I think it's fairly well established that voiced stops tend to depress f0 while unvoiced stops tend to raise it, but I'm less clear on other types of consonant. I'm especially interested in sonorant consonants and semivowels, and whether they have any systematic effect on the f0 of the following vowel. More specifically, is there any reason why initial [j] should depress the f0 of the following vowel?

  • In historical Chinese phonology, sonorant consonants (次濁, voiced initial nasals like /m/, /n/, /ŋ/ as well as lateral /l/) are well-established as patterning with voiced initial stops and fricatives ("obstruents", 全濁). This is also true of /j/ (the 以 or 喻 initial).
    – Michaelyus
    Dec 10, 2021 at 12:48

1 Answer 1


You tag this as a phonology question but refer to effect on F0, which is confusing. There are substantive factors in the phonetics which contribute to raising and lowering of F0, but there are also categorial phonological rules involving that change from one tone category to another. I assume that you are talking about continuous F0 effects in phonetics and not phonological tone effects. Sonorant effects in phonology are known, and have a range of historical explanations.

Especially if this only involves [j], that would be very surprising and unexpected, so well worth firmly establishing. There are also vowel height effects where [i,u] have higher intrinsic F0 effects, and if the corpus has no cases of [w] then this could be the reason. The effect has long been known for high vowels but I think it has not been systematically studied for glides since the effect would be harder to measure (you can sompare [i,u] to [e,o] but what do you compare [j,w] to?).

  • 1
    Sorry – poor choice of tag / use of slashes. What I’m ultimately trying to explain is a pitch contour in a recording that falls when you would expect a rise (H tone between 2 M tones). So far, native speakers produce a rise when asked to pronounce the same sentence, but at the same time they perceive the clip as correct/natural. I don’t think it’s down to accent as other tokens are normal. It’s as though the expected pitch contour has been superimposed on a rapid fall, but I can’t explain where the fall is coming from. Not prosody, and AFAICS not downward momentum from the previous word...
    – rchivers
    Mar 8, 2021 at 16:45
  • That only seems to leave the initial consonant. Initial [j] doesn’t normally seem to have this effect so there would have to be some special pleading along the lines of there is a downward impetus when it is released - this can be reined in, and it usually is, but on an unstressed word in a casual context it can noticeably change the pitch contour. From what you say though this is not a plausible explanation either, so it’s back to the drawing board. Couldn’t you compare [j,w] to [r,l]?
    – rchivers
    Mar 8, 2021 at 16:46
  • Romanian has diphthongs /e̯a/ and /o̯a/ as well as exhibiting semivowels /j/ and /w/. Couldn't that be suitable for such a study?
    – LjL
    Mar 8, 2021 at 16:47
  • What language is this about?
    – user6726
    Mar 8, 2021 at 16:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.