Language Learning Difficulty for English Speakers | Hacker News Reader

Language Learning Difficulty for English Speakers

en.wikibooks.org

22 pointstshtf34 comments

Four out of five of the most difficult languages are east Asian; not too surprising there. The fifth, however, is Arabic, which I didn't anticipate.

Anyone know why it's apparently dramatically more difficult for a native English speaker to learn Arabic compared to other languages which would seem (to me, at least) equally remote from English in terms of geography, syntax, phonology, and lexicography?

I might hazard some informed response on that, as a native speaker of English and student in classroom instruction at various times of German, Russian, Modern Standard Chinese, Cantonese, Literary Chinese, Biblical Hebrew, Attic Greek, Biblical Greek, Taiwanese, Japanese, and Hakka. But what other languages did you have in mind that seemed to be rated as surprisingly easy by contrast with Arabic?

P.S. Being quite the language hobbyist, I have also self-studied a variety of other languages, including several more that are on the list included in the submitted article. Modern Standard Chinese is by far my strongest acquired language, but I would agree that it does not come easily for a native speaker of English--and arguably not for most native speakers of Sinitic languages.

http://news.xinhuanet.com/english/2007-03/07/content_5812838...

Disclaimer: I am almost completely uninformed about other languages beyond assorted superficial details about a handful. That said, I would have expected Arabic to be of no greater relative difficulty than, say, Hindi, Persian, Hebrew, Mongolian, or such.

I mostly expected Mandarin, Cantonese, and Japanese to be of greatest relative difficulty, from what little I know about those languages, and would have expected Arabic (and many other languages) to be in the second category.

Hindi and Persian are in the Indo-European language family, which provides a ton of similarities and possibilities for innate understanding of the underlying grammatical structure, which is not so much the case for Semitic languages (Arabic in particular)...

What's brutal about Chinese for a native speaker of English is phonemic tone for speech and lack of cognate vocabulary, coupled with a writing system that reflects a stage of development of writing from 2,000 years ago. But what is forgiving about Chinese is that the grammatical relationships are mostly shown by word order and function words, as in English, rather than by inflection of words. That, to me at least, makes speaking Chinese "easier" than speaking Russian, even though Russian is much more cognate with English.

Arabic may be especially difficult (even as compared to Hebrew, a language I have studied much more) because the standard form of the language is so divergent from the day by day vernacular forms, as noted by a more informed participant who has kindly posted another comment.

The United States federal government language proficiency rating system attempts to identify equal levels of real-world functional proficiency for ratings of interpreters and the like. When I was last tested, the rating scale went from 0 (with a readily apparent meaning) to 5 (a rating in principle reserved for educated NATIVE speakers of the language) in the functional domains of speaking, listening, reading, and writing. I was quite proud of my 4+ rating in speaking and listening to Modern Standard Chinese, but I took quite a few years out of my life to get to that level, as have most westerners I have ever met who have done the same.

Is the state of development of writing from 2000 years ago referring to the Chinese or Latin alphabet?

The Latin alphabet, or rather Phoenician, from which Hebrew, Arabic, Latin, Greek and Cyrillic scripts are descended, is more like 3500 years old.

But both the specific letter forms of the Roman alphabet and, more relevantly, the general spelling patterns of English are much more recent than that (as they would have to be). But, yes, writing is more ancient in the Near East than it is in China, for sure. It's just that it has gone through more language-specific adaptations, especially in the direction of exhaustively spelling all speech sounds with a minimal number of writing characters, than writing in China has.

It's a reference to Chinese characters not having been further developed in the direction of sound-indicating writing (as was done later in Japan and Korea). Full historical details, based on quite recent archeology in China, can be found in the writings of William Boltz. I searched for some examples of his writings that are online and not behind a paywall, but didn't find any convenient examples. His observation is that Chinese almost went in the direction of explicit syllable-writing (as Japanese has had for several centuries in the katakana and hiragana writing systems) but development in that direction ceased during the Han dynasty.

P.S. I just found, by browsing, that the Wikipedia articles about Chinese writing still need a lot of clean-up. Too bad.

The real hard part of Arabic is the morphology. It's "irregular", both in the human sense that there's all sorts of horrible exceptions to ridiculous numbers of rules, and in the compsci sense that you cannot recognise/produce the inflected forms of nouns and verbs using a finite state automaton/transducer. (Though, that's true for all the Afroasiatic languages including Amharic, Hebrew, and Hausa, but the FSI claims those are only "Level II" languages).

Also IMO Arabic phonology is a lot harder than either Chinese or Korean. (Chinese has tones, Korean has tensed versions of the same consonants we have in English, but Arabic has the pharyngeal and glottal consonants, completely unlike anything in English). On the other hand, Georgian's worse.

Finally, if you take Arabic in a school, they're going to teach you Modern Standard Arabic AKA fusha, which is basically in the same position that Latin was in 14th-century Europe: useful if you like to do nothing but read, useless if you actually want to understand what people are saying.

Having acquired Finnish (twelve years ago) and German (in the last two years), coming from a Bulgarian background, I've been thinking about language transfer. I don't think this list is very specific to English.

My own limited inquiry into the matter (collecting general opinions from people who have learned foreign languages and ordering them by subjective difficulty) seems to generate a list where difficulty correlates strongly with distance from Latin. I suspect this is related to the way languages are taught, rather than the way they are structured. Our formal grammars appear to fit Latin very well, probably for historical reasons. Here's some ideas:

1. Finnish is known worldwide as a difficult language to learn. It has an extremely complex grammar (when put in Latin terms). Finnish-born children acquire Finnish in about the same time Spanish-born children acquire Spanish, but Spanish adults have much more difficulty learning Finnish than Finnish adults have learning Spanish. This suggests that teaching methods might favour Latin-like languages. A further supporting data point is that foreign people living in Finland who take Finnish courses do better at tests than people who don't, but take much longer (many years) to start actively using Finnish in casual situations (conversation) and are never as comfortable as people who acquire it without courses. This might of course be a factor of what kind of people take courses.

2. Finnish sentences are structurally some of the simplest, possibly the simplest of any European language.

3. The European languages furthest away from Latin appear to have the most complex case systems. Cases appear to be a hack to fit a language into a Latin grammar. Cases in point are Russian, Hungarian and Finnish.

4. Constructing a Finnish grammar that is almost entirely correct is possible without the use of cases.

5. This sort of grammar is more than sufficient for everyday casual communication.

This leads me to think that people who take Finnish courses might be too afraid of making mistakes to actually communicate. People who don't know they're "doing it wrong" focus more on the result of communication rather than the formal correctness. I know I am consciously ignoring a lot of the grammatical rules in German, but I can communicate in it freely. Native German speakers may get confused by me switching cases and articles at times, but I am always understood. I tend to prefer German to all other languages I know (all of which I have used for many more years than German) when speaking to native German speakers. I wonder how the list in the OP would look like if the target was a "conversation-sufficient" subset of the language. This would be much more meaningful than "general proficiency" (and much less testable).

When you go to school in Sweden, you get taught a Swedish grammar that is also slightly wrong, but one that makes Swedish to be closer to Latin than what it is. The purpose of this is to make it easier to learn the other big European languages like English, German, French or Spanish.

However, if you take linguistics at university level in Sweden, the first thing they tell you is to forget everything you were taught earlier, because a lot of it is wrong.

I'm really curious if this actually works, if it really makes it easier for kids to learn the other languages by learning a wrong version of grammar for their own language.

Actually, I meant it the other way around. Teaching a grammar that is constructed from the language, not one that fits better to Latin. I think it would help if each language was thought with a specialized grammar for that language rather than through subsets and hacks on top of a "universal" one.

It would be very interesting to see this list from the perspective of other languages. This list only shows their "distance" from English and not from each other, but would it for example be easier for a Chinese person to learn Japanese or English?

And if you take it further, could you make a generic ranking list and see which languages are generally the easiest to learn?

Here in Taiwan, I meet tons of Mandarin speakers learning Japanese. In general, I'd say it's pretty easy for them. There are a limited number of sounds in Japanese only a few of which pose difficulty for Mandarin speakers and the kanji are a slightly simplified version of Chinese characters.

There are a lot of cognates between Chinese and Japanese, there's a good deal of overlap in terms of history, food and religion and best of all there's nearly as much Japanese media available for those who want it as there is English media.

However, the Japanese people I've met here studying Mandarin usually have a tough time of it. They can usually read pretty well, but their pronunciation and general communication abilities tend to be far worse than those of the students of every other nationality I've met at the language centers. The difference was absolutely shocking when I first started studying here.

One sorta starting point is the list of average TOEFL scores by native language of test-takers, which should tell you who finds English easy and who finds it hard. Page 9 in the link [1]. Of course, they say "use of average TOEFL scores to infer the overall English-language proficiency of a population or of a country is not appropriate due to a variety of variables that influence the test-taking population" =)

20 languages whose speakers get the worst scores: Lao (56), Bemba (58), Wolof (61), Mossi (62), Javanese (64), Khmer (65), Japanese (66), Arabic, Fula, Uighur, Uzbek (69), Burmese, Hausa, Kurdish, Vietnamese (70), Kazakh (71), Amharic, Oromo, Somali, Thai, Turkmen (72)

I wonder how much of this is due to how "complex" your own language's morphology and syntax are. For example, Mongolians tend to be better at English than either Japanese or Koreans who have studied for the same amount of time. And Mongolian's morphological and grammatical rules are much more complex than Japanese or Korean. (Anecdotally, Mongolians also find Japanese and Korean fairly easy to learn due to the structural similarities [2]).

For example, all those languages mark the role of nouns in a sentence by tacking on extra syllables ("particles"/"case markers"). For any given particle, Japanese invariably has one form, Korean one or two (some change if the word ends in a consonant), and Turkish usually two (vowel harmony). And the forms with pronouns follow the same rules. But Mongolian regularly has as many as eight forms depending on vowel harmony, what kind of consonant the word ends in, random rules for inserting "secret n"s and "hidden g"s, plus the forms for pronouns are irregular, plus there's rules for dropping vowels sometimes.

[1] http://www.ets.org/Media/Tests/TOEFL/pdf/test_score_data_sum...

[2] http://www.nytimes.com/2003/10/12/world/mongolians-and-korea... --- though contrary to what it says, linguists don't actually believe the two languages are related

Their "languages closely related" is way off. They should have just labeled it "easy languages", since that's apparently what they mean. Icelandic, in the third category, is uncontroversially more related to English than most of the (predominantly romance) languages in the first category, as English and Icelandic are both Germanic languages.

While grammatically, English is Germanic, it did borrow a ton of vocabulary from French after the Norman invasion, so from a lexical perspective it's related to the romance languages by proxy.

Yes -- and it also inherited French-style plurals (blessedly), and I believe is the only Germanic language with a gerund.

I like the way they've arranged all the vocab, pronounciation, grammar links together for each language in a nice organised fashion.

I'm a little skeptic though of their classification of difficulty. I understand that they're categorising it based on English being a starting point, but after a certain level of proficiancy (what most people would classify as "intermediate" level), language learning just becomes a matter of constant exposure. I'm sure this would hold equally for both French or Japanese (apart from writing).

Yes, but it still requires more exposure for the more difficult languages.

I am by no means proficient in anything other than my first language, but I found other languages much easier to start on once I'd learned lojban. it "untaught" lots of things that I took for granted, and gave me some new structures and ideas that served me in good stead.

It is said that your first acquired language takes four years, subsequent ones take two each. Lojban takes about six months. Perhaps if you want to learn a language, learn lojban first.

They've actually done studies on that - not with lojban, but with another simple constructed language: Esperanto. They found that students studying Esperanto for a year followed by French for 3 years fared better than students studying French for 4 years:

http://en.wikipedia.org/wiki/Esperanto#Language_acquisition

Any such claim about Esperanto should be regarded as a claim that needs further verification.

http://norvig.com/experiment-design.html

http://www.xibalba.demon.co.uk/jbr/ranto/

Anyway, it is very doubtful that such a claim, even if true for speakers of English learning French in England, would be true for speakers of Chinese learning Japanese in China (or the other way around) or for any number of other likely pairs of native and acquired languages.

I only mention the Esperanto studies because (1) they seem to confirm RiderOfGiraffes' intuition (2) AFAIK Esperanto is the only auxlang for which such studies have been done.

I suspect similar effects could be found with lojban, and would love to see more research in that area. However AFAIK no such research currently exists.

I read the ranto ages ago and while it makes many good points, none of them are particularly relevant to the suggestion that the studies were faked. In fact imperfections like sexism could be advantageous if the student then proceeds to a similarly sexist European language (i.e. most of them).

I wonder what is meant by general proficiency?

I'm surprised that Arabic and Hebrew aren't in the same category.

(disclaimer/foonote: I'm a native speaker of Arabic)

A relative lack of dialectical variation in Hebrew might explain it.

Thanks, hadn't even considered that!

Everybody in India can speak minimum 3 languages.

Interesting list. I can certainly agree that Russian is a difficult language to acquire for native english speakers; far harder than Spanish.

I am curious, however: has anyone found nonobvious ways to make money with language skills (obvious meaning translator, etc)? While some of these, esp eg Arabic or Russian, might be very interesting to learn, and it would probably be a great experience living in an Arabic speaking country, there are high costs -- the education and the time living in a country and not getting much done but language courses. What are good ways of making this a financially sound thing to do, or will it always be solely a cultural experience?

Some time ago, I read an article by an English teacher living in Russia. She said you can make a good living there by privately teaching English.

I've met some multiligual people. One spoke about a dozen languages (fluently, but far from perfectly, AFAICT), and I think he had a college degree in the humanities even. Worked as a car mechanic, so I guess the answer might be no.

In general, the return on investment for a native speaker of English learning any language as a second language is lower than the return for the same person learning a lot of mathematics. People who don't know English usually find better gains by learning English, but they still have to learn some other career skill as well. The least expensive and most effective way to learn a language is to learn a bit of it before foreign residence (so that you'll live in fear of not knowing the language, and study hard) following that by foreign residence and continued part-time focused language study while working and participating in the culture in the new country. And that's just how a lot of second-language speakers of English reach very high levels of proficiency in English.

Reply on news.ycombinator.com