"I love them very much!"
Translation:Eu le iubesc foarte mult!
one of the translation options was "Eu le iubesc alt mult pe ele!" Unless I'm wrong, that's a completely meaningless sentence. Something like, "I love them other much." From a learning perspective, I think it's counterproductive to expose learners to nonsense sentences, because when we see sentences in a foreign tongue, they become paradigms for correct speech in than language. If the incorrect sentences were good, grammatical Romanian, that would be a real improvement. "Eu le urasc foarte mult" would be an incorrect match, but still a functioning sentence. Doesn't that make more sense? Is it that the incorrect sentences are computer generated?
I've seen many nonsensical ones in Romanian too (this kind of random words making the whole sentence syntactically invalid), but I'm not sure if other languages have them too.
Well, I'm not sure how duolingo works.
If they are randomly generated, then it would be difficult to ensure they are syntactically valid. But it would be even more difficult to ensure you've not randomly generated something that would be a correct answer, reason for which I doubt they are randomly generated. It could pick randomly between more than the options though.
Personally I do not like them in Spanish either because I think I cheat at them. Instead of actually looking at each individually to see if it fits, I look for reasons to prove them wrong (if the English version talks about a dog, but I see no perro in the Spanish one, then I have one less translation to worry about). When I get to have to focus more it's mostly because there are two correct options.
So either way I don't like it pedagogically, but I agree they should at least all be valid constructs.