Jump to content

Wikifunctions:Type proposals/IPA phone base glyph

From Wikifunctions

Summary

A type for working with the values of IPA transcription (P898) statements.

Uses

The IPA is a universal* transcription system for human speech. It's also replaced most other systems for pronunciation guides.
There are 100-odd characters which each represent a particular place and manner of articulation (for consonants; for vowels it's a height, a backness, and a roundedness), and the meaning is narrowed further with diacritics or superscript modifier characters.
The specificity is needed because a lot of human languages have distinctions between pairs of similar sounds. But where similar sounds are considered allophonic in the context of a language, one, often basic, character is chosen to represent that phoneme.
(This does mean that characters in the IPA technically don't correspond to the sounds associated with themthe signifier is not the signified. For our purposes I'm proposing an enum of phones as below, not some enum of characters. We can have parsing/conversion functions, such as IPA symbol to Wikidata item (Z28474).)

My current proposal is that we pick (but how?) for the enum a set of base phones that are identified in and distinguished by the IPA.
If there's a use case that requires those small distinctions within a language, or perfect round-tripping of IPA, the extra information can be associated with them using some outer data structure, analogous to the IPA's diacritics and superscripts.
Meanwhile the enum on its own can be used for romanisation (or the reverse e.g. gairaigo).

Elements of the enumeration

These are encoded as subclasses of phone (Q202064). Exactly which to include is an open question.
has phoneme (P2587) links Items for languages to the Items for their phonemic inventory.

Alternatives

Just use Z6s (not Z86s since some phones like ⟨d͡ʒ⟩ are transcribed with digraphs).

Comments

For general comments, please reply to the proposer.