Re: RATE

From: Steven M Smith (smsmith@usgs.gov)
Date: Wed Oct 15 2003 - 19:41:47 EDT

  • Next message: Sarah Berel-Harrop: "Re: Phillip Johnson (and Methodological Naturalism)"

    Allen Roy wrote:
    > Steven M Smith wrote:
    >> Allen Roy wrote:
    >>> 1. The acceptance or rejection of isometrically acquired ages for rock
    >>> depends upon factors other than the science/technology of isometric
    >>> dating.
    >>
    >> I agree with this point ... but so what? Radiometric dating based on
    >> isotopic compositions in rocks is a tool, not a panacea. Radiometric
    >> dates are only one piece of evidence concerning the age of a rock.

         ... Snipped list of other types of evidence ...

    >> When the radiometric date doesn't agree
    >> with the other lines of evidence, then it may be rejected. But that
    >> is not the end of the story, it is just the beginning of trying to
    >> figure out why the test failed.

         ... Snipped list of new research ideas based upon a test failure ...

    > All of it [the "explanations" above] are nothing but pure rationalization
    > needed to try to save the fatally flawed methodology. As Woodmorappe
    says,
    > "CMBN" -- If it works, Credit Methodology, it if doesn't, Blame Nature to
    > save methodology. If it doesn't work here, why should it work somewhere
    > else? Why trust it anywhere?

    Allen,
    What you appear to be demanding is a perfect method that can be used by
    anyone, anywhere, and on anything and that always gives an exact
    non-ambiguous answer. Unfortunately, geology (and probably all other
    sciences) has yet to develop such a fool proof method. I guess you must
    have the perfect car (if it doesn't work sometimes, why trust it
    anywhere?), the perfect auto mechanic with the perfect auto diagnostic
    tools, the perfect doctor with the perfect medical tests, and the perfect
    investment broker with the perfect investment strategy. Regrettably that
    is not my experience. My experience is that all methods and tests have
    flaws, interferences, or conditions under which they do not work -
    regardless of whether these are geological, chemical, medical, financial,
    or mechanical procedures. Perfection is an ideal. In the real world, we
    generally use methods that work most of the time and then try to identify
    those conditions under which they won't work so that we don't waste our
    time and resources.

    Allen wrote:
    > You say, "Is the radiometric date telling us something other than simply
    > the time the rock was emplaced or extruded?" But it is the BACKBONE
    > assumption of isometric dating that the emplacement and extrusion resets
    > the clock. If it doesn't reset the clock here, then why should I
    believed
    > that it reset the clock elsewhere. Rationalizing it away when it doesn't
    > work is pseudo-science.

    Wrong BACKBONE assumption. Some isotopic clocks are reset during
    emplacement or extrusion but not all. Some may additionally be reset by
    subsequent reheating or other processes. It is important to know the
    details for each isotopic clock. The assumption behind Strontium
    (Sr)/Strontium isochron dating is that the clock is totally reset when a
    melt is isotopically homogenized and that may not necessarily be the time
    of extrusion.

    I wrote:
    >> So, in effect, it appears that you are presenting the perfection of
    >> radiometric dating as a strawman argument. Should a radiometric dating
    >> procedure ever give us a result that doesn't agree with the other
    evidence,
    >> then all radiometric dating is invalid. We can then ignore all
    radiometric
    >> dates regardless of corroborating or conflicting evidence.

    Allen replied:
    > If your assumptions are wrong, your conclusions are wrong. If your
    > assumptions are wrong part of the time, your conclusions are wrong ALL of
    > the time, because there is no way to know, using the method, when they
    are
    > correct or when they are wrong. (T AND T = T) AND ( T AND F = F) = False

    Your conclusion here is wrong because your assumption is wrong. There ARE
    ways of knowing, using the method, when the assumptions of the method are
    correct or when they are wrong.

    Let's examine the next point:
    >>> 3. If any rock unit is believed or known to be young, then any old
    >>> isometric age is rejected, not as inaccurate, but as irrelevant.
    >>> EVERYONE recognizes that the computed age of the Unikaret basalts
    >>> based on Rb-Sr cannot be the age of the crystallizing of the lava
    >>> flows in the "Cenozoic" because the "law of superposition" overrides
    >>> the isotopically derived age.
    >>
    >> This is incorrect. The radiometric age may be inaccurate but it is not
    >> irrelevant. It is telling us that there are additional factors to be
    >> considered. Your example of the young Uinkaret basalts and Austin's 1.3
    >> billion year age is an excellent example of what I am referring to (as
    well
    >> as evidence that you did not completely read or consider the criticisms
    of
    >> Austin's work at <www.talkorigins.org/faqs/icr-science.html>). The date
    >> that Austin obtained may be a valid radiometric test result but the
    method
    >> he used does not give the age of the lava flow sitting at the surface
    but
    >> more likely the age of the source material in the mantle below the Grand
    >> Canyon. (See the criticism labeled "The wrong meaning is assigned to
    the
    >> dates.") The tests and the result may be valid but Austin has
    >> misinterpreted the data to mean that the radiometric data is giving the
    age
    >> of the multiple flows and, since that age is stratigraphically
    >> unreasonable, then all radiometric dating is invalid.

    Allen replied:
    > So 1.3 billion years for rock that could only have crystallized a few
    > "million years ago" is only "inaccurate?" Any technicians who processed
    > the Uinkaret lavas so inaccurately would not be working much longer in
    > that field.

    Does that last statement include the work of Steven Austin?

    > There is nothing inaccurate in the methodology of measuring ratios of
    > isotopes and computing ages. No one is going around saying that the
    > measurements are inaccurate. Rather, they say that the computed dates
    > do not have anything to do with the actual age of the crystallization
    > of the extruded lava. This simply means that the computed ages are
    > irrelevant to the age of the lava.

    And this may very well be the case sometimes. I would argue that these
    cases are relatively rare while your argument appears to be that ALL
    computed ages are irrelevant.

    > You said, "The date that Austin obtained may be a valid radiometric
    > test result but the method he used does not give the age of the lava
    > flow sitting at the surface but more likely the age of the source
    > material in the mantle below the Grand Canyon."
    >
    > That is nothing but pure rationalization.

    You call this a rationalization only because you insist that Sr isotopes
    can never tell us anything except for a date of emplacement or extrusion.
    The above quote is not actually mine but a summary given in the TalkOrigins
    (T.O.) FAQ on Austin's Grand Canyon Dating project. <
    http://www.talkorigins.org/faqs/icr-science.html>.

    I wrote:
    >> Let me give another non-geological analogy. The Uinkaret basalts are
    >> like a fruitcake (except that the fruitcake is probably older! <grin>).
    >> Just like the fruit in the fruitcake, these basalts include pieces of
    >> rock from other sources (known as xenoliths or literally "foreign
    >> rocks"). According to the TalkOrigins FAQ, Austin has deliberately
    >> chosen a sampling method that dates the fruit and then interpreted
    >> that as the age of the cake.

    Allen replied:
    > Sorry: The T.O. FAQ says "phenocryst" not "xenolith"

    See my previous post entitled, "Fruitcake analogy retraction (was RATE)"

    > And, as is typical of T.O., they misinform and mislead by saying that
    > a phenocryst is a "large mineral which likely formed in the magma
    > chamber." That is not the definition of phenocryst --i.e. "a large
    > mineral crystal surrounded by smaller ones." It is pure speculation
    > by T.O. that the larger crystal was formed in the magma chamber.

    T.O. was following the accepted genesis of phenocrysts. It is a textural
    term (Phenocryst: "A relatively large and ordinarily conspicuous crystal of
    the earliest generation in a porphyritic igneous rock." Dictionary of
    Geologic Terms, Doubleday Anchor Publishers) with an implied mode of origin
    ("... phenocrysts originated at depth and belong to the early stage of
    crystallization, when the magma was hot and thin ...", Huang, W.T., 1962,
    Petrology, McGraw-Hill, p. 55.)

    > Austin did not go looking for old ages. He did the very same thing
    > that any other geologist would do in selecting crystals to find a
    > single mineral crystal age of the rock. T.O. implies that if Austin
    > had used some on the smaller mineral crystals he would have not
    > gotten the old age. However, 5 whole rock samples (made up
    > primarily of the smaller mineral crystals) were also dated. They
    > all are in close concordance with themselves and the single crystal.

    While following up on the xenoliths/fruitcake problem I went back and read
    the original papers on which Austin based his first IMPACT article <
    http://www.icr.org/pubs/imp/imp-178.htm> and saw immediately how he culled
    the data to obtain the results he wanted. That data also suggests an easy
    way to obtain a false isochron by choosing samples from selected lava
    flows. I will expand this statement.

    The original paper by Leeman (1974) was a study of strontium isotopes as a
    method of determining the source material for a series of basalt flows. It
    was not intended to be used to date any flow. In fact, Leeman specifically
    says, "There is no systematic temporal variation in Sr87/Sr86 ratios."
    Leeman collected samples from 20 basalt flows: 12 Hawaiite flows, 5
    Basanite flows, and 3 Quartz-bearing Basaltic Andesite flows. (Note: to
    actually determine a valid Sr87/Sr86 isochron date of extrusion one would
    want to collect several rock samples from only ONE flow and not multiple
    flows.) Leeman measured the Sr isotope compositions for his 20 flows and
    concluded that:

    "All but four of the analyzed western Grand Canyon basalt samples have
    Sr87/Sr86 ratios (0.7028 to 0.7041) and chemical compositions that are
    clearly compatible with a mantle origin for their parental magmas. ....
    Four Grand Canyon basalt flows have relatively high Sr87/Sr86 (0.7046 to
    0.7069), Rb/Sr, and K/Rb ratios that provide permissive, though not
    conclusive, evidence for crustal contamination of their parent magmas."

    Leeman included a table and a series of X-Y plots for his data. One plot
    was a Sr87/Sr86 v. Rb/Sr plot similar to that used to construct isochrons.
    The data all fall pretty much in a cluster except for 4 samples (3 Hawaiite
    flows and 1 Basanite flow). These are the 4 outliers that he concludes
    were contaminated by crustal material. On the diagram, Leeman added a
    reference line for a 1.5 B.Y. isochron and pointed out that the 4
    contaminated flows crudely fell along that line. He hypothesized that
    "This relation could mean that these four flows have been contaminated by a
    crust ~1.5 B.Y. old ...."

    Guess which samples Steve Austin chose to include in his diagram in IMPACT
    178 <http://www.icr.org/pubs/imp/imp-178.htm>? Yep, the 3 Hawaiite outlier
    flows plus 3 other Hawaiite flows from the cluster! These 6 points roughly
    approximate an isochron line. Austin deliberately left out the other 6
    Hawaiite points which would have shown a significant deviation from the
    sketched isochron line!

    All anyone would need to do to collect new samples and create a
    "fictitious" isochron is to specifically choose flows originally sampled by
    Leeman that fall along a desired line. From Leeman's plot, you could
    easily select flows to create a "fictitious" isochron for 1.5 B.Y. or 0
    (zero) years or even for a date in the future! (a negative isochron).

    This kind of picking and choosing of data points (obvious when comparing
    Leeman to IMPACT 178) with the possibility of deliberately skewing a future
    isochron by selectively picking flows to sample does not inspire any trust
    in the results of RATE project with respect to dating Grand Canyon basalts.

    Steve

    References:
    Leeman, W.P., 1974, Late Cenozoic alkali-rich basalt from the Western Grand
    Canyon area, Utah and Arizona: Isotopic composition of strontium, Geol.
    Soc. of America Bull., v. 85, p. 1691-1696.

    [Disclaimer: Opinions expressed herein are my own and are not to be
    ascribed to my employer]
    _____________
     Steven M. Smith, Geologist, U.S. Geological Survey
     Box 25046, M.S. 973, DFC, Denver, CO 80225
     Office: (303)236-1192, Fax: (303)236-3200
     Email: smsmith@usgs.gov
     -USGS Nat'l Geochem. Database NURE HSSR Web Site-
      http://pubs.usgs.gov/of/1997/ofr-97-0492/



    This archive was generated by hypermail 2.1.4 : Wed Oct 15 2003 - 19:41:56 EDT