I thought to myself, why not check the Arabic on LingQ just for fun (I am a native Arabic speaker) and to my surprise, I was counting correct sentences instead of counting errors in the audio (MSA only, dialects seemed fine), please bare in mind that MSA is not the language we use every day, so not everyone speaks it fluently and not any Arabic speaker will be able to speak it accurately, and please be careful because some official channels have mistakes as well (not as many as in LingQ’s stories: short stories, who is she, etc…) those mistakes mainly consist of omitting case endlings, pronouncing ة when you are not supposed to, adding the case ending at the end of words (at the ends of words, it’s supposed to be omitted) and lastly using the wrong case ending.
materials that are produced by LingQ should be recorded by professionals and materials imported to LingQ as well should be reviewed by professionals, and learners importing media should try to be careful
an example would be (if you speak German)
“Ich spreche mit ein Mann” where it should be “Ich spreche mit einem Mann” or “Sie geht zu die Arbeit” where it should be “Sie geht zu der Arbeit”
Hi Ahmad, thanks for your post! we have been actually working on improving the mini stories and other basic courses in different libraries recently. It would be great to hear more details about MSA stories from you, could you please send an email to me at sahra @ lingq .com ?
thanks!
In my experience Lingq often seems to have wrong vowels in pronunciation, when tapping on the word. When you listen the whole text though, the pronunciation is correct. Probably different speech synthesiser, which analyses the whole sentence, is used. I recommend to switch off the sound, except when you listen the whole text.
I’ve used to learn Arabic before, just forgot it a little, so I can handle these errors, but for beginners they are harmful.
Basically, automated tools like Lingq are just not so good in handling complex languages, with their quirks. Languages like Arabic, in which you need to “guess” the correct form and missing vowels. And you can forget about learning “dialects”, which have no official written forms altogether.
As for proofreading, it would be great, but probably cost a lot of money, especially if they need to check all user generated content. So it doesn’t fit into Lingq business model.
There is an app which does what you propose for Chinese though, using a lot of proofreaded content. And it costs like Lingq which has 20 languages.
Unfortunately there’s probably just not enough demand for Arabic to create a dedicated service like this.