More about Potassium Argon Dating....

[youngborean]

http://id-archserve.ucsb.edu/Anth3/Cour ... ating.html


There is something that I suspected of these dating methods, that the study posted by the Creation researchers seemed to point out. It is best illustrated by some information on the website I posted. We move out of statistical relavance when using this method to date anything less than 100,000 years. That coupled with the idea that precentage errors can mean a lot of time when we are talking exponential decay. So my point is this. How can we even say that a volcanic rock is less than 100,000 years compared to 1,000,000 years old assuming some error, etc.? Is then a fair method to use to refute suggested dates for an early earth since the method is unable to detect a volcanic rock that is early, like 7000 years ago? Choosing this method must therefore assume that a rock is older than 100,000 years old. Now I would argue that the research project on answersingenesis that shows the volcano from 60 years ago being analyzed as being 1.5 million years old is not novel science. That is becasue that result is inherent in the margin of error. It only clearly showed the limitation of this method.

[juliod]

We move out of statistical relavance when using this method to date anything less than 100,000 years.

That a date like 100,000 BP could exist is sufficient proof that creationism is wrong.

Even if the method is wildly inaccurate, creationism is still wrong.

DanZ

[youngborean]

That a date like 100,000 BP could exist is sufficient proof that creationism is wrong.

That would be true if the logic of the method was inclusive. I could be way off and there could be another method that is inclusive. But if you and I were to have a method that could disprove Creationism by dating methods, shouldn't the method be able to measure the dates that creation claims as well as the dates that you or geologists are claiming? So if all the volcanic rocks were less than 100,000 years it would be impossible to tell from this method. Also that date "exists" only by extrapolation since we are unable to prove it, or the actual decay rate of Potassium-40 to Argon-40. The fact remains you must first assume that the rocks are "old" when testing them, because we couldn't even measure them if they were young.

[juliod]

But if you and I were to have a method that could disprove Creationism by dating methods, shouldn't the method be able to measure the dates that creation claims as well as the dates that you or geologists are claiming?

No. Real geologists don't give a fig what creationists think. The young earth was falsified a century ago by geologists and none of them (0%) are still considering the issue.

It's a simple point. You shouldn't use a technique that is only accurate for dates over 100,000 years to examine samples that are less than that age. It's not possibel to go from that point to "creationism is true". No matter how you view it, creationism is still wrong.

DanZ

[youngborean]

It's not possibel to go from that point to "creationism is true".

I am not trying to prove creationism. I am only saying that K-Ar dating cannot disprove it. That would be as impossible as it is to prove creationism.

No. Real geologists don't give a fig what creationists think. The young earth was falsified a century ago by geologists and none of them (0%) are still considering the issue.

And what evidence are they using to establish their belief? Hopefully new ways of analysis can be developed so we can truly ask the age of the Earth.

[ENIGMA]

http://id-archserve.ucsb.edu/Anth3/Cour ... ating.html


There is something that I suspected of these dating methods, that the study posted by the Creation researchers seemed to point out. It is best illustrated by some information on the website I posted. We move out of statistical relavance when using this method to date anything less than 100,000 years. That coupled with the idea that precentage errors can mean a lot of time when we are talking exponential decay. So my point is this. How can we even say that a volcanic rock is less than 100,000 years compared to 1,000,000 years old assuming some error, etc.? Is then a fair method to use to refute suggested dates for an early earth since the method is unable to detect a volcanic rock that is early, like 7000 years ago? Choosing this method must therefore assume that a rock is older than 100,000 years old. Now I would argue that the research project on answersingenesis that shows the volcano from 60 years ago being analyzed as being 1.5 million years old is not novel science. That is becasue that result is inherent in the margin of error. It only clearly showed the limitation of this method.

The thing about the margin of error when dealing with half-lives is that the same margin of error for the daughter/parent isotope ratio can yield many different margin of errors for chronological dating depending on the observed ratio.

The reason that Carbon dating becomes unreliable after 60k (Off the top of my head so don't sue me if it's off) years is that while a margin of error of say 1% leads to a fairly small margin of error dating wise when only half the sample as decayed, but a massively huge one when the sample is almost completely depleted (1% is unacceptably large when attempting to determine whether 1/256 or 1/512 is closer to the correct ratio of the sample which makes a few thousand years worth of difference time wise). Similarly attempting to get measurements from isotopes with little such decay runs into similar problems, just flip the ratios.

What needs to be presented to make a determination of accuracy is not only the date specified but a margin of error that is translated into differing time measurements. For the volcano sample I would suspect that the latter is huge.

[youngborean]

The reason that Carbon dating becomes unreliable after 60k (Off the top of my head so don't sue me if it's off) years is that while a margin of error of say 1% leads to a fairly small margin of error dating wise when only half the sample as decayed, but a massively huge one when the sample is almost completely depleted (1% is unacceptably large when attempting to determine whether 1/256 or 1/512 is closer to the correct ratio of the sample which makes a few thousand years worth of difference time wise). Similarly attempting to get measurements from isotopes with little such decay runs into similar problems, just flip the ratios.

That is the correct logic. It deals mostly with noise in instrumentation. It is mostly based on a statistical rule that invadlidates results when measuring differences that are .001% of the original sample size. Because the confidence interval becomes essentially nil at this point. Depending on instrumentation, the rules move a bit as in the page I posted K-Ar used .005%. I was using the example of K-Ar dating to show that there is no statistical relavance for early dates within this method that is all. The margin of error would be huge in these early range and would not barely effect the result like Carbon, but would result in a diference that could be off by millions of years. Ideally we would have a method in my mind that could be used at all time that could tell the age from 0-10billion years old exactly, but we are not perfect and can't live long enough to check our estimates.

[juliod]

I am not trying to prove creationism. I am only saying that K-Ar dating cannot disprove it.

But it does. (We need an icon for jumping up and down screaming.)

There mere fact that something could be dated to 100,000 BP is sufficient evidence that the earth is greater than about 6000 years old. YECism is completely wiped out.

It doesn't matter that this form of dating cannot resolve an age of 6000 years from 7000 years. If it can resolve an age of 200,000 from 300,000 that is sufficient.

Let me try an analogy. Imagine you have a device for measureing the speed of rifle bullets. You find, after much effort, that it is incapable of measuring the speed of inchworms. Someone says to you that 5 mm/s is the maximum possible speed for any object. You say "nonsense, here's my evidence of bullets moving at >1000 m/s". They come back with "But your machine can't measure the speed of an inchworm, which is the fastest land animal."

Would accept that they might be correct?

DanZ

[youngborean]

How can you ensure that those rocks were not less than 100,000 years? I don't think you understand what happens when you test around control limits. But usually becasue of saturation of whatever probe you are usuing you get anomolies. But either way you must first assume that the date of the rock is over 100,000 years old. It proves nothing, because it is based first on an assumption. Let's say you have a method for dating people, but it only works for people beyond 80 years. However, if all the test subjects were less than 80 years, the results would not work, they would show an age beyond 80.

A creationists assertion is that all rocks are under 100,000 years old. The only way you could disprove that theory is to have a test that could test in that range and past that range to establish a control. If no control is established, i.e a verifiable date through this method, how can it be accepted as an accurate mesurement of time? You can have evidence that these rocks may be older than 100,000 years, assuming that they are not less than that age. You can call it theorethical date but it doesn't disprove dates as you claim. Maybe you are supposing that there would be a less than 100,000 year measurement that would be obvious on the data plot. But I highly doubt it would. How can you disprove a suggested date without being able to measure that date? I am not arguing that this method can give theoretical dates beyond 100,000 years, because it clearly does. But if you choose this method to measure a rock, you must first assume that the rock is over 100,000 years old.

[bigmrpig]

It proves nothing, because it is based first on an assumption. Let's say you have a method for dating people, but it only works for people beyond 80 years. However, if all the test subjects were less than 80 years, the results would not work, they would show an age beyond 80.

Well, the problem lies in that there's no way to actually know without a shadow of a doubt how old a rock or anything is without watching it being created and timing it.

If you want to be able to date something at all, you have to accept the results and not say they're wrong, because let's say I have something that dates 10,000 years ago with perfect accuracy. How can you confirm that it's 10,000 years old, or that it's 8,000 and I'm just getting a "perfectly" false reading. Any system could theoretically be completely wrong without us know. You either accept that the systems are not possible to be proven without a shadow of a doubt, or not date stuff.

But where does it say that the K-Ar dating system would falsely show higher? The website orginally linked says that the amount of Argon in a sample younger than 100,000 years would be too small to detect. Meaning that we would be able to tell it's a fresh rock, because there's almost no Argon. An old one will have the Argon in it, a new one wouldn't. How would you get a false positive for an old rock? You could get a rock that could look new while actually being old, but how would you mistakenly read that the rock is millions of years old when it's new? The argon wouldn't be there...