Radioactive Dating: Science or Alchemy?

Richard Kerr had some surprising things to say about uranium-lead dating in the Sept. 17 issue of Science¹ – surprising, because as a believer in the method and an evolutionist, he admitted there is a fair amount of unscientific methodology and controversy involved.  “For years, different laboratories using uranium-lead radiometric dating—the gold standard of geochronology—have been getting entirely different ages for the P-T extinction,” he says.  His comments stemmed from a paper in the same issue by Mundil et al.² that touted a new method for getting the bugs out of U-Pb zircon samples.  But the way Kerr worded his subtitle, he sounds at best tentative about its benefits: “A new, apparently improved, way to date the greatest mass extinction points to a volcanic cause but fails to resolve geochronologists’ long-running differences.”  (Emphasis added in all quotes.)
    Mundil’s team, from the Berkeley Geochronology Center, admits right off that “The age and timing of the Permian-Triassic mass extinction have been difficult to determine because zircon populations from the type sections are typically affected by pervasive lead loss and contamination by indistinguishable older xenocrysts.”  In order to date samples from China, they “adopted a technique recently developed by James Mattinson of the University of California, Santa Barbara, Kerr says.  “ They baked the southern China zircons at 850�C for 36 hours and then leached them with hydrofluoric acid under pressure at 220�C for 16 hours, with the intention of removing the parts most weakened by radiation damage.”
    This harsh treatment of the samples was intended to eliminate some of the “picking and choosing” that commonly goes on by researchers, who discard samples that don’t give them the results they expect.  Samuel Bowring (MIT), for instance, got a date for the P-T extinction that, while it seemed to match some dates for massive Siberian lava flows, disagreed with the age Mundil prefers:

Mundil, however, doesn’t believe that either the eruption or the extinction happened that recently.  He thinks Bowring engaged in “arbitrary data culling” by throwing out more than half his zircon ages before averaging the rest of them together.  But Bowring says his choices were judicious, although “necessarily somewhat subjective.”  In some of his zircons, the two different uranium-lead ratios gave different ages, suggesting that lead had leaked out of those zircons during the past quarter-billion years.  And other zircon ages looked distinctly old, as if those zircons had crystallized earlier than the rest and had later gotten mixed in with them.  By taking into account how volcanic ash beds are stacked around the rock layer that shows the extinction, Bowring believes he can confidently select the reliable zircon ages and discard the rest.
    Mundil set out to take this “picking and choosing” out of uranium-lead dating.

Thus the heat, pressure and acid treatments.  With this method, Mundil claims he only had to throw out three out of 79 of his zircon samples which were “obviously too old.”  He arrived at a date for the extinction a million years older.  It was also coincident with an argon-argon date for the Siberian lava flows made by others, “after making a 2-million-year correction to it”  The goal of this tweaking is to fix the timing: “The professional timekeepers—the geochronologists—are trying to place a volcanic catastrophe at the moment of the extinction, thus linking cause and effect to explain an event that wiped out 95% of animal species on Earth,” Kerr explains.  The challenge is that “P-T daters must draw their conclusions from vanishingly small isotopic remains of radioactive decay.”  Though the antagonists try to keep a positive spin on the controversy, Kerr indicates that geochronology may not be the exact science we have been led to believe:

The new preprocessing technique “is very promising,” says Drew Coleman of the University of North Carolina, Chapel Hill.  “It appears to be very fruitful.”  Bowring agrees.  “This is a step in the right direction,” he says.  “Mattinson’s annealing is the big breakthrough, though I have no idea why it works.  But Bowring points to the later date that his group estimated for the P-T extinction in China and Kamo’s group independently got for zircon and other minerals from the lavas of the Siberian Traps.  Mundil hasn’t explained how subjective interpretation could have produced such a coincidence, he says.

All can agree on one thing.  Better cooperation might help.  Speaking of the geochronologists, Randall Parrish of the British Geological Survey paints them like a secret society: “They’ve been competitive and secretive for decades,” he said.  With a meeting of geochronologists in Boston coming up next month, Kerr hopes for a “frank and open discussion of all those little details that don’t make it into the literature.”

¹Richard Kerr, “Geochemistry: In Mass Extinction, Timing Is All,” Science, Vol 305, Issue 5691, 1705, 17 September 2004, [DOI: 10.1126/science.305.5691.1705].
²Mundil et al., “Age and Timing of the Permian Mass Extinctions: U/Pb Dating of Closed-System Zircons,” Science, Vol 305, Issue 5691, 1760-1763, 17 September 2004, [DOI: 10.1126/science.1101012].

Now I’m worried.  What are all those little details that don’t make it into the literature?  You can count on it: the “picking and choosing” that Kerr admitted is only the tip of the iceberg.  To be fair, the U-Pb differences between the teams only amount to a small percent.  But to arrive at the millions-of-years dates at all, dates that justify the modern consensus for the geologic column (see 05/21/2004 headline), they have to toss out many other dating methods that produce far younger dates by orders of magnitude.  Those dates are not interesting because they do not support the Darwinian evolutionary timescale; therefore they are “obviously wrong.”  Obviously.  That’s why they must pick and choose.
    Radioactive dating was supposed to be as solid as the Rock of Gibraltar, a highly constrained, well understood physical mechanism for dating old things.  That’s what we learned in high school and on the Discovery Channel.  Now they tell us they have been picking and choosing the samples they want and tossing out over half the rest?  Even if Mundil threw out only 3 of his 79 samples, we want to know if those three had a story to tell: on what basis did he assume they were “obviously too old”?  How can we know the 79 he used were not also obviously too old, at least to someone without Darwin glasses on?  Sometimes the most interesting science is in the data the conventional wisdom tosses out.  The stone that the builders rejected sometimes becomes the cornerstone of a new paradigm.  Instead, the subjective practices of researchers, secretly trying to turn lead into the “gold standard of geochronology,” appear more like a modern form of alchemy than modern science.
    When they don’t get what they want, they turn up the heat.  What did two days of red-hot heating, pressure and treatment with hydrofluoric acid do to the samples?  How can we be convinced this was “the big breakthrough” if an admirer admits he has no idea how it works? (which, being translated, means “I have no idea if it works”).  Remember, they are talking about “vanishingly small” bits of radioactive material to begin with, and then heating and acid-washing some of it away.  What can samples tell you under this kind of torture?  Here’s what we suggest they are saying: “Stop!  (Gasp!  Ouch!)  I give up!  I’ll give you any date you want!”