And geneticists are using DNA to uncover relationships in populations all the time. Jobling’s colleague, Turi King, profiled the Y chromosomes of 150 men with random surnames and compared them with 150 men who shared surnames. Unexpectedly, she found that sharing a surname means you are highly likely also to share a Y chromosome.
— Alok Jha, The Adam and Eve of genetics, Salon.com Technology, 2005/4/29
Okay, what am I missing here? Isn’t this, well, obvious? Maybe with extremely common surnames (Smith, Garcia, or Chang/Zheng) the expectation is that there is virtually no relationship at all among those so named, because the names independently developed multiple times. But surely it’s statistically likely that Juan Garcia is more closely related to Tomas Garcia than to, say, Leon Martinez? And in the case of less common surnames, passed down patrilineally, in almost all cases with the actual Y chromosome, wouldn’t we really expect to find a high concordance of common Y chromosomes?