the hardest natural language

[gusl]

to learn is perhaps Ubykh, with 80 consonants, and extremely fast communication.

The verb basically reiterates the entire syntactic structure of the sentence, agreeing with subject, object, indirect object, benefactive and oblique objects (English, by contrast, has agreement only with the subject of the verb, and that only in the third person singular).

However, I do not doubt the difficulty of learning tonal languages. Hoa showed me her beginner's Vietnamese tape, and I was completely hopeless.

Now, going away from phonetic and towards deeper, more conscious aspects of language:
On Dec 5, I saw a DIP lecture on the evolution and complexity of languages, Linguistic complexity, by Wouter Kusters. He argued convincingly that it is highly unlikely that all languages have equal complexity, for this would require a mechanism for adding complexity somewhere else whenever something got simplified in evolution and vice-versa.
I learned that languages can have more than one function: communication is always the primary function, but showing speaker identity is also a function in some languages. Isolated languages (type I, I think) tend to have "illogical", useless features that only make it hard to learn. e.g. although Dutch is not a good example of an isolated language, adjectives are marked by both gender AND article definiteness, e.g. "de grote kamer", "een grote kamer", "het grote huis", "een groot huis".

Comments

[jcreed.livejournal.com]

I think the idea of language-for-speaker-identity is really important. I keep getting in arguments with people about whether TV is really likely to level dialect differences in, say, American English. The fact that people hear a "standard midwestern english" in every CNN broadcast is only going to make it more clear to them how their dialect differs from that particular presented standard, and if they want to identify with their region, I imagine they will even (un)consciously actively diverge from that standard.