Loanwords are words adopted from one language into another. Some loanwords have diacritic letters like the é in résumé.
In this article:
Why is the AI mispronouncing words in my Pronunciation Library?
What are diacritics?
Diacritics are “accent marks” placed above or below a letter in a word to indicate a particular pronunciation.
- The current list of diacritic marks that are sometimes used in English words:
- â Â à À á Á ê Ê é É è È ë Ë î Î ï Ï ô Ô ù Ù û Û ç Ç ä ö ü Ä Ö Ü ñ Ñ
- Words that use these diacritics are typically loanwords, like crème brûlée, cliché, voilà, attaché, and doppelgänger.
If your script doesn’t heavily use diacritic words, another option would be to simply add the word to your pronunciation library. You can also try creative spellings to create the correct pronunciation. The following examples vary by voice (one respelling doesn’t always work for all voices).
- Rezzm-"A"
- Rez-emmay
- Rez-uh-may
- Rezzm-eigh
Why isn’t ñ working?
The letter ñ is still not recognized as its corresponding phoneme is outside standard English phonetic sounds.
Why is the AI mispronouncing words in my pronunciation library?
If you have previously added Respellings to your Pronunciation Library that include words with diacritic marks, the AI may have a hard time pronouncing these replacements.
To avoid this, you may want to toggle the Pronunciation off or remove these entries.