://linguisticmaps.tumblr.com/tagge ... istic-maps - several maps showing the occurrence of various linguistic features.

alice wrote: ↑Sun Oct 25, 2020 6:32 am https://linguisticmaps.tumblr.com/tagge ... istic-maps - several maps showing the occurrence of various linguistic features.

Some of the data seems questionable—e.g. Malagasy is listed as a language that doesn't mark tense, but it definitely does (on its verbs as well as its demonstratives!). They also keep referring to Aymara as 'Ayamara'.

Risla wrote: ↑Sun Oct 25, 2020 5:50 pm

alice wrote: ↑Sun Oct 25, 2020 6:32 am https://linguisticmaps.tumblr.com/tagge ... istic-maps - several maps showing the occurrence of various linguistic features.

Some of the data seems questionable—e.g. Malagasy is listed as a language that doesn't mark tense, but it definitely does (on its verbs as well as its demonstratives!). They also keep referring to Aymara as 'Ayamara'.

Besides, most of their maps just look like cleaned-up copies of WALS’s ones.

Brought to you bearing alice's Iinfallible Mark Of Quality™. Never knowingly compromised.

Just found this webpage with lots of stuff on Nuristani languages:

http://nuristan.info/lngFrameL.html

Lots of links are dead, but its a nice resource nonetheless.

Edit: the more general web page index is here: http://nuristan.info/index.html#TOP.

Yes, that is Richard Strand's page. I once managed to (inadvertently) piss him off by writing (almost all of) this as part of my job at the time.

Here's the Grammar Pile: https://mega.nz/#F!x4VG3DRL!lqecF4q2ywojGLE0O8cu4A

Is it me or is the door to the fun house shut?

Raholeun wrote: ↑Wed Nov 16, 2022 3:51 am

Here's the Grammar Pile: https://mega.nz/#F!x4VG3DRL!lqecF4q2ywojGLE0O8cu4A

Is it me or is the door to the fun house shut?

Same for me

I discovered the Glottobank via Martin Haspelmath's Researchgate page and thought it may interest others: https://glottobank.org/

MacAnDàil wrote: ↑Fri Jan 20, 2023 7:51 am I discovered the Glottobank via Martin Haspelmath's Researchgate page and thought it may interest others: https://glottobank.org/

And I have to ask, how is anything accessed? I don't see any online UI, or download links, or Github links or anything for Grambank and Phonobank...

Good point. I can't even get my head around Lexibank: https://github.com/lexibank/lexibank-analysed/. THere is this I managed to get to, but they admit they excluded Continental Celtic which screws their model. https://language.cs.auckland.ac.nz/

MacAnDàil wrote: ↑Mon Jan 23, 2023 12:36 pmhttps://language.cs.auckland.ac.nz/

Ah, those notorious New Zealand computer scientists who think they can make a meaningful contribution to Indo-European historical linguistics, but what they propose is falsified nonsense based on glottochronology, a method proven not to work decades ago because it is based on an obviously false assumption, namely that the rate of lexicon replacement is constant across time and languages - just look at Icelandic vs. English.

IIRC, their idea was to look at the cognate retention of various languages as a percentage, and map this against their physical location to find a point of origin. Quelle surprise, Anatolian had the best cognate retention, with languages retaining fewer cognates as they fan out into the Atlantic margin or South Asia. Ignoring the fact that languages lose cognates at different rates could be a forgivable error, especially if the data set were large enough (it isn't). But the real frustration is that the possibility of an entirely unattested language on the Pontic steppe doesn't seem to have occurred to them, even though that's exactly the question they were trying to answer!

They apparently used a methematical model used to reverse engineer the spread path of a virus. And I guess it makes sense that when looking for the origin of an Ebola outbreak, you're not looking too closely at any place that has no reported cases of Ebola. It's more a game of "which of these Ebola-riddled villages got riddled first?" But even then, if you presented this as serious Bayesian inference at a computer science conference, people would laugh you out of the room for not considering the imperfections in the reported data. It seems like they figured when doing non-STEM, the first thing you have to do is turn your brain off, because if you had a working brain, you'd be good enough for STEM.