Comparative language difficulty for English speakers

This morning I found a copy of the chart the Foreign Services Institute uses to grade the comparative difficulty of world languages for acquisition by an adult monoglot English speaker.

I have an unusual perspective on this list for an American. I’m a low-grade polyglot; I have spoken three languages other than my birth English and can read a couple others with Google Translate. I have studied comparative linguistics; I know a bit about the morphology and phonology of many of these languages. I have received street-level exposure to over a dozen of them in my extensive travels, and I have a good ear.

So, I’m going to try to add some value to the list with additional notes and comments.

First, the list itself:

Category I: 23-24 weeks (575-600 hours)
Languages closely related to English
Afrikaans
Danish
Dutch
French
Italian
Norwegian
Portuguese
Romanian
Spanish
Swedish
Category II: 30 weeks (750 hours)
Languages similar to English
German
Category III: 36 weeks (900 hours)
Languages with linguistic and/or cultural differences from English
Indonesian
Malaysian
Swahili
Category IV: 44 weeks (1100 hours)
Languages with significant linguistic and/or cultural differences from English
Albanian
Amharic
Armenian
Azerbaijani
Bengali
Bosnian
Bulgarian
Burmese
Croatian
Czech
*Estonian
*Finnish
*Georgian
Greek
Hebrew
Hindi
*Hungarian
Icelandic
Khmer
Lao
Latvian
Lithuanian
Macedonian
*Mongolian
Nepali
Pashto
Persian (Dari, Farsi, Tajik)
Polish
Russian
Serbian
Sinhala
Slovak
Slovenian
Tagalog
*Thai
Turkish
Ukrainian
Urdu
Uzbek
*Vietnamese
Xhosa
Zulu
Category V: 88 weeks (2200 hours)
Languages which are exceptionally difficult for native English speakers
Arabic
Cantonese (Chinese)
Mandarin (Chinese)
*Japanese
Korean
* Languages preceded by asterisks are usually more difficult for native English speakers to learn than other languages in the same category.

Generally I think this list is dead on target. I agree with most of it even to the level of which languages should be starred in their category.

Before I launch into specific discussion of the exceptions, a minor caveat that I think I might have a tendency to underweight phonetic difficulty because of my Frodo ear – there are very few phonemes that really throw me for a loop, and I can hear tones with little effort.

That said, here we go:

Arabic should be starred. It is grammatically and phonologically extremely difficult; from my exposure to both I’d say significantly more difficult than Chinese without traditional Chinese writing.

If I were to add a mark for “easy in its category”, I would put that on Spanish and Italian. These are significantly easier than the other Category I languages and (I believe) the easiest of all world languages for English speakers to learn.

Portuguese is a freebie if you learn Spanish and can get used to an odd but relatively consistent shift in the phonology, mainly heavy nasalization of everything.

I think putting German in an intermediate category II above other major European languages in difficulty is very shrewd. Phonology is easy (I can pronounce German with a good accent even though I don’t speak it) but the grammar is significantly more difficult than say, French.

If not for the help German gets from having Indo-European cognates, I think Category III (Indonesian/Malay and Swahili) might be easier than German. These are areal trade languages, quite probably creolized from ancestral trade pidgins, and have retained a simplicity that makes them relatively easily acquired by adult speakers. (English has a similar history.)

Chinese is a strange case. It is indeed brutally difficult in toto, but if you (a) are willing to settle for spoken fluency only, and (b) have a musician’s ear for tone (as I do) I suspect it falls to category IV and possibly to a hard category III. The positional grammar of Chinese is simple and easy to acquire.

Given an “easy” mark I would also rate Polish an easy class IV for an educated English speaker, mainly due to a much heavier infiltration of Latin roots than in other Slavic languages – this make recognizing cognates easier.

Persian/Farsi/Dari rates an “easy” mark in its category due to simple and regular grammar. If I were going to try to grok the Indo/Persian group I think it’s a tossup whether Farsi or Hindi would be the best point of entry.

Lithuanian deserves a star, I think. Phonology isn’t bad but the grammar will break your brain.

Xhosa definitely deserves a star. When the phonology is difficult enough that I have trouble retaining the distinctions, most English-speakers would be lost beyond hope. Clicks and implosives, man!

On the blog where I found this list, a commenter opined that there ought to be a category zero for Esperanto. Yep.

Published
Categorized as Science

240 comments

  1. Come now, Eric. I’m disappointed. An SF fan like you who didn’t even try to rate Klingon?

    1. >Come now, Eric. I’m disappointed. An SF fan like you who didn’t even try to rate Klingon?

      Category VI. Designed to be weird and difficult for English-speakers. It’s the INTERCAL of conlangs.

  2. esr:

    I would disagree with this list some. Generally, Turkic languages are much easier than languages like German grammatically. They don’t share much vocabulary with English and Latin-based languages but I’d estimate those languages easier to learn. There is also no gender there, so an English speaker will find the unisex atmosphere of those languages familiar.

    I think you are correct about Chinese. Once you get over the tonal nature of language, the grammar is easy, and if you’re only interested in speaking but not reading/writing I think that is easy. I have seen a few Americans who work for Taiwanese tech companies in Taipei manage to have done exactly that.

    I have heard mixed opinions on Arabic’s difficulty. Every once in a while I teach an Arabic class at a nearby university for fun. From some people, I have heard that it is no more difficult than German. From others, I have heard it is much more difficult than that. Surprisingly, female students do extremely well in-spite of the mathematical nature of the grammar. Also, I know from experience that native German speakers find it much easier to learn Arabic than native English speakers. It is also much easier to learn the spoken forms of Arabic (particularly Egyptian form) than modern standard Arabic (the literary language of the media). Speaking of google translate, the intermediate language it uses is somewhat modeled after Arabic in its template-based mathematical derivation, though more limited in its ability to derive words and precise meanings.

    Arabic (and semitic languages in general) imo, are the lisps of human languages. Arabic is the clojure of the semitic languages (in its beauty and grace), with the conciseness/correctness of a haskell. It has even more rhythm built into the language than Italian but does not sound as beautiful as Italian to an English speaker who doesn’t know both languages.

    1. >Generally, Turkic languages are much easier than languages like German grammatically.

      English is not your birth language, so your judgment about where a language belongs on this list is unreliable.

      That said, I agree with you that some aspects of Turkic morphology are refreshingly simple and regular compared to a medium-complex inflectional language like German. For a native speaker of English, though, the near-complete absence of cognates is a huge barrier. So are some phonological features like vowel harmony.

    1. >How exactly is Japanese the most difficult language? The sounds are few and simple.

      I agree. Phonologically Japanese is dead-easy. It’s the grammar that’s a nightmare for native English speakers. Perhaps not as bad as Arabic (which is also phonologically difficult), but really bad.

  3. @esr –

    > I have spoken three languages other than my birth English

    You’ve mentioned elsewhere that you were (are) functionally a crib bilingual, Spanish and English (a Venezuelan nanny, IIRC) – which two other languages did (can) you speak?

    1. >which two other languages did (can) you speak?

      French and Italian. I’m very rusty at the spoken forms but retain substantial reading comprehension. A few years back, when I had to translate idiomatic French on the fly during an IRC session I found I could still do it.

  4. While one could argue about the comparative difficulty of German (except I probably can’t; suppose I’m disqualified from that discussion by having German as my first language), the first two categories are definitely mislabeled: As they stand now, they strongly imply that Italian and Spanish are more closely related to English than German is.

  5. @Christian R. Conrad –

    > [T]hey strongly imply that Italian and Spanish are more closely related
    > to English than German is.

    Assumes facts not in evidence. The OP called this a table of “comparative difficulty of world languages for acquisition“. That is a function of many factors – similar phonology (less new sounds to have to learn how to hear and produce), cognate vocabulary, similar grammar, etc.

    I have a T-shirt whose caption goes something like this: “English doesn’t borrow from other languages. It follows them down dark alleys, mugs them, and then rifles through their pockets for loose vocabulary.” (Also grammar.) English may have started out cousins with German, but it has pillaged enough from many other languages of the world to bear, in many cases, a closer relationship to them than to its roots.

  6. Indonesian is a little like chess; easy to learn and incredibly difficult to master. As Eric noted, Indonesian is a trade language which is descended from a language called “Pasar Malayu” which literally translates a “Market Malay.” As such, you can use it as a sort of pidgin, with a vocabulary of perhaps 1000 words – essentially Market Malay – and get by very nicely in Indonesia, but you can also write a mathematics textbook in the language, in which case it has the same resemblance to Market Malay as chess has to checkers… Depending on where in Indonesia I was, I sometimes got complimented on my Indonesian (at least the accent) and was sometimes told my accent was abominable.

    I do have one funny story about the way the written language works. In Indonesian, the “c” is pronounced as “ch.” Thus, an untutored Indonesian looking at the word “city” will pronounce it as “chity.” On the island of Sulawesi they turn the “ch” into an “sh,” and I was asked by a self-taught English speaker “What shitty do you come from?”

  7. As a native Lithuanian speaker, I’m wondering what exactly it is that makes the grammar seem so difficult to foreigners. (The declensions? Pretty sure some languages from the same category have even more.)

    1. >As a native Lithuanian speaker, I’m wondering what exactly it is that makes the grammar seem so difficult to foreigners.

      Yeah, it’s the declensions. Lithuanian is near the high end of Indo-European languages in number and usage complexity of declensions.

  8. “It’s the INTERCAL of conlangs.”

    Which makes it even more delicious that it is the most widely spoken conlang, something that pisses off Esperantists no end. Considering the political leanings of those, however…

    Mantas, any language with 11 noun cases to deal with declensions for is …going to be tough for a native English speaker to handle. That’s worse than Latin, which has earned its notoriety in that area well.

  9. German isn’t RPN. It’d be somewhat easier if it were. Verbs only come at the end if they have helper verbs (“Ich will schlafen”); if not, they go before the direct and indirect objects (“Wir geben ihm den Tisch”). And prepositional phrases put the prep first (“Er geht am Strand”). Also, other modifiers have a best ordering, which shouldn’t matter that much in RPN: “Er geht heute am Strand” looks right, but “Er geht am Strand heute” looks weird, IIRC.

  10. I wonder how hard it would be to use the above difficulty chart to derive a least-effort language set that covers the optimal area of the earth, and/or optimal number of speakers.

    1. >I wonder how hard it would be to use the above difficulty chart to derive a least-effort language set that covers the optimal area of the earth, and/or optimal number of speakers.

      Maximizing coverage in megapersons divided by hours of learning time, that you’d do with English plus Chinese.

      Speaker area divided by learning time: Probably English plus Spanish, with rapid fall-off in gains per hour from additional languages.

  11. @esr:


    English is not your birth language, so your judgment about where a language belongs on this list is unreliable.

    No doubt. But I was educated in English-medium schools (International Schools, the schools diplomats, expatriates etc send their children to). So I do have some perspective on this.


    So are some phonological features like vowel harmony.

    Vowel harmony cannot be more difficult or alien a concept for English speakers than it is for Persian, and Arabic speakers who all find Turkish and Turkic languages very easy to learn even though those languages are from a completely different language family. Persian is the simplest indo-euro language in terms of vowels. Arabic is even more extreme where vowels mainly serve the role of adding music/tuning to words/meanings that largely encoded in consonantal roots.

  12. I’m reminded of the effort to build a directed graph based on the following relation: language A is “harder” than language B if language B contains an idiom referencing language A and meaning incomprehensibility, e.g., English “It’s all Greek to me”.

    Turns out, the graph converges on Chinese as the “hardest” language, being “harder”, directly or indirectly, than any other. And Chinese doesn’t reference other human languages, having idioms like “heavenly script” and “chicken intestines”.

    http://knowmore.washingtonpost.com/2015/03/25/the-equivalent-of-its-all-greek-to-me-in-30-other-languages/

    Even Japanese has chinpunkanpun meaning roughly “mumbo-jumbo” but with a Chinese-ish sound imitative of scholarly language.

  13. > I agree. Phonologically Japanese is dead-easy. It’s the grammar that’s a nightmare for native English speakers.

    I like to think that the year and a half of it I took in college left me with a reasonably good understanding of the grammar – the difficulty for me was the completely alien vocabulary.

    You’ve sort of encapsulated this issue in the difference between category I and II (being exclusively Romance and Germanic languages) and the higher categories, but you don’t have any separation between Indo-European and non-IE languages.

  14. > On the blog where I found this list, a commenter opined that there ought to be a category zero for Esperanto. Yep.

    f.s.v.o “ought to”, maybe. I don’t think that Esperanto is any more similar to English than, say, Italian is. Its position (along with Interlingua, etc) as a language designed to be reasonably easy for all Europeans to learn doesn’t really earn it a place on a list of languages easy for English-speakers in particular

    I’d say if there were a category zero it should include the “easy” Germanic languages (e.g. Dutch, Flemish, and maybe Danish), which can at times sound like heavily accented English… and British dialects like Scots, which is often actually mistaken for heavily accented English.

    1. >I don’t think that Esperanto is any more similar to English than, say, Italian is.

      Similarity isn’t the metric the chart is expressing. Ease of acquisition for an adult native speaker of English is. Esperanto happens to be extremely easy, enough so that a few international conferences have successfully addressed language barriers between English Speakers and others by handing out Esperanto primers and saying “Take a couple of hours to learn this.”

      >The “easy” Germanic languages (e.g. Dutch, Flemish, and maybe Danish

      I speak from experience when I say that Swedish and Norwegian are easier than Danish, which is rapidly degenerating into a glottal mush that is increasingly difficult for non-natives to even hear. (Written Danish remains relatively easy.)

      The trouble with these “easy” Germanic languages is that while recognition competence is not a lot of work to pick up (I can sometimes puzzle out the meaning in Swedish newspaper stories) production competence is elusive. You will come up to production speed in Spanish or Italian after a lot fewer hours than in Dutch or Afrikaans.

    2. >British dialects like Scots, which is often actually mistaken for heavily accented English.

      Yeah, that’s…complicated.

      Having read up on the linguistic situation and been to Scotland twice, here is what I think is going on. There are three languages being spoken in Scotland today. One is Scottish Gaelic, an insular Celtic language closely descended from Irish Gaelic. Another is Scots (proper) an Anglic-family language closely related to Modern English but with a separate line of descent from Middle English (this is the language Robert Burns’s poetry was written in). The third is a Scots-influenced dialect of Modern English.

      What makes the situation complicated is that the boundary between Scots and Scots English is vague, porous, and shifting, and both are sprinkled with lexical borrowing from Scottish Gaelic. It’s not even certain that anyone speaks “pure” Scots anymore, rather than a contact language formed between Scots and Scots English – and especially difficult because Scots and English were about as similar as English and (say) Frisian even at the time of maximum divergence, which was at least several centuries ago now.

      So your “Scots” may in fact be heavily accented English. Or not. There are some seriously headache-inducing definitional issues in play.

  15. I generally agree with the list, with a few major exceptions. I am the opposite end of the spectrum — native English speaker with no talent for languages, have never lived overseas, total monoglot with minimal academic exposure. I have taken a few language classes and briefly studies some written languages from introductory texts.

    I agree that Spanish is the easiest. Cognates, simple grammar, no weird phonemes. I really want to make the effort to learn this someday and I don’t expect it to be very difficult.

    French is slightly harder. It has a few phonemes that don’t map. I don’t think I will ever be able to pronounce a French “R” properly. The grammar is also a bit more complex than Spanish, but it’s not difficult. Most verbs are regular and fall into one of three patterns. There are quite a few verb tenses, but they generally map back to English. Cognates are very common thanks to the Norman invasion.

    German has the cognates and in some ways feels that it should be easy without actually being easy. Part of the problem is the German technique of creating new words by jamming multiple words together to form long compound words. These probably make intuitive sense to a native speaker, but sound like organic chemistry to English-speakers. (And in fact that’s not a coincidence; organic chemistry was German near-monopoly in the formative 19th century.) In this respect it is similar to some Indian languages such as Mohawk. (The Mohawk word for wine is oneharadesehoengtseragherie, which literally translates as “a liquid made from the juice of grapes”. That’s very German.)

    Thai easier than Japanese? Sorry, I have to strongly disagree. I took 2 semesters of conversational Thai in business school, taught by a native speaker, and I’ve visited Japan and looked at introductory books on the written language. Japanese has no tones, no verb tenses, and a general sound very distinct from other Asian languages. If you put a Japanese, Chinese, and Thai in a room and had them talk, I could very quickly tell you which language was which, though without being able to understand any of them. But I could never say “sawat dee kha, bai de mai?” without getting a native speaker giggling, and there’s a reason for that.

    Eric seriously underestimates the difficulty that tonal languages give to most English speakers, who are *not* musicians and do *not* have trick ears. Even four different tones, as in Thai, is enough to make words that are very different sound the same to us. Whereas more complex grammar is just a thing to practice.

    Mandarin and closely-related languages are undoubtedly the hardest. No cognates, many different tones, made worse by obscure characters. In some ways I’m surprised that Mandarin didn’t benefit from a creole effect as the culture was spread out first across China proper and later by the Mongol conquest. The characters were simplified under Communist rule, but that was done by fiat rather than cultural evolution.

    1. >In some ways I’m surprised that Mandarin didn’t benefit from a creole effect as the culture was spread out first across China proper and later by the Mongol conquest.

      Interesting that you mention this, because the personal theory of mine that is most likely to make historical linguists go “What the fuck? Well…maybe…” is related.

      I strongly suspect that Mandarin Chinese was originally an Imperial-court creole formed by collision between dozens of different related Chinese languages – which then, as the prestige language, of the capitol, influenced those regional languages to converge. But I think the creolization event well predated the Manchu reign, possibly as far back as 1000CE.

      The reason I think this is that Mandarin has the combination of stripped-down grammar and elaborate relexification you see in old creoles like…English. And other languages that probably have a far past as trade creoles: Malay/Indonesian and Swahili are examples already mentioned in this thread.

  16. KMFDM comments on English’s wholesale adoption of German words:

    https://www.youtube.com/watch?v=IIt3_7u6A9s&ab_channel=Zahav

    Cathy: “Part of the problem is the German technique of creating new words by jamming multiple words together to form long compound words.”
    I claim that German words have magnets coated with superglue on the ends. Get them too close together, and they stick and can’t be pried apart.


  17. >> The reason I think this is that Mandarin has the combination of stripped-down grammar and elaborate relexification you see in old creoles like…English.

    The grammar of Chinglish resembles that of Mandarin from what I was told. If China takes over, that will be the future of English.

  18. I agree. Phonologically Japanese is dead-easy. It’s the grammar that’s a nightmare for native English speakers. Perhaps not as bad as Arabic (which is also phonologically difficult), but really bad.

    Japanese grammar is not so bad as long as you remember the verb-terminal word order. Japanese is very roughly SOV; I say “roughly” because the actual order of the other components doesn’t matter too much if they’re marked with special particles. Tellingly, Japanese grammar lacks a lot of the features and complexities of European languages, such as noun declensions, articles, noun-verb agreement, adjective-noun agreement in gender or number, and so forth. Japanese is also “pro-drop”, meaning “if it can be omitted, it should be”. So such things as plural markers and the subject of a sentence will be omitted if understood from context.

    I’ve never had much of a problem with the grammar of the language, especially when compared to the likes of French (and German I won’t even countenance). Maybe I’m just really smart like that, maybe my exposure to Lisp and Forth made me more receptive to human languages with a different word order, I dunno. But I suppose it’s possible that Japanese’s verb-terminal-ness will prove a big hurdle for most English native speakers to overcome.

    But if you throw in two different writing systems, two different pronunciation systems (on-yomi and kun-yomi; any Japanese kanji word can actually have arbitrary reading but this is rarely used except by too-clever-for-their-own-good manga artists who give their characters’ special move names English or Spanish readings that may not have anything to do with the kanji), the fact that there are many different registers depending on whom you’re talking to and your social relation to them and whom you are referring to in your speech, and of course regional variations and dialects, and… YOUR HEAD ASPLODE.

    English grammar tends to trip up native Japanese speakers something fierce. Go to Japan and try to converse in English. Realize that just about every single person you talk to had several years of mandatory English education in grammar and/or high school. The ones who do acquire greater than basic English proficiency did so by spending a long time overseas. The state of English education over there is really, really bad; English is hard and Japanese insularity is still such that there’s not much emphasis on getting English students over the major hurdles.

    Our phonology is a total mindscrew for them too, as bad as Arabic is for us; not only because of the famous l/r merger, but Japanese has no labiodental or interdental fricatives, so they stumble on /f/, /v/, /?/, and /ð/ sounds. The vast majority of Japanese cannot pronounce the sentence “Darth Vader is Luke’s father” correctly to save their lives.

    (But what about Mount Fuji, you say? Well the Japanese “hu” is pronounced with lips tightly pursed as to almost approximate “fu”, which is why it is often romanized as such; but again there is no actual labiodental consonant in spoken Japanese.)

    1. >there is no actual labiodental consonant in spoken Japanese.)

      True, that. Japenese ‘f’ is like blowing out a candle in English. Fortunately that is about the oddest feature of the phonology and the rest is extremely simple, comparable to Italian.

  19. Go to Japan and try to converse in English. Realize that just about every single person you talk to had several years of mandatory English education in grammar and/or high school. The ones who do acquire greater than basic English proficiency did so by spending a long time overseas. The state of English education over there is really, really bad; English is hard and Japanese insularity is still such that there’s not much emphasis on getting English students over the major hurdles.

    If you’re in Japan and don’t speak Japanese, try writing in English with a Japanese person. This strategy can frequently make things much easier.

  20. In middle school and high school I had several years of Spanish and one year of Latin. Having had no opportunity to use my Spanish since moving away from Florida in 2007, I’ve essentially forgotten it. Nonetheless, I can eventually make sense of most writing in most of the Romance languages if I stare at them long enough. However, I find French considerably easier than the others even though I’ve spent exactly zero time studying it. Therefore, I think French perhaps deserves an “easy” mark, at least with respect to its written form.

  21. WTF? Japanese is easy. In fact, it makes way more sense than English.

    However, I am more inclined towards the polyglot…I actually enjoy figuring out the abstract structures underpinning languages, shaping their communicative power.

    1. >WTF? Japanese is easy. In fact, it makes way more sense than English.

      You’re odd. Even my friends who speak Japanese think it’s a difficult language.

  22. @esr:
    >You will come up to production speed in Spanish or Italian after a lot fewer hours than in Dutch or Afrikaans.

    Are you sure that your childhood Spanish exposure hasn’t biased you on this?

    1. >Are you sure that your childhood Spanish exposure hasn’t biased you on this?

      Yes. The time requirements for immersion courses from outfits like Berlitz and Pimsleur seem to show the same scaling. Well, with respect to Spanish vs. Germanic languages, anyway; Italian is not a sufficiently popular choice that I have stumbled over indications.

    1. >A friend posted this in response:

      Ah, yes, a classic. I had read the written romaji-Chinese and English translation, but not heard it recited before.

  23. (I’d kinda like to know what causes a video link to get inlined and what causes it to just be mentioned, so as to avoid surprises like the above…)

    1. >(I’d kinda like to know what causes a video link to get inlined and what causes it to just be mentioned, so as to avoid surprises like the above…)

      I’m not sure what the rules are – it seems to interpret a bare URL as an URL with empty link text. But you can write <a href=”…”>link text</a> and the right thing will happen.

  24. >You’re odd. Even my friends who speak Japanese think it’s a difficult language.

    I’ve had native Japanese speakers tell me so. One friend who says it’s rough for a little while going back to Japan when he’s been in the US for too long.

  25. I am a native German speaker, an almost-native American-English speaker (been speaking it since age 3), and a competent if slowish French speaker. When I travel to the Netherlands, I am hearing something that, but for political and historical reasons, might as well be called a strong German dialect. Dutch seems no more different from Hochdeutsch to me than the local dialects I hear in Switzerland. I’d say it’s comparable to the difference between Appalachian English and TV-anchor English.

    So in what sense is Dutch easier for Americans than German is?

    1. >So in what sense is Dutch easier for Americans than German is?

      The Dutch case system is simpler. Also, it (like English) never underwent the High German consonant shift; this makes a lot of lexical items more familiar.

  26. @Thomas Blankenhorn:


    So in what sense is Dutch easier for Americans than German is?

    It is all imagined. Possibly because the Dutch on average speak better English than native English speakers, they assume they reverse will be true and they’d learn Dutch so quickly.

  27. “So in what sense is Dutch easier for Americans than German is?”

    For me, at least, it’s closer to English, being pretty much a fusion of English, French, and German. Even the German bits are closer to English.

    Also, it helps in reading it if you know how to pronounce Dutch diphthongs: “uit” is pronounced the same way as English “out”, as one example.

    In Dutch, “IJ” is one letter (it’s the y with an umlaut at the end of the PC extended ASCII set), pronounced not very differently from how a Texan pronounces a long I. I got a surprised compliment at a Hercules gathering a long time ago from one Dutchman with an “ij” in his name by nailing the pronunciation the first time.

  28. As someone else who grew up with Lithuanian courtesy of parents on both sides, it is a crazy-making language to describe to english only or native speakers.

    Yes, “Darius” is my first name, it’s actually third person. It, like all nouns in the language are actually conjugated based on person, tense, etc.

    For example – “for whom”, address forms (speaking directly to) , etc. all have different conjugations.

  29. Interested in your opinion on English. Obviously it doesn’t fit well in this chart since English speakers obviously have already acquired English (some in the Chicago public school system accepted.) But this chart is a measure of two things really — the intrinsic complexity of the language and the distance from the particulars of English. So you can’t measure the second really, but you can rate the first against others.

    To me, much as with your comment on Mandarin, English is intrinsically a pretty simple language because it is a whore, and whores are easy since they have to accommodate the needs of many different suitors. However, English’s biggest complexity is its dreadful spelling, which again comes from its whorish behavior and the fact that the suitors just wanted to use it, so writing stuff down came way late in the game of linguistic formation. And it has a few other forms where you seem to have multiple ways to do the same thing, with subtly different meanings (the English Genitive and article for example.)

    But I’m curious to know your view.

    1. >But I’m curious to know your view.

      I think the question you are trying to form is something like: If you rank languages by ease of acquisition for adult speakers, where does English fall?

      This question has been studied, most notably in Nicolas Ostler’s excellent book Empires of the Word. And the answer is that English is exceptionally, almost ridiculously easy compared to most languages. Even the spelling of English, often cited as a bugaboo, is about 75% regular; it has a bad reputation because the irregularities are concentrated in the most common 400 words.

      (There is a well-established rule of language evolution that the rate of which irregularities are smoothed out is inversely proportional to their usage frequency; common irregularities are more likely used, and thus reinforced, often enough to be remembered, while rare ones are prone to be forgotten. Thus, the irregular verbs in French are the common ones.)

      Ostler’s book attributes English’s ease of acquisition to a specific set of morphological features it shares with trade pidgins, languages effectively designed for ease of acquisition: SVO isolational grammar. He notes that two other extremely successful trade languages, Chinese and Malay/Indonesian, also have this structure.

  30. Japanese pronunciation isn’t ridiculously tough, nor is memorizing two “alphabets”, (ok remembering a million kanji is hard), but yes, grammar probably puts it far and away. A lot of languages can be learned by watching their TV shows, but I’m pretty sure, while you might be understood, you’d make a fool of yourself doing so with jap, what with all the customs they have like uchi-soto in and out groups. That one sounds like a useful one to me, we have to contextualize with phrases like “and I’m telling you this as a friend”, wheras it’s built into the phrasing for them.

  31. However, English’s biggest complexity is its dreadful spelling, which again comes from its whorish behavior and the fact that the suitors just wanted to use it, so writing stuff down came way late in the game of linguistic formation.

    Specifically, the primary issue with English spelling is that the Great Vowel Shift happened at the same time as the spread of the printing press with movable type, so that the spelling of different words crystallized at different times during the upheaval.

  32. Apropos of your remarks on Xhosa, Eric, I should like to make an observation. You say:
    “Xhosa definitely deserves a star. When the phonology is difficult enough that I have trouble retaining the distinctions, most English-speakers would be lost beyond hope. Clicks and implosives, man!”
    I am inclined to agree that Xhosa needs a star for its click phonemes. However, that is not its most difficult aspect. It is closely related to Zulu, a language with which I am much more familiar, and I think the system of concordance between nouns and verbal and adjectival constructs is the part of these languages’ grammars which causes the most difficulty to English speakers, not to mention the fact that they are also semantically tonal and syntactically intonational.
    It is arguable that there are 17 ways of saying “the” in Zulu. It is also confusing for English speakers that the concordial prefix “u-” means “you,singular” when pronounced with a low tone and “he/she” when pronounced with a high tone.
    When you add the fact that there are many past and future tenses,eg. remote past continuous, remote past completive, (ordinary) past completive/continuous, remote past future completive, future completive, future continuous, etc., it is very strange to an English speaker.
    Also, the fact that predicates are constructed as single strings of prefixes, stems and suffixes in which the elements are agglutinative (like German multiple words) makes many English (and Afrikaans) speakers opt out and resort to a bastardised hyper-simple system (I hesitate to call it a language) of issuing instructions or requests called Fanagalo (the name is derived from Zulu for “resembles all”) which is merely the imperative case of Zulu with a few pronouns and adjectives and adverbs.
    Having said all that, I have to add this: Zulu and Xhosa and Sotho, etc., are extremely logical languages and exceedingly regular with regard to orthography. What you read is the way it’s said.

  33. Tamil is my native language. I know basic Hindi (including reading / writing) as well as English. I am learning Hindi through an English book, but I think Hindi cannot be learned purely with English grammatical knowledge. Significant re-learning of rules are required.

    Tamil is the toughest for a foriegner to learn I think. It is not easily learnable using Hindi as a base, since Tamil has different roots than other Indian languages. Most Hindi speakers find it harder to pick up Tamil than Malayalam or Telugu native speakers.

    Also I wish I had the time to learn German or other European languages. Learning a new language online or through books is very frustrating – need to go to a proper class.

    1. >Most Hindi speakers find it harder to pick up Tamil than Malayalam or Telugu native speakers.

      Of course! Tamil, Malayalam and Telugu are all closely related, and none of them are related to Hindi; they’re not even Indo-European languages, so Hindi is more akin to even English or German than it is to what you speak.

      The whole Dravidian group (Telugu, Tamil, Kannada, Malayalam, and related minor languages) has a reputation for difficulty, and not just to English- and Hindi-speakers.

    1. >* I should have said: significant re-learning of rules is required.

      Fine point of English usage: “are” is a little odd here but not necessarily wrong. If you say “significant re-learning of rules is required” you are using the form appropriate when treating “rules” as a singular thing (that is, you are not speaking of more than one set of rules).

      But because “rules” is a plural being used as a collective noun, “Significant re-learning of rules are required” can have the meaning “Significant re-learning of multiple individual rules in the set is required.” The first construction emphasizes the set as a whole, the second the individual members, especially the particular members that are the topic of the sentence.

      For this particular case a native speaker would be unlikely to choose the second option, but there are other cases where it is deliberately done for that slight shift of emphasis.

      There are even a few cases (though I can’t think of any just now) in which native speakers might disagree on whether a word should be viewed as a plural (use “are”) or a singular collective noun (use “is”). And anyway, this is not one of the rules educated speakers are fussy about.

  34. I have an interesting experience with Chinese tones. I say and hear Chinese words right, but the tone thing bypasses conscious cognition entirely, so I can’t easily e.g. write something in Pinyin that I can say perfectly, for the same reasons that I can’t easily write an English word I know well in IPA. I’ve just never lost my childish ability of easy phoneme mimicry.

    (I also cannot tune instruments or recognize keys, but I ace every pitch recognition test I’ve tried. I think this is more normal and just “lacking absolute pitch”, but I can’t even tune a guitar with one of those things that lights up when you have it right.)

  35. @esr: Unless I’m missing something, you misparsed that sentence. The subject is “re-learning” and is singular regardless of the number of the prepositional object.

    1. >@esr: Unless I’m missing something, you misparsed that sentence. The subject is “re-learning” and is singular regardless of the number of the prepositional object.

      Which is why I used I used “topic”, not “subject”, after thinking about the sentence structure. There’s a little slipperiness in the rules here, as in other cases where topic and grammatical subject do not coincide.

      Adding to the murk, prescriptive grammars do their best to ignore this kind of slipperiness in order to present a neatly packaged set of rules. You have to notice it in the actual production rules of extremely fluent speakers to know it exists.

  36. It’s not directly related to the topic on-hand, but one thing I’ve always thought interesting was that Arabic’s writing system appears far more aesthetically-pleasing than almost any other language. I don’t know a bit of the language, written or spoken, but the writing seems to have a beauty built into it as if it was carefully-constructed to be just as such. I think the art and architecture of the Arabic world reflects the writing system as well.

    The only example I can even think of that might best Arabic at this quality is the conlang writing system of Tengwar. Which should be hardly surprising given it was expressly designed to be beautiful.

    This is totally in constrast to almost all else. Latin and Cryllic alphabets have a cold, pragmatic approach (and the Latin alphabet seems to be abused often for languages that don’t fit well into it. The English language spelling inconsistencies are probably a result of this). They work and pass simplicity tests, but regardless of how you write them or the typeface, there’s little beauty in them. East-Asian languages are sprawing examples in over-complexity … some of it does have beauty, but mushed among all the characters, to an eye without the skill to read the languages, it often looks like a mess.

  37. The German order for modifiers seems absurdly strict to a native English speaker, but it bears comparison with that in standard Chinese!

  38. Arabic writing looks like English cursive handwriting in a mirror: try it sometime!

  39. @esr “If you say “significant re-learning of rules is required” you are using the form appropriate when treating “rules” as a singular thing”

    Er, what? The subject of the phrase is “re-learning”, not “rules”.

  40. “Adding to the murk, prescriptive grammars do their best to ignore this kind of slipperiness in order to present a neatly packaged set of rules. You have to notice it in the actual production rules of extremely fluent speakers to know it exists.” – I think you’re imagining this specific case. As a native speaker I would never say “[verb -ing] of [plural noun] are [adjective]”, nor would I ever not notice it as incorrect if someone else says it.

    There’s a difference between proper descriptivism [which still recognizes that some things can be mistake] vs the “nothing is true; everything is permitted” attitude you are pushing.

  41. English spelling inconsistencies have many causes: regional dialect differences, obviously; class variations; educational attainment differences.

    But much of that arises from the multi-origin nature of the language. Think what happened when each wave of conquerors arrived. Historically: Celts, Romans, Saxons, Normans/Bretons/Flemings/”French”.

    I put French in ironic quotes because they’re hardly Franks! They’re Romanized Gauls.

    William I’s proclamations were translated into English for the benefit of the locals. What stands out is that people in the North (Anglo-Danes) were still speaking like Beowulf, but in the South the language was already recognisably Modern English, more so than Chaucer or even Shakespeare!

    So the language of the capital and most of Wessex was already heavily hybridised before the Normans came. There are signs of this in the continuing use of the Non-Germanic “Mum” and “Dad”. These are Brythonic (Breton, Cornish and Welsh) usages.

    The people continued to be “British”, but with a Saxon overlay due to the necessity of communicating with the newer rulers, just as the languages of Gaul and Spain came to look much more like Latin. A similar effect is observable in the Gallicizing of English following 1066.

    Had the Normans and Angevins migrated in larger numbers, or been sufficiently determined to convert everyone to their language, we might have only traces of Anglo-Saxon remaining in “English”.

  42. There’s a little slipperiness in the rules here, as in other cases where topic and grammatical subject do not coincide. […] You have to notice it in the actual production rules of extremely fluent speakers to know it exists.

    Hmm. The only times I’ve seen a mismatch under these circumstances is where the speaker has clearly lost track of the far-away grammatical subject and attached to the number of the closest noun phrase. On the other hand, I occasionally run into an almost intractable problem when dealing with disjunctions of differing cardinality and frequently end up rewriting the entire clause.

  43. Jay Maynard:
    “Mantas, any language with 11 noun cases to deal with declensions for is …going to be tough for a native English speaker to handle. That’s worse than Latin, which has earned its notoriety in that area well”

    Try 14 in Finnish (formally 15) and 18 in Hungarian. They did get a star in the ranking.

    @Paul Brinkley:

    I’m a Finnish speaker who hasn’t studied Mandarin apart from exchanging “ni hao ma”s with a few Taiwanese people, I have no ear for the tones, and the video you linked is just frightening to me. This Finnish tongue-twister that learners are sometimes made to repeat seems very pedestrian in comparison: Kokoo koko kokko kokoon. – Koko kokkoko kokoon? Koko kokko kokoon. (Explanation behind the link.) That’s somewhat contrived, but perfectly understandable to a Finnish speaker when pronounced correctly (distinct single and double vowels and consonants.)

  44. @Mike Swanson:

    It’s not directly related to the topic on-hand, but one thing I’ve always thought interesting was that Arabic’s writing system appears far more aesthetically-pleasing than almost any other language. I don’t know a bit of the language, written or spoken, but the writing seems to have a beauty built into it as if it was carefully-constructed to be just as such. I think the art and architecture of the Arabic world reflects the writing system as well.

    It is an alphabetic, cursive, phonemic and shorthand script (all in one). You can almost write as fast as you speak. It is also extremely fast to read. Fastest Arabic readers are about 2x the fastest English readers. And English is far better than other European languages in that department.

    This is totally in constrast to almost all else. Latin and Cryllic alphabets have a cold, pragmatic approach (and the Latin alphabet seems to be abused often for languages that don’t fit well into it.

    This cold approach is an accident of history. The latin alphabet is descended from the Phoenician alphabet (semitic language, different language family). Semitic languages are ideally-suited for alphabetic orthography. Vowels play a minor role in semitic languages, and can often flipped to another vowel, or omitted altogether without loss of meaning. Earliest alphabets mapped every consonantal sound to an associated symbol (effortless invention) and voila humans managed to write/read afterwards.

    The “functional” nature of Latin orthography, and block by block representation of every letter (similar to Phoenician alphabet) is what resulted in the printing press being (re-)invented in Europe. It is also the reason why the Islamic world missed out on the renaissance age.

    For English orthography, I like the Quickscript orthography invented by George Bernard Shaw. It is very much in the spirit of the Arabic script and mimics some key features there. However it’s left-to-right nature would make it far less suited for beautiful calligraphy which imo would require a right-to-left script since most people are right-handed. Think forehands (right-to-left) and backhands (left-to-right) in Tennis.

  45. @Jay Maynard:

    “[Klingon] is the most widely spoken conlang, something that pisses off Esperantists no end.”

    If by “most widely spoken” you mean “used in the most movies” then it is. But Klingon is not even close to Esperanto in number of speakers.

  46. Mike Swanson: “East-Asian languages are sprawing examples in over-complexity … some of it does have beauty, but mushed among all the characters, to an eye without the skill to read the languages, it often looks like a mess.”

    This is certainly true of Japanese and Chinese, but Korean’s system is actually a phonetic alphabet that describes its sounds so well that different dialects wind up with different spellings.

    Aaron: A bit of Googling seems that you are correct. It’s a shame, actually; Esperanto was invented for the kind of people who should be opposed at every turn in the name of liberty.

  47. @Jay Maynard –

    > Esperanto was invented for the kind of people who should be opposed
    > at every turn in the name of liberty.

    Huh? Please explain. I know that many of the original proponents of Esperanto were what I suspect you would call “fuzzy-minded one-worlders”, but does the language itself promote that kind of thinking? Or is it just a potentially convenient “universal language”?

    /me veers off in the direction of the Sapir-Whorf hypothesis….

    1. >a native German speaker dissecting the German accent.

      OK, I’ve just learned that I have some truly strange neural crosswiring. I found Rabbit-girl’s exaggerated pronunciation of phonemes English has that German does not extremely sexy. WTF? How does that work? What is being superstimulated here?

      The video contains one error about English phonotactics. Our final s tends to soften to /z/, so it’s not true that English has only one ‘s’ sound, in her terms.

  48. Mike Swanson wrote:
    “…one thing I’ve always thought interesting was that Arabic’s writing system appears far more aesthetically-pleasing than almost any other language.”

    I would put written Thai in that category as well. I can’t read it, but the lovely looping curves are more aesthetically-pleasing than kanji or katakana.

    Where would you put English *cursive* writing on that scale? We tend to think of printed letters in English, but for hundreds of years cursive was the norm. Now that it is no longer being taught and computers are taking over, I expect it will die out within a generation except among a small niche of users.

  49. “Oh, no. That’s not my attitude at all.”

    What you said here – the idea that anything that might come out of a native speaker’s mouth is automatically correct (that native speakers are incapable of making mistakes) – is damn close, no matter what post-hoc justifications you apply to invent a subtle distinction of meaning between the mistake and the correct construction. It’s the assumption that seems to be embedded in statements like “You have to notice it in the actual production rules of extremely fluent speakers to know it exists.”

    1. >It’s the assumption that seems to be embedded in statements like “You have to notice it in the actual production rules of extremely fluent speakers to know it exists.”

      You’re missing the functionalist constraint.

      It is a fact that fluent native speakers sometimes attach syntactic agreement to a noun or noun phrase that is the topic of the sentence even when it’s not the formal grammatical subject. My own theory about this is that it’s a maneuver to shift listener attention to the actual semantic subject of the sentence, away from the phrase-structure subject.

      It is a fact that prescriptive grammars flag this kind of shifting as incorrect, when they notice it at all. I note a tendency for prescriptivists to stay away from analyzing cases complex enough where they might have to deal with this sort of thing.

      Now, let’s examine this in terms of the typology I recommended in my article.

      A prescriptivist would say “wrong”. A pure descriptivist (aka language populist) which is what you are incorrectly binning me as) would say “OK, anything goes, it is fluent-speaker production therefore it is correct.”

      A functionalist, which is what I actually am, asks the following primary question: “Does the rule, or the noted relaxation of the rule, increase or decrease or not alter the ability of the language to convey distinctions?” If the production feature is neutral on this axis, we may ask other questions, like “Does it reduce the complexity of the production rules?”

      In this particular case I think the prescriptive violation is functionally neutral, so I am willing to say “OK, fluent-speaker production, benign language change, no harm no foul.” It’s only time to sound the klaxon and thunder “INCORRECT!” when the change damages or compromises the ability to convey meaning.

  50. Christopher Smith wrote:
    “…the primary issue with English spelling is that the Great Vowel Shift happened at the same time as the spread of the printing press with movable type, so that the spelling of different words crystallized at different times during the upheaval.”

    Fascinating. I had never heard this before, but it sounds plausible.

    I wonder how long it will take for convenient shortened-spellings (e.g., “thru”) to replace the traditional-but-ridiculously-complicated spellings now that we see text-messaging and Twitter replacing long written letters?

    “Though the tough cough through the river slough, the ice sloughs off quickly…”

    If you can read that sentence with correct pronunciation and without pausing to think, you are almost certainly a native speaker!

  51. “Where would you put English *cursive* writing on that scale?”

    Or, on the other end, what do you think of experimental non-cursive Arabic typefaces e.g. the UA Neo family?

  52. @esr I’d argue that despite being spelled as “s” that is not an “‘s’ sound”, precisely because it is identical to “z” and has no minimal pairs against an actual z in the same position.

    1. >@esr I’d argue that despite being spelled as “s” that is not an “‘s’ sound”, precisely because it is identical to “z” and has no minimal pairs against an actual z in the same position.

      In the abstract you have a point. It collides with the way Rabbit-girl was talking, though. I heard her as not recognizing the existence of the /z/ allophone for orthographic ‘s’ in English. So did my wife Cathy, who found the video hilarious because she’s studying German.

  53. > Our final s tends to soften to /z/, so it’s not true that English has only one ‘s’ sound, in her terms.

    That caught my attention, too. Seems to me that English does have all of the three sounds that she talks about (“shit”, “sit”, “zit”).

  54. > It is a fact that fluent native speakers sometimes attach syntactic agreement to a noun or noun phrase that is the topic of the sentence even when it’s not the formal grammatical subject. My own theory about this is that it’s a maneuver to shift listener attention to the actual semantic subject of the sentence, away from the phrase-structure subject.

    I believe that your theory is incorrect, and that where this appears it is normally a result of confusion and having lost track of the subject of an excessively complex sentence. I think this is supported by the fact that when this happens it is not accompanied by any conventional expression of emphasis (punctuation or bold/italics in writing; tone, volume, or pauses to set it off in speech… some of which, admittedly, would also change the structure and therefore the formal grammatical subject of the phrase, and some which would not.).

    1. >I believe that your theory is incorrect, and that where this appears it is normally a result of confusion and having lost track of the subject of an excessively complex sentence.

      You might well be be right in some cases. But is that a plausible interpretation of an utterance as simple as “Significant re-learning of rules are required”?

      hari is an impressively fluent ESL speaker. I think he “fumbled” that one because his brain has noticed that in the production of native speakers syntactic subject and semantic one don’t always coincide, and I chose to respond in accordance with my observations of usage.

  55. “I found Rabbit-girl’s exaggerated pronunciation of phonemes English has that German does not extremely sexy.”

    YKIOK.

    1. >AcronymFinder.com is your Good Friend

      Well, fine, but I was hoping for some kind of plausible theory.

      Cathy thinks it’s because the phoneme pronunciations tend to focus attention on the girl’s lips, which she has made up for sexual exaggeration anyway. I guess that makes sense; maybe I’ll have to go back and watch the video to find out if I can see through my own limbic-system reaction. Maybe after I’ve watched it a few (hundred) times… :-)

  56. “Well, fine, but I was hoping for some kind of plausible theory.”

    I’ve long since quit trying to analyze the source of various sexual attractions. I simply indulge them.

  57. Way too many contextual cues in that video. She’s physically pretty, intelligent, geeky, not-arrogant, kinda funny, and the exaggerated German sounds like a stereotypical Biergarten vamp. I should’ve listened to it without watching at first, but that wouldn’t have addressed at least four clues.

  58. Where would you put English *cursive* writing on that scale? We tend to think of printed letters in English, but for hundreds of years cursive was the norm. Now that it is no longer being taught and computers are taking over, I expect it will die out within a generation except among a small niche of users.

    It honestly wasn’t in my mind when I wrote my thoughts earlier. I’m old enough to have been trained to read and write in cursive, but it’s a skill I’ve not used for a long time. It does look better, especially if you view legal or important documents from the 18th/19th centuries where there is far more emphasis on beauty. I don’t hold a particular fondness for it, and getting too fancy, such as the old documents I mentioned, almost becomes unrecognizable these days and takes quite a trained eye to read them.

    Or, on the other end, what do you think of experimental non-cursive Arabic typefaces e.g. the UA Neo family?

    Neither having knowledge in the language nor constant exposure, I hadn’t heard of them until now and don’t have a pre-existing opinion. My initial impression is that it, indeed, is less pleasing to look at than standard Arabic writing.

    Perhaps it is the cursive systems that make all the difference.

  59. Rabbit-girl has extraordinarily precise diction, which should not be surprising since she is fluent in at least 3 languages with remarkably good accents in each. Skilled precise diction is subconsciously perceived as pleasant and attractive, perhaps because it indicates both intelligence and a strong desire to communicate.

    There’s a YouTuber named XhosaKhaya who supplies video instruction in Xhosa. His speech is VERY pleasant to hear and watch — to both sexes and all orientations.

  60. @Random832:
    > Er, what? The subject of the phrase is “re-learning”, not “rules”.

    Which is what I also thought at the time I made the slip-up. Thanks for the clarification.

    @esr:

    I am not sure how I could justify using “are” in the place of “is” in that sentence. I mentally applied parentheses to the phrase when correcting it as: “Re-learning (of rules) is required.”

  61. @esr:
    >The Dutch case system is simpler. Also, it (like English) never underwent the High German consonant shift; this makes a lot of lexical items more familiar.

    On the subject of English, German, and Dutch, one thing I’ve found is that if you can speak both English and German, comprehension of Dutch (though not production competence, of course) comes almost automatically.

    Completely off topic, something I’ve been wondering about: There seems to be a fair degree of correlation between the hacker mindset and traits (such as maleness, ADD, and the autism spectrum) that suppress multitasking and focus-shifting abilities, as well as “I/O bandwidth”. Is there any evidence that this affects the kind of code that actually gets written? For example, serial, synchronous algorithms appearing more than might be optimal because hackers (or programmers in general, whether inside or outside the hacker culture) have more trouble forming mental models of parallel and asynchronous algorithms? Do pioneers in fields involving lots of parallelism and asynchronicity tend to be “non-standard nerds” more than in other fields? Do wannabe programmers that don’t understand computers well enough to code well tend to do better with parallelism and asynchronicity than would be expected from their general level of skill?

  62. This got me thinking about Sumerian*. The easy way out is to create category VI: Impossible. Impossible because the only descriptions we have of the sounds are from a different long-dead language, so no one really knows what it sounded like.

    Passing over all of the practical problems, I would put it in category IV. Pretty much zero cognates with English, The grammar isn’t horrible, but the rules structure is quite foreign to us. We are familiar with the process of building complex words by sticking prefixes and suffixes onto roots, but the when, where, why and how are all very different. For example, some things that we would do as clauses are done as one or two complicated words. There are two different verb-stem categories with different rules, suggesting that Sumerian is actually the merger of two unrelated languages. There is also a secondary “women’s dialect”, emesal.

    * I dabble a bit, and can grind through the laborious process of reading inscriptions, but that’s about all. I can’t produce anything beyond a short phrase, and even those I don’t imagine to be correct.

    1. >The only descriptions we have of the sounds are from a different long-dead language, so no one really knows what it sounded like.

      There’s a similar problem with respect to Ancient Egyptian. The vowel-sign assignments are guesses based on Old Church Coptic, which is continuously descended from but not identical to Ancient Egyptian (compare the evolution of Old Latin into Italian over a shorter timespan).

      Just to complicate matters, the pronunciation of Egyptian we are all used to hearing in words like “Tutenkhamen” and “Thutmose” is not the best available reconstruction, but a simple and almost arbitrary set of assignments created by early Egyptologists in order to make hieroglyphic inscriptions pronounceable, well before there was good evidence for an actual phonetic reconstruction.

      There’s a third layer of obfuscation. Many words we think of as “Egyption”, notably god-names, are actually rather mutated Greek or Greek-to-Latin renderings of the Egyptian from Classical times. Here’s a handy example I plucked from Wikipedia: The name “Osiris” is a Latin transliteration of a a Greek rendering of the hieroglyphic word “Wsjr”; this is variously reconstructed as Asar, Yasar, Aser, Asaru, Ausar, Ausir, Wesir, Usir, Usire or Ausare.

  63. @John Brase, re: parallel

    In practice, I think we often find the exact opposite. People that can’t handle serial processing tend to totally come apart when asked to do parallel. More often, people that are competent, maybe even “good” programmers will fail at parallel. One of the early warnings, in my opinion, is when you start to see synchronization points strewn about the code, like magic talismans to keep demons at bay. Left untreated, you end up with buggy functions that the author tried to fix by shotgunning mutexes in, sometimes achieving (or surpassing) parity between mutexs lines and functional lines.

    In traditional CS courses, there is an early (1xx level usually) required class that involves pointers. This is usually the dividing line between “real” CS students and the people who need to transfer to the MBA program ASAP. I’m not aware of any such mercy course for parallel processing.

    1. >One of the early warnings, in my opinion, is when you start to see synchronization points strewn about the code, like magic talismans to keep demons at bay.

      Yup. The difficulty of reasoning about the code rises roughly as the square of the number of mutexes. If you make too many mutexes you will be literally unable to model the interactions correctly in your head; that way lies bugs and madness.

  64. @kjj:

    The most convincing theory I have seen on Sumerian is that it is the language of the people of the Arabian plate before they acquired afro-asiatic from south-west Arabia. To this day, Sumerian words still exist in the vocabulary of southern mesopotamian dialects in Iraq.

    The people who descended down on the Arabian plate were caucasians (meaning from Caucasus region) who spoke a language-isolate similar to the many diverse languages that still exist in that region that are not related to any of the bigger language families.

    If this theory about Sumerian origins holds true (and it seems to be backed up by a lot of evidence) then there is no possibility of learning the Sumerian language.

  65. ESR wrote: The name “Osiris” is a Latin transliteration of a a Greek rendering of the hieroglyphic word “Wsjr”; this is variously reconstructed as Asar, Yasar, Aser, Asaru, Ausar, Ausir, Wesir, Usir, Usire or Ausare.

    …Aesir?

  66. I’m sure you guys were all turned on because she is pretty and has great boobs.

    However, what I experienced listing to her is a version of the uncanny valley. As a non native speaker gets a better and better accent she sounds better and better, but when she is very close I find it a little bothersome. I guess it is because her excellence sets my expectation of perfect fluency, and she is mostly perfect, but makes these tiny little errors that are dissonant.

    I see it a lot with the many Indian programmers I work with. Although their English is great and perfectly functional, the thing that really sets them apart (aside from the incorrect use of the definite article) is the syllabic emphasis and pauses in unusual places:

    “We are going TO (pause) write a feed program TO (pause) solve this problem.”

    Of course since I can barely scrape by in French and a little Spanish I am not a one to criticize, However, I wonder if anyone else experiences this linguistic equivalent of the uncanny valley.

    1. >I’m sure you guys were all turned on because she is pretty and has great boobs.

      If it were that simple I wouldn’t have remarked on it. Yes, I noticed the pretty straight off, as what heterosexual male would not? But after the first second or so that went to background until she did a specific thing, which was exaggerate pronunciations of English phonemes that German doesn’t, at which point “sexy” moved back to the foreground in a big way. I still find this odd.

  67. “As a non native speaker gets a better and better accent she sounds better and better, but when she is very close I find it a little bothersome.”

    One of the Hercules developers is a Frenchman whose father is an American serviceman, and who spent several years growing up in South Carolina. He speaks English like a Southern good old boy, but when he runs into a word he has no English pronunciation for, he reverts to French. I find it very jarring.

  68. >>* I should have said: significant re-learning of rules is required.

    >* Fine point of English usage: “are” is a little odd here but not necessarily wrong. If you say “significant re-learning of rules is required” you are using the form appropriate when treating “rules” as a singular thing (that is, you are not speaking of more than one set of rules).

    IS in this sentence is referencing/referenced by re-learning, not rules or rule, so ARE would never be correct in the sentence IMO.

  69. > You might well be be right in some cases. But is that a plausible interpretation of an utterance as simple as “Significant re-learning of rules are required”?

    Sometimes a mistake is just a mistake. In a written, informal environment such as this, I’ve made mistakes (not necessarily of the same particular class, so don’t go taking this as proof of your theory) on similarly apparently simple phrases due to e.g. the composition process being chopped up by starting with one phrasing and then going back and re-editing it into a different one.

    For example, I could imagine myself typing something similar by typing “[something since-deleted] rules are [can’t think of non-awkward phrasing for ‘things that require re-learning’, go back and restructure, forget to change ‘are’]”

    —-

    In general, your view seems like an extreme version of descriptivism, where correctness is defined purely by prevalence among native/fluent/”highly fluent” speakers, and which does not admit the possible existence of common mistakes as a concept.

    My own view is a somewhat more moderate version of descriptivism, where things that other native/fluent speakers would generally notice something wrong with*, or that someone who produced one would regret it if they noticed it (on re-reading, for instance), are excluded from this “usage === correctness” principle.

    (One case that I’m not sure about is “highly contentious” usage, such as “irregardless” or “literally”… it doesn’t help that for “literally”, objectors have a substantive point in that the new usage erases a useful distinction of meaning, rather than just not liking how the word sounds. But that’s a whole ‘nother argument entirely.)

    *regardless of whether they can identify a formal rule it breaks, or if one even exists – therefore something can be descriptively incorrect even if it is prescriptively correct. One example of a rule I’ve never seen a 100% clear formal description of is the possessive for singular nouns that end in “s”. They tend to descend into murky concepts like whether a proper name “sounds classical”. Another, though I’m not sure of it since I’m not a native speaker of such a dialect, is the usage of plural or singular agreement (in British dialects, where plural agreement is acceptable at all) for “group” nouns like corporations, sports teams, etc.

    1. >In general, your view seems like an extreme version of descriptivism

      But as I keep telling you, I’m not a descriptivist at all, I’m a functionalist. You’ve actually reminded me of a perfect pair of examples to illustrate the difference. Consider two common mistakes related to the apostrophe of possession:

      1. Not getting the “classical names” exception right.

      2. Using the apostrophe of possession in plurals.

      An extreme descriptivist has no warrant to treat these differently – the only question he or she can ask is whether either has achieved majority use, at which point game over.

      On the other hand, a functionalist like myself can draw a principled distinction. I can analyze as follows.

      The “classical names” exception has no informative value; discarding it does not significantly change the meaning of speech acts. We also observe that it is the kind of rare irregularity that tends to get smoothed out over time because people forget it or fail to transmit it to children, like an uncommon irregular verb – this is a natural process of Zipf’s-Law pruning. Because the exception has no informational value, we can say “Natural language change, no harm no foul,”, and let it go. (I would not bother trying to teach it to an ESL speaker).

      The adventitious apostrophe in plurals, on the other hand, is a damaging blunder because it erodes the distinction between possessive and plural formation. Information-carrying capacity is actually jeopardized here, so I smite it with great wrath wherever I encounter it and will continue to do so even if it increases in usage.

  70. One instance of anomalous plural agreement that is common in American English is e.g. “a bunch of people [are doing something]”, where the formal grammatical subject is “bunch”. Rather than being related to some esoteric concept of ‘people’ being the topic, my analysis of this usage is that word sequences such as “a bunch of” have morphed into adjective phrases and that “people” is the real grammatical subject of the sentence.

  71. > 2. Using the apostrophe of possession in plurals.

    At the risk of getting even further afield, how do you deal with the most common case of this, its use in cases where it is undeniably orthographically awkward to append an “s” without one? Such as pluralizing letters, digits, punctuation marks, abbreviations, etc.

    I’m also not fully convinced of the informative value of an orthographic distinction between a possessive and a plural* in general. There’s obviously no analogue in speech, so grammatical context [a plural is a noun which usually has a plural verb to agree with; a possessive is an adjective] is clearly sufficient. It merely looks ugly, in the same way as appending ‘s to a classical name, or the bare apostrophe to a non-classical one, does to some people.

    * or, for that matter, a singular present-tense verb… for similarly orthographically awkward verb phrases I sometimes see an apostrophe on ‘ed (or ‘d) and ‘ing as well.

    1. >how do you deal with the most common case of this, its use in cases where it is undeniably orthographically awkward to append an “s” without one?

      Very simply: I never use an apostrophe in plurals, ever. Mistake, wrong, crash landing.

  72. I’m mildly surprised to see Danish and Swedish on the Category I list; I’d assumed that the North Germanic languages were even more removed from English than German or the Romance languages. Though reflecting on this, I realize that assumption is invalid. Having seen some written Swedish and listened to a native Swedish speaker, I’d had an easier time picking up on the cognates than I had with German.

    >a native German speaker dissecting the German accent.

    Interesting, and amusing. Her description of the stops between words, comparing German, English and French, with waveforms as illustration is especially amusing. There’s “quite some breaks” in the German phrase, English being “middle between the two”, and French being “one long blublublubleh”. That accurately captures my frustration with trying to speak or understand spoken French. “Whoah. Hold on, slow down and separate your words so I can understand them.” I notice something similar when listening to native Spanish speakers, though it’s not quite as pronounced as with French, IMO.

    1. >Though reflecting on this, I realize that assumption is invalid. Having seen some written Swedish and listened to a native Swedish speaker, I’d had an easier time picking up on the cognates than I had with German.

      It’s not just you; I’ve noticed the same thing, and mentioned it on this blog before.

      I don’t know of any detailed explanation; indeed, Swedish is at slightly greater phyletic distance from English than High German and one would expect less mutual intelligibiliy, not more.

      I think what is going on is that (a) the North Germanic group has been evolving in parallel to or slightly convergent with English as the Old Norse case system breaks down into prepositional/positional grammar, but (b) High German has actually diverged from English, wandering off in a direction less like our language than surviving Low German dialects.

    2. >There’s “quite some breaks” in the German phrase

      Germans speaking English love the phrase “quite some” and use it where a native speaker would say “quite a few of” or “quite a lot of”. Can any Germanophone here tell me what German construction they’re importing?

  73. @Jeremy Well, it makes sense to me considering that Standard German, or “High German”, evolved from a set of dialects whose geographical home is the southern half of Germany.

  74. > Very simply: I never use an apostrophe in plurals, ever. Mistake, wrong, crash landing.

    Yeah but what I was asking was 1. how do you pluralize those things 2. how do you defend your position when that is what the person you are criticizing was doing? It’s hard to argue that you’re in favor of functional expressiveness in the language when you’re saying that certain words/phrases simply aren’t allowed to be pluralized.

    1. >1. how do you pluralize those things

      I glue on an s without an apostrophe, and accept that this sometimes looks awkward.

      I don’t quite understand your second objection.

      >sufficient information is available from context

      I don’t think so. Compare “1970’s idea of a good time” with “1970s idea of a good time”. One refers to a single year, the other to a span of 10 years. If the apostrophe for plural formation became the rule we would no longer be able to make this distinction – so kill it with fire!

  75. Also do you have any answer for my argument that it’s not a necessary feature because sufficient information is available from context and there is no (and has never been any) analogue in speech?

  76. > cursive

    I was required to write it in elementary school, but that was an output-only task; I never had to read it then, and have never needed to since. My sister-in-law insists on writing us letters in cursive; they might as well be random scribbles. I pass them to my wife, who’s a few years older than I am and can actually read cursive.

    By the time I got to junior high assignments in real printed letters were acceptable, and I’ve never written a word in cursive since.

    “If cursive is so wonderful, why are all our schoolbooks in block letters?”

    Cursive existed due to quill and fountain pens, which tended to drool and blot if you stopped a stroke. So you joined all the letters up into a continuous line and (theoretically) only got a blot at the end of each line. Cursive was dead with the invention of the pencil, but it took a long time before it finally keeled over.

  77. Native German speaker here. Naively translated both “quite some of” and “quite a few of” sound quite awkward in German (ziemlich manche der / ziemlich ein paar der), but “quite many of” maps cleanly to “ziemlich viele der”. So it might be that Germans prefer “quite some of” because it’s closer to “quite many of”.

    Was the omission of the ‘of’ after the ‘some’ deliberate? It doesn’t seem to me like the kind of error Germans would usually make.

    1. >Was the omission of the ‘of’ after the ‘some’ deliberate? It doesn’t seem to me like the kind of error Germans would usually make.

      I often hear or see it without the “of”. I first noticed this when I was active on Battle For Wesnoth, were a lot of the developers are German ESL speakers.

  78. @esr the latter cannot be expressed in speech or in a grammatically correct way without an article [a “1970s” idea of a good time; the 1970s’ idea of a good time], so I’m not sure why you want to express it in writing without one. If I saw this sentence in the wild I would probably assume it was a simple mistake and was intended to refer to a single year.

  79. > Was the omission of the ‘of’ after the ‘some’ deliberate? It doesn’t seem to me like the kind of error Germans would usually make.

    It’s not actually incorrect without “of”. As far as I can tell its presence depends on whether the object is a simple noun (“a few computers”) or identifies a set which the phrase is intended to identify a subset taken from (“a few of my computers”).

  80. “quite some” is unusual with or without of, but “some” by itself can likewise be used either way.

  81. A bit of sidenote: I have read an article in Scientific American about trying to prove or disprove Sapir-Whorf hypothesis, and in it there was example of language (unfortunately I don’t remember the name) where you didn’t say “on the left, on the right, ahead, behind” but “north of, south of, east of, west of”… You had to keep your direction to use this language…

    1. >(unfortunately I don’t remember the name)

      I remember reading about this. Central American Indian, IIRC. I can’t remember the name either.

  82. Re Jessica:

    > I see it a lot with the many Indian programmers I work with. Although their English is great and perfectly functional, the thing that really sets them apart (aside from the incorrect use of the definite article) is the syllabic emphasis and pauses in unusual places:

    I’m not clear what you mean by “incorrect use of the definite article”.

    I think what you’re hearing there is 100% fluent native speakers of an English dialect different from yours, not “incorrect” usage by speakers of English as a foreign language. Indian English is an odd case though; as a first approximation it isn’t anybody’s mother tongue, but I suspect the majority of people who make it into and through college and into programming grow up in bi- (at least) lingual households.

    As a native speaker of English English, I do notice some pretty significant differences in Indian English, but not necessarily more than I do in American English.

  83. > Yup. The difficulty of reasoning about the code rises roughly as the square of the number of mutexes. If you make too many mutexes you will be literally unable to model the interactions correctly in your head; that way lies bugs and madness.

    Or as I refer to it: my day job.

    On another note, I grew up in Canada and participated in French Immersion for ~12 years. I won’t claim it was well done or that I put a lot of effort in. Still, that’s considerably more than the number of hours listed above. And yet I’d be surprised if at my peak I had better mastery than a native 8-year old.
    Memorizing the gender associated with words drove me nuts.
    Conjugating ‘regular’ verbs was easy enough.
    But there were large numbers of exceptions. Horrible numbers.
    We had one essay in 10th grade we had to write where we were allowed up to 2 grammar or spelling mistakes per page, and then half a point off per error after that. I think I ended up with a score of 2/50. The lowest score was -8/50. (That’s right: negative 8).
    And this was still with the benefits of computerized spell-checkers.

  84. > However, I wonder if anyone else experiences this linguistic equivalent of the uncanny valley.

    I find this with native Swedish speakers speaking English. I’ve met some people who are convinced that their English is flawless, and it almost is, but when they make a little mistake, I can recognize the bit of Swedish where it came from and want to scream “that’s Swedish!” at them. Part of my problem with this type of a person is that they’re also convinced that Swedes always speak English better than Finns do, which is mostly true, for obvious reasons, but makes me want scream louder.

  85. Old Church Coptic, which is continuously descended from but not identical to Ancient Egyptian (compare the evolution of Old Latin into Italian over a shorter timespan)

    Yabbut Italian is a living speech. Coptic, AFAIK, has been a purely liturgical and scriptural language for the last several hundred years. That would tend to preserve it.

  86. BTW, last night I tried to post a comment and failed. It had a link in it, but posting failed even with the link obfuscated.

  87. My experience is with Spanish, Korean, and Russian.

    Couple things:

    – One of the more interesting observations I made in the process of learning Korean was the fact that when speaking, any time I needed a word I didn’t know my brain’s natural reflex was to reach for the equivalent in Spanish, *not* in English. I believe I’ve read somewhere that acquired languages are stored in the same region of the brain, and this introspective evidence bolsters that claim.

    Further, I found that when I came back to the States and started casually reading the news in Spanish again, my ability had only atrophied slightly, despite the fact that it had been years since I’d studied Spanish at all. I’m a fairly talented language learner and I have a good memory, but maybe, if Spanish and Korean are both stored in overlapping circuits in my brain, practicing Korean led to a certain amount of phantom neural activation in the place where Spanish was lurking.

    – I didn’t like the Korean script very much when I first encountered, but after a while I came to realize that, ease of learning notwithstanding, it’s actually a pretty good metaphor for Korea and Her culture.

    Korea, Korean food, and the Korean language all have a sort of craggy, sharp, almost desolate beauty. I mean, it’s not Antartica, but you aren’t going to find much in the way of lush forests or breathtaking aquascapes, and the food is sampled from an alien palette and often served still-boiling.

    Still though, once you spend a weak or two listening to and reading spoken Korean, and watching the sun rose over the mountains that surround your town, you start to see the charm.

    – Korean grammar is pretty slippery to a native English speaker, there being no reference point for the “particles” sprinkled throughout Korean. But it’s not *that* fearsome.

    – I think Russian hits a sweet spot for me. It’s sexy and it’s the primary vehicle of communication for one of the most towering and resilient cultures of all time, but it’s not so drastically different as, say, Cherokee.

    -Speaking of, why are we ignoring indigenous American languages?

  88. > I believe I’ve read somewhere that acquired languages are stored in the same region of the brain, and this introspective evidence bolsters that claim.

    +1

    Before I moved to Germany I could speak French reasonably well. Now I speak fluent German but can barely get a sentence of French out to save my life, although I can still read it pretty easily.

  89. Meta level: difficult grammar matters only if you want to pass official exams or present yourself in a high-status way. Being very fluent with broken grammar is possible and often workable in many situations. Which raises the question essentially what do you want, how you measure various kinds of proficiency. Would you rather speak slowly but correctly or quickly but incorrectly? Most people tend to be impatient, they prefer quick but incorrect. I think you, too, when a Japanese tourist ask directions from you, you would prefer him quickly, if incorrectly, spitting out what he wants, and not wrestle too long with trying to pronounce “the” three times until he gets that elusive and correct sound that is 70% de and 30% se.

    Object level: German harder for English-speakers than Romanian? Hm. The only good reason I can imagine for that is the German habit to translate Latin words. Like, “television” is good enough for almost any languages, but they just had to literally translate that to Fernseher. Farseeer. And computers to reckoners i.e. Rechner.

    This is also why the Anglish Moot project tends to rely on German. It tries to present a form of English, how it would look like the Franco-Latin influence. So “translation” in Anglish becomes “oversetting”, quite likely influenced from Übersetzung. Give it a look one day, it is fun. The better your Anglish – or Old / Middle English – is, the easier German looks like.

  90. I can converse a bit in Spanish. I deal a great deal with my opposite number… a Spanish speaker who can converse a bit in English. We end up talking in Spanglish, and I tend to forget I’m doing so.

    I don’t have much to add, other than when I do so I just blot out the fact that I’m doing so. I find it somewhat natural. We do sometimes have to repeat ourselves sometimes a bit until we find the entomology that we both understand.

    Word order, plural form and other crap is unimportant when you’re on a job-site trying to understand where the conduit runs, and what’s in it. Though, to be honest, what is in a conduit is pretty important, and communication errors lead to people jumping around shouting ‘Shit, that wire is hot!’. Unless it’s 440v, and then you just have police to deal with. And a funeral. (I very nearly severed a 440v line with a cheap pair of wirecutters once.)

    On an unrelated note: Space-X successfully landed a Stage-1 Falcon 6 rocket about 9 miles (um…15 km) from the original launch site. I watched the live stream and got chills when the stage -one came back down and… just landed. And sat there. And didn’t topple over.

    Where were YOU when the first recoverable stage-1 booster (that was part of an orbital flight) was successfully recovered?

  91. @kjj:
    >In practice, I think we often find the exact opposite. People that can’t handle serial processing tend to totally come apart when asked to do parallel.

    Whether people with low programing competency do better with parallel wasn’t my main question, just speculation on the kind of evidence that might accompany a “yes” answer to my main question.

    The main question was, somewhat rephrased, “Is there any evidence that the traits that seem to most typically accompany good programing ability interfere with the ability to work with parallel or asynchronous code?”

    So the next sub-question to ask would be “Does the set of people who can work well with parallel and asynchronous code tend not to exhibit the traits that are considered to correlate with general coding ability, while still having good general coding ability? For example, are members of this set more likely to be female and less likely to exhibit signs of ADD or autism spectrum disorders than members of the broader set of people with good general coding ability?”

  92. @esr:
    >I don’t think so. Compare “1970’s idea of a good time” with “1970s idea of a good time”. One refers to a single year, the other to a span of 10 years. If the apostrophe for plural formation became the rule we would no longer be able to make this distinction – so kill it with fire!

    The phrase can only have the second meaning if you prepend an article, e.g, “A 1970?’s idea of a good time”, which would mean “an idea of a good time that would have fit in the 1970?’s”, or “The 1970?’s idea of a good time”, meaning “The idea of a good time that was current in the 1970?’s”.

    “1970?’s idea of a good time”, without an article, can only mean “The idea of a good time that was current in the year 1970”, so there’s enough context to resolve the ambiguity whether you omit or include an apostrophe in any of the three phrases.

    @Emanuel Rylke:
    >Was the omission of the ‘of’ after the ‘some’ deliberate? It doesn’t seem to me like the kind of error Germans would usually make.

    At least in my dialect of English “quite some” is valid and “*quite some of” is a grammatical error. However, “quite some ___” cannot be used to mean “quite a few (of) ___”, but rather means “An impressive ___”. For example “Quite some dog” means “An impressive dog”.

    BTW, you translate “quite” as “ziemlich”, but my impression was more that “ziemlich” means something more like “somewhat” and “quite” is closer to “echt”.

    For example “The car is somewhat good” would be something like “Das Auto ist ziemlich gut / Das Auto ist gut, aber nicht wunderbar”, while “the car is quite good” would be “Das Auto ist echt gut”.

    1. >so there’s enough context to resolve the ambiguity whether you omit or include an apostrophe in any of the three phrases.

      Only if the writer is careful about the article placement. I can easily imagine contexts where this wouldn’t be the case, especially in advertising followed by a ‘!’. It makes me twitchy.

  93. @Jon Brase
    Interesting, it feels to me like “quite some of” is an acceptable (but not as slick sounding) substitute for “quite a few of”, so I assumed that ESR meant it was odd but not wrong.

    Isn’t ‘quite’ rather more ambiguous than ‘echt’? But yes, in many cases it’s the better translation. Also I’m prone to understatements which brings my usage of ‘ziemlich’ closer to my usage of ‘quite’ than it should be.

  94. Regarding “quite some”:

    When we German-speakers use the phrase “eine Menge” (ie Er hat eine Menge Geld), then it conveys that “Er” (he) not only has some money, but quite a lot of it. So i guess we subconsciously pack that into the “quite some” construct.

  95. @AlanL
    > I’m not clear what you mean by “incorrect use of the definite article”.

    I mean it is common for them to omit it when a native speaker would use it. So it would be more accurate to say the incorrect non use of the definite article. In fairness, the English definite article (and the article in all languages as far as I can see) is often tricky and subtle. It has a lot of meanings and bristles with non obvious idiomatic quirks, so it is, to me, at the edge of fluency.

    However, someone mentioned Sapir Whorf above. It made me think of something I think is really cool. There is a language spoken by the aboriginals in the Peru region called Quechuan. In this language the word qhipa means both ‘behind’ and ‘future’, where as ñawpa means ‘ahead’ and ‘past’. This is of course entirely the opposite of the languages I am familiar with when we consider our past behind us and our future ahead of us. The speakers of this language think of themselves walking backward into the future, because, of course, you can see the past just as you can see what is behind you if you walk backwards, and you cannot see the future, just as you cannot see ahead in you walk backward.

    Which I think is an interesting perspective that is embedded in the linguistic character of the language itself (whether is is causal of, or consequential from, or coincidental with the cultural meme, I don’t know.)

    (All info from Wikipedia.)

  96. @esr:
    >Only if the writer is careful about the article placement. I can easily imagine contexts where this wouldn’t be the case, especially in advertising followed by a ‘!’. It makes me twitchy.

    For me, at least, changes of article placement in this phrase make enough difference that there’s significant potential for, if not a certainty of, misinterpretation whatever you do with the apostrophe. I can’t think of any context where I’d interpret “1970?’s idea of a good time” in such a way that “1970?’s” meant the decade and not the year. If you’re referring to the decade and don’t use an article, you’re committing a grammatical error that’s going to cause a misparse.

  97. The list seems to overestimate the difficulty of learning a non-indoeuropean language, while it ignores the fact that some languages are difficult simply because they are highly irregular. For example, Hungarian with its highly regular grammar is starred, while Slavic languages, with irregular phonetic changes inside roots and hundreds of suffices to be learnt by heart are not. It doesn’t feel right.

    1. >The list seems to overestimate the difficulty of learning a non-indoeuropean language, while it ignores the fact that some languages are difficult simply because they are highly irregular.

      The alternative hypothesis is that you are underestimating the helpfulness of cognate forms.

      Given that the list is based on measured time to fluency over a large sample, I think FSI is more likely to be right about this than you. Sorry…

  98. > I mean it is common for them to omit it when a native speaker would use it.

    That’s what I thought you probably meant. My point being that they *are* native speakers of their dialect, in which the definite article is no longer as compulsory as it is in either of ours. There are already pretty big grammatical drifts between English dialects and that’s one of them. (The far more frequent use of the simple past tense in American in situations where a British speaker would naturally use other past tenses is another)

    1. >My point being that they *are* native speakers of their dialect, in which the definite article is no longer as compulsory as it is in either of ours.

      Which accidentally reinforces my point about resisting tooth and nail degenerative changes like pluralization with aposttrophe-s. Someone else pointed out that you can get some disambiguation from whether or not the plural-or-possessive is preceded by an article, but you never know when some other kind of language drift will cut your disambiguation out from under you.

      So, if you’re a functionalist, always always always resist the normalization of usage mistakes that lose information. They are evil, the enemies of communication and clarity of thought.

  99. I don’t necessarily agree that loss of the definite article represents loss of information, Russian and many other languages get along just fine without it.

    Nor do I see as my business to “resist changes” in Indian English, a dialect of which I am an interested observer but not a speaker.

  100. > Dutch
    From knowing a few speakers of Dutch, I’ve learned that even casual or high-school-taught speakers of English have near-impeccable English grammar and could pass for natives with a little accent coaching.

    Also am I alone in thinking Dutch sounds and reads like an English person doing a bad impression of a German and vice-versa? (As somebody who was once semi-fluent in German, I’ve noticed that I rarely need Google Translate when reading my Dutch-speaking friends’ posts online…)

  101. > However, someone mentioned Sapir Whorf above. It made me think of something I think is really cool. There is a language spoken by the aboriginals in the Peru region called Quechuan. In this language the word qhipa means both ‘behind’ and ‘future’, where as ñawpa means ‘ahead’ and ‘past’. This is of course entirely the opposite of the languages I am familiar with when we consider our past behind us and our future ahead of us. The speakers of this language think of themselves walking backward into the future, because, of course, you can see the past just as you can see what is behind you if you walk backwards, and you cannot see the future, just as you cannot see ahead in you walk backward.

    This was actually common in most languages until fairly recently. For example Latin had this property. Heck English used to have this property the two meanings of the word “before” are a vestige of this.

  102. >“Though the tough cough through the river slough, the ice sloughs off quickly…”
    >If you can read that sentence with correct pronunciation and without pausing to think, you are almost certainly a native speaker!

    Hah! No. Most native speakers would have trouble with that one. “-ough” is possibly the worst bit of otrthography in the English language. A minimum of six different pronunciations. We’d get through “through” easily enough, but slough is not a common word. Using it in two different senses with two of its three pronounciations? Nope. That’s effectively a tongue twister for your brain, even to natives. As much as I generally find our arcane spelling variations appealing (etymology geek) “ough” is a worthless construction that needs to go. Thru cannot replace through fast enough for me.

    https://en.wikipedia.org/wiki/Ough_(orthography)

  103. >If it were that simple I wouldn’t have remarked on it. Yes, I noticed the pretty straight off, as what heterosexual male would not? But after the first second or so that went to background until she did a specific thing, which was exaggerate pronunciations of English phonemes that German doesn’t, at which point “sexy” moved back to the foreground in a big way. I still find this odd.

    Hey, I was told there would be boobs. You have to watch some of her *other* videos to get any look at the boobs. So don’t be misleading. :)

    I suspect it’s all about the ‘intelligence = sexy’ trigger. Aside from her being very cute (not conventionally pretty I think, but very cute), and the boobs, and the fact that she flirts outrageously with the camera and makes jokes that include sexual references… (“let’s talk about the D” while making an exaggerated leer at the camera, oh my) ALL that aside, it’s the intelligence thing.

  104. There are dialects in north England that omit the definite article.

    Irregularity occurs in Chinese dialects, even in the names of numbers. Standard Chinese is a regularisation, just as standard Bahasa Malaya/Indonesia is.

  105. @esr:


    Given that the list is based on measured time to fluency over a large sample, I think FSI is more likely to be right about this than you. Sorry…

    We don’t know how the list was compiled. There is plenty of material that seems to contradict the list. I found this video, by a native English speaker and professor of Arabic, that contradicts ranking of Arabic as most difficult.

    I would look for a list compiled by bureaucrats of the british empire 150yrs ago and take it over this one any time of the day.

    1. >I would look for a list compiled by bureaucrats of the british empire 150yrs ago and take it over this one any time of the day.

      Why? Their breadth of experience was less than the FSI’s. They, for example, never has to deal with Lithuanian.

      One reason I believe the FSI’s ranking of Arabic as most difficult is that it’s supported by the line of evidence laid out in Empires of the Word, where Ostler ranked languages by difficulty of acquisition by adults. He pointed out that even Arabic’s difficulty of acquisition is such that even immense religious prestige has been insufficient you cause it to replace other languages except where incoming invaders have essentially wiped out the native population.

    2. >I found this video, by a native English speaker and professor of Arabic, that contradicts ranking of Arabic as most difficult.

      That video is full of shit. Not only does it make false claims about the simplicity of the language, but the author must have known those claims to be false when he made them.

      I’ll just hit the most obviously bogus one…

      Just three phonemes in Arabic that don’t appear in English? Yeah, right – when Modern Standard Arabic has a phonemic contrast between velarized and unvelarized consonants and not one but two different velar/uvular fricatives. Even the uvular stop is extremely difficult for an English-speaker, matter of fact there’s no way to pre-train for that one in the entire northern branch of Indo-European. (Probably not in the Indic languages either, but something might have snuck in from Dravidian that I don’t know about.)

      I count no fewer than eight phonemes in MSA that don’t occur in English, plus at least one other (the voiceless glottal stop) that only occurs allophonically (with medial /d/ and /t/). Four of those (the velo-uvular and pharyngeal fricatives) use articulation points that English never touches.

  106. > On the blog where I found this list, a commenter opined that there ought to be a category zero for Esperanto. Yep.

    Where would you place Lojban?

  107. >Hah! No. Most native speakers would have trouble with that one. “-ough” is possibly the worst bit of otrthography in the English language. A minimum of six different pronunciations. We’d get through “through” easily enough, but slough is not a common word. Using it in two different senses with two of its three pronounciations? Nope. That’s effectively a tongue twister for your brain, even to natives.

    Is true. Also, the variant spellings ‘slew’ or ‘slue’ are common for the few occasions that ‘deep mud, backwater, creek in a mudflat’ usage occurs.

    ‘Slough’ in the sense of ‘cast off’ is never (yes, not ever never) heard or used except by bridge players, as it’s part of the jargon.

  108. I might add, I wasn’t aware of a third pronunciation of ‘slough’ until a few minutes ago.

    ‘Slough’ as a place name in England does nothing for this native American speaker, and while I’m aware of the reference ‘Slough of Despond’, the few times I’ve ever heard a spoken reference to it, it was pronounced like ‘slew’. Every time. I am actually surprised to learn that isn’t correct.

    1. >‘Slough of Despond’, the few times I’ve ever heard a spoken reference to it, it was pronounced like ‘slew’. Every time. I am actually surprised to learn that isn’t correct.

      Apparently both /sloo/ and /slau/ occur as regional variations in the U.S., but /slau/ (the only pronunciation I knew before looking this up up) is normal in Great Britain. My own idiolect is a bit more British-influenced than that of most Americans because I lived there at a formative age; to this day I am sometimes unaware of American-variant pronunciations of rare words.

      In the U.S., I’m used to seeing “sluff” rather than “slough” for the bridge term.

  109. @Greg
    > ‘Slough’ in the sense of ‘cast off’ is never (yes, not ever never) heard or used except by bridge players, as it’s part of the jargon.

    That isn’t true at all. It is a common verb used to describe the process of a snake shedding its skin.

  110. @Geoffrey Tobin
    > There are dialects in north England that omit the definite article.

    Can you offer examples from written works? There are a few places where the def art is variable in its usage between BrEng and AmEng, but from what I know that northern English dialect more combines the article in pronunciation such as “I’ll see you int the pub”. where “int” is a combination “of” in and “the.”

  111. Pardon me, obviously the example of northern English i gave was meant to be “I’ll see you int pub” and int is “in + the”.

  112. >That isn’t true at all. It is a common verb used to describe the process of a snake shedding its skin.

    Maybe where you are…. Where I am people talk about snakes shedding their skin, on the precious few occasions (as in, fairly indistinguishable from none) discussion turns to snakes and their skin. Some other critters ‘molt’ – that’s a very uncommon usage as well.

  113. @Greg
    > Maybe where you are….

    My objection was not that it wasn’t common, but that it was never ever used except in the context of bridge. That is a pretty emphatic statement you made. Curiously so since I have never heard it in the context of bridge, and have frequently heard it in the context of our serpentine friends. Of course I hang out a lot more places where there are snakes than I do bridge players, never having played.

    Having said that, I hate snakes, I am ambivalent on bridge.

    In regards to Northern English accents, I think even if the preposition ends in t they still merge the two, but it is pretty subtle vocally “I were stood against’t wall”. The second t is very brief indeed. But YMMV. That is just what I hear in my encounters with Geordies and Yorkshire folk.

    1. >In regards to Northern English accents, I think even if the preposition ends in t they still merge the two, but it is pretty subtle vocally “I were stood against’t wall”. The second t is very brief indeed. But YMMV. That is just what I hear in my encounters with Geordies and Yorkshire folk.

      Well spotted. My Frodo ear confirms.

      A lovely accent, Geordie is. I could listen to it all day just for the music of it.

  114. “Moult” (so spelt) is very common usage.
    Seems we speaking from one global minimum across one or more potential barriers to another.

  115. @esr:


    >> Why? Their breadth of experience was less than the FSI’s. They, for example, never has to deal with Lithuanian.

    British empire agents had to become proficient in local/world languages to further the goals of their empire among other things. Far more than any American diplomat is required to be proficient in a foreign language these days. They were far more experienced on what it took to learn language X back in their day than we are today.


    >> Just three phonemes in Arabic that don’t appear in English?

    That is absolutely correct. Read on.


    >> when Modern Standard Arabic has a phonemic contrast between velarized and unvelarized consonants

    Sure. but all those consonants can be approximated by their English counterparts. One characteristic of Arabic is that velarized/unvelarized are always separated from one another by several phonemes in speech. So 99.99% of the time if you flip velarized to unvelarized you’ll be perfectly understood.

    In “semitic” people think in terms of roots, templates, and derivation, not in terms of mapping words to meanings. Throw in the approximate consonant sounds according to the right template and you’ll be perfectly understood. The exact sound of the consonant is not as important as to how it is mathematically employed in the template used to construct the word. The template packs far more information density as to the possible meaning of any given word than the exact pronunciation of the word. In other languages, unless you pronounce the word (consonants) perfectly you won’t be understood and your word cannot be mapped properly to its meaning in the mind of the audience that hears it. Not so in semitic languages!


    >> Even the uvular stop is extremely difficult for an English-speaker

    Can be approximated to [k] or [g]. If you approximate to [k] or [g], you will be perfectly understood, and that is in fact what happens in the majority of Arabic dialects.


    >> I count no fewer than eight phonemes in MSA that don’t occur in English

    If you learn how to imitate the /?/ and /?/ sounds, along with the french “R” you’re done. The rest is easy. And once you’ve figured out the method behind being able to poem a million verses in perfectly rhyming prose, and the logic of packing so much meaning in so few phonemes, you’ll wonder why humanity has wasted so much of its potential trying to communicate its ideas with that ‘shi shi shi’ insanity. Imagine the cathedral and the bazaar, written in iliad form, infused with layer upon layer of mysticism and wisdom, and where every verse on cathedral software-development is as memorable as the dialog of a Quentin Tarantino movie. And needless to mention of course garnering followers with the zeal of Al-Qaeda and ISIS. That is what things would have looked like, if Arabic was the international language, and if the cathedral and the bazaar was written in Arabic by someone with your level of mastery of English.

  116. uma: Instead, all that beauty has given rise to the most virulent, hideous ideology in the world today. Forgive me if I don’t see the beauty in a book that commands its followers to kill the infidel.

  117. @Jay Maynard:


    >> uma: Instead, all that beauty has given rise to the most virulent, hideous ideology in the world today. Forgive me if I don’t see the beauty in a book that commands its followers to kill the infidel.

    Have I mentioned the dreaded book, or its followers, anywhere in my posts? I think not.

    Since christmas eve is upon us, and upon us is christmas eve, here is your present. A gift from santa uma’s childhood memories. A poem in Syriac (language of Jesus) with some Arabic at the end. One out of so many like it. Most likely by Ephrem of the Syriac, whose output is estimated at over a million such verse. Recited by some random (Syrian) girl in some random church in Italy. Not related to Eric’s topic, but not at all unrelated to your misplaced comment. To not appreciate the beauty in that, even by non-believing infidels such as myself, is to not appreciate beauty anywhere else in this world.

    Now peace!

  118. @ esr

    > A lovely accent, Geordie is.

    I’m not familiar with it. Is this video accurate?

    And since there are different British accents, which one of them is the “standard” one from the American standpoint? Because there seems to be a stereotypical “British accent” portrayed in American movies and TV shows.

    1. >I’m not familiar with it. Is this video accurate?

      Reasonably, given that he’s describing a generalized northeastern (Anglo-Scottish border) accent. The more specific accent of the Tyneside around Newcastle is what people usually mean by “Geordie”; this has some features he doesn’t describe, including a very specific speech cadence that is difficult to mistake.

      >And since there are different British accents, which one of them is the “standard” one from the American standpoint?

      The South-Eastern English one that the voice coach speaks; it’s called British Received Pronunciation. His version is very light and modern. I suspect it’s not his birth accent and that he’s deliberately neutralizing region- and class-specific features in his native speech.

    1. >The “standard” British accent, to us Americans, is what the British call “Received Pronunciation”

      Look on the Wikipedia Received Pronunciation page for the recorded spoken sample, which is representative. The woman speaking exhibits a version typical for an upper-middle-class female born in the 1950s; more recent versions (like the voice coach in the Geordie-accent video) sound a little different, being more influenced by American TV and movies and thus a touch trans-Atlantic by the standards of previous generations. Having lived in England in the 1960s I really noticed the difference returning in the 1990s.

  119. You might also want to listen to Margret Thatcher, who speaks in RP, which is ironic since she grew up with a Lancaster accent, a very northern English accent related to the Yorkshire accent with similarly clipped articles. The fact that she felt the need to change her accent to succeed in British politics I guess says something about the Britain of her time. Nonetheless, she is as ever fabulous: What a contrast to the vile Hilary Clinton.

    https://www.youtube.com/watch?v=okHGCz6xxiw

  120. @esr:


    >>One reason I believe the FSI’s ranking of Arabic as most difficult is that it’s supported by the line of evidence laid out in Empires of the Word, where Ostler ranked languages by difficulty of acquisition by adults. He pointed out that even Arabic’s difficulty of acquisition is such that even immense religious prestige has been insufficient you cause it to replace other languages except where incoming invaders have essentially wiped out the native population.

    The statement above is full of factual errors. I will still read the book though because it appears that you recommend it. A quick glance at Amazon suggests it might be intellectually stimulating, and not lacking in useful information after filtering out the usual WASPo hubris.

    Now to the errors. Arabs and semites have never had empires in the traditional sense of Empire (ie vast territories acquired by military conquest, followed by shoving their culture down everybody’s throat). Quite the opposite in fact. Their empires, if they can even be called that, were little more than trading city states, and autonomous enclaves, usually spread along coastlines with little interest to venture to the interior. Nor does the concept of the modern nation state (a territory with a largely homogenous population speaking one national language and having one national culture) have any basis in semitic culture. In fact, much of the problems with the failure of the modern day middle eastern nation states is because the concept of homogenizing people into one nation state speaking one language and having one culture is so alien to their traditions.

    Arabic primarily spread through trade and sufi orders, not conquest and genocide (the caricature of islam in the west). Genocide in fact has not happened in Muslim history prior to the Armenian genocide, which was largely perpetrated by Turkish nationalists who were nationalists in the European sense of the word, and not islamic militants. The vast majority of Muslims even classified themselves as sufis till very very recently, when the wahabbi and shiite militant cults started spreading like wildfire after the Khomeini islamic revolution.

    The entire north Africa is still populated by its indigenous people, with those descended from the Arabian plate in the absolute minority. In fact, a country like Italy or Greece has far more people descended from the Arabian plate (albeit via pre-islamic migrations going as far back as neolithic) than a country like Morocco, or Mauritania. France has wiped out its linguistic diversity in the last 200 yrs yet the indigenous languages of north Africa are still spoken today. The arabs and arabic speakers have had 1400 years to do in north Africa what France did in the last 200 years on its territory and didn’t.

    Even at the height of the Arab empires (Ummayad and Abbassid) no Arab soldier even ventured south of the Egyptian delta (Egypt was equal to the delta at that time). And yet the vast space of territory south of delta all the way to south sudan speaks Arabic, if not as a native language then as a link language, and it mostly happened in the last 300 yrs.

    Semitic influences and links to north Africa were very strong well before Islam and semitic tongues were widely spoken there. In fact, one could argue that the reason Arabic today is far less spread than it could have been is exactly because the process of making it sacred also made it extremely difficult to learn. The grammar that Persian logicians created for the language over a 1000 years ago became as sacred as its script, even though any modern-day linguist will tell you that grammar is mostly shit. The grammar started with the assumption that everything in the Quran is grammatically correct. How could it not be if it is from God?!. If it comes from God it must be absolutely correct. Hence, what the professor called the fiendishly complicated grammar. In reality, the Quran is not only full of grammatical errors, but in fact cannot even be understood without good knowledge of Syriac. Syriac is to Arabic, what old English is to English, but if you tell that to any Muslim they will be offended, because their Persian scholars of a 1000 years ago told them otherwise, and told them that the Quran is the supreme example of ‘pure Arabic’ and of course also told them what everything in the Quran means, including most famously the 72 virgins which is the most wacko crazy shit meaning of the syriac word for “white grapes”.

  121. Uma,

    Arabs and semites have never had empires in the traditional sense of Empire (ie vast territories acquired by military conquest, followed by shoving their culture down everybody’s throat).

    The Rashidoun and the Umayyad might disagree with you.

    You could, I suppose, justify your statement by a delicate parsing of “shoving their culture down everyone’s throat,” but in fact most empires did nothing of the kind. Starting from the Persian Empire, many empires found it more convenient to leave alone local cultural practices, so long as political supremacy was assured. The Romans did the same, and the Austro-Hungarians even codified a degree of local religious and cultural autonomy until late in the 19th Century.

    Indeed, in the political-science sense, “empire” means precisely the opposite of what you think it means. It means precisely the subjugation of different cultural groups within the same political jurisdiction, with little attempt at creating a unifying nationalism or ethnicity. So to say that the Arabs never had an empire is absurd on its face.

  122. Hi,

    I live in Brazil, so have Portuguese as my language. I’ve studied Arabic for some time, and was able to successfully master its phonemes, mostly because I’m a musician, have sang at a choir during my high school and have learnt Esperanto, so I’ve learned to pay close atention to phonetics’ details. Nonetheless, I completely agree it’s an exotic phonology, yet in my case the main problem is just vocabulary.

    Although you’re discussing living languages, I’ve heard that the most difficult language that ever existed and is still used in some religious contexts is Sanskrit, due to its number of declensions and too subtile differences between certain vowels. Indeed, according to Wikipedia, the Proto-Indoeuropean language was also like this.

    As for English, unfortunately it is indeed difficult to speakers of latin-derived languages like us, partly because of the extra phonemes it contains (especially vowels), and also the difficulty to deduce the correct pronounciation upon reading a wrote text. Although English has a large number of words from greco-latin origins, the pronounciation of such words in English is too shifted away from the originals…

    Greetings.

  123. ESR: Arabic’s difficulty of acquisition is such that even immense religious prestige has been insufficient to cause it to replace other languages except where incoming invaders have essentially wiped out the native population.

    Not true – Arabic has substantially replaced previous languages throughout North Africa, but not completely. Berber is still widely spoken in the Maghreb, and Coptic in Egypt, though the dominant language in both areas is Arabic.

    Arabic replaced other languages in the Fertile Crescent, but AFAIK there is no evidence that “[Arab] invaders… wiped out the native population.”

    It is interesting that non-Arab madrissa students memorize long passages from the Koran without understanding any of it. Perhaps like opera singers?

    OTOH, Arabic made no impression at all in Iran.

    There have been other cases of language replacement without extermination. Genetic profiling indicates that Celtic-speaking Britons (Welsh, Highland Scots, Irish) and Germanic-speaking Britons (English, Lowland Scots) are descended from the same “Britannic” ancestors. The Danish and Saxon conquerors of the Middle Ages changed the speech, not the DNA.

    1. >Not true – Arabic has substantially replaced previous languages throughout North Africa, but not completely

      That’s because the Islamic genocide in North Africa was incomplete. And yes, I use the term “genocide” advisedly here; we can reconstruct what happened by haplotyping the local populations, and the picture is not pretty. Before the conquest there was a large Latin-speaking population there; they were slaughtered and replaced by Arabic-speaking incomers. We know this because the Arabic-speaking population has genetic markers continuous with Arabs farther east. The Berbers kept both their language and their non-Arab haplotypes. What didn’t happen was any general adoption of Arabic by the surviving indigenes.

  124. Jay Maynard on 2015-12-23 at 21:30:34 said: Forgive me if I don’t see the beauty in a book that commands its followers to kill the infidel.

    But it doesn’t. It commands the Faithful to conquer the infidel. AFAIK, conversion by the sword has been rare in Islam – it was much more common in medieval Christianity. Moslems have ruled over infidel populations for centuries without killing them. The Coptic population of Egypt is “proof of concept” of this.

  125. uma: Arabs … have never had empires in the traditional sense of Empire (ie vast territories acquired by military conquest)…

    That would be news to the nations conquered by Arabs at Qadisiya (636), Yarmuk (636), Alexandria (641), Guadalete (712), and Palermo (831).

    no Arab soldier even ventured south of the Egyptian delta…

    This would be news to the Nubians, who were invaded by Arabs from Egypt in 642.

  126. The grammar started with the assumption that everything in the Quran is grammatically > correct. How could it not be if it is from God?!. If it comes from God it must be absolutely correct. > Hence, what the professor called the fiendishly complicated grammar. In reality, the Quran is not only full of grammatical errors, but in fact cannot even be understood without good knowledge of Syriac. Syriac is to Arabic, what old English is to English, but if you tell that to any Muslim they will be offended, because their Persian scholars of a 1000 years ago told them otherwise, and told them that the Quran is the supreme example of ‘pure Arabic’ and of course also told them what everything in the Quran means, including most famously the 72 virgins which is the most wacko crazy shit meaning of the syriac word for “white grapes”.

    Hi, please can you give me sources for more details about all this? A very interesting subject! It can be in English, Spanish or Portuguese, the languages I understand.

  127. @Cleverson


    >>Hi, please can you give me sources for more details about all this? A very interesting subject! It can be in English, Spanish or Portuguese, the languages I understand.

    Yes. A good western source would be “The Syro-Aramaic Reading of the Koran”, by Christoph Luxenberg.

  128. LOL my above error ends up teaching a little Portuguese… In fact, we have a single verb (“poder”) which may mean both “to can” and “to may”. In other words, we don’t necessarily distinguish between not being able to do something due to not having physical or other material means, or not being able to do something due to ethical or moral concerns.

  129. @esr:


    >> That’s because the Islamic genocide in North Africa was incomplete. And yes, I use the term “genocide” advisedly here;

    What genocide are you talking about? The vast majority of Arabic speakers are natives of north africa by DNA haplos. If anything, DNA contradicts just about everything Europeans/Westerners imagine world history to be including their very own history.

    The fact that most English are not germanic invaders and instead arrived on the island after ice melted contradicts their own version of history and what they taught in their schools until recently. Including oh btw the stolen generations of indigenous children from thoroughly-genocided continents, who were brainwashed with the history of the la-la-universe of the anglo man.

    The percentage of males descended from Arabia in north africa is far less than the percentage of Mexican males descended from spanish invaders who have either mated with or raped the Amerindian women of Mexico.


    >> Before the conquest there was a large Latin-speaking population there; they were slaughtered and replaced by Arabic-speaking incomers.

    Are you joking or are are you joking here? Assuming this is correct (which is debatable of course) why not think of it as an alliance between natives (berber) and Arabs (with their superior religion and way of life) against an elite that is latin-alien to that region?

    The same is true of the Ummayyad conquest of Hispania which liberated the local population from Visigoth control and was definitely popular with the local population.

  130. >My objection was not that it wasn’t common, but that it was never ever used except in the context of bridge. That is a pretty emphatic statement you made.

    It’s a true one, but you can take it as anecdote rather than data. I’m not sure how well either of us represents the general population.

    >Curiously so since I have never heard it in the context of bridge, and have frequently heard it in the context of our serpentine friends. Of course I hang out a lot more places where there are snakes than I do bridge players, never having played.

    Other way around for me. I played quite a lot in school, in my teens and twenties.

    >Having said that, I hate snakes, I am ambivalent on bridge.

    I was tempted earlier to make a joke about you needing to stop hanging out with reptiles, but bridge players have their downsides, too. :)

  131. @Mastiff:


    >> The Rashidoun and the Umayyad might disagree with you.

    How so? Most accounts of the Rashidoun and Ummayads and were little more than political propaganda and legends written by the Persian chronichlers working for their enemies (e.g. the Abbasids). They are contradicted by a mountain of what historians call primary sources.


    >> You could, I suppose, justify your statement by a delicate parsing of “shoving their culture down everyone’s throat,”

    What is it that I wrote, that leads you to believe that I am into delicate parsing of words?


    Indeed, in the political-science sense, “empire” means precisely the opposite of what you think it means. It means precisely the subjugation of different cultural groups within the same political jurisdiction, with little attempt at creating a unifying nationalism or ethnicity.

    That is perhaps the anglo-saxon flavor of empire, which indeed explains the British success relative to the French for example who seeked to impose their culture everywhere they went and failed miserably. And the tolerant flavor only applied to those territories where the British did not send their settlers to displace local population.

  132. @esr

    I did a quick search and found this article that is written by a native English speaker (Australian), and someone who learned both Arabic and hebrew. It goes into detail and explains why the FSI’s ranking of Arabic is “absurd” and it largely agrees with what I said here minus the history lesson of course. You will find it interesting especially the grammar part.

    1. >I did a quick search and found this article

      Second comment, from a Levantine native speaker “But i have to mention you are making arabic way too easy and it is not true.” I will still believe the FSI’s measurements over anyone’s anecdotes.

  133. I have to agree with Uma’s most recent link. I am a bit of a biased observer, having studied Hebrew to some extent since childhood, but in particular the three-letter root system makes learning a Semitic language much, much easier than other languages without strong root systems. I can take a single root such as ??? and effortlessly transform it into seven different verbs by applying the different constructions, ranging from the simple “he guarded” to the intensified passive “he was continuously guarded”. Thus, Hebrew takes a relatively compact set of roots, a few thousand, and can create a flexible yet succinct language.

    I find this far easier than learning French or German, both of which I am doing now. In particular, Hebrew is far less fussy about prepositions and articles.

    Now, it is true that Semitic languages read right to left, and also use a non-Latinate alphabet. Perhaps most Americans have far more trouble with the initial transition than I appreciate.

  134. @Mastiff:


    >> Thus, Hebrew takes a relatively compact set of roots, a few thousand, and can create a flexible yet succinct language.

    What you describe can be alternately explained as this: If one were to simply devise a figure of merit for languages whereby information density is normalized to the number of phonemes semitic languages would be way at the top.

    This is in fact the reason why Arabic poetry is untranslatable whereas Persian poetry can to some extent be translated. It has to do with information density per unit phoneme. Poetry in general does not lend itself to translation, and translations themselves can be thought of as original works in their own right, but the number of phonemes allowed to express an idea is absolutely crucial as to whether a translation can even be attempted.

  135. @Mastiff:
    >I am a bit of a biased observer, having studied Hebrew to some extent since childhood, but in particular the three-letter root system makes learning a Semitic language much, much easier than other languages without strong root systems.

    I think you’ll find that the average native English speaker with no childhood exposure to Semitic languages will find triconsonantal roots to be a stumbling block.

    A particularly bad bit in that article is where he mentions lack of a neuter gender as an advantage for learnability. For English speakers, it’s not.

  136. Meh, to a native English speaker without the concept of gender for anything that doesn’t have genitalia, neuter gender isn’t as much of a stumbling block as gendered nouns in general.

  137. @esr
    > I will still believe the FSI’s measurements over anyone’s anecdotes.

    I am with Eric. The FSI has taught hundreds of thousands of people to speak foreign languages fluently. Functionally fluent that is, so that these students can live in the countries and speak the language and communicate at the very highest levels of government. As far as I know they have no hidden agenda. In fact given the super soft — love everything Muslim — attitude of the US Government, perhaps they have a bias in favor of softening the ratings of Arabic.This is quality data, and we should rely on it. It is the classic debate of theory over data. Good scientists take the latter over the former every time (take note climate scientists.)

    You can argue over individual parts of the language all you like. I agree, constonantal roots are cool, and remind me a lot of Esperanto, but the binyanim method of varying the verb meaning isn’t that much different than conjugations in other languages. Also many languages do have an informal process for converting different parts of speech into another., encapsulated perfectly in a quote I still remember from a lecture given by Mike Lesk of “lex” fame who said “the great thing about English is that nouns verb so easily.”

    However, fluency in a language is a combinatorial explosion of all the different parts, so one significant part being easy is readily offset by other things being hard. In the case of Semitic languages a lack of cognates for English speakers would be an example. And in fact the further you get from classic Latin and Germanic languages, the more the ugly head of phonology raises its head. And that is perhaps the thing that betrays fluency more any any other. Our pretty girlfriend’s struggle with German to English phonology is a walk in the park compared to English to Mandarin phonology. Remember that this is a measure difficulty of language acquisition for English speakers, not a measure of the intrinsic complexity of the language.

    1. >Our pretty girlfriend’s struggle with German to English phonology is a walk in the park compared to English to Mandarin phonology.

      Actually, I think Mandarin phonology might be a touch easier than that of English, if she hears tones naturally (of course that’s a big if). The aspirated/unaspirated distinction central to Mandarin consonants is not difficult, and the two retroflexes are required only in a rather exact Beijing dialect.

      It’s the written form of Chinese that’s brutal, not the spoken form, which is IMO rather simple as world languages go. Empires of the Word backs me up on this.

      But it is also possible that, being unusually good at acquiring alien phonologies myself, I am underestimating the difficulties here.

  138. > The aspirated/unaspirated distinction central to Mandarin consonants is not difficult

    > But it is also possible that, being unusually good at acquiring alien phonologies myself, I am underestimating the difficulties here.

    Honestly, in light of this, I’m not even sure why you’re willing to even make assertions about aspects of phonology not being difficult in the first place. Unaspirated unvoiced consonants do exist in English, but they’re exclusively allophones, and are therefore difficult to hear or produce in positions not matching where they actually appear in English. I’m not sure I can even reliably hear the difference between one and the voiced version even though minimal pairs definitely exist in principle [rating vs raiding].

    To say nothing, of course, of the tone issue.

    1. >Honestly, in light of this, I’m not even sure why you’re willing to even make assertions about aspects of phonology not being difficult in the first place.

      Because I can tell what’s difficult for me, and I figure I compress the difficulty scale for monoglot English speakers. That is, because an uvular fricative is more difficult for me to pronounce that a retroflex stop, I think I’m pretty safe ground expecting the relative difficulty to be the same even if I have an easier time with both in some absolute sense. And phonemic aspiration is towards the low end of the difficulty scale.

      Tones kind of stick out to one side as an orthogonal complication – you either get them easily due to innate pitch sense or they’re nastily difficult in a way rather different from forcing your vocal tract into novel contortions.

  139. @esr:


    >> Second comment, from a Levantine native speaker “But i have to mention you are making arabic way too easy and it is not true.” I will still believe the FSI’s measurements over anyone’s anecdotes.

    I checked the comment. Nothing in the comment related to the grammar at all. It is mostly referring to the intelligibility of the dialects and the pharyngeal consonants which are in fact not pharyngeal at all.

    The video on that link of the girl teaching the pronunciation of the A’yn pharyngeal consonant, one of the 3 phonemes that cannot be approximated by any sound in English, is a funny video that teaches a good technique.

    I would agree with that comment in the sense that levantine arabic would be a bad starting place for anyone who wants to learn Arabic. Levantine Arabic to Egyptian Arabic, is what British English is to American English. There is a lot more diversity, and complexity in the levant.

    The comment though does bring attention to another feature of Arabic which might explain the widely different opinions: While some forms of Arabic might be as difficult as category IV and category V, other forms may be as easy as a category II. Levantine would perhaps be a category IV. Egyptian would be category II. And the type of convoluted arabic taught in Islamic seminaries is a category V*.

    The different forms of Arabic (referred to inaccurately dialects) can to an extent be thought of as languages in their own right with quite different grammars at times but with the same key concepts of roots, derivations, and templates remaining intact. In spite of being different languages they are all mutually intelligible to a very large extent and also intelligible with modern standard arabic. It is the 3-letter-roots and templates that make them intelligible. Think the myriad of lisp dialects all sharing the power of S-expressions, macros, and tail recursion and the same ‘style’ of expressing ideas.

    I can look for a good video on what most forms of levantine arab sound like. Outside of the cosmopolitan Beirut-speak, and the very sleek but very complex Damascus-speak, it is really rough with the consonants you object to pronounced very boldly and crisply.

  140. >Esperanto was invented for the kind of people who should be opposed at every turn in the name of liberty.

    Man, this is one of the dumbest things I have ever read. Esperantists are some of the most intelligent and open-minded people on the face of the earth. The irony is, there are Klingon speakers among Esperantists as well. Heck, there are even Volapük speakers among Esperantists. From where did you get that juicy idea?

    Many of the Esperantists I have met were polyglots by the way. The Esperanto community is diverse and multifaceted. It is not as unified as you might think. Read this for instance:
    https://en.wikipedia.org/wiki/Raumism

  141. “Because I can tell what’s difficult for me, and I figure I compress the difficulty scale for monoglot English speakers.”

    I agree with your rankings of relative phonetic difficulty, and think they are also correct for those of us who are crib monoglot.

    Where I think you err is when you weigh phonetic difficulty vs. grammar and vocabulary difficulty. I know that for me, learning new grammar rules is pretty easy, and while cognates are helpful and aid learning, I would always choose simple phonology with lack of cognates over complex phonology with cognates.

    I think the reason that the FSI list rates Japanese as so difficult is that they are looking for true fluency at the diplomatic level, rather than ability to function as a tourist. I doubt there any other language in the world with so much potential to unintentionally insult the listener through errors in grammar and where choice. But for those us traveling who just want to ask where the museum is, these words are more harmless and will just be put down to gaijin stupidity.

    When I try to learn weird in a final language, the experience is so frustrating as to nearly reduce me to tears. I *can’t hear* the subtle differences between similar-pitched tones! I’m not at all sure that additional practice would fix this.

    1. >Where I think you err is when you weigh phonetic difficulty vs. grammar and vocabulary difficulty.

      I admit this is plausible.

  142. Esperantist: “From where did you get that juicy idea?”

    Fromt he article you quote:

    Essentially, the “Internal Idea” of the Esperanto movement (according to the Enciklopedio de Esperanto) has been that an easy-to-learn common language can be a significant influence for peace by promoting interpersonal understanding across cultural and national boundaries. The movement’s main goal has been to promote Esperanto as that common language with a focus on the goal of world peace.

    The kind of people who think world peace can be promoted by promoting interpersonal understanding across cultural and national boundaries are the kind of fuzzy-headed leftists who freedom-loving people must oppose at every opportunity. No amount of understanding will prevent many who see conquest as their national interest from pursuing it with every means at their disposal.

  143. @ Jay Maynard, esr, and Jessica Boxer

    Thanks for your replies. I wasn’t aware of Received Pronunciation. As for Thatcher: in my book, she gets extra credit for humiliating the Argentine military junta. :-)

    @ Jay Maynard

    > No amount of understanding will prevent many who see conquest as their national interest from pursuing it with every means at their disposal.

    A couple of days after the Paris attacks, I read a letter to a newspaper that exemplifies what you describe. The letter encouraged us to be loving and peaceful in our everyday actions, in order to defeat hatred and thus eventually end terrorism. How well-meaning, and how misled!

    To be fair, the sender mentioned he’s sixteen years old – or so – and isn’t necessarily a leftist (the newspaper in question, while pluralistic, is essentialy right-wing). But I hope he, and many others like him, soon realize that it’s all wishful thinking. And that stuff like lighting candles, while noble, is perceived by the enemy as an auspicious sign of Western weakness.

  144. “When I try to learn weird in a final language, the experience is so frustrating as to nearly reduce me to tears. I *can’t hear* the subtle differences between similar-pitched tones! I’m not at all sure that additional practice would fix this.”

    Let’s try that again:

    “When I try to learn words in a tonal language, the experience is so frustrating as to nearly reduce me to tears.”

    This is a great blog, but I sure wish WordPress supported editing. My Android Swype interpretations of my intended keying can get pretty crazy sometimes.

  145. Frognot:

    I have posted a lot of stuff that backs up my viewpoint. I have also taught modern standard arabic (much more difficult than Egyptian arabic) on a number of occasions to students here in the US. So I do have some perspective on this.

    If you want to believe what the government tells you, then by all means.

    And no the US government does not have an incentive in making Arabic sound easier than it is because of its “super soft — love everything Muslim — attitude” (what Jessica Boxer suggests). It is the other way around. The only way they can explain their repeated fuckups is by classifying a language with an easily-readable phonemic script to be as inaccessible as Chinese.

    In fact your defense institute classifies pashto along there with Chinese. Anyone who knows anything about languages will tell you that Pashto is an indo-european language written in an alphabetic script and is in no way as insane as a tonal language or chinese characters. But because of the Afghanistan fuckup it, too, must be placed in the same category as Chinese!

  146. @ Jay Maynard:
    >Fromt he article you quote
    The article I quote is about a group of people within the Esperanto speaking community whose ideas are radically different from traditional Esperantists. The point that you keep missing, is that Esperanto is not and never has been a political movement or an ideology and it accommodates all sorts of people who share the ability to use the language. It is just a language with a small but interesting community of speakers. Different people with various beliefs and interests are part of the group. It is true that it gravitates more to intelligent people but the fact that you try to label all of them as leftists just doesn’t make sense.

    >No amount of understanding will prevent many who see conquest as their national interest from pursuing it with every means at their disposal.

    Of course. That’s exactly why the idea of a universal language is not enough to bring about world peace. Peace is dependent on many other factors as well. I would oppose those Esperanto speakers who would religiously defend Esperanto as a “solution” to human conflicts throughout the world, but then there are very few people who think like that. Nowadays, Esperanto is just a small linguistic community with its own unique literature and open culture. It is a good tool for finding international friends. I can understand why many people would be uninterested in learning it but I honestly see no reason why any sane person would go out of their way to oppose it under any pretext.

    Anyways. A little off-topic, but I couldn’t help but notice this in Eric’s old archives:
    http://www.catb.org/esr/tengwar/esperanto-tengwar.html
    Very interesting !

  147. > world peace can be promoted by promoting interpersonal understanding across cultural and national boundaries

    This was definitely true _at the time Esperanto was developed_, though. One reason that the world today is so much more peaceful than in L. L. Zamenhof’s time is the kind of “interpersonal understanding” that has been fostered by such things as global communications and the spread of free trade. (Although Esperanto has had no role in this; its main interest nowadays may be in its propaedeutic value to monolingual speakers.) Indeed, the very reason that we now have to worry about such things as militant fundamentalist Islam is the way it _actively hinders intercultural competence_ in its followers, e.g. by praising hatred and violence towards the “infidel” outsiders. But this is a very unusual memeplex, overall; and most people are sane enough to realize that aggressive “conquest” is almost always a bad deal.

    1. >But this is a very unusual memeplex, overall;

      Nonsense. And dangerous nonsense, blinding you to reality.

      Xenophobia and institutionalized tribal violence towards outsiders is what’s historically normal for human cultures. Before the modern era exceptions were rare and largely confined to “middleman” mercantile polities on long-distance trade routes.

      These days we are nearly all merchants (as opposed to subsistence farmers dominated by warlords) which has created the illusion that the older pattern of societies with a rigid and violent self/other distinction is obsolete and will not recur. ISIS puts the lie to this; so do the street gangs in our own cities.

  148. “most people are sane enough to realize that aggressive “conquest” is almost always a bad deal.”

    Tell that to, say, Vladimir Putin.

  149. > which has created the illusion that the older pattern of societies
    > with a rigid and violent self/other distinction is obsolete and will not recur.

    To amplify this from a different direction –

    Humans evolved while living in small (by modern standards) groups. In the EEA, you might only see a couple of hundred different people in your entire lifetime, and most of them were related to you, or at least people you lived and worked with day after day. All others were outsiders, who were potentially after your food, water, or mates. Distrust of the other was not only the norm, it was evolutionarily prefered.

    Today, we see hundreds of thousands of others via the media, and may well meet up (or at least be in the same physical space) with many thousands over our lifetimes. Most of those are not part of our “tribe” or “clan”. Our inner angels may wish to treat all of them as potential allies and friends; our inner killer ape still regards them with fear and hate.

  150. Street gangs are not “societies with a rigid and violent self/other distinction”. They’re criminal groups which _rationally_ use violence to defend their “turf’, because they lack the benefits of government-enforced property rights. Overall, the modern world may not have “obsoleted” older patterns outright, but it has definitely given ‘merchant’-oriented societies the upper hand. The Chinese (certainly no strangers to rigid self/other dichotomies, historically) understand this quite well – and I have to assume that Vladimir Putin does too; his sabre-rattling and anti-Western stance, while indeed concerning, is to some extent a concession to popular sentiment. But he remains one of the _most_ market-oriented political leaders in that country, and these attitudes are what matters most in the longer run.

    1. >They’re criminal groups which _rationally_ use violence to defend their “turf’, because they lack the benefits of government-enforced property rights.

      Sometimes.

      We could argue over where the boundary between criminal group within a functioning society and autonomous micro-society is all day long. but that would miss the point. The armature of social instincts on which the criminal groups are built is identical to the armature by which tribal-band-sized societies maintain themselves. Both count as a recurrence of pre-market patterns of organization.

  151. @Esperantist

    Something like Esperanto would have made a lot of sense back in the 20s as a language of scientific publications. It was WWII that gave us the current situation where English is the language of science.

    The amount of scientific publications coming out of China and Asia with really poor English illustrates the problem.

  152. Riight, uma. The US government really, really wants to spend 64 weeks teaching someone (who’s proven to have the aptitude and interest) effective fluency in a language in a total immersion environment, when they could do it in 24 weeks — I guess just for giggles and grins? It’s taught by native speakers, so I guess they must be incompetent? I take it that English is not your native language, which makes me wonder how you became an expert on the difficulty of coming from English to another language? More so than those who actually specialize in teaching that to significant numbers of English-speaking students each and every year? I strongly suspect that Britain’s language training courses will be broadly similar, though. Google brings up so much chaff that I can’t sort through it in the time I have. You did post a link that agreed with you, but I’m pretty sure I can find others claiming that Klingon is by far the easiest language because of it’s harsh, primal bestiality that is a pure form of nature’s true beauty or somesuch nonsense. Find some groups that specialize in teaching truly significant numbers of native English speakers (i.e., something more than anecdote) and I’ll attach more weight to it. Otherwise, I shall cut and paste blather from 19th Century English imperialists about how English is the purest, most beautiful, and utterly easiest language to learn. An utter joy to behold, and a jewel to be clasped to the heart. Blech.

    1. >Otherwise, I shall cut and paste blather from 19th Century English imperialists about how English is the purest, most beautiful, and utterly easiest language to learn.

      They had a better point. My ESL-speaker friends who speak other languages as well are pretty unanimous about English being very easy relative to their other languages. It is even pretty well understood why: nearly isolational SVO grammar is very like the way trade pidgins work and seems to facilitate adult acquisition.

  153. Frognot:


    >> really wants to spend 64 weeks teaching someone (who’s proven to have the aptitude and interest) effective fluency in a language in a total immersion environment, when they could do it in 24 weeks

    So explain to me then the difference in Pashto listing between its FSI listing here, and its listing on the defense institute that groups it with Chinese. And give me a logical explanation.

    FYI, the mere suggestion that there is an immersion environment for modern standard arabic is idiotic, when the language is not used outside of formal media outlets and newspapers. The arabic dialects are not even standardized as to how you write them. They are generally “looked down” on by a certain smug literary class as not as beautiful as modern standard arabic (MSA). There is no formalized way for example of writing down say the Iraqi dialect. So here you have a written language that is understood by everybody that is not spoken by people in their day to day life. Imagine if the whole country spoke in ebonics, but the media is in the standard English that we read. The distance (MSA vs dialects) is even much further than that between standard English and ebonics. It is only due to the unique features of the semitic language family that makes them easily intelligible.


    >> I guess just for giggles and grins? It’s taught by native speakers, so I guess they must be incompetent?

    That could very well be. I can count only one US diplomat that I saw on Arabic TV on youtube that I considered speaks good arabic. Not so with the Brits. So clearly somebody is incompetent in the government.


    >> You did post a link that agreed with you

    Not so. I posted a video and a link. The video, by a british professor and arabist who is involved with the arabiconline.eu service, an EU-funded arabic education program based out of Britain, which so far as I know speaks English as its native language


    >> I take it that English is not your native language, which makes me wonder how you became an expert on the difficulty of coming from English to another language

    It is called ‘teaching’. That is teaching students whose first language is English.

    If you are good with google, you’ll find that there many problems teaching arabic in its native countries to its native speakers. Because the religious nature of the language has thoroughly retarded its teaching. I personally picked up my near-perfect command of modern standard arabic, watching cartoons dubbed into Arabic when I was growing up. Not out of any book/curriculum I studied in primary school.

  154. “These days we are nearly all merchants (as opposed to subsistence farmers dominated by warlords) which has created the illusion that the older pattern of societies with a rigid and violent self/other distinction is obsolete and will not recur.”

    TL;DR — merchant behavior is so difficult for a society to create, it typically needs to be imported rather than developing organically.

    I read an excellent history of the Early Middle Ages in northern Europe. I don’t have the title handy, but can look it up next time I’m in the library if anyone is interested.

    With the fall of civil authority,, after the collapse of the Western Roman Empire, long-distance travel became too dangerous, so trade declined. The northern European areas that had once been part of Roman trade networks regressed to subsistence agriculture.

    This lasted until political stability was restored in the medieval period, with feudalism replacing ancient empire. But even when trade was restored, it was not locals who recreated the networks, but traders coming in from regions where merchants had flourished, such as the Eastern Roman Empire (Byzantine).

    What makes this especially interesting is that some local residents did become involved with trade, but they did this by joining with outside merchants to learn the business. This was a commonly-taken economic opportunity for former slaves who had earned their freedom, but had no land to farm. So it wasn’t lack of people willing to engage in trade that held back local development, so much as lack of knowledge of how to make it work, and lack of interest in trade by those with capital (wealthy feudal aristocracy).

    You can see these challenges in the ancient Greek world as well. While trade was widespread in the Aegean, the metics who ran the trade were not granted citizenship in the city-states where they lived. They could grow rich, but this did not give them any political role in the Greek city-state. In late Middle Ages, these merchants formed their own city-states oriented to their life, cities like Genoa, Venice, Antwerp, and Augsburg.

    Merchants were never the normal class until very late in human history. And populist movements today are rooted in this same suspicion of trade and rich merchants, who are seen as outsiders. The multinational nature of everything from Greek metics to corporations doing tax inversions strengthens the sense of Other.

  155. Possibly off-topic, but what’s with phonetically doubled but ortographically single consonants in English. ESL speaker here, and, for example, the s in “disappear” sounds double. I’ve noticed other ESL speakers having the same problem (a dead giveaway is spelling it as “dissapear”).

    1. >Possibly off-topic, but what’s with phonetically doubled but ortographically single consonants in English.

      English doesn’t have phonemic consonant lengthening in root words, though something similar to it occasionally happens around word boundaries and in a handful of compound words like “lamppost”. Thus, this is mainly just an orthographic quirk.

  156. > Possibly off-topic, but what’s with phonetically doubled but
    > ortographically single consonants in English. ESL speaker here, and, for
    > example, the s in “disappear” sounds double. I’ve noticed other ESL
    > speakers having the same problem (a dead giveaway is spelling it as
    > “dissapear”).

    Orthographically doubling the consonant in English usually serves a purpose other than lengthening its pronunciation. It often causes the preceding vowel to take the “short” pronunciation (compare biding [first i long] to bidding [first i short]). With certain consonants, it also differentiates between different phonemes (compare fusing [s pronounced as z] to fussing [s pronounced as s]). Only rarely is its purpose to lengthen the consonant.

    It is only because “disappear” is recognised as the prefix “dis” plus the root verb “appear” that we pronounce it as we do (compare “disable”, but contrast “disaster” [first s pronounced as z]). This is why understanding root words, suffixes, and prefixes is emphasised heavily in primary school. It’s not uncommon for people to mispronounce unfamiliar words when they incorrectly parse prefixes. For example, you might first encounter the relatively uncommon prefix “infra” in the word “infrared”. You might correctly hypothesise the existence of the prefix “infra” if you recognise the root adjective “red”, but I suspect more learners (certainly true among those I know) recognise the common prefix “in” and the common suffix “ed” and incorrectly hypothesise the existence of a root verb “frare”. This leads them to mispronounce “infrared”, usually rhyming with “impaired”.

    1. >Orthographically doubling the consonant in English […] often causes the preceding vowel to take the “short” pronunciation

      That is true. So the orthographic quirk is not entirely meaningless.

  157. While we’re at it: why is “bass” pronounced like “base”? We ESL speakers would expect it to rhyme with “grass”.

    1. >While we’re at it: why is “bass” pronounced like “base”? We ESL speakers would expect it to rhyme with “grass”.

      And it does, if you’re talking about a fish rather than a musical instrument.

  158. Fascinating digression.

    >>They’re criminal groups which _rationally_ use violence to defend their “turf’, because they lack the benefits of government-enforced property rights.

    >Sometimes.

    Much of traditional ‘organized crime’ business depends on the inability of the government to properly protect or enforce property rights. Think ‘protection racket’.

    I’ve always been of the opinion that a territorial gang/organized crime outfit was effectively a low-level insurgency, because to succeed they must undermine the power of the ‘proper’ government and effectively shadow or replace it with their own. (Certainly it’s common in history for the life cycle of a politically motivated insurgency to end as an organized crime syndicate/criminal gang.)

    >We could argue over where the boundary between criminal group within a functioning society and autonomous micro-society is all day long. but that would miss the point. The armature of social instincts on which the criminal groups are built is identical to the armature by which tribal-band-sized societies maintain themselves. Both count as a recurrence of pre-market patterns of organization.

    Well that’s definitely true. Pretty sure they share a lot of rituals too, and use them for the same purpose.

  159. That article to which Jay posted a link makes the point well. Yes, English speakers expect the first noun in the sentence to be the subject in most cases, regardlesd of the form or ending of the noun. There’s that SVO isolational grammar sticking it’s nose under the tent again.

  160. Have you tried to learn Polish? A couple of people I’ve talked with about it have said that it’s significantly more difficult than Russian (the cases and genders have extra exceptions and complexities).

    1. >Have you tried to learn Polish?

      Not seriously, though I did pick up on a a little of it when I visited Poland. A fair number of compound words have recognizable Latin roots.

  161. The word Indo in “Indo”-European refers specifically to India and in particular to Sanskrit and it’s descendants, Hindi, etc (akin to Latin and it’s descendent, Italian).

    There is something off about your analysis re: Hindi to a native English speaker.

    Hindi is maybe 5 min more work than Italian, even the numbers and pronouns are almost the same, “tu” is you in Hindu (exactly same as “tu” in sanskrit and latin), “do” is two in Hindu (exactly same as Italian), etc, “panni” is water, up/above is “uper”, “dad” is “pita”, “que” is “que/queya” etc. The grammar is almost the same too.

Leave a Reply to stille Cancel reply

Your email address will not be published. Required fields are marked *