Somali orthography

[inariksit]

Great to have Somali coverage in GF!

I'm curious to know where the data is from. It looks rather linguistic, with lots of IPA characters.

lin question_1_V2 = mkV2 (mkV "suʔal") noPrep ; --guessed
lin put_6_V2 = mkV2 (mkV "ɖɪg") noPrep ; --guessed
lin squeeze_1_V2 = mkV2 (mkV "tuːǯi") noPrep ; --guessed

The sources I've read that try to adhere to Somali conventions (like this one), would use ' for the glottal stop, and dh for the retroflex. The word squeeze would be written tuuji. There are many other examples as well, and I'm not quite sure about all of them. So I'd like to know the source, so I can make a more educated guess.

I would like the GF resource to reflect more the prevalent usage than grammar books. If you agree, I can take care of the transliteration.

Smart paradigms for verbs

This is a TODO item for myself.

When I wrote the fragments of the Somali RG 2 years ago, I left my verb morphology quite unfinished. I was trying to decide which form would be the dictionary form, and found different conventions in different lexica and grammars. I didn't come to a conclusion, so my current implementation takes an artificial stem, which is not any actual wordform. I can adjust my smart paradigms to work with different forms, but that's good to keep in mind if we get words from different sources: their dictionary form may not be the same.

[krangelov]

Afrikans, Maltese, Romanian, Somali and Swahili are new additions which are just proof of concept to show that new languages can be added even without any existing WordNet. Basically I search for translations in PanLex from all existing languages to the target language. After that I use different features to decide which one is the best translation. For the morphology I imported external sources when I can find one or I used the smart paradigms if the word cannot be found otherwise. None of the above languages have an existing morphological lexicon but for instance for German and French there were. PanLex contains thousands of bilingual lexicons in hundreds of languages and they all come from different sources. In particular for Somali, I listed the used sources below. Feel free to make any adjustments necessary. I can also easily regenerate the dictionary if you find that something in general is wrong. CLDR Version 6.1, By-Type Chart: names.language Unicode Common Locale Data Repository Languages in their original form Geonames: Language codes PanLex meta-source List of ISO 639-3 codes Dbnary: Wiktionary as Linguistic Linked Open Data Wiktionary Wikiwörterbuch Wîkîferheng Wiktionary tiếng Việt Wikisanakirja Wiktionnaire Lexicon Enhancement via the GOLD Ontology Wordlists The ASJP Database Reading in Somali: An Elementary Cultural Reader Geonames: Glossaries Wikiorðabók HeiNER - the Heidelberg Named Entity Resource Continents Unicode Common Locale Data Repository — Territories Wikizionario The days of the week in various languages വിക്കിനിഘണ്ടു Vikisõnastik International Organizations Викисловарь Linked Languages Resources The months of the Gregorian (Christian) calendar in various languages Oceans Astronomia Terminaro Numbers from 1 to 10 in over 5000 languages Some Fields of Science Seas and Straits Mountains Solar System Peace in various languages The months of other calendars in various languages Circles of latitude Some Historical People Times in History Some Works of Literature and a Ballet Football World Cup The books of the New Testament

…

On Thu, 20 May 2021 at 11:01, Inari Listenmaa ***@***.***> wrote: Great to have Somali coverage in GF! I'm curious to know where the data is from. It looks rather linguistic, with lots of IPA characters. lin question_1_V2 = mkV2 (mkV "suʔal") noPrep ; --guessed lin put_6_V2 = mkV2 (mkV "ɖɪg") noPrep ; --guessed lin squeeze_1_V2 = mkV2 (mkV "tuːǯi") noPrep ; --guessed The sources I've read that try to adhere to Somali conventions (like this one <http://morgannilsson.se/BeginnersSomaliGrammar.pdf>), would use ' for the glottal stop, and dh for the retroflex. The word *squeeze* would be written tuuji. There are many other examples as well, and I'm not quite sure about all of them. So I'd like to know the source, so I can make a more educated guess. I would like the GF resource to reflect more the prevalent usage than grammar books. If you agree, I can take care of the transliteration. Smart paradigms for verbs This is a TODO item for myself. When I wrote the fragments of the Somali RG 2 years ago, I left my verb morphology quite unfinished. I was trying to decide which form would be the dictionary form, and found different conventions in different lexica and grammars. I didn't come to a conclusion, so my current implementation takes an artificial stem, which is not any actual wordform. I can adjust my smart paradigms to work with different forms, but that's good to keep in mind if we get words from different sources: their dictionary form may not be the same. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#28>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AEYFSZEFUQRASOEU3SEW2WDTOTFVZANCNFSM45GQKRCA> .

[inariksit]

Okay, I understand! So it's likely that already the words in there currently have different base forms. Unfortunately I don't know Somali in any meaningful way, just random anecdotes about morphosyntax. :-P But I noticed that the word "ulayn" appears as a translation for 890 different words, which seem to be pretty arbitrary (here's a party game: what's in common with accept, chop, paste, seduce, trash… and 800+ more words?)

I think I'll contact someone who actually knows Somali, and try to organise a meeting to go through the lexicon. I'll get back to you after I have some more information.

[krangelov]

My knowledge of Somali is even weaker, so if you could check with someone that would be great. I wonder if "ulayn" might be a light verb like "do" in English. Maybe accept is translated as "do acceptance". Or maybe it is just a nonsense translation that should be simply removed.

…

On Thu, 20 May 2021 at 12:21, Inari Listenmaa ***@***.***> wrote: Okay, I understand! So it's likely that already the words in there currently have different base forms. Unfortunately I don't know Somali in any meaningful way, just random anecdotes about morphosyntax. :-P But I noticed that the word "ulayn" appears as a translation for 890 different words, which seem to be pretty arbitrary (here's a party game: what's in common with *accept*, *chop*, *paste*, *seduce*, *trash*… and 800+ more words?) I think I'll contact someone who actually knows Somali, and try to organise a meeting to go through the lexicon. I'll get back to you after I have some more information. — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#28 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AEYFSZHWHELYKSCM4EJL7DDTOTPDTANCNFSM45GQKRCA> .