Several of
Dr. Snelling's examples came up in talk.origins last year
(1998). In
Deja
News I found some old articles of mine discussing
Austin's K-Ar results,
Dalrymple's survey, and the
Hualalei lavas, all of which are referenced by the
Impact
article.
Most of the
Impact examples fall into two categories: (1) rocks
which were never
expected to be datable by K-Ar and understandably yield
nonsensical results (e.g., minerals formed in the mantle) account for most
of the large errors in K-Ar age; and (2) rocks which
will yield accurate
long-term K-Ar dates (though they don't yield accurate short-term K-Ar dates)
because there is a small amount of inherited argon present.
I'd recommend that folks examine the data which Snelling himself
references. In particular, Dalrymple's paper, which reported K and Ar
analyses of 26 samples from lava flows observed in historic times. Of
those 26, one contained obvious (upon visual inspection) mantle inclusions
and was therefore unsuitable for the method. Not a single one of
the remaining 25 samples contained enough excess argon to interfere with
an old-age K-Ar assessment.
For example, Snelling listed the Mt. Etna basalt from Dalrymple's paper,
which had among the highest levels of inherited argon found in that study.
It gave a K-Ar age of about 300,000 years, even though it is less than 2,000
years old. In 60 million years (when it will still be younger than any
mesozoic, paleozoic, or precambrian formation is today), that same rock
will give a K-Ar age of 60.3 million years -- an error of one-half of
one percent due to the inherited argon.
Snelling knows that hundred-thousand-year errors due to initial argon
are insignificant in the context of a hundred-million-year measurement.
Dalrymple's study actually demonstrates that K-Ar dating is quite
dependable for long-term isotopic age determinations -- because the error
due to inherited argon in every one of Dalrymple's samples is far smaller
than the spans of time that the dating method is regularly used to measure.
That is precisely the opposite of the the position that Snelling tried
to use it to support. Snelling has to know better, but I'd bet that most
of his Impact readers don't.
The issue of whether it's intentionally misleading aside, I see three
main issues with Snelling's Impact:
- The data simply do not support the position that he wishes to promote --
that all old isotopic results based on the decay of
40K can simply be ignored as untrustworthy. The bulk
of the data suggests that the methods are reliable (more detail below).
Snelling can try to create the opposite impression with a slanted presentation
of carefully selected data, but even then (as noted above) the data is so
consistent that he is stuck badly stretching the meaning of his own
hand-picked examples.
- If Snelling wishes to make a general case against dating based on
the decay of 40K, it's not going to be
done by dealing in laundry-lists of marginal cases, deliberate misapplications
of the dating methods, and concentrating on the least reliable methodologies.
Instead he should be trying to bring down the opposition's best evidence
-- the most reliable methodologies (Ar-Ar and stepwise heating for example),
the samples which by all other tests appear the most suitable, the results
whose interpretation isn't unclear. Snelling and other creationists never
touch that data. I believe that they know they don't have a solid argument
against it.
- And, finally, even if Snelling succeeds with the first two items, his
case is still only half-finished. It is fairly easy to make up excuses for
ignoring the evidence -- that's all Snelling's Impact article
amounts to. Snelling can't turn his own desired timescale into a legitimate
alternative until he stops merely "explaining the evidence away"
and begins to "explain the evidence." Snelling would have to show how the
observed pattern of K-Ar results is a necessary and expected
consequence of the age and history of the Earth which he accepts.
That final item is very important, and the remainder of this feedback
response will expand on it.
As an example, consider the Albian Stage, which sits roughly in the middle
of the Cretaceous. It was identified by distinctive fossil composition in the
1840s, more than a century before isotopic methods were applied to it.
The identification was performed by geologists who believed in fixity of
species, decades before Darwin published Origin of Species. It
cannot be argued that the fossil content or relative position in the geologic
column of the Albian Stage was driven by either "evolutionary" concerns or
knowledge of its isotopic age results.
Harland et al. (A Geologic Time Scale 1989, pp. 89-90) report more
than 30 dates for samples from the Albian Stage.
The number of dates for just that one
stage is greater than the number of bad ages that Snelling produces. Unlike
Snelling's list, these samples are ones which have the highest appearance of
suitability -- for example, least evidence of weathering or later metamorphism.
Several of the reported individual numbers are actually the aggregate result of
a suite of several samples and several measurements. The results are (values
in millions of years, dates by K-Ar dating except
red which are Rb-Sr):
| 95.00 ± 1.00 |
98.70 ± 2.50 |
100.00 ± 0.80 |
104.40 ± 0.75 |
| 96.18 ± 3.11 |
98.90 ± 1.23 |
100.27 ± 3.00 |
105.36 ± 0.91 |
| 96.18 ± 3.14 |
99.00 ± 1.12 |
100.60 ± 0.50 |
106.00 ± 0.50 |
| 96.50 ± 1.35 |
99.24 ± 3.38 |
100.60 ± 2.50 |
107.45 ± 5.00 |
| 97.50 ± 1.00 |
99.25 ± 1.39 |
100.62 ± 4.02 |
110.48 ± 3.87 |
| 97.60 ± 0.48 [2] |
99.40 ± 0.65 |
100.62 ± 4.00 |
114.76 ± 4.01 |
| 97.60 ± 1.00 |
99.60 ± 2.50 |
102.57 ± 4.10 |
116.05 ± 1.24 |
| 98.22 ± 2.00 [1] |
99.70 ± 1.10 |
103.10 ± 0.95 |
|
| 98.22 ± 3.22 |
99.72 ± 0.76 [3] |
103.55 ± 4.00 |
|
| 98.35 ± 1.16 |
99.77 ± 0.98 |
103.58 ± 0.72 |
|
The correlations are even more significant than the above list would suggest
on its own. Formations sitting on top of Albian formations date to less than
97 million years; formations sitting below Albian formations date to more than
110 million years.
Not only do the list of Albian ages fall into a consistent range; that range
is in agreement with the ages of formations which were necessarily deposited
before and after -- indicated by simple geological relationships that even
Snelling would agree with. Harland et al. is an entire book of nothing but
this sort of data, and the Albian Stage is just a tiny fraction that I chose
at random.
Further, Harland et al. is merely a top-level summary of the data, packed
with references to technical papers containing the actual measurements. If we
dig into the detail behind these numbers, it gets even worse for the
young-Earth cause. We have volcanic sanidine and biotite from Montana and
Wyoming which sit with late Albian fossils (marked "[1]" above). These contain
a range of concentrations of potassium, and yet give a series of almost-identical
ages around 98 million years. We also have glauconite (a mineral that forms in
clays where deposition is slow, often replacing fecal pellets, shells, and the
like) from Germany which sits with late Albian fossils (marked "[2]" above).
These contain a range of concentrations of potassium, and yet give a series of
almost-identical ages around 98 million years. We also have glauconite from
France which sits with late Albian fossils (marked "[3]" above). Multiple
samples give Rb/Sr ages of 97 to 102 million years, each with about 3 million
years uncertainty.
Why do these samples from all over the world -- matched up by
distinctive fossil composition -- consistently date to similar values by
multiple isotopic methods?
The mainstream scientists' answer to that question is simple: the results
consistently agree because the methods work, and the Albian Stage
represents a span of about 15 million years of time, roughly 100 million
years ago. This answer cleanly explains all of the data discussed above --
the agreement of mutiple samples per location, the agreement of sample
suites from distant locations, the agreement across different dating methods
-- and in fact requires such a patten of results to be observed.
But what is Snelling's answer to the same question? He suggests that the
methods are wildly unreliable, prone to giving random results that are off by
a factor of a thousand or more. He believes that all Albian formations were
deposited a few thousand years ago. What is his explanation for why
there is a consistent pattern of results agreeing on ages that he is certain
are off by over four orders of magnitude? If we are to take his
Impact essay ("it's excess argon") as a response, it begs a
number of additional questions that are going to be quite difficult for him
to answer:
Did all of the dated samples inherit "excess argon" from a mysterious,
unnamed source? How did the samples with the most potassium end up with the
most excess argon so every sample in the suite of ages would yield the same
result? How did the igneous samples ([1]) inherit exactly the same proportion
of excess argon as the sedimentary ones ([2]) on a different continent, so
that both suites of samples would agree? How did the samples with the most
87Rb end up with the most excess
87Sr so that the whole suite of Rb-Sr ages would
agree? Even if we grant the assumption that there is some systematic answer
to matching the K-Ar ages ([1] and [2]), how do the Rb-Sr ages ([3]) -- which
depend on elements with chemistry very unlike that of argon -- get set to
exactly the same numbers? Why do these far-flung groups of rocks yield
"fictitious" ages that agree on the same value for multiple samples per site,
and how did they each get buried with late Albian fossils?
Snelling's generic handwaving about excess argon does not explain the
pattern of results. The young-Earth crowd doesn't have
a sensible explanation for this data. A formally trained geologist like
Snelling must know this, which makes his obfuscation about
K-Ar dating all the less excusable.