There have long been Middle Chinese and Old Chinese reconstructions on paper, but since the Chinese script is not phonetic (although syllabic to a degree), it has been difficult to ascertain pronunciations. If one takes Classical Latin as an example — that is a reconstruction of fairly normal and believable speech of about 2000 years ago if read aloud, yet there is nothing approaching that for Middle Chinese (about >1000 years ago) much less for Old Chinese (>2000 years ago). Recently though, a couple of funny videos cropped up on Youtube showing people making overly academic attempts at reading classical texts using reconstructed archaic pronunications.

Here’s a guy attempting to read Tang Dynasty poetry in Middle Chinese.

Here’s another guy attempting to read some of the oldest written poetry from the Zhou era in Old Chinese.

Generally I find these unbelievable and unnatural, probaby due to a combination of reconstruction error and bad or exaggerated intonation. Nobody can speak like this. For Middle Chinese it has been argued that the rhyme book scheme does not give one speech pattern and in any case the reconstruction has severe survivor bias toward modern southern dialects. For Old Chinese, I’m just speechless… I don’t think this guy even knows how to read Zheng Zhang Shang Fang’s phonetic notation, so he produces hilariously impractical initial and ending clusters with bizarre tonal variations to boot. Anyway, it’s a start.

---

Here’s an interesting article on the origins of court dialects.