User:Elector Dark/Sandbox/6

Phonology
All vowels can be nasalised except for /æ/ and /ɒ/; nasalised vowels are romanised with an ogonek (/ã:/ = ą́). In situations where nasalisation occurs, a nasalised /æ/ shifts to /ɒ/ while /ɒ/ remains unchanged and unnasalised. Even though the romanisation doesn't show it well, the vowel /o/ can be nasalised (/õ/ vs. /ɒ/ will anyhow be romanised <ǫ>). The romanisation doesn't reflect this well, even though their placement is usually predictable; on the other hand, the runic script shows nasalisation on /o/ well.

Umlaut
The language features two forms of productive umlaut: i-umlaut which fronts non-front vowels and u-umlaut which backs or lowers non-back vowels. Where i-umlaut is triggered by the presence of /i/ in the following syllable, u-umlaut is triggered by the presence of /u/. The table below demonstrates the effects of both umlaut types:

When /a/ is followed by the triggering vowels but is outside the first syllable, it does not undergo classic umlaut and instead acquires the quality of said triggering vowels. The vowels /u i/, on the other hand, do not change their quality outside the first syllable.

The umlaut is applied mostly regularly: it is more-or-less consistent when there is only one intervening consonant, but is sometimes blocked by a cluster or long consonant. Long vowels can also avoid umlaut.