>>9748582Racialists argue that while overall genetic variation between continental population divisions is <10%, this rather little human population structure still supports race classification (contra Lewontin, 1972[48]).
They quote Edwards (2003) who found by looking at how gene loci are correlated: "probability of misclassification falls off as the number of gene loci increases". It is notable Edwards in his paper does not dispute Lewontin's statistical data on various blood polymorphisms, writing: "There is nothing wrong with Lewontin's statistical analysis of variation, only with the belief that it is relevant to classification". As a result, as Marks (2010) explains, there is no "Lewontin fallacy":[119]
“”Geographical correlations are far weaker hypotheses than genetically discrete races, and they obviously exist in the human species (whether studied somatically or genetically).
What Lewontin (or Marks) and Edwards are discussing are two completely different things. Of course genetic correlations exist which can pinpoint someone's geographical ancestry, but as Marks asks: "What is unclear is what this has to do with race", and concludes: "Lewontin's analysis shows that such groups [races] do not exist in the human species, and Edwards's critique does not contradict that interpretation" (emphasis added). What this means is that Edwards is re-defining the race concept to a far weaker hypothesis, which is not how race is commonly understood in biological taxonomy: Fujimura 2014 writes:[114]
“”
f one wants to measure the meaningfulness of differences between groups with different geographic ancestries, one needs to use a “proportion of variation” approach. This approach compares genetic variation among individuals within groups to the genetic variation between these groups. Ironically, Edwards’s (2003) reported findings confirm Lewontin’s (1972). 
