Beam Me Up

Lacking anything resembling an adequate number of Arabic-speaker, the military is apparently trying to build devices reminiscent of the old Universal Translator to fill the gap. It's rarely discussed, but you've got to think that the severe paucity of people who speak Arabic (to say nothing of Farsi, Pashto, etc.) is incredibly crippling to a whole range of military, diplomatic, intelligence, and law enforcement endeavors. Indeed, as best I can tell it's a fairly serious impediment to American journalism, and therefore to the general public understanding of what's even happening.

Clearly, it would take a long time for a massive investment in building skills in Middle Eastern languages to pay off, but all that goes to show is that we couldn't start such an effort soon enough. Indeed, had we really done it seriously in fall 2001, it might be paying off at least a little already.

Comments

In a country as diverse as this, it will usually be quicker and faster to use the skills of our first- and second-generation immigrants than to train the skills up from scratch, as good as places like Monterrey are.

Another reason not to race-bait or stigmatize the Arabs and Persians and Afghans and what have you who are already Americans...

Posted by Davis X. Machina | November 1, 2006 1:09 AM

In a country as diverse as this, it will usually be quicker and faster to use the skills of our first- and second-generation immigrants than to train the skills up from scratch, as good as places like Monterrey are.

But surely one good use of first- and second-generation immigrant skills would be teaching some of those skills to other people.

Posted by Matthew Yglesias | November 1, 2006 1:14 AM

The idea that we didn't go pedal to metal with Middle East language training on 9/12/06 is a shocking eye opener.

I hope you read the pieces claiming that hundreds (or more) of innocent Iraqi civilians have been shot and killed at vehicle checkpoints because the troops set up warning signs which don't clearly convey "stop."

Posted by gar | November 1, 2006 2:32 AM

Indeed, had we really done it seriously in fall 2001, it might be paying off at least a little already.

This is certainly true. On the other hand, Arabic is reportedly a difficult language to learn for folks with European-derived mother tongues. Now, this doesn't mean we shouldn't be putting effort into it of course, but it might point to why we have such a paucity of speakers even after making efforts to train folks in the language.

Posted by moonbiter | November 1, 2006 3:59 AM

Not that difficult to find Arabic speakers in Dearborn and Newark. Same goes for Farsi in LA. Probably time to start looking for Bengali speakers. While we are at it, perhaps now is the time to reexamine the clearance process since students who learned Arabic and other foreign languages while they studied abroad are often 'weeded out' of the applicant pool.

Posted by suggestions | November 1, 2006 4:31 AM

Not that difficult to find Arabic speakers in Dearborn and Newark.

More difficult for them to get security clearances. If you've never left your home state, you're more likely to get through vetting than if you've travelled abroad, let alone lived abroad. And having relatives in Scary Brown Foreign isn't going to help you get cleared: much better to be ninth-generation Kansan as far as the vetting rules are concerned.

And that's even before you address the different regional variants and local dialects. So what you end up with is English As She Is Spoke. And I must say that the website for the guy behind the Phraselator doesn't fill me with confidence.

Posted by pseudonymous in nc | November 1, 2006 5:18 AM

See, the problem with recruiting Arabic speakers is that some of them turn out to be gay--can't have that . . .

http://www.cbsnews.com/stories/2002/11/14/attack/main529418.shtml

Posted by rea | November 1, 2006 6:46 AM

A massive investment in Middle East language skills would a massive transfer of resources to Juan Cole and his friends. The organised groups who influence US ME policy do not favour it.

Posted by otto | November 1, 2006 7:05 AM

The military did work harder to get more Arabic speakers.

Getting a clearance is time consuming. Scoring high enough to take a Cat IV language like Arabic is difficult. Passing the intensive 63 week course is difficult. PLUS you get the joy of learning a dialect like Iraqi which is only sort of similiar to actual Arabic.

Getting native speakers to pass clearance is almost impossible. You could use them as a teacher, but you really need an educated person who CAN TEACH otherwise it is a waste of money.
Another problem is morale is low and hardly anyone stays in. So there is a tremendous turnover.

I should know, Arabic linguist is my job.

Posted by guachi | November 1, 2006 7:23 AM

Guachi -- having just seen "Iraq For Sale", I was under the impression that the extraordinary incompetence of the private contractor hired to supply translators in Iraq -- Titan -- was also a factor? Do you work for a private contractor, or the US gov't, or elsewhere -- I'd be curious to hear more.

And, definitely, Arabic is a hard language, but having learned Russian and Vietnamese, it's my impression that 3 years can get you pretty far, no? Not if you're trying to pass for native, but if you just need to do effective, idiomatically correct translation?

Posted by brooksfoe | November 1, 2006 8:17 AM

I am in the Navy. All the contractors I work with are very good. And all are ex-military linguists. They are far more dedicated than the surly military linguists I work with. I have not personally encountered any bad contract linguists.

The biggest problem with Arabic is that the language they teach us at school isn't what Iraqis speak. We are taught MSA (Modern Standard Arabic) and it is astonishing how much easier MSA is than Iraqi. Broadcasters and politicians use MSA. Almost no one else does.

For example - barha is yesterday in Iraqi. In MSA it is ams.

Imagine being taught Shakespearean English with a British accent and having to listen to ebonics and hip-hop music. It's technically the same "language" but only barely.

Posted by guachi | November 1, 2006 8:46 AM

"But surely one good use of first- and second-generation immigrant skills would be teaching some of those skills to other people."

And...? Your "one good use" in no way precludes the more immediate use suggested by Davis. We have plenty of Arabic speakers. Some can qualify for security clearance more readily than others. Some are better suited to teach than others, as Guachi notes. Given the resources we are pouring into the Iraq war and the intense security focus on all things Arabic, this should not come down to a question of financial resources. One month's spending on Iraq would go a long way toward hiring us some linguistic proficiency.

We also need to consider the results, in terms of linguistic proficiency, of our efforts. We are going to hand a paycheck to the designated Arabic speaker, no matter who that Arabic speaker is. We can have native proficiency, or we can have 63-week proficiency. Same paycheck. One is ready to go right now. One is ready to go in 63 weeks. One learned at a time when language acquisition was easy, the other when it's hard. Brooksfoe's 3-year assessment is not all that reassuring.

Time to hire some native speakers not just to teach, but to do the job. The added advantage is that we might get a higher level of exposure for some of the less-well-represented segments of our own population.

Posted by kharris | November 1, 2006 8:46 AM

it's never too late for this; now is the time to start.

Arabic is a very hard language, but young-ish minds (e.g. late teens early twenties) can still pick up languages in fairly short order.

I mean, if we'd started in 2001, the pay-off would have come in 2002, 2003 at the latest, not only now, five almost six years later.

Further evidence that the Bush regime was never serious about terrorism.

Posted by kid bitzer | November 1, 2006 9:29 AM

A massive investment in Middle East language skills would a massive transfer of resources to Juan Cole and his friends. The organised groups who influence US ME policy do not favour it.

This, I think, is the real problem. Powerful actors would do just about anything to prevent big federal dollars from flowing to Mideast Studies departments.

Posted by Matthew Yglesias | November 1, 2006 9:41 AM

If we had started, seriously, in January 2002, we'd have had five-year people rolling out of the program as of January, 2006; four-year people in January, 2005, etc.

When the Air Force is trying to build a supersonic stealth fighter, which is hard, they take 'hard' to mean 'spend more money, with a long lead time'.

Posted by Barry | November 1, 2006 9:53 AM

Well, if you relax clearance requirements and citizenship requirements, the exodus of Iraqi professionals is likely to provide enough to get things going, by which time the priorities may have switched to Persian, for which there is an older 'exile' community with second-gen speakers. And it's an easier language to learn, which is a bonus.

The hoo-hah over the guy from State who gives interviews on al-Jazeera was sad: as Marc Lynch pointed out, he's sufficiently fluent to master not just the content but the context.

Posted by pseudonymous in nc | November 1, 2006 10:30 AM

Gadgets bring in money for The Syndicate, translators don't. I'm waiting for Lockheed Martin and the Air Force to start talking about how the F-22 and the JSF can translate Arabic text while flying supersonically and evading al Qaeda radar. For a little while, they were actually trying to claim that the white elephant would be an effective counter to roadside bombs.

Posted by sglover | November 1, 2006 10:47 AM

A universal translator.

It's like building a ladder to the moon. You make real good progress at first ...

Posted by Jeffrey Davis | November 1, 2006 10:49 AM

A note about timing: I remember seeing a prototype and meeting people working on a Pashto phraselator at SRI when I was in college in 2002.

My experience with machine translation suggests hat anyone telling you today that they're building a universal translator i crazy, lying, or both, but they have been working at it a long time.

Posted by Trevor | November 1, 2006 12:02 PM

My experience with machine translation suggests hat anyone telling you today that they're building a universal translator i crazy, lying, or both, but they have been working at it a long time.
Another way of putting it might be:
Invent a universal translator, get workable AI for free!

Posted by max | November 1, 2006 1:14 PM

Let's use a little deduction here- Putting aside the basic challenges of voice recognition, to get to the point of basic understanding, interpreting the spoken word is a more difficult endeavor than translating the written word. Done in a professional context, the written translator must be far more precise, learned and consistent, but to the point of mutual understanding, written translations deal with a much smaller set of variables. We are still very far away from functioning software to facilitate automatic written translation between English and semi-related languages like Spanish, French and German. It will be many years longer before such technology could be adapted for spoken conversations. One day, this may very well be the best tool in the world, but it's far too early to embrace any current device as a solution, especially if (and I don't know that it is) it is done at the expense of human linguist programs

Posted by mic dee | November 1, 2006 7:31 PM

'Universal translator?" HAH!

I remember reading an article in the inhouse STA magazine pointing out how difficult any sort of machine translation was. Japanese has an incredible number of homonyms, as well as differences according to context. The example they gave was "watashi wa kitune desu." Which could either mean "I am a fox" or "I'll have the kitune noodles."

Posted by grumpy realist | November 2, 2006 4:34 PM

Yeah, and in Mondern Standard Arabic "damit deqeeqa fuqit" means "he only lasted a minute." Applies to our virile codpiece encased Gorgie Boobya Puddin Pie in several senses, at the very least.

Learned that essential phrase at DLI in 2002.

Many of our teachers where, in fact, rather less than fully engaged in teaching us their language. Why? La arif, yannee. I don't doubt most of them had other much more primary concerns & loyalties than advancing American interests. My favorite prof was a Palestinian. Exiled from his home in Hebron since 1948. A former Arab Leauge ambassador to Japan. He was unabashed in expressing his political views. I love him for it, by the way.

Still, using him- or most of my profs- as a military linguist were secrets and lives - and American & Israeli interests are on the line- would be less than wise. Unless your sympathies lie more with the Munazzamat al-Tahrir al-Filastiniyyah /Al-Ikhwān Al-Muslimun/Hezb'Allah/Harakat al-Muqawama al-Islamiyya/Hezb Al-Ba'ath Al-Arabee Al-Ashtrakee, or suchlike.

I also know a few of my friends "gayed out" on purpose. Not to say that they weren't really gay. No. I'd never judge. I could really give a s**t, honestly. I do admire them for playing the Army hard, though. Tens of thousands of college loans relieved, a 100k+ education in Arabic recieved free, over a year spent living well on the Big Sur. Nice work, if you can get it (and most of you reading this could- it's still availible for any US citizen under roughly 35 with no criminal record in good health with a good ear and "120+ IQ" (4/5 of DLI students have at least that- nice stat, eh?)

Anyhoo, talk to you local recruiter. Give it a shot. You could try to fat out (etc. etc.,) if the gay thing is too risque for you. That's a little harder, but still feasible & still honorable (in contemporary parlence, anyway.) If you are smart (and very few at DLI are not) you will never see Iraq. Or Afganistan. Or anywhere else than Ft. Jackson (lots of deep fried southern charm, cute Drill SGTs of both sexes, and lots of very discrete fraternization- coed training is such fun!) Monterey (more mild reigndeer games to bitch about, then lots of fun in class, on the beach, diving or sailing- drinking your ass off- just keep half an eye and mind on the UCMJ and the cadre, and you should be fine.) If you really suck at getting out, you might end up at Goodfellow AFB, in San Angelo, TX. The reigndeer games get a little more annoying here, and if you don't get flushed here, you will end up at a duty station (old fashioned word, that, but there you are) such as Ft. Campbell, KY (Home of the 101st ARB, HOOAH!) or FT. Stewart (Home of the 3rd ID, HOOAH!) or Ft. Carson, CO (Home of the and the Munazzamat al-Tahrir al-Filastiniyyah

Matt, maybe you could do an expose on the real reason we have so many homosexual discharges from the military, concentrated especially in MOS's (job specialties) inhabited by smart, often liberal minded soldiers.

It's a very juicy, if highly un-PC (from every angle) story. You should give it a shot.

Posted by lone_striker | November 19, 2006 11:01 PM

Jacked that post up- blame it on the Merlot! Ranting will usually get you nowhere. Or else somewhere you'd rather not be. Living proof, here.

Just for the record, the PLO is not based at Ft. Carson.

Anyhow, the sex & gender politics thing in the military is a currently a minor disaster. I have no solutions or positions on that. I could tell you all sorts of funny (to me) stories about gays and teenyboopers of all genders in the military, but you might not all enjoy them.

My only comment is that when the crap really hits the fan we will all have other things on our mind. Andrew Sullivan and Dr. James Dobson & Co. aside.

Posted by lone_striker | November 19, 2006 11:19 PM