Frame Game

by The New Republic Staff | November 4, 2006

by Geoffrey Nunberg

For linguists of mature age, it's hard to read the exchanges between Steven Pinker and George Lakoff over Pinker's review of Lakoff's new book without a sense of déjà vu all over again--they recall the heated confrontations of the "Grammar Wars" of the 1970s, when Lakoff and other linguists of the Generative Semantics school were going to the mattresses with Chomsky and his followers over the nature of language structure. True, Pinker hardly shares Chomsky's politics, but like Chomsky he's a take-no-prisoners polemicist. And Lakoff is still displaying the rhetorical irrepressibility that has always been a bit exasperating even to linguists who like and respect him. He frames his claims grandly, often downplaying or ignoring his intellectual predecessors, so that his own contributions come off as revolutionary rather than usefully accretive. And he isn't above caricaturing his opponent's views or raising ad hominem arguments (what do Pinker's views on women in science have to do with any of this?). The tone of those exchanges has a lot to do with the personalities of the protagonists, of course, but it also reflects the intensely Oedipal character of the debate. You have to go back to Freud to find a figure who has dominated the thinking in his field as pervasively as Chomsky has, and his influence is evident not just in the work of Pinker, who happily acknowledges it, but in the challenges mounted by his renegade former disciples. The fact is that in his way, Lakoff is as much a Chomskian as Pinker is: There's the same psychologism, the same reductiveness, the same stress on cognitive universals and downplaying of merely social and historical factors. In the end, the two men are playing the same game by the same rules. Which may explain why, in the end, Pinker hasn't really bearded Lakoff on the most problematic consequences of his theory of political morality and political language, even if he scores some very telling points along the way. To hear Lakoff tell it--and Pinker more-or-less goes along with this--the debate is really about the nature of reasoning, or more specifically, political reasoning. As Lakoff puts it, with characteristic expansiveness:

We can no longer conduct 21st century politics with a 17th century understanding of the mind.... In thinking, the old view comes originally from Descartes' 17th Century rationalism. A view of thought as symbolic logic was formalized by Bertrand Russell and Gottlob Frege around the turn of the 20th Century, and a rationalist interpretation was revived by Chomsky in the 1950's. In that view, thought is a matter of (as Pinker puts it) "old-fashioned ... universal disembodied reason." Here reason is seen as the manipulation of meaningless symbols, as in symbolic logic.

The new view is that reason is embodied in a nontrivial way. The brain gives rise to thought in the form of conceptual frames, image-schemas, prototypes, conceptual metaphors, and conceptual blends. The process of thinking is not algorithmic symbol manipulation, but rather neural computation, using brain mechanisms...

These questions matter in progressive politics, because many progressives were brought up with the old 17th Century rationalist view of reason that implies that, if you just tell people the facts, they will reason to the right conclusion--since reason is universal. We know from recent elections that this is just false. "Old-fashioned ... universal disembodied reason" also claims that everyone reasons the same way, that differences in world-view don't matter. But anybody tuning in to contemporary talk shows will notice that not everybody reasons the same way and that world-view does matter.
Now if you took this all at face value, you might conclude that from the seventeenth century onwards, everybody thought that political decision-making involved purely abstract ratiocination about brute facts, until Lakoff and some of his contemporaries invented cognitive science and discovered that people actually make decisions on the basis of their emotions and their social identities, and that metaphors and cognitive shortcuts play an important role in the way we reason about politics. But it isn't exactly news that emotionally charged symbols and metaphors play a big role people's political reasoning--not to politicians, certainly, and not to scholars, either. Philosophers have been emphasizing the role of metaphor in cognition since the eighteenth century, and linguists and psychologists since the nineteenth. And there's no shortage of 20th-century writers who have looked specifically at the role of metaphor in political argument--you think of Ernst Cassirer, Quentin Skinner, J. G. A. Pocock, Walter Lippmann, and Murray Edelman, to name just a few (none of them cited by Lakoff, as it happens). Indeed, if the attacks on Lakoff involved simply a refusal to acknowledge the obvious importance of metaphors, frames, and the like in political reasoning, as he charges, you might conclude that he's right to say his critics are motivated chiefly by hostility to his progressive views (though that would make all the more puzzling the attacks on Lakoff by the likes of Marc Cooper and Kevin Drum, who wouldn't seem to have any DLC axes to grind.) It's true that modern cognitive scientists have learned a lot about the cognitive mechanics of metaphorical processing and the way people use cognitive shortcuts in their reasoning. And Lakoff and his collaborators have made important contributions in cataloguing the metaphorical schemas that systematically run through ordinary language. As both Pinker and Lakoff acknowledge, there are lively debates about just how much one can conclude about cognition from the metaphoricity of natural language (for what it's worth, I tend to take the same side here that Lakoff does--in fact I've argued that even stone-dead metaphors like kick the bucket retain a psychological reality that affects their syntactic behavior). But notwithstanding Pinker and Lakoff's sparring over these issues, they're all beside the point, because there's nothing in any of this research that could remotely prove or disprove Lakoff's grander claims about the cognitive basis of our political orientations. Lakoff isn't claiming simply that metaphorical thinking often shapes our political views, after all, but that our political reasoning is dominated by a single metaphorical schema of "the nation as family." From that assumption follow several sweeping claims: First, our allegiance to one or the two poles of American political life, liberal and conservative, depends on which of two idealized models of the family we subscribe to, the "strict father" model or the "nurturant parent" model. And second, these models of the family shape people's positions on all the major issues. Opposition to gun control, environmental regulation and abortion rights; support for the flat tax, tort reform, capital punishment, and a belligerent foreign policy--all of those positions follow like logical theorems from the strict father model, just as the logic of the nurturant parent model entails support for civil liberties, progressive taxation, a broad social safety net, an accommodating foreign policy, anti-smoking ordinances, restrictions on pesticides, and strict rules on seat-belt use. Now it isn't unreasonable to suppose that political orientations have their genesis in early socialization, as Maslow, Lasswell, Erikson, and Adorno among many others have argued. But Lakoff is claiming more than that. He reduces the model of the family to two broad types, which determine one's position on a broad range of specific political issues, and which coincide precisely with the poles of early-twenty-first-century American politics. Why should the "nation-as-family" metaphor be paramount in dividing the sides, and why should everything follow from that one schema? Lakoff points out that we often talk about the nation as a family--we speak of the Founding Fathers (or used to, anyway), send our sons off to war, and so on. But as Pinker notes, people choose their metaphors judiciously, and can bail out on them or switch to a new one when they don't fit a particular situation. In fact there are any number of metaphorical schemes that we use to talk about the nation, many of them with very deep roots in the language and in history. As I put the point in Talking Right:

The nation can be a body with a head, stomach, heart, and arms, or a person who is young, grows old and sick, and dies. The nation can be a ship, as poet Walt Whitman portrayed it, which sails on, loses its moorings, drifts, or has to be righted. It can be a theater, where people and issues wait in the wings, take center stage, or lay an egg. It can be a house, crumbling at the foundations or built to withstand the buffeting of the winds. It can be a city, as both Ronald Reagan and Mario Cuomo described it, though with different images of what was going on in its various neighborhoods. It can be a party (to which everyone must be invited). Or as David Brooks suggests, it can be a high-school cafeteria where each clique has its table. And sometimes it's just a nation: not all the vocabulary we use to talk about national life has its conceptual origin in some other domain.
So why should we give primacy to the nation-as-family metaphor? Lakoff doesn't give any direct evidence for that hypothesis: no surveys, interviews, case studies or ethnographic investigations; no database counts or empirical investigations of language use; no historical or contrastive analyses; no experiments that support the centrality of the family metaphor over others. His analysis rests entirely on a kind of rational reconstruction: by systematically working out the entailments of the two pictures of the family, Lakoff says, he can show how each position follows from the basic model.

In 1994, I dutifully read the "Contract with America" and found myself unable to comprehend how conservative views formed a coherent set of political positions. What, I asked myself, did opposition to abortion have to do with the flat tax? What did the flat tax have to do with opposition to environmental regulations? What did defense of gun ownership have to do with tort reform? Or tort reform with opposition to affirmative action?... The answer is that there are distinct conservative and progressive worldviews. The two groups simply see the world in different ways.

... I worked backward. I took the various positions on the conservative side and on the progressive side and I said, "Let's put them through the [family] metaphor from the opposite direction and see what comes out." I put in the two different views of the nation, and out popped two different models of the family: a strict father family and a nurturant parent family.
The argument, in short, rests on two assumptions: first, that there is a conceptual coherence to the conservative worldview, and second, that you can uncover it if you can tell a story (or "build a model," as linguists like to put it) where everything fits into place. That's the methodology which Chomsky taught linguists to use in explaining the facts of English syntax, and which Lakoff himself deployed with great skill in his early research on syntax. But its application to the meaning-saturated facts of political culture is more questionable. To take one example, Lakoff's suggests that conservatives believe abortion is wrong because "pregnant teenagers have violated the commandments of the strict father" and "career women [who have abortions] challenge the power and authority of the strict father," and hence "should be punished by bearing the child...." But he gives no evidence at all for that claim, which as it happens runs counter to the much more complex and sympathetic picture that emerges from ethnographic studies and interviews with rank-and-file pro-lifers that have been reported by liberal writers and researchers like Tom Frank, William Saletan, and Carol Maxwell. (As Maxwell observes, those people are less likely to think of pregnant women seeking abortions as sinners who must be punished by carrying their children to term than as women who have been "abandoned by the men who impregnated them or by the men who employed them without options for maternity--and duped by society into believing their best, or their only option was to abort.") It's enough that Lakoff's account of conservative motives fits his picture of how they would think if they conformed to his strict father model in a systematic way. QED. Lakoff's account of the motives of anti-abortion conservatives exhibits one other unfortunate feature of the way he goes about his analysis. As Pinker puts it, "cannot stop himself from drawing horns on the conservative portrait and a halo on the progressive one." To take another example, he describes conservatives as being "against nurturance and care. They are against social programs that take care of people." It's a picture that's calculated to flatter the self-righteousness of many liberals, particularly if they number few actual conservatives among their acquaintances. And it goes at least part of the way to explaining the appeal that Lakoff's work has had among many liberals who haven't looked at it critically or who are dazzled by terms like "cognitive" and "neuro-." But in its way, the story as loaded and reductive as the stereotypes of pretentious and supercilious liberals that you hear from the likes of Ann Coulter and Laura Ingraham. But that by itself isn't a criticism of Lakoff's method, just of the unfair way deploys it. And there are deeper problems here. Even if Lakoff had characterized the two worldviews more evenhandedly, the reductiveness of his analysis would disqualify it as a serious account of American political divisions. I suspect that one reason for Lakoff's insistence on reducing political orientations to two basic family models is that he shares the theoretical ambitions of modern Chomskian linguistics, where the name of the game is to come up with universals of grammar that underlie the structure of every language. To most linguists, a fact about English or Yoruba is of no theoretical interest unless it's rooted either in some specifically linguistic mental mechanism or in some universal and eternal principle of human cognition. And clearly a model of political morality based in family experience is a better candidate for universal or at least very wide cross-cultural status than one based on houses or theaters (in Don't Think of an Elephant, Lakoff traces the origins of his contrasting models of the family to prebiblical times). The reductiveness of Lakoff's story is compounded by the implicit determism. It doesn't simply turn political orientation into a wholly owned subsidiary of moral character. It also implies that, locked as we are into our worldviews, we'll respond automatically to language that triggers the right moral frames. This helps to explain why Lakoff makes such a big deal out of the claim that the frames that underlie political reasoning "are in the synapses of our brains, physically present in the form of neural circuitry." As Pinker points out, that's a singularly uninteresting observation, since

every thought we think--permanent or transient, rational or irrational--is instantiated physically in the brain. The implication that frames, by being "physically fixed" in the brain, are especially insidious or hard to change, is gratuitous.
To put this in other terms, there's nothing you can say about the neural basis of our moral frames that you can't also say about our knowledge of the infield fly rule or the lyrics to "Zippity-Doo-Dah." But until you think it through, the claim leaves you with the impression that political language works on us in an unconscious and irresistible way. That isn't an unusual point of view in the age of Orwell, not to mention of linguistic consultants who promise marketers that they can come up with strings of syllables that can magically put a new brand on the supermarket shelves. And that picture is a lot more compelling if you think that our political reasoning involves the application of a single model, rather than a welter of competing metaphors and schemas that people can pick and choose among. Even if it were plausible to assume that Lakoff's two models of the family were universal or eternal and that they underlay all our political reasoning, you'd have a problem getting from those general models to very specific positions that Lakoff associates with modern liberal and conservative worldviews. After all, the positions that define the two poles have shifted considerably since the modern liberal-conservative opposition was first fixed in the New Deal period, and still more compared to the way the corresponding sides were defined in the Progressive era before that. And the mix of cultural, political, and economic features that define contemporary American conservatism and liberalism diverge considerably from the way the left-right poles are sorted out in other nations. If being pro-choice follows from adopting the nurturant parent model, why has it been a defining characteristic of the "liberal worldview" only for a few decades, and why does abortion play no role whatsoever in defining the sides in British politics? The effect of this is to make the positions that people associate with each of the poles look a lot more consistent and universal than they appear when you consider them either historically or comparatively. As Lakoff tells it, what passes for "conservatism" isn't simply an opportunistic amalgam of positions that reflect the various constituencies that were awkwardly joined under Nixon's Southern Strategy--Southern whites, working-class urban ethnics, religious conservatives, corporate interests, and country-club Republicans. It's a coherent conceptual scheme. And similarly for the various constituencies associated with the liberal label. True, Lakoff acknowledges that not all Americans subscribe to all of the positions that we associate with either pole. But his answer is to claim that people who support conservative positions on some issues and liberal ones on others are "biconceptuals" who are operating with "mixed models," unlike both the unremittingly nurturant progressives who populate the blue-state cities and the relentlessly strict-father conservatives of the red-state evangelical redoubts. (For example, he speaks of "blue-collar workers [who] are strict fathers at home, but nurturant toward their coworkers.") In fact Lakoff denies that there can be any such thing as a conceptually coherent centrism:

There is a common belief that there is an ideological "center"--a large group of voters either with a consistent ideology of their own or lined up left to right on the issues or forming a "mainstream," all with the same positions on issues. In fact, the so-called center is actually made up of 'biconceptuals', people who are conservative in some aspects of life and progressive in others.
You don't have to be an ardent devoté of Third Way politics--I'm certainly not--to be suspicious of the implication that only left-liberal Democrats and conservative Republicans have the right to claim ideological purity or moral consistency. And it's notable that Lakoff's picture here is the perfect mirror image of what the right says in its criticisms of moderate Republicans--that there's something inconsistent and wishy-washy about supporting free trade, say, while being strongly pro-environmental. That isn't the only resemblance between Lakoff's story and the one that the right tells about American political polarization. What he's saying, in the end, is that liberals and conservatives are distinct kinds of people, whose differences are embedded in personal psychology. That's exactly what the right has been saying all along--that the country is divided not along lines of class or power, but by "cultural" differences. And like the right, Lakoff locates the prototypical liberal among upscale urban professionals. True, he never says that explicitly; in fact he doesn't tie the difference he draws between strict-father conservatives and nurturant-parent liberals to any social factors at all. But he depicts nurturant-parent liberals as people who don't believe in spanking children, find work a means of realizing their contribution to their community and developing their potential, believe that both parents should share equally in child rearing, and want to protect their children from cigarettes, cars without seat belts, dangerous toys, inflammable clothing, pesticides in food, and unscrupulous businessmen--all of them attitudes more typical of the Berkeley Hills than of working-class, black, or Latino neighborhoods. In effect, this confirms the stereotypes that conservatives have been successfully hawking for the last 30 years, which have made the phrases "working-class liberal" or "black liberal" nearly inexistent in the American media. In fact you wouldn't guess from Lakoff's analysis that the liberal-conservative divide might have its roots in divisions of class or power, or that wealth is still the best predictor of political orientation, as Nolan McCarty, Keith Poole, and Howard Rosenthal have recently documented in Polarized America. (In this connection, it's worth noting that Pinker doesn't raise any objection to Lakoff's general claim that the liberal-conservative distinction is fundamentally rooted in differences in cognitive attitudes, though he suggests it might be better drawn in terms of a "discipline-compassion dimension." As good cognitive scientists--and in particular, good Chomskians--both writers naturally look for the primary genesis of political orientation in underlying psychological dispositions, rather than the social and economic conditions that shape us politically.) True, Lakoff notes at various points that the positions that follow from the strict-father morality and the view of freedom that it entails work to the benefit of the rich:

What [conservatives] have done is to create, via framing and language, a link between strict father morality in the family and religion on the one hand and conservative politics on the other. This conceptual link must be so emotionally strong that it can overcome economic self-interest....

Conservative political and intellectual leaders ... represented an economic and political elite, but they were seeking the votes of middle- and lower-class working people. They needed, therefore, to identify conservative ideas as populist and liberal/progressive ideas as elitist--even though the reverse was true. They faced a massive framing problem, a problem that required a change in everyday language and thought. But strict father morality gave them an important advantage: It suggests that the wealthy have earned their wealth, that they are good people who deserve it.
Now if you followed that line of argument rigorously, you might conclude that the entire "strict-father" morality business was just a line of patter aimed at getting working-class Americans to sign on to Republican economic programs. But in that case the unity that Lakoff professes to discern in conservative thought would be merely rhetorical, not "conceptual"--what we'd be talking about isn't "deep moral frames" or broadly competing worldviews but simply specious connections among positions that have nothing essentially in common. Probably for this reason, Lakoff never addresses the difference between ideas and ideology--between sincere moral tenets and the rationalizations and pretexts that Republicans offer for self-serving policies. In fact he often seems to want to have it both ways, occasionally talking about "conservative morality" as a shell game, but more often presenting it as a deep-seated conviction. The real reason for Republicans' success, he says, is that "they say what they idealistically believe." And again: "Many of the ideas that outrage progressives are what conservatives see as truths-presented from their point of view." In fact Lakoff's picture of conservative rhetoric isn't really like the one which Tom Frank gives in What's the Matter with Kansas? or which I outline in Talking Right. What he is suggesting is not simply that working-class and middle-class Americans often vote for Republicans because their concerns about values issues and national security trump their concerns about their own immediate economic problems, but that they actually buy into the Republicans' rhetoric about the virtues of wealth. As he puts it in a revealing passage in Don't Think of an Elephant:

In the 2000 election Gore kept saying that Bush's tax cuts would go only to the top 1 percent, and he thought that everyone else would follow their self-interest and support him. But poor conservatives still opposed him, because as conservatives they believed that those who had the most money-the "good" people-deserved to keep it as their reward for being disciplined.

But there's no evidence at all that this is true, and lots to suggest that it's simply wrong. In polls, a great majority of working- and middle-class Americans say that Democrats will do a better job on issues like taxes and social security, and associate the Democrats with values like fairness. And if some of those people wind up voting the Republicans, it isn't because they believe that the rich are morally entitled to keep their wealth. It may be because they don't have much faith in the power of the Democrats to address their concerns, as Karl Agne and Stanley Greenberg have argued, or because they personally don't stand to gain much from existing government programs, as Stephen Rose has suggested. Or it may be because they simply find values issues more compelling or comprehensible than economic ones. But until now, it's only Republican ideologues who have argued that working-class Americans really take moral satisfaction in the increasing wealth of the top one percent. The difference is that when that claim comes from the right, it's a politically self-serving canard, whereas coming from Lakoff it's simply an effort to make the facts fit his theory. Why does any of this matter? Pinker suggests that the danger is that Democratic politicians might actually take Lakoff at his word and build their strategies around his ideas. But as best I can tell, Lakoff's direct influence on the language of the Democrats has been negligible. He may have had the ear of some prominent Democrats, but you couldn't tell it by what comes out of their mouths. And no wonder--as Pinker and a number of other people have observed, Lakoff's own framing suggestions are pretty lame. Democratic politicians don't need to know anything about cognitive science to realize that referring to taxes as "membership fees" or to trial lawyers as "public protection attorneys" would make them easy targets of Republican ridicule. And as for his proposal that Democrats should reframe "activist judges" as "freedom judges," a Google search turns up no instances of the phrase apart from remarks that make fun of the suggestion. True, linguists coin slogans about as well as physicists ride bicycles. And the fact that Lakoff has a tin ear for political phrasing doesn't negate his indirect influence in drawing Democrats' attention to the importance of framing. That's all to the good. It's easy to say that what matters is ideas, not language. But while people often exaggerate the effect of Republican slogans and bumper stickers, there's no question that a well-turned catchphrase can do a lot of work in shaping public opinion--think of "cut and run." As Walter Lippmann pointed out in Public Opinion, American political life is saturated with verbal symbols that "assemble emotions after they have been detached from their ideas." However compelling the ideas that Democrats come up with are, they'll have a hard time packaging them unless they can do a better job confecting the wrapping paper. (My own sense is that liberal Democrats would do better revisiting the populist language that brought them to the ball in the first place than invoking the labored moral frames that Lakoff proposes. But that's for another conversation.) But ultimately, I think Lakoff's political importance has to be reckoned not by his effect on the party professionals, but by his appeal to rank-and-file liberals, progressive activists, liberal bloggers, and the like. (In the press, Lakoff has gotten only about a quarter as many mentions as James Carville over the past two years. On the Internet, the two run neck-and-neck.) And here the effects of Lakoff's thinking may not be so salubrious. Over the last few decades, the right has managed to reconfigure the polarities of American politics so that economic divisions are trumped by the bogus cultural distinctions of the "red-blue" divide, and in the process "liberal" and "conservative" have been redefined as opposing social styles or personality types, rather than as contrasting philosophies of government. Indeed, listening to the talk shows on Fox News, you might have the impression that the two sides are really distinct political genders. As it happens, that picture of an America riven into two distinct nations--"more divided than at any time since the Civil War," as people sometimes say--has no empirical reality for the mass of ordinary Americans, as researchers as politically diverse as Alan Wolfe and Morris Fiorina have shown at some length. But once you start thinking of liberals and conservatives as distinct kinds of people, divided by deep moral differences that grow out of their early family experience, then it's easy to fall into the hyper-moralizing rhetoric of political polarization. True, many liberals have always been prone to this tone of argument. But Lakoff's writings seem to give a scientific imprimatur to the idea that liberalism and conservatism are distinct mentalities--that we're the ones who are "for nurturance and care," for example. And while liberals may find that picture flattering, it also plays into the rhetorical hands of conservatives, who are happy to reframe ideological divisions as warring personalities and lifestyles and to obscure the economic roots of political divisions. In fact the most damning thing you can say about Lakoff is that he too often takes the right at its word.

Source URL: