After my online research on the subject, I learnt that, biologically speaking, many scientists believe that there is no such thing as a race. Homo sapiens as a species is only 200,000 years old, which has not allowed for any significant genetic diversification yet, and our DNA is 99.99% similar. I've read statements that there can be more genetic variation inside a racial group than between different racial groups, meaning that, for example, two individuals from the same "race" can have less in common with each other than with an individual from another "race".

Wikipedia on Race (human classification) quote:

Scientists consider biological essentialism obsolete, and generally discourage racial explanations for collective differentiation in both physical and behavioral traits

Q1: If Homo sapiens has no races (according to biologists), why are we so different morphologically? (hair/eyes/skin colour and even athletic performance seem to differ between human populations)

Q2: Is it common for other species too, when genetically close populations have very different morphological traits? Are there any other mammal or animal species that exhibit biological diversity comparable to human diversity, and how do taxonomists treat these species? (excluding intentionally bred domestic species to keep the comparison fair)

The question has been paraphrased to emphasize that it is the biological debate that is in question, not the sociopolitical. I.e., why is there no consensus in evidence and opinions of scientists?

  • 2
    $\begingroup$ Why did you remove "to the point that we cannot tell an individual of one racial background from another based on a DNA sample" from the question? This makes the first sentence in the answer obsolete. $\endgroup$ Commented Apr 16, 2015 at 16:06
  • 6
    $\begingroup$ You have to note that taxonomy is artificial $\endgroup$
    – One Face
    Commented Apr 17, 2015 at 1:57
  • 3
    $\begingroup$ Here a recent paper that mixes philosophy and population genetics that is probably of interest. I haven't read it though. $\endgroup$
    – Remi.b
    Commented May 12, 2015 at 2:29
  • $\begingroup$ See also Does it make sense to classify all humans in a single species? $\endgroup$
    – Remi.b
    Commented Sep 21, 2017 at 15:22
  • $\begingroup$ Some random googling gives this article on the topic (more searching might give other articles with different conclusions; I haven't made any comprehensive effort) which finds that ...human races/continental groups or clusters have no natural meaning or objective biological reality. $\endgroup$
    – Tgr
    Commented Dec 19, 2017 at 7:47

6 Answers 6


Firstly, it's not true that you can't tell racial background from DNA. You most certainly can; it's quite possible to give fairly accurate phenotypic reconstruction of the features we choose as racial markers from DNA samples alone and also possible to identify real geographic ancestral populations from suitable markers.

The reason that human races aren't useful is that they're actually only looking at a couple of phenotypic markers and (a) these phenotypes don't map well to underlying genetics and (b) don't usefully model the underlying populations. The big thing that racial typing is based on is skin colour, but skin colour is controlled by only a small number of alleles. On the basis of skin colour you'd think the big division in human diversity is (and I simplify) between white Europeans and black Africans. However, there is vastly more genetic diversity within Africa than there is anywhere else. Two randomly chosen Africans will be, on average, more diverse from each other than two randomly chosen Europeans. What's more Europeans are no more genetically distinct overall from a randomly chosen African than two randomly chosen Africans are from each other.

This makes perfectly decent sense if you consider the deep roots of diversity within Africa (where humans originally evolved) to the more recent separation of Europeans from an African sub-population.

It's also worth noting that the phenotypic markers of race don't actually tell you much about underlying heredity; for example there's a famous photo of twin daughters one of whom is completely fair skinned, the other of whom is completely dark skinned; yet these two are sisters. This is, of course, an extreme example but it should tell you something about the usefulness of skin colour as a real genetic marker.

  • 10
    $\begingroup$ @Herman Suppose that in some sport it is an advantage to be very tall. Then because all genetic variation of humans is in Africa, you would expect the tallest people of the world to be in Africa. But you would also expect the smallest people in the world to be an African people. Similarly, in Africa you would find those who can sprint fastest. But you would also find in Africa those which are the slowest runners. So expect all variation to be on that continent. Also remember that those Africans who were deported to the Americas as slaves, all came from a small area on the west coast of Africa. $\endgroup$ Commented Jan 11, 2014 at 20:24
  • 1
    $\begingroup$ @Jack Aidley, how do "deep roots" of African populations and "more recent separation of Europeans" explain higher genetic diversity in Africa? Sorry, it doesn't make "perfectly decent sense" to me. Europeans inherited these deep roots too when separated, no? From my ecology background I know that there are theories that claim that evolution occurs faster in the tropics, but I never thought about it in the context of human populations. That could explain your point, but I don't think that's what you meant. $\endgroup$
    – Th334
    Commented Jan 12, 2014 at 21:27
  • 6
    $\begingroup$ @Herman: Evolution follows a branching, hierarchical structure (broadly, it's more complicated than that) and Europeans group within the various African populations, i.e., they're essentially a subpopulation of a particular African population. $\endgroup$ Commented Jan 13, 2014 at 8:22
  • 2
    $\begingroup$ @JackAidley, do you mean the founder effect? A subset of the original African population that founded European nations did not represent the whole genetic pool of the original African population, is it right? $\endgroup$
    – Th334
    Commented Jan 13, 2014 at 9:39
  • 3
    $\begingroup$ @Herman: No, although that can also be a factor. Have a look at this image - bit.ly/1iHOMss - and note how there is no clade that includes both Africans but not the Europeans. This is because the European population is a subpopulation of the African population. The two African populations shown are as different from each other as the European population can be from the first African population. In reality, there are many more distinct African populations than shown here. $\endgroup$ Commented Jan 13, 2014 at 15:43

Well, that's just it, we don't actually have much phenotypic variation. For example, compare this:

Collection of human faces

to this:

Collection of dogs

or this:

Collection of cat faces

Or this:

Six human babies

to this:

Five dogs

This is phenotypic variation:

Collection of pigs

So, as I hope is clear from the images above, phenotypic variation among humans is tiny compared to other species. We just notice small differences much more because, well, it's us so little differences are much more noticeable.

  • 13
    $\begingroup$ Hmm... I'm not sure how useful comparing natural variation (human phenotypes) to artificial variation (dogs, cats and pigs) really is. Are you humans really less phenotypically variable than wild animal species? I guess ring species might make a good comparison? $\endgroup$ Commented Jan 11, 2014 at 12:52
  • 9
    $\begingroup$ @terdon: The bird of paradise are not a species, as you note in your answer and I don't think it's clear - at all - that humans show relatively little phenotypic variation. I would say that humans are probably on the high side (likely due to the huge range that humans occupy). $\endgroup$ Commented Jan 11, 2014 at 13:03
  • 1
    $\begingroup$ I also think it is misleading to compare human phenotypical variation and phenotypical variation that is the outcome of artificial breeding. To make this argument you would have to place human phenotypic variation within the range of phenotypic variation from a large number of "natural/wild" species. And comparing variation in birds of paradise (family Paradisaeidae) would only be valid if compared to Hominidae, i.e. including chimps, orangutans, extinct Hominini etc (still ignoring that taxonomic families cannot really be considered equal). $\endgroup$ Commented Jan 13, 2014 at 14:05
  • 1
    $\begingroup$ @Rodrigo yes, that was added after I posted this answer. $\endgroup$
    – terdon
    Commented Jan 14, 2014 at 19:57
  • 2
    $\begingroup$ @Alex I'm normally a stickler for quantification, but I think the point is made quite well qualitatively. Your rebuttal is neat -- the jackdaw has similar variation to us. But that itself is justification for ignoring the subspecies divisions in jackdaws as much as imposing subspecies labels in humans! @bshane offered a great summary of subspecies here that demonstrates how arbitrary a term "subspecies" is. $\endgroup$
    – James
    Commented Apr 11, 2016 at 9:36

I decided to summarize a competing hypothesis to make our answers more balanced. I also tried to address the question about the degree of human morphological diversity compared to other animals.

According to Woodley (2010), it is plausible that H. sapiens does not belong to one species and subspecies (i.e. is polytypic). Some of the data he uses to support this hypothesis could be useful for answering our question. He claims that H. sapiens, which is often considered monotypic, posses higher levels of morphological diversity, genetic heterozygosity and differentiation than many animal species which are considered polytypic.

Woodley cites a study by Sarich and Miele, who claimed that morphological differences between humans, on average, are equal to the differences among species within other mammalian genera (excluding species bred for domestic purposes), and are typically more strongly marked than in other animals.

However, morphological differences are known to be caused by little genetic differences too, like in the case of domestic dogs, which are still considered to be one species. Therefore, Woodley presented further evidence that looked on these inconsistencies in classification using allele frequencies and genetic diversity.

He presented data from a wide range of studies, which compares genetic diversity of various mammalian species based on heterozygosity (H), which is a common indicator for genetic diversity, and describes whether both alleles are the same or not on a studied locus. According to this data (which you can find in the linked paper):

  • Chimpanzees exhibited H of 0.63-0.73, which is very similar to H found in humans (0.588 - 0.807), however, chimpanzees are divided into four subspecies.
  • Some species like the grey wolf even exhibited a lower H (corresponding to lower genetic diversity) than humans (0.528 vs 0.588 - 0.807), while the grey wolf has been divided into as many as 37 subspecies.

This data suggests that humans are more diverse both morphologically and genetically than some of the other mammalian species that have been divided into subspecies.

Woodley, M. A. (2010). Is Homo sapiens polytypic? Human taxonomic diversity and its implications, Medical Hypotheses, 74, 195-201. doi:10.1016/j.mehy.2009.07.046 (full-text PDF)

  • 7
    $\begingroup$ It's a political question. They think (or pretend?) they're fighting racism denying human races. At the same time, a country with more species and subspecies receive more funds for conservational programs (see my answer). $\endgroup$
    – Rodrigo
    Commented Jan 14, 2014 at 19:32
  • $\begingroup$ How would that translate into multiple species? From my uninformed perspective it looks like a species with a high but uniform heterozygosity is possible, so that there would be absolutely no clear cluster or even that any attempt to build clusters based on this would lead to weird effects such as people ending up into a different species than their siblings. $\endgroup$
    – agemO
    Commented Jan 5, 2023 at 14:58

It seems to me that many answers to this question suffer from the nasty habit of "political correctness". As a zoologist, I never heard of somebody sequencing the whole DNA of any species to decide when to use or not the term "race". If a group of animals comes from a side of a river, and the other comes from the other side, and they have one or a few distinctive features (color of chest pelage, tuffs of hair in the sides of head, etc), that's enough to call them both different races (or even subspecies). Of course some geographical isolation has taken place, although it hasn't been long enough to divide the two (or more) populations into full species. The same logic should be used to humans, right? Maybe some folks believe that, pretending there are no human races, then the question of racism is "solved"? Bad logic, to me.

Look at http://www.worldbirdnames.org: they cite 10,518 extant species of birds and 20,976 subspecies. About 2 subspecies per species. How do they do that? (No DNA for most of them.) Subspecies/races are usually from different regions (like human races), their vocalizations are usually different (like human races) and their colors sometimes vary (not as much as in human races). My point is: there is NO SCIENTIFIC REASON to say there are no human races, if we're using zoological reasoning. (Unless we are not animals anymore!)

For mammals, http://www.catalogueoflife.org/col/browse/classification/kingdom/Animalia/phylum/Chordata/class/Mammalia/match/1 cite: 4,843 species, 2,998 infraspecific taxa. Again, if someone show me they did complex DNA analysis for most of those infraspecific taxa, and those analysis showed that variation inside races is smaller than between races, THEN I will be forced to agree with them.

  • 3
    $\begingroup$ I just inverted your question: "If human populations have several distinct phenotypes, why (they say) there are no human races?" Or, from other point of view, I questioned your first "If" in the title. It's not just a biological question, to me it seems much more a political one. How many biological races have been described? Or subspecies? Take a look at IUCN Redlist, for instance, or Zoological Records in many universities. Do you really think they scanned their DNA in depth? There's no biological reason for a different approach on human beings. As I said, it's just politics. $\endgroup$
    – Rodrigo
    Commented Jan 15, 2014 at 0:51
  • 3
    $\begingroup$ I think you make a fair point... I might paraphrase my question. Why paraphrasing? Although intuitively I agree with you that it might have something to do with politics, it's not really something that I wanted to discuss, and it's probably not even allowed to be discussed on Biology SE. If we do, someone will ask: "Well, how do you know that it's all just politics", and here we go, it becomes a debate with no science behind it, just opinions. $\endgroup$
    – Th334
    Commented Jan 15, 2014 at 1:00
  • 7
    $\begingroup$ @Rodrigo However, generally, I would argue that when classifying non-human animals into races we either know or imply a very low level of gene flow between races/subspecies. We know that this is not the case for humans. $\endgroup$ Commented Jan 24, 2014 at 11:23
  • 7
    $\begingroup$ While I see your point about there being races, I very much doubt you'll be able to support the use of the term as it tends to be used today. With the criteria you state, the Scots and the English would be different races, for example. If we take variations in vocalizations and geography as indications of race, each country will represent a race of its own. So while your argument does indeed support the existence of human races, it does not support the idea that those races are defined by color alone. $\endgroup$
    – terdon
    Commented Feb 2, 2014 at 20:00
  • 5
    $\begingroup$ Here is a source that might be what you're looking for: "The biological definition of race is a geographically isolated breeding population that shares certain characteristics in higher frequencies than other populations of that species, but has not become reproductively isolated from other populations of the same species." - biologyreference.com/Ar-Bi/Biology-of-Race.html#ixzz4dVXjcmSb $\endgroup$
    – Kent.Green
    Commented Apr 6, 2017 at 21:38


When you say phenotype you mostly mean "skin color", "size of the nose", "hair color", "shape of the eyes", "height", and some others. All these traits that we manage to find to explain population structure among humans. But you forget all the rest of the phenotypic diversity. If you would choose 1000 randomly chosen traits (external morphology and other stuff) and make a PCA. Will the main axes explain much of the inter-subpopulation (or inter-racial) diversity? I am not sure about that. One cannot use one subset of the total phenotypic or genetic variance and use it in order to define several species within what was previously thought to be one. It is not because two genes are associated with the same population structure that one can define two different species.

scientific observations vs biased intuitions

You say:

[..] is it reasonable to use a genetic approach to races and claim that biological races do not exist, while it seems to be a poor indicator, when it comes to comparison of phenotypes of human populations.

Similarly, one could say

Bats look like birds. We consider them as being mammals just because that is what scientific observations (genetic data or in-depth observation of the phenotypic variance) says. Should we follow scientific explanations when my personal bias intuition tells me that these observations are poor indicators?

The answer is yes if you want to increase your knowledge and no if you just want to be comforted into what you think you already know.

I have to confess though that I don't know much about phenotypic or morphological variations among humans. And I'd be curious if someone could give some words about that and whether or not much of the phenotypic and morphological variance is explained by what we consider being racial groups. It might be possible that very little of the total phenotypic variance is explained by racial groups but quite a lot of the face morphological variance is explained by racial groups.

Concept of species

For species that can reproduce exclusively with sex, we tend to use the concept of reproduction isolation to define a species and I really don't think that there is a pair of racial groups are sexually incompatible (however you delimit the racial groups). Eventually, there might have some slight inbreeding depression but I am not sure.

You might want to have a look at this answer to understand the semantic difficulties behind the concept of species

  • $\begingroup$ I would be very, very surprised to find out that any two racial groups cannot reproduce: not only we are one species, we are also one subspecies, and we share over 99.9% of the same genome. The world would be a totally different place if this was true, from the sociopolitical perspective. It might happen one day though. I hope collectively we'll be much wiser when and if it does happen. $\endgroup$
    – Th334
    Commented Jan 11, 2014 at 12:49
  • $\begingroup$ Yes, that would be immensely surprising indeed! But I don't think any study ever reviewed cases of reproduction between all "considered social groups". $\endgroup$
    – Remi.b
    Commented Jan 11, 2014 at 12:59
  • $\begingroup$ On the other hand, it also means that no one has ever noticed any suspicious trends in child mortalities from cross-racial parents to think about such a survey. I also believe that gentlemen from the Third Reich would be very interested in this question, and they conceivably could do some research, which wasn't published because the results were negative. It's just my guess though, and I'm no specialist in history. $\endgroup$
    – Th334
    Commented Jan 11, 2014 at 13:08
  • $\begingroup$ Remi, sorry, I'm not familiar with PCA, and the Wikipedia page is intimidating. Was your first point that we share so much traits in common that the ones that distinguish us from each other are statistically non-significant? $\endgroup$
    – Th334
    Commented Jan 11, 2014 at 14:04
  • $\begingroup$ "For species that can reproduce exclusively with sex, we tend to use the concept of reproduction isolation to define species and I really don't think that there is a pair of racial groups are sexually incompatible" But this reproductive isolation is not only due to genetic incapability, but to purely practical geographic isolation as well, right? $\endgroup$
    – Alex
    Commented Jun 7, 2015 at 10:18

First you can take a look at this diagram:

enter image description here

You can make a google image search for "Cavalli Sforza" and get a lot of similar diagrams. This diagram is using a concept known as genetic distance by fixation index. This is a way to measure how different different ethnic groups are genetically.

Basically what you do is you compare how many differences there are on average between two humans from the same population and between two humans from two different populations. The same method can be applied to animals. There is no "threshold value" where you can say that two groups of people or animals have become different enough that they can be defined as "different races".

Traits such as skin colour are regulated by a very small handfull of genes so even if we have different skin colours that does not mean that we differ much genetically speaking.

Lions and Leopards can have fertile offspring and live basically in the same habitat in Africa. I do not know if crossbreeding is known to have occured in the wild. Probably a hybrid has lower fitness than the original species.

  • 1
    $\begingroup$ So, what's your point about human races? $\endgroup$
    – Rodrigo
    Commented Jun 6, 2019 at 20:48
  • 1
    $\begingroup$ I would say focus on the actual differences instead of the name. Differences as far as Fst, IQ etc. goes. "These are the differences between group A and group B". I like the term "ancestral group", nobody can really deny that they have an ancestry. $\endgroup$
    – Agerhell
    Commented Jun 7, 2019 at 4:24

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .