Monday, July 18, 2005

Lots to Say About Lakoff

Lakoff talk is back, this time in the New York Times (via Language Log). The article, titled "The Framing Wars," is long, and full of good, bad, and just plain false points about Lakoff, framing, and the uses of framing. The best parts of the article are the discussion of some successful real-world applications of framing by Democrats, in the social security and filibuster debates, and the point that one of the most important aspects of the Democrats embracing framing is their newfound discipline in staying on message and using the frames that their research has shown will work. While the article points out that much of the discussion of framing in the media, among politicians, and I should add, on blogs, has been overly simplistic, the fact that Democrats are sticking to their frames shows that they've grasped at least one of the more complex points about framing, namely that in order for it to work, it has to be consistent and coherent enough to change the way people actually think about an issue. If you switch back and forth between multiple frames, or alternate between your frames and your opposition's frames, you will have a hard altering anyone's schemas.

These two good aspects of the article are overshadowed, however, by much that isn't so good. The popular press simply can't get it right when it comes to science, even when that science is as shitty as Lakoff's. I'm glad that framing itself is still getting attention, even in a non-election year, but I hope that more liberal cognitive scientists who study concepts, metaphor and analogy, rhetoric/communication, and related topics will start to speak up within earshot of politicians and the press (i.e., not on podunk blogs like this one). For now, I'm going to speak up on a podunk blog like this one (identical with it, in fact) about the NYT article.

The article doesn't get to Lakoff until the third page (of the online version), and when it does, it starts off on the wrong foot, calling Lakoff "the father of framing." This will come as a shock to anyone who read Erving Goffman's work on framing in the mid-1970s, or to anyone who wrote on framing between the publication of Goffman's book and Lakoff's work on the topic in the 1990s (or even the 1980s, if we are generous and describe Lakoff's early work on metaphor as work on framing). Lakoff has popularized framing, however, and for that he does deserve recognition. I can't help but think that Matt Bai, the author of the New York Times piece, has been buying into Lakoff's framing of himself, though. He follows the bogus "father of framing" metaphor with this:
A year ago, Lakoff was an obscure linguistics professor at Berkeley, renowned as one of the great, if controversial, minds in cognitive science but largely unknown outside of it.
As far as linguistic professors go, Lakoff hasn't been obscure for a long time. After Noam Chomsky, he has probably been the most famous linguistics professor in the U.S. for at least two decades. Metaphors We Live By has been one of the most popular cognitive science books among non-cognitive scientists for 20 years, and I'd wager that Women, Fire, and Dangerous Things has sold more copies that virtually any other cog sci book of similar length as well. This is not because cognitive scientists are reading it (most of the ones I know haven't). Furthermore, among cognitive scientists, Lakoff is certainly not considered one of the great minds of our discipline. He was controversial at one time, though now he's more recalcitrant in the face of overwhelming data against his view than he is controversial. But this article appears to be a sort of myth-building exercise, and "largely ignored by cognitive scientists, Lakoff has begun to gain a great deal of influence in politics" doesn't seem like a good start to a myth.

To Bai's credit, he does try to hint at some of the deeper aspects of framing. He writes:
According to Lakoff, Democrats have been wrong to assume that people are rational actors who make their decisions based on facts; in reality, he says, cognitive science has proved that all of us are programmed to respond to the frames that have been embedded deep in our unconscious minds, and if the facts don't fit the frame, our brains simply reject them. Lakoff explained to me that the frames in our brains can be ''activated'' by the right combination of words and imagery, and only then, once the brain has been unlocked, can we process the facts being thrown at us.
This could be an OK short description of how framing works. Framing is about changing and utilizing largely unconscious representations. (A terminological note: I call these representations schemas, even though Lakoff calls them frames. I think it's important to call the representations "schemas" because if we call both the language and the conceptual representations that framing analysis deals with "frames," then we might be tempted to treat "framing" as a strictly linguistic phenomenon, which is what many, including most of the politicians the article discusses, appear to do.) The article at least attempts to make it clear that this is about changing how people think about issues, not just changing how we talk about them, and that's a good thing. Still, as Bai himself shows, the use of the word "unconscious" immediately activates common misconceptions of the cognitive unconscious. For example, Bai writes:
This notion of ''activating'' unconscious thought sounded like something out of ''The Manchurian Candidate'' (''Raymond, why don't you pass the time by playing a little solitaire?''), and I asked Lakoff if he was suggesting that Americans voted for conservatives because they had been brainwashed.
Lakoff quickly points out how this is wrong, saying:
That's true, but that's different from brainwashing, and it's a very important thing. Brainwashing has to do with physical control, capturing people and giving them messages over and over under conditions of physical deprivation or torture. What conservatives have done is not brainwashing in this way. They've done something that's perfectly legal. What they've done is find ways to set their frames into words over many years and have them repeated over and over again and have everybody say it the same way and get their journalists to repeat them, until they became part of normal English.
What Lakoff doesn't really make clear is that framing in politics is not about changing people's entire world-views by talking about things differently. Instead, it's about changing the way we talk about things so that people will represent them within particular schemas that they already have. So, for instance, we might want to reframe abortion so that, instead of being about life and death, it is represented utilizing the schemas we use to represent freedom and choice. This is nothing like brainwashing (it's not, as Lakoff suggests, different simply because it's legal). It involves showing people new relations between domains, and then utilizing the schemas from one domain (e.g., freedom) to represent another domain (abortion). If we want to use framing successfully, we have to realize this. It's important that, instead of going out and conducting surveys and focus groups to figure out which frames work best for each individual issue, we go out and look for the base domains (the schemas that people already have) that work the best, and then develop analogies (Lakoff would call them metaphors, but in order to avoid the appearance of endorsing his bogus cognitive theories, I'm going to call them analogies) from those. Sometimes this will require figuring out which base domains can best be used to represent individual domains, but it will also require developing large-scale narratives (or systems of analogies) from a set of base domains, which is what Lakoff argues Republicans have done so well.

Finally, the article includes the gratuitous "other perspective" (you know, to make it "objective"). There are two versions of this, one from Republicans and one from anti-framing Democrats. The Republican critique comes from Frank Luntz. Here's what Bai writes of his objections:
The problem with Lakoff, Luntz said, is that the professor's ideology seemed to be driving his science. Luntz, after all, has never made for a terribly convincing conservative ideologue. (During our conversation, he volunteered that the man he admired most was the actor Peter Sellers, for his ability to disappear into whatever role he was given.) Luntz sees Lakoff, by contrast, as a doctrinaire liberal who believes viscerally that if Democrats are losing, it has to be because of the words they use rather than the substance of the argument they make. What Lakoff didn't realize, Luntz said, was that poll-tested phrases like ''tax relief'' were successful only because they reflected the values of voters to begin with; no one could sell ideas like higher taxes and more government to the American voter, no matter how they were framed. To prove it, Luntz, as part of his recent polling for the U.S. Chamber of Commerce, specifically tested some of Lakoff's proposed language on taxation. He said he found that even when voters were reminded of the government's need to invest in education, health care, national security and retirement security, 66 percent of them said the United States overtaxed its citizenry and only 14 percent said we were undertaxed. Luntz presented this data to chamber officials on a slide with the headline ''George Lakoff Is Wrong!!''
I'll be damned if I know what Luntz' admiration of Peter Sellers has to do with any of this, but I can say for sure that Luntz' doesn't understand Lakoff's work. First of all, while it is highly likely that Lakoff's politics are guiding his representation of conservatives and liberals (through his descriptions of the STRICT FATHER and NURTURANT PARENT metaphors), that's not science. That's Lakoff's own idiosyncratic informal cognitive linguistic analysis of political discourse. The science, which I will admit isn't very good science, but which makes many of the same points that good science does, was around long before Lakoff got political. In addition, Luntz' data doesn't show that Lakoff's view of framing (or framing analysis itself) is wrong. It does show that Lakoff might be wrong about which frames to use for taxation, but even that's not clear. Lakoff himself notes that simply presenting frames once or twice over a short period of time is not sufficient to change people's deeply-entrenched and largely unconscious (and therefore difficult to consciously change) representations. It takes the extended and consistent use of frames to create large-scale representational change. As Bai puts it, Lakoff "does not suggest in his writing that a few catchy slogans can turn the political order on its head by the next election."

The Democratic criticisms are slightly better, but completely misdirected. Here is how Bai describes it:
Lakoff's detractors say that it is he who resembles the traveling elixir salesman, peddling comforting answers at a time when desperate Democrats should be admitting some hard truths about their failure to generate new ideas. ''Every election defeat has a charlatan, some guy who shows up and says, 'Hey, I marketed the lava lamp, and I can market Democratic politics,''' says Kenneth Baer, a former White House speechwriter who wrote an early article attacking Lakoff's ideas in The Washington Monthly. ''At its most basic, it represents the Democratic desire to find a messiah.''
This is simply unfair to Lakoff. He is not a charlatan, even if he's not a good cognitive scientist. His central points, about the importance of morality in political discourse, and the use of what we know about how people think to help communicate our ideas better while counteracting the deceptive use of language by Republicans, are dead on. Yet, it is obviously true that many Democrats have latched onto Lakoff without fully understanding what framing entails. It does not mean that we can call taxation "dues for services rendered," and think that people will immediately vote for our candidates.

It's a shame that the article included either of these criticisms, at least as directed toward Lakoff (the latter is pretty dead on when directed toward Democrats who don't understand framing). It's also a shame that the article doesn't do more to correct the misrepresentations that it itself laments, including the ones that it tries to perpetuate (e.g., with talk of brainwashing). It would be really nice if the article actually mentioned the criticisms of cognitive scientists (you know, the people who actually know what they're talking about), too. But all of that is probably too much to expect from an article in the New York Times, even if it is 12 frickin' pages long.

11 comments:

Bora Zivkovic said...

I saw somewhere the previous night that the article was coming out yesterday, so I went out and bought NYT - who is going to read such a long article on screen - an I have to say, I agree 100% with your commentary. That is exactly how I thought about it. As far as quality of his science, I defer to you, but I think it is inexcusable that so many Democrats do not even understand what framing is all about. I think Lakoff is impoartnt in pushing the idea of framing. I also think that "Elephant" is bad and that people who have read it do not understand what framing is all about, i.e., think it is branding or sloganeering. Even if "Moral Politics" is bad science as you say, I think it does a much better job at explaining what framing is and how to go about it.

Have you noticed that other "language experts" mentioned in the article come from marketing or polling business. What on Earth do they know about framing?

I think that Lakoff is aware that he is not a "framer" but a scholar. He is pushing himself and Rockridge as a first step to get framing into the Democratic machine, and would, undoubtedly, hire other good people to do the polling, testing, research and writing Talking Points.

I am glad that Lakoff at least got the Dems serious enough to realize the importance of "staying on message" and frequent repetition. I was intrigued with their focus-group testing about Social Security. I see the first, originally favored wording, which failed the test, as something that figures prominently on liberal blogs, while the most succesful wording, one that ended up being successfully used, is not so commonly seen on blogs. Which, in a way, suggests that Lakoff is correct.

Chris said...

I'm glad you saw it the way I did. My first thought was, "I'm glad it's getting this much attention, but I wish it wasn't getting this attention."

If Lakoff were a good, honest scientist, instead of an overly self-conscious burgeoning celebrity, he would invite other cognitive scientists weigh in on how to do this. In true Lakoffian fashion, he's presented his way as the only way, and as a scientifically proven way, and it is neither of those things.

There is no shortage of liberal cognitive scientists who are working on the same cognitive phenomena that Lakoff studies, and very, very few of them agree with him. It would take very little for him to invite one to work with Rockridge, for instance.

Razib Khan said...

i don't know what you're going to do about lakoff. i have several friends who are in the humanities and eat his shit up as an alternative to post-modernism (they love philosophy in the flesh). he's going to be the face of "cognitive science" pretty soon. pinker or him, pick your poison ;)

Chris said...

Oh man, I know the love of Lakoff in the humanities all-too-well. I came out of philosophy, and it was there that I first encountered him. I was (and am) interested in embodiment, because I was heavily influenced by the writing of Merleau-Ponty, Heidegger, Husserl, J. J. Gibson, and to a lesser extent, Hubert Dreyfus, so I was eager to read Lakoff, Johnson, and some of the other cognitive linguists who relied heavily on embodiment in their theories. I was so disappointed that to this day I remain noticably bitter. Reading Philosophy in the Flesh (or Metaphors We Live By, for that matter), it's quite clear that Lakoff and Johnson have had little exposure to work outside of the analytic tradition, and are simply rebelling against that tradition. Their ignorance means that they ignore the wealth of philosophical literature on embodiment (they explicitly say that philosophy has been objectivist since Aristotle, disembodied since Descartes, and that they are the first to go against it), and their rebellion means that they ignore the methods of science in their work.

Honestly, if people in the humanities took the time to read more of the work on concepts, analogy and similarity, and reasoning in cognitive science, they'd learn that mainstream cognitive science presents a picture of the mind that is pretty postmodern, and wouldn't have to latch on to the simplest version of those ideas that they came across.

I also wish that the idea of embodiment hadn't been so firmly connected to Lakoff, Johnson, and their epigones, because it has made the acceptance of embodiment in cognitive science much more difficult. To many cognitive scientists, embodiment is just another whacky fringe idea, not because they actually believe that the body is not important, but because Lakoff and his ilk tout it. That's changing (you know it's changing when Gordon Bower, who cognitive scientists all know as one of the old guard, a student of George Miller's even, starts giving talks about embodiment), but it's changing very slowly.

Bora Zivkovic said...

Also, thanks for the reminder to read Guffman. My brother told me I had to read Guffman several years ago. I know I have at least one of the books here somehwere in this mess - perhaps this is a good time to find it and read it.

Razib Khan said...
This comment has been removed by a blog administrator.
Razib Khan said...
This comment has been removed by a blog administrator.
Razib Khan said...

the NSF or someone should fund a yearly project which surveys people doing research in various disciplines on their opinions on topics of note. these surveys should be widely distributed, so that people can figure out what the consensus is in other fields.

right now, you have self-promoters like lakoff (in "cog sci") setting the terms of the debate of what cog sci is outside of cog sci because people doing real research are too busy. this is common to other fields of course...but i won't bring the disputes bora and i have had as relates to the "levels of selection" controversy on to this blog :)

and on something we can all agree on it, if this practice became more common the lay public would have a better grasp of the reality that there is no point in "teaching the controversy" as regards to evolutionary science as there isn't a controversy, since evolutionary biologists don't dispute evolution, and there isn't a field of biology called "design biology" to survey right now.

Chris said...

That would be an interesting project. I bet the findings would surprise some people, even me.

I also bet that you'd find that among many sciences, there are as many practitioners who believe shitty ideas as there are non-practitioners. The fact that scientists are so specialized these days means that they often don't have the time or motivation to thoroughly evaluate ideas in their general field thoroughly.

Razib Khan said...

The fact that scientists are so specialized these days means that they often don't have the time or motivation to thoroughly evaluate ideas in their general field thoroughly.

well, i'm sure some psych people have investigated this. i know many scientists who find it frustrating, but don't know what to do about it, they have a hard time as it is keeping up on the literature in their small patch of their subfield. there are some really bizarre dynamics that ensue because of this. a common one is

person A has a science X, person B is related to person A somehow (friend, significant other, relative), and person B tells person C about the "findings in science X1." person C might actuall have a degree in science X1, while person A has a degree in science X3, so person B is in a really embarrassing spot all of the sudden because the argument-from-authority just collapsed.

i am simply distilling the general experience i've had with anyone who is somehow connected to someone in a medical related field, and who don't know that i have a biological science background myself. they will invariably tell me something about evolution that is a really embarrassing summation which is worthy of a lay person who hasn't picked up even any of the popular works, as if of course person A would know since they are a vetrinarian, doctor, dentist, etc.

this sort of scenario is generalizable of course. people are narrow-focused specialists, but when the other people they are discussing shit with are far enough, they can pretend they would know something about organic chemistry when they're really biochemists.

Razib Khan said...

to elaborate on the organic-biochemist divide: i am skeptical that the typical biochemist knows anything more about organic chemistry than what anyone who has taken intro organic chemistry (all premeds, all biology majors, many other science majors) would know. but most lay people don't know they, organic and biochemistry sound like they are should be right next to each other.