Thursday, January 26, 2006

The branch stub

An anonymous person emailed me this:
To me, the following is the most significant quote in Dr. Jackson’s Auk article:

Figure S5A is likely a branch stub (J. Fitzpatrick pers. comm., 29 July 2005)

That means Fitzpatrick, leader of the Cornell team, has acknowledged that a claimed “Ivory-Bill” image from the famous Luneau video, the centerpiece of their evidence, is apparently a branch stub!

The text of Cornell’s Supporting Materials explains that an “indistinct object” was noticed on a tree in the video. When they put am IBWO model in the same spot, and filmed it out-of-focus in similar way to the poor quality original video, it looked much the same as “the object” in the video. Cornell concluded “We interpret the object in the Luneau video as an Ivory-Bill Woodpecker.”

Last September 6th, your blog examined the Luneau video and that object, and noted that there are numerous objects in the DVD that look indistinguishable from the “Ivory-Bill,” and that it was probable that they were all just vegetation and/or artifacts. You looked for a reasonable alternate explanation for Cornell’s “extraordinary claim” and your simple explanation was correct.

The revelation that Cornell is now saying that one of their “Ivory-Bills” was likely a branch stub is extremely important because it shows:

1. Cornell made at least one major blunder in their study, but have yet to admit it publicly.

2. Reenactments with models prove nothing if the test is flawed and results aren’t objectively evaluated. (Cornell's “flying Ivory-Bill” model with inflexible wings is another example of flawed models and tests.)

3. Blurry video and photos are easily misinterpreted.

4. Groupthink was definitely a factor in the study. Only groupthink would allow a team of intelligent people, including professional ornithologists, to identify a tree branch as “one of the rarest birds alive.”

5. The evidence as a whole is terribly weak if Cornell attempted to “spin” such a poor image into an Ivory-Bill.

6 If Cornell mistook a tree branch for an IBWO, surely it would be easier to mistake a Pileated for an IBWO, and easier yet to mistake a glimpse of an aberrant Pileated for an IBWO.

7. It illustrates that “the desires and expectations people possess influence their perceptions and interpretations of what they observe.”

In short, Cornell saw what they wanted to see, something that wasn't there, in that blurry image. I’m afraid that history will conclude that that was the case with all their Ivory-Bill sightings.

6 comments:

Anonymous said...

Excellent post.

The Jackson paper, without something extraordinary happening, surely marks the beginning of the end of credibility in the Science paper.

Anonymous said...

Actually, it demonstrates that Cornell will retract conclusions that they feel were erroneous. This is good science, not bad. And they have retracted nothing else in the paper.

Tom said...

Why did we have to hear about this retraction from Jerome Jackson, rather than directly from Cornell?

Jackson's paper says that Fitzpatrick admitted this on July 29, 2005. Why didn't Fitzpatrick mention this retraction at the AOU meeting the following month?

Anonymous said...

Did he present the branch stub at that meeting? Did anyone ask about it? I can't view the video of the presentations.

Tom said...

The so-called "six-pixel bird" was mentioned multiple times at the meeting, including at least a couple of times in Fitzpatrick's plenary. No one asked if it was likely a branch stub, and no one publicly volunteered that information.

Anonymous said...

"Actually, it demonstrates that Cornell will retract conclusions that they feel were erroneous."

It demonstrates exactly the opposite. They haven't retracted that statement at all. They've only admitted their error "privately." That's not science, that's an organization trying to prop up a shaky story.