What's the exact realization of Chinese glides (in medial position)?

Question

While in some transcriptions you might encounter the IPA [j], [ɥ], and [w] for the medial glides, there are also transcriptions with IPA [ʲ], [ʷ] and [ᶣ].

Glides can have many interpretations whichever notation you use. For example, in Russian, palatalized consonants can have an audible glide or not depending on the environment where the palatalized consonant is and the palatalized consonant itself. Now, that glide can also be very audible, like a true [j] or not that much, like a true [ʲ].

I'd like to know which of these or other interpretations is the correct for Mandarin Chinese. I've been listening to a little Chinese but from apps where they don't necessarily speak at conversational speed, so at least when they speak clearly or slower the glides are very audible to me, like [j], but sometimes it could be closer to [ʲ] I guess, I'm not sure.

However, I've found this statement from Duanmu: "A consonant-glide combination at the start of a syllable is articulated as a single sound – the glide is not in fact pronounced after the consonant, but is realized as palatalization [ʲ], labialization [ʷ], or both [ᶣ], of the consonant". So that can also have many interpretations since a palatalized consonant can have a following audible glide and I haven't found the book for more details.

What pronunciation is more accurate for Standard Mandarin Chinese, and/or more popular?

Answer 1

The distinction between [pʲu] пю and [pju] пью for a Russian speaker may be very salient, but for the majority of Mandarin learners of Russian, it is a very difficult one to pick up.

That is because the equivalent phonetic difference is not phonemic in Mandarin, and it does not matter much for comprehensible Mandarin. It is thus unsurprising that the glide is more audible in slow Mandarin than faster-paced Mandarin.

Nonetheless, acoustic analysis from several sources as well as psycholinguistic tests e.g. involving rhyming and game languages seem to all agree that in Modern Standard Mandarin, the glide is part of the onset.

Note that diachronically, the glides of Middle Chinese were firmly in the final, not the onset, and this is still the case for many other varieties in Chinese, e.g. in Taiwanese Hokkien.