Print

Print


In addition to what Christophe and Kou said:


I think there's one more option for limiting the number of syllables per
block.

Languages with dynamic accent often don't permit long sequences of
unaccented syllables; to avoid them, they put secondary stresses on
certain syllables.

If you formulate explicit rules for secondary stress position, you can
unite in one block either (1) a syllable with (secondary) stress plus
all unaccented syllables that precede it, or (2) a syllable with
(secondary) stress plus all unaccented syllables that follow it (these
two options seem simplest; you can prefer something more sophisticated
and/or matching your phonetics better).

I guess the words having high textual frequencies will mostly be
relatively short, thus fitting in one block, while relatively rare
longer words could be written with two or more blocks.

But with accent types other than dynamic this won't look natural (IMO).


Basilius