Anne Trafton of the MIT News Office had a report last July on an interesting study:
In a new study, a team of neuroscientists and psychologists led by Amy Finn, a postdoc at MIT’s McGovern Institute for Brain Research, has found evidence for another factor that contributes to adults’ language difficulties: When learning certain elements of language, adults’ more highly developed cognitive skills actually get in the way. The researchers discovered that the harder adults tried to learn an artificial language, the worse they were at deciphering the language’s morphology — the structure and deployment of linguistic units such as root words, suffixes, and prefixes.
“We found that effort helps you in most situations, for things like figuring out what the units of language that you need to know are, and basic ordering of elements. But when trying to learn morphology, at least in this artificial language we created, it’s actually worse when you try,” Finn says. […]
Linguists have known for decades that children are skilled at absorbing certain tricky elements of language, such as irregular past participles (examples of which, in English, include “gone” and “been”) or complicated verb tenses like the subjunctive.
“Children will ultimately perform better than adults in terms of their command of the grammar and the structural components of language — some of the more idiosyncratic, difficult-to-articulate aspects of language that even most native speakers don’t have conscious awareness of,” Finn says.
In 1990, linguist Elissa Newport hypothesized that adults have trouble learning those nuances because they try to analyze too much information at once. Adults have a much more highly developed prefrontal cortex than children, and they tend to throw all of that brainpower at learning a second language. This high-powered processing may actually interfere with certain elements of learning language.
You can read more about how the study worked at the link; unfortunately, “Still unresolved is the question of whether adults can overcome this language-learning obstacle.”
“Still unresolved is the question of whether adults can overcome this language-learning obstacle.”
I agree with most of the excerpt except that in this last statement I would use “when” rather than “whether”. Obviously many adults have been able to overcome the obstacle. But many others stop learning when they are satisfied with a functional command sufficient for their purposes.
I guess what’s unresolved is whether you can come up with a teaching/learning strategy that can finesse the prefrontal engagement, and make it as easy for adults as for children. (Lots of people who advertise to me on social media assure me that they’ve already solved this, so I don’t know what these MIT folks are on about.)
declarative memory, which stores knowledge and facts, would be more useful for learning vocabulary and certain rules of grammar. Procedural memory, which guides tasks we perform without conscious awareness of how we learned them, would be more useful for learning subtle rules related to language morphology.
This meshes with my experience. Over-intellectualisation hinders language acquisition. Exercises that force you to “just say it” are much more helpful. That’s why so-called “conversation classes” are so sought after — learning the grammar isn’t the same as learning the language.
Lots of linguists find that their most productive language-learning comes during fieldwork among hospitable locals who are constantly plying them with the local firewater. Alcohol helps lower the social inhibitions that prevent making an active effort; you don’t feel so ashamed to make mistakes any more. Would booze also help with this overthinking that the authors here describe?
Hehe, the problem with booze is that you perform great at the time but forget it all the next day!
I have extremely simple strategy for learning languages.
It’s exposure to large portions of annotated/translated text with audio playing simultaneously over relatively short period of time.
Say, 1 million characters of text in 10-12 hours. Doable over a weekend and you end up with ability to read fluently in new language.
Of course, this approach gives only passive knowledge and you’ll have to make more effort to learn to converse in this language. But it’s always easier to learn to speak a language when you can already read it fluently.
the problem with booze is that you perform great at the time but forget it all the next day!
I’d attribute a similar effect to another substance . . .
Would booze also help with this overthinking that the authors here describe?
In my experience – definitely. I have even managed to hold long conversations in Mandarin after drinking enough baijiu, and I find when sober I get far too hung up on the tones. The problem is, while alcohol may help you speak a foreign language better, it reduces your ability to accomplish anything productive in that language so I don’t know that it is a useful fix.
Ah the famous Listening-Reading thing. My personal verdict is that it works either for a cognate language (another Romance, Germanic, Slavic or Chinese language when you already know one well) or for the kind of faux débutant situation, when you know a lot about the grammar but cannot put that information to use.
Not necessarily.
I once did this with Vietnamese, the language which I never studied before and not related to any language I know.
Granted, like Chinese, Vietnamese doesn’t really have a grammar to speak of and with languages with more challenging grammar such approach would have less effect.
I tell you what, let’s do an experiment. Name a really challenging language and I am willing to spend a few hours to check how this method works with it.
Polish.
I can read Polish fluently, so we have to chose some other language to check the method. Preferably one based on Latin, Cyrillic or Greek script or at least with easy way of obtaining reliable machine transliteration.
Would Hungarian do?
What languages don’t you know?
And anyway, I should try it again myself with the languages that I have always wished to learn.
Ten years ago, I was trying to learn new language every week or so and I ended up with dozens of half-learned languages which inevitably I forgot afterwards quite quickly.
Good thing is that a language acquired (doesn’t matter if it’s incompletely learned) is never forgotten completely.
The stuff still sits in your head somewhere and when you need it again, it comes up really fast.
Last week I decided to brush up Tajik, a language which I studied for a week some eight years ago and never had a chance to use again since then. It’s unbelievable how much of it I still recall.
AJP Crown,
Regarding Polish. I used somewhat modified method with Polish. I downloaded audiobooks of Andrzej Sapkowski’s Trylogia husycka (total playing time – 70 hours) and listened to them while reading simultaneous Russian translation.
It’s hard to imagine more effortless and enjoyable experience in learning foreign language.
Ten years ago, I was trying to learn new language every week or so and I ended up with dozens of half-learned languages which inevitably I forgot afterwards quite quickly.
Yes, I was doing that kind of thing a couple of decades ago, studying Georgian, Persian, Hungarian, etc., just long enough to get a basic grasp and then moving on. And you’re right, it’s surprising how much sticks with you.
Would booze also help with this overthinking that the authors here describe?
Another million-dollar idea, as a friend of mine used to say! Now, that would make for a very popular language class.
On booze in language learning, there’s a classic study on the effects of alcohol on pronunciation ability, Guiora et al. 1972:
http://deepblue.lib.umich.edu/bitstream/handle/2027.42/34041/0000318.pdf
I’ve heard about it many times, though I still haven’t read it. But I feel positively disposed towards any experiment whose apparatus includes cocktail napkins, stemmed glasses garnished with a cherry and a twist of lemon peel, and “varying amounts of a punch known for its deceptive potency”.
any experiment whose apparatus includes cocktail napkins, stemmed glasses garnished with a cherry and a twist of lemon peel, and “varying amounts of a punch known for its deceptive potency”
I’m sort of a pail and siphon hose guy myself.
Paul Ogden: “varying amounts of a punch known for its deceptive potency”
The optimal solution is where the struggle for fluency gives way to full immersion.
SFReader’s system is intriguing but I’m not sure I understand it — so the audio is in the new language, and you’re simultaneously reading a translation in a familiar language?
(If so, this is more or less how I learned most of my Greek and Latin, except that the left-hand pages of Loeb Classical Library volumes had to stand in for the audio.)
I can see that working with Russian and Polish (“essentially a light form of Russian that even Germans can master”), but I hardly think it would work with, say, English and Japanese, or ordinary anime fans would know a lot more Japanese than they commonly do.
@Breffni: So in other words, there’s a Ballmer Peak for language learning.
Adults certainly can pick up languages by immersion in child-like fashion.
When I lived in NE Ghana I spent a lot of time learning the major local language, Kusaal. Quite apart from the fact that (like all natural languages) the language itself is beautiful and fascinating, the experience itself was a great intellectual adventure as it was my first experience of learning a language without any existing learning aids like grammars and dictionaries, and having to do all the analysis myself.
I probably know more *about* Kusaal as a system than anyone else in the world; I am one of only a handful of Europeans ever to be able to manage a conversation in the language.
Almost at the same time as I arrived, so did a new colleague from the south of Ghana. She was about 30. Her native language was related to Kusaal (if at all) more distantly than Welsh to Urdu. She had no linguistic training or indeed particular interest in languages as such. Her work however meant that she spent all her time with Kusaal speakers. After a year or so she was incomparably better at the language than I was, with no studying (there’s nothing you *can* study) or conscious effort at all.
This is not a particularly unusual story in West Africa, where unless you plan to spend all your days near your own birthplace, you will often need to pick up another language to a degree where you can function completely in it. Or two languages. Or three …
“Now, that would make for a very popular language class”, except, for example, in Saudi Arabia to improve your Arabian, as, in the past, in the USA during Prohibition to improve your AmEn.
Well, I’m going to drink some beers just in case.
“as it was my first experience of learning a language without any existing learning aids like grammars and dictionaries”
… well, not the first, obviously. But you know what I mean …
I suppose mothers are existing learning aids …
Ah yes, one of those typical Niger-Congo languages, only with the nouns spoken backwards.
Indeed so. Suffixed noun class markers instead of prefixed (though in Kusaal itself the agreement system is defunct, and there’s only natural gender: human/non-human. The system lives only in the pairing of particular singular markers with particular plural markers, and the fact that as is usual with noun-class systems there’s some correlation of class membership with meaning, but a far from complete one.)
TR
Left and right parallel columns don’t work for me. I prefer to break the text in sentences and generate translation below.
Like this
And then I listen to the audio, usually at x1.5 or x2 speed. This helps to achieve fluent reading skill as fast as possible.
http://i.imgur.com/vhgGlwZ.png
Re: Kusaal
These days learning Kusaal is incredibly easy.
Just read text with translation
http://i.imgur.com/R6PNLr8.png
and listen to the audio
http://www.bible.is/ENGESV/Matt/2/D
right link http://www.bible.is/KUSTBL/Matt/2/D
Just listen to the audio and read the text with translation
http://i.imgur.com/R6PNLr8.png
Is tone marked in the Kusaal orthography?
Aside from Kusaal, good place for armchair language tourism. The Hebrew version is funny: the narrator and the locals speak Ashkenazi Israeli Hebrew, the Eastern Magi speak with a Mizrahi (Near Eastern) accent, properly pronouncing their ħ and ʕ.
By the way, king Herod is referred in Kusaal translation as na’ab Herod.
Presumably this is a borrowing from Arabic, same as English “nabob”.
Presumably this is a borrowing from Arabic, same as English “nabob”.
But na’ab / nabob does not mean king in Arabic. The word for king is ملك , m-l-k, cognate of Hebrew מלך melekh.
Per AHD, English nabob is from “Hindi nawāb, nabāb, from Arabic nuwwāb, pl. of nā’ib, deputy, active participle of nāba, to represent.”
Tone isn’t marked in Kusaal standard orthography (which in practice means the Bible translation.) It is distinctive, though there aren’t all that many single words distinguished by tone alone; in terms of comprehensibility in speaking it’s more important to get the many syntactically determined changes of tone correct. The system is basically the area-typical terracing type with two levels and emic downsteps, but it interacts with stress to produce extra-high realisations of H! which confused the only previous person to look at the system into thinking it was a three-level system.
Well, you asked.
Na’ab(a) “chief, king” is definitely not borrowed from Arabic. The -b(a) is a class suffix, segmentally identical with the usual human-class plural as in nidib(a) “people.” It turns up on a few other words for high-status people. The word is found in all the closely related languages, eg Moore: Moog Naaba = King of the Mossi.
The brackets in na’ab(a) are because word-final short vowels are normally deleted unless the word is the last in a negated phrase or question. The vowels which resurface are not predictable from the usual word form in isolation and mostly match the final short vowels seen in the cognate forms in related languages.
The Kusaasi are still overwhelmingly adherents of their traditional religious worldview, though there are (as in all the languages of that zone) some Arabic loans, often mediated through Hausa, the major source of identifiable (to me, anyway) loanwords. For example “arazana” “heaven” and the very common “alaafi” “health.”
… also a word that you will hear shouted after you everywhere by small boys if you have a complexion like mine
Nasaara “European”
Nasaaranam all speak the Nasaal “Europeanish.” I had a doubletake on the Burkina side of the border when I first heard it used to mean “French.”
SF, you are a genius if you can learn a language that way.
It only looks hard, but actually it’s great fun and takes much less effort than any other method.
Rapid reading skills in English are also needed, otherwise you will not be able to read the text and translation simultaneously.
– Nasaara “European”
I wonder how they call African Christians
@SFReader
You’re right that Nasaara is ultimately from the Arabic plural for “Christians.”
As actual Christianity is a very recent arrival in that part of the world (annexed by the British only in 1904, and not really visited by missionaries until times within living memory) there is no old traditional word for “Christian.” The Bible translation uses “Christ nidib” ie “people of Christ.”
(The Kusaal Bible translation also uses the word for “hail” in place of “snow” and has to resort to paraphrase for “island.”)
The various West African words for “European” are interesting. “Nasaara” is quite widespread in that area. Hausa has “Bature” which is supposed to be ultimately from “Turan” via “not-Iranian” through “not-part-of-the-regular-Islamic-world” to “foreign outsider.” Twi has “obroni” which (I think, I know very little Twi) is basically “person from the bush” ie from outside the settled civilized world. The more northerly Sahel languages often have “tubab” which looks vaguely Arabic, but I know no more …
Wikipedia:
Some of those are more convincing than others. (“Two bob”? Seriously?)
Baturki is a name used by Ugandans in late 19th century to refer to Egyptians who were advancing from the north under command of Colonel Gordon.
Central Alaskan Yupik term for white man is “Kass’aq”.
Since there are no Cossacks around after 1867, the term is now applied to all white Americans.
@SFReader:
Baturki would presumably be just Turki (ie Ottoman) with a Bantu ba- plural prefix.
The Hausa ba- in “Bature” is a singular prefix which just happens to look similar (the plural is Turawa.)
@Hat:
Should have thought of the Wkipedia …
I share your incredulity re “two bob.” Quite apart from the instrinsic silliness, the word seems vastly commoner in the old French bits of West Africa than the anglophone.
The ‘pedia links for “Toubab” and (even more so) “Oyinbo” seem remarkably ill-informed, not least in having little idea of the linguistic complexity of Africa.
“In Ghana the word used for a ‘white’ person or foreigner is ‘Obroni’ in the local languages, those of the Akan family.”
Ugh. There are dozens of languages in Ghana, only a minority of them Akan even in the broadest sense.
Ignorance on these matters is sadly not confined to Europeans. Unfortunately, many southern Ghanaians are under the misapprehension that practically everyone in Ghana outside Accra speaks Twi/Fante/Akuapem too.
In French le toubib is a slangish word for ‘doctor’. I assumed it was from North African Arabic.
Mechanically repeated across sites is this:
“The most likely earliest derivation is from the Wolof word for Europe “Tougal”. In the same way that Wolof means the people of Jollof, Toubab means the people of Tougal (Europe).”
Eh?
Looking at an actual resource
http://www.linguistics.ucla.edu/publications/OPL_19.pdf
… I find that Tugël truly is Wolof for “France”; the same dictionary reckons that Wolof “tubaab” is from the Arabic tabiib “doctor.”
It lists the “convert” word too, incidentally: tubbi “to leave Islam and convert to another religion.” As far as I can tell there’s no likely way to derive tubaab from tubbi within Wolof, even if you overlook the semantic implausibility.
I suppose word-play could be a possible origin of tubaab from Tugël and tabiib along the lines of “European know-it-all”: West Africans are no strangers to linguistic inventiveness and language jokes.
For SFReader: apparently “turki” means “shirt” in Wolof.
Another piece of the puzzle: Wolof “Tugël”/”Tugal” as the name for “France”/”Europe” is itself a loanword, from “(Por)tugal”: the Portuguese having been the first Europeans to explore coastal West Africa, the name of their country became a generic term to designate “Europe”.
The -gal reminded me of Fulfulde, so I searched and found
http://files.eric.ed.gov/fulltext/ED403752.pdf
which muddies the waters further by giving
tuubaako (pl tuubaakoobe, where -be is a human-plural ending; should be implosive b) as both “Christian” and “European”
and also lists the verb tuubude “repent”; there is also a Fulfulde verb tuuba “repent”
A lot of this must ultimately go back to Maghribene Arabic. We need Lameen Souag …
tuubude is the infinitive while tuuba is the present tense. This is indeed Arabic: tuub- is the imperative/imperfect stem of taaba ‘he repented’.
@George Gibbard:
That’s illuminating – thanks.
Although the meaning might have mutated from “repent” to “convert” in some key West African language (and there isn’t a lot of semantic distance between) it seems unlikely that Muslims would coin a term for Christians based on a verb primarily meaning “repent.” So if Arabic tuub- underlies any word for “Christian” it would be more likely a Christian self-designation, and so (I would imagine) a good bit less likely to end up as a Muslim West African designation for Europeans.
Personally I incline to the smart-alec foreign doctor theory.
I suppose the idea of tubaab being linked to “convert” is that the word would mean “converter” rather than “convertee”; but this seems if anything even less plausible, not only on linguistic grounds but also because it’s frankly (alas) incredible that Africans’ most striking impression of Europeans has ever been that they were mostly interested in Christian evangelization.
I must admit it’s pretty much equally implausible that Africans have ever thought of the archetypal European as a medical doctor. I think this is indeed the core sense of ʈabiib; is there any evidence the word has anywhere performed the opposite trick to Latin “doctor” and changed from “physician” to “educated person”?
Depending on what “need” means.
I (practically) stopped learning Russian 15 years ago. I tend to recognize words or grammar when I encounter them, but I tend not to recall them when I wonder how to say something in Russian. It’s all still there, but not easily accessible. I can, with an effort, read scientific papers in Russian, but my last two attempts at conversations (years ago, and years apart) were just pathetic – even though anxiety was not a factor.
SFReader: Where do you find translations with the two languages interleaved like that?
I was wondering the same thing about SFReader’s method – for how many languages is it easy to find materials with interleaved translations and audio versions of the same texts? (I’d like to give it a try though – finding an audio version of the Genji monogatari shouldn’t be too hard, and if anyone can point me to an interleaved translation, the idea of taking just one weekend to finally become a fluent reader of Heian Japanese is very attractive!)
Mostly, I make interleaved translations myself.
First I break text into sentences (using Replace function in Word to add paragraph mark after every dot, question or exclamation mark).
Then I run the broken text through Google translate. Put both texts in parallel columns. Select first column and paint it red. Then convert table into text with formatting preserved. And save as PDF.
Obviously Google translate is not available for every language (and for some it is unreliable to the point of being useless).
Then I simply have to find structured text broken into sentences with available translation.
Mostly this happens to be Bible. And then I run the same process skipping the Google Translate part.
For some languages, adding machine transliteration might be necessary. They are not always accurate, but if you have audio available, then it’s no big problem.
-for how many languages is it easy to find materials with interleaved translations and audio versions of the same texts?
Nobody does interleaved translations, you’ll have to make them yourself as I said.
But Bible translations and audio books are available for free in many languages.
Bible.is site offers free audio Bible translations in 1076 languages. And there are other websites out there with even more languages.
-Depending on what “need” means.
For me that means that when I want to study the language again, the previous effort does not disappear completely and you start very close from the point you left last time, not from the scratch.
After I realized this, I became very relaxed about learning languages. There is no deadline, the bits you learned are not going anywhere, so just learn language at leisurely pace wherever you have a mood for it.
That’s what happened to my Chinese. I learned to read Chinese characters and then I forgot them and this happened several times over the past ten years. It really doesn’t matter, if there is a need, I know that I can restore this skill within a few days of study.
@SFReader
When I tried converting the table to text, I get the entire L1 text, followed by the L2 text, rather than interleaved sentences. I don’t see any “formatting preserved” option, but maybe that’s because I’m using a different version of Word. Do you copy and paste the sentences by hand?
From table, click Layout, click Convert to text, select Separate text with paragraph marks.
Works every time (though might take somewhat long with very large text)
Strange, doesn’t work for me. Instead of:
L1.
L2.
L1.
L2.
I get:
L1. L1.
L2. L2.
Looks like you are selecting Separate text with Tabs option (it’s default option). Select Separate text with Paragraph marks instead (it’s the option above Tabs) and click OK.
Ah, I found the problem. I pasted both texts in a 2×1 table rather than creating a new row for each sentence pair. I can’t wait to start using it; thanks!
-Would Hungarian do?
I studied Hungarian a few years ago. I didn’t find the language too difficult, though I haven’t reached fluency at the time.
Perhaps it’s a good time to start again.
Hungarian doesn’t translate well with Google Translate, but there are some pretty good interleaved translation available for this language (to Russian).
Makes learning very easy, have a look.
http://i.imgur.com/YXFbtbv.png
The idea is not new and linguists have knew for a very long time how to teach languages effortlessly.
Perhaps there is some hidden conspiracy in action to make poor students suffer…
Here is a good example, from New Arabic Grammar published back in 1965.
http://i.imgur.com/zabO9qn.png
Several hundred pages of such text – readable within couple of days – would easily provide student with vocabulary which he would have spent years to acquire by any other method.
While we’re on the subject, here’s a Latin edition of Caesar’s Gallic War, with interlinear translation in English: http://www.gettingstartedwithlatin.com/docs/Caesar_Interlinear.pdf
(There are also various books by Franz Boas on archive.org that include stories in Native American languages with interlinear translation, too. I think searching his name there would be enough.)
Thanks, very nice. Though I didn’t like this “The original text reduced to natural English order” warning.
It’s not authentic Caesar!
Here is another website offering interlinear text (with transliteration) for entire Old Testament in Hebrew and New Testament in Greek.
http://biblehub.com/interlinear/genesis/1.htm
Depending on how good the student’s memory is. “A brain like a colander” is a common saying in Austria. From reading several hundred pages once, I’d just get overload and fail to remember most of the words; reading them more than once would take longer, because you can’t read an interlinear translation very fast if you’re actually trying to figure out the grammar (including word order).
Number of words used in any text is quite limited. Several hundred pages (say, about 100 thousand words in total) would use up at most 4000-5000 unique words, hence, most of these words would be encountered several times (most common ones even hundreds of times).
You don’t need to read anything twice. If a word is important, you will encounter it again and again and again in a variety of contexts, until you stop forgetting.
That’s how memory works – continuous repetition in different contexts.
As for overload, obviously there is no need to torture yourself (unless you have an exam tomorrow morning).
Reading two or three hours per day is more than sufficient for most purposes. I assure you, reading a novel in a new language in two or three days is an excellent achievement by any standard, no need to try to accelerate it to superhuman speeds.
The Caesar version linked above gives the Latin and interlinear English in what the publisher calls “natural”, meaning English, order. This produces grammatical but not very idiomatic Latin.
Original, i 1-4:
Latin from the 1893 interlinear, diacritics omitted:
1869 English translation:
1865 French translation:
And finally a translation into Interlingua de IALA:
Sam Ogilvy: There are also various books by Franz Boas on archive.org that include stories in Native American languages with interlinear translation, too
Indeed, and I have studied some of them. Too bad there were no voice recordings in his time.
But the translations are not always very reliable. The “free” English versions are rather flat-footed and contain some errors, and the interlinear ones are even more full of errors: the original native translators, often speaking in a lingua franca (such as Chinook Jargon) rather than English, gave Boas an idea of the general meaning of sentences, while he himself tried (not always successfully) to match the actual words to the alleged meaning.
Oops, in the interlinear (which I had to retype) for gellum read bellum, and chalk it up to scribal contamination from gerere.
“All Gaul is quartered into three halves.” —Anon.
“Man is an animal who suckles his young.”
“Omnia Gallia e devisa en parte trei, quaro una encolont Belge, alia…” – Martin Padway
Incidentally… does anyone know where did the “Omnia Gallia in tres partes divisa est” version come from? (Could it have been some similar interlinear?)
[It must have certainly been fairly early: “’Then,’ says I, ‘we’ll export canned music to the Latins; but I’m mindful of Mr. Julius Cæsar’s account of ’em where he says: “Omnia Gallia in tres partes divisa est”; which is the same as to say, “We will need all of our gall in devising means to tree them parties.”‘“ – from Kings and Cabbages, O’Henry]
[Or, in Russian, “Умного галла в три партии не обставишь – вот мой девиз!” Can’t recall how well it fits the context, but it certainly fits the original Latin lettering much better.]
[Update: here’s the full Russian quotation… “— В таком случае, — говорю я, — будем экспортировать латинцам музыкальные консервы. Но я вспоминаю, что мистер Юлий Цезарь в своем отчете о них сказал: «Omnia Gallia in tres partes divisa est», что означает: «Умного галла в три партии не обставишь – вот мой девиз».”]
I suspect that people felt the urge to adapt it to a more “unremarkable” Latin word order with the verb at the end. And modern languages almost universally seem to render the wording as “All Gaul is divided” rather than “Gaul is all divided”, abandoning whatever nuance Caesar may have been going for by placing “all” where he did. (And those who use “omnia” are simply not well versed in Latin declension and naively follow the analogy of “Gallia”.)
Also note that this phrase wasn’t written as a whole sentence, but it’s often quoted as if it were; it might be that Caesar’s choice of word order, with “partes tres” at the end, is specifically intended to set up the list of regions that follows. If so, then this would be one of many famous quotations which have been adapted to sound better in isolation, like “Luke, I am your father” or “Do you feel lucky, punk?”.
Ok, my challenge to SFReader is to learn Somali using his method
While I’m not an expert on Latin prosody and may be getting it wrong a bit, I think that July’s version simply flows better than any of the ‘fixes’. I assume it was written to be read aloud, not silently.
“A brain like a colander” is a common saying in Austria.
This corresponds to “a mind like a sieve.”
My Latin teacher explained it as meaning “Gaul as a whole” and said it could be translated “Greater Gaul”, as opposed to Gaul proper, which was just one of the three parts along with Aquitaine and Belgium.
Somali seems pretty easy language. No new script to learn, for starters. And its Afro-Asiatic language like Arabic or Hebrew, so grammar is familiar. If started today, I could probably finish reading entire book in Somali within a week.
Like this
2:1—-Goortii–when–Ciise–Jesus–ku–to–dhashay–birth–Beytlaxamtii–Bethlehem–Yahuudiya,–Judea,–waagii–days–Herodos–Herod–boqorkii–king–ahaa,–of–waxaa–it–Yeruusaalem–Jerusalem–bari–east–kaga–from–yimid–from–niman–men–xigmad–wisdom–leh–with–iyagoo–they–leh,–saying,–2:2–2:–2–Xaggee–Where–buu–and–joogaa–is–boqorka–king–Yuhuudda–Jewish–ee–of–dhashay?–born?–Waayo,–For,–bari–east–ayaannu–we–xiddigtiisii–star–ka–from–aragnay,–seen,–oo–and–waxaannu–we–u–to–nimid–come–inaannu–to–caabudno.–worship.–2:3–2:–3–Goortuu–When–boqor–chief–Herodos–Herod–waxan–it–maqlay,–heard,–ayuu–he–welwelay,–troubled,–isaga–his–iyo–and–reer–the–Yeruusaalem–Jerusalem–oo–and–dhan.–all.–2:4–2:–4–Oo–The–intuu–while–isu–together–ururiyey–collected–wadaaddadii–priests–sare–up–iyo–and–culimadii–scribes–dadka–people–oo–and–dhan,–all–ayuu–he–weydiiyey–asked–meeshii–place–Masiixu–Christ–ku–to–dhalan–born–lahaa.–would.–
Heh. I searched on “будем экспортировать латинцам,” and Google asked, “Did you mean: ‘будем экспортировать латвийцам'” [‘we will export to the Latvians’]?
In fact there seem to have been at least three new unique scripts proposed for Somali during the 20th century, but they all lost out to Latin.
Kaddare alphabet looks like little dancing men code.
SFReader: pretty good result for Somali. But one won’t be able to really learn the language from texts in standard orthography, because firstly each vowel symbol can mean either of two vowels (in different vowel harmony classes), and also because accent is not marked (it can fall on either the last or the penultimate mora of a word, if we assume certain written suffixes are really clitics, and meanwhile it can also fail to appear at all, in the case of noun phrases consisting of a single noun in the subject case, or non-progressive verbs in non-subject-focused main clauses).
As to the idea that Somali grammar is familiar, being Afro-Asiatic:
To me, Somali grammar is quite exotic, and all that was familiar from Semitic when I started reading about Somali grammar, was:
1) pronominal things: verb agreement, to some extent pronouns (including clitics), also t ~ d as a feminine marker in articles/demonstratives/possessives;
and 2) verbal valence changers: -(a)m- ~ -an passive (cf. Arabic measure VII (ʔi)n-), -t ~ -ad- reflexive (cf. Arabic (ʔi)Ct-, ta-), causative -i(s)- (cf. Arabic (ʔa)- < Semitic *š- where the whole proto-Semitic prefix is lost in Arabic when there is an additional prefix).
In your glosses, you have missed that definite articles (some claim that these mark present vs. past tense) are suffixes, and that b(aa) and ay(aa) are focus particles for a noun phrase preceding the verb, while wax(aa) literally 'the thing (that)' is a focus particle for a noun phrase following the verb (and both kinds of focus particle can exist in a single sentence, meanwhile a main clause must contain a focus particle, if only w(aa) 'verb focus'). One will not easily deduce that the focus particles and a following subject clitic are written as one word (b-uu, ay-uu 'he'), and for spoken Somali it's worse: the b- of the focus particle will often disappear, leaving only a falling accent marking its presence. I could go on but I won't.
Syntax is entirely different from familiar Semitic languages, including preverbal clitics, including the applicative clitics like 'to', 'from' in the examples, which most of us would expect to find adjacent to the argument they govern, but which instead come before the verb, while the ordering of the complement argument is quite free. In fact I learned in graduate school that when there are several arguments to a verb, Somali allows a wide range of ambiguity as to which argument plays which semantic role.
So I'm wondering how quickly one would pick up such things without a crutch such as Saeed or Andrzejewski.
Somali oo dhan ‘all’ (at the end of the noun phrase) sounds like “all done” in Cockney.
I don’t get it about the accent. Every Russian word has an accent, which is usually not marked, so essentially it has to be learned by listening to people. But as far as I know, nobody yet complained that Russian couldn’t be learned from texts in standard orthography!
Anyway, I have Somali New Testament audiobook on my computer with text read by native speakers, so it won’t matter.
Articles and this focus particle thing are easy to adjust. Now it’s pretty readable actually. (blog platform here removes all text formatting, usually I put the original text in red)
2:1-2: 1-Goortii-when-Ciise ku-Jesus- dhashay- birth-Beytlaxamtii-Bethlehem-Yahuudiya,-Judea,-waagii-days-Herodos-Herod-boqorkii ahaa,-King,-waxaa Yeruusaalem-Jerusalem-bari kaga-east-yimid-from-niman-men-xigmad leh-wise-iyagoo leh,-saying,-2:2-2: 2-Xaggee buu-Where is he-joogaa-is-boqorka-king-Yuhuudda ee-Jewish-dhashay?-born?-Waayo,-For,-bari-east-ayaannu-we-xiddigtiisii ka-star- aragnay,- seen,-oo-and-waxaannu u-we-nimid-come-inaannu-to-caabudno.-worship.-2:3-2: 3-Goortuu-When-boqor-chief-Herodos -Herod-Waxan maqlay,-I heard,-ayuu-he-welwelay,-troubled,-isaga-his-iyo-and-reer-the-Yeruusaalem-Jerusalem-oo-and-dhan.-all.-2:4-2: 4-Oo-The-intuu-while-isu-together-ururiyey-collected-wadaaddadii-priests-sare-up-iyo-and-culimadii-scribes-dadka-people-oo-and-dhan,-all-ayuu-he-weydiiyey-asked-meeshii-place-Masiixu ku-Christ- dhalan- born-lahaa.-would.-2:5-2: 5-Oo waxay ku-And- yidhaahdeen,- said,-Beytlaxamtii-Bethlehem-Yahuudiya,-Judea,-waayo,-for-sidaasaa-so-nebigii u-the prophet-qoray,-wrote,
Of course Russian can’t be learned (properly) from texts in standard orthography. That’s why all teaching texts have accent marks.
You two seem to be talking at cross purposes. Rodger, you seem to be talking as if SFReader was saying “All I need is a few simple texts, and in a week I will be speaking the language fluently,” when I understand him to be actually saying “All I need is a few simple texts, and in a week I will have a decent start on learning the language [though of course much more, including exposure to actual speakers, will be necessary to really know it].”
And of course people mean different things by “knowing” a language; for some purposes pronunciation is irrelevant (and for some languages, like Ancient Egyptian, it’s impossible to know).
It gets even better. With text glossed like this and aided by audio, you can start reading texts in a language you don’t know immediately.
Audio also speeds reading immensely. Somali Audio New Testament (big novel sized text) is 22 hours long. Can be easily consumed in less than a week (reading the glossed text while listening to audio, say, for 3-4 hours a day).
Before advent of Google Translate, I used parallel texts (aligned verse by verse) for this purpose. It also works, but you’ll get new vocabulary memorized much faster if it’s glossed word by word.
As for results, usually, a book of this size read in such manner gets you more or less fluent reading ability with passive vocabulary of about 5 thousand words.
That’s pretty good result for a week of work. Using traditional methods you’d get to this point after several years of study (or never, if you are learning language in school).
However, I should warn you that this method hits the diminishing returns very quickly.
The second, third, fourth, etc books read in this method will improve your reading fluency and vocabulary only slightly. (I think I told elsewhere about my experiment with French – over 20 French novels read in a month and a half. My French has improved greatly. Before, I could understand about 98% of French text, now I can understand some 99% or maybe even 99.5%. Not bad for almost two months of effort…)
Of course, the method doesn’t work with scripts other Roman, Cyrillic or Greek.
I can read devanagari or arabic, but very slowly, at such speeds this method is impractical. So you need to get the text transliterated. And transliteration needs to be readable (for example, currently available Burmese machine transliteration is awful. I am refusing to learn Burmese until they fix it!)
Of course, it might sound strange to talk of reading fluency if it’s reading fluency of transliterated text.
I have such “Pinyin fluency” in Chinese. I can read Pinyin annotated Chinese texts easily (without English translation), but I need access to my computer first and I’d fail miserably if I am asked to read Chinese on paper (embarassingly, this happened to me a few times). This is because of unique feature of Chinese characters – they are apparently designed to be forgotten if you don’t read Chinese regularly – each and every day! I think I acquired and lost ability to read fluently in Chinese characters two or three times. Now I don’t even bother.
Having large passive vocabulary also helps to understand spoken Chinese too (I can follow TV news in Mandarin easily, but people on the street are different matter)
I don’t know how would you call such level of language mastery. It satisfies me for now. And frankly, it’s the best I can practically hope to achieve without moving to China or getting full-time job requiring reading Chinese everyday.
If it works for you, then it works for you; everybody has different needs. I read Russian fluently but my speaking ability is embarrassingly bad, because I never get a chance to practice. Obviously I could change that if it was important to me to be able to speak it, but at the moment it isn’t. And when I was studying Sanskrit I avoided devanagari as much as possible, since I only wanted to know how the language worked, I didn’t care about reading the literature in the original.
At times I get a wanderlust for Japanese, but I’m always deterred by the kanji (and I have little interest in gaining a spoken-only proficiency). I keep reminding myself to focus on the handful of languages in which I’m already half-proficient, but sometimes it’s hard to stay focused on them.
Somali u, ku, ka, which you’ve grouped with the preceding word, are actually the preverbal applicative clitics that do the work of prepositions. I think you thought they were the articles I mentioned, but those are written as part of the preceding word: ‘the’ is -ka/-ga (masculine), -ta/-da (feminine),-ku/-gu/-tu/-du (the same in subject case), -kii/-gii/-tii/-dii (with what is claimed to be past tense marking, in either case). xiddig-tiis-ii is ‘his star’, with a possessive suffix and then an article without the -t-.
I read Russian fluently but my speaking ability is embarrassingly bad, because I never get a chance to practice. Obviously I could change that if it was important to me to be able to speak it, but at the moment it isn’t.
This was basically my English up to about 2007. I could read no problem through all sort of Wikipedia and blogs and even interesting books, but my written text was awful, and don’t even ask about speaking!
Since 2009 or so, I’ve tried to practice in forum comments, and, less significantly, also in actual talking (one of our university department’s professors comes from England).
I withdraw my hasty comment.
Revised statement on unaccented words in Somali: I said that the accent fails to show up “in the case of noun phrases consisting of a single noun in the subject case”. But first, this needs to be interpreted in light of the previous statement “if we assume certain written suffixes are really clitics”, which will include articles, demonstratives and possessives. So I was thinking of the case of an NP consisting of a single noun without any of those, therefore an indefinite single noun. But there are other instances of unaccented words in noun phrases, the correct generalization (subsuming the previous one) being: now treating the articles, demonstratives and possessives as separate words, don’t put an accent on the last word of a noun phrase in the subject case. The last word can be any number of things since noun phrases have the noun at the beginning. (Subject case also sometimes has a suffix or other word-final segmental change marking it, but only on the last word of the NP, and what that suffix or change is depends on what that last word is. I find this nifty.) But also, -(C)a, the non-subject-case-marked non-past definite article, if we are treating it as a separate word, now has to also be mentioned as a word in an NP that has no accent.
As to said articles/demonstratives/possessives, I neglected to mention, for the masculine, allophones with -h- rather than -k- or -g-: this happens after a non-high vowel (which then changes to agree with the next vowel, even becoming high: madaxweyne > madaxweyni-hii ‘the president’ (in the past tense) : guri > guri-gii ‘the house’ (in the past tense). Also for the masculine articles/demonstratives/possessives, the consonant is lost after certain consonants, maybe all gutturals: wax-a (wix-ii) ‘the thing’ is one example.
Meanwhile preverbal clitics are written separate from the verb, but sometimes more than one preverbal clitic gets written as a single word, as in u + ku > ugu (note the consonant lenition, which by contrast doesn’t apply in the root of a verb after a preverbal clitic). I remember that there are some accent rules for preverbal clitics which I don’t remember. SFReader may know by now what they are.
To summarize, Somali clitic articles/demonstratives/possessives have to count as separate words for purposes of accent deletion in subject case marking (nín-ka, past tense nín-kíi ‘the man’ > subject-case-marked nín-ku, nín-kii), but have to not count as separate words for purposes of the phonology of their initial consonants, and also in that -(C)a, the non-subject case marked, non-past definite article, doesn’t have an accent.
Man, Somali sounds like an interesting language — I may have to check it out myself.
Back when this thread was still fresh, I wrote a little program to automatically align corresponding sentences between source and translation texts.
Here is an example of the output. Embarassingly, it gets the first sentence wrong, but then it recovers and only makes a few more mistakes.
I never got around to trying out the SFReader method, either because I was too lazy, or I couldn’t find any decent, free audiobooks with corresponding free texts (in Spanish or Russian, the two languages I am perpetually trying to learn) or possibly both.
The code is here, at the moment it’s horrifyingly un-user-friendly. If people are interested, I might be able to make a web interface for it, but I’m not sure how useful it is with it’s current error rate.
In my experience, longer sentences are difficult to follow. It’s better to split them (at each punctuation mark or even some joining words like ‘and’ or ‘that’)
Like this.
http://i.imgur.com/edGj5rQ.png
Of course, no computer program will be able to align two texts like that, so I recommend Google Translate. For European languages it is good enough.
And also it’s better to pick some easy and entertaining reading, which won’t feel like a chore.
Get a book which you would normally read in a day in your native language and expect to read it over several days using this method.
In my experience, reading Dostoyevsky novel takes weeks in any language (and reading him in the original Russian would probably take you months).
He is that boring…
If available, books of short stories (by one or several authors) are a good choice: the texts are indeed short, and unlike with chapters in a novel, you get to some form of closure at the end of each. If you get bored before finishing one story, you can always switch to another one without feeling guilty that you could not get through a whole novel. Another possibility is to find a translation of a work you already know and like in your own language.
Also if available, general interest magazines are also good choices. They deal with a variety of topics, some of which are bound to be already interesting for you and make it easy to guess the meaning of some key words. If they mention internationally known people, you have heard of some of them and of their lives and situations, something which again helps in guessing.
I agree with the person above who said not to worry about learning words: if they are very common, they will recur and the context will gradually reveal their meaning. I used to tell students not to look up a word in the dictionary unless they had seen it at least six times and they still could not figure out its meaning.
About audio: With a language I don’t know well I try to read the material aloud, and I still do so in Spanish even though I am quite fluent orally. I became so partly through this method as well as knowing a group of Latin Americans. One advantage of reading a work aloud is that you are pronouncing words and sentences that you would not have the opportunity (or the knowledge yet) of using in the course of interactions with native speakers. This includes for instance verb forms in their variety (I deliberately did not set out to memorize Spanish conjugations).
My spoken Italian is minimal but I can read Italian fairly well (relying on French, Spanish and Latin) but since the spelling rarely indicates stress I am often frustrated by having to try two or three alternate stress patterns, without being able to decide which is the correct one. I took Russian some years ago, I know the letters, but the unpredictable stress is often a problem. My vocabulary is very limited and I can’t rely on related languages as I can in Italian. I would very much like to find audio versions of texts in Italian and Russian.
I would very much like to find audio versions of texts in Italian and Russian.
Thanks to the internet, that should be no problem. Here are some sites that offer Russian audio books, also for download. I haven’t tried any of them, so click with due caution. 😉
Thank you Hans! I tried to find some a few years ago, but things must have improved since then.
For Italian, try
http://www.libroaudio.it/libri-audio/i-promessi-sposi/
big-sized classical Italian novel. audio runs 24 hours.
But Italian of the novel might resemble the Illiad a bit….
I agree with the person above who said not to worry about learning words: if they are very common, they will recur and the context will gradually reveal their meaning. I used to tell students not to look up a word in the dictionary unless they had seen it at least six times and they still could not figure out its meaning.
That’s good advice for most people, but I love looking words up — I enjoyed doing so as a kid as soon as I discovered dictionaries, and I have not lost the taste for it. Which is good, because I’m a fan of Russian:
I took Russian some years ago, I know the letters, but the unpredictable stress is often a problem.
That’s why I not only look up words as soon as I encounter them, but mutter the various forms aloud (if the stress shifts, as it does in so many words) and frequently half-vocalize the texts that I’m reading. I’m encouraged in this habit by the knowledge that Russians find wrong stresses one of the most common and easily mockable errors made by foreign speakers.
LH, yes, reading a dictionary can be a delightful and rewarding occupation, but most of the discussion in this thread is about how to read and understand running texts. I personally find that constantly interrupting my reading to look up words goes against my “learning style”.
Russians find wrong stresses one of the most common and easily mockable errors made by foreign speakers.
Similarly, English speakers find wrong stresses one of the most common and easily mockable errors made by French speakers, who tend to either ignore stress (and vowel length) altogether or to place stress haphazardly. Apparently I rarely make such errors, causing many strangers to think I am German or Dutch. But my vowels and consonants give me away as a non-native speaker.
Of course, English speakers make opposite errors in trying to speak French, introducing stress on some vowels and consequently distorting the unstressed vowels.
@Trond: flotation and submersion, you mean?
It seems so. I obviously didn’t heed the advice in the post title.