[cbr]

Let's say we pick a random person and make them learn a completely unrelated language. Sometimes they will learn faster than others, like Spanish or Swahili vs English or Mandarin. Why would we expect them all to be the same difficulty?

[umanwizard]

An easy-to-learn writing system is a massive help when learning a language, but the writing system isn't part of the "language" as linguists understand the term. Just a representation of it.

So I totally believe that it's easier to learn English than Mandarin (and maybe easier to learn Spanish than English) because the writing system gives you more hints, but that doesn't have anything to do with the (spoken) language, strictly speaking.

On another note, it's not clear to me that "as phonetic as possible" makes a (written) language easier to read, since our brains process words/morphemes as chunks, rather than sounding them out letter-by-letter (this is how reading Chinese is even possible). So semantic vs. phonetic writing should be thought of as a tradeoff. As an easy example: spaces between words are not phonetically justifiable (there is no pause between words in natural speech), but they sure help reading comprehension a lot.

[cbr]

    I totally believe that it's easier to learn English than
    Mandarin (and maybe easier to learn Spanish than English)
    because the writing system gives you more hints, but that
    doesn't have anything to do with the (spoken) language,
    strictly speaking.

Sorry, I wasn't actually trying to bring up the writing system at all! I put both English and Mandarin in the "harder" category, even though English's writing system is simpler (though still not as simple as many).

[chaoky]

It's not a question of inherent difficulty, which doesn't make sense to quantify absolutely. It's all about how similar it is to the random person's native language, phonologically, morphologically, syntactically etc... Obviously Portuguese would be faster for a Spanish speaker to learn, since they are very similar in many aspects (in fact, they are both very conservative Iberian Romance languages) than it would be for a Spanish speaker to learn Polish. On the other hand, Polish would be easier to learn than Spanish for a Czech speaker. None of these are objectively more "difficult" or "complex" than one another. They are just "different".

[cbr]

Why doesn't it make sense to talk about the inherent difficulty of a language? I agree that some languages are related and this eases learning (PT to ES, PL to CZ) but why can't we look at learning speed on completely unrelated languages? For example, get some monolingual Quechua speakers and some monolingual Tamil speakers, and have them each learn the other language. Which group learns faster? The group that learns faster is learning the easier language.

I realize that this isn't practical to actually do as an experiment with all the world's language pairs, but I do think it demonstrates that inherent difficulty is a thing.

One thing that sort of shows this, though it's kind of distorted by writing systems, would be charts like this from language learning companies: http://www.effectivelanguagelearning.com/language-guide/lang... For example, they think for an English speaker Swahili is easier than Nepali is easier than Finnish.