August 06, 2010

A rare genomic look at Aboriginal Australians

How strange that modern genetics is supposed to have invalidated the concept of race, yet, at every turn, it confirms most of the basic racial taxonomic observations of people working only with their eyes and, much later, their calipers.

On the left is the frappe analysis from the supplementary material, the Oceanian populations are seen on the far right.

The Australasid cluster emerges as an entity at K=5, showing Caucasoid admixture (AUR), Mongoloid admixture (MEL), and no apparent admixture (PAP).

At K=8 it is evident that the Caucasoid admixture in Aboriginal Australians is specifically European in origin, certainly the result of colonization in very recent times.

What can account for the Mongoloid admixture in Melanesians? It is probably the recent spread of Austronesian languages, arguably the most epic maritime language spread before Columbus, which affected a good deal of the southern hemisphere from Madagascar through Indonesia, Micronesia, Melanesia, and all the way to Polynesia on the far end.

As for the unadmixed Papuans, the indigenous inhabitants of New Guinea, their results are not surprising: there is a lack of admixture of East Asian Y chromosomes on the island, even in its most affected NW corner (Bird's head) where this admixture runs only to about 2.5%.

The American Journal of Human Genetics, doi:10.1016/j.ajhg.2010.07.008

Whole-Genome Genetic Diversity in a Sample of Australians with Deep Aboriginal Ancestry

Brian P. McEvoy et al.

Australia was probably settled soon after modern humans left Africa, but details of this ancient migration are not well understood. Debate centers on whether the Pleistocene Sahul continent (composed of New Guinea, Australia, and Tasmania) was first settled by a single wave followed by regional divergence into Aboriginal Australian and New Guinean populations (common origin) or whether different parts of the continent were initially populated independently. Australia has been the subject of relatively few DNA studies even though understanding regional variation in genomic structure and diversity will be important if disease-association mapping methods are to be successfully evaluated and applied across populations. We report on a genome-wide investigation of Australian Aboriginal SNP diversity in a sample of participants from the Riverine region. The phylogenetic relationship of these Aboriginal Australians to a range of other global populations demonstrates a deep common origin with Papuan New Guineans and Melanesians, with little evidence of substantial later migration until the very recent arrival of European colonists. The study provides valuable and robust insights into an early and important phase of human colonization of the globe. A broader survey of Australia, including diverse geographic sample populations, will be required to fully appreciate the continent's unique population history and consequent genetic heritage, as well as the importance of both to the understanding of health issues.

Link

82 comments:

German Dziebel said...

"Australia was probably settled soon after modern humans left Africa, but details of this ancient migration are not well understood."

This paper doesn't clarify it either. And without these "details," you can't say if Australians and Papua New Guineans came from Africa. According to the PCAs, Sahul could've been peopled from America just as likely as from Africa. Notably, the authors highlight the lack of connection to India, which is consistent with haploid studies and flies in the face of a "southern route" idea. The qualifier "recent" makes even this observation somewhat uncertain.

Spy said...

I notice that when you leap from K=3 to K=4, yellow goes from representing Caucasians to representing Amerindians, whereas Africans become blue and Mongoloids take the orange. Is this completely arbitrary in frappe or does it signal a rearrangement of which breaks account for the most variance?

Unknown said...

"Notably, the authors highlight the lack of connection to India, which is consistent"

By that reason you can not even find Australoid connection to any other place than Australia itself from the same Map.

Can anybody give more details on this.

Australoids provided the core group in South East Asia, Andaman and India also in the beginning. In current genetic makeup ratio of Australoids in India is very less. Still the Macro haplogroups clearly show the connection.

http://en.wikipedia.org/wiki/Australoid_race.

Some of the researchers like to talk what they want to talk and make small irrelevant inferences like these . Not sure why these comments are not in mainstream.

Andrew Oh-Willeke said...

This study simply confirms the ancestry of people in the area in a way that has been a settled understanding for many decades. It would, of course, be quite troubling and cast doubt in the discipline if the "Caucasoid" admixed in Aboriginal Australians were specifically Chechnyian or North African given what we know about the history of Australia. This study validates the conclusions of prior population genetic studies and studies of pre-history by other means in this area, but it doesn't really do much of anything for the concept of race one way or the other.

While "basic racial taxonomic observations of people working only with their eyes" are sometimes right, particularly when the situation is simple (and tens of millenia of population genetic isolation in Australia until the 19th century is certainly one of the simplest stories in the world on that count), those eyeball observations are less reliable in more complex situations. South Asians sometimes look like Egyptians. Mestizo and a recent admixture of Northern Europe and East Asian look very much alike. Different people draw clusters in different places -- for example distinguishing subgroups within mixed descent African populations to different degrees.

Non-African people working only with their eyes also tend to have difficulty distinguishing between African population clusters despite the fact that taxonomically, they should seem more distinctly genetically separated populations there than the closely related populations of Europe or Asia, which locals there can often easily distinguish.

The more useful insight from population genetics about Papuan and Australian Aboriginal peoples is not the one made in this paper, but the cladistic connections between their genes and those of the Asia bound Out of Africa population, as opposed to the Europe bound Out of Africa populations or the didn't leave Africa populations. Frappe analysis doesn't show those cladistic relationships, it simply shows admixture clusters.

Moreover, it necessarily follows from the mathematics of the method used, that if you make K a high enough integer, that any group other than a completely homogeneous population is going to show structure and sub-populations. Part of the case against race is that you have to get to K=5 to see it in this case.

Those skeptical of the usefulness of the concept of race certainly don't deny that people from different ancestries look different. They simply deny that those differences in appearance are very important and that the right level of specificity at which to draw lines around clusters is self-evident in the hard cases.

terryt said...

"Notably, the authors highlight the lack of connection to India, which is consistent with haploid studies and flies in the face of a "southern route" idea".

Interesting. The connection is certainly more with East Asians, which is what Alan Thorne claimed way back, before he was howled down. I have not seen any evidence that connects the Australian Aborigines with any population in India, ancient or modern. Just a 'belief'.

Old Wombat said...

I was a bit mystified by the reference to the "Riverine" region in the abstract, where the sample was said to be from. After some thought, I concluded that it must be a typo for the Riverina in southern NSW. This area was settled by whites early in the 1800s and those identifying as Australian Aborigines there probably have more European than Australian in their genetic heritage -- more so than nearly anywhere else in rural Australia. It's a pity they didn't pick one of the many Australian populations that are less mixed.

Onur Dincer said...

Razib Khan has access to the full paper. According to the results of the full paper he presents in his blog, one of the Australian Aborigines tested in this study is %100 Australian Aborigine genetically. If you look at the frappe and STRUCTURE analyses carefully (better seen by zooming in), you will notice that one man among the Australian Aborigines tested in this study has indeed %100 Australoid component and thus no other component. Before the arrival of Europeans all or almost all of Australian Aborigines were %100 Australoid, but today such pure Australoid Australian Aborigines are in minority among Australian Aborigines (when I am saying Australia, I am always including Tasmania) due to widespread admixture with Europeans (and then also with other people like Chinese however much less widespread) in the last a few centuries.

According to the full paper, Australian Aborigines appear genetically very very close to New Guineans (Papuans, etc.) and Melanesians when the European components of Australian Aborigines are subtracted from the equation (thus leaving behind only the Australoid component to examine).

You should all take a look at Razib's thread about the full paper:

http://blogs.discovermagazine.com/gnxp/2010/08/50000-years-of-dreamtime/

Onur Dincer said...

Different people draw clusters in different places -- for example distinguishing subgroups within mixed descent African populations to different degrees.

I think that is the biggest problem of racial (subspecies) classifications as there is no gold standard for classifying subspecies (races). That brings some subjectivity to the subspecies (race) categories. But irrespective of how you define and classify subspecies (races), the genetic and physical differences are there and no one can deny them.

German Dziebel said...

"I have not seen any evidence that connects the Australian Aborigines with any population in India, ancient or modern. Just a 'belief'."

Yes. Looks like the connection if spurious. I gave more background at http://blogs.discovermagazine.com/gnxp/2010/08/and-they-came-from-central-asia/.

But this paper argues that the only identifiable connection Australians have is with Papuans and non-Austronesian island Melanesians. East Asia doesn't figure either. This implies what we already know, namely that Australia and PNG were populated early on before the formation of mainland East Asian populations. I do believe that the colonization of Australia followed the coast, only the Pacific Coast. What I wonder is whether there was a later backflow of Oceanic people back north and west in the direction of the Andaman islands and India. The fact that Oceania harbors various Y-DNA C lineages (C4 in Australia, C2 in Melanesia and Polynesia, C6 in Papua New Guinea), while Andaman islanders and Southeast Asians have closely related D lineages may point to this kind of backflow carrying C, D and K (?) lineages. India does have C5, which is noteworthy. And we do have a puzzling Ongan-Austronesian linguistic connection.

German Dziebel said...

"What can account for the Mongoloid admixture in Melanesians? It is probably the recent spread of Austronesian languages, arguably the most epic maritime language spread before Columbus, which affected a good deal of the southern hemisphere from Madagascar through Indonesia, Micronesia, Melanesia, and all the way to Polynesia on the far end."

Do we really know that "Mongoloid" genes came to Melanesia with the speakers of Austronesians languages? Or the Mongoloid admixture occurred AFTER Austronesians have been already in place at least in some parts of SE Asia and Oceania. I understand there's quite a bit of authoritative writing behind the former scenario but do we know for certain that the north-to-south Mongoloid expansion and Austronesian expansion are the same? If we look at teeth, the Ami, Yami and Bunun of Taiwan are Sinodonts, which is Northern Mongoloid. But all Malayo-Polynesians are Sundadonts, which is very different, and in fact, according to Turner, ancestral to Sinodonty. I'm somewhat uncomfortable with the prevalent, but ill-supported idea that Austronesian-speakers replaced a great number of indigenous Melanesians who adopted Austronesian languages but claim the majority of Y-DNA chromosomes now found in Austronesians. Could pre-existing Austronesians have absorbed incoming Mongoloids without the latter having any impact on their languages?

Mongols said...

Malay Mongoloid skulls appeared in SE Asia after that neolithic culture expanded there, while the natives were australoid-like people. I believe that most of non-negrito SE asians migrated from south China during the neolithic or bronze time.

Amanda S said...

I think that it would be interesting to see how the Australoids compare genetically to some of the Indian tribal and various negrito ethnic groups from SE Asia.

If I just saw a picture of the young rugby player that is featured in this newspaper article, it would have assumed that he was of Melanesian descent but infact he's Indian.

http://www.guardian.co.uk/sport/2010/may/21/scrumdog-millionaire-india-rugby

I think that there is a component in the ancestry of Indian and SE Asians which descends from the original people who migrated from Africa but that, on the whole, it's overwhelmed by other genetic influences.

Onur Dincer said...

Wish they also studied populations from more Dravidoid places like India (both north and south) and Bangladesh (I guess Bengalis would have also some clearly visible Mongoloid component in addition to the Dravidoid and Caucasoid components looking at the resuts of the previous genetic studies about them). Xing et al.'s recent paper is important in its focus on Indian populations:

http://dienekes.blogspot.com/2010/07/more-uniform-sampling-of-human-genetic.html

BTW, in one of my above posts I accidentally wrote "%100" instead of "100%". This is a natural result of reading and writing in Turkish since the age of 5.

Onur Dincer said...

Amanda, the Dravidoid race I am talking about may be or may not be most closely related to the Australoid race among all human races, but I don't know which one is true. I am sure more studies, more populations, and more details and pecision (like whole genome studies) will ultimately give the final answer to such questions.

terryt said...

"Do we really know that 'Mongoloid' genes came to Melanesia with the speakers of Austronesians languages? Or the Mongoloid admixture occurred AFTER Austronesians have been already in place at least in some parts of SE Asia and Oceania".

It's quite possible that the Mongoloid move south was a prolonged process. It contributed to the Austronesians, but carried on long after they'd begun their expansion.

Andrew Lancaster said...

Onur said: "I think that is the biggest problem of racial (subspecies) classifications as there is no gold standard for classifying subspecies (races)."

Actually, I think this is the objection to the concept of human races which Dienekes can rightly say is collapsing. In other words, there are now more objective ways to divide up the species.

But what Dienekes is missing with his frequent remarks on this is that being able to divide up a species objectively into relatively homogenous inter-breeding sub-populations, for example (a favorite example of Dienekes) even the regions around Reykyavik, is doable now, and no one is debating that, but this is quite simply NOT what most people mean when they refer to human races.

So to save the concept, if it is worth saving, the meaning has had to change for Dienekes, away from what it means for other people. So when he says others are proven wrong and he is proven right, this is where he is missing the point.

To me it would seem far less confusing to simply say that races, as the term is generally understood, do not exist, and to call sub-populations sub-populations, and to state clearly what characteristics those sub-populations have.

Avoiding confusion seems to me to be the main goal when deciding how to use words.

Best Regards
Andrew

princenuadha said...

"While "basic racial taxonomic observations of people working only with their eyes" are sometimes right, particularly when the situation is simple (and tens of millenia of population genetic isolation in Australia until the 19th century is certainly one of the simplest stories in the world on that count), those eyeball observations are less reliable in more complex situations.

I'm amazed at how often older physical anthropology is consistent with new genetic studies.

They had a pretty good picture of the races which captured a lot of the population clusterings and racial archetypes. (an archetype would be European or northeast Asian; a non-archetype would be Mongolian which is a mix of two previous. Older anthropology could even precisely drew a divide between the Caucasians and Negroids and the Southeast Asians and the Australiods. They knew the NA were closest to East Asians. And of course they thought we all came from Africa.

Some important things old anthropology got wrong or failed to answer, which genetic studies clarified, was that Indians are not just a simple mix of Caucasians, East Asians, and Australiods. They are instead another racelike group. Also that the Irish are just like the English and that both cluster with other Northwest Europeans and not the Southwest Europeans. There's the genetic impact from the Neolithic revolution on Europe. Genetics does also got rid of a credible Nordic, Alpine, and Mediterranean scheme of Caucasians. Old anthropology did not know how far Africans are from non-africans. I also don't think they knew how genetically distant the NA are from the East Asians.

Onur Dincer said...

If we will continue to use the term "race" (subspecies) in humans, we should first demonstate what difference it has from the term "genetic cluster".

My proposal is that races are genetic clusters separated from each other to a high degree and long enough to have clearly and easily identifiable consequences on the physical traits (especially craniofacial traits, which are largely hereditary) of humans. This usage of the concept of race is perfectly in line with its usage in other animals with the name "subspecies", as subspecies are defined not solely based on genetics but also physical traits.

Being clearly and easily identifiable means that you can identify them only with your eyes (as already happened to a high accuracy rate before the emergence of anthropometric practices).

Andrew Lancaster said...

Onur said:
"If we will continue to use the term "race" (subspecies) in humans, we should first demonstate what difference it has from the term "genetic cluster".

So what is the difference?

"My proposal is that races are genetic clusters separated from each other to a high degree and long enough to have clearly and easily identifiable consequences on the physical traits (especially craniofacial traits, which are largely hereditary) of humans. This usage of the concept of race is perfectly in line with its usage in other animals with the name "subspecies", as subspecies are defined not solely based on genetics but also physical traits."

So you are proposing a new way of using an existing English word? Very ambitious.

But I think you have to think about this a bit more.

1. Why do we in fact need a different word (race) for sub-species only when we talk about humans?

2. You seem to say that any genetic cluster which can be found by looking at more than genetics is a "race". So that raises some questions. Why is this method being used especially? How many layers of race can there be within one species?

But let's not ignore the bigger issue which is that normal English usage demands that there is only one level of race within humanity. Darwinian science on the other hand says that there is nothing fixed or special about races at all. They are just inter-breeding populations. Nothing in nature means that the boundaries have to be clearly defined, and often they clearly are not? If that is a problem for language as currently used, maybe we should drop words which reflect a pre-Darwinian misunderstanding.

The basic problem, even if you find a new definition for race which makes sense, is that the word race as really used is in conflict with modern biology. The word race is not used to mean "sub-species" or "genetic cluster". It is normally used quite specifically in very specific ways which are fully intended to be different from those Darwinian terms.

Best Regards
Andrew

Onur Dincer said...

I think marginal and quite isolated groups like the Kalash, Basques and Sardinians somewhat distort the genetic picture in such large-scale studies by causing distinct non-marginal populations to appear closer to each other than they really are (as, I guess, is happening between non-marginal West Eurasians and non-marginal South/Central Asians because of the marginal populations of those regions). So I think marginal populations should be employed less frequently or completely excluded in such studies.

Onur Dincer said...

Why do we in fact need a different word (race) for sub-species only when we talk about humans?

We don't. That human-other animal dichotomy (including the terminology used) is completely artificial and anthropocentric in nature. So we should drop the word "race" and use "subspecies" exclusively for the branches of every species.

Why is this method being used especially?

I don't want to repeat myself, but I already explained that with these words: "This usage of the concept of race is perfectly in line with its usage in other animals with the name "subspecies", as subspecies are defined not solely based on genetics but also physical traits."

How many layers of race can there be within one species?

One. As globally recognized biological conventions only allow for one layer of subspecies for every species. So the prime races I listed in my first post in this thread should actually be the only races (subspecies) in humans, so I must have used the word "race" or "subspecies" instead of "prime race". BTW, I think I should have counted American Mongoloids ("Americoid" from now on) as separate from Asian Mongoloids (just "Mongoloid" from now on) and Pygmes ("Pygmoid" from now on) and Capoids as separate from Negroids proper (I will call Negroids proper just "Negroid" from now on) in the same list, thus as different races (subspecies), this makes the number of races (subspecies) in humans nine instead of the five I mentioned, ten if we consider Negritos ("Negritoid" from now on) as a separate race (subspecies).

But let's not ignore the bigger issue which is that normal English usage demands that there is only one level of race within humanity. Darwinian science on the other hand says that there is nothing fixed or special about races at all. They are just inter-breeding populations. Nothing in nature means that the boundaries have to be clearly defined, and often they clearly are not? If that is a problem for language as currently used, maybe we should drop words which reflect a pre-Darwinian misunderstanding.

Modern post-Darwinian biology uses only one layer of subspecies (race), so I don't understand what you mean here.

The basic problem, even if you find a new definition for race which makes sense, is that the word race as really used is in conflict with modern biology. The word race is not used to mean "sub-species" or "genetic cluster". It is normally used quite specifically in very specific ways which are fully intended to be different from those Darwinian terms.

What do you mean by "the word race as really used"? Also do you think my definition of race (subspecies) is also in conflict with modern biology?

princenuadha said...

here's my definition of a race; its pretty basic.

A race needs to be a cluster on a global scale:

ex.; while Iberians form a cluster on one the European scale, at the global scale Iberians are simply European and therefore not a race.

A race cannot be a cluster that is made up of other clusters:

ex.; mestizo is mortality a race. Nor would Okinawans of mixed European and Japanese.

Note that I don't about physical traits only relative differences in genetics.

Onur Dincer said...

this makes the number of races (subspecies) in humans nine instead of the five I mentioned, ten if we consider Negritos ("Negritoid" from now on) as a separate race (subspecies).

I accidentally miscalculated the total number of human subspecies (races): They are eight without Negritoids, and nine with Negritoids.

Onur Dincer said...

A race needs to be a cluster on a global scale

All of the races (subspecies) of humans I have presented are clusters on a global scale. But global clusters like the Kalash cluster in Rosenberg's studies cannot constitute a race (subspecies), as the Kalash cluster is a result of recent genetic isolation, indeed the Kalash are physically indistinguishable from the other highly admixed Caucasoid-Dravidoid hybrids in the greater region. On the other hand, Amerindians are clearly physically distinguishable from Mongoloids in Asia because of their extremely long isolation from them (and also from the rest of the world), so they constitute a separate race (subspecies), which I call "Americoid".

A race cannot be a cluster that is made up of other clusters

But of course, there will be sub-clusters within races (subspecies) as you go further in Ks even in global studies (as in this and the other studies), which can easily be explained as results of the clinal (thus in a continuum) geographical distribution of within-racial genetic and physical traits, so they cannot constitute a race (subspecies). On the other hand, clusters of the initial K levels correlate very well with the conventional race categories, and their combinations should point to racial admixture. For instance, many, if not most, Subcontinentals, have Caucasoid admixture (especially those in the northwest and those from the high castes) and those in Nepal and northeastern regions like Bangladesh have Mongoloid admixture, but there are also many full Dravidoid Subcontinentals (especially in the south and among lower castes and tribals), which all correlate well with phenotypes. Look at the Subcontinental populations in this study:

http://dienekes.blogspot.com/2010/07/more-uniform-sampling-of-human-genetic.html

In conclusion, populations that are results of relatively recent racial admixture like Central Asians, some South Asians (see above), Saharans, some Arabs, Pacific Islanders (I am excluding New Guinea and its adjacent islands here), some East Africans, and of course, Mestizos, Mulattos - and all sorts of modern racial combinations - cannot constitute a race (subspecies), they are all racial hybrids.

terryt said...

"Darwinian science on the other hand says that there is nothing fixed or special about races at all. They are just inter-breeding populations. Nothing in nature means that the boundaries have to be clearly defined, and often they clearly are not?"

Everything you've said there holds for subspecies as well. It's just that most other species are not as mobile as the human species, so they tend to give rise to more discrete populations, although exceptions do occur.

"If that is a problem for language as currently used, maybe we should drop words which reflect a pre-Darwinian misunderstanding".

And, as Onur suggests, use 'subspecies' for humans rather than 'race'?

Andrew Lancaster said...

AL: Why is this method being used especially?

Onur: I don't want to repeat myself, but I already explained that with these words: "This usage of the concept of race is perfectly in line with its usage in other animals with the name "subspecies", as subspecies are defined not solely based on genetics but also physical traits."

Yes, so why use another word for humans? But I can now understand that this is not really your preference?

AL: How many layers of race can there be within one species?

Onur: One. As globally recognized biological conventions only allow for one layer of subspecies for every species.

Yes, but that is a Linnaean/Aristotelian i.e. pre-Darwinian tradition, and in reality it is now pretty much ignored in biology. To force a system of categorization to recognize fixed numbers of clusters, is to imply that this is also how nature is. But Darwin proved that this is not how nature is. There are not fixed species and nor are there fixed ways in which they cluster, and what's more he proved that there is no real difference between clustering of species and clustering of sub-species.

AL. The basic problem, even if you find a new definition for race which makes sense, is that the word race as really used is in conflict with modern biology. The word race is not used to mean "sub-species" or "genetic cluster". It is normally used quite specifically in very specific ways which are fully intended to be different from those Darwinian terms.

Onur. What do you mean by "the word race as really used"?

I simply mean that race is a normal English word used every day by non scientists. The way it is used connects to old and wrong scientific or pseudo scientific theories. It is not used in a way which is the same as what biologists mean by sub-species. Biologists seeking to use the everyday word are of course free to pretend that race is just a word for sub-species, but as explained above, this is not correct.

Onur. Also do you think my definition of race (subspecies) is also in conflict with modern biology?

It depends. Are you really insisting that there is only one layer of sub-speciation possible in nature? You do not say this outright, but your approach implies it.

Best Regards
Andrew

Onur Dincer said...

Everything you've said there holds for subspecies as well. It's just that most other species are not as mobile as the human species, so they tend to give rise to more discrete populations, although exceptions do occur.

Yes, but we should differentiate between various forms of the modern and pre-modern human mobility.

But I can now understand that this is not really your preference?

Yes.

I simply mean that race is a normal English word used every day by non scientists. The way it is used connects to old and wrong scientific or pseudo scientific theories.

What do you mean by "old and wrong scientific or pseudo scientific theories"?

Yes, but that is a Linnaean/Aristotelian i.e. pre-Darwinian tradition, and in reality it is now pretty much ignored in biology. To force a system of categorization to recognize fixed numbers of clusters, is to imply that this is also how nature is. But Darwin proved that this is not how nature is. There are not fixed species and nor are there fixed ways in which they cluster, and what's more he proved that there is no real difference between clustering of species and clustering of sub-species.

Are you really insisting that there is only one layer of sub-speciation possible in nature? You do not say this outright, but your approach implies it.


I think you don't understand the subspecies categorization conventions. Allowing one layer of subspecies doesn't mean there can't be any other layer below; it only means that only the members of the first (uppermost) layer can be called subspecies. Wouldn't it be nice and more scientific to have more than one layer of subspecies? Maybe, but that would also cause some confusion and irregularities in the classification (in addition to the already existing ones) of subspecies, so only the first layer is accepted as the layer of subspecies just for the sake of ease and clarity, not because of any scientific limitation to the number of layers; so it is just a matter of convention, not a matter of some absolute scientific facts.

princenuadha said...

@ onur

"On the other hand, Amerindians are clearly physically distinguishable from Mongoloids in Asia because of their extremely long isolation from them (and also from the rest of the world), so they constitute a separate race (subspecies), which I call "Americoid""

My definition does not care about physical traits, only genetics matter. Genetics are the fullest evolutionary record in the human body. In fact, recognized "race traits" are merely a subset of expressed genes.

In response to your second paragraph, when I said "a race cannot be a cluster that is made up of other clusters", I should have said "a race cannot be a cluster that could be recreated from other clusters". When I say recreated I'm talking about theoretical combinations (of information) of clusters to produce a synthesized cluster. This synthesizing process must reflect real world biological/inheritance rules such that if those cluster populations were to mate they could produce the synthesized cluster population. (Also, "other clusters" means clusters outside the given cluster in question).

Examples:

Mestizos are an example of a cluster that could be recreated by a combination of other clusters (mostly Spanish and NA).

Caucasians are a global cluster which cannot be recreated using other clusters; so Caucasian is a race.

Here is a succinct definition:
A race is a cluster on a global scale which cannot be recreated from other clusters.

princenuadha said...

@ onur


"On the other hand, Amerindians are clearly physically distinguishable from Mongoloids in Asia because of their extremely long isolation from them (and also from the rest of the world), so they constitute a separate race (subspecies), which I call "Americoid""

My definition does not care about physical traits, only genetics matter. Genetics are the fullest evolutionary record in the human body. In fact, recognized "race traits" are merely a subset of expressed genes.

In response to your second paragraph, when I said "a race cannot be a cluster that is made up of other clusters", I should have said "a race cannot be a cluster that could be recreated from other clusters". When I say recreated I'm talking about theoretical combinations (of information) of clusters to produce a synthesized cluster. This synthesizing process must reflect real world biological/inheritance rules such that if those cluster populations were to mate they could produce the synthesized cluster population. (Also, "other clusters" means clusters outside the given cluster in question).

Examples:

Mestizos are an example of a cluster that could be recreated by a combination of other clusters (mostly Spanish and NA).

Caucasians are a global cluster which cannot be recreated using other clusters; so Caucasian is a race.

Here is a succinct definition:
A race is a cluster on a global scale which cannot be recreated from other clusters.


sorry if this is a repost but I did not get a success message.
Dienekes, please allow us instant posts or take a poll of what people want.

Onur Dincer said...

Nuadha, I will keep in mind your definition in order to see how robust it is. Until then, I will continue to use my own definition.

terryt said...

"I should have said "a race cannot be a cluster that could be recreated from other clusters". When I say recreated I'm talking about theoretical combinations (of information) of clusters to produce a synthesized cluster. This synthesizing process must reflect real world biological/inheritance rules such that if those cluster populations were to mate they could produce the synthesized cluster population".

It is quite apparent that several subspecies of dabbling ducks have actually formed from hybrids between two other subspecies. Therefore it is surely quite possible that the same could happen for humans.

Gregory76 said...

I have long objected to classifying Australians and Papuans together as one race. They have phenotypical differences, such as the hair form, and their differences in language further gave reason to suspect that genetics would not show them to be close (aside from some mixture of each into the other). I believed that the Papuans were closest to the Negrito and saw called them both “Negritoid” (with the understanding that the Negrito were purer and the Papuans differed in having some Australian admixture).
Now when I look a genetic studies of Y chromosomes and mtDNA it seems to me to mostly confirm this view (though the degree of admixture is somewhat greater than I thought), except for one major complication.
This complication seems to make the overall picture confirm something like the trihybrid hypothesis of J. B. Birdsell. They identified 3 components in Australia: Carpentarians (Veddoids), Murrayans (archaic Caucasoids) and Tasmanoids (Negritoes). This seems to parallel the 3 major Y haplotypes and 3 major mtDNA haplotypes:
(1) mainly Australian (Carpentarians, male C, and female N); (2) mainly New Guinean (Negritoes, male M, and female Q); and (3) having comparably large minorities in both areas (Murrayans, male K, female P). The only major flaw is that male M is descended from male K rather than the male D of the Negritoes; this suggests that male M Murrayans killed off the Negrito males (presumably D or a descendant) and married their widows, and later K evolved from P in their descendants, or that K had already evolved from M and male K Murrayans killed off the Negrito males and married the widows. The rest can fit with the hypothesis (female Q is mainly confined to New Guinea but is descended from the female M, which prevail among Negrito women; female P is mainly confined to New Guinea and Australia but is descended from female N, and N or its descendants prevail among Caucasoid women; female N is mainly shared with Caucasoids rather than Veddoids, but that merely means that the strongest strain among the Australians—the so called “Carpentarians”—shares a father with the Veddoids but a mother with the Caucasoids).

Andrew Lancaster said...

"I think you don't understand the subspecies categorization conventions. Allowing one layer of subspecies doesn't mean there can't be any other layer below; it only means that only the members of the first (uppermost) layer can be called subspecies."

I understand the tradition you are talking about, and it has pre-Darwinian roots.

What you are saying is that we have to follow the tradition. (But then again you seem to want to follow it, because it suits you. You are happy to propose changes to the tradition, for example in how to define a single level of sub-species, but you want the system to stick with pretending there is only one level in nature?)

I do not agree that it is important to keep to a tradition in science if it is known to be wrong. If the tradition does not fit the reality we find in nature then it should not be followed.

This is science right? These naming systems are ONLY for trying to describe reality. To the extent they fail, they can and will be changed.

You can see why wrong traditions should not be respected when you consider the confusion they can cause. For example in your case you are not clear, I think also not to yourself, about whether there really is one most important level of distinction within the human species as it really exists in nature. If you do not think this is real, then why do you propose trying to create a new tradition (your proposed way of defining a single level of sub-species in humans) which pretends it is so in nature?

Best Regards
Andrew

Onur Dincer said...

Andrew,

One shouldn't take current subspecies too seriously. They are subject to re-classification with increasing information. And even if they remain stable, their branches and branches of their branches may be re-classified with increasing information.

As to naming conventions, to hell with them. They also shouldn't be taken so seriously. Even just among Caucasoids there are many sub-subspecies groups as demonstrated by physical anthropologists and recently by geneticists. But there is no natural limit to further dividing the extant categories (you can go down to sub-village and family levels!), so one should put a limit somewhere just for the sake of practicality in cataloguing taxa, and not because that sub-subspecies categories are scientifically meaningless or invalid. For instance, DNA Tribes prefers a categorization of humans that consists of sub-subspecies and inter-subspecies hybrids, and their categorization is also scientifically valid (though more artificial than mine):

http://www.dnatribes.com/sample-results/dnatribes-global-survey-regional-affinities.pdf

terryt said...

"You are happy to propose changes to the tradition, for example in how to define a single level of sub-species, but you want the system to stick with pretending there is only one level in nature?"

I don't think anyone would claim just one level beyond the species level. Arguments continue concerning dabbling ducks, for example, as the whether there are three subspecies of Pacific black duck or just one species, whether the several variations in North America are separate species, subspecies or sub-subspecies, and so on. As onur said, 'They are subject to re-classification with increasing information'.

"The only major flaw is that male M is descended from male K rather than the male D of the Negritoes"

I'm fairly sure that Y-hap D is not characteristic of Negritos. It is not particularly common in SE Asia for a start, being found only in parts of Malaysia, Borneo and Sumatra. The last two having no Negitos anyway. D is common in Tibet, and was most probably introduced to SE Asia by movement of Tibeto-Burman speakers from near there, long after Negritos had become isolated in their jungle hideouts. The haplogroup presumably spread further than the language did.

"I have long objected to classifying Australians and Papuans together as one race".

I agree. Especially seeing that in Australia Y-hap C appears to be an early arrival (C4) whereas in New Guinea C (in the form of C2 and C2a) seems to have arrived less than 10,000 years ago (or even more recently) from Wallacea. Y-hap K and its derivative M are more characteristic of New Guinea. Separate settlement events, although K later spread to Australia too. The mtDNA haplogroups are also relatively separate.

Andrew Lancaster said...

Onur said: One shouldn't take current subspecies too seriously.

That is my point. So...

When you say sub-species or interbreeding sub-population is sounds as "un serious" as it is. These are clear terms.

What I can not understand is the "need" people seem to feel to use a very serious word like race to replace those clear scientific words.

Onur Dincer said...

When you say sub-species or interbreeding sub-population is sounds as "un serious" as it is. These are clear terms.

What I can not understand is the "need" people seem to feel to use a very serious word like race to replace those clear scientific words.


I always prefer the term subspecies to race, too.

As to the seriousness of the word race, I don't agree with you on that matter; race is obviously a more ambiguous (because of its very different historical usages) thus less serious word than subspecies (one of the reasons why I always prefer subspecies).

As to the clarity issue, I don't think subspecies is as clear a concept as it seems from afar. Different species can be subjected to vastly different criteria in defining their subspecies, so there is some subjectivity here (that brings us back to the gold standard issue).

Unknown said...

I wish to commend the comments made by Andrew Lancaster and Andrew Oh-Willeke.

There is no reason, nor any ethical justification, for trying to salvage either the word or the concept of race. It is an ugly word and an ugly concept, with an ugly history.

Race has never been a neutral, benign concept; and attempts to resurrect it as such are, quite frankly, ill-considered and offensive. Moreover, as Mr. Lancaster has aptly pointed out the word in English, both connotatively and denotationally, is unscientific.

Does this mean that there are not groups of more closely related humans and groups of less closely related humans? Of course not. But this has never been at issue. Even the most ignorant fool understands that some people are more closely related to each other than they are to other people, and that often (though by no means always) people who look alike are more closely related than people who do not. A more trite observation could hardly be made.

The fascination and importance of genetic mapping is not that it proves that 19th century bigots were right, but that it shows us how wrong they were--it shows us how closely related we all are to people who superficially seem so different. It shows us the paths our forebears took as they wandered the globe. It shows us the story of "us."

But "race" is not a concept not about "us," it is and always has been a concept about "them."

If you need a special word to denote the levels and degrees of interrelatedness among and between human populations, a far better word than "race" already exists, one which is also more descriptively accurate. That word is family (family has the added benefit of highlighting the interconnectedness of the human species and the fact that the differences we are talking about are minute).

I have blue eyes; a blued-eyed aborigine is more closely related to me, than not. Perhaps a blue-eyed swede is even more closely related to me than the aborigine, but we are all still, quite literally, family, sharing the same grandparent 6-10,000 years ago. And attempts counter this comment with, "well, but a pure blood aborigine..." only serve to highlight the profound problem with the concept of "race:" the only "pure" people are those who were, and are, physically prevented from having sex with other populations (genetic mapping shows that cultural restraints were, and are, absolutely useless at doing so). Any population that was, or is, able to have sex with members of another population, did so, and does so. So whatever "race" is, it doesn't seem to have any real bearing on the most intimate act two humans can engage in.

Like the concept of "breed" in dogs the concept of "race" is artificial and largely pointless (other than to denote inbreddedness and a high probability of genetic unfitness). If given the chance, all dogs, just like all people, will breed with each other reverting to "mutts." Perhaps the correct observation is that the mutt/multi-racial is the true phenotype, and the breed/pure racial is the aberration? If this is so, do you still feel so wed to the notion of "race?"

Onur Dincer said...

Kelly,

Majority of current dog breeds were developed by humans within the last two centuries, the rest have been with us for several thousand years at most. On the other hand, current human races (subspecies) have been around (not necessarily in the same form with their modern descendants) from a minimum of 10,000 or more years to a maximum of 100,000 or more years. So your dog breed analogy is irrelevant and wrong.


Your overall arguments have already been refuted years ago by someone whom we are all familiar with:

"There are two kinds of people in the world who think about race. The first kind are the ones I call, the Scary. They are the neo-Nazis, KKK, Black Panthers, etc. who espouse racial hatred and prejudice. At the opposite corner, there are the Scared ones, who -in opposition to the former- are too scared to even bring up the topic of race, unless they couple it with endless platitudes about the brotherhood of mankind and the "Human Race".

I won't say that the two are equally bad. Of course, the Scared ones are not as bad, or as dangerous as the Scary ones. But they both trivialize the subject of biological differences between human populations, by making these out to be either too important (Scary), or completely unimportant (Scared).

It's about time there was a Third Way, for the scientific study of racial diversity. But of course, "racial science" has not been popular for half a century now, and it shows no signs of coming into vogue any time soon.

In the end, it all boils up to accepting that racial differences have nothing to do with the socio-cultural and constitutional notions of equality. Only when we accept this basic truth, will we be able to study racial variation without worrying about what such research might lead to." [emphasis mine]

Dienekes Pontikos

http://dienekes.awardspace.com/blog/archives/000005.html


I also advice you to read this much more recent (from today) statement of Razib Khan for some pondering:

"I can see where the individual is coming from, but I think more people should just come out say that evolution is just science, and has no deeper moral implications besides those which humans impute to it. No one cares about the ethical implications of physics, and fundamentally biology is no different. I realize that since it’s closer to the phenomenon of humanity than physics it won’t play out that way, and philosophers and theologians can do with it what they will. But really, if you look at a specific question like cross-racial mating among humans in 1910 and 2010 you learn a lot more about human values of scientists than about the science of human evolution and genetics." [all emphases mine except the last one, though I myself would still emphasize also the last one even if it was originally unemphasized]

Razib Khan

http://blogs.discovermagazine.com/gnxp/2010/08/daily-data-dump-tuesday-15/

terryt said...

"It shows us the paths our forebears took as they wandered the globe. It shows us the story of 'us'"

Quite. And the really interesting thing is that, as for most species and subspecies, the 'populations' at the margins are the most different. This difference can be very informative if we're prepared to admit the difference, and then consider it.

"I have blue eyes; a blued-eyed aborigine is more closely related to me"

The phenomenon of 'a blued-eyed aborigine' is only possible if the individual has some European ancestry. Therefore, as you say, he is likely to be more closely related to you than would other Aborigines be (assuming you're of European ancestry).

"we are all still, quite literally, family, sharing the same grandparent 6-10,000 years ago".

You'd have to add at least another zero to that of course.

Gregory76 said...

Terryt said:

“I'm fairly sure that Y-hap D is not characteristic of Negritos.”

It is strong among the Andamanese, and they seem to be the purest Negritoes phenotypically.
After reading physical descriptions of the various Negrito groups and Papuans some years ago, it seemed that one could arrange them in a scale as follows: the Andamanese, the Semang of Malaysia, the Aeta of the Philippines, the Tapiro of New Guinea, and the Papuans. As one moved along the scale, the people became hairier and taller, and had less uniformly woolly hair. This suggested to me back then there was increasing Australoid admixture as one moved along the scale, but now I think that the admixture cold be Caucasoid or both. In any case, I would expect more diversity in the genotype too.

Onur Dincer said...

"It shows us the paths our forebears took as they wandered the globe. It shows us the story of 'us'"

Quite. And the really interesting thing is that, as for most species and subspecies, the 'populations' at the margins are the most different. This difference can be very informative if we're prepared to admit the difference, and then consider it.

"I have blue eyes; a blued-eyed aborigine is more closely related to me"

The phenomenon of 'a blued-eyed aborigine' is only possible if the individual has some European ancestry. Therefore, as you say, he is likely to be more closely related to you than would other Aborigines be (assuming you're of European ancestry).

"we are all still, quite literally, family, sharing the same grandparent 6-10,000 years ago".

You'd have to add at least another zero to that of course.


I agree with everything you say here, Terry.

BTW, I want to make an addition to my last post:

The word race I use is synonymous with the word subspecies, so in no way it is unscientific. For instance, when I say Caucasoid race I always mean the Caucasoid subspecies, whose trinominal name should be something like Homo sapiens caucasoidus.

Andrew Lancaster said...

Onur writes:
"As to the seriousness of the word race, I don't agree with you on that matter; race is obviously a more ambiguous (because of its very different historical usages) thus less serious word than subspecies (one of the reasons why I always prefer subspecies)."

"As to the clarity issue, I don't think subspecies is as clear a concept as it seems from afar. Different species can be subjected to vastly different criteria in defining their subspecies, so there is some subjectivity here (that brings us back to the gold standard issue)."

That is a very watery reply I'm afraid. You just don't agree? You see the word race is ambiguous, but so is the word sub-species?

It is equivalent to arguing that you admit murder is wrong, but so is swearing, and the guy who got murdered swore. If you think it is not equivalent, you should have some better argument than you just don't agree.

The seriousness, not just ambiguity, of the word race is obvious. Try googling "race" if you want to see how the word is normally used.

Best Regards
Andrew

Andrew Lancaster said...

Onur quote Dienekes:
"In the end, it all boils up to accepting that racial differences have nothing to do with the socio-cultural and constitutional notions of equality. Only when we accept this basic truth, will we be able to study racial variation without worrying about what such research might lead to."

I fully understand the logic, which is effectively saying that we can give race a new meaning. The logic is logical, but that does not make the aim being proposed a good aim.

Someone should explain WHY it is a good thing to try to use this word mixed in with scientific discussion.

All the practical results I have seen have shown that even very clever people find it hard not to have their thinking twisted by the original meaning of this word.

Why do you need an exciting word? This is science. Use a boring word?

Best Regards
Andrew

Onur Dincer said...

This discussion has turned into a semantics discussion, so I won't go on to debate the meanings of words. I think the only thing we should focus on is whether modern humans have subspecies or not. I think they (=we) have, but I will not discuss whether modern human subspecies can also be called race (a scientifically useless term today) as that would have nothing to do with science but semantics.

terryt said...

"it seemed that one could arrange them in a scale as follows: the Andamanese, the Semang of Malaysia, the Aeta of the Philippines, the Tapiro of New Guinea, and the Papuans".

And the only one of these with any Y-hap D at all are Andamanese. And the connection between various Negritos may be more apparrent than real. They may each have developed fairly independently.

"but now I think that the admixture cold be Caucasoid or both".

I tend to agree with that. And I suspect there is a connection between the Ainu and Australian Aborigines, although it's unfashionable to see such a connection.

Andrew Lancaster said...

Onur: This discussion has turned into a semantics discussion, so I won't go on to debate the meanings of words.

That's odd. You were proposing a definition, a way of defining sub-species. How can you propose a definition without discussing semantics and the meanings of word?

Personally I think this is the problem: people say things without realizing all the assumptions and indeed assertions included in their own statements.

You proposed defining words in a way which you think would suit an unproven theory about what you think are the important dividing lines between the sub-populations of humanity. That's what it comes down to.

Back to reality and away from mere words though, I think humanity, like most species, does not have clear sub-species, just relatively fuzzy and complex sub-populations. The old traditional way of defining sub-species does not fit - especially not with a species which is so very mobile and ever-changing in its interactions.

...And if someone forced scientists to name ONE level of division which is most important within humanity it is most likely they would have to drawn it WITHIN Africa. Caucasoids, the group you are apparently interested in seeing as a primary division within humanity fit phylogenetically as a branch of a branch (at best) stemming out of an AFRICAN tree.

Onur Dincer said...

Back to reality and away from mere words though, I think humanity, like most species, does not have clear sub-species, just relatively fuzzy and complex sub-populations. The old traditional way of defining sub-species does not fit - especially not with a species which is so very mobile and ever-changing in its interactions.

I see in that newly genetically discovered fuzzyness (not just for human subspecies, but for all subspecies in general) a different trend. More and more, subspecies of all species are being defined as less clear-cut and more loosely coupled groups. So today the genetic/physical divisions of humanity in the highest scales more than ever deserve to be classified as subspecies. In other words, any meaningfully apparent genetic/physical division of humanity can be called subspecies today.

Andrew Lancaster said...

Onur: I see in that newly genetically discovered fuzzyness (not just for human subspecies, but for all subspecies in general) a different trend. More and more, subspecies of all species are being defined as less clear-cut and more loosely coupled groups.

More and more all fitting into one layer? That's the problem with the type of classification system we were discussing. And I believe it is absolutely clear that the more and more sub-populations are not going to all fit in one layer?

Onur Dincer said...

More and more all fitting into one layer? That's the problem with the type of classification system we were discussing. And I believe it is absolutely clear that the more and more sub-populations are not going to all fit in one layer?

As I told you, there is no limit in layering well into the village and family levels, so it is best to put a limit to layering in subspecies classifications just for convenience and clarity.

terryt said...

"I think humanity, like most species, does not have clear sub-species, just relatively fuzzy and complex sub-populations. The old traditional way of defining sub-species does not fit"

True. But the 'old traditional way of defining sub-species' is still used for all species apart from the human species. Why would that be so, I wonder?

"And I believe it is absolutely clear that the more and more sub-populations are not going to all fit in one layer?"

But if you're going to aggue that position you can hardly justify keeping the term 'species' as a meaningful concept either. Try to come up with a realistic definition of species.

Andrew Lancaster said...

Onur wrote:
As I told you, there is no limit in layering well into the village and family levels, so it is best to put a limit to layering in subspecies classifications just for convenience and clarity.

If the clarity is not in nature we should not pretend it is there.

Terryt wrote:

True. But the 'old traditional way of defining sub-species' is still used for all species apart from the human species. Why would that be so, I wonder?

I do not think you are right about this. Occasionally biologists say they can define a sub-species, i.e. when that corresponds to what they really see in nature, but when it comes to the study of non-human animals, I do not detect any of the same obsession with needing to find sub-species which we see amongst some people when it comes to humans.

Terryt wrote:
if you're going to aggue that position you can hardly justify keeping the term 'species' as a meaningful concept either. Try to come up with a realistic definition of species.

It is my repeated position that the definition of a species is indeed no longer critical to biology, since Darwin. A species is just a relatively very distinct inter-breeding sub-population. No single definition is possible or needed anymore.


However, note that this does not cause enormous problems because there are at least very often quite distinct and isolated populations at some level of relatedness, and those are what tend to get called species now. That is pretty much the rough working definition of species today: the lowest level of very clear isolation. Note that this does not always work perfectly well, but to repeat, the concept of a species is no longer pivotal and these definitions can and are considered as rough guides today, not basic and fixed properties of nature.

The difference with the concept of sub-species is that at this level the correspondence between old style classification system with reality becomes very weak, so that MOST real cases are problematic, and so while the concept of a species is today also properly understood as a rough guide, the roughness and nominalness of modern classification systems becomes more clear at levels lower than the species.

...And as a result any attempt to take classification system too seriously also become more jarring compared to reality at levels under the species.

Onur Dincer said...

Andrew, are you proposing that we should use subspecies classifications only for a very limited number of species, or that we should abolish the concept of subspecies entirely?

Andrew Lancaster said...

Hi Onur

I think most biologists already do what works, and sometimes they can define a relatively clear sub-species and sometimes they can not. When they can not however, this does NOT stop them studying the MORE COMPLEX inter-connections they find.

There seems to be a kind of implied argument that to give up on the word "race" when discussing humans means that we'd be banning discussion of human diversity?

Obviously that is not true. It is also not true in biology that scientists can not write about the diversity of populations that do not fit in the traditional classification system.

Onur Dincer said...

There seems to be a kind of implied argument that to give up on the word "race" when discussing humans means that we'd be banning discussion of human diversity?

I have never had such an argument. It would be silly to have such an argument. Whether there are human subspecies or not, human biological diversity (whether genetic or physical) is undeniable and will always be so (and, of course, scientists in biology can write about the biological diversity of populations that do not fit in the traditional classification system). What is really important is the degree of the human biological diversity and whether it is in a degree to allow scientists to make subspecies classifications for humans. I think it is and I think that I've said enough to explain why I think so.

terryt said...

"I do not detect any of the same obsession with needing to find sub-species which we see amongst some people when it comes to humans".

From that comment I can only conclude you have not read very widely on the subject. If I find time I'll link to as many examples of the phenomenon as you wish.

"No single definition is possible or needed anymore".

But the term is certainly used widely. For example the continuing arguments as to whether Neanderthals and modern humans were the same 'species'.

"What is really important is the degree of the human biological diversity and whether it is in a degree to allow scientists to make subspecies classifications for humans. I think it is and I think that I've said enough to explain why I think so".

I agree.

Onur Dincer said...

From that comment I can only conclude you have not read very widely on the subject. If I find time I'll link to as many examples of the phenomenon as you wish.

I see the exact opposite of what Andrew sees: there is a widespread tendency to ignore subspecies when it comes to humans.

But the term is certainly used widely. For example the continuing arguments as to whether Neanderthals and modern humans were the same 'species'.

If Neanderthals and modern humans could interbreed and produce fertile offspring of both sexes, then they were the same species. If they really interbred and produced fertile offspring of both sexes, then there can be no doubt that they were the same species. Just like modern humans, I think of Neanderthals as a biological group of relatively close subspecies. If Neanderthals and modern humans were the same species, they were probably two of the main branches (there were probably others as well) of the same species.

Onur Dincer said...

I have revised my definition of subspecies. My new definition of subspecies is so: lowermost genetic clusters separated from each other to a high degree and long enough to have clearly and easily identifiable consequences on the physical traits in a species.

So even if Neanderthals and modern humans were the same species, they wouldn't be subspecies but groups of subspecies.

Andrew Lancaster said...

1. Onur and Terry agree:
What is really important is the degree of the human biological diversity and whether it is in a degree to allow scientists to make subspecies classifications for humans. I think it is and I think that I've said enough to explain why I think so

But I think what was said was that actually there is too much complexity to fit in a one layer system? Was there another discussion?

2. terryt says that the term "species" is still widely used.

Yes, but I just said that no simple and clear definition is possible or necessary. I did not say that this stops the term being possible to use. Please note that I also explained why the problems this word has are nothing compared to the problems "sub-species" has.

3. Onur says:
I have revised my definition of subspecies. My new definition of subspecies is so: lowermost genetic clusters separated from each other to a high degree and long enough to have clearly and easily identifiable consequences on the physical traits in a species.

First of all it is inconsistent to be telling me that this term is already widely used in a correct and meaningful way, and then to also be saying that you have several proposals for your own definition.

Secondly, I agree with Dienekes about how interesting the example of the suburbs of Reykyavik is for this subject of discussion. Haven't you just given a definition which says that the sub-species of humanity are now either that big or smaller?

4. onur says
I see the exact opposite of what Andrew sees: there is a widespread tendency to ignore subspecies when it comes to humans.

Not on the internet though. :)

terryt said...

"Please note that I also explained why the problems this word has are nothing compared to the problems 'sub-species' has".

I'm sure it does have the same sort of pfoblems. The boundary between various species is often very ill-defined unless there has been extinction of intermediates.

Onur Dincer said...

First of all it is inconsistent to be telling me that this term is already widely used in a correct and meaningful way, and then to also be saying that you have several proposals for your own definition.

Actually my first definition was a little haphazard, as I hadn't considered all the possibilities, so I corrected it in the revised version.

Anyway, as I have explained enough of my thoughts, I won't further the same debate by answering your questions but will ask you a simple question. In your first post on this thread, regarding human racial classifications you said, "there are now more objective ways to divide up the species" . What is your proposal for dividing up the human species?

Andrew Lancaster said...

Onur asks:
In your first post on this thread, regarding human racial classifications you said, "there are now more objective ways to divide up the species" . What is your proposal for dividing up the human species?

Maybe a misunderstanding here. I do not propose any particular new classifications. My point is about the method and I just say that there are better ways of doing it than trying to fit all cases into a traditional system which only allows one level.

Examples of real scientists getting on with real biology in just this way are cited all over this website for example.

It effectively comes down to me saying that reality is messy and so we should not pretend otherwise, which is what biologists are already doing.

In contrast, as you yourself pointed out, published biologists "ignore subspecies".

I guess there is a reason?

They might ignore subspecies, but you can not say they are ignoring human diversity itself. OTOH, if they had to fit their results into subspecies then they really would have to ignore reality.

terryt said
The boundary between various species is often very ill-defined unless there has been extinction of intermediates.

Well there are always extinctions of intermediates if you go back far enough, but I agree with what you are saying. Defining species, as I said, is no longer possible or necessary to do in any simple or clear way.

But relatively speaking, the concept of a species is more often useable than the concept of a sub-species. In the case of sub-species, by definition, there is never a clear gap provided by the extinction of all potential intermediates.

Onur Dincer said...

In the case of sub-species, by definition, there is never a clear gap provided by the extinction of all potential intermediates.

Then what is the difference of human subspecies from subspecies of other species?

Andrew Lancaster said...

Onur, It is not only homo sapiens.

I think relatively few clearly defined species can usefully be broken into clearly defined sub-species in only one level.

It simply requires very specific circumstances for this to come about.

I guess that if sub-species means something useful these days it is in practice something like speciation which which is not yet finished.

Onur Dincer said...

clearly defined sub-species in only one level

I am skeptical that allowing more than one level of subspecies will make them more clearly defined. Also there is the problem of putting a limit to the number of subspecies levels, as theoretically there is no limit.

Andrew Lancaster said...

I am skeptical that allowing more than one level of subspecies will make them more clearly defined. Also there is the problem of putting a limit to the number of subspecies levels, as theoretically there is no limit.

Exactly. And what's more the populations move so quickly. One day two populations never interbreed, and then suddenly they do, but they stop breeding with another. So in other words the whole "tree" shape of relatedness breaks up completely in most cases that not either species, or on their way to be species.

But if you look at the real articles being published they do not need to assume a tree of categories, unless there is a tree, for example in the case of male lines on their own.

Discussion has gotten messy, and that is good! :)

Onur Dincer said...

Andrew,

Though I respect your scientific perfectionism in taxonomic classifications, I think your solution would only bring more confusion and messiness to the already existent ones. We need more regulization instead of more deregulization in this already messy and confusing chaos. So we shouldn't refrain from assigning subspecies status to certain sub-groups of species when they are distinct enough despite the imperfections in drawing lines between subspecies (some hybridization in intermediate zones should be seen normal). Also we should stick to the one-level rule, as it isn't less scientific than using more than one level in any way, as in both cases we are dealing with conventions, not absolute scientific facts. And lastly, it shouldn't be forgotten that subspecies are much more open to change than species as a result of being open to admixture and being generally a much less clearly defined and much more flexible category than species, so for instance, too much hybridization between two or more subspecies may eventually bring about their fusion under a new subspecies category, or some level of isolation of the members of a single subspecies may eventually bring about their split into two or more subspecies without any speciation.

terryt said...

"Though I respect your scientific perfectionism in taxonomic classifications, I think your solution would only bring more confusion and messiness to the already existent ones".

Perhaps we should just talk of 'clades'?

"too much hybridization between two or more subspecies may eventually bring about their fusion under a new subspecies category, or some level of isolation of the members of a single subspecies may eventually bring about their split into two or more subspecies without any speciation".

I think that sums up the situation quite well.

Onur Dincer said...
This comment has been removed by the author.
Onur Dincer said...

Perhaps we should just talk of 'clades'?

Clades don't help us in infra-species levels (including subspecies), as there is no barrier - at least theoretically - to admixture at those levels.

Andrew Lancaster said...

Onur:
Clades don't help us in infra-species levels (including subspecies), as there is no barrier - at least theoretically - to admixture at those levels.

Clades do help, in their messy way, to describe reality. Describing reality is the aim right?

What you see biologist do, as I am sure you must realize, is break the question up and talk about clades of mitochondrial lines, Y chromosomal lines, and of specific genes.

We need more regulization instead of more deregulization in this already messy and confusing chaos

Well there is the disagreement in a nutshell.

What I am saying on the other hand is that trying to describe nature in a way which fits old theories which are known to be wrong, and not the best models of reality, is what is creating confusion.

Biologists is not confused. It is has never been so clear and consistent and important as it is today.

It is people who try to fit real biological conclusions into pre-Darwinian models in order to "help" and "regulate" who are the confused people, and they are also the ones responsible for confusing the public.

Science should not be more regulated.

Onur Dincer said...

Clades do help, in their messy way, to describe reality. Describing reality is the aim right?

What you see biologist do, as I am sure you must realize, is break the question up and talk about clades of mitochondrial lines, Y chromosomal lines, and of specific genes.


You misunderstood me. I wasn't talking about clades of mitochondrial lines, Y chromosomal lines, specific genes or any other genomic piece, as I already acknowledge that they are helpful at all taxonomically recognized or unrecognized taxonomic levels, whether they be species level, ultra-species levels or infra-species levels. I was only talking about clades of taxonomic levels themselves whether they are taxonomically recognized or unrecognized. At the level of species you can talk about taxonomic clades due to the inter-species barriers to admixture or fertile offspring of both sexes. But at the level of subspecies or any other infra-species taxonomic level (I already said that the only taxonomically recognized infra-species taxonomic level is subspecies BTW) there is no such barrier, so admixture usually makes taxonomic clade-based classifications meaningless at infra-species taxonomic levels (whether they are taxonomically recognized or not).

Onur Dincer said...

Well there is the disagreement in a nutshell.

What I am saying on the other hand is that trying to describe nature in a way which fits old theories which are known to be wrong, and not the best models of reality, is what is creating confusion.

Biologists is not confused. It is has never been so clear and consistent and important as it is today.

It is people who try to fit real biological conclusions into pre-Darwinian models in order to "help" and "regulate" who are the confused people, and they are also the ones responsible for confusing the public.

Science should not be more regulated.


I think what we need is bringing current taxonomic classifications completely into consonance with Darwinism and evolution in general. To this end, we should make all species and ultra-species (=supra-species) taxonomic classifications taxonomic clade-based. Unfortunately, as I wrote in my previous post, there is usually no such possibility at infra-species taxonomic levels due to admixture, so at those levels we should turn to a more relaxed but again genetic-based method of classification.

Andrew Lancaster said...

Sounds about right: use what works basically.

By the way, it seems like we might have exhausted this, so if so, thanks for the interesting discussion guys.

Onur Dincer said...

Sounds about right: use what works basically.

Absolutely. Science isn't just about some absolute facts or truths, it is also an art of practical conventions. After all, which of the taxonomic categories have any meaning in the lives of living beings that are classified? Maybe the species and subspecies categories, though they are both far less important than small clans or family groups in the lives of animals.

thanks for the interesting discussion guys

You’re welcome. I'll always be ready for discussion unless there is a special condition.

terryt said...

"I was only talking about clades of taxonomic levels themselves whether they are taxonomically recognized or unrecognized".

I have seen some very useful use of clades in the classification of dabbling ducks. So it's not just applicable to individual genes. If you're interested I'll find and post, but I'm sure you're not as interested in ducks as I am.

"thanks for the interesting discussion guys".

Thanks to you too, and Onur.

Onur Dincer said...

If you're interested I'll find and post, but I'm sure you're not as interested in ducks as I am.

If it has relationship with the subject of our discussion, I would be interested.

terryt said...

Sorry for the delay. I was waiting until I had time to look it up. Here's my essay on the implications from some time ago (which includes the relevant reference):

http://humanevolutionontrial.blogspot.com/2009/06/human-evolution-on-trial-species.html

And here is the link itself:

http://elibrary.unm.edu/sora/Auk/v108n03/p0471-p0507.pdf

You'll see form the trees on pages 6, 7 and 9 how impossible it is to reduce the complexity to just teo levels. In fact towards the end he talks about 'superspecies', 'allospecies', 'subgenus', 'supergenus' and 'infragenus'.

Onur Dincer said...

Terry, I read your essay and found it well studied and documentented. As you might have noticed, I am a lumper when it comes to classifying species. Some splitters go too far as to classify human races (I define them as subspecies) as different species. I don't agree with them, as so much splitting would make the category of species, which is a category defined primarily based on absolute or practical impossibility of sustainable admixture between groups of living beings, meaningless.

Onur Dincer said...

documentented -> documented

terryt said...

Thanks Onur. Did you read the link on duck classification?

Onur Dincer said...

Did you read the link on duck classification?

Not much of it, it is apparently much more about >=species classifications than <=species classifications (the area I am more interested in).

But I read your essay from beginning to end.