In your opinion, how would the most difficult language be?

[Squall]

How would a language that is very difficult be and you would not like to learn it?


I would say that a terribly-difficult language uses ideograms as the only writing system.

A difficult phoneme inventory is one that includes phonemes that sound too resembling.
/i y ɪ ʏ e ø/
/k kʰ q x kx/
/ɸ f θ ʃ ʂ ɕ ɬ/

/i: y: ɪ: ʏ: e: ø:/
/k: kʰ: q: x: kx:/
/ɸ: f: θ: ʃ: ʂ: ɕ: ɬ:/

And at least 3 tones.

The grammar would have many cases and genders.
Nouns, adjectives and determiners inflect according to case, gender and number (singular, dual, three and plural).
Verbs inflect according to person, gender and number.
At least 3 categories of inflection.
And many irregular inflections.

Rules are also other problems: auxiliary verb and word order in interrogation and negation, rules regarding object pronouns and other rules (Castilian: la agua -> el agua).

Many ambiguous words.

[Theternitend]

I think it depends absolutely on your first language.

Japanese people find English really difficult to learn, since its grammar is totally different and concerning its phonology, they probably think that it comes from Mars.

The more different, the more difficult. These issues always turn out to be relative.

[Micamo]

Well, first, as always with the question of "language difficulty", we need to decide whether we're framing the question in terms of first language acquisition by children (where it's debatable whether there really is such a thing as a "difficult language" within this context), or second language acquisition by adults (where it's very much dependent on what languages they already know).

We also need to decide what role language universals have to play: Do linguistic universals exist at all (and if they do, which ones), and what happens when we try to teach someone a conlang that violates them?

Otherwise, the answer to this question is not interesting: Define an arbitrary bijective function, F, to turn english sentences into unique numbers. Then define an arbitrary bijective function, G, to turn any number into a unique, pronounceable (with "pronounceable" being whatever phonology you feel like defining) string of sounds. For a given english utterance we want to translate into our "impossilang", U, compute G(F(U)*N), where N is an arbitrary number (though make sure it's the same number every time). To translate one of our impossilang sentences U back into english, just compute F(G(U)/N). Basically we've just made a cipher that's waaaay to complex for the human brain to ever possibly do spontaneously, with the exact complexity depending on the definitions of F and G.

[Valosken]

Making a "most difficult language" needs a reference. It seems you've used generally isolating/modern European languages or something as yours. There's no absolute.

[Squall]

I mean the opinion of each one, and not a universal difficult language.
I asked how would be a language (or conlang) that would be nightmare if you were forced to learn it.

[Thrice Xandvii]

So what you are asking is: What characteristics would a language have that would be difficult for YOU to learn?

[Thakowsaizmu]

I think it depends absolutely on your first language.

Japanese people find English really difficult to learn, since its grammar is totally different and concerning its phonology, they probably think that it comes from Mars.

The more different, the more difficult. These issues always turn out to be relative.

Not really. I am sure some Japanese people find it more difficult, but I highly doubt it is a universal. I find Mandarin Chinese much easier to learn than I do French, for example, and French is not only in the same language family as English, but I have been passively and actively exposed to a lot more French than I have Chinese. But I find Chinese easier. This is me, though, and should hardly be taken as stating that English speakers have a better time learning Chinese than they do French.

[Znex]

I asked what would be a language (or conlang) that would be a nightmare if you were forced to learn it.

Fixed the grammar for you. It sounds like you're asking in what way would a language be difficult - hence some people are saying in return that the difficulty of a language depends on the person - as opposed to asking for examples of languages that we think would be difficult to learn.

For me, I don't really have any one language that I think would be super-difficult - any of those of which I am most unfamiliar with are those that I'll need the most practice for and hence are the ones I consider super-difficult. For example, I'd need lots of practice for languages like Russian, Japanese, or any of many of the native American languages, because these all use agglutination, unfamiliar phonetic contrasts, unfamiliar vocabulary, and unfamiliar idioms, among other things.

[Xing]

How would be a language that is very difficult and you would not like to learn it.

I would say that a terribly-difficult language uses ideograms as the only writing system.

A difficult phoneme inventory is one that includes phonemes that sound too resembling.
/i y ɪ ʏ e ø/
/k kʰ q x kx/
/ɸ f θ ʃ ʂ ɕ ɬ/

/i: y: ɪ: ʏ: e: ø:/
/k: kʰ: q: x: kx:/
/ɸ: f: θ: ʃ: ʂ: ɕ: ɬ:/

And at least 3 tones.

The grammar would have many cases and genders.
Nouns, adjectives and determiners inflect according to case, gender and number (singular, dual, three and plural).
Verbs inflect according to person, gender and number.
At least 3 categories of inflection.
And many irregular inflections.

Rules are also other problems: auxiliary verb and word order in interrogation and negation, rules regarding object pronouns and other rules (Castilian: la agua -> el agua).

I think none of these features need to make a language particularly difficult. At least not if you compare with what Micamo suggested...

The stock answer seems to be that difficulty depends on your L1 (and possibly which other languages you have been exposed to). To this we could perhaps add the learner's motivation and willingness to learn a specific language. A slightly more controversial claim is that some features might be "objectively" more difficult to learn. It seems like a plausible plausible hypothesis that languages are more or less equally difficult to acquire for native speakers. (Or do we have any evidence that certain languages take substantially longer time to acquire for native children??) This does not entail that all languages are equally difficult/easy for L2 speakers to learn. Their situation is, after all, quite different from that of natives.

When it comes to written languages, things are a bit different. Some writing systems requires a lot more of rote memorisation than others (for example, a logographic system, or an otherwise highly irregular writing system - though such systems might have other advantages).

[Micamo]

A slightly more controversial claim is that some features might be "objectively" more difficult to learn. It seems like a plausible plausible hypothesis that languages are more or less equally difficult to acquire for native speakers.

There's actually some fascinating research from artificial language learning experiments that humans have a cognitive bias towards acquiring certain language features: There has to be overwhelming evidence that this feature is the correct one before learners will accept it and begin to use it themselves. It's unclear whether this constitutes "difficulty" or not.

[Yačay256]

Not really. I am sure some Japanese people find it more difficult, but I highly doubt it is a universal. I find Mandarin Chinese much easier to learn than I do French, for example, and French is not only in the same language family as English, but I have been passively and actively exposed to a lot more French than I have Chinese. But I find Chinese easier. This is me, though, and should hardly be taken as stating that English speakers have a better time learning Chinese than they do French.

Agreed: My mother speaks French and so once I was curious and I asked her to teach me some: I can't remember a single word from her hour-long lesson, though I can recall several dozen Simplified Chinese Characters on the fly. Also, Chinese languages have no gender !

Anyway, I found this from the Economist on Tuyuca; frankly, I think they don't know what they are talking about (which is fine, being that they are not linguists but, well, economists). Also, I don't know if the so-called noun "classes" are not just affixial classifiers; perhaps somewhat like Jakaltek's?

IMHO, the hardest language I have yet studied in any depth is certainly Sumerian, and by far - though of course our fragmentary knowledge of the language may mean that I, and everyone else for the last two millennia, are gravely mistaken (though I strongly doubt this).

First, Sumerian, in addition to being just insanely polysynthetic, is ridiculously complicated and pretty darn irregular. Heck, I've studied Classical Nahuatl in the past and it was vastly easier than Sumerian, despite a similar level of synthesis, thanks to the former's extreme regularity, lack of Sumerian's insane number of homophones, way more straightforward conjugation system and a simple system of compounding (just put -ti- between the two verbstems and you're done). Meanwhile, Sumerian has: A huge number of verbs not in the main - affixiation - class (AKA verbs where you derive the marû stem via suffixing -e to them), not to mention that you just can't tell what class of three (really more, but simplified to three for convenience's sake in my textbook) a verb stem is in (and this is not helped by the language's swaths of homophones); a requirement that you mark all arguments on the verb (including all its oblique arguments), including their respective person, case and number, including using a large number of sets with highly defective paradigms; the split-ergativity means that the core arguments completely flip how they are marked on the verb whenever one wants to say something in the marû (e.g. active imperfective) voice; and the fact that even nouns, though vastly simpler than verbs, still have the very confusing genitive construction.

Finally, its insanely defective and polyvalent logosyllabic orthography is even harder to learning the Maya Script, though I must admit, I have not intensively studied the latter. That said, in the words of my favourite linguistics professor (a Mayanist who has studied the Popol Vuh and is fluent in Yukatek and can read and write in the Maya Script: "The Maya glyphs are great, they look just like fancy versions of what they stand for; this [pointing at a slide with an Ur III building inscription text in cuneiform] just looks like a bunch of lines to me. Kudos to the Sumerologists and Assyriologists who deciphered this, I never could have."

[hadad]

Navajo for me. That's because I'm still getting used to the concept of polysynthesis, and to throw onto that many grammatical forms and complex sound changes, it would be much to get accustomed to, and be a confusing start.

IndoEuropean languages if I wasn't born speaking them, and wasn't already familiar with them. Just trying to write down the sound changes for ProtoIndoEuropean, or even writing a grammar for english, has shown me that. I appreciate the simplicity found in many other languages, like those of Chinese and Japanese, that either use affixes (Japanese) or particles (Chinese) to modify nouns.

To make things even better, the IndoEuropean languages have trouble deciding where you are to conjugate in different meanings. The safe assumption would to say things like grammatical voice are conjugated into verbs, but you'll find them in pronouns as well (For example, the reflexive voice - myself).
Then there's the use of words like "have" ("I have a drink") used to mark tense instead of the concept of "having", but that's where english gets easier/simpler....

[Micamo]

Anyway, I found this from the Economist on Tuyuca; frankly, I think they don't know what they are talking about (which is fine, being that they are not linguists but, well, economists). Also, I don't know if the so-called noun "classes" are not just affixial classifiers; perhaps somewhat like Jakaltek's?

Have a book!

[Lambuzhao]

This is a slippery-slope question. The things that wig me out might be someone else's "cup of tea". But here goes.

I think, for me, an L2 that uses irregular declensions (like Greek masculine 1st declension, or feminine 2nd declension) would be tough to learn again from scratch.

An L2 that exhibits many deponent verbs, or relictual verb forms, especially for common (let's say Swadesh List) verbs, that really take a left-turn from how other verbs (are expected to) behave/conjugate, would present challenges to learn.

Insane, stackable, repeatable agglutinations, like in Sumerian (and Etruscan, if I remember rightly). If I were a vampire, this alone would be a barrel of pickled garlic bound with silver nails and barrel-rings, with lovely crucifices woodburned into the grain. Yeah, I'm not yer secular, Twilight vampire.

Just plain silly, historically-fossilized orthographies as in French, English or (somewhat) Gaelic. The kind of orthography that had to have been invented in a long-forgotten Monty Python sketch. On the other hand, the shorthand-like writing system of Arabic was a really big hindrance to me learning more of that language, despite a very patient Arabic-speaking girlfriend.

Personally, I think writing systems like Chinese and Egyptian Hieroglyphic make a lot more sense than some "alphabetic" or "abugida" systems.

But this is just for me, personally. I've never had, as Xing suggested, a proper motivation to learn any languages like the above (except Greek and Latin and Gothic).

And, as Xing and Micamo suggested, if I had to learn an L1 like any of the above, I wouldn't be complaining: I'd be used to it (Whoa. "used to" Sumerian - now there's an interesting thought ).

Heck, my L1 is English, and as an L2, English freaks enough people out over the world on an hourly basis.

[Yačay256]

Anyway, I found this from the Economist on Tuyuca; frankly, I think they don't know what they are talking about (which is fine, being that they are not linguists but, well, economists). Also, I don't know if the so-called noun "classes" are not just affixial classifiers; perhaps somewhat like Jakaltek's?

Have a book!

Thanks!

[Omzinesý]

Difficulties also depend on what level of proficiency you are going to achieve.
English is said to be easy at the beginning. You can form complete sentences just putting words in line. But when it comes to styles and registers and numerous French loan words and their connotations, English gets hard.
(For me, English was the language with which I had to learn the concept of a foreign language, because I always do things theoretically. And they didn't teach general linguistics to a seven-year-old I hope I'll not see so horrible language studying anymore.)
Learning Arabic gave me the challenge of a new vocabulary. It just doesn't have common European words and learning how Arabic forms its vocabulary takes time, but now that I (more or less) manage the system of triliteral (now well typed) patterns Arabic gets easier. If I one day begin to learn a new "exotic" language I have to prepare to learn the whole vocabulary.

I have never studied a polysynthetic language, and English is the only analytic language I have studied (and it still has some synthetic derivational processes). So I cannot say how it is to learn a polysynthetic or analytic language, a language with a different morphological type.
But I think that studies in linguistics help with grammar, but the vocabulary must be learned again and again for every language.

Pronunciation could (maybe) be called difficult more objectively.
I guess pronouncing Halkomelem would be more difficult for a Japanese than the other way around.

[Solarius]

For me, irregularity is the killer, whether it be in spelling, morphology, or whatever- that's why I find French more difficult than Spanish- French orthography is less ordered.