Fair warning in advance: this whole post is just meant as an eloborate thought experiment. Let’s imagine that somehow for a fictional Afro-Diasporan we ARE able to trace back and identify the ethnic origins of ALL of his African born ancestors for the last 7 generations. How would that look like on paper? And what does it imply for anyone in the real world who’s faced with a missing paper trail in most cases and usually in need of tons of luck and perseverance to even identify 1 single African ancestor?
Well behold my own modest attempt at producing such a fictional family tree! The graphics might not be state of the art 😉 but seeing this fictional family tree all filled out might be useful to visualize what tracing one’s African roots really entails. Also it could help to clarify some common misconceptions if we apply it in some hypothetical scenario’s. All the blank cells represent American-born ancestors for a person born around 1960 while the coloured ones are African-born. Each generation is assumed to start in the same year and the number of ancestors is mentioned in parentheses. The African-born ancestors born around 1810 & 1780 are shown as having parents and grandparents of the same ethnical/regional background and these would naturally not have been enslaved.
***(click to enlarge)
. 
***
Judging from this family tree (completely made up and simplified on purpose), the owner of it has the following regional/ethnic breakdown going back seven generations:
- 12,5% Gambia/Mandenka (16 ancestors out of 128 living around 1750)
- 14% Ghana/Akan (18 ancestors out of 128 living around 1750)
- 24,2% Nigeria/Igbo (31 ancestors out of 128 living around 1750)
- 27,3% Congo/Bakongo (35 ancestors out of 128 living around 1750)
- 20,4% American born (26 ancestors out of 128 living around 1750)
That would be in total 98,4% of his complete ancestry. I left out his direct maternal and his direct paternal lineage (marked in blue) on purpose. They would represent the missing 1,6% (2/128) of his family tree. Let’s assume these 2 ancestors were both somehow ethnically Mende from Sierra Leone. (the actual distribution frequencies of his haplogroups L1b and E1b1a within Africa are irrelevant as i picked them completely at random 😉 ) Such a hypothetical ancestral breakdown could potentially be very misleading for this person if he were NOT to have access to the information contained in this fictional family tree but was forced, like many Afro-Diasporans nowadays, to rely on DNA testing instead.
Let’s assume he first started out with testing only his mt-DNA & Y-DNA. Provided his DNA testing company owns an extensive African database he might be told that both of his parental lineages are matching with Mende persons from Sierra Leone, which would in fact be corresponding with the truth. Based on solely this information he might then come to the conclusion that his origins are mainly or even exclusively Mende/Sierra Leonean. However in reality he would then be out of touch with about 98% of his total ancestry and fully unaware of his far more diverse ethnic/regional background as shown in the fictional family tree!
Now let’s assume his DNA testing company has a less extensive database which doesn’t contain any Mende persons or other Sierra Leoneans or even more realistically what happens if we assume that his haplogroups are not exclusive to the Mende or Sierra Leone and can also be found among other ethnicities not even per se found in neighbouring countries (as implied by this DNA study from 2006)? In this scenario he could then be told rather randomly that his mt-DNA was found to match a woman from Senegal while his Y-DNA was matching with a man from Benin. Meaning he would be completely in the dark even about only two of his 128 ancestors!
Obviously many more scenario’s might be considered. Including autosomal DNA testing (BGA admixture) which provides geographical predictions not corresponding with our fictional breakdown from the beginning because of a database that might not be fully covering all of Africa’s regional/ethnic diversity. Or random recombination of ancestral markers causing a genetic breakdown that’s disproportionate to the genealogical breakdown as shown in the fictional family tree. Or DNA cousins being reported from a database that’s biased towards only a limited amount of African ethnicities, not including all of the ones appearing in the fictional family tree. Also the DNA matches being reported might actually just be random false positives instead of IBD (Identical By Descent) merely reflicting shared DNA segments because of ancient migrations and not having any modernday ethnic origins in common etc.etc.
The bottomline is that there are many, many potential discrepancies if somehow you could magically verify any DNA test results or compare isolated genealogical findings with your fully complete family tree (going back 7 generations) incl. all African born ancestors. Even if the results/findings are correct or at worst approximate they will still usually only provide a partial or distorted overview which can possibly be misleading if you’re not aware of the complete picture. To some this might not be a major revelation but for others it could be food for thought 😉 Just to finish the excercise i’ll tease out some of the other assumptions i made when creating this imaginary and strongly simplified family tree:
- The fictional person whose ancestors are being displayed is assumed to be of 100% African descent. If he were not (which is the case for most Afro-descendants in the Americas) you would have to make a correction for non-African ancestors according to the degree of non-African admixture if you want to get a regional breakdown.
- In the family tree i mentioned mt-DNA & Y-DNA each being 1/128 of total ancestry. This is of course strictly genealogically speaking and not genetically. My intention being to show the relative contribution of the first African born ancestors carrying your haplogroups. If instead of 7 generations ago they would have arrived 8 generations ago it would be 2/256 of total ancestry just as another possibility (out of many).
- In reality of course generations might be shorter or longer than 30 years and separate family lines will not run parallel. Also I only provided this seven generational family tree as i couldn’t find any suitable template containing more generations. Plus i didn’t want to make it too complex. Many Afro-descendants might actually have a higher proportion of American-born / “Creole” ancestors dating from before 1750, which would be increasing the amount and diversity of African born ancestors who arrived in the Americas early on. See also this previous post.
- To keep it simple i also minimized the number of ethnicities. Even if the regional breakdown could be somewhat realistic, it’s very likely that for example his fictional 31 Nigerian ancestors would be of more than just 1 ethnicity. Even if they were all shipped through only one slave port somewhere on the Bight of Biafra, there could be many more ethnicities involved than just Igbo (see this map i posted earlier on the origins of African recaptives in Sierra Leone). The same could be said for his 16 ancestors exported via Gambia who might have had various Upper Guinean etnic origins instead of just simply “Mandingo”, a term which is known to have obscured much underlying ethnic diversity (see this map). Also his 18 ancestors brought over via Ghana/Gold Coast might actually have been Ewe or from northern Ghana (Gur speaking) (see this map). His 35 Congolese ancestors might have been known at that time as “Congo” a much (ab)used lumping category. However also they could have been of various ethnic background despite commonalities (see this map).
Tracing African Roots 