Last week I challenged a particularly nasty commentary by David Paul Kuhn that alleged that Democrats were even nuttier than the Republicans who believe the ‘birther” narrative because many Dems believe that George W Bush had “advance knowledge of the 9/11 attacks”. Here’s what I said:

There are two different ways that a survey respondent could interpret the Rasmussen question about Bush’s possible “advance knowledge of the 9/11 attacks” – one of which is entirely rational and, in fact, undeniably true.
Gee whiz, come on. Doesn’t everyone still remember the warnings Bush received about the potential use of airplanes as terrorist weapons in the summer of 2001 – warnings Condi Rice admitted did not get followed up? Don’t we all remember the CIA memos saying that “something big” was in the works in September? Don’t we all remember the 9/11 Commission and Richard Clarkes’ dramatic statement that “We failed the American people”
These were not hallucinations or the product of fevered, paranoid Democratic brains. They were component elements of the undeniable fact that there were indeed significant advance warnings that a terrorist attack was in the works for the fall of 2001 – a fact that was the central subject of the 9/11 commission hearings, 10 or 15 books and hundreds of articles.
One would have to throw out every single academic study of the past 30 or 40 years about the effects of question wording on survey response not to recognize that, for many survey respondents who remembered the 9/11 Commission Report and other media coverage, the phrase “advanced knowledge of the 9/11 attacks” could be cognitively processed as meaning:
“The Bush administration had substantial advance knowledge from U.S. intelligence sources that a terrorist attack on the U.S. was being predicted as imminent in the fall of 2001”
Rather than interpreting the question as saying:
“The Bush Administration had specific and detailed advance knowledge about a particular group of 19 Saudi Arabian terrorists armed with box cutters and trained to fly commercial jet aircraft who planned to hijack four U.S. airliners at 9:45 in the morning on September 11th 2001 and attempt to crash two of them into the New York World Trade Center”

This week, in his blog, Brendan Nyhan offers essentially the same criticism of the Rasmussen question. As he says:

The problem, as Media Matters points out, is that the wording of the Rasmussen poll (“Did Bush know about the 9/11 attacks in advance?”) almost surely conflates people who believe Bush intentionally allowed an attack to occur with those who think the administration was negligent in its attention to the potential threat from Al Qaeda. Even National Review Online’s Jonah Goldberg conceded this point in a column published soon after the poll was released.

But then he goes on to say the following:

Another, lesser-known poll used less ambiguous wording and found similar results. A July 2006 Scripps Howard/Ohio University poll asked the following question:

There are also accusations being made following the 9/11 terrorist attack. One of these is: People in the federal government either assisted in the 9/11 attacks or took no action to stop the attacks because they wanted to United States to go to war in the Middle East….

In short, using a more appropriate comparison poll, the primary conclusion stands — both party’s bases are disturbingly receptive to wild conspiracy theories.

Nyhan is clearly right that this question is better than the previous one. But – when viewed in light of the academic literature about how ordinary respondents actually cognitively process poll questions – the fact is that it is still insufficient to support his conclusion.

Why? Because the question presents a reader with three fundamentally different kinds of assertions about unnamed “people in the federal government” in a single sentence and then asks them to judge if the sentence as a whole is “likely”.
The question asserts that the unnamed people either

(1) “…assisted in the 9/11 attacks”
Or
(2) “…took no action to stop the attacks”
Because they
(3) “…wanted the United States to go to war in the Middle East”.

Now, if these three different assertions about “people in the federal government” were presented to the survey respondents as three separate items, I suspect most Democrats would classify them as follows
Item one – “…assisted in the 9/11 attacks” — false and indeed a “wild conspiracy” accusation.
Item three – “…wanted the United States to go to war in the Middle East” — an essentially accurate statement about the attitude of Dick Cheney and the leading neoconservatives at the time. It is a statement that is specifically supported by evidence in the 9/11 report and Bob Woodward’s books, among many other sources, and is therefore in no way proof of a person’s belief in “a wild conspiracy”
Item two — “…took no action to stop the attacks” this is a phrase that can be interpreted in two entirely different ways depending on whether one believes that the administration actually had in its possession sufficiently detailed and specific knowledge to have been able to prevent the hijacking of the four airplanes if it had really wished to. If one assumes – as most first-hand accounts suggest — that the administration did not actually have such detailed knowledge in their possession then a belief that they failed to act cannot be interpreted as belief in a “wild conspiracy”. Instead the phrase could easily be interpreted by someone familiar with the 9/11 commission report and the views of Richard Clarke and others as intended by the writers of the survey question to suggest negligence on the part of the administration – negligence that Clarke and other critics specifically alleged — rather than a conspiracy.
Using the most strict and literal reading of the Scripts/Howard question, it could be argued that the overall wording does tend to imply the first possibility — that the government did indeed have sufficient knowledge to have forestalled the attack. On this basis it could be argued that – if one considers this implication false — an ideal logician should also reject the overall conclusion as false as well. But I suspect that no-one familiar with the literature on the haphazard way survey respondents actually fumble about trying to parse, process and construct on-the-spot answers to survey questions will seriously want to defend the proposition that this linguistic train wreck of a survey question is really sufficiently clear and unambiguous to justify its use as a psychiatric tool for the diagnosis of delusions and paranoia in America’s Democrats — or anybody else for that matter.
So, once again, the case for Democratic lunacy remains a “case not proved”. Heck, let’s be honest. If people seriously want to find a 9/11 conspiracy narrative that is comparable to the “birther” narrative, they will have to find a question that clearly asks the following:

“Did the Bush Administration have sufficient advance knowledge of the 9/11 attacks to have prevented them had they really wished to but nonetheless deliberately allowed the attacks to take place?”

This wording clearly and unambiguously expresses a conspiracy narrative and most Dems, I think, would indeed agree that it is so fundamentally false that any Dems who believe it can be compared without injustice to the Republican believers in the “birther” narrative.
I would also bet that a substantially – very substantially — lower percentage of Dems would sign on to this notion than do Republicans to the ‘Birther” narrative – a narrative which now appears to have attained the depressingly high level of support that opinion polls have always shown for belief in such things as alien abductions and the notion that professional wrestling is not fixed.
Sure, I’ll concede that there are some Dems who are really nutty — just a whole lot fewer, I suspect, than Republicans.