Joined: Oct. 2005
Sorry for overkill, but as I said, I like radiometric dating, and over at Smilodon's Retreat, Cryptoguru is wrong out of the gate and rapidly gets worse:
|There are 3 faulty assumptions with radiometric dating ... all 3 have been proven to factor into dating inaccuracies|
1) we know the initial conditions
2) there has never been any contamination
3) the decay rate is constant
Three out of three statements that are misleading or wrong right up front! #2 is not an assumption; #3 is a conclusion that is known to work in crustal conditions but is not claimed for other situations; and #1 is a testable inference that is re-investigated with each study.
First, decay rates, like any other process, can of course be sped up or slowed down. This is not news. Some types of acceleration of decay rates are well known: we call them nuclear bombs and nuclear power plants. The trouble is that to accelerate decays appreciably require the sorts of temperatures and pressures found inside stars, or massive bombardment by decay products nearby concentrations of other radioactive isotopes, as at the Oklo natural nuclear reactor site. On the basis of experiments and calculations, the comparatively mild conditions of metamorphism (2,000 degrees K and 100 kbar should melt pretty much anything in the crust) aren't sufficient to influence the process. Intense gamma ray bombardment can trigger decays in some isotopes (as with neutrons in other isotopes). Embedding in metal and cooling to a few degrees above 0K can do in alpha emitters. You get the idea: there are ways to speed these up, as with any physical or chemical process, but there aren't any that will operate in crustal conditions without leaving evidence of an Oklo type reaction or a nuclear bomb going off. In short, in crustal environments, they're constant, so a) decay rates are known to vary but b) this doesn't happen under crustal conditions absent an Oklo reactor.
Second, there's always contamination. It's inevitable. This is not news either but again it is not a fatal flaw for all of radiometric dating. The questions are how much, can we evaluate it, and can we adjust for it? If you handle a dinosaur bone with bare hands, oil and sweat and dandruff can contaminate it and give the bone a very old radiocarbon age (30,000 years or older for trace contamination by modern organic carbon). If the bone was prepared with shellac or glue or cleaned with detergent or carried in a burlap bag or picked up with leather or cotton gloves, or got mold growing in it in the museum basement, or (most commonly) was invaded by plant root hairs while lying in the soil prior to discovery, that's potentially significant contamination. But those are known issues and are easily avoided: don't do those things, and extract the sample for analysis by drilling a sample out of a pristine part of the interior, treat it with peroxide, and inspect it for any root hairs, fungal hyphae, etc. If the sample is in bad shape, then you can't date it. However, there are even more subtle and pernicious sources of contamination: the lubricant that you use in your instrument's vacuum pump; the grease that you use to seal your access port; does your instrument have any plastic piping?; modern carbon in the acids you used to dissolve the sample, and so on. These can all be mitigated to a degree, but ultimately there will always be enough contamination to give some kind of a very old date to something like diamond dust or ancient coal. The nature of exponential decay is that a little contamination will not throw off a date for a sample that only a few half-lives off, but as you get to 40kyr to 60 kyrs, there is so little original C14 left that contamination will be a problem. Several decades ago, people hoped to get improved C14 dates on materials 100 to even 200 kyr by individually counting C14 atoms with a cyclotron (accelerator mass spectroscopy) dating, which did extend dating ranges a little, but contamination issues dashed those hopes. 40-60 kyr is about the limit, depending on the quality of the sample and other issues.
A reverse example of contamination is if you have a modern lacustrine clam or snail living in a lake on limestone bedrock, it may well be getting some or most of its carbon not from atmospheric CO2 that becomes dissolved in the lake water but from "dead" carbon i.e. carbonate via dissolution by the lake water of the calcium carbonate in the bedrock. That way, living organisms can easily date as 10-20 thousand years old. Moral, you usually can't reliably date things living on limestone.
Other decay systems have other contamination issues and other solutions. K-Ar for igneous rocks is very nice for this: not only is argon a natural gas, which won’t link up in crystals, but argon in the magma will bubble out before it has a chance to get trapped in a crystal. However, once in a while argon will flood up a hotspring pipe or a fracture or an apatite vein and will soak into adjacent crystals, but those issues are expected in those situations and can be tested for by testing samples progressively farther from the fracture (and then throwing out the contaminated samples, or simply not sampling near veins and the like in the first place). For U-Pb and Th-Pb, we rarely any longer try to do whole-rock analyses because there’s there’s too high a possibility of some old lead from previous decays that got into the magma from melted host rock. Magmas do in part melt their way up to the surface, so a lava that erupted 10 million years ago can have picked up chunks of 2 b.y. old rock that it passed through (the chunks may be visible xenoliths) or a few crystals that melted out of the host rock but floated around in the magma (more on those in a moment), or simply melt products that included some old radiogenic lead from long-ago decay events. The last is resolved by no longer doing whole-rock dates. Instead, we pick out crystals of minerals that for various reasons will incorporate the parent isotope but won’t naturally incorporate the daughter element, such as zircon and monazite. Zircon has the added advantage of forming after most of the lead has gone into other minerals, of having a very high melting point, of being chemically and physically very resistant to weathering and breakage. This means that it won’t leak daughter radon during the decay process, and is comparatively hard to mess up during mild metamorphism. As I mentioned, it is possible for old zircons to melt out of host rock, not dissolve in the magma, and get included in the new igneous rock, or to be recycled though an even older rock cycle. However, when this happens, you get a corrosion rim around the old crystal and then a new growth band around the old core with nice new crystal faces: these are easily recognized, and we date several sample spots across the crystal and can see date the old core and the new rim separately
It’s an extremely cool technique.
Third, ‘it is wrong to assume that you know the initial conditions’ Well, that’s rather vague. We can melt granites and gabbros and so forth in the lab, so we know conditions of crystallization. Crystal chemistry and mineralogy give us constraints on whether daughter isotope could be incorporated in a growing crystal or not. Sure, if you let a zircon grow in a puddle of lead, some lead would get incorporated even though incorporation is thermodynamically strongly disfavored. On the other hand, your sample would be in a lead ore body, and you’d know that. You can count atoms of uranium-235 and atoms of lead-207, but also atoms of other isotopes of lead as well. If we don’t have any nonradiogenic lead (isotope 204) then we wouldn’t have had any original 207 either. If by chance we did have some PB 204, we could use the amount and typical ratios between the isotopes to correct the radiogenic lead contents by an appropriate ratio and then attribute the excess to decay of uranium (I’m oversimplifying a bit here). There are some other techniques like Pb-Pb dating that can be used.
Still on #3: It is always possible that something else went wrong that we haven’t yet anticipated or corrected for. Most typically for Precambrian rocks, we had a little later metamorphism that caused partial to total recrystallization, kicking out some to all of the daughter isotope formed since the original crystallization. We have multiple lines of defense here including one truly awesome technique that gives us two useful bits of information for the price of one. First, we don’t just date one sample, and we don’t just date by one technique. All radiometric decay chains decay at their own unique decay rates. This means that it is impossible to mess up a crystal in such a way as to force a wrong date*, while having two different decay chains agree on the same wrong result. (*If you completely melt the crystal, you will force out all the accumulated daughters for all decay chains, resetting the clocks to 0, but the new crystal is brand new, so 0 is now the correct age.) So if different isotopes in the same crystal give you the same dates, you’re golden. We typically date multiple different decay chains in multiple different crystals in a rock, which always give different dates because different minerals crystallize at different temperatures, and a granite can easily take tens of millions of years to cool from crystallizing a temperature with a high melting point to another with a low melting point. That’s a colling history for the granite, and we expect it to show a logical progression, so if that is present, good. Most important, we can calculate how the U-235/Pb-207 ratio should be changing over time and how the U-238/Pb-206 should also be changing over time, if everything worked. This gives you a concordia curve, and if your sample has exactly the right pair of ratios then everything is peachy keen
When rocks experience a little metamorphism, the crystals can be partially reset, by losing a fraction of their daughter isotope. The exact fraction is going to vary: more from little crystals than big ones (it’s harder for lead to diffuse out of the center of a large crystal); more from broken crystals or crystals with cracks than from pristine ones; and so forth. This results in discordant dates. However, the result of different percents of losses is typically a line (or long narrow field) of dates on the concave side of the concordia curve.
The math for this is inexorable: where the line through the discordant points hits the older part of the concordia curve, that would be the age for crystals that have lost 0% of the daughter, i.e. the original age of crstallization. Where the line hits the younger end of the concordia curve, that would be the age of crystals that lost 100% of the daughter atoms, i.e. the age of metamorphism.
|We can't possibly assert that we can prove 1 or 2 because neither is observable, except we can maybe enforce 2 in the case of diamonds. We have cases where we have later proved that our initial assumptions for 1) and 2) were wrong ... so how do we know when we've got it right?|
For 3 we do know that decay rates can be effected by external factors
Well, that’s you making bogus arguments because they sound good to you, rather than because you understand what’s going on.
|So answer me the following|
A) how is there C14 in diamonds?
B) how is there C14 in dino bones?
C) why do professionally processed rock datings on the same sample give WILDLY varying results?
D) why do professionally processed rock datings on known age samples (100 years) give billions of years?
I’ve answered most of these already, but there’s a little more to say on all of them, especially diamonds.
C & D) Stratigraphically young lavas that passed through much older rocks on the way up can get contaminated in terms of chunks of old rock (recognizable in outcrop and thin section, so don’t take dates of xenoliths as dates of the rock), & old crystals that get melted out of the host rock and end up in the new rock (SEM / EDAX examination, & date the cores and the rims separately). Different minerals in a single rock can and should give different ages when they have different crystallization temperatures, meaning that they formed at different times, and one with the lower melting temperature can get reset by low-grade metamorphism that will not affect the more refractory crystal.
There are some excellent and famous examples in Hawaii, where comparatively young lavas have yielded ancient dates. Hawaiian volcanoes are “hot spot” volcanoes, with magmas coming up from mantle depths, and they have a habit of picking up chunks of mantle and lower crust as they come up. These are easily identified, and can be avoided or dated as you wish.
http://academic.emporia.edu/abersus....ith.JPG (not from Hawaii)
Creationists implied that these are legitimately confusing dates: they cited a study of the xenoliths as if it were a study of the lava, and they failed to point out that these sorts of studies are done routinely to see if a rock can be dated, and the authors concluded that their data showed that dating could not be done and that their results had no geological meaning.
Most of your “data” here is likely to come from the creationist literature, which involves a truly mind-blowing quantity of lies, misrepresentations, misunderstandings, and mistakes regarding legitimate work done by legitimate sciencists. These are dealt with in a wonderfully readable essay by Brent Dalrymple, available at
Note in particular how creationists use data such as tests of a potential dating system where the scientists decided it was too vulnerable to getting screwed up so they concluded that it should not be used, but the creationists used it as evidence of geologists being unable to justify radiometric dating:
|The two ages from gulf coast localities (Table 2) are from a report by Evernden and others (43). These are K-Ar data obtained on glauconite, a potassium-bearing clay mineral that forms in some marine sediment. Woodmorappe (134) fails to mention, however, that these data were obtained as part of a controlled experiment to test, on samples of known age, the applicability of the K-Ar method to glauconite and to illite, another clay mineral. He also neglects to mention that most of the 89 K-Ar ages reported in their study agree very well with the expected ages. Evernden and others (43) found that these clay minerals are extremely susceptible to argon loss when heated even slightly, such as occurs when sedimentary rocks are deeply buried. As a result, glauconite is used for dating only with extreme caution. Woodmorappe’s gulf coast examples are, in fact, examples from a carefully designed experiment to test the validity of a new technique on an untried material.|
Another classic example of abuse of this sort involves the Plateau and Cardenas Basalts in the Grand Canyon: see http://www.talkorigins.org/faqs.......ce.html for details.
B) How is there C-14 in dinosaur bones? First, see the answers for diamonds below. Second, contamination: a very little modern Carbon can give a very old age to a rock with no carbon of its own. Contamination is hard to avoid in the best of times. There is also an issue of trust: people who want young ages and fib about other aspects of science perhaps shouldn’t be trusted to have handled the samples carefully enough to avoid contamination, when just a little contamination could appear to validate their beliefs. Third, misuse of data. I attended a creationist talk in the late sixties where the creationist presented a C14 date on a geologically ancient fossilized wood and showed a lab report that he claimed said the fossil was only 36000 years old. Unfortunately, the photo of the lab report showed the lab report saying “>36000 yrs" (or something close to that), which I presume was radiocarbon infinity for that lab at that time, and in any case just says “>”. However, I haven’t seen any creationists do that since, and this guy was a jaw-droppingly sorry example (he also complained about paleontologists not being able to produce any transitional fossils between a fish and a starfish).
A) How about C14 ages on diamonds? Scientists use industrial diamonds to test their C14 instruments and methods precisely because diamond should have very little C14 in it. (They have also used coal and oil.) This is how they learn how much contamination is entering their samples and whether it’s a problem. They publish their results to show the resolution achievable in their labs (and bear in mind that if you have so little contamination that you only get a date of 50,000 years, you have done a very good job of avoiding contamination, and dates only a little less than that will be trustworthy). Typically, chemical preparation of samples adds about 1 microgram of modern carbon, so a standard sample of 1 milligram will inevitably have 0.1% modern carbon in it, which will give a date younger than radiocarbon infinity (57136 yrs, to be precise). It is therefore dishonest of creationists to say that this proves that all C14 dating is problematic, because the work shows that dates are reliable nearly up to those ages, but not beyond. Because decay is exponential, “backing up to the recent” a little means that any samples of a slightly younger age will have a LOT more of their own original C14, so the trace amount of C14 that your procedures or machine introduced will be insignificant. There is a great treatment of these issues by Karl Bertsche:
Bertsche has an interesting discussion of “instrument background” (the values reported by an instrument with no sample in it). This includes
ion source “memory” of previous samples, due to radiocarbon sticking to the walls of the ion source, thermally desorbing, and then sticking to another sample
mass spectrometer background, non-radiocarbon ions that are misidentified as radiocarbon, sometimes through unexpected mechanisms
detector background, including cosmic rays and electronics noise
I would also add that coal and diamonds contain nitrogen as a trace component (up to 1% of a diamond is nitrogen). If that gets bombarded by beta particles (neutrons, from nearby radioactive atoms), then some nitrogen could get converted to C14. I don’t know if the amounts are significant, however.
These topics are complicated, way more than I presented here, so there is always endless scope for trying to throw doubt by bringing up more complications. Geochronologists are always testing their systems, investigating situations where something didn't work, and trying out new techniques and new dating methods, where some of them work while others turn out to get discarded as unworkable or problematic. Regardless, creationists have made a dishonorable but standard practice of taking those complex results where scientists determined that a potential technique was unworkable or where they determined why an enigmatic result happened and then representing them as evidence that dating overall does not work.