« Rhetorica Podcast... | Main | Long past stupid... »
December 20, 2005
Guilt by citation...
Tim Schmoyer has asked that I comment on this study of bias in the news media. My overall impression is this study is a good attempt to quantify something of cultural and political importance. I do not think it measures or proves media bias. How could it?
The study measures citations of "think tanks" in articles and compares those citations to the ADA scores of politicians who cite the same organizations in speeches. Despite an excellent attempt to bridge the gap between citation and intention, the study fails to show a necessary link between citation and bias--even by the authors' bracketing of the definition of "bias." How might they have accomplished this? I suggest applying my (revised) formula for the illocutionary act: Fr(p) / C ->PE (one way to get at it, not the only way...and BTW, my thanks to Tim for his help in refining this formula). F = the force of a statement--what we are doing when we utter it, i.e. asserting, directing, commiserating, expressing, or declaring. The exponent r represents the "rheme"--the unit(s) of rhetoric, i.e. those rhetorical forms chosen by the speaker to make the message persuasive. (p) = the propositional content of the statement. And, finally, we must divide by the context--the rhetorical situation--to separate the speech act from other potential situations. That leads us towards a perlocutionary effect, i.e. what happens in regard to the statement.
Okay, that was a long-winded and overly-academic bit of horn blowing. The point is this: A causal connection must be proven by more than circumstance. Groseclose and Milyo make a good attempt. I don't think it's good enough. This doesn't mean their conclusions are wrong, merely not proven.
There are all kinds of biases in the news media. All of them are political in the sense that someone, some interest, or some group is harmed or hurt politically by every message. There's no such thing as a message that doesn't help or hurt politically. Think "Happy Holidays" here if you're looking for a particularly odd and exasperating recent example.
Can we identify liberal and conservative bias in various news outlets? Of course we can. As I have always said, political bias is a very real local phenomenon--local to a person, an issue, a news organization, etc. But this does not suggest that the news media in general are liberal or conservative or anything else. The "news media" is an awfully big and complex thing. It includes The New York Times and the Podunk Weekly Bugle.
The study was peer reviewed and published. But I doubt someone with my credentials reviewed it. But, then, it was published in an economics journal, and I'll be the first to tell you I haven't a clue about that discipline.
I think it is, however, a valuable contribution. These guys really try to connect a textual feature with something like intention--which means they are doing exactly the kind of work I think is valuable.
Charges of political bias don't get you very far toward understanding the news media. Saying The New York Times is a liberal paper (and by Daniel Okrent's definition I agree to a certain extent) doesn't help you understand its culture or behavior in a way that's predictive of journalistic behavior. Want to know what makes journalism really tick? Click here.
Posted by acline at December 20, 2005 10:12 AM | | Spotlight