Let me bring people up to speed: In his movie, Al Gore shows a chart spanning hundreds of thousands of years, on which is plotted temperatures and atmospheric CO2 concentrations. The two lines move in lockstep. Gore says something like, "Now some people will deny that there is any connection here, but I'm not so sure about that...heh heh," and the audience laughs with him at all those evil deniers.
Well, it turns out that if you blow up the scale on the timeline, then you can easily see that the temperature increases lead the CO2 increases, by up to 800 years. So that's a bit odd, if your current explanation for global warming in the 20th century is that humans have released a bunch of CO2.
The anti-skeptic, professional climatologist site, RealClimate, has addressed this very issue. They explain that other things (such as changes in the earth's orbit) triggered an initial rise in temperatures, which then initiated a CO2 feedback cycle (i.e. more CO2, more warming, more CO2, etc.).
What's interesting is that in the letter exchange at the end of the blog post, it comes up that there are lags in the coolings as well. To repeat: The temperature goes up first, then hundreds of years later CO2 goes up. Then the temperature drops, and up to thousands of years later the CO2 goes down.
My question: Are there any warming/cooling periods without such lags? If not, does the consensus climatologist view require that large moves in CO2 have only been due to temperature variations initially set in motion by other forces?
(It seems to me they would have to assume that, for if some non-temperature-changing event caused CO2 to change, then this should have been at least one episode without a lag. E.g. if a CO2-rich asteroid smacked into the earth and increased concentrations in the air, then that should have been one episode where CO2-increases led global temperature increases.)