The Metaphor Program
June 27, 2011 8:05 AM   Subscribe

Daniel Soar on the militarisation of metaphor: Spies aren’t known for their cultural sensitivity. So it was a surprise when news broke last month that IARPA, a US government agency that funds ‘high-risk/high-payoff research’ into areas of interest to the ‘intelligence community’, had put out a call for contributions to its Metaphor Program, a five-year project to discover what a foreign culture’s metaphors can reveal about its beliefs.
posted by jack_mo (40 comments total) 13 users marked this as a favorite
This plan will go over about as well as Darmok and Jalad at Tanagra.
posted by griphus at 8:10 AM on June 27, 2011 [19 favorites]

Soon, a bureau sub-division to handle the use of Similes
posted by Postroad at 8:12 AM on June 27, 2011

There are some happy sociologists out there somewhere.
posted by maryr at 8:14 AM on June 27, 2011

Does this mean all the literature PhDs can get jobs?
posted by jeffburdges at 8:15 AM on June 27, 2011 [1 favorite]

You know if I read this in a fiction book I'd be all "This is paranoid and slightly absurd" but here we are.
posted by The Whelk at 8:16 AM on June 27, 2011 [1 favorite]

[Requisite coffeeshop joke.]
posted by maryr at 8:16 AM on June 27, 2011

Well, someone had to provide the 101st Fighting Keyboardists with appropriate weapons.
posted by Skeptic at 8:16 AM on June 27, 2011 [2 favorites]

posted by DU at 8:18 AM on June 27, 2011 [6 favorites]

And wait until they discover hyperbole.
posted by Skeptic at 8:18 AM on June 27, 2011 [1 favorite]

I want to eat your children.
posted by swift at 8:20 AM on June 27, 2011 [2 favorites]

Dude! The IARPA logo spells "iARPA"!!
posted by NoMich at 8:24 AM on June 27, 2011

Dude! The IARPA logo spells "iARPA"!!

Steve Jobs is not going to be pleased.
posted by Skeptic at 8:28 AM on June 27, 2011

"If, say, Pashto speakers in Waziristan tend to describe democracy in terms of a tool used indiscriminately by a predator to beat its prey into submission, then this might help intelligence analysts understand their point of view. I admit I don’t know if that is how Pashto speakers do describe democracy."

There are plenty of real examples out there; no need to use an unexplained made-up one. It's not a good way to open and I am also not a fan of the implicit not-so-nice connotative evaluations it has buried in there. GRRRR.

"With the help of heavy-duty computer analysis, they will spend the first couple of years identifying conceptual metaphors in each language and listing them in a ‘metaphor repository’ along with their associated ‘affect’, so that LIFE IS A JOURNEY, for example, would be ‘neutral’ (‘That was the road not taken,’ ‘He really is going places’), and LIFE IS A STRUGGLE would be ‘negative’ (‘You have to fight for what you want,’ ‘They’re on a sinking ship with no lifeboat’)."

I wish the author elaborated on this more. I see an obvious logical error here in that, as it is stated above, the methodology assumes that the targets have inherent universally-recognized affect values of positivity/negativity/neutrality, when the reality is that they are contextually dependent (i.e. in some cultures 'journeys' or 'struggles' might not always be neutral or negative, and there's a LOT of problems in arriving at a conclusion like that in comparing a single concept's cross-linguistic representations).

"But there’s precious little evidence that [metaphors] tell you what people think."

I'm assuming the author means that "there's precious little evidence that metaphors tell you about what people are thinking" (and not the other read, implying that metaphors tell people what to think.). That ambiguity aside, I don't even understand why this author would say this. It's simply not true in that there are countless reputable studies that show how the use of certain metaphor frames and language reveal things about the mental and social organization of language in the mind. I feel like it's unnecessary and unfair problemetization of the current state of metaphor research and it weakens the argument for why this type of work is really important.

"Among those present were representatives from Olson Zaltman Associates – a consulting firm that uses ‘metaphor elicitation’ to help global brands (clients include GlaxoSmithKline, Pfizer, Chevron, DuPont, Procter & Gamble, Coca-Cola and the World Bank) ‘target consumers’ implicit thoughts, feelings and knowledge’ – and from Charles River Analytics, which provides expertise in ‘psychology of narratives, cultural language patterns and semantics’ for the Department of Defense."

Punctuation-wise, this sentence is sillypants to parse.
posted by iamkimiam at 8:34 AM on June 27, 2011 [3 favorites]

Everything I know so far about this project, which is admittedly not very much, suggests that it has limited true usefulness, but unlimited potential for misuse and misinterpretation. Take a look at this entence from the linked LRB essay:
What would it tell us if it turned out that encoded in the very language of the Iranian people is the concept that LIFE IS A BLAST?
Presumably, about as much as finding out that LIFE IS A BLAST is a common-enough metaphor in English. Or, you know, what about English expressions like "break a leg"? So, as far as I can see, this thing can easily fall into (and has to meticulously avoid) several extremely pernicious misconceptions:
  1. Strong Sapir-Worf linguistic relativity (i.e., language directly shapes perception and thought).
  2. The language is the people (i.e., identifying Iranians with Farsi and Farsi wth Iranians).
  3. Hence, divorcing language from the daily reality people inhabit (thus externalizing and essentializing social practices).
Again, using the example of Farsi, were I to speak Farsi like a native, would I get the benefit of its cultural payload? (See what I did there?) Or is it only "real" Iranians, i.e. the kind who might do us harm, who are unconsciously and mindlessly guided by their tongue into specific modes of thought and action?
posted by Nomyte at 8:39 AM on June 27, 2011 [2 favorites]

Footnote: it looks like the first link is based on Fauconnier and Turner's work

War on drugs
War on terror
War on poverty
War on waste
War on cancer
War on women
War on journalists
War on cops
War on fun
War on everything

From the first three pages of a google search for "war on"

Now, war on metaphor

Where's my war on war?
posted by justsomebodythatyouusedtoknow at 8:50 AM on June 27, 2011 [1 favorite]

First, I challenge the premise: that spies are not traditionally very 'culturally sensitive.' If they're any good at spying, they bloody well are. Don't confuse:
  1. Spies with the people who manage spies
  2. Either of those with their political masters
  3. Methods with results (see [2])
That said, second: if the project is meant to have any kind of broad applicability, then I'm with Nomyte: it's not likely to end well.
posted by lodurr at 8:52 AM on June 27, 2011 [2 favorites]

Maybe with some measure of success in this program we'll be able to get the cow off the ice.
posted by Meatbomb at 8:59 AM on June 27, 2011

Presumably, about as much as finding out that LIFE IS A BLAST is a common-enough metaphor in English. Or, you know, what about English expressions like "break a leg"?

I seriously doubt even the US military is dumb enough to go this route. It's more like how in English (and other languages) use "see" to mean "understand". Not just in a single stock phrase, but throughout our language: We make this "clearer" when we explain them. I "see" what you mean. Let's try to understand his "point of view". Etc.

An alien could use that information to realize that we our eyes are a major source of information. Now do the same thing, but on a cultural rather than species level. That "don't think of an elephant" linguist guy has written books on this. Lakoff.
posted by DU at 9:00 AM on June 27, 2011 [1 favorite]

Where's my war on war?

We had one in the '60s. War won.
posted by griphus at 9:01 AM on June 27, 2011 [7 favorites]

This is a fascinating idea; I only came across Lakoff and his work on metaphor fairly recently (thanks to another mefite), and I'm intruiged to know to what extent analysis like this could prove useful. It would, presumably, also be a powerful tool when designing propaganda cultural exchange materials.

My first instinct was to doubt its utility. Not because I dismiss Lakoff's ideas, but because even if you can identify metaphors which reliably carry the same overtones when considered by a large cross-section of the population, there doesn't seem to be much information available from the technique that couldn't be gained by just interviewing someone familiar with the society. Say, a recent ex-pat or people on the ambassadorial team.* Further, if someone don't know the society well enough to already have an intuitive sense of the answers they're looking for, I seriously doubt their ability to fully understand the baggage of their metaphors that they're trying to analyse, much less usefully use them in propaganda communications.

With that said, there was a lovely bit of work recently showing that choosing to describe crime as a "beast" or a "virus" changes people's opinions on how crime should be tackled and what punishments should be meted out. It's a much more powerful effect than I would've expected, so maybe there's some mileage in this sort of project?

*This reminded me of an irrelevant but interesting conversation I had with the leaders of two labs responsible for monitoring viral disease outbreaks in US Army troops. In addition to studying the outbreaks that inevitably spread like wildfire when you bring young people from all over the country into constant close contact and subject them to immune-suppressing levels of stress, they had a side line in gathering epidemiological data on countries who didn't release any (or any trustworthy) epidemiological data to the international community. Their approach, basically, was to wait for healthy troops to go out on leave in whatever country they were stationed in, eat the street food and consort with the locals (wink, nudge), then come back to base to be prodded and probed to see what interesting diseases they've managed to pick up. Not a perfect system but, with some stats work and a bit of judicious fudging, they get much more reliable information about the state of a region's health than certain governments are willing or able to publish.
posted by metaBugs at 9:01 AM on June 27, 2011

This plan will go over about as well as Darmok and Jalad at Tanagra.
Shaka, when the walls fell.

FTFY. (Unless you meant to express that you think this plan will produce a new partnership through shared struggle.)
posted by The Tensor at 9:01 AM on June 27, 2011


You know, at first it was a poorly-research, nerdy reference. Then I realized that anything that gets the US to study a foreign culture instead of just delivering democracy via carpet bombs is a step in the right direction. Students in Cold War Russia were taught English from the get-go because why shouldn't you know the language of the enemy? How is ignorance a tactical advantage?
posted by griphus at 9:05 AM on June 27, 2011 [1 favorite]

Yikes, I *want* to like and support this project so much, but it is sorta doomed in the way it is presented. For reasons Nomyte stated above, and more.

From the contribution-call:
"The Metaphor Program will exploit the fact that metaphors are pervasive in everyday talk and reveal the underlying beliefs and worldviews of members of a culture. In the first phase of the two-phase program, performers will develop automated tools and techniques for recognizing, defining and categorizing linguistic metaphors associated with target concepts and found in large amounts of native-language text. The resulting conceptual metaphors will be validated using empirical social science methods. In the second phase, the program will characterize differing cultural perspectives associated with case studies of the types of interest to the Intelligence Community. Performers will apply the methodology established in the first phase and will identify the conceptual metaphors used by the various protagonists, organizing and structuring them to reveal the contrastive stances."
So many things...
(I put metaphor frames containing evaluations in bold; I put problematic phrasing in italics)

exploit the pervasiveness of metaphors? Are we trying to kill them? Overwork them? Why the attack language? Why a negative word like 'pervasive'?

The use of 'reveal' in the first sentence is missing an antecedent and so it is unclear if the second clause is about The Metaphor Program revealing underlying beliefs or if metaphors reveal underlying beliefs, i.e. who/what's the agent here? Also, the last bit of the first sentence implies that individuals of a culture may be treated as a monolithic entity. For many readers I think, this will immediately set off alarm bells, even if the aim of the project is specifically not to do that.

Performers? Why this metaphor frame? Is this work to be seen as a show, a magic trick, or is this simply stylistic writing flair's hard to tell.

Automated tools are great, but I have some concerns about how that could work for metaphor, which is sometimes idiomatic, often stylistic, usually semantically ambiguous and always context dependent.

large amounts [...] of text. Is this text-based research? Speech that has been transcribed? Two very different speech genres there. "Large amounts"...what does this even mean? Are they talking about frequency of occurence across 'texts' or size/length/time of communciation of a speech or piece of text?

"The resulting conceptual metaphors..." Resulting from what? Are they taking 'linguistic metaphors' (as they mentioned above) and then breaking them down into component conceptual/cognitive metaphors? Or are they just conflating concepts here (linguistic metaphors and conceptual metaphors are different things)? If they're breaking down linguistic metaphors, then 'validating' them makes sense. If they're conflating terms, then 'validating' refers to the validation of a metaphor (as a metaphor) and we are back to exploting metaphors and their pervasive ways again.

"characterizing differing cultural perspectives" is again a bit shaky...characterizing individuals within a culture or lumping all of them together by their metaphor frames, treating them as a monolithic entity and comparing them to other monolithic entities and characterizing the differences? In other words, it really matters where and in which order the lumping and characterizing comes in.

the Intelligence Community...this is getting a bit scary to me, to be honest. Especially now that we've stopped referring to people as members of a community, but rather protagonists characterized by us as performers for the benefit of the IC audience on some world stage. (Yes, I elaborated the metaphor frame at the end of the sentence there.)

'them' and 'contrastive stances' are ambiguous like 'reveal' was. Are the protagonists being organized? Or are they organized indirectly via their conceptual metaphors (via their linguistic metaphors)? Also, 'contrastive' stances assumes opposition. Not necessarily a bad thing, but just another point to ponder.

Again, I think there's a huge amount of importance to the work, but I'm a weee bit skeptical in the way it's been framed. I can't tell if it was careless or intentional or both, and that scares me.
posted by iamkimiam at 9:24 AM on June 27, 2011 [3 favorites]

Sounds like IARPA is just thinking about the immortality of the crab.
posted by RobotVoodooPower at 9:26 AM on June 27, 2011

I is in yr base using yr words.
posted by adamvasco at 9:40 AM on June 27, 2011

The way to stop nonsense like this is not to play whackamole with the intelligence agencies, trying to shut down each ridiculous, pointless, or downright counter-productive program that pops up. The way to stop this is to REDUCE THEIR BUDGET UNTIL THEY DON'T HAVE THE MONEY TO DO THIS ANYMORE.

They clearly have way, way, WAY too much money on their hands. The only way IARPA could be redeemed is if it's actually a secret sting program designed to lure the deadweight employees out of the woodwork so they can all be fired at once. But who am I kidding - they don't have the organizational capability to put together something that complicated. High-risk, high-payoff programs indeed. Sounds more like zero-risk, negative-payoff to me.

The "Intelligence" community has shown itself time and time again to be far too organizationally incompetent to do the most basic things you would want an intelligence community to accomplish. Even if you allow that it might be useful to "analyze foreign metaphors", any useful application is miles over the heads of those brainiacs. I'd rather go back to funding men who stare at goats.
posted by Salvor Hardin at 9:44 AM on June 27, 2011

posted by idiopath at 10:21 AM on June 27, 2011 [1 favorite]

This is guerrilla marketing for Miéville's "Embassytown", isn't it?
posted by Slap*Happy at 10:23 AM on June 27, 2011

iamkimiam: 'Performers' is a term of art in government contracting meaning "those who have been awarded a contract to perform the work". Nothing more.

I wouldn't overanalyze the specifics of the BAA language too much. Research BAAs must tread a fine line between specificity and generality - and in a task like this, generality is probably more important. A winning proposal would probably address all the questions you raise, and TELL the government what they should be doing, why your proposed solution is correct, and expose a number of important issues they haven't thought about yet.

Salvor Hardin: interesting response for someone with that handle. I personally think this is an excellent research area. If you want to comprehend real-world language, you MUST address the issue of metaphor. The government is finally understanding, after a decade-plus of highly cross-cultural conflict influenced by rhetoric, that they better start addressing these issues.
posted by scolbath at 10:53 AM on June 27, 2011 [1 favorite]

Oh thanks, scolbath! That makes a bit more sense there. I didn't know that 'performers' had a specific industry meaning. It's a perfect example of the context-dependent classification problem of metaphor and why it's so hard to automate it though.
posted by iamkimiam at 11:06 AM on June 27, 2011

slap*happy, that's not exactly a crazy notion.
posted by lodurr at 12:09 PM on June 27, 2011

Just wait until they learn to weaponize irony.
posted by pmugowsky at 12:23 PM on June 27, 2011

Salvor Hardin: interesting response for someone with that handle.

It is interesting, I suppose, but it illustrates my understanding of this as pure fantasy. Perhaps, in the distant future of our galactic empire, sociolinguistics will attain the rigor to make this work useful, and our intelligence agencies will gain the operational coherency to make use of it. But not now. Even if this were more than a fanciful piece of icing for the DoD budget, US intelligence agencies do not have the capability to translate this into useful policy.

If I may use a metaphor, this project is the swordsman who studies the finer points of sword-sharpening when he hasn't yet learned to swing his sword without cutting off his own limbs.

I sound a bit like small-government republicans when I talk about military & intelligence spending. The government is just terrible at collecting intelligence, analyzing it meaningfully, distributing it to the people who can act on it, and then acting on it.

The thing that distinguishes me from small-government republicans in this respect is that I'm pretty sure that private contractors would be at least twice as terrible at it as our government is now.
posted by Salvor Hardin at 1:20 PM on June 27, 2011

it illustrates my understanding of this as pure fantasy

Hmm, I didn't mean to be quite as self-deprecating as naturally parses.

I meant, it illustrates my understanding that this project is pure fantasy
posted by Salvor Hardin at 1:22 PM on June 27, 2011

I didn't mean to be as self-deprecating as that sentence naturally parses. I give up.
posted by Salvor Hardin at 1:23 PM on June 27, 2011

Let me get this straight -- US Intelligence and Military castigated by Metafilter for not knowing anything about other countries. US tries to initiate programs to learn more about other countries and ... gets castigated by Metafilter. Cripes.
posted by proj at 1:38 PM on June 27, 2011

US Intelligence and Military castigated by Metafilter for not knowing anything about other countries. US tries to initiate programs to learn more about other countries and ... gets castigated by Metafilter.

The only winning move is not to play.
posted by never used baby shoes at 2:56 PM on June 27, 2011

Just wait until they learn to weaponize irony.

I don't know about irony, but I have an idea about where they can find some weapon-grade snark.
posted by Skeptic at 4:06 PM on June 27, 2011

US tries to initiate programs to learn more about other countries and ... gets castigated by Metafilter.

I react this way to Metafilter often, but the Metaphor Program seems especially dubious. Leaving aside the very controversial Sapir-Whorf assumptions built in here, the hypothesis that it would operate on a societal level raises a whole host of other theoretical concerns.

But the NLP applications are exciting! Computers are lagging in this area big time, as I understand it.
posted by stroke_count at 4:46 PM on June 27, 2011

Do spies use sarcasm?
posted by BlueHorse at 8:00 PM on June 27, 2011

« Older Reveals Reality?   |   SCOTUS, Video Games, & Violence Newer »

This thread has been archived and is closed to new comments