Can't get over how WRONG the pinyin is on the Chinese Mini Stories

Not only are there misspellings in the pinyin, but the pinyin itself is incorrect!
学校 xué xiào
很棒 hěn bàng
女儿 nǚ ér
是 shì
在 zài


Thanks, and sorry about that. We are aware that pinyin isn’t perfect and we are looking into improving it.

please do. maybe for now just give us the option of editing the pinyin in lingqs. the wrong pinyin is really making me question whether I want to continue using this app in spite of its obvious strengths.

I see the issue with the 3-character word highlighted above, but I don’t see the other incorrect pinyin at the bottom of the post. What device were you using when you took that screenshot?


I’ve joined the party late. Definitely agree that LingQ has strengths but perhaps rather than spending too much time developing the Gucci side of its software, it should perhaps be better to concentrate on the basics - i.e. a congruence between translations. An analogy (with not such severe consequences) is taking a map out into the mountains whose symbols are inaccurate and therefore does not correspond with the terrain - but where the map is housed in a bright sparkling cover that looks great to other hikers and which protects, what has turned out to be an essentially useless map, from the wind and rain. We need to have confidence that our maps are accurate in order to place trust in them and help to provide comfort when taking leaps of faith into the unknown.

1 Like

Please do! As you know I have also already posted several times, and sent several e-mails to support[at]lingq, to report mistakes in pinyin.
I can understand LingQ is not perfect, but PLEASE implement the corrections as soon as someone reports a mistake!

The voice translation is also wrong (listen to hai) LingQ Mini Stories 32 . LingQ is our map - we should be able to trust it. Not only are customers supplying feedback of basic things that are wrong (like a car that does not have a steering wheel) it seems that the feedback does not register. I’ve sent numerous emails about other issues but they were not answered so I’ve had to address this in an open forum (where I would have preferred to acknowledge shortcomings in private and give time for people to come up with solutions)-still if this is the only communication channel.

I’ve tried to explain this before, but this and similar request bark at the wrong tree in my opinion. The Pinyin is just computer generated. It will fail in all sorts of circumstances. Further, it will never satisfy all tastes and preferences. Pinyin is also to some extent subject to debate[1], see for example tone sandhi: Some people claim these should be included, while native sources, like monolingual dictionaries, do not.
My recommendation is therefore:
If you want to use Pinyin you need to add the correct Pinyin to the word definition itself. This way you will always have the expected result and you’re in control.

[1] Tone change rules - Chinese Pronunciation Wiki

OK- thank you for your explanation. From what you say I understand that the translation process is very personal and to a large degree idiothetic (and that words/symbols do not attach to an ontological reality-perhaps I now understand that is why it is better to learn language from Chracters rather than Pinyin??). I get that. What I find hard in this argument, and certainly as a novice of Mandarin, that if LingQ’s premise of language learning is about comprehensible input (that largely disregards rules etc) then what you are implying is that there must be some knowledge to begin with in order to apply (what you seem to be as alluding to as contextualisation). Do you therefore not think your suggestion is a big ask of a novice (especially when reading basic entry level text)?

PS- also thank you for the link. I’d be grateful to know if language learning is more effective through character recognition rather than pinyin.

I don’t think it is a big ask, most of the user definitions already contain Pinyin. Also, most online dictionaries should provide Pinyin as well, e.g. Word dictionary - 耽误 - MDBG Chinese Dictionary

I myself don’t use Pinyin much, whenever possible I try to use audio and text in conjunction. The idea being, that when looking at a character or word I hear the native speaker pronounce it, obviating any need for transcription. I also didn’t mean to imply any need for explicit rules, I just tried to provide some background information. I believe @Erik8970 asked for tone sandhi to be included in the transcription in another thread.

I am aware that some people try to learn the language by using Pinyin alone and avoid learning characters, I have no opinion and no experience with this approach, so I cannot comment. I learned the characters from the start.

Regarding LingQ, my personal view is that it is a really great concept and it works. The company however, is a bit special and the development works in strange ways. My experience is that you can either make feature requests, or you can leave it be - the difference is, the latter option leaves you more time to study languages. Nowadays, I try to accept the product as is and don’t place much hope in them improving the product significantly - if they do it’ll be a happy surprise.

Further to my earlier comment - I was not stating that there is always a direct congruence between the internal and external world (Korzbyski- the map is not the terrain) but I was merely saying that there should be some congruence between basic entry level stuff and their translations. If I understand you correctly what you seem to be saying is that the process of congruence is a little more complicated than a direct correspondence between the external reality and the symbols associated with it - then; would it not be the case that the more symbols we have to represent something the more nuanced our perception of reality (like Chinese characters to fine tune perception). For example-I read that the Inuit tribe has 50 different names for snow because their survival depends on it. Am I just seeing the basic white stuff mediated through the convenience of pinyin - and would my powers of observation be better by understanding Chinese characters (but with the caveat that usage depends on necessity). Any insight would be appreciated.

Hi - sorry, I think I cross posted.

TBH, from what I have gleaned so far (and hearing somewhere) that learning by characters can be more effective (gulp-deep breath). I agree, it’s probably worth spending more time on learning languages rather than the development. Personally, I find streaks, coins etc a distraction-though they probably work for some. Thanks for your perspective - although it feels like launching into the deepend - it’s something that I feel may be necessary sooner rather than later.

1 Like

“Further, it will never satisfy all tastes and preferences.” What does that mean?
There is only one pinyin for a character/meaning combination, and that’s what can be found in the dictionary. A character can have a different pinyin (pronunciation or tone) but then the character will have a different meaning.
Pinyin is necessary if you want to learn Chinese.
I’m not talking about the few cases where tones change when combined with another character.
I wish LingQ would implement the corrections for the errors that subscribers report. It should not be too difficult. Saying ‘we will look into it’ is not enough.

Searching the correct pinyin for every word, and add it in the word definition? that must be a joke? LingQ should make learning languages easy and fun.

Just for your information: I have learned Chinese for 8 years (almost 20 years ago) and I lived and worked in China for 6 years. I know from experience how important pinyin is when learning Chinese.

1 Like

TO LingQ - Eric 8970 has put forward an entirely different perspective to bamboozled. So with the expertise in your organization (and within your networks) is there not someway that you are able to bring a (and careful not to bring Hegel into this) a synthesizing between these two perspectives (at least where synthesizing doesn’t distort that outcome). As a novice what I cannot do and what I will not do is spend/waste time looking under the bonnet instead of driving the car (because as Erik8970 says) that doesn’t make language learning easy and fun. After paying for your product I am left questioning the fair exchange of value (because I’m still finding glaring errors, as a novice, that should really not be there and which does not inspire confidence in my map - regardless of the ontological/epistemological debate). The least you (LingQ) should do is to address the concerns that have been raised here months ago and provide some perspective on them. Perhaps it’s a product related problem - putting too many out in the market place rather than concentrating on a few niche products (languages). Perhaps less efforts on freebies, glitz and shine and more into product development (content over style).

Hello Erik8970!

we will look into it
When Zoran says they will look into this, it means he will contact the developers and suggest they improve the transcription algorithm, nothing more nothing less. Your misunderstanding might be that you assume that the Pinyin is saved alongside each lesson, this is not the case. It gets generated automatically. Feel free to verify this by creating a lesson for yourself. So asking LingQ to correct single instances of incorrect Pinyin isn’t expedient in my opinion.

add it in the word definition
This is no joke.
Adding the Pinyin to the definitions has the following advantages:

  1. The user is in control and is thus not subject to the vicissitudes of an algorithm
  2. Ergonomic positioning together with the translation
  3. Able to capture phrases and non-dictionary words

It is easy and intuitive to add Pinyin to the definitions for two reasons:

  1. The vast majority of existing user hints already contain Pinyin
  2. The vast majority of dictionaries already contain Pinyin, so when a user adds a hint they can just copy the Pinyin together with the translation

There is only one pinyin for a character/meaning combination
Please consider the following three points, how do you suggest solving these cases using a dictionary based algorithm?

  1. Three common characters: 地 的 着

  2. The issue of tone sandhi, dictionaries typically do not include those, however a learner may prefer to see those for didactic reasons.

  3. What if a user prefers, for example Taiwanese Mandarin pronunciation?

Please also take this additional factor into account when assessing the situation: Languages such as Chinese and Japanese are written without natural delimiters like spaces, so word segmentation is essential for a platform like LingQ. This process is however by no means perfect, resulting in many words that will not appear in any dictionary (maybe 10-15%). How do intend to solve this problem?

Just for your information
Thank you for clearing this up. As for myself, I’m just a humble student of this endlessly fascinating language and I do hope you will continue to share your valuable expertise here on the forum so that I can learn from it. It is great to have you.

1 Like