Radiometric dating is not the only technique available to establish that the earth is much more than 6,000 years old.  Anyone trying to build a history of the earth from the chronology given in Genesis will have a lot of trouble with any of these, let alone all of them. The correlations between these methods and the radiometric methods is much too high to be dismissed as coincidental .

1) DENDROCHRONOLOGY. This is dating from counting tree rings and correlating the patterns of wide and narrow rings between trees, thus working back farther than the age of a single individual tree. This methodology does not gain an order of magnitude, but patterns of tree rings have been traced back to ages that are at least too old for the Ussher dates. For example, Becker et al., Nature, 353:647-649, 1991 discusses a sequence of oak trees from Europe that yields a CONTINUOUS chronology from the present back 9,928 (by now 9930) years. The same authors also have a 1600 year old pine chronology which overlaps the end of the oak chronology. As interspecies comparison is difficult they have not yet fixed an absolute age for the pine chronology. If they are able to we will probably have a chronology going back 11,500 years. The oldest living trees are bristlcone pine trees from the White Mountains of California/Nevada. The oldest known live ones are between 4500 and 5000 years old and dead trees have been matched with live ones, extending the scale of bristlecone pine tree rings of known age back 7000-9000 years (there seems to be some discrepancy in the posted sources, but all of the figures are certainly more than 6000 years) before the present. Note, too, that even the ages of the LIVING trees extend back beyond the usual dates cited (and calculated from Biblical references) for the purported flood. Another study, finished in 1984 in Ireland, established an unbroken record of tree rings in oak trees from the present [i.e., 1984] to 7,272 years ago. Creationists sometimes object that some trees have been known to form multiple rings in a single year, but this can be and is controlled for.

2) ICE CORES There is definitely a FAQ on this one. Matt Brinkman has compiled a beautiful summary of the data. Briefly, though, people have counted annual layers of ice in places like Greenland and Antarctica. Naturally, the layers become harder to distinguish as you go deeper in the ice, but the principles are comparatively simple. It is comparatively easy to show (e.g., using inclusions of ash from known volcanic eruptions) that the layers do correspond to years rather than snowfalls (as Ted Holden has often asserted). As the ice gets compacted, the layers do become hard to distinguish and there are various theoretical methods used to test them, so someone who is inclined to disbelieve in them may feel uncomfortable accepting figures from beyond the point where people can actually *see* them reliably. However, even within this more restricted range, the news is bad for a "young earth." For example, in the Greenland ice divide (summit) core, it is possible to distinguish 40,000 annual layers before they grow too thin. Note that in those 40,000 years, there is no sign of a flood. Deeper in the core we get to ice laid down in the last interglacial (circa 120,000 years ago). According to both observations and theory (Milankovitch) this was a time in which the northern hemisphere was much warmer than today. Sure enough, at this point the annual layers reappear. Precipitation was much higher at this time (a consequence of the greater capacity of air to hold water vapour as temperature increases) and the annual layers are still 6mm thick despite the thinning induced by the flow of ice. Roughly another ten thousand layers can be counted. It is possible, though not yet confirmed, that more sophisticated isotopic analysis may allow annual layers to be counted between these two zones, perhaps back to around 70,000 years before present. This will be quite difficult because these layers have been considerably thinned by ice flow, and unlike the last interglacial layers, they were never that thick to begin with. Very preliminary results on this core were published in Nature, volume 359, page 311.

3) VARVES. Varves are annual layers of sediment laid down on lake bottoms. Depending on the climate and environment, you may get different numbers of layers per year, but in any case they cycle as 2, 3, or 4 distinct types of sediment and then repeat the same cycle again. In the Green River formation of what is now Wyoming, there are places with 20,000,000 thin varves, each varve consisting of a thin layer of fine light sediment and an even thinner layer of finer dark sediment. According to the conventional geologic interpretation, the layers are sediments laid down in a complex of ancient freshwater lakes. The coarser light sediments were laid down during the summer, when streams poured run-off water into the lake. The fine dark sediments were laid down in the winter when there was less run-off. (This process can be observed in modern freshwater lakes.) If this interpretation is correct, the varves of the Green River formation must have formed over a period of about twenty million years. Young-earth creationists insist that the earth is not much more than 10,000 years old and that the geologic strata were laid down by the Flood. Whitcomb and Morris therefore attempt to attribute the Green River varves to "a complex of shallow turbidity currents..." (p. 427). Turbidity currents--flows of mud-laden water--generally occur in the ocean, resulting from underwater landslides. If the Green River shales were laid down during the Flood, there must have been forty million turbidity currents, alternatively light and dark, over about three hundred days. A simple calculation (which Creationists have avoided for twenty years) shows that the layers must have formed at the rate of about three layers every two seconds. A sequence of forty million turbidity currents covering tens of thousands of square miles every two-thirds of a second seems ridiculously unlikely.

4) CORAL CLOCKS. Short explanation: the moon is slowly sapping the earth's rotational energy. The earth should have rotated more quickly in the distant past, meaning that a day would have been less than 24 hours, and there would have been more days per year. There is an exceedingly strong correlation between the "supposed age" of a wide range of fossils (corals, stromatolites, and a few others -- collected from geologic formations throughout the column and from locations all over the world) and the number of days per year that their growth pattern shows. Modern corals deposit a single, very thin layer of lime once a day. It is possible with some difficulty, to count these diurnal (day-night) growth lines and to determine how old the coral is in days. More important seasonal fluctuations will cause the growth lines to change their spacing yearly so that annual increments can also be recognized much as in growth rings of trees. Professor John Wells of Cornell University began looking for diurnal lines of fossil corals. He found several Devonian and Pennsylvanian corals that do show both annual and daily growth patterns. He found that the Pennsylvanian forms had an average of 387 daily growth lines per year cycles, and that the Devonian corals had about 400 growth lines. The agreement between these clocks, and radiometric dating, and the theory of superposition is a little hard to explain away as the result of a number of unlucky coincidences in a 300-day-long flood.

[The above excerpts, somewhat edited, were posted to talk.origins over the past several years by a number of people. Chris Stassen, Matt Brinkman, Bill Hyde, Karl Kluge and Bill Jefferys are certainly among them, though.]