I wanted to write a post about forensic anthropology, identification techniques, and the legacy of scientific racism as soon as I saw the article about scientific concepts that need to be abandoned (Link here). This post has been compressed in the interests of space and readability and is not an exhaustive discussion of the topic

Picture credit: Wikimedia Commons

The history of race in science

So what is a "race" of people? Essentially, the idea was that humans exist as a variety of fixed types that all share the same suite of physical traits and are different from the other types. The classification of human variation into different types has been around for quite a long time. Most discussions of this topic start with Carolus Linneaus, but the idea predates him. Linneaus, as most of you know, is the father of modern taxonomy and is responsible for the adoption of binomial nomenclature. In addition to classifying plants and animals, Linneaus came up with classifications for humans. In Systema Naturae Linneaus identified five varieties of humans: Americanus, Europeanus, Asiaticus, Africanus, and Monstrosus. These varieties were based not only on perceived morphological differences, but also on behavior and culture. Africans were unsurprisingly described as " black, phlegmatic, relaxed; black, frizzled hair; silky skin, flat nose, tumid lips; females without shame; mammary glands give milk abundantly; crafty, sly, careless; anoints himself with grease; & regulated by will." While no explicit hierarchy was described for these types, the descriptions made it quite clear that Europeans were supposed to be the best.


This racial view of human variation was widely accepted in the sciences. Anthropologists, unfortunately, also held on to these ideas for quite a long time. The father of physical anthropology, Ales Hrdlicka, measured skeletal material from all over the globe and used these data to reinforce ideas about the validity of race and the supremacy of whites. Anthropometric measurements were important to the Nazi Party in their quest for "racial purity", as the science of the time told them that they could detect Jewish people by measuring their heads and faces (link).

In the 1960s most anthropologists shifted away from a typological view of human variation and moved to a clinal view. Frank Livingstone famously said "There are no races, there are only clines" (On the non-existence of human races). Clinal variation basically refers to gradual change in traits across geographic distance. Skin color is the best known example, especially since skin color makes up so much of how we perceive race. Skin color is darker closer to the equator due to selective pressure to protect the body from ultraviolet radiation (not to stop cancer which affects old people, but to prevent the destruction of folate which fetuses need to live).

Skin color gradually lightens the further north or south of the equator you go. Altitude also has a moderating effect on skin color, as higher elevations also have a higher exposure to UV radiation. There are no natural breaks in skin color across the globe, despite the fact that it's easy enough to see the difference in skin color extremes. Any line drawn would be arbitrary. Many human traits are continuous in this way.


Need more proof? Further evidence that race is not a biological reality came when large scale studies of genetic diversity were done. All humans on the planet share somewhere around 99.99% of their DNA. Within the genetic differences between humans it was found that the genetic diversity across the globe doesn't match a racial view of human variation. As the diagram here illustrates, there is more genetic diversity within the continent of Africa than anywhere else in the world. All other populations are subsets of this variation. Another way to read that diagram is that a random European and a random Asian individual likely are more genetically similar than any two random African individuals.

How people define race is primarily culturally determined. It's also extremely fluid, changing even within a person's lifetime. Imagine waking up one day and suddenly being considered a different race. Is African-American the same thing as Black? Is Black the same thing as colored? Are Hispanics colored? If they're not are they still Hispanic? In this country there was a time when Italians, Irish, and other European groups were not considered white (see the lead image). This is a nice discussion of the topic. You can quickly see that humans simply do not fit into the neat categories we've made for them and that these categories are constantly changing.

There are some discontinuous traits, which can likely be explained by integrating a populational model of human variation, where there is more mating within a group due to cultural norms, geographic distance, or a combination thereof causing a higher frequency of heritable traits within that population than in other populations. Some of these traits will be important later in this article.

How Race Becomes Biology

I hear some people already saying, "but what about differential health outcomes across races?" Well, that's a good question. It's well known that there is a large gap in health outcomes for people of different "races." These differences have often been thought by biomedical researchers to represent genetic differences between distinct biological races. The reality is much more complicated and interesting. The science is too long to get into here, but I recommend anyone interested please read this article here. The short version is that racism directly results in poor health outcomes due to a variety of factors including epigenetic effects.

The toxic effects of exposure to racism in one's own lifetime include a higher risk of hypertension, diabetes, stroke, and other conditions. These conditions, in turn, affect the health of the next generation, because they alter the quality of the fetal and early postnatal environment. The immediate consequence of this intergenerational effect is a higher risk of adverse birth outcomes, but there is also a lingering effect into adulthood, as adult chronic diseases like heart disease and diabetes can be traced in part to prenatal and early life conditions. Thus, the cycle begins again. (Gravlee 2009:52)

Further proof that these differential health outcomes are a result of lived experience when it was found that there is a stark health advantage to children born in Africa compared to their peers in America. What this means is that on average, Black children born in America are less healthy at birth than Black children born in Africa. According to one study, this pattern in replicated in direct mother-to-daughter measured outcomes. The cultural construct of race is being written onto people's biological selves.

What does this have to do with forensics?

I'm guessing at least some of you have watched Bones. Let's ignore the scientific inaccuracies in the show and just deal with something the show has exposed you to, the biological profile. The biological profile consists of sex, stature, age at death, and ancestry. The goal of the biological profile is to limit the list of all possible individuals a set of unidentified skeletal remains could belong to down to a manageable number. Forensic anthropologists have been using the word ancestry to avoid the negative connotations surrounding the word race, but the actual goal is to estimate what an individual's perceived race would have been. Forensic anthropologists are looking at the small amount of discontinuous traits that exist and imperfectly mapping this variation onto the human cultural construction called race.


So how do forensic anthropologists determine the ancestry of skeletal remains and how does this translate into race? The first question is the easiest to answer. If you watch Bones, it appears as if you just glance at a skull and know immediately. Thankfully there's more science behind it than that. The primary method is what's called craniometric analysis. Using a set of calipers the linear distance between anatomical landmarks is collected and compared to other skulls. We know that cranial traits, especially of the face, are highly heritable (so less effected by the environment) and are somewhat discontinuous across global populations. That is, there is some amount of overlap, but also a statistically significant amount of difference. Populations that share a close genetic ancestry will have craniometric traits more similar to each other than to other populations.

Credit: Smithsonian Institution

The way the comparison is done these days is using a program called FORDISC. FORDISC compares the measurements of an unknown skull to measurements from reference samples using a statistical technique called discriminant function analysis. Without going into the details this allows you to visualize multivariate space and compare your unidentified skull to reference groups. It will also indicate how many of the reference sample skulls were classified correctly, giving you an estimate of how accurate your classification is. The output of the analysis will tell you how likely your skull belongs to one of the reference sample groups (i.e. which group's centroid it's closest to) and how typical, or similar, it is to the other skulls in that group.


How these reference samples are constructed is an issue that forensic anthropologists grapple with and are attempting to improve. This issue leads us to our second question, how do me map "ancestry" to social race? The most commonly used samples in FORDISC come from the Forensic Anthropology Data Bank (FDB). These samples come from casework and donated bodies. In FORDISC the samples are American Blacks, American Whites, American Indians, Chinese males, Guatemalan males, Hispanics, Japanese, and Vietnamese males (link).

There is some overlap between the Asian populations as well as with American Indians due to their closely share genetic heritage. For instance, when running a skull through FORDISC that I knew to be Vietnamese it came back as being most similar to Guatemalan males. In order to deal with this overlap some anthropologists will simply report a skull as East Asian/Native American if it comes back as one of those groups. What about more clear cut cases where there isn't much overlap? Say a skull comes back as American Black. As the anthropologist, do you report that as a Black person or as an African-American? How many people from the Caribbean classify as American Black? What does society consider them?

Why do we still estimate race?

So forensic anthropologists know race isn't a valid biological unit. Why do we still bother? No good forensic anthropologist truly believes in race as it's classically defined, but population level differences do exist in some form and can be measured. To that end, mapping these populational differences onto the social classification of race is useful in identifying skeletal remains because society still uses race to characterize people. Despite the fact that it is difficult to connect ancestry with social race it can be successful and has been important to the resolution of many unidentified individuals.


I hope this discussion wasn't confusing or rambling. I'd encourage everyone to check out the further reading section to see some articles by people much smarter and more articulate than I am.

Further Reading

Gravlee, 2009. How Race Becomes Biology (links directly to pdf)

Ousley et al. 2009. Understanding Race and Human Variation: Why Forensic Anthropologists are Good at Identifying Race

Understanding Race