Sunday, February 21, 2010

Crank Science: Worse than Wrong

There is a popular quote that summarizes the philosophy of scientific modeling:

“Essentially, all models are wrong, but some are useful.” - George E. P. Box (Wikiquote)

One characteristic of pseudo-science is that it tries to focus on the first part when attacking mainstream models while blatantly ignoring the second part of this quote when presenting their own.

What it means in the scientific sense
Models are a key product of scientific research. A model is basically a 'recipe' for other researchers to apply, test, and extend the work. A model also simplifies the task of implementing new science into technologies.  The best models are presented in a mathematical form, as mathematics provides an independent set of rules which can be used for exploring other consequences of a scientific model.

An example of a useful model would certainly be Schrodinger’s equation (Wikipedia).  This equation became important because other researchers could apply, test, and extend this work, using it to understand the structure of atoms and of matter itself.  Initially, much of the work was applied in understanding the spectra of atoms and molecules, a capability that has provided a powerful diagnostic of conditions in the distant universe. But this same equation made semiconductor electronics and lasers possible today. In that sense, Schrodinger's equation built our modern world of microelectronics.

But models are not absolute.

Consider the example of the THEORY of Gravity (and it IS just a theory).
  • Galileo's theory of gravity: All objects fall with the same acceleration, is useful for constructing things close to the Earth's surface, but it becomes measurably wrong at higher altitudes, and speeds. It is wrong for launching an orbiting satellite.
  • Newton's theory of gravity: This revision to Galileo's theory can generate Galileo's theory of gravity close to the Earth's surface, but has the additional advantage that it is useful for computing the behavior of objects much further away, such as orbiting satellites and predicting motions of the planets as well as stars in our galaxy.
  • Einstein's theory of gravity (General Relativity): In the case of weak gravitational fields, such as those created around large masses like the planets and Sun, it generates results identical to Newton's theory, unless you use very precise instruments. The difference between Newton and Einstein's theory gets larger, with larger values of the mass/radius of the object.
Notice that as we move from Galileo's theory of gravity to Einstein's, the theories not only apply to a wider range of conditions, but they also contain the results of the previous, less precise theory. Theories are subject to refinement as our knowledge grows. For another slightly different perspective on this, see Bob Grumbine's “Successive Approximations”.
What it tells us about how and why crank science fails
Crank models are often worse than wrong. They are useless. The advocates of these models are the only ones who can obtain the ‘desired’ results, as they will rely on ad hoc physical assumptions and/or invalid mathematical manipulations. This creates the 'priesthood' mentality of many pseudosciences, where the 'Truth' of a given claim can only come from specific individuals. Such priesthoods are only overthrown by political means. Consider the example of the recent falling out between Answers in Genesis (AiG) and it's Australian counterpart, Creation Ministries International (CMI) (NCSE: Trouble in Paradise) .

In scientific circles, the type of 'priesthood' might appear to exist, and some individuals may intentionally or unintentionally encourage this image. The difference is that in scientific circles, successful models and experiments can overturn that apparent priesthood and the knowledge base is generally available (with effort) for others to develop the expertise.
Crank science emphasizes the problems with mainstream models while ignoring where those mainstream models work.  Meanwhile, crank science ignores cases where their models fail entirely.
Crank science models, in the rare cases when they present one in a form that others can actually use, often have severe problems in cases where we have lots of experimental data. Consider two examples that I have explored:
  • Look at Barry Setterfield's (latest) failed attempt to fix the problems with his model for a changing speed of light. As I note in “Setterfield & c-Decay: “Data and Creation: The ZPE-Plasma Model”, Mr. Setterfield concentrates on his model 'predicting' the value of the speed of light in the past and far from Earth, yet doesn't talk about this model's predictions today, near the Earth, where we can easily test it.  Near the Earth, we see that it predicts a value for the speed of light almost 2,000,000 times larger than the value we actually measure! Why would anyone else use Mr. Setterfield's theory when it clearly generates nonsensical predictions?!
  • The Electric Sun (ES) model claims there are streams of electrons inbound towards the Sun, sufficient to explain the total energy output of the Sun. Yet the ES models don't explain anything about the fluxes and energies of electrons & protons this model requires. We've flown satellites with plasma detectors in this region for over 40 years (see "Video: Sentinels of the Heliosphere"), but have yet to detect inbound particle flows with sufficient energy to power the Sun.  Storms from the Sun, called coronal mass ejections (CMEs) and solar flares, threaten the integrity of electrical grids on the Earth, satellites in space, as well as the lives of astronauts in space. However, no one uses the ES model for dealing with these issues since the ES models are incapable of making the necessary predictions on the conditions of such solar storms. Why should anyone use the ES model when the ES supporters themselves try to evade this experimental requirement?
Why would anyone use such models? They tell us nothing about how the universe actually works in even an approximate level, and in that regard, they are totally USELESS.

Galileo's theory of gravity was wrong, yet we still use it for designing structures near the surface of the Earth. Newton's theory of gravity was wrong, yet it is still accurate enough for us to launch and navigate most missions in space.

What can you call a theory that generates NO useful information?

Worse than wrong!

Saturday, February 20, 2010

Things slowly returning to normal...

I've been a little busy of late.  Between the first and second waves of “Snowmageddon”, a roof leak developed that rendered my home office and our dining room unusable.  This has disrupted my usual sessions of quiet time where I would compose these posts.  I hope to return to regular posting in the coming week as the leak has been patched and the floor and carpet are finally drying out sufficient to return furniture to a usable arrangement.   Then we need to make arrangements for more robust repairs.

For those who want to claim that these snowfalls are 'proof' that Global Warming (AKA Climate Change) are a hoax, one should note that warmer air holds more water (Wikipedia: Relative Humidity)  so precipitation events will tend to be more intense.  The concerns that such warming would favor more intense storm formation (of all types) is something I recall hearing back in the early 1990s with hurricanes Hugo and Andrew.  Back then, insurance companies became interested in climate change for its potential impact on their bottom-lines.

You also might enjoy:
Bob Grumbine's “Three Feet of Global Warming” and “Cloud-Temperature Feedback”

A couple of other worthwhile items on the topic:
"Heavy snowfall in a warming world"
and on a more light-hearted note:
"It’s Snowing Today, Therefore Climate Change Is a Myth and Al Gore Is a Fraud"

Monday, February 8, 2010

Updates may be a bit spotty this week or so...

I've been caught in "Snowmageddon" this week so updates may be sparse for a time as I've got nothing in the auto-release queue.

We spent 30 hours with no power from about 5:30 Saturday morning and inside overnight temperatures dropped to about 42 degrees F, below the operating range of most home electronics so I dared not even use a battery-powered laptop.  However, there was enough light that I was able to catch up on some backlogged hardcopy reading.  Power returned yesterday about lunch time and a snowplow reached us today.  I've got Internet access so long as I have power but a couple of trees may change that as the next storm moves through on Tuesday-Wednesday of this week.

Be safe.

Friday, February 5, 2010

ICR: “It’s Official: Radioactive Isotope Dating Is Fallible.”

by Gerard Jellison, Ph.D.  Physics

    “It’s Official: Radioactive Isotope Dating Is Fallible.”
    That was the dramatic headline of a news article posted on the website of the Institute for Creation Research (ICR) on January 21, 2010[1]. In the first sentence, ICR Science Writer “Brian Thomas, M.S.” delivered the shocking news: “New data collected by secular researchers has confirmed what creation scientists discovered decades ago – geologists’ cornerstone assumption that the rate of radioactive decay is constant over time is not correct.” A geology professor was quoted: “Everybody was sitting on this two legged stool claiming it was very stable, but it’s not.” Evidently, creation scientists were right all along! Their claims about the unreliability of radiometric dating, and the reality of accelerated nuclear decay that compressed millions of years of apparent radiometric history into less than a year, were vindicated.
    Shocking news, but there’s just one problem. Not a word of it was true.

    The research that was reported – or rather distorted – by ICR was performed by G. A. Brennecka, a graduate student at Arizona State University, along with colleagues at ASU and two institutions in Germany. Published in the January 22, 2010 issue of Science[2], the study investigated whether two uranium isotopes always occur in the same ratio in meteorites. In previous measurements on samples from the Earth, the Moon, and meteorites, the ratio between U-238 and U-235 had generally been found to have the same value, 137.88. This is an important number, because it’s used in Pb-Pb radiometric dating, a crucial technique used to find the ages of meteorites, as well as lunar and terrestrial samples. In Pb-Pb dating, scientists measure the concentrations of two lead isotopes, one derived from the decay of U-235, and the other from U-238. The equation used to derive the age from Pb-Pb measurements requires scientists to know the ratio between the two uranium isotopes.
    Brennecka and his colleagues measured the uranium isotope ratio in thirteen calcium-aluminum-rich inclusions (CAIs) from the Allende meteorite, which fell on Mexico in 1969. Probably the best-studied meteorite in the world, Allende is an unusually large example of a primitive class of meteorites called carbonaceous chrondrites. Evidence shows that its CAIs include some of the first solids to condense during the birth of the Solar System, and their age is thought to represent the age of the Solar System itself.
    Using refinements of existing techniques, Brennecka et al. found that uranium ratios in the thirteen inclusions varied from 137.409 to 137.885. This small variation, if confirmed in further studies, is enough to bring about a change in the radiometric dates of meteorites, and of the Solar System, of about 0.1%. That’s small, but significant for astronomers who want to date events like planet formation to within a few million years.
    The work by Brannecka et al. is significant for another reason. The variation they found is apparently due to an excess of U-235, and the most likely explanation is enrichment of this isotope by decay of a curium isotope, Cm-247. But no one has detected this short-lived isotope in the Solar System before, and it is only created in certain types of supernovae. This new work, suggesting the existence of supernova-derived atoms in the protoplanetary disk, may have important implications for the evolution of the Solar System, and its relationship to its galactic environment.

    So much for real science. What did ICR make of all this?
    Brian Thomas told his readers that the “variation” detected by Brannecka et al. destroyed the “assumption” of constant decay rates. “The clock is broken,” he declared. Assuming for no good reason that the “variation” meant changes over time, he quoted Brannecka: “This variation implies substantial uncertainties in the ages previously determined by Pb-Pb dating of CAIs.” Thomas’s bias about the invalidity of radiometric dating caused him to imagine something that was never even implied in the Science report. Thomas never discussed the actual magnitude of these “substantial uncertainties” (a reference to the difference as “small” was buried at the end of the article). Much of the article was devoted to ICR’s familiar claims that its RATE project had “nullified the idea that the decay rate has been constant.” Thomas even misinterpreted the RATE results, claiming that only with accelerated decay could “microscopic scars called ‘fission tracks’ have formed” (fission tracks are a standard tool for radiometric dating, and their existence and concentration in rocks, even as measured by RATE, are completely consistent with old-Earth science).
    This article was so incompetent that even ICR couldn’t keep it on the site in its original form. On January 26, it was quietly revised (“for accuracy” according to the fine print at the bottom). 
I don’t know who told ICR about the debacle, but the new version tells me all I need to know about ICR’s atrophied sense of integrity. The honest thing, of course, would have been to withdraw the article, with a full explanation and a public apology to the scientists whose work was so blatantly misrepresented. Instead, ICR kept the article and the references to Brannecka et al., but just diluted some of the claims. For example, “geologists’ cornerstone assumption that the rate of radioactive decay is constant over time is not correct” was changed to the vaguer “geologists’ assumptions about radioactive decay are not always correct.” Although the new article doesn’t say explicitly that the work of Brannecka et al. confirmed accelerated decay, about 40% of it is still devoted to Thomas’s credulous (and now laughably out of context) descriptions of the RATE work on (what else?) accelerated decay. And the new article still falsely says that the scientists’ work “dovetails” with creationist research that undermines the validity of radiometric dating.
The message of ICR’s article hasn’t changed. Their conclusions are immune to facts.
    The first version of ICR’s article can, charitably, be explained away as an act of honest incompetence by “Science Writer” Thomas. The new version demonstrates ICR’s inability to admit a mistake, and their willingness to guide readers to conclusions that have nothing to do with the scientific research they cite.
    ICR’s current article says that the varying isotope ratios “call into question the calculated age of the solar system.” In truth, the work of Brannecka et al. suggests a change in this age estimate – from 4.567 to 4.562 billion years. This tiny correction testifies to the precision and reliability of current radiometric dating technology. ICR’s online description of this work testifies to their willingness to deceive their readers.

[1] "It's Official: Radioactive Isotope Dating Is Fallible" by Brian Thomas.
[2] G. A. Brennecka, S. Weyer, M. Wadhwa, P. E. Janney, J. Zipfel, and A. D. Anbar, “238U/235U Variations in Meteorites: Extant  247Cm and Implications for Pb-Pb Dating”, Science, 327, pp. 449-451 (22 Jan. 2010).  For less-technical online explanations of the research, see the ASU press release and the article by L. Grossman, “Age of Solar System Needs to Be Recalculated,” Wired Science.