Module:rad-IPA/documentation

This module is used to get the pronunciation estimation of a Radestrian word from a single parameter. The base generation runs backwards, due to Radestrian's 'latter bias' regarding vocalic polygraphs.

{{#invoke:rad-IPA|generate|aies}} will generate ˈa.ies, not *ˈai.ɛs, which would be spelt *aiès.

Irregular spellings are recognised in data:

{{#invoke:rad-IPA|generate|heņrecjes}} will generate ˈhɛŋɡrɛtʃɛs.

It can also handle multiple words, but they must be in the same parameter.

{{#invoke:rad-IPA|generate|ryņles u aieg}}}} will generate ˈriŋɡlɛs u ˈa.ieɡ.

If you would like the hjádvanþs standard, add the hj parameter.

{{#invoke:rad-IPA|generate|hjas}}}} will generate ʃas.

{{#invoke:rad-IPA|generate|hjas|hj}}}} will generate ças.

If you would like to account for allophony (the phonetic transcription), add the phon parameter.

{{#invoke:rad-IPA|generate|hmanþs}}}} will generate hmanθs.

{{#invoke:rad-IPA|generate|hmanþs|phon}}}} will generate xmãˑθs.

The generation runs in 6 ‘stages’:

1. Base generation (one-to-one grapheme to phoneme conversion, adhering to the observed orthographical finality bias, but may still contain pseudo-phonemes)
2. Vowel resolution (vowel alternations and combinations are resolved)
3. Consonant resolution (consonant alternations and combinations are resolved)
4. Syllabification
   1. Syllable tagging (marking syllables for length, checkedness etc.)
5. Stress assignment
6. Syllable resolution