Well, let's see:
We can take a sample of air, count how many 12C atoms there are for every 14C atom, and calculate the 14C/12C ratio. Because 14C is so well mixed up with 12C, we expect to find that this ratio is the same if we sample a leaf from a tree, or a part of your body.
In living things, although 14C atoms are constantly changing back to 14N, they are still exchanging carbon with their surroundings, so the mixture remains about the same as in the atmosphere. However, as soon as a plant or animal dies, the 14C atoms which decay are no longer replaced, so the amount of 14C in that once-living thing decreases as time goes on. In other words, the 14C/12C ratio gets smaller. So, we have a "clock" which starts ticking the moment something dies.
Obviously, this works only for things which were once living. It cannot be used to date volcanic rocks, for example.
The rate of decay of 14C is such that half of an amount will convert back to 14N in 5,730 years (plus or minus 40 years). This is the "half-life." So, in two half-lives, or 11,460 years, only one-quarter of that in living organisms at present, then it has a theoretical age of 11,460 years. Anything over about 50,000 years old, should theoretically have no detectable 14C left. That is why radiocarbon dating cannot give millions of years. In fact, if a sample contains 14C, it is good evidence that it is not millions of years old.
However, things are not quite so simple. First, plants discriminate against carbon dioxide containing 14C. That is, they take up less than would be expected and so they test older than they really are. Furthermore, different types of plants discriminate differently. This also has to be corrected for.
Second, the ratio of 14C/12C in the atmosphere has not been constant -- for example, it was higher before the industrial era when the massive burning of fossil fuels released a lot of carbon dioxide that was depleted in 14C. This would make things which died at that time appear older in terms of carbon dating. Then there was a rise in 14CO2 with the advent of atmospheric testing of atomic bombs in the 1950s. This would make things carbon-dated from that time appear younger than their true age.
Measurement of 14C in historically dated objects (e.g., seeds in the graves of historically dated tombs) enables the level of 14C in the atmosphere at that time to be estimated, and so partial calibration of the "clock" is possible. Accordingly, carbon dating carefully applied to items from historical times can be useful. However, even with such historical calibration, archaeologists do not regard 14C dates as absolute because of frequent anomalies. They rely more on dating methods that link into historical records.
Outside the range of recorded history, calibration of the 14C "clock is not possible.
Also, the amount of cosmic rays penetrating the earth's atmosphere affects the amount of 14C produced and therefore dating the system. The amount of cosmic rays reaching the earth varies with the sun's activity, and with the earth's passage through magnetic clouds as the solar system travels around the Milky Way galaxy.
The strength of the earth's magnetic field affects the amount of cosmic rays entering the atmosphere. A stronger magnetic field deflects more cosmic rays away from the earth. Overall, the energy of the earth's magnetic field has been decreasing, so more 14C is being produced now than in the past. This will make old things look older than they really are.
In addition, the Genesis flood would have greatly upset the carbon balance. The flood buried a huge amount of carbon, which became coal, oil, etc., lowering the total 12C in the biosphere (including the atmosphere -- plants regrowing after the flood absorb CO2, which is not replaced by the decay of the buried vegetation). Total 14C is also proportionately lowered at this time, but whereas no terrestrial process generates any more 12C, 14C is continually being produced, and at a rate which does not depend on carbon levels (it comes from nitrogen). Therefore, the 14C/12C ratio in plants/animals/the atmosphere before the flood had to be lower than what it is now.
Unless this effect (which is additional to the magnetic field issue just discussed) were corrected for, carbon dating of fossils formed in the flood would give ages much older than the true ages.
Creationist researchers have suggested that dates of 35,000 - 45,000 years should be re-calibrated to the biblical date of the flood.[6] Such a re-calibration makes sense of anomalous data from carbon dating -- for example, very discordant "dates" for different parts of a frozen musk ox carcass from Alaska and an inordinately slow rate of accumulation of ground sloth dung pellets in the older layers of a cave where the layers were carbon dated.
Also, volcanoes emit much CO2 depleted in 14C. Since the flood was accompanied by much volcanism (see Noah's Flood..., How did animals get from the Ark to isolated places?, and What About Continental Drift?), fossils formed in the early post-flood period would give radiocarbon ages older than they really are.
In summary, the carbon-14 method, when corrected for the effects of the flood, can give useful results, but needs to be applied carefully. It does not give dates of millions of years and when corrected properly fits well with the biblical flood.
