ipa-x-sampa: Fix some mistakes #217
Conversation
* Wrong input codes for ɝ, ʛ, ʄ, and ̥ * Wrong characters for _X and _l Fixes uim#216
| ((("1" )) ("ɨ")) | ||
| ((("2" )) ("ø")) | ||
| ((("3" )) ("ɜ")) | ||
| ((("3" "`" )) ("ɝ")) |
There was a problem hiding this comment.
I confirmed that your changes are correct except this based on https://en.wikipedia.org/wiki/X-SAMPA .
It seems that 3` doesn't exist. (E` doesn't exist too.)
Could you share a definition of 3` ?
There was a problem hiding this comment.
Wikipedia doesn't list 3` as a separate entry, but it has an example of [n3`s] for GenAm pronunciation of nurse which would be spelled nɝs in Unicode IPA. In the original paper introducing the notation (linked on Wikipedia), Wells gave the example of GenAm bird in Kenyon and Knott transcription (ie. bɝd) as /b3`d/.
Wells defined ` as a diacritic and all entries that have ` in them can be analysed as a code for some base symbol followed by `. In case of vowels, this directly corresponds to the rhotic diacritic of IPA. See https://en.wikipedia.org/wiki/R-colored_vowel#Notation. Compare how ɚ @` doesn't even appear on the current IPA chart as a standalone symbol, but only as an example of the usage of the rhoticity diacritic.
So even if it's not listed explicitly, E` unambiguously stands for ɛ˞ while ɝ is 3` (as seen in examples). The former can be entered as ɛ E followed by ◌˞ `, but 3` is virtually always written with a single character ɝ rather than as a sequence ɜ˞ (ie. ɜ 3 followed by ◌˞ `). Although Unicode doesn't define any equivalence, from the IPA point of view ɝ is functionally just like a precomposed character for ɜ˞.
Fixes #216