I've been looking around for answers to fey's question. One relevant paper would seem to be Francisco J. Ayala, Ananias Escalante, Colm O'hUigin and Jan Klein, "Molecular Genetics of Speciation and Human Origins," Proc. Natl. Acad. Sci, USA, 91: pp6787-6794, July 1994, p. 6787-6788. Addresssing the question of whether there was a bottleneck 100,000-200,000 years ago, they calculate from the present level of heterozygosity that there can't have been fewer than 4000 breeding individuals at that time.
Now the effective population would have to be larger the closer in time the bottleneck was, so if we put it at a mere ~4000 years ago, it would have to be so large it would hardly be a bottleneck at all ... certainly not as small as 8 people.