I guess it's now officially a Media Meme: Obama's "royal we has flowered into the naked 'I'". First Terence Jeffrey ("I, Barack Obama"), then George Will ("Have We Got a Deal for You"), now Stanley Fish ("Yes I can"):
By the time of the address to the Congress on Feb. 24, the royal we has flowered into the naked “I”: “As soon as I took office, I asked this Congress.” “I called for action.” “I pushed for quick action.” “I have told each of my cabinet.” “I’ve appointed a proven and aggressive inspector general.” “I refuse to let that happen.” “I will not spend a single penny.” “I reject the view that says our problems will simply take care of themselves.” “I held a fiscal summit where I pledged to cut the deficit in half by the end of my first term.” That last is particularly telling: it says, there’s going to be a second term, I’m already moving fast, and if you don’t want to be left in the dust, you’d better fall in line.
There’s no mistaking what’s going on in the speech delivered last week. No preliminary niceties; just a rehearsal of Obama’s actions and expectations. Eight “I”’s right off the bat: “Just over two months ago I spoke with you… and I laid out what needed to be done.” “From the beginning I made it clear that I would not put any more tax dollars on the line.” “I refused to let those companies become permanent wards of the state.” “I refused to kick the can down the road. But I also recognized the importance of a viable auto industry.” “I decided then…” (He is really the decider.)
The trouble with this idea, as often with the insights of the punditocracy, is that there's no evidence that it's true. Worse, evidence is easily available to disconfirm it.
In yesterday's post ("Fact-checking George F. Will"), I counted, and discovered that in that "speech delivered last week" — the one about the auto bailout — President Obama used forms of the first-person pronoun at a rate of 1.7% (i.e. 42 instances in 2423 words). Compared to Obama's press conference of 2/9/2009 — before the 2/24/2009 date that Prof. Fish identifies as the pronominal turning-point — this is a lower rate, not a higher one: the 2/29/2009 press conference exhibited a rate of 2.6% (205 of 7,775).
Furthermore, if we compare the first press conferences of the previous two presidents, we find higher rates yet: 3.9% for Bill Clinton (275 in 6,935) and 4.5% for George W. Bush (300 in 6,681).
You'd think that Prof. Fish would have done some counts of this sort before displaying his analysis in the pages of the New York Times, especially since he's explicit about the fact that he's inferring the president's attitudes by counting (and analyzing) pronouns rather than by using ESP. Unfortunately, at some point in the past few decades, literary scholars seem to have abandoned the assumption that claims, even quantitative ones, ought to be testable.
But maybe I'm being unfair. Fish explicitly compares Obama against Obama's own earlier practice, and only implicitly against other recent presidents. (And speeches are a different genre from press conferences.) So I guess I should go back and look at some of Obama's pre-presidential transcripts as well. Any bets on what I'll find?
Here goes. Fish mentions 8 of Obama's speeches. I've arranged them in chronological order below, and given each one a link to a transcript. (I've added one that Fish didn't mention: Obama's first presidential press conference, which I analyzed yesterday, because it provides a useful point of comparison to the first press conferences of previous presidents.) The rightmost column gives the percentage of words in the transcript that are first-person singular pronouns (I, me, my, myself, mine):
|1||the presidential announcement||2/10/2007||2.1%|
|2||the Iowa victory address||1/3/2008||1.9%|
|3||the Philadelphia speech on race||3/18/2008||1.8%|
|4||the nomination acceptance||8/28/2008||2.5%|
|5||the Grant Park victory speech||11/4/2008||1.9%|
|6||the inaugural address||1/20/2009||0.2%|
|7||the first presidential press conference||2/9/2009||2.6%|
|8||the address to congress||2/24/2009||1.6%|
|9||the GM bankruptcy speech||6/1/2009||1.7%|
(Recall again that in comparison, William J. Clinton's first two presidential press conferences had a rate of 3.9%, while George W. Bush's first two presidential press conferences had a rate of 4.5%.)
Prof. Fish tells us that there's a "pattern of pronouns" in Barack Obama's speeches, whereby an increase in the use of 'I' and 'me' and 'my' marks his transition from the "restraint and modesty" of a politician whose "self-assertion is immediately muted" to the "imperial possession" of a "guy [who]'s completely in charge, making decisions, giving instructions, deploying resources, assigning tasks — a combination point guard, quarterback and clean-up hitter." And according to Fish, the "flowering" of the "naked 'I'" is first evident in the 2/24/2009 address to Congress, and "the note of imperial possession" is fully on display in the 6/1/2009 GM bankruptcy speech. Maybe so — but in fact, the rate of use of first-person singular pronouns in those two speeches is lower than in *any* of the six earlier speeches that Fish cites — except for the inaugural address, where first-person singular pronouns were almost completely absent, presumably for stylistic reasons.
Maybe president Obama is now "completely in charge, making decisions, giving instructions, deploying resources, assigning tasks". That's more or less what the leader of the executive branch of government is supposed to do, after all. But his (or his speechwriters') rate of use of first-person singular pronouns hasn't increased — in fact, maybe it's gone down.
[In fairness to Fish, he doesn't make his case only or even mainly on a simple count. He suggests a distinction between the earlier speeches where "Both syntactically and substantively, the 'I' is subordinated to the projects to which it is dedicated", and the recent speeches, where the "royal we" emerges in the inaugural, and then "flowered into the naked 'I'" in the address to Congress.
On the other hand, he does wave some numbers around ("Eight “I”’s right off the bat … Accompanying the “I”’s are a bevy of “my”’s"), so it's worth establishing that simple pronoun counts undermine his argument rather than supporting it.
And after giving the texts a modest amount of scrutiny, I can't see any form of (inter-subjectively stable) linguistic analysis that would support Fish's notion that Obama's earlier uses of 'I' are "syntactically and substantively … subordinated to the projects to which it is dedicated". Perhaps some reader can figure out how to turn this idea of Fish's into a testable claim, in a way that would support his argument.
Until then, I'm going to assume that Stanley Fish has joined Terence Jeffrey and George F. Will in a classic display of confirmation bias: they feel that president Obama is arrogant and uppity, and so his uses of the first-person singular pronoun are striking and salient to them, in a way that previous presidents' pronouns were not. I freely admit, of course, that I have only a little more evidence for this opinion than they do for their views on Obama's egotism. ]