Sunday, April 30, 2006

In praise of Cornell's science?!

I respectfully disagree with just about everything in the 23-minute lies.com podcast available here.

1. The podcaster (John Callender) seems to think that (according to sound spectrum analysis) some recorded Arkansas kent-like calls are a good match for real (recorded 1935) Ivory-bill calls.

That's just not true. Check out the two spectogram pictures available here. Do those two spectograms look like a good match to you?

Since this mismatch was a big problem, Cornell "degraded" the original calls by broadcasting them through foliage at 145 meters, then re-recording them. Of course, if you degrade the "target" evidence sufficiently, you can always find field evidence that appears to match. It's not unlike claiming that your video (of a Pileated and branch stub) match highly-degraded video of Ivory-bill models.

2. Callender seems to agree that Cornell's stiff-winged models are seriously flawed, but then suggests that Sibley should create models with wings that do move in a lifelike manner!

3. Callender still seems impressed by Cornell's "wingbeat frequency" argument (if you're a Cornell grad, this argument in particular should make you cringe).

For one thing, Cornell's claim about the Luneau bird's wingbeat frequency is clearly bogus.

For another thing, Cornell has a single data point for the wingbeat frequency of a fleeing Ivory-bill, and that claim may also be bogus.

The 1935 wingbeats (analyzed by Cornell here) may not be the wingbeats of a fleeing Ivory-bill. The recorded sounds may have been made by an agitated Ivory-bill fluttering near the nest hole.

Listen to the wingbeats here (WAV format). Why do the wingbeats stop abruptly? Note that the eighth wingbeat is still fairly loud, and although the recording continues, no further wingbeats are heard. If the bird was really fleeing, why don't we hear (and why doesn't the spectrogram show) wingbeats 9, 10, etc?

16 comments:

Anonymous said...

Tom asks: If the bird was really fleeing, why don't we hear (and why doesn't the spectrogram show) wingbeats 9, 10, etc?

Interesting question, do parabolic reflectors have a "field of view"? What would you expect the field of view to be if there is one? How far would an IBWO travel in 8 wingbeats? What would you expect the audio drop off of a bird departing from the focal direction using a parabolic reflector?

anonanon

Anonymous said...

Why, Why, Why, and Why thousands & thousands of times over. That seems to be the first word that begins every issue that applies to this so-called discovery. It would seem that a find like this one that has been front & center would be able to answer at least a few of the "Why's" does it not.

But in this case, not only are the original "Why's" still there to be answered, but new "Why's" are being added to the long list everyday.

How many more "why's" need to be added before a few get answered? I, like most others gave this issue a fair chance. We believed when told by experts that 1+1=3. We knew it didn't add up, but we still went along. But has it been shown to us where 1+1=3? No! But just the oppisite. We have been told that 1+1= as many as they want on this issue. The math is no longer adding up. I for one can add & I am still waiting on the first #1 to be answered. I believed because of the creditability of those of ask me to believe. But now I no longer believe because the creditability of those who ask me to has failed! 1+1=2 in the real world!

Anonymous said...

Blue-Jays do make dead-on IBWO kent impersonations. Yesterday I did finally hear these sounds with the occasional characteristic "queedle" mixed in.
They are spookily close to IBWO calls. This personal and admittedly anecdotal account happened yesterday morning... near the Hudson River. Maybe this is how a Blue-Jay sometimes imitates a nuthatch. The calls were cadenced and sometimes "paired".
Nothing like the weak Blue Jay calls Cornell put on their website. I admit like most birders, I usually pay little attention to Blue Jays. Now somewhere I imagine Cornell probably has a Blue Jay recording that is much closer to what I heard yesterday than the weak calls on their website. I've resisted Cornell bashing but this seems like evidence designed to mislead... as if to say... with all our recordings of bird sounds
this is the closest we came to an Kent-Kent sequence. I know I shouldn't challenge Cornell. Afterall I'm a mere 7-sisters grad throwing pebbles against a vast wall of ivy-laden reputation. ;-)

Paul Sutera, New Paltz, NY

Anonymous said...

Welcome, Paul. Welcome to the Skeptics.

But in truth, I've read all your posts. You were always a Skeptic at heart.

Anonymous said...

Can men graduate from the 7 sisters?

Do I sense intense rivalry and envy between these elitist institutions?

Personally, I love blue jays and their mimicry.

olivacea said...

Paul,

I am glad that you have heard that specific Blue Jay call. It is not common, but I hear it regularly when family groups are visiting my dripper pond, which several do on a daily basis. It's not a loud call, and I have no guess as to its "social" context. It is given by nearby birds, when one bird is in the bath...

I have listened to the Cornell and Mike Collins audios over and over, and have not heard anything, except for the mysterious double-knocks, that I cannot hear in my yard.

Olivacea
deep in the woods of Florida

Anonymous said...

Paul, I love you more than words can describe.....an Ivory-bill....on a blurry video....half-hidden by a Tupelo.

And I mean it.

John Callender said...

It's not so much a critique of the specific issues that I was making in my podcast. It's more a critique of Sibley's et al.'s whole approach, which is to basically treat the question as a very, very high-stakes hearing regarding a claim of seeing an extreme rarity in the world of competitive birding.

Well, on one level, that's what it is, which makes their approach appropriate *within that context*. But what they're doing isn't science. Or, as I said in the podcast, it's only the first of many steps that would be required for it to qualify as science.

What they've done is to hypothesize some alternative explanations for the Cornell team's data. What they've failed to do is to make any meaningful attempt to test those hypotheses.

Sibley's position basically amounts to, "If I can come up with a way of explaining away the data without requiring that I give up my starting assumption (that the Ivory-billed Woodpecker is actually extinct), then I'm done. Game over."

Great. But it's not an objective search for truth. It's an emotionally driven act of self-deception, just like the medieval astronomers who devised elaborate explanations of orbital mechanics to preserve the fiction that celestial bodies moved in perfect circles.

Anonymous said...

John;
The burden of proof is on the CLO not Sibley. All Sibley had to show was an alternative explanation to the 4 second fuzzy wuzzy video.

The major TRUTH/SCIENCE point as restated in Tom Nelson's signature line, Miraculous Claims require 100% proof; not suppositions, vague sightings, Zapruder film videos, and unknown source vocalizations that are more likely Blue Jays.

Nobody to date can come close to a 100% verification that the Lord God bird still exists.

Believing something is alive is not proof it is belief.

An objective birder would clearly see another case of a mis identified Pileated.

John Callender said...

Well, again, while the position you've stated may satistfy your own sense of what's required to mount a convincing case that an Ivory-billed Woodpecker was seen in Arkansas in the past few years, it falls well short of the standard required by good science.

It's actually kind of helpful that you've taken the Sagan quote ("extraordinary claims require extraordinary proof") even further ("miraculous claims require 100% proof") because it helps to emphasize how little your position has to do with science.

Science doesn't deal with "100% proof." No scientific explanation can ever be considered 100% "proven", because no explanation gets to be immune to falsifiability. It is precisely that weakness that separates miraculous claims by creationists (for example) from being real science.

Sure, I'll stipulate the obvious: the Luneau video is ambiguous. The eyewitness accounts of recent Ivory-billed sightings are not verifiable in the same sense that good photographic evidence would be. The audio recordings of the "kent" calls and double raps could have been produced by something other than Ivory-billeds.

It's possible that someone could come up with an alternative explanation that actually makes more sense than the one the Cornell team offers. But to do that would require doing actual scientific testing and evaluation on a par with what the Cornell team has done already. And for all the disparaging comments about that work that you and Mr. Nelson have made, what Sibley and his co-authors have offered so far falls well short of it, at least in the sense of representing serious scientific analysis.

Hypotheses are fine as starting points. But Sibley et al. haven't tested theirs. They simply appear to like them better for emotional reasons. Which is nothing unusual; it's human nature, and we're all subject to it. But again, at the risk of being tedious, it isn't science.

Anonymous said...

John;

You gotta believe what you gotta believe. Cornell makes the miraculous statement that the IBWO lives again. Their proof is vague. Yet, you don't have a problem with that because their "proof" is better than Sibley's "disproof"?

Anonymous said...

Given that the blog on which this podcast was posted frequently faults the Bush administration for stubborn denial of facts, misinformation, deception, and persistent bold claims based on insufficient, and often biased evidence, I would have placed Fitzpatrick et al. right there with them. Sibley et al. presented visual evidence, empirical data, and sound reasoning to reject the hypothesis that the bird in the video could be an Ivory-billed. Fitzpatrick et al. present only suggestive evidence to support their belief that they have "convincing proof the Ivory-billed Woodpecker has survived into the 21st Century” (J. W. Fitzpatrick, Big Woods video) or that the sightings and the video evidence confirm the existence of at least one male Ivory-billed Woodpecker (Science, 2005). Sibley et al. effectively challenged these erroneous conclusions. They undertook an independent analysis and demonstrated that Fitzpatrick et al.’s conclusions were not repeatable. That is doing science.

To support the opinion that only untested hypotheses are presented by Sibley et al., the podcast cites the relationship between wingbeat frequency and mass—a heavier bird is expected to flap at a slower rate. Empirical data--real, observational evidence--was cited by Sibley et al. to support this. The regression analysis for that correlation is significant. When someone finds an Ivory-bill, then we can test this. And why should Sibley et al. have made articulated models when visual data of real birds presented by them disconfirms Cornell’s phony and deceptive models and reconstructions (see, for example, Tom’s post on figure 1). It seems clear to me that all reports remain unsubstantiated and that the claims of convincing proof and confirmation are false.

John Callender said...

While you've certainly done an admirable job of using all the right words in your response, I remain unimpressed with the actual quantity and quality of what Sibley et al. have done in terms of scientific analysis.

As far as I can see, they really haven't done much more than to draw some interpretive illustrations of a number of frames from the Luneau video, and to offer alternative explanations (with little by way of supporting evidence) for other aspects of the Cornell case.

Since we have no shortage of Pileated Woodpeckers, I'd be more impressed if the Sibley team would present a video of a Pileated Woodpecker taken under similar conditions that comes close to matching the Luneau video. Regarding claims that the sightings in the area are likely of a leucistic Pileated, again, I'd be more impressed if the extensive searching in the area had turned up more evidence for that bird.

I'm not saying that the case for the Ivory-billed's existences is a slam dunk at this point. I'm saying that it's much stronger than this weblog's author, and many of the commenters here, give it credit for. And that's fine; people can choose skepticism as a perceptual filter if they want to, and clearly a lot of obsessive birders choose to do just that. It makes them much less likely to be taken in by a bogus sighting. It also makes them much more likely to dismiss a valid sighting that is supported by limited evidence.

Anonymous said...

I think John exhibits a normal but fallible tendency in human reasoning—the tendency to be excessively influenced by supportive information and the tendency to seek confirmatory evidence. The fact that the bird in the Luneau video exhibits black in the wings is pretty clear, and this rejects the notion that it is an Ivory-bill. Seeking videos of Pileateds showing a lot of white is not as informative (there are examples in the Cornell videos themselves; see an earlier post on this blog). A fine example of how a bird with a similar pattern may look all white in the wing is shown by this Magellanic Woodpecker (http://www.cimicorp.com/DI/Collection14/pages/2005_Patagonia_05_0119.htm), a species which has black secondaries and white wing linings similar to a Pileated.

I don’t think one needs to look for an aberrant Pileated to explain the sightings. They were more likely innocent errors—misperceptions due to bias or expectations—confirmed in the observer’s own mind because they lacked opportunity to reconfirm what they thought they saw (all the sightings fall into this category). Even still, white patches on birds can be transitory; so any aberrant Pileated may look “normal” now and not findable. The interpretation of the sightings again places emphasis on confirmatory information, overlooking features that disconfirm them. Both Gallagher and Harrison, for example, independently sketched a black back, a feature that indicates the bird was not an Ivory-bill. (Beyond that, the white in the wings they did depict is too extensive and curving at the wingtip, unlike an Ivory-bill.)

Lastly, I think John fundamentally misunderstands the problem here: how do we know something. Sightings of rarities, even those possibly or probably valid but not well-documented, merit searches. They are not dismissed. All reports remain on record. Sightings supported by only limited evidence are simply less informative and less important for the historical record and the body of knowledge to which they are admitted. Many people encounter birds that they think are rarities but end up with little to show for it (me included). Well-documented sightings do come along, however, and those are the ones that contribute to the ornithological record. Reports of Ivory-bills in Arkansas have been misrepresented as definitive, and the record distorted by hype and beliefs based on evidence that cannot be validated scientifically.

John Callender said...

Well, one could just as easily make the case that it is the skeptics who are focusing on information that confirms their belief (that the Ivory-bill was _not_ seen) while overlooking disconfirming evidence.

My main point, to the extent I have one, is that reality is under no obligation to twist itself to conform to our a priori beliefs. The bird either exists or it doesn't, and that existence (or lack thereof) is not influenced by human beings' tendency to pick one explanation or another, and then fight to defend that position by belittling or ignoring contradictory evidence.

Green Man said...

Why were "they" looking in this spot?
Years ago, "they" discovered a rare fish and shut down construction of a dam and nuclear power plant. Eventually, the fish was discovered to be as rare as dandruff. It just cost the taxpayers years and $billions.
Now, "they" are at the site of a proposed irrigation project. "They" knew another rare fish story would be hard to swallow. So, "they" came up with a rare bird that could not possibly prove to be less than rare because it does not exist. All "they" had to do was find a camera capable of taking blurred video. I think that finally bought a used camera from a Big Foot hunter. Now the project is on hold and costing the taxpayers untold $$$.
If anyone wanted to know if the IBWP really exists all you need to do is set a season and bag limit then publish a recipe. It won't take long.