The debate last February between creationist Ken Ham and science educator Bill Nye has been widely discussed (see also the video). Both sides were rather an embarrassment, but one interesting aspect was Ham’s distinction between “observational science” and “historical science.” This has been called an “inane and baseless fallacy” – but is it?
In fact, all watchers of the CSI franchise know that there is a clear distinction between (on the one hand) applying known science to the past – in order to decide who did what – and (on the other hand) developing new knowledge of scientific principles. There is, of course, an interplay between the two. For example, forensic entomology draws on experimental work in a specific aspect of insect ecology. Experimental work in ballistics (popularised by the MythBusters) is used to decide what conclusions can be drawn from bullets and bullet wounds.
Observational science tends to be restricted to the here-and-now, where confounding factors can be dealt with. NASA and ESA justifiably spend a lot of money sending probes around the solar system (e.g. the probe above) so that the reach of observational science can be extended to objects which humans cannot visit. Events which are outside the solar system, or are distant in time, are outside the scope of direct observation altogether, which means that some degree of inference is inevitable.
Of course, this does not mean that scientists throw up their hands in despair, and say “we’ll never know.” Astronomers routinely investigate the same phenomenon at multiple wavelengths (e.g. radio waves and visible light), in order to get a clearer picture of what’s been going on. The supernova of last April (see image below) is one example, having been investigated at gamma-ray and optical wavelengths.
Carbon dating involves several assumptions about the past – but from the very beginning those assumptions were cross-checked using other dating techniques, such as tree rings and historical methods (the diagram below is redrawn from the Arnold & Libby paper of 1949). In practice, carbon dating is adjusted for multiple confounding factors, and provides a moderately accurate dating method for carbon-containing objects with ages up to tens of thousands of years.
In summary, then, a distinction can indeed be drawn between “observational science” and “historical science.” The latter draws on the scientific principles established by the former. Scientists tackle the problem of not being able to directly observe the past by using multiple independent methods to infer what happened, and this can allow very solid conclusions to be drawn. That’s precisely what makes books, films, and television shows about forensic science so compelling.
Update: see also this 2008 post from the National Center for Science Education on the topic.