Racing To the Wrong Conclusion

Check out the story behind these
cool twins HERE

We've all seen that post: "I have this photo of a relative that looks like [insert race]." The poster and comments following start pointing out features. Suddenly everyone is an anthropologist! Hair texture, eye colour, cheekbones, foreheads, nose shapes..... Everyone seems to ignore a simple fact: there is no characteristic that belongs to one race. Moreover, there aren't different races genetically. Race is a human construct to divide and define, and it does a very poor job!

Let's Play a Game

I found this really fun game on PBS's website. There are 4 categories of race (White, Black, Native American, Hispanic/Latino) and you take 16 people and place them in the race category that you most think they look like. I did really well on Asians....... but not so much on anyone else (I didn't get a single White one right). After you play the game, go ahead and check out the other links, especially the timeline on race. Can you imagine a time when race wasn't a category to put people into? Well, there was a time just like that.

Of course, humans love labelling things, especially other people. The problem with generic labels is that it never really fits what you're trying to label. Most race categories we think of today were made to better define humans for a census (i.e. "Hispanic"). Caucasian really means people from the Caucasus mountain region, but has morphed to mean "White" in common use. There are three basic class distinctions for forensic anthropology: Caucasoid, Negroid, Mongoloid. Because of negative connotations for the last two, many people have automatic stereotypes as to what these classifications mean based on skin colour (which is not at all part of the consideration). And because they are so general, many people don't fit one or the other which has led to some experts dividing them even more (i.e. Mongoloid has been divided from Australoid to mark the very distinct differences in Australia's Aborigines; Negroid has been divided into Congoid and Capoid for West Africa and South Africa respectively). In reality, no matter how you divide it, humans don't have any populations that are isolated enough to create a truly different genetic makeup. I could say that a Caucasian/European/Caucasoid has a long narrow face, cheekbones that don't protrude, and a narrow nose with a high bridge; a Negroid/African has a wide nose, rounded forehead, and protruding lower portion of the face; a Mongoloid has a moon shaped face, prominent cheekbones, and "shovel shaped" teeth. I could say these things. But the problems are myriad with these generalisations.

The picture below was made by taking several photos of different women from the same country or region and making a composite average. I'll refer to this photo while we discuss the problems with generalising (note that this composite doesn't take into account mixed races).


Couldn't find who originally posted this, but found a good blog while I was at it!

First, the problem with averages. In order to make any average of anything, even the average of five numbers, there has to be some below the average and some above. Go ahead and find the average African American woman in the photo above. Pretty average, even reminds me of a few actresses. But to make that photo, someone had to have a larger face, smaller nose, larger eyes and have those corrected to the mathematical average. Second, do you see any Native Americans? According to the 3 categories discussed above, Native Americans fall into the Mongoloid group. The stereotype we've seen on television would make me lean towards the Samoan more than the Peruvian. But what about the Indian, Mexican, Puerto Rican or Burmese photos? I have full blood friends and kin I'd place closer to those faces. Which brings us to problem number three: skin color. I'm sure you scanned the faces for darker skin for the Native Americans. If I asked you to pick out the Whites, you'd look for the light skin.... what if I asked you to pick out the Caucasoids? Go ahead and try that: look for narrow noses, cheekbones that aren't prominent, long faces...did you pick out the Ethiopian? No? You should do. How about the Afghan and Iranian? And is the South Indian a Caucasoid and the Indian a Mongoloid? Problem number four: the characteristics that anthropologists use to identify the general categories are best applied to a skull. How can I tell whether the Filipino has the more oval eye sockets of the Mongoloid or the round ones of the Caucasoid? Does the picture of the Central African show the square sockets of the Negroid? Is she a Congoid or Capoid?

Special Fun With the Census!

The US census currently divides humans into the races of White, Black/African American, American Indian/Alaska Native, Asian, and Native Hawaiian/Other Pacific Islander. It wasn't always that way. From 1790-1810 asked for Whites and Slaves. 1830 asked for White, Black, or Mulatto. The 1870 census was the first time East Asians were listed (all as Chinese) as well as American Indians. 1890 got crazy with it: White, Black, Quadroon (1/4 African ancestry), Octoroon (1/8 African ancestry), Chinese, Japanese and Indian. 1910 added an "other race" category for anyone that didn't fit anywhere else to write in their affiliation. The 1920 census finally recognised that Koreans, Filipinos and Hindus (Indians) were separate from Chinese or Japanese people but still Asians.

1930 saw a few wild changes. They removed Mulatto and decided that anyone with any African ancestry or mixed Black/Native Americans with no official tribal affiliation were to be listed along with full blood Blacks as "Negro". A mixed White/Native American was to be Indian if they had a tribal affiliation, White if they were accepted in White society. "Mexican" became a category if a person or their parents were born in Mexico. In 1940, Mexicans went back to being categorised as White. In 1950, Hindu and Korean were removed. 1960, Indian was changed to American Indian and Hawaiian, part-Hawaiian, Aleut, and Eskimo were added. The "other" category was also removed. 1970 changed "Negro" to "Negro or Black" and added Korean and "other" back in. 1980 added in Indian, (East) Guamanian, Vietnamese and Samoan.

Add into this the recent confusion of adding "Hispanic" in 2000. The Census Bureau defines "Hispanic or Latino" as "a person of Cuban, Mexican, Puerto Rican, South or Central American or other Spanish culture or origin regardless of race." So on the census, every person answers whether they have Hispanic ancestry or not regardless of what category (White, Black, Asian, Indian) they would place their race. So the woman in the above photo from Mexico would be Hispanic and White or American Indian depending on her family history and if she has tribal affiliation. With that said, I know a man from Spain who could say he's Hispanic by that definition (having a Spanish culture), but mistakenly regards it as a race of peoples from Central and South America that are Native Americans mixed with other races. And what about my niece? Her father is from Puerto Rico, her mother Indiana...... Does her father's influence give her enough cultural knowledge to be Hispanic? Or would she say she isn't Hispanic as she is growing up in a more Midwest American culture? Well, there's talk of allowing people to check both "Hispanic" and "Not Hispanic" on a census....... then what's the point of asking the question??? Which brings me too....

General Genetic Confusion

 

From the Human Race Machine this is one woman's face in 6 skin shades

 If I were to say someone had thick lips, hooded eyes (the fold of the upper lid hides the lower lid), a broad nose, and brown eyes... what do you think their skin color is? The characteristics I listed are all dominant genes. When your parents made you, dad gave you one half, mom the other. So, say, dad gave you the gene for thick lips and mom thin lips..... you'll have thick lips. If you marry someone who has thin lips, there is a possibility that your child will have thin lips as you carry the gene for it, but if your child gets the thick lip gene then they'll have thick lips. Skin and eye color aren't dominant/recessive. Both are physical manifestations of the amount of melanin in our system (as is hair color!). They are actually controlled by more than one gene pair (skin has 7), so they can be in a range of shades. Brown eyes are considered dominant as originally all humans had brown eyes. It only takes one gene (OCA2) to be mutated to shut off a portion of melanin production to make blue eyes. (For a cool relationship fact: Blue eyed folk may all have the same ancestor!) In point of fact, dark hair, skin and eyes are all dominant. Since there are several genes involved, any one or more mutation can lead to a change in the color. As the photo of the twins at the top shows, skin doesn't equal ethnicity/race.

Now, my good friend Loretta's blog "Barking Up the Wrong Tree" covered some interesting responses to Ancestry.com's posts regarding it's new autosomal DNA test. What it proves is that many people don't understand what genes are about. I'll have to make a post (or ten) just on genealogy and genetics, but today I'm going to give you a rough outline to explain ethnicity markers and why your identified race/ethnicity isn't what you'll always see in your DNA. So a bit of the basics on DNA: your DNA is made up of 46 chromosomes in 23 pairs. One pair is well known as it determines your sex: you'll be a female if you are XX, male if you are XY. The rest of the 22 pairs determines most everything else about you (there are some specific mutations found on the sex chromosomes for some characteristics, but again, this is a broad stroke today).

Chromosomes are made up of genes. Genes are strings of molecules that are basically instructions for our cells on how to do whatever it is they do. An allele is what we call different forms of genes. For example, the gene for number of fingers has two alleles: the dominant 6-fingered option and the recessive 5-fingered one. If I wanted to trace the history of the recessive mutation, I would look for the earliest concentrated occurrence of that allele. Since there is no gene shouting "I AM WEST AFRICAN!", scientists need to look for alleles that are commonly found in West Africans. If an allele is almost exclusive to, say, Indonesians, then that mutation can be used to map migration of the Indonesians to neighboring islands. I just read a great book "Guns, Germs and Steel" which, while a little repetitive on the conclusions, explains how genetics, language, food production, etc. can be traced from a concentrated "start point" to other areas by conquering civilisations. If you want a really detailed explanation, though, I would recommend the read.



Please read Ancestry.com's blog post HERE for their explanation



So about DNA tests: When you take a Y-DNA test, you must be a male. Women don't have Y chromosomes. Since women don't have Y Chromosomes, a Y-DNA test will tell a man about his father, grandfather, great grandfather, etc......in a straight line of descendancy. The X Chromosome is found in both men and women, so to do a test exclusive to women, scientists came up with mt-DNA testing. That test is for the DNA in your mitochondria. Mitochondria is a small organelle found in cells. It has it's own DNA with it's own mutations. Women pass mitochondria down to both their female and male children, so a man or a woman can take the test. The test will tell you about your mom, grandmother, great grandmother...... both these tests give you a very clear idea of your family migration as far as the mutations to genes on the Y chromosome and mitochondrial DNA.... but ethnicity isn't among them. To test for ethnicity, you must take an autosomal test. That tests the other 22 pairs of chromosomes. It looks for specific alleles that are found in specific areas and have clear migratory patterns to prove a point of origin. It's less accurate than the other tests and does NOT tell you whether it's your  mother's or father's side that gives you a particular allele. And there isn't a time on these mutations, so your tree may go back 7 generations, but the DNA is going back 10 or 20..... so just because your grandmother was Spanish, doesn't mean your DNA won't say you're Turkish. None of the alleles that have been identified to be used in an autosomal test are for a physical appearance trait. You can "look like" an Italian, and share the alleles associated with a South Indian. And since your parents' DNA doesn't divide evenly to give you their half, you may lose an allele that is specific to one group or another. So your 4th great grandfather may in fact be Native American, but it's so far back in the DNA and has been recombined and divided so many times that you don't have a marker that says you're Native American.

It's all about human migration. People conquering the inhabitants of neighboring areas, being conquered, travelling and intermarrying for millenia..... that's why German ancestry is so hard to pin down and West African is not! Germany is the cross roads of a large land mass called Eurasia (the largest one Earth). Romans tried and failed to conquer it; Germans tried and succeeded in conquering most of the British Isles; Scandinavians tried and succeeded several conquests of most of the areas of Europe including the British Isles and Germany. West Africa was able to conquer most of South Africa, leaving only pockets of native populations there. Most African Americans will find West African links, because most slaves were collected from the West African coast. West Africans took the mutations specific to their area and bred them into their conquests. We can see a clear migratory pattern. Germany.... not so much. As more people sign up for the autosomal test, we'll get more data and a clearer picture of what mutation started where. Even still, it is not an accurate system. It's the genetic equivalent of "looks like".

To Confuse Us Further

Most people having problems understanding their DNA results, or asking what a relative "looks like", are actually just making one big problem: defining words incorrectly. Here's what Dictionary.com says about nationality, race and ethnicity.

• Nationality- the status of belonging to a particular nation, whether by birth or naturalization: the nationality of an immigrant.
• Race- a group of persons related by common descent or heredity..... an arbitrary classification of modern humans, sometimes, especially formerly, based on any or a combination of various physical characteristics, as skin color, facial form, or eye shape, and now frequently based on such genetic markers as blood groups....a human population partially isolated reproductively from other populations, whose members share a greater degree of physical and genetic similarity with one another than with other humans. (This is what is meant by the Caucasoid, Negroid, Mongoloid classifications by anthropologists)
• Ethnicity- relating to or characteristic of a human group having racial, religious, linguistic, and certain other traits in common.

So when we ask what someone looks like, or do the autosomal test, most of us are expecting an answer of Nationality. Grandpa was from Norway, so we're hoping to hear Norwegian. When someone says that their relatives looks like a Native American, they are using an arbitrary race stereotype (dark hair, high cheekbones, dark skin). When someone finds out that their relative who was documented as from England carries Scandinavian DNA, they call the test bunk. They don't look Scandinavian! Well, what does a Scandinavian look like? The alleles are specific to an ethnicity and they're confusing it with race (both England and Scandinavia are considered to average as Caucasoid) or just nationality.

Show a person a picture of a baby and an unrelated man and tell them that the man is the child's father, and most will start to see "obvious" physical traits to prove it. (I read this in a psychology magazine, but couldn't find it again to give you the exact percentage. I know it was somewhere around 80%) If you are looking for proof of a particular ethnicity, you'll probably see it, even if it isn't there. I'm not saying everyone is wrong when they think their relative looks like a particular group.... I'm saying that unless you're an expert, you're most likely wrong. I did a two part article on how to determine the age of a photo based on photograph technique and the clothing of the subject. Even at my best, I can get a range rather than an exact date. If I tried to determine what someone "looked like", I wouldn't be any more accurate. Even an expert would be uncomfortable with an absolute ethnicity. They'd say that feature suggested this and this one suggested that.... but no, they won't look at your photo of great aunt Ruth and tell you she definitely had slave ancestors and just "passed for white".

I'd rather decide if one photo of my great grandfather "looks like" the unlabelled earlier photo and leave it at that.

-Ana