Esperanto Language Stack Exchange is a question and answer site for teachers and students of the Esperanto language. It only takes a minute to sign up.
Sign up to join this community
I am principally thinking of compound words (though other counter/examples are welcome).
To illustrate: is ŝatokupo normally pronounced as ŝa-to-ku-po ("easier" to pronounce, in terms of syllable weight/etc.?), or ŝat-o-ku-po (therefore separating the roots ŝat- and okup-)?
Dankegon.
As Fliomu already stated, there is no rule that morpheme boundaries determine syllable boundaries.
When it comes to postfixes, they are pronounced by syllable or as you said the "easier" way: be-le-ga instead of any other way.
Compound words without a risk of misunderstanding are pronounced without break at the word boundary: vaporŝipo.
There are common compound words which can get hard to understand if pronounced "by syllable". For instance
Most opt for a pause (which can be a very short) between the parts and only very few retain the last letter of the first word: ŝatookupo resp. sanoasekura.
When it comes to uncommon compound or prefixed words, like konkludo in the sense of ludo per konkoj, I say that most speakers will opt for another expression, konkoludo, konka ludo, ludo de/per konkoj… rather than making the non-common meaning clear by repeated use of pauses.
There is no rule that morpheme boundaries determine syllable boundaries. They are independent of each other. Rules of phonotactics and syllabification, however, do apply, like in all natural languages. Those rules alone determine what can constitute a syllable. Consequently, if two words differ only in the locations of morpheme boundaries, these words are pronounced in exactly the same way.
PMEG claims that it is possible and customary to demonstrate the morpheme boundaries in ambiguous words by dividing the words in syllables at the morpheme boundaries. But it does not tell how to do this:
Kiam oni volas elparoli kunmetitan vorton tre klare, aparte montrante la limojn inter la vortopartoj, oni nature uzas silabodividon laŭ tiuj limoj. La malofta kunmetaĵo konkludo (= ludo per konkoj) povas do soni kiel kon-klu-do en rapida elparolo, sed kiel konk-lu-do, kiam oni volas klare montri, kiel tiu vorto estas kunmetita. Kontraste la ofta vorto konkludo (= rezulto de rezonado) kredeble ĉiam sonas kiel kon-klu-do.
I suppose what is meant is that speakers can mark morpheme boundaries by inserting a break. In my opinion, this is not regular speaking of Esperanto. I would pronounce konklud'o and konk'lud'o absolutely identically. (I would not say konk'lud'o, in the first place, but konkoludo.) To insert a break at morpheme boundaries is an auxiliary device to make something explicit that normally isn't, in case it is requested, for example.
