*** this post was first published in “Issues with Chinese” forum, maybe it was misplaced there ***
for sure these issues have been addressed somewhere. Maybe you can answer here or direct me to the appropriate discussion.
a Hanzi has several pronunciations depending on the usage. An example is 着 which can be zhe5, zhao1, zhao2 or zhuo2. In the LingQ Pinyin transcription, it only has “zhao1”, although the audio is saying it correctly. I guess, there is no chance to have it correct in the transcript as well, is this correct?
Same goes for 女, which is nü3 but is transcribed with ru3. Yes, that is another reading of this Hanzi, but mostly it is nü3, isn’t it?
This has certainly been discussed multiple times before, here are some discussions I remember:
The important thing to know about the Pinyin on LingQ is that it is auto-generated by a software running on the server. This means, you cannot edit or correct it and it is unrelated to a specific lesson.
As you have noticed, the software has a tendency to select the most obscure reading of a word. But in the case of 着 it is impossible to guess the reading without context. I would recommend you rely on native audio to learn the correct pronunciation whenever possible, all lessons in the library should come with it. Text-to-speech software can be wrong as well.
The only way to assure that you always have the correct Pinyin is to copy it from a dictionary when you create a definition. If your dictionary language is English you should see that many, if not most, definitions already contain the Pinyin.
I have continued thinking about the Pinyin issue.
I know about Hanzi having several different Pinyin transcriptions/pronunciations. I also understand that it is almost impossible (just to be on the safe side ) to make a 100% correct automatic Pinyin transcription under these circumstances.
Consequently, I have the two following questions:
why not make the Pinyin transcription accessible for edition after being created? Really impossible?
why not remove all automatic Pinyin and make it an additional input field for vocabulary, if we need to input it in many cases anyways.
Have these points been discussed with the developers? If yes, what have been the resolutions?
I believe before the automatic system was introduced, users had to save the Pinyin in the meaning field. Personally, I think this is still a good way because it puts the user in control and causes fewer surprises. But others might be more casual learners and enjoy the convenience of having a mostly correct automatic solution.
I don’t know why LingQ doesn’t allow editing the transcription. Maybe the system would get too complex? Also if they decide to improve the algorithm, all lessons benefit immediately.
Also, I’m just a user and not in a position to contact the developers. You can ping @Zoran, he can look into it and raise this internally.