Many words, e.g. 東西 (dong1xi1 / dong1xi), 朋友 (peng2you3 / peng2you) have no tone on the last, depending on who you ask, and which dictionary you use (CC-CEDICT lists both words without tone on the second).
Personally I learned those words without tone on the second syllable, so I have to exert mental effort to tell myself to write them with tones.
Is there a chance this can be supported in the future, at least as a secondary accepted input?
Great question, and thanks for the feedback!
This is something that has always tripped me up as well. It is further compounded by the fact that the CC-CEDICT and Pleco dictionaries often do not include the Taiwan pronunciation for words. Even the Ministry of Education dictionary in Taiwan that many Taiwanese themselves use, and can be considered the most authoritative dictionary on this side of the strait, can sometimes have an outdated pronunciation. Dictionaries ultimately can only inaccurately describe the reality on the ground, and are not prescriptive by any means.
I can see us supporting this case, especially for something like 東西 which is indeed pronounced both ways here in Taiwan.
I’ll add an issue for this to our backlog and think about it a bit more. Thinking aloud, here are some things we could do to address this confusion:
- Add a answer mistake prevention case for this. If one types in this well-known alternative pronunciation, we will remind them to use the other one. This is the easiest for us on our end to implement.
- Allow the alternate pronunciation. In such a case, we would probably still have some indicator that indicates the definitive pronunciation that our mnemonic uses. This way we don’t mark you wrong for something that is clearly right, while giving hints on a potentially more accurate pronunciation for Taiwan. This will take more time.
What do you think about the first option above?
I think even the first option would be a massive help if you’re already familiar with the word and have internalized it through other means. It may even make one more familiar with alternate pronunciations instead of silently accepting both (one which could be answered through luck/error).
I’ve been thinking this over the past couple of days and finally settled on a design that I’m happy with. We’ll allow these alternate pronunciations and accept them, but show a pop-up within the review quiz that says:
This pronunciation is less common in Taiwan and does not match our audio .
For the traditional course.
This pronunciation is less common in China and does not match our audio .
For the simplified course.
This way users can be aware that our mnemonic and audio do not match what they typed in so they can move towards the “slightly better” answer. But it doesn’t mark them wrong for an answer that is frankly not wrong. This aligns with our philosophy of being flexible with our answers as long as they are in the right ballpark.
When we add it, we’ll have an announcement about it. We hope to do some other quality-of-life improvements for quiz answers sometime in the next month or two, and will likely get to it then.
Thanks again for your suggestion!