Oh? Let's start here... The following is from a debate between Chris Stassen (the author) and a young-earth creationist:
(B) Methods scientists use to give an age for the earth/universe.
I will present three ways to derive an age for the earth:
- We can try to find the oldest rocks on the earth. While this doesn't guarantee an absolute age (for the original rocks need not be available), it can at least give a lower limit for the age of the earth. (Unlike Bob's limits, these are derived by dating a specific object.)
The oldest rocks exposed on the surface of the earth are 3.5 to 3.8 billion years in age. Consider the various dating methods applied to the Greenland Amitsoq Gneiss:
Rb-Sr isochron 3.70 +- 0.14 billion years Pb-Pb isochron 3.80 +- 0.12 billion years U-Pb discordia 3.65 +- 0.05 billion years Th-Pb discordia 3.65 +- 0.08 billion years Lu-Hf isochron 3.55 +- 0.22 billion yearsNote that all of the methods agree (3.68-3.70 is within all of their ranges of error). Isochron and discordia methods also have an internal check which identifies undateable samples. Similar formations which give similar ages can be found as well in North America, India, Russia, Australia, and Africa. This date therefore merits some confidence.If Bob wishes to object to these dates, he will have to explain why a 10,000-year-old rock was "created" so that five independent dating methods would all yield the same fictitious age.
- We can try to date other objects in the solar system. Both sides of the debate believe that other objects in the solar system formed at about the same time as the earth, and therefore an age for one of those objects is an age for the earth.
The moon is not as geologically active (dating should be more reliable, as rocks have less complex "histories"). Again, the original rocks need not be available, so the age will only be a lower limit; the moon must be at least as old as the oldest rocks we've found on it.
Lunar basalts were collected by six different Apollo expeditions, from six different sites. These samples all give ages ranging from 3.16 to 3.96 billion years, by both Rb-Sr isochron and Ar-Ar dating methods. When both methods are applied to one sample, the results agree to within 3%.
Meteorites are not geologically active at all; there is good reason to expect that most are undisturbed since their formation with the rest of the solar system. Faure has a chapter on meteorite dating in Principles of Isotope Geology (this book is a must-read for anyone who wishes to understand radiometric dating).
Chondritic meteorites consistently give an Rb-Sr isochron age of 4.49 +- 0.07 billion years. Achondritic meteorites consistently give an Rb-Sr isochron age of 4.36 +- 0.11 billion years. A combined method using samples of minerals from many different meteorites gives an Rb-Sr isochron age of 4.46 +- 0.08 billion years.
Note that a small percentage of meteorites give ages younger than 4.5 billion years. This is to be expected when events such as collisions cause melting and recrystallization, which would "reset" the radiometric "clocks." Still, most meteorites give the same age, and none give ages older than that.
This arrangement of data is expected if the solar system is indeed 4.5 billion years old. I can't imagine how to explain it if the actual age is 10,000 years. But that is Bob's task - not mine.
Again, if Bob wishes to disagree with the methods, he will have to give specific objections. He will have to explain why meteorites were created to give isochron ages of 4.5 billion years rather than, say, 91 billion years. He ought to have a reason why a 10,000-year-old sample could be expected to give an isochron at all.
- Finally, since we figure all of the objects in the solar system formed at about the same time (as do the creationists), we can construct a "model lead" age. This is a calculation which is performed on various Pb isotopes (some of which are the result of uranium decay, and others which are not). We will plot [207]Pb/[204]Pb vs. [206]Pb/[204]Pb of samples from several different objects (meteorites and earth sites).
If these objects were all formed at the same time from a shared pool of materials, these points should lie on a straight line, and the slope of the line should give the age at which these objects became separated.
If these objects instead had separate origins (for example if they were created out of nothing), then there is no reason to expect the data points to lie on a straight line. Since the age is determined from the slope of the line, a scattering of points prevents any age from being determined at all.
In addition, if some of the samples were contaminated after the separation event, then those points should be moved away from the straight line, and again a meaningful age could not be determined. The fact that the samples do indeed lie on a straight line provides evidence that the resulting date is accurate, and that the samples have not been contaminated.
Y-axis: ratio of Pb[207]/Pb[204]; X-axis: ratio of Pb[206]/Pb[204]. Data points: (1) Iron Meteorites; (2) Beardsley; (3) Modern sediments and young Galenas; (4) Saratov; (5) Elenovka; (6) Richardton; (7) Nuevo Laredo. I can't really do it justice in ASCII, I recommend interested parties to get the original. All of the points lie on (or very near) a straight line. The slope of the line represents an age of 4.55 billion years. The only reasonable explanation for this arrangement of the data is that (1) the objects in the solar system all formed from a common pool of matter, and (2) they became isolated from each other about 4.55 billion years ago.
I doubt Bob has a convincing explanation for how young, "independently created" objects from all over the solar system could have their lead contents form an isochron. I wonder how he will account for the fact that the resulting age matches other dating methods' results for the solar system.
[...]
Summary
We have examined Bob's method for dating the earth. The method has many insurmountable problems, yet it is repeated in several creationist books which argue for a young earth. We have also examined some creationist complaints about the validity of the scientific methods for finding an age for the earth. All are either irrelevant or simply wrong. These are not the hallmarks of an enterprise which has reached its conclusions on careful study of the evidence. These are not the hallmarks of an enterprise which even understands the evidence.
Some accuse "scientific" creationists of ignorance or dishonesty. That may be true in some cases (e.g. Morris trying to palm off wildly inaccurate dust influx rates as reasonable values). A more reasonable explanation is that most of these mistakes are born out of desperation to support a view that the evidence flatly contradicts.
Perhaps Bob can do better than the leaders of the movement. Since Bob usually seems content to work directly and trustingly from their books, I do not expect it. Bob may surprise me and produce a reasonably convincing method which gives a young age for the earth. But if he manages to do so, it will not be a method found in popular creationist literature.
Bob's age for the earth differs by about six orders of magnitude from the value that scientists propose. This is not a minor difference. One of our two positions is like arguing that Alpha Centauri is closer to the Earth than the Sun is, or that one can buy a nice house in California for a quarter. The positions are so far apart that it should be trivial to choose the one that the evidence supports. It is. I have presented some solid pieces of this evidence. Bob needs to propose some evidence of his own, but he must also have testable explanations for how my evidence fits into his position.
Bob will have to offer an explanation for how a collection of young rocks from different parts of the solar system could form an isochron giving an age of 4.5 billion years. He will also have to offer an explanation for how the Amsitoq Gneiss could give the same (incorrect) date by 5 independent methods, all of which passed internal checks.
No matter how many "rocks" Bob throws at radiometric dating, he will need to present an explanation for the curious agreement of all of the varying methods. Proposing a deceptive creator is an admission of failure to do so. [...]
In my opening statement, the dating methods I used were all radiometric. I used radiometric dating because it is the only quantitative method I know for giving an age. There are plenty of geological formations which I could discuss (e.g. varves, fossil reefs and stromatolites, limestone and chalk deposits) that are very difficult to explain as features of a young Earth. I mentioned a few in my rebuttal (e.g. ocean floor sediments). But these formations cannot be used to date the entire earth - which is our topic.
I am somewhat disappointed that Bob didn't provide better objections to the radiometric dating methods. He mostly reused materials that he had posted previously to talk.origins. (I was loaded for bear, but faced gnats. :-) )
I presented only isochron and discordia methods. Bob gave objections to K-Ar methods, which he incorrectly generalized to all methods. For K-Ar dating methods, there is an "assumption that no parent or daughter product has entered or left the system" (technically, the assumption is that it won't happen without leaving detectable evidence of contamination). There is no analogous assumption for isochron methods, as a systematic contamination is the only reasonable explanation for a bad date which keeps the points on a line. That sort of contamination ("the mixing model") is detectable, was tested for, and was found not to be present.
Bob makes the claim that "radiometric dating methods are known to give very inaccurate (6-7 orders of magnitude) results." But this is in a tiny minority of cases. If these methods are so wildly inaccurate, how do five of them agree on the same age? If the five methods gave results similar to the metals-in-the-oceans method (scattered randomly), I would be first in line to call the Amsitoq Gneiss sample "undateable." (By the way, I presented the "raw" results of the methods; no "corrections" necessary.)
Bob also objects that the dates "have never been shown to be correct." They show a strong correlation with each other, with position in the geologic column, and with dates derived by other means. The methods work practically all the time on samples of known ("historical") age which pass contamination tests. I don't know what more Bob expects. He argues we can't "know" the age, but lack of absolute certainty doesn't make 10k years a palatable alternative. (Bob would need to explain how the dates could consistently be so far out of whack. Without that explanation, such an argument is worthless.)
I also want to take issue with Bob's argument against deceptive creation, which was to say that "the Creator didn't design the dating methods." All radiometric dating methods I presented are straightforward mathematical equations derived directly from half-life and isotope measurements (yes, even isochron methods). Agreement of the five methods is "appearance of age" just as surely as if the rock were labeled "3.7 billion years old" (perhaps even more surely, as a label is easier to fake). Bob pleads "misinterpretation," but fails to present any intepretation at all which could account for the sample's actual age being nearly six orders of magnitude lower.
Bob is asking me to believe that the Creator (without deception - by accident?) "initialized" or an unidentified process (acting for <10,000 years) "changed" isotope levels in the Amsitoq Gneiss so that five self-checking methods would yield the same (wrong) age. If this is creationist "science," it's unteachable in public schools even without religious reference. Arguments like that would justifiably get laughed out of any respectable refereed journal. But one can find lots of arguments like that in creationist "scientific" journals. :-(
An old earth provides the best and simplest explanation for the Amsitoq Gneiss dates, the Solar System model Lead results, the pattern of ocean-floor sediments, and the youth of short-period comets (see my rebuttal for the last two). A 10,000-year-old earth does not explain any of these things easily. [Note that I can use Bob's own evidence to contradict his proposed age!]
If the Earth were young, and the Creator wanted us to believe it, then all five methods applied to the Amsitoq Gneiss would give an age of 10k years. Anyone examining the evidence independent of religious conviction cannot escape the conclusion that the Earth is very old (Harold Coffin, a creationist witness at the Arkansas trial, admitted that under cross-examination).
In summary, Bob dismisses radiometric dates mainly "because they are not known to be correct." This argument holds no water because he failed to explain how the dates could systematically be wrong. It is merely a naked handwave, without any "scientific" hypothesized mechanism to support it.
Furthermore, this dismissal (on the grounds of "lack of certainty") implies that Bob's belief in a young earth is based on something which will override "non-certain" - but solid - evidence. I wish Bob had discussed whatever it is he finds so convincing. His seven dating methods certainly don't merit such an investment of confidence; I bet he would still believe in a young Earth even if he were forced to admit his methods to be unreliable.
