September 04, 2004

Frames and messages

George Lakoff's ideas about "the framing wars" have started to find their way into public discourse, though I've complained that the media (at least Bill Moyers) presents the issue as being about the choice of words rather than about the choice of ideas. You can find a clearer statement of George's perspective in this 9/1/03 article from the American Prospect, and in these two interviews from the UC Berkeley News. He thinks that the Republicans have "out-framed" the Democrats over the past couple of decades, and that seems to be the truth of the matter.

However, Lakoff is not talking about ideas as rationally-constructed opinions, convictions or principles, but rather about metaphors, images, and evoked scenarios with emotion-laden roles like victim, hero, villain, crime, strict father and so on. Although he has partisan goals, this mode of analysis is politically neutral, at least with respect to current American political parties. It emphasizes attention to emotion instead of substance, but you can apply that emphasis to promoting any goals you want.

I'm somewhat skeptical about the degree of separation between substance and presentation that this approach assumes. It's traditional in the advertising industry, and for that matter in politics. But it seems to me that political movements -- and for that matter, advertising campaigns -- succeed best when style and substance are integrated. That's certainly what Michael Silverstein suggests in his analysis of how Abraham Lincoln's rhetoric came to form part of America's "civil religion".

Like Lakoff, Silverstein is certainly partisan. His pamphlet Talking Politics contains a certain amount of bile directed at George W. Bush, as the author of Semantic Compositions discovered by reading a few pages via Amazon's "Read Inside the Book" feature. However, I think this misses the point. Like Lakoff, Silverstein is promoting a mode of analysis that applies to any human communication, political or otherwise. What he has to say about image, style and message can be applied to positive or negative evaluations of any politician, or for that matter to your relatives, your friends or yourself.

In order to give a sense of what we're talking about, I've typed in a highly elliptical version of Silverstein's introduction to what he means by "message" -- this represents about a fifth of the stretch of text from which the quotes are derived, with most of the elaborations, exemplications and asides left out:

Those not attuned to politicoglossia may at first think that someone's "message" is the topic, or theme, or cetnral proposition he or she is trying to communicate. ... You could paraphrase someone's "point" as a kind of assertion that such-and-such is the case about something-or-other.

You would be wrong. ... "Message," we can discern from the study of political communication, is really much more complicated than that. If successful, a person comes to inhabit "message" in the act of communicating.

... In order to understand "message," ... we have to think about the several different kinds of meaningfulness always present -- though not always recognized -- when language is used. ...

In our own intellectual tradition of understanding how people use language, the most salient -- the official -- "what" of communication lies in how words and expressions describe, or in technical terms, denote. ...

So officially we describe things and states-of-affairs so that others can also identify those things and states-of-affairs. ...

But additionally, there are principles based in developing information-structure itself, distinct from the grammar of sentences, that determine what expressions we can and do use at various points in communication ... while communication proceeds, sender and receiver can rely more and more on what has already been communicated about a topic, information about it that cumulates between them. ...

In this way, discourse is always being evaluated as description for how it achieves a kind of cumulative coherence as information. ... people can use language to construct collectively reached and collectively consequential knowledge, opinion, and belief about all manner of things. ...

But, having mentioned both grammar and information structure, is there anything else to communicative use of words and expressions? ... [I]n every discourse a large number of extra-verbal contextual factors leave their determinate traces in the forms we use -- what are termed in the trade indexical (pointing) traces. These traces inform us about, they point to, the who-what-where-when-why of discourse by subtle loadings of the "how", the actual forms, of discourse. ("Democratic" or "Democrat"? ...) ... Indexical values of language forms locate and identify the parties to the communication ... the way a good pantomime gives the impression of taking place in a comprehensible surround.

These indexical factors in language seem to crosscut the information structure always emerging via grammar and denotational coherence as speakers add to the words and expressions in a text. masters of political "message," just like other users of languages, have intuitively known all along about the indexical power of the words they use, and especially about the cumulative indexical poetry of poperly arranged words. Such masters have a knack for indexical design that has shaped each era's political communication -- at least as much as the descriptive content of it ... -- thus creating a true rhetorician's art form. ...

... In communicating we ... rely on social arrangements already in place, and the expectations we can then have about what form talk should take between two socially locatable individuals. But as well, each time we deploy specific forms of langauge we create social arrangements as consequences of using these forms; we bring new social arrangements into being.

... [T]he act of communication itself, that is, the emergence of certain indexically potent message forms, can always transform the intuitive classifications we apply to one another, new ones suddenly pointed to as now operative and consequential ...

What indicative signs and signals, for example, were you relying on in your aunt's talk when you concluded, the other day, that she was "stressed"? ... Again, how did I come to know that the prospective student in my office last week was gay? He did not announce this to me as a self-description, explicit or implicit. He just talked about -- described, in the sense I discussed earlier -- why he was interested in a particular educationational degree program. These kinds of inferential processes go on constantly in interaction, as we all know, on the basis of indexical signals that work like gestures in pantomime.

In essense we continuously point to our own -- and, relationally, then, to our interlocutor's -- transient and more enduring identities. Interactions as events develop these relational identities as consequences of communicative behavior. The clarity of identities comes in phases, punctuated by shifts over interactional time: what-you-and-I-are in a moment of interaction strives to become what-you-and-I-will-be. ...

Over multiple indexical channels, then, there comes into being a kind of poetry of identities-in-motion as the flow of communicative forms projects around the participants complex patterns -- let's say "images" -- not onto Plato's case wall, but onto the potentially inhabitable and then actually inhabited context. So there is image. There is style. There is "message". Image is not necessarily visual; it is an abstract portrait of identity ... Style -- the way image is communicated -- has degree and depth of organization... "Message", then, strategically deploys style to create image in a consequential way. ...

So being "on message" contributes to that consistent, cumulative, and consequential image that a public person has among his or her addressed audience. A really powerful "message" ascribes to me -- as opposed to describes -- my reality.

Leaving politics for a while, Silverstein points out that "these demonstrations of and inferences about identities" are central to human communication, but "have been largely out of the aware consciousness of communicators". And he observes that

[A]ll of the institutionalized technologies of languages have cumulatively reinforced this intuitive difficulty of explicit recognition by concentrating on its descriptive functions. ... I mean everything from the writing and printing conventions to the personnel and paraphernalia of enforcing standard languages: dictionaries, thesauruses, grammars, manuals of style, and the people who create them and insist that they are authoritative." ... The biases, built into our institutional forms across the board, keep telling us to discount what is actually indispensible to normal and effective human communication.

Although I'm just as interested in "the poetry of identities in motion" as Silverstein is, I do think that there are some good reasons for the bias in favor of the "descriptive functions" of language. The key factor is the psychological phenomenon of word constancy. We can all pretty much agree on what words someone said, and to a lesser degree on how the words go together, and what states of affairs are consistent with them. As we get further away from that level of description, things tend to get fuzzier and fuzzier. And I'd contend that this is the cause, not the consequence, of the way that writing systems work.

However, Silverstein is absolutely correct to observe that we often

.. hyperemphasize the use of language for descriptive purposes, sometimes foolishly and vainly attempting to disregard the inevitable, simultaneous use of language for inhabiting identities. And, of course, just such ways of fashioning inhabitable identities in communication give our messages whatever life-like appeal they may have.

Returning to the analysis of political discourse, it's worth observing that there is another way to think about these things, due originally to Aristotle: effective rhetoric is an amalgam of ethos (the character of the speaker), pathos (the emotions of the audience) and logos (the content of the argument).

Lakoff's "frame wars" are all about pathos: choosing metaphors that line the terms of the debate up with the emotions of the audience in a favorable way.

"Message", for Silverstein, "strategically deploys style to create image", and thus is basically about the projection of ethos.

Silverstein worries that there is an on-going "adjustment of the ratio of operative meanings ..., the denotational and the context-indicating", such that "[t]he key expressions are no longer experienced ... as signals of concepts with which we communicate denotational truth-and-falsity", but are merely "[conjuring] up a kind-of-'who' at a certain cultural 'where'." He associates this with "a corporate-standard language register".

It seems unfair to me to associate this kind of pointillistic display of identity-indexicals with modern corporations. They often use it, sure enough, but so do protest groups, NGOs and scientific societies, or musicians, computer programmers and university professors -- at least some of the time. At other times, the same organizations and individuals may achieve the integration of ethos, pathos and logos that Aristotle (and Lakoff and Silverstein) would recommend.


Posted by Mark Liberman at September 4, 2004 10:59 AM