Back in March, I started a DNA experiment with five friends. I invited each of them to take a genetic ‘ancestry’ test – something that most of them had barely heard of, let alone thought about doing. After sending off their swab or spit sample, I asked each person to describe their expectations regarding their results: What did they expect or hope to find out? What would their ideal result be, based on their own family history, physical appearance and cultural background?
I also participated in the experiment, although in the interests of scientific curiosity, I took a test with five different US-based companies specializing in ‘genealogical’ or ‘anthropological’ DNA tests: 23andMe, AncestryDNA, Family Tree DNA, the Genographic Project, and African Ancestry. This mini-experiment is inspired by my PhD research on the way genetic data interact with local notions of identity, ethnicity, kinship and nation in the US and Brazil.
Sarah, born and raised in the East Midlands, England
The first test results I received were from 23andMe, a California-based company whose tests are tailored to a mainly American customer base. Being English, fair skinned and with no family stories of anyone living outside of Britain for the past few generations, I was hardly expecting any African or East Asian ancestry to show up on the test – although when I told this to an African American colleague she raised her eyebrows and told me, wryly, that black Americans have long known that fair skin is no guarantee of anything. Nevertheless, my expectations were confirmed by my results, and the only spot of colour on my blue-grey ancestryscape was a green dot indicating 0.6% Ashkenazi Jewish genetic heritage – possibly statistical ‘noise’, or possibly a relic of our lost Polish family connection. There’s no way of knowing.
The most striking thing about my 23andMe results was that my ‘British & Irish’ genetic ancestry estimate was only 6.1%; the rest of my result was made up of ‘Broadly Northern European’ and ‘Broadly European’ heritage. Rather than bringing on an identity crisis about my British identity, however, this merely indicated to me that there has been so much migration within Europe (not to mention with the rest of the world) that it is near impossible to identify genetic markers that are limited to specific regional or national populations.
The second set of results came from the Genographic Project. Linked to National Geographic, the Geno 2.0 test has an ‘anthropological’ focus and a humanitarian feel. The product has its roots in an international scientific endeavour, which aims to to collect genetic samples from contemporary indigenous groups all over the world, in order to trace the ancient migrations of humankind out of Africa, starting some 60,000 years ago. To spare you a long-winded explanation, I’ll just recommend you this documentary (followed by this excellent paper by Kim TallBear as complementary reading). Noticing the general public enthusiasm for all things genetics-related, National Geographic savvily set up an online company, making their tests available to anyone with $160 to spare.
Unlike 23andMe (and most other companies), which attempt to split an individual’s genome into something that looks more or less like a blood quantum chart composed of two or three founder populations (e.g. Europeans, Africans and indigenous Americans for New World customers), the Genographic Project assumes that all national ‘ethnic’ identities are based on older genetic mixes of substrate populations. So although I don’t recognize anything Mediterranean or Southwest Asian in my own family (and nor would most Brits, I think), this turns out to be a very common mixture for the UK, according to how the Genographic scientists visualize the islands’ ancient genetic makeup. Substitute ‘Mediterranean’ for ‘Roman’; ‘Southwest Asian’ for ‘Celtic’; and ‘Northern European’ for ‘Saxons’, ‘Vikings’ and ‘Normans’, and with a stretch of the imagination you get something vaguely representing the country’s history.
I admire the Genographic Project’s universalist approach (in spite of their conspicuously Euro-American-centric outlook), although I’ll confess that I found these results the most difficult to connect with on a personal level, for the very reason that they don’t reproduce my own sense of identity or ancestry in the way I would expect them to. This is one conundrum of genetic testing: on the one hand, if the company gives us a chart that looks recognisably like a blood quantum graph, we accuse the scientists of reinscribing race in genetic technologies. Yet, when the scientists try to introduce a new paradigm (even when it is only slightly removed from the blood quantum model), customers remain unsatisfied because don’t see anything that relates to our notions of ‘racial’ or ethnic identity in the results.
Next came my results from AncestryDNA and Family Tree DNA – the newest and oldest competitors in the DNA ancestry testing market, respectively. Both boldly proclaim me to be ‘100% European’, with genetic heritage from four broadly defined regions within the continent. I find it striking that both companies choose the word ‘ethnicity’ to describe the concept they are selling: a statistical means of breaking down a broad ‘racial’ or geopolitical category like ‘white’/European into distinct biogeographical regions. None of the areas highlighted on either map could be considered to correspond to ethnicities or ethnic groups by any currently recognised social or political boundaries, although there is some congruence with familiar European ‘types’: the Scandinavian; the Mediterranean; the Eastern European, and the Brit.
These ancestral identities would be recognisable for many Americans who may not have concrete knowledge of the exact national origin of their European great-great-grandparents, beyond a vague sense of these broad regional categories. For European customers, on the other hand, my sense is that these ‘ethnic’ regions do not hold much water. For instance, if I know myself to be British (with Polish and French ancestors going further back in time), then what can it mean to be told that I am also genetically Scandinavian and Mediterranean, against all common sense and family tradition? Being British is not something that I feel in a genetic or hereditary sense – but rather an identity that revolves around a shared language, regional accents, a niche sense of humour, a mutual fascination with Marmite, and so on… In comparison, this genetic information is something I know to be true on an abstract level – but I don’t feel it, and it’s definitely not news to me.
The final test was by African Ancestry. As you might guess from the name, the company was set up in 2003, offering DNA tests to African Americans as a way to recover the tribal identities of their ancestors who were forcibly displaced by the transatlantic trade in enslaved Africans. African Ancestry’s main products are DNA tests that analyse only one maternal or paternal genetic lineage (rather than your autosomes, which contain genetic material inherited from all of your direct ancestors); these are then associated with one or more ethnic groups or tribes in Africa, and customers are presented with a certificate showing their ancestral tribal identity. Since I already knew my maternal lineage was mainly found in Europe (and therefore suspected I wouldn’t be eligible for a certificate), I opted for the slightly cheaper, less popular myDNAmix test.
Of all the different companies I tested with, these results are the most rudimentary in appearance. The only overtly Afrocentric company out of the five, African Ancestry expresses its results using five broad ‘genetic’ regions: ‘Indigenous Americas’, ‘East Asia’, ‘Sub-Saharan Africa’, ‘Europe’ and ‘India Subcontinent’. In the form I filled in to return with my DNA sample, I was asked to estimate my expected genomic ancestry percentages – a standard practice that nonetheless makes me slightly suspicious of whether this might influence the results given out to customers, although there is no reason to think that this should be the case.
As it turned out, these were the most surprising results of all: 73% ‘Europe’ and 27% ‘India Subcontinent’ (give or take an error margin of 7%). There is no mention in my family of anyone ever even going to India, let alone living there or having Indian ancestors. Given the stark difference between these results and those of the other companies I am forced to conclude, on the basis of sheer good sense, that there has been an error somewhere down the line – although whether in terms of a sample mix-up or a category mistake, it is hard to know.
Yet this brings up the problem of the unknowability of genetic ‘truths’. Say, for the sake of argument, that there is no error, and 27% of markers analysed in my DNA sample really did match most closely to samples in the African Ancestry database originating in the Indian Subcontinent. That may be perfectly true, according to the criteria of this genetic test, and these sampling, labelling and statistical standards. But it is so untrue in terms of what I know about myself and my family (admittedly, not much past three generations), that I am forced to reject it, and to seek solace in one of the other tests that reinforced my preexisting sense of who I am. And, ironically, this brings me right back to where I was before I even took my first DNA test.
You can read my original post here.
Thoughts, questions or constructive criticism about this post? Help turn this monologue into a discussion by leaving a comment (use the speech bubble button next to the title of this post, or write one in the comments box below).
If you have a similar or related experience you would like to share or talk about, or if you are interested in writing a guest post for this blog, please get in touch using the Contact tab at the top of the page.