« Bit of Madness: Likely Upcoming Silence | On disasters and meetings »


October 09, 2005

On Arabic II: Training, Translation & Intelligence

I nipped by "Liberals Against Terrorism" (an atrocious name I may add that never ceases to irritate me) and found Pratike commenting on Arabic again, on indeed financing of Arabic studies by the United States.

That incited me to comment.

First on the financing issues, given what I saw when learning Arabic in the dark ages there certainly could be (and here I refer to the Anglo world not being in any way conversant with actual teaching materials elsewhere - except in terms of in MENA region, which are regardless of language (including Arabic), risible) better finaning of efforts to develop better pedagogical materials - preprepared texts, targetted vocabularies and all the sorts of things I recall from German (although this was wasted on me, after I decided I loathed German) and French (although in this case I was young and impressionable).

Second, of course, in terms of instruction, it takes time to build up a base of instructors - the native base is as poor as the non-native base in my opinion, with extremely primative pedagogy (nothing gets me gnashing in anger more than recalling my forced march through great gobs of Arabic literature, which I loathed, and without prepared texts).

Moving on, let me assure Pratike and his commentor that Arabic is not the hardest langauge. Having had some exposure to Chinese I think we can reliably say Chinese is worlds away harder.

But Arabic is indeed a bloody toothache that takes a really bloody long time to get up to practical levels of usage, and often involves learning modes and grammar utterly irrelevant to the non-academic world.

This aside, the concept of hiring hordes of unemployed university grads to translate for the US government is a fucking crock.

First, having had professional exposure to the "English" of most Egyptian school leavers, I assure Pratike that one would need a retranslation into English for much of the product (think three steps down from Arab News website).

Second, I very much doubt hordes would really sign up for such work, and further to that, Pratike innocently assumes in a student/NGO manner that one could ipso facto trust the translations to be without editing or spin. For intelligence purposes, I would presume, one would have a concern with respect to that issue. Again, never assume the best - it gets one fucked over right fast in this part of the world my dear innocents.

Now, of course, covert financing to generic translation houses could very well work. Pretend indepedence - like MEMRI except without the explicit agitprop function. Indeed I rather like the idea as I reflect on it. Might even be workable, although without too much snubbing of the United States, I would suggest letting the British run the effort and handle the money. The US Agencies seem to have a problem in this area.

This aside, relying on the Egyptians would be a mistake. They're lumps. I'd sooner tap Jordanian talent, but Jordanians are more anti-American. Still their education levels are better, all around. Lebanese too of course, but Lebanon is less stable.

Regardless, the covert (not, not, not Overt) translation house would actually be a brilliant idea and there is indeed enough low level talent to hire that it could be rather economical. More so than mass training of Americans, which will take a decade to really ramp up.

Now, one more comment, to the underlying article or blog blithering that our man in Cairo originally commented on.

This paragraph struck me as stupid in part although I rather agree with the general sense that there is far too much ignorant and clueless yammering on by "pundits" (I rather despise that word as well, come to think of it) who read a MEMRI translation and think they're getting a real sense of the Arab world discourse, etc:


Our punditry would be much better if at least some of us could muddle through an Arabic op-ed or a Persian television broadcast; that would make for a better country. Give grants to low-income school districts to teach Asian languages to third graders. Hire 10 guys to sit in a Foggy Bottom basement, translate newspaper articles and post them on a website. Pay Google to expand their autotranslation program's capacities. Do something.

Grants to low income schools for hard languages seem silly to me. The return on your investment is likely to be negative (unless one has a concommittant hiring and assured posting program).

As for Google (or others) to expand machine translation, there are in fact such services. Sakhr, an Egyptian company has one such. Perhaps someone can alert the author he's wishing for what already exists.

However, while machine translation for say French or German can get you for basic texts a decent enough idea of what is going on (although I should say that a word of caution to the naive user, the renditions are piss-poor and just barely useful enough to get you a sense of what is going one in a very unnuanced manner. A very unnuanced manner. I raise this because my moron of a CEO actually sent out a letter to a high ranking official in the Maghreb which he apparently wrote in English and ran through a machine translator, thinking it worked just fine. To my office's collective horror when we saw the farcical piece of insulting crap. My director actually rang up New York asking them if they had actually faxed it.), for Arabic it is even less sufficient.

Now, I have used the Sakhr tool once in a while for my own purposes when I needed to render a text and was having a block on how to render things into English (and this is straight forward investment language: X produces Y), as often as not it was useful more for annoying me and unblocking my mind as for actually giving me a useful phrase. Blocks of text? Forget it.

No, afraid machine translation despite the hype is not yet here, not even for Indo-European langauges, never mind Arabic.

Now, I have to get back to the fun of trying to analyse the potential of carbon project credit generation in the MENA region.

Posted by The Lounsbury at October 9, 2005 10:28 PM
Filed Under: Perso , Politics - EU FP , Politics - Foreign Policy , Politics - US FP , Society & Culture

Trackback Pings

TrackBack URL for this entry:
http://www.aqoul.com/movabletype/mt-tb.cgi/2078


Comments

Chinese isn't really all that hard. Its grammar is rather simplistic. Characters, although there are tens of thousands of them, only a few thousand are in common use and they are surprisingly systematic. Pronunciation is the only really hard part about Chinese.

On the other hand, I always thought Russian was impossibly hard nut to crack. (I have no experience in Arabic or other Semitic languages, however.).

Posted by: kao_hsien_chih at October 10, 2005 05:08 AM

Well, I might say the same about Arabic.

The grammar, while complex for some literary applications, is quite regular and systematic. It might appear daunting at first look, but with a few exceptions of 'irregular' verbs (themselves following quite regular patterns), it's quite straight forward.

Sounds can be a bitch, but nothing like the Chinese tones.

The real problem comes down to (i) vocabulary, vocabulary, vocabulary, (ii) diglossias, (iii) vocabulary. The stuff that appears hard up front, writing system, learning the sounds; if you have a modicum of talent it's not that awful. Bloody vocabulary and regional differences....

Learning a few thousand signs (also vocabulary), a tonal system rather far off from IE sound, and so forth is hard.

In retrospect one can always say X was easy, but frankly that is retrospect.

Posted by: The Lounsbury at October 10, 2005 10:42 AM

You raise some good caveats that are worth considering. Perhaps you could have a trustworthy editor-type to look out for signs of bias or subtle manipulation, but otherwise I think it's a very workable idea. Maybe you could even do something like hire Copts to translate jihadi websites, so that you would be assured of having an unsympathetic audience.

All I need now is a hockey-stick graph and I'll start to work on my pitch ...

Posted by: praktike at October 10, 2005 01:05 PM

BTW, I also commented on your earlier remarks here. Didn't see them until now.

Posted by: praktike at October 10, 2005 01:30 PM

Spin is spin. Coptic anti Jihadi spin is every bit as bad as pro or sympathetic spin. One needs neutral, un-heated renditions.

A translation house openly funded will be a disaster, I assure you of that.

Covertly, well, that could work. Indeed it could be brilliant.

Posted by: The Lounsbury at October 10, 2005 01:38 PM

I note also that no matter who you use, one always has to have editorial review. That is a serious bottle neck in any type of operation. Given the low quality of most Egyptian's English, one will be faced with serious up front costs (and those with excellent English usually have better prospects).

Posted by: The Lounsbury at October 10, 2005 01:40 PM

My Arabic is minimal (only dabbled no formal study except a nonacademic class) and yes vocabulary is the deal. At some point, maybe soon, I will produce some comentary on the subject based on a W Post magazine article which showed the FOreign Service labelled Arabic among the harder languages and Hebrew a grade lower! The challenges are the same -- similar grammar and vocabularies, and a new script. Hebrew has fewer letter forms in writing but if you count the indecipherable modern cursive it is as new and alien as Arabic.

I think there are two main reasons why Arabic's difficulty, while real, is exaggerated, and will adddress that.

Also, I studied French for years, up to Baudelariean poetry, can read it and translate professionally but am as mystified as Chinese when it is spoken by metropolitans especially in film. I can pick up as much Arabic when spoken (well maybe not as much). I think that is partially due to the fact that my hearing is bad, and Arabic isnt shy on consonants while spoken French is basically a bunch of vowels with consonants sprinkled in now and then (kind of like African-American English).

Posted by: matthew hogan at October 10, 2005 03:13 PM

i've actually been offered a translation job before. i laughed so hard that the man finally got the idea it would be a bad idea for me to translate anything important. despite the two years of MSA in university (funded by the defence department, of all things) and a year of practical exposure, i feel woefully underskilled to even tackle a children's story book.

and they wanted me translating (mildly) important things.

maybe this is why bad intel keeps floating around the world.

Posted by: drdougfir at October 10, 2005 03:17 PM

The greatest irony must surely be that there are millions of people in the world who speak and write Arabic with native fluency... but unfortunately they're all Ay-rabs ;)

Seriously: if Dubya hadn't set it up so Them and Us, it wouldn't be such a problem.

Posted by: secretdubai at October 10, 2005 06:25 PM

My personal irritation has been (though not lately) at translators in the US media. I actually saw a bin-Laden speech translated with him saying "we will not put up with this humility" instead of humiliation. Or even Saddam's the "mother of all battles" literalness. Or not translating Allah as God, something that many hardcore Muslims also separately conspire to keep distinct by not translating it also. (It's funny that when GW Bush was recently (mis) quoted from an internal Arabic PAlestinian document as telling Pal leaders that God told him to do this or that, no one in the media translated it Allah even though I am sure that was the word used in the Arabic text.)

Anyway, oral translations of anything on the media should not have accents as the person is not speaking in a foreign accent, or in incoherent syntax, in his own language. And no translator should be given a microphone who cannot say the letter p.

Posted by: matthew hogan at October 10, 2005 07:55 PM

The US government has had a very efficient mechanism for translating Arabic and other foreign language media for the past several decades: the Foreign Broadcast Information Service (FBIS), to which non-government consumers can also subscribe. FBIS utilizes any number of qualified part- and full-time translators, living all over the world, who don't need security clearances. The only thing keeping them from translating even more of the Middle Eastern media is the amount of funding provided in a given year. No new mechanisms are needed, as the system is already up and running very effectively.

The government's real problem in terms of Arabic language capability is finding qualified people to do classified work, whether it's transcribing phone conversations or e-mails, interrogating suspect terrorists, etc. For this purpose, it's seen (rightly or wrongly) as being too difficult to clear many foreigners (and Americans with suspect foreign connections). In addition, such people generally have to work in secure facilities, usually located in big cities, which eliminates otherwise qualified people who aren't willing or able to move to Washington, Guantanamo, etc. It's in the classified area where the bottleneck lies, as FBIS already translates more of the Arabic press in a day than most analysts would be able to read.

Posted by: Anonymous at October 11, 2005 02:50 AM

Posted by: Alex at October 11, 2005 11:46 AM

I am sure there is an issue for the finding qualified people to do classified work.

A rather long while back I made use of the US FBIS materials for research/work and found them fairly decent. However unless practice has changed, I don't believe that FBIS covers much, if any, of non-traditional medias. Nor is it well diseminated to the public - although that is a seperate question than pratike's US Government centered idea.

In connexion with BBC monitoring, again the same issue with respect to coverage of materials and volume - although I would suggest that BBC is underused in American circles due to a odd phobia among the Bolshy Right over BBC.

As a general matter, a covert mass translation and public dissemination program might well be useful. Better, one rather suspects that what is being translated is not being reviewed and packaged well enough for people like Ms Hughes to review and get a sense of what Arab region discourse is really like.

In place of such efforts, you get MEMRI, which is a cleverly package intel op if nothing else. Slimey but very clever indeed.

Posted by: The Lounsbury at October 11, 2005 02:18 PM

You are right in terms of a shortage of coverage of non-traditional media, although in part this is a function of the proliferation of media of every sort, so that selection becomes key. In addition, there is an issue of covertly acqired information being given more weight than media reporting, so that there is an upper limit to what many analysts feel they can read in a given workday in terms of media reporting. Most actual policymakers, I am confident, spend next to no time reading the foreign media, unless something is specifically brought to their attention and marked as important.

As for what reaches or makes an impact on "power hitters" like Ms Hughes, I think the basic problem is more attitudinal rather than a lack of translators or translations. Having seen a bit of this world from the inside (but at a substantially lower level), I have no doubt but that she spends minimal, if any, time in her workday actually reading what foreigners - particularly Ay-rabs - have to say about the US - even though coming to terms with just this question should be a major part of her job. The "Washington-centrism" of such people is unbelievable, and there is near-total disdain for "lesser breeds without the law", except, of course, when one is on a "listening tour" to show how much one "loves children".

You may not be overly impressed with the sensitivity and/or local knowledge of the diplomats you meet "on the ground" where you are, but despite their shortcomings they are almost certainly more appreciative of local and regional realities, by several orders of magnitude, than their masters and mistresses in Washington. This is true in the best of times, but is greatly exacerbated by an administration like that of George the Younger, whose Secretary of State is a "Soviet expert" who never lived in the Soviet Union and who recently proved unable to respond correctly to simple questions in Russian on a radio interview.

These people are somehow programmed always to broadcast and never to receive, and it is not the lack of translation resources or capabilities that gives them such a tin ear to the realities of the Middle East or the broader Islamic world. Having read and appreciated your earlier blog for some time, I find it strange to say, but feel I must, that you simply aren't cynical enough...

Posted by: Anonymous at October 11, 2005 07:18 PM

Well, first, I should make a rule against signing as anonymous. It irritates me immensely. I could care less about identities but responding to Anonmymous rather rubs me the wrong way. Bloody well make up a motherfucking moniker you lazy twit. (Nothing personal of course)

Now, leaving that aside, this requires some commentary. Pity I am irritated, but nevertheless with regards to Sr. policy makers and reading original material, I would never expect them to. That is, barring a personal obsession or expertise, a waste of their time. They should not do so, there is too much to read and its the function of the bloody analysts to do the original screening.

Imperfect I am sure it is, but regardless, that is the only way these things can work. I am sure we agree on this.

Now then, with respect to the Washington (or other) centered-ness of Hughes level people, I am well aware of this. I've had the misfortune of having dinner with such people as Ms Cheney, who is the head of the US MENA program with the absurd name. An experience that gave me the desire to slit my own throat. Or hers. Same same really.

That being said, my comment should not be taken as mere naivete - in reality I do prefer that we all be realistic about real (which means immediate and substantive) incentives, not merely wishful thinking. As such I would not expect that most Sr. level people without experience in region will be equipped. The sole mitigator, such as it is, is reputational risk. A thin cover really.

However, in commenting abstractly, I did not, obviously, delve into my deep cynicism with respect to this. At the moment, as Hughes has not demonstrated her utter idiocy (although I suspect it, ceteris paribus, given this present American administration), I have tried not to wack away.

At the same time, realism as to the real structural incentives with respect to these kind of programs is necessary if one is to aspire to something moderately effective.

Posted by: The Lounsbury at October 11, 2005 08:57 PM

Comment Subscription

Email Address: