Essay on "Race"

  Ronald Kephart

University of North Florida

Perhaps no other aspect of human existence has been so mystified, so clouded by half-truths, falsehoods, and fantasy, as "race." Only language, the other obvious and readily visible human trait, comes close. And in fact, it is no accident that language and race are the two human characteristics that have been used, over the centuries, to sort the people into categories which can be labeled, stereotyped, and then used by those in control for their benefit. 

One of the functions of language is to facilitate communication with other people, and to make possible the transmission of a complex culture to the next generation.  But this is not language's only function.  Language also provides us with the main tool by which we categorize and make sense of the world around us.  Language does this by providing the oral symbols (words) which we use to label the things we need or want to pay attention to, think about, and talk about (the study of how language does this is called semantics).

Let's take a simple example to illustrate.  Everyone has arms and legs, and in every culture, at some point, there is a need to think and talk about arms and legs.  Every language provides its speakers with ways to do this. In North American English we label this part of the body with the words arm, hand, and fingers (we also have labels for the joints: wrist, elbow; however, since these refer to joints we will ignore them for our present purposes). In Arawak, a language spoken in parts of northern South America, the labels used to refer to the same "reality" are: daduna (shoulder to elbow); dakabo (elbow thru hand); and dakabosi (fingers).  Notice that there is no single label for what we refer to as the "arm."  At the same time, unlike us, they have separate labels for parts that we have to point with adjectives: upper and lower arm.

In Caribbean Creole English, the word han is generally used to refer to everything from shoulder to fingers; then we have finga on the end of our han.  (Note that this is not the only language that does this; Tohono O'odham, a language from the Southwest US, uses an identical system.) Similarly in Creole English, fut refers to everything from the hip down.

We can easily find other differences in semantics, as for example the use of the word tea to refer to any hot drink including hot cocoa, or the word evening to refer to any time after the midday meal, which means that you can hear "Good evening" as early as noon or so.  However, in this section I want to focus on how people label and categorize other people, something you all will agree that we certainly do in our culture.  Probably the most important way that we do this is when we place people into the categories we call races, and we have seen on the videos and elsewhere that Caribbean people do this also.  In order to compare our system of racial labeling with that used in the Caribbean, we need first to be sure that we understand the Northamerican system.

In this section, I want to discuss the ways in which language provides us with the symbols which we use to categorize our fellow humans, and how the ways in which those symbols are organized correlate with other aspects of our culture. But before we do that, I want to discuss the biological background to "race."

Folk models of race

If the classification of people into racial groups does not reflect an analytic model, then we have to assume that it reflects a folk model. But, our folk models are constructs of our cultures; we can expect that there may be more than one folk model regarding "race" and as it turns out this is true.

The North American folk model of race

In North America, the folk model of race is relatively simple. There are just two categories: "black" and "white." Although we say that our classification is based on skin color, and we even use color terms (white, black), in fact it isn't. We do not classify, for example, Indians (from India) who may be just as dark skinned as any African, as "black" (we are dealing here with the issue of African ancestry versus European ancestry; we certainly have other "racial "categories," but here in the US these are tangential to the black-white distinction).

The Northamerican folk model of race classifies people according to descent; people who have any trace of African ancestry, no matter how distant, can be classified as "black". A person can only be classified as "white" if they either have no African ancestry or if nobody knows about it. This system is sometimes referred to as a system of hypodescent, because people born of "mixed" matings are put in the category with the lower socioeconomic standing. The following table summarizes how this has worked traditionally in the US:

"black" mother

"white" mother

"black" father

"black" child

"black" child

"white" father

"black" child

"white" child


Notice the biologically illogical consequences of hypodescent; white women can give birth to either a white or a black child, while black women can give birth only to a black child. Analytically, this makes no sense.

Other folk models

Anthropologist Marvin Harris set out some years ago to compare the Brazilian folk model of race with the Northamerican one.  He had noticed that Brazilians use a much greater variety of terms to refer to people of different physical types; he also noticed that sometimes, the same person could be categorized differently depending on the context in which they were observed.  For example, a person categorized as preto (black) when wearing old clothes and standing in front of a poor-looking house might be called moreno (brown) or even branco (white) when seen all dressed up     in front of a wealthy-looking house.

To attack this problem systematically, Harris designed a set of cards each of which contained a portrait of a person.  There were 36 male and 36 female portraits.  The portraits differed by 3 skin tones, 3 hair types, 2 nose widths, and 2 lip types.  All 36 combinations were represented for each sex.  Harris gave people the sets of cards and asked them to sort them into racial categories.

Harris found that in Brazil, unlike the United States, people are not classified rigidly by descent into a small number of racial categories.  Instead, there is a wide variety of terms available to Brazilians which allows them to call attention to a combination of physical traits (skin color, hair form, etc.) and social class information.  The Brazilian sys   }tem is more like a continuum, in contrast to the North American system, which is a dichotomy.

Harris' explanation for the difference between US and Brazilian folk models of race involves taking into account the different socioeconomic conditions that prevailed in these places during colonial times.  In other words, it is a cultural materialist explanation.

In the US, during the period of slavery, the ratio of slaves to Europeans was relatively small; only in places like coastal South Carolina, where rice (to help feed Caribbean slaves) was grown on large plantations, were the ratios anything like those of Brazil or the Caribbean.  There were plenty of European immigrants available to fill the middle level jobs such as blacksmith, carpenter, wheelwright, and so on, as wel as to work on small farms providing food for the plantations, and even to serve as militia.  The strict cat    egorization of both free and unfree slaves into a "black" (Negro, Colored) category helped keep them from competing for these jobs with the European working class.  At the same time it helped keep the power of the planter class secure, by insuring that poor whites saw themselves as belonging to a category distinct and separate from both slaves and ex-slaves,  and thus preventing the two groups of exploited people from joinging together against the planters.

In Brazil, in contrast, there were relatively fewer Europeans available to do these sorts of jobs, which still needed to be done.  Harris thinks that this led to a more open system which allowed people to move around and occupy a greater variety of roles in the society.  In Brazil, and in other parts of Plantation America, there was no need to maintain strictly defined castes.

This is not  to suggest that Brazilians are "color-blind" or that skin color and other physical features do not make any difference at all to Brazilians.  In Brazil, light skin and Euopean features are valued over darker skin and non-European features and, in fact, Brazilians are probably more aware of these differences than we Northamericans are.  What it does mean is that in the Brazilian view of race people are spread out along a continuum with many more points on it, rather than being squeezed into two mutually exclusive categories.  The discrimination that does take place in Brasil and other places which follow the Brasilian pattern is more accurately defined in socioeconomic, rather than color-caste, terms.

The Caribbean folk model

As a graduate student at UF, I heard Harris lecture on his results and I thought it would be interesting to investigate Caribbean racial categories in the same way.  Most places in the Caribbean ought to follow a pattern more like the Brasilian one than the Northamerican pattern, since the Caribbean belongs mostly to Plantation America.  Of course, as we already know, the Caribbean contains a good deal of variety with regard to social stratification and social and economic conditions.  Before beginning my own investigation, I made the following hypotheses:

Societies which are stratified both by color and social class, like Trinidad and Barbados, should show the most racial categories.

Homogeneous or essentially unstratified societies, like Carriacou, should show the fewest categories.

Societies stratified by either class or color, but not both, should fall somewhere in the middle.

I borrowed Harris' cards from him and made photos of them.  Then I began by trying the portraits out on North Americans.  Without exception, they placed most of the portraits into discrete categories labelled White, Black, and Indian or Asian.  What happens after that varies; usually people have categories like Hispanic,Aamerican Indian, Australian Aborigine, etc. into which they put the portraits that don't fit the basic categories.  Interestingly, both black and white Northamericans did this.

I used the portraits on students at the University of Florida from Barbados, Trinidad, Jamaica, Bahamas, Dominica, and Grenada; I also tried them out on people in Carriacou.  In general, I found an overall pattern that more closely resembled Harris' Brazil pattern than the Northamerican pattern.  I also found that my hypothesis about the correlation between social stratification and intensity of categorization was confirmed.  Trinidad and Barbados ha
d the most categories, Grenada and Dominica had fewer; Carriacou had the fewest of all.



Consultants from Trinidad offered the following seven basic terms in response to the portraits:



Dogla (Indian + African)


Panyol (Venezuelan)



In addition, the following nine compound terms occured:





French-Creole (= Local White)


Koko-Panyol (African + Venezuelan)

Half-White (=Mulato)




I was not able to find a consultant from a society stratified by color but not by class.  A consultant from Dominica, a society stratified by economic class rather than by color, offered the following four terms in response to the portraits:

Nèg (= Black)



Blan or Beke (= White)

Besides these four terms, the consultant offered the descriptive phrase fairskin- which can be used to modify the first three terms already given.



On Carriacou, a relatively homogeneous society, no single-word terms such as White and Black were used in identifying the portraits.  Instead, color terms were used as descriptive adjectives: they must be accompanied by a noun like people, woman, man, girl, boy, etc.




pink- (=albino)

white- (Note that "white people" generally refers to outsiders; no one born and raised in Carriacou is normally thought of as "white.")

The implication of the finding for Carriacou is that in Carriacou, unlike both Trinidad and Dominica, people are described, but not categorized.


On Carriacou, people are also classified traditionally by the African nation (ethnic group) from which they claim ancestry.  This classification was more important in the traditional society than it is today, because it helped set limits on possible marriage partners: a person was not supposed to marry someone from the same nation as their mother.

We have discovered that, just as different languages and cultures categorize and label parts of the human arm-hand differently, so they may also classify and label people differently.  What does this say about the "truth" of any particular culture's racial classification model?  It suggests that these models— Brasilian, North American, Trinidadian, or whatever— are folk models of race, not analytic models.  If they were analytic models, presumably they would be a lot more similar since they describe the "same" reality.  What is the analytic model of race?  To answer that, we have to take a short trip into the realms of physical anthropology, to see if we can find an analytic model of race.

"Race": some biological background

Several facts about human evolution and biology may serve as background to a discussion and demystification of the "race" concept...
All living humans belong to a single biological species, called Homo sapiens. We know it is a species because we can all interbreed and produce fertile offspring, although the early slaveowners insisted that the offspring of Blacks and Whites were infertile, hence the term "mulatto" from "mule". We only need to recall the films we have seen, or walk down any street in any population center in the Caribbean, to know that they were wrong.
The species Homo sapiens, to the best of our knowledge, appeared first in East-central Africa about 200,000 years ago. From there, we spread out in waves to Europe, Southern Asia and the Pacific, and Northern Asia and eventually the Americas. This is supported by two different kinds of evidence: the fossil record, and comparative genetics of living populations.
Modern humans share nearly 99% of their genetic material with chimpanzees, our closest living relative. Apparently, we have been separated from chimps for between 8 and 10 million years. We have the same number of body hairs as chimps. The differences between us and chimps are more developmental than structural. (Chimps actually have one more pair of chromosomes than we have.)

Given the perspective which these facts suggest, it is obvious that modern humans, of whatever group, are extremely closely related. After all, if we are 98% chimpanzee (and chimps are 98% us), and chimps and people appear so different, how much difference can exist within our own species? Indeen, in some recent recent on chimp and human DNA, it was discovered that there is more genetic variation in a single band of chimpanzees than there is across the entire human species!

Nevertheless, people from different parts of the world do look very different. If we line up an Australian aborigine, a Masai cattle herder, and an Arawak Indian, we can definitely tell which is which. What does that mean? Before we decide, we need to know something else from biology...

Traditionally, for biologists, "race" is the same as "subspecies". Biologists, following the model set out by Linnaeus in the 18th century, were very concerned with classifying any plant or animal specimen into the tightest possible category, which meant defining it down to the level of subspecies or race. But what eventually happened was that animals and plants were being put into categories which varied from other categories by only one trait.

At this point, biologists began to realize that there was really a much more interesting problem going on. While they were busy classifying living things into static groups, the living things themselves sometimes refused to be classified easily; the boundaries between groups were very difficuly to pin down. Biologists began shifting their emphasis from classification on the basis of characteristics, such as skin color, to study of the reasons for the variation in these characteristics.

Here another factor comes into play: the nature of variation itself. Most variation is not simply on-off, but rather a continuum of possibilities between two extremes. Variation of this kind is called clinal variation. When this kind of variation occurs in nature, any division into discrete units is arbitrary.

Thus, biologists have for the most part dropped the whole idea of race or subspecies. These categories simply are not meaningful in terms of the biology of variation. In other words, the category "race" has been abandoned as an analytic model in biology; however, it remains a very important folk model for many people.


Skin color

The paramount example of clinal variation, for us, is skin color, because it is in fact by skin color that so much "racial" categorization takes place. Why do people have different skin colors? To answer that question, we have to understand that skin color has played, and continues to play, an important role in the adaptation of human groups to new environments, as outlined in the box below...

Human beings' bodies are composed of a certain amount of the mineral calcium. The amount has to be right; too little calcium results in soft bones, or ricketts; excess calcium results in kidney stones, calcium deposits, etc. The amount of calcium retained in the body is controlled by a hormone, calciferol ("Vitamin D").

The production of calciferol, in turn, is regulated by the amount of sunlight which penetrates the skins and stimulates glands which are located below the skin's surface. A dark pigmented skin impedes the penetration of sunlight and slows down calcium retention; a light skin lets more sunlight through and increases body calcium. This is why we all get darker if we spend a lot time out in the sun, say in the summer, but we lighten up during the cold months when we spend more time indoors. Our bodies are controlling for sunlight penetration and calcium production.

Homo sapiens evolved, by best evidence, originally in eastern Africa, in tropical or subtropical conditions. A darkly pigmented skin would have been necessary, under those conditions, in order for the early hominids not to have too much calcium. As they spread out into the world, however, these early Homo sapiens moved into more northern and southern climates where less sunlight was available, so they developed lighter skin to compensate. Meanwhile, those who remained in tropical areas retained their darker color. Biologists call this adaptation to local conditions.

 Once humans were more or less in place, say by around 20,000 BP, most populations were probably relatively well adapted to their own local conditions. Starting at the Equator, one would find relatively dark people. As one moved north, the people encountered would be lighter and lighter, until we reach Scandinavia, for example, where we find very light-skinned, blond-haired peoples. The variation in skin color would be represented by a cline, and there would be no place where "dark" ended and "light" began. Any division of human beings into "white" and "black" subgroups would be arbitrary.

There are some things to add to this picture:

Skin color is an extremely obvious trait, but only because our skin covers almost our entire body. The genetic material involved is about 6 pairs of genes, an infinitesimally small amount of our total genotype.

Skin color does not correlate well with other traits, despite the pervasiveness of our folk model regarding dark skin and kinky hair. For example, some Australians are very dark, but have wavy, blonde hair. South African !Kung San peoples have extremely tightly curled hair, but their skin color is quite light. These two traits, skin color and hair form, simply do not always go together; they do not define a coherent package which we can call a "race" or "subspecies". When we add invisible traits, such as blood groups, to the picture, it becomes even more clear why "race" as a biological category is not very satisfactory.

The relatively light skin color of most tropical American peoples reflects their relative late arrival from northeastern Asia where their light skin would have been advantageous.

The movement of peoples into areas which they are not locally adapted to can cause physical problems. Europeans moving into tropical areas, for example, tend to suffer from kidney stones. Darker skinned peoples moving into high latitudes tend to have problems with ricketts, especially before the introduction of "vitamin D milk".

We find, then that "race" is not an analytic model after all, even though it seems so "scientific" and many scientists have made a living from the study of it. If the classification of people into races does not follow an analytic model, what is it? It is a folk model.

"Race" and social structure

Why do the languages of different places, such as North America and the Caribbean, represent racial categories in such different ways? Marvin Harris suggests that the difference has its origin in the demographic and social conditions that existed in the North America and the Caribbean during the period of slavery.

Recall that from the earliest times, what eventually became the U.S. was essentially a settler colony; Europeans came from the Old World with the intention of staying and recreating their culture(s) in the colonies. As a result, there were almost everywhere in North America enough Europeans to fill all the middle-level jobs available (blacksmith, gunsmith, cabinetmaker, wheelwright, etc.), even in the southern colonies where slaves provided most of the labor on the largest plantations. In a situation like this, a mechanism must be found to prevent slaves and freed slaves from competing with Europeans for those jobs; to accomplish, the culture developed a system of hypodescent which effectively reinforced the privilege of the European segment of the population.

Contrast this with the situation on the islands of the Caribbean, which were basically exploitation colonies. In these regions there were never enough Europeans to fill the middle-level jobs; most Europeans were estate owners and managers, merchants, government officials, and the like. But somebody had to do those jobs; the people who did them were very often newly freed (or even not-yet-freed) slaves. Here a hypodescent rule was not needed, because there was no need to protect Europeans from Africans in the labor market; indeed, the Africans were needed, because those jobs had to be done.

So, in the Caribbean, as in Brasil and other parts of Plantation America, the linguistic system for representing people developed as a continuum which allowed for description of people according to their actual physical traits. In contrast, the North American representational system developed into a binary opposition between "black" and "white," with placement based on the hypodescent rule.

"Race" and intelligence

Over the years, many researchers, especially in psychology, have attempted to find correlations between "race" and intelligence. In every case, so-called "black" people have been found to have lower intelligence than so-called "white" people. These findings are nearly always based on the administration of "IQ" tests which purport to measure intelligence fairly and objectively. There are many problems with this research, a few of the principal ones being:
Despite the claims of IQ test enthusiasts that their tests measure "intelligence," there is only thing that IQ tests can be without objection said to measure: a performance on an IQ test. To what extent a performance on an IQ test is a reliable indication of some abstract mental trait called "intelligence" is open to question.

Despite the usual claims, IQ tests are unavoidably biased culturally. Indeed, the very act of test-taking reflects a cultural bias.

The categories of "black" and "white" into which people are sorted for analysis of IQ test scores are not reliable indicators of underlying biology. Research has shown that at least 90% of US "blacks" have a European ancestor somewhere in their background; amazingly, around 60% of all the descendants of Africans brought to the US during the slave trade are now in the "white" population, and about 1 in 6 people considered "white" in the US have an African ancestor.

These problems, taken together, suggest that the debate over "race" and "intelligence" ought to be a non-issue. But, one of the legacies of African slavery has been, and continues to be, a search on the part of some to construct narratives which justify continued abuse and exploitation of the descendants of those ten million or so slaves brought to the New World between the sixteenth and nineteenth centuries.



Ethnicity is often popularly equated with "race," but for anthropologists the two are very different. For anthropologists, "race" (if it exists) is a strictly biological classification. Ethnicity, on the other hand, is centered on language and culture. People who share a language and culture can be said to belong to the same ethnic group.

People who differ biologically may belong to the same ethnic group, and people who are biologically very similar may belong to different ethnic groups. In other words, ethnicity and biology are, for anthropologists, independent variables which may or may not correlate.

Copyright © 2000 by Ronald Kephart