Monday, August 28, 2006

Cherry-picking passives

Last week I went to a seminar by Edward Tufte on Presenting Data and Information. I’ve been interested in Tufte for a couple of years now. Especially since much of my job involves communicating statistical evidence. Despite the fact that the Boston Globe referred to one of his books as "A visual Strunk and White,” I am still impressed by his books. (In fact, much to my chagrin, Tufte seems particularly enamored of Strunk and White—but I’m getting ahead of myself.) I particularly like his stuff on why Power Point is a bad presentation tool.

I want to say up front that I still really like his work and the presentation was well worth it. If you ever get the chance to see this guy, definitely do it.

That said, the presentation kind of veered between a pep rally, a lecture, and a kind of witch hunt. I’m not sure how to describe this last bit. Basically it was a kind of in-group sneer at people who should have known better. Together, we got to look down on some poor writer with the implicit understanding that WE would never make that mistake. This type of group think is one of the more difficult things for me to sit through.

And right in the middle of one of these sections, we lookied at the chapter, Corruption in Evidence Presentations: Effects without causes, cherry-picking, overreaching, chartjunk, and the rage to conclude in his latest book, Beautiful Evidence. In this chapter, Tufte makes the argument that writers are morally responsible for getting their facts right and not spinning the data. This is something I’m sympathetic too. But on p. 142, at the beginning of the chapter, he makes an analogy with writing style. In particular he goes after the oft-maligned passive:

…passive verbs also advance effects without causes, an immaculate
conception. To speak of ends without means, agency without agents, actions
without actors is contrary to clear thinking.
...

The technique of evasion by passive voice is well-known and widely noted—for example, in Strunk and White, Elements of Style—and yet some reviewers of The 9/11 Commission Report failed to detect its overt evasive deflections of responsibility.

[Despite the CIA’s numerous warnings, America’s] domestic agencies never mobilized in response to the threat. They did not have direction, and did not have a plan to institute. The borders were not hardened. Transportation systems were not fortified. Electronic surveillance was not targeted against a domestic threat.54 State and local law enforcement were not marshaled to augment the FBI’s efforts. The public was not warned. [From The 9/11 Commission Report (p. 282 of the online PDF)]

Here and elsewhere in the Report, conspicuously absent is the agent of inaction. above, 5 verbs are passive. The 3 active verbs take the utterly vague subject “domestic agencies.” Exactly who did not make a plan, who did not follow up, who failed to warn the public? “These things that were not done must have been not done by somebody, and the somebodies reporting to him.”2 By means of the passive voice, the 9/11 commission evades attributing responsibility for security lapses. Of course, if the passive didn’t exist, they would have done it some other way.
My initial reaction to his expert from the 9/11 report was that this paragraph wasn’t about the somebodies who failed to act; it was about the things that were not done. So Tufte jumping on the passive as a means to hide culpability was stretching it a bit. Not to mention the fact that the passive voice switches subject and object, not agent and patient and that the passive voice does allow the subject to be expressed in an optional by-phrase.

I wondered if Tufte wasn’t guilty of a little cherry-picking himself. Sure enough a quick look at the actual report showed he was.

The piece he excerpted is in the chapter “The System Was Blinking Red” in a section called “Government Response to the Threats.” It is in fact a summary paragraph for that section. In the pages before the paragraph, the following government agencies are mentioned by name as having failed to act: CSG, NSC, FBI, CIA, FAA, and INS. Given the context of the piece, it’s clear who those somebodies were that didn’t do anything.

Here are the two paragraphs preceding Tufte’s paragraph.
Attorney General John Ashcroft was briefed by the CIA in May and by Pickard in early July about the danger. Pickard said he met with Ashcroft once a week in late June, through July, and twice in August. There is a dispute regarding Ashcroft’s interest in Pickard’s briefings about the terrorist threat situation. Pickard told us that after two such briefings Ashcroft told him that he did not want to hear about the threats anymore. Ashcroft denies Pickard’s charges. Pickard says he continued to present terrorism information during further briefings that summer, but nothing further on the “chatter” the U.S. government was receiving.

The Attorney General told us he asked Pickard whether there was intelligence about attacks in the United States and that Pickard said no. Pickard said he replied that he could not assure Ashcroft that there would be no attacks in the United States, although the reports of threats were related to overseas targets. Ashcroft said he therefore assumed the FBI was doing what it needed to do. He acknowledged that in retrospect, this was a dangerous assumption. He did not ask the FBI what it was doing in response to the threats and did not task it to take any specific action. He also did not direct the INS, then still part of the Department of Justice, to take any specific action.

In sum, the domestic agencies never mobilized in response to the threat. They did not have direction, and did not have a plan to institute. The borders were not hardened. Transportation systems were not fortified. Electronic surveillance was not targeted against a domestic threat.54 State and local law enforcement were not marshaled to augment the FBI’s efforts. The public was not warned.
Another interesting point is that Tufte replaced the original “In sum, the” with “[Despite the CIA’s numerous warnings, America’s]”. That change completely alters context of the paragraph.

Here’s a Tufte quote from the same chapter that sums up what I feel.
From scientific reports to political speeches, few things are more appalling than listening to inept and specious arguments made by one’s allies.

2 comments:

Anonymous said...

It disappoints me to see that Tufte engages in this kind of ... well, hypocrisy. How many of the other seminar attendees (or Tufte book readers) will check up on his sources and find stuff like this? Not any, likely, and certainly not many.

boredoom said...

Good spotting! I think the aversion to passive voice is sometimes taken to silly extremes in journalism in particular. In some instances, we don't know who did something, so we're forced to invent words that carry no information, like "gunmen," as in "Gunmen shoot 10 Iraqis." Well, duh, if they shoot somebody, we can conclude they have guns.
In my own field, I'm forbidden from writing "shares were up." It has to be "shares rose," even though the shares sure didn't propel themselves.

Site meter

Search This Blog