phoible sounds that cannot be mapped (among others)

LinguList commented 3 years ago

ɾ̪ ɾ̪ ʐ͇ ʐ͇ ɫ̪ ɫ̪ ɾ̪ ᴅ ð̪̺ ð̪̺ ɹ̪̩ ɹ̪̩ z͇ z͇ ʂ͇ ʂ͇ ts͇ ts͇ z͇ z̪͇|z͇ ɹ̪̹̩ ɹ̪̹̩ ɾ̪ˠ ɾ̪ˠ ɯ͓ ɯ͓ k̙ k̙ ɗʒ ɗʒ ɡǀ͓ ɡǀ͓ i̯au̯ i̯au̯ iou iou ǁ͓ʰ kǁ͓ʰ l̠˞ l̠˞ ʀʁ ʀʁ s͇θ s̻θ d̪ʒ d̪ʒ ˀt̪ɬ ˀt̪ɬ kɬ kɬ ɖr̠ ɖr̠ ɖr̠͓ ɖr̠͓ ð͇ˠ ð͇ˠ dz͇ d͇z͇ ⁿt͇s͇ʰ ⁿt͇s͇ʰ s͇ s͇ ʃ͇ ʃ͇ t͇s͇ t͇s͇ ts͇ʰ ts͇ʰ t͇s͇ʰ t͇s͇ʰ ʈʂ͇ ʈʂ͇ ʒ͇ ʒ͇ t̪ʙ t̪ʙ xʀ̥ xʀ̥ ŋ̥m̥ ŋ̥m̥ ð̙ˤ ð̪̙ˤ ɹ̪̰̩ ɹ̪̰̩ ɹ̪ˠ ɹ̪̩ˠ ɾ̪̊ ɾ̪̊ ɾ̪̊ʰ ɾ̪̊ʰ ɾ̪ʲ ɾ̪ʲ i͓ i͓ ʟ͓̥ ʟ͓̥ ɭ͓ ɭ͓ r͓ r̪͓|r͓ ɾ͓ ɾ͓ u͓ u͓ d̪l̪ d̪l̪ d̙ˤ d̙ˤ t̪̙ˤ t̪̙ˤ ʕ̙ ʕ̙ ɾ̪ ᴅ̪ ɾ̪̰ ᴅ̪̰ fʃ fʃ ŋǂx ɡ̰ǂx ŋǃx ɡ̰ǃx ld ld ʟ̝̊ɬ ɬʟ͓̥ ⁿbz mbz ⁿd͇z͇ ⁿd͇z͇ r̠̙ r̠̙ st st s̙ˤ s̙ˤ ʃt ʃt tθ̪ t̪θ̪ ˀy ˀy

LinguList commented 3 years ago

These sounds were wrongly specified, @cormacanderson, like, st is not possible as a cluster, as we do not allow these combinations, etc. If you want these to be included (I am against it, as I don't consider "st" as a valid sound), I'd need explicit feature values. The little "x", for example, is a feature that you did not introduce, and the like.

cormacanderson commented 3 years ago

In four places I would replace these as it's an error on my part: ð̪̺ ð̪̺ mistake for ð̺ voiced apical dental fricative consonant d̪ʒ d̪ʒ replace by dʒ tθ̪ t̪θ̪ mistake for tθ voiceless dental affricate consonant ˀy ˀy mistake for ˀj pre-glottalized voiced palatal approximant consonant

I suppose we leave blank the following: ɗʒ ɗʒ kɬ kɬ ɖr̠ ɖr̠ i̯au̯ i̯au̯ iou iou l̠˞ l̠˞ d̪l̪ d̪l̪ fʃ fʃ ŋǂx ɡ̰ǂx ŋǃx ɡ̰ǃx ld ld st st ʃt ʃt

cormacanderson commented 3 years ago

The following should be resolved after implementation of https://github.com/cldf-clts/clts/issues/61: s͇ > θ̠ z͇ > ð̠ ʃ͇ > ɹ̠̊˔ ʒ͇ > ɹ̠˔ ʂ͇ > ɻ̝̊ ʐ͇ > ɻ̝

t͇s͇ > tθ̠ d͇z͇ > dð̠ t̠ʃ͇ > tɹ̠̊˔ d̠ʒ͇ > dɹ̠˔ ʈʂ͇ > ʈɻ̝̊ ɖʐ͇ > ɖɻ̝

My understanding then is that we get these for free: ⁿt͇s͇ʰ > ⁿtθ̠ ⁿd͇z͇ > ⁿdð̠

A PHOIBLE-specific mapping: z̪͇|z͇ > ð̠

Implemented as aliases are ð͇ > ð̠ θ͇ > θ̠ (not currently in PHOIBLE)

LinguList commented 3 years ago

so we add them as alias sounds?

cormacanderson commented 3 years ago

No, I think it better to do this as mapping from PHOIBLE, because this is a PHOIBLE-specific convention, not standard IPA usage. This is more of a reminder to myself. I can change the mappings in the .tsv file, but maybe best we resolve the issues first and then I can make the changes to the .tsv and then we run it again to see what still isn't accepted.

LinguList commented 3 years ago

But that means that you do not get anything "for free". That is important! Anyway, there arent' more than the six sounds here.

cormacanderson commented 3 years ago

But surely if we have the base sound tθ̠, then the system will also parse ⁿtθ̠, right? Or should I specify that too?

LinguList commented 3 years ago

Sure, but if it is a correction for a transcription system the alias form won't be detected.

cormacanderson commented 3 years ago

There's no alias here though: tθ̠ is a new base form that we defined in the discussion in https://github.com/cldf-clts/clts/issues/61 (the alias is rather tθ͇). ⁿtθ̠ is just that base form with prenasalisation, so it should be recognised, right? By "free" I mean that I don't specifically have to define ⁿtθ̠ as a new sound, because we have already defined its base.

cormacanderson commented 3 years ago

Or am I missing something? I realise I still have to map these to the sounds we have now defined, which is why I put up the reminder to myself...

LinguList commented 3 years ago

No, you are right, I thought you MEANT the aliases.

cldf-clts / clts

phoible sounds that cannot be mapped (among others) #59