The Atlantic Monthly reports that a tiny secret U.S. intelligence group, the Intelligence Advanced Research Project Activity (IARPA), has inaugurated The Metaphor Program with the mandate to develop a computer program which can scan large chunks of text and, regardless of the text’s language, generate an evaluation of the author/speaker’s mindset based upon their use of metaphor. IARPA is well-funded and those who tender for research grants under the auspices of The Metaphor Program will share in a purse worth nine figures. (Did you notice my use of a personal accessory metaphor? Clever of me, huh?) IARPA services organizations like the CIA and NSA, so presumably the objective of this project is to enhance U.S. domestic security (rather than expand our knowledge of how we think and communicate). Although the project sounds fascinating, it faces many hurdles, which is my mixed-metaphorical way of saying that it’s the lamest idea I’ve ever heard. It’s right up there with Acoustic Kitty.
Here are some reasons why:
1. Experts don’t agree about what counts as a metaphor.
Metaphor is like art. When we look at visual creations, we have confidence that we know art when we see it. The Mona Lisa is art; graffiti is vandalism. The same holds for verbal creations, and for the same reason. We have been taught what counts as a metaphor, just as we have been taught what counts as art. In high school, we read Catcher In The Rye and understand that Holden Caulfield isn’t really standing in a field of rye grabbing kids as they run past; it’s just a metaphor for an attitude—a desire to protect innocence from the phony world of adults. Or something like that. In fact, most of the time, we resist the radical equivalency of the metaphor (Thou art Peter, and upon this rock…) and interpose the word “like” between the metaphor and the referent (Thou art Peter, like a rock…). We like to translate our metaphors into similes. I use the word “translate” deliberately. It is the Latinate version of the Hellenistic “metaphor”—to carry across.
We are taught the metaphor as a formal creature:
“O rose, thou art sick!” —Rose as Love.
“Oh happy horse, to bear the weight of Antony!” —Riding as Sex.
“Nuns fret not at their convent’s narrow room” —Cloister as Poem.
But, as Michael Rosen points out, our language (every language) is rife with “hidden” metaphors. Metaphors become hidden in two ways:
1) We use them so often that we leap immediately to the meaning without noticing the radical association the metaphor has forced us to make. e.g. “You’re crazy!” Crazy refers to odd angles, and we apply that figure to describe the mind of someone who is mentally ill. But we have grown so accustomed to the phrase that we ignore the visual image which the word “crazy” suggests and have ascribed the meaning “mentally ill” directly to the word “crazy”.
2) The metaphor is embedded in the word’s etymology and, especially now, with the general demise of classical education, we are no longer aware that the etymology even exists. “Metaphor” itself is an example of a hidden metaphor—its components come from ancient Greek words and express in metaphorical terms a particular theory of language.
Rosen is rightly skeptical that The Metaphor Program will be able to uncover all our hidden metaphors and in all languages. But this first hurdle is compounded (to further mix my metaphors) by the fact that many of our etymological roots are in turn hidden metaphors within their own languages, which pushes us back through the history of language to a time which predates writing. We have been verbal creatures for perhaps 100,000 years; only five percent of that linguistic history has been recorded—less when one considers that many cultures have remained oral cultures even down to modern times. Mapping hidden metaphors becomes an exercise in paleolinguistics, whose most practical application thus far has been to inspire the Anthony Burgess film snippets for Quest For Fire which allowed Rae Dawn Chong to grunt convincingly.
2. How does a metaphor work?
Here are some suggestions. Take your pick.
i) We automatically translate metaphors into similes by supplying the word “like”. Peter is like a rock.
ii) We intuitively apply syllogistic reasoning when we encounter a metaphor. Instead of supplying “like”, we proceed this way:
All rocks are stable and strong.
Peter is a rock.
Therefore Peter is stable and strong.
iii) We make a leap of equivalency in which Peter is both like a rock and is in fact a rock, but we recognize this as a cognitive event in which we see overlapping “pictures” of Peter simultaneously.
iv) Same as iii) but we are religious fanatics who affirm an equivalency between Peter and rocks which is not a cognitive event but an event in the physical world.
v) Same as iii) but we are psychotic. Peter is a rock and we use him as a doorstop.
3. Experts don’t even agree about whether metaphors are important.
In Poetic Logic: The Role of Metaphor in Thought, Language, and Culture, Marcel Danesi identifies two main strands of thought about the role of metaphor in language. The dominant strand is the oldest, stretching back to Aristotle and reasserting itself today in the work of Noam Chomsky. It holds that language has a literal foundation. Words have primary meanings and their referents are easily found in the real world. We know what a horse is even though horses in the real world come in a thousand different varieties Metaphor is grafted onto this foundation. It is accidental; not essential. When we speak about “a horse of a different colour”, we assume the givenness of the horse. The expression takes the given (literal) meaning as its foundation and adds a fanciful turn of phrase (metaphor). People who follow this strand of thought tend to denigrate the metaphor. For example, the so-called New Atheists adopt this approach and apply it as their primary critique of Biblical literalism, saying: “But it’s only a metaphor.” In linguistic terms, metaphor is a second-class citizen.
A more recent strand finds its source in the writings of the 18th century Neapolitan philosopher, Giambattista Vico, and it gained renewed interest with the 1980 publication of Metaphors We Live By by George Lakoff and Mark Johnson. This has solidified into Conceptual Metaphor Theory (CMT) and challenges the “accidental” nature of metaphor, claiming instead that metaphor is integral to the way we think.
IARPA has thrown its hat in with the latter strand (mixing my metaphors again), rejecting the traditional view for which Chomsky is the chief proponent. One wonders if Chomsky’s personal politics influenced this decision. One also wonders at the irony of adopting CMT as a theoretical position to support the development of a metaphor algorithm when CMT holds that the reduction of metaphor analysis to algorithms is impossible; metaphor is fundamentally irrational. (Danesi explicitly addresses this issue.)
4. Metaphors are not an exclusively verbal phenomenon.
Metaphors are a cognitive phenomenon which manifest themselves in our communication. While Western literate societies tend to privilege the verbal, that often blinds us to the importance of other media.
In visual arts, metaphors can be hidden, just as in literary arts. For example, religious art tends to place heaven/angels/Christ “up” in relation to hell/devils/Satan. Or they can be overt. Salvador Dali’s watches scream for us to interpret them as meaning-laden objects which stand in for something else.
In music, we often hear auditory references which stand in for something else. We hear the guillotine drop in Hector Berlioz’s Symphonie Fantastique. We hear snatches of Yankee Doodle in Dvorak’s New World Symphony and “read” them as “America”.
But don’t think that metaphors are creatures only of “high” art. Metaphors are just as pervasive in pop culture. Although it is possible to view the Nike swoosh as an abstract visual design, consumers tend to see it as a check mark or as the motion of a runner leaping from the starting blocks. And in the Fulsome Prison Blues, Johnny Cash doesn’t choose his underlying rhythm because his hand happens to twitch that way; he wants to evoke the sound of a train passing the prison.
It is absurd to suggest that a linguistic analysis, no matter how nuanced, can reveal how a people thinks when much of that thinking isn’t even readable as text.
5. In the hands of the military, description becomes prescription.
IARPA is selling The Metaphor Program in benign terms. It merely seeks to describe the use of metaphor with a view to developing a tool to gauge potential security threats. Description, after all, is the primary role of science. This is a scientific project whose aim is to enlarge our knowledge. Science and knowledge. These are good things. And yet the Atlantic Monthly article cites an example which raises a chilling possibility. It cites a presentation slide offered by The Metaphor Program’s manager, Heather McCallum-Bayliss:
Metaphors shape how people think about complex topics and can influence beliefs. A study presented participants with a report on crime in a city; they were asked how crime should be addressed in the city. The report contained statistics, including crime and murder rates, as well as one of two metaphors, CRIME AS A WILD BEAST or CRIME AS A VIRUS. The participants were influenced by the embedded metaphor…
Look at the first sentence. The words “shape” and “influence”—active verbs—suggest that The Metaphor Program will not stop at mere description any more than The Manhattan Project stopped at abstract nuclear physics.
The shift from the descriptive to the prescriptive happens all the time. For example, economics is the science of collective behaviour, most often human behaviour as measured through the transfer of wealth. But its descriptive function has been co-opted by ideologues who convert descriptions of what is into prescriptions of what ought to be: our markets ought to be unregulated to produce a Darwinian free-for-all in which the fittest economic actors prevail. In the same way, a program to describe the metaphorical use of language seems already to incorporate the aim of manipulating metaphor to produce prescribed outcomes.
Maybe it should describe the word ethics.
The example of economics is instructive. Economists have graduated from mathematically inclined social scientists to politically inclined social engineers who tinker with people’s lives by applying a syncretistic ethic that combines a strange mix of social Darwinism and Benthamite utilitarianism. It sounds as if The Metaphor Program promises something analogous, converting the science of cognitive linguistics into an engine of political propaganda. Or, to say the same thing minus all the big words: it wants to clamp our tongues in a vice and fuck us while we squirm.
6. Metaphor is too organic to prescribe in any event.
There is good news: The Metaphor Project aspires to something impossible.
i) Whatever control power imposes from above, language itself provides the tools to subvert from below. There are countless examples of this. Here just a couple. Earlier I had cited the opening line from a poem in William Blake’s Songs of Experience. Here is another:
A little black thing among the snow,
Crying weep! weep! in notes of woe!
“Where are thy father and mother? say?”—
“They are both gone up to the church to pray.
“Because I was happy upon the heath,
And smiled among the winter’s snow,
They clothed me in the clothes of death,
And taught me to sing the notes of woe.
“And because I am happy and dance and sing,
They think they have done me no injury,
And are gone to praise God and his Priest and King,
Who make up a heaven of our misery.”
In the words, weep! weep!, we hear the chimney sweep’s cry. However, half of the metaphorical association is missing from the poem because it isn’t verbal; it is auditory. Blake’s contemporaries would have recognized in the words “weep! weep!” the cry of “sweep!, sweep!” which chimney sweeps made as they wandered through the streets of London soliciting business. (As an aside: our current image of chimney sweeps is influenced by Disney’s Mary Poppins. In fact, the mortality rate of chimney sweeps was extraordinarily high and, for all the soot in their lungs, could never have danced like Dick Van Dyke. In addition, a sweep of Dick Van Dyke’s age would have purchased child slaves to works as apprentices.)
Another example is Symphony No. 5 by Dmitri Shostakovich. Completed in 1937 during the Stalinist purges, it is widely accepted that Shostakovich embedded in the music elements which ran directly counter to the required purpose of the work. Like Blake’s weep! weep! he spoke with two voices at the same time. Those who listened with the ears of suffering heard suffering, and those who listened with the ears of celebration heard celebration. In this way, Shostakovich could critique the Stalinist regime without losing his life.
Shifting to our own time, one need only browse through the urban dictionary to recognize that the practice of subverting “authoritative” usages through metaphorical euphemism is alive and well. You will see what I mean if you try to compile a list of terms to describe male masturbation.
ii) The Metaphor Project appears to presume that language is acontextual. One would expect this of an American-sponsored project since, as we all know, there is only one English which is American, just as there is only one French which is Parisian, and one Russian, and one Spanish, and one Farsi, and one Chinese.
But would The Metaphor Program capture local usages? Because I live in Toronto, when I hear the phrase “Ford nation”, I hear something different than someone who lives in Detroit, as I do when I hear someone speak of pride. The Basket Weave is a section of highway. We don’t look up when someone mentions Vaughan, even though that township claims to be above us. The red rocket doesn’t blast anywhere (it isn’t even red). The King of Kensington doesn’t wear a crown. And when people say “Gumby Goes to Heaven”, they’re not offering a theological opinion on the fate of a green stop-motion plasticine character. These are ways of speaking that are geographically circumscribed and have local referents.
iii) In terms of context, metaphors can be temporally local just as they can be spatially local. Even as one metaphor is catalogued, a new one sprouts. In New York City, a clothier called Akademiks ran ads on MTA buses with the caption: “Read books, get brain.” Officials at MTA didn’t know that “get brain” is street slang for getting oral sex. How could they know this? Kids on the street had only just come up with it.
Emergent technologies are fertile ground for new metaphors. People tweet, surf, spam, chat, dump, poke. Screens freeze. Hard drives crash. We download and upload. Our ISP’s choke our bandwidth. Web sites can be forums or rooms. We load them in our browser even though they have addresses but direct us to a landing page. We ping like a submarine, trace like a spy, and log like a ship’s captain. All of these are metaphorical expressions and several of them didn’t exist until we entered the 3rd millennium. The curious thing is that many of these metaphors will lose their currency, and like Blake’s weep! weep!, will become historical artifacts. As PDA’s and tablets rise in prominence, the app will supersede the browser and a new set of metaphors will emerge. Their appearance won’t be imposed by a central authority; they will “bloom” (more like weeds than flowers, but organic nonetheless).
iv) Those who believe in and claim to know the nature of the metaphor are like those who believe in and claim to know the nature of God. They are true believers. I have no doubt that the people at IARPA believe absolutely in the possibility of their undertaking. Like the God of Reverend Phelps, the Metaphor of the IARPA is a Republican. To put the matter more diplomatically, because metaphor is a cognitive phenomenon, and because our attempts to understand metaphor are mediated by this very cognitive phenomenon, we cannot help but interpret it using tools already embedded in our consciousness. The IARPA frames its study using the very tools which are to become the object of its study. The snake swallows its tail. The vacuum sucks itself out of existence.