[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.