cldf-datasets / phoible

Creative Commons Attribution 4.0 International
2 stars 2 forks source link

Savara (PH 998) is a nonexistent language #15

Closed Taivo55 closed 4 years ago

Taivo55 commented 4 years ago

Savara (PH 998) has been removed from both Glottolog and ISO 639-3 as a non-existent language

bambooforest commented 4 years ago

@drammock -- here's a new one. non-existent language:

https://iso639-3.sil.org/sites/iso639-3/files/change_requests/2015/2015-010.pdf

i suppose we should remove it from the raw data, but what should we do about the InventoryID?

drammock commented 4 years ago

I think the right conclusion here is (as mentioned at the bottom of page 2 of the SIL change request) that Savara is an alternate name for Sora (an austroasiatic lx, not dravidian). It is listed as an alternate name for Sora here, here, and as an alternate name for the Sora peoples here, in addition to the P.C. cited in the SIL change request. Looking at the inventories for sora1254 and sava1244 they are similar enough to plausibly be different analyses of the same lect (or two closely related lects):

phoible %>% 
    filter(InventoryID == 998 | LanguageName == "Sora") %>%
    select(InventoryID, Phoneme) %>%
    group_split(InventoryID) %>%
    lapply(function(tib) pull(tib, Phoneme))

result is this:

[[1]]
 [1] "b" "b̻" "d̪" "d̻" "j" "k" "k̻" "l" "m" "n̪" "p" "p̻" "r" "s" "t̪" "ŋ" "ɡ" "ɲ" "ɾ" "e" "i" "o" "u" "æ" "ɑ" "ə" "ɛ" "ɪ" "ʊ" "ʌ"

[[2]]
 [1] "b" "c" "d" "j" "k" "l" "m" "n" "p" "r" "s" "t" "ŋ" "ɟ" "ɡ" "ɲ" "ɽ" "ʔ" "a" "e" "i" "o" "u" "ə" "ɨ" "ʊ"

So I'd be inclined to keep the inventory and its InventoryID, and change the isocode and glottocode with which we associate it.

bambooforest commented 4 years ago

@drammock -- thanks!

Taivo55 commented 4 years ago

Good call

drammock commented 4 years ago

@bambooforest can you send me the anonymous1927.pdf from which this inventory was extracted, so I can double-check that my proposal is not contradicted by anything the author says?