1

Understanding Reich, David et al, 2009

by P. Priyadarshi

Some findings and their explanations from the research paper: Reich, D. et al 2009,

Reconstructing Indian population history, Nature 461:489-494. link:

http://www.genome.duke.edu/seminars/journal-club/documents/nature08365.pdf

“Paradigms, especially old ones, die harder than Bruce Willis.” (James

Adovasio 1999)1

Reich et al (2009) was a landmark work for human history. Yet

unfortunately very few people gave it an attentive reading.

Moreover, there seems to be a wide understanding-gap between

the statisticians who worked up the research, and the biologists

who wrote the script, and perhaps could not grasp many of the

hardcore statistical results (vide infra). This may be called inter-

disciplinary comprehension gap.

1 Quoted by Oppenheimer, S. J., 2010, “Comment” on Donohue, M. and Denham, T.,

Farming and Language in Island Southeast Asia, Current Anthropology 2010, 51(2): 243-

244.

2

Indian castes are only 600 years old:

One of the important findings of the article is that the Indian

caste system is not very old. Reich et al examined six Indo-

European and Dravidian speaking groups (caste or tribe), and

found that their founder event dated back to about thirty

generations back (p. 1 pdf). Each generation is considered about

18 to 20 years. Hence these castes were founded about 600

years back. This figure is consistent with findings from history

that endogamous caste system started only during the late

medieval period in India (Basham:148; “It was only in late

medieval times that it was finally recognized that exogamy and

sharing meals with members of other classes were quite

impossible for respectable people.”)

However the study noted that the Vaishya group had a founder

event 100 generations back, i.e. about 2,000 years back. It

should be understood that Vaishya is not an endogamous caste,

but is a group of about a thousand or more trading groups, each

unit group being an independent endogamous caste. Hence this

result about the Vaishyas does not tell us anything about the

date of origin of the caste system. If more people from Vaishya

groups were included, e.g. a few from Tamil Nadu, and a few

others from Bengal, their founding event might go back to

20,000 or 30,000 years back.

However, from the present result we may assume that the figure

2,000 years reflects the time of founding of trading and

commerce guilds, which the term vaishya connotes in modern

times. Max Weber and many other authors think that such guilds

3

came into existence once the emancipatory religions (like

Buddhism and Jainism) became strong enough in India (Max

Weber:399; “'Guilds' of merchants, and of traders figuring as

merchants by selling their own produce, as well as 'craft-guilds,'

existed in India during the period of the development of cities

and especially during the period in which the great salvation

religions originated. As we shall see, the salvation religions and

the guilds were related.”)

Indian Population is Indigenous:

Reich et al 2009 provided a robust support to Metspalu et al

(2004) and Sahoo et al (2006). However because of inattentive,

casual and cursory readings of their article by some evolutionary

biologists, it has been claimed that Reich et al 2009 contradict

some of the findings of Metspalu et al and Sahoo et al. (e.g.

Romero et al 2011:21). Thus Romero et al (2011) wrote about

Reich et al’s findings:

(a) “A recent study has argued for a sizeable contribution

from western Eurasia to the ancestral north Indian gene

pool (Reich et al. 2009), although it has not provided a

date or a precise origin location for this event, and the

observation is compatible with low continuous gene flow

over a long time period.” (Robero et al:20-21 pdf).

This reading is atrocious. Reich et al do not even suggest

any gene-flow from western Eurasia to South Asia. See

discussion below.

4

(b) Another such reading of Reich et al is: “However, there is

a paucity of mtDNA and Y chromosome lineages in the

Indian gene pool that have been assigned a Central Asian

origin (Metspalu et al. 2004; Sahoo et al. 2006),

suggesting that the west Eurasian genetic contribution

identified by Reich et al (2009) principally reflects gene

flow from Iran and the Middle East.” (Robero et al:21

pdf).

This conclusion too is atrocious. This statement (by

Robero et al) implies presence of west Eurasian genes in

Iran and Middle East, and also their flow into India to the

extent that they are detectable today in India: a

conjecture (see Renfrew 1989; Bellwood 2005 etc),

which has been explicitly ruled out by Sahoo et al

(2006:847), and many other authors.

Sahoo et al had actually written the following words:

“The perennial concept of people, language, and

agriculture arriving to India together through the

northwest corridor does not hold up to close scrutiny.

Recent claims for a linkage of haplogroups J2, L, R1a,

and R2 with a contemporaneous origin for the majority of

the Indian castes’ paternal lineages from outside the

subcontinent are rejected, although our findings do

support a local origin of haplogroups F* and H.” (p. 847).

They also rule out arrivals from Southwest Asia because

West Asian haplogroups (like Y-Hg G) are not found in

India.

5

Kivisild’s findings (2003:322, column 2) too had shown that

humans could not have arrived from West Asia into India

because of lack of West Asian Y-hgs E, G, I, J* and J2f. Kivisild et

al wrote, “When compared with European and Middle Eastern

populations (Semino et al. 2000), Indians (i) share with them

clades J2 and M173 derived sister groups R1b and R1a, the latter

of which is particularly frequent in India; and (ii) lack or show a

marginal frequency of clades E, G, I, J*, and J2f.” Thus Romero

et al have invented something and labelled it to have been

written by Reich et al, and then arrived at a compromise that

there was a migration of West Eurasian genes from Iran/ Middle

East to north India.

Thus any discord between Sahoo et al and Reich et al, or that

between Kivisild et al and Reich et al etc is in fact nonexistent,

and is a product of misreading by a fertile brain. Incidentally,

Reich is a co-author of Robero et al! Hence it is important that

Reich et al must be well explained to readers. An attempt to

which is being made below.

Understanding Reich et al 2009

The things we just saw have happened only because “Paradigms,

especially old ones, die harder than Bruce Willis” (James

Adovasio 1999; quoted by Oppenheimer 2010). It is therefore

desirable that ‘Reich et al 2009’ should be explained in more

simple English once more so that casual readers may not get

confused.

6

Reich et al never said that there was any flow of west Eurasian

(European) gene into South Asia. To prevent such misleading

readings of their article, they explicitly mentioned: “These results

do not mean that the Indian groups descend from mixtures of

European and Austro-Asiatic speakers, but only that they derive

from at least two different groups that are (distantly) related to

CEU and Santhal.” (p. 4 pdf, col 1).

Relatedness does not at all imply gene flow from CEU (Europe),

however it means a common ancestor for both ANI and CEU.

This common ancestor had existed after split from the African

main trunk, and after the Ancestral South Indians too had been

split. Reich et al depict this situation by means of a figure (Fig. 4

of Reich et al). The figure can be seen at the link below:

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2842210/figure/F4/ or,

http://www.nature.com/nature/journal/v461/n7263/fig_tab/nature08365_F4.html

or, see Fig 4, page 4 of pdf of the article.

Figures for family tree can be constructed from the text itself

actually. Reich et al write, “the tree (YRI,((CEU,ANI),(ASI,

Onge))) provides an excellent fit to the data”. This gives us a

family tree for five populations, viz. 1. YRI (West African), 2. CEU

(Modern Europeans), 3. ANI (Ancestral North Indians), 4. ASI

(Ancestral South Indians) and 5. Onge (a Modern Andamanese

tribe)—if drawn graphically would be like this:

7

Original Human Population in Africa

Africa(YRI) contemporary

Andaman (Onge, contemporary)

Europe(CEU)

ANI Hypothetical, remote past Populations of North India and South India respectively

ASI

The top inset from Stephen Oppenheimer’s The Journey of Mankind, Bradshaw

Foundation has been added as a ready-reminder to African-Eurasian split and

relationship. Source: http://www.bradshawfoundation.com/journey/

8

This is clarified further in the text of Reich et al: “the fact that

different Indian groups have inherited different proportions of

ancestry from the ‘Ancestral North Indians’ (ANI) who are related

to western Eurasians, and the ‘Ancestral South Indians’ (ASI).” It

does not say that different Indian groups have inherited different

proportions of ancestry from the CEU and … However, this latter

is the impression, Romelo et al have derived from Reich’s article!

Here we note that, in spite of the fact that ANI were related to

Western Europeans in remote antiquity, the current Indian

populations (both north and south Indian) derive from admixture

of ANI and ASI, and not by admixture from any third population.

Contrary to this clear finding, Romero et al read from this

paragraph, “the observation is compatible with low continuous

gene flow over a long time period” (p. 21 pdf, last line) [into

South Asia from Europe or West Asia: square bracket added.]

This is atrocious. It may be noted that neither Reich et al, nor

any other worker has ever claimed “low continuous gene-flow

into Indian sub-continent” from any outside source/place.

Now we can come back to the family tree.

The family-tree (Fig 4 of Reich) is good. Yet it does not take into

account the dates. It should be noted that ANI and ASI are past

populations, and hence need to be placed nearer the source than

the YRI, CEU and Onge populations, which are contemporary

living populations. Therefore the picture needs to be corrected, to

adjust placing for time. Then the picture would look like this:

9

Original Population

CEUEuropeModern

However, from Metspalu (2004) and many other studies, we know, that the

stem of the non-African limb of the graphic was located in South Asia. Hence

the real picture would be like this:

10

Original Population

CEUEurope

ASI

However, this model ignores the fact that many other populations than the CEU

and Ongan had been derived from ANI and ASI. Hence the picture needs to be

modified further to accommodate modern north Indian and South Indian populations:

11

Original Population

Ancestral African Ancestral IndianFirst Out of Africa arrivalinto India/ or the relic Indian Population after Toba eruption

YRIWest Africaetc

Ancestral NorthIndian; in northIndia

Ancestral SouthIndian; in south India

CEU Modern

Eur

ModernSouth

Indian Population

ONGANAndamanese Tribes

This picture is consistent with both the important conclusions of the Reich’s

article:

12

1. “These results do not mean that the Indian groups descend from mixtures

of European and Austro-Asiatic speakers, but only that they derive from

at least two different groups that are (distantly) related to CEU and

Santhal.” (Riech et al, p. 4, column 1). This rules out any admixture of

Europeans and Austro-Asiatics.

2. “Applying our model-fitting procedure, we find that the tree

(YRI,((CEU,ANI),(ASI, Onge))) provides an excellent fit to the data from Indian groups.”the tree (YRI,((CEU,ANI),(ASI, Onge))) provides

an excellent fit to the data” (Ibid., p. 4, col 1).

Hence a full synthetic figure would be like this:

Original Human Population

in Africa

Africa

(YRI)

contemporary

Andaman

(Onge)

contemporary

Europe

(CEU)

ASI

(Located in

South India

long back)

Modern North and

South Indian populations

as an admixture

of ASI and ANI

Pathan, Vaishya,

Meghawal, Bhil,

Contemporary.

Location:

India

Location:

India Location:

India /East Iran

13

Problem of Adyegi:

Reich et al note (p. 4 pdf): “the ANI and CEU form a clade, and

further analysis shows that the Adygei, a Caucasian group, are

an out-group… Indian and European groups speak Indo-

European languages, whereas the Adygei speak a Northwest

Caucasian language.” This issue can be resolved by assuming

that the Adyegi broke off from north Indian population just

before split between ANI/CEU. This split as well as the ANI/CEU

split must have occurred before LGM.

It is also helpful to remember that the modern languages of

Europe are products of language change from elite dominance by

incoming farmers. There has been population replacements too,

either coming through northeastern corridor of Europe, or

through the Balkans-West Asia route. Hence the “ANI-CEU” clade

may not have spoken the Proto-Indo-European, rather they

possibly spoke the Proto-Nostratic. Ancestral Europeans spoke

Semitic and other languages, but not the Indo-European (Baldi

and Page, 2003). Proto-Indo-European evolved within the ANI

population after split from CEU. ANI-ASI admixtures took place

after LGM. If we add these few understandings to the findings of

Reich et al 2009, it becomes much more comprehensible and fits

other stories well.

At the end it is useful to add that often statements by great

authors are best not forgotten. Western Eurasia was formed of

populations migrating from Asia and Africa. Hence if plotted,

European genes cannot form any cline towards Asia or Africa,

while African and Asian genes will always show clinal pattern of

expansion into Europe.

14

The seminal words of Cavalli-Sforza remain valid even today,

“...both Africans and Asians contributed to the settlement of

Europe, which began about 40,000 years ago. It seems very

reasonable to assume that both continents nearest to Europe

contributed to its settlement, even if perhaps at different times

and maybe repeatedly. It is reassuring that the analysis of other

markers also consistently gives the same results in this case.

Moreover, a specific evolutionary model tested, i.e., that Europe

is formed by contributions from Asia and Africa, fits the distance

matrix perfectly (6). In this simplified model, the migrations

postulated to have populated Europe are estimated to have

occurred at an early date (30,000 years ago), but it is impossible

to distinguish, on the basis of these data, this model from that of

several migrations at different times. The overall contributions

from Asia and Africa were estimated to be around two-thirds and

one-third, respectively". (Cavalli-Sforza 1997:7720).

This explains Reich et al’s finding that “The most remarkable

feature of the PCA is a gradient of proximity to western Eurasians (Supplementary Fig. 5), an analogous PCA in Europeans did not produce a gradient of proximity to India” (p. 3 pdf) This finding too is consistent with our knowledge from other sources. European population has been derived from north African, West Asian, Central Asian, and through the latter the Indian source. Caucasian and Palaeolithic genes too have contributed to making of modern Europe. Hence its clines are distorted, and a gradient of proximity to India cannot be produced. However, from India, genes spread like advancing waves, hence clinal rings could be formed. However if we are imaginative enough to visualize the wave- like picture of a clinal wave, we can appreciate that only the source

15

population can have the waves of expansion as a cline. This makes obvious that Europe was not the source of any population in the world, but rather a recipient of DNAs from other parts of the world.

References:

Baldi, P. and Page, B.R., Europa Vasconica-Europa Semitica Theo Vennemann, Gen.

Nierfeld, in: Patrizia Noel Aziz Hanna (Ed.), Trends in Linguistics, Studies and Monographs

138, Mouton de Gruyter, Berlin, 2003, pp. xxii + 977

Basham, A.L., The wonder that was India, Third revised ed (1967), thirty-fifth impression,

1999, Rupa and Co., Bombay.

Cavalli-Sforza, L.L., 1997, Genes, Peoples, and Languages, PNAS, 94(15):7719-7724.

Kivisild, T. et al, The genetic heritage of the earliest settlers persists both in Indian tribal

and caste populations, Am J Hum Genet, 2003 Feb, 72 (2) : 313-32.

16

Metspalu, M. et al, 2004, Most of the extant mtDNA boundaries in South and Southwest

Asia were likely shaped during the initial settlement of Eurasia by anatomically modern

humans, BMC Genetics 2004, 5: 26doi:1186/1471-2156-5-26

Reich, D. et al, 2009, Reconstructing Indian population history, Nature 461:489-494.

Romero, I.G. et al, 2011, Herders of Indian and European cattle share their predominant

allele for lactase persistence, Molecular Biology and Evolution, Advance Access Published

August 11, 2011.

Sahoo, Sanghmitra et al; A prehistory of Indian Y chromosomes: Evaluating demic diffusion

scenarios, PNAS 2006, 103(4): 843-848.

Weber, Max, Gerth, H. H. and Turner, B. S., “India: The Brahman and the castes”, in From

Max Weber: Essays in Sociology, Routledge, 1991. First published in 1921 in German as

Part 3, Chapter 4 of Wirtschaft und Gesellschaft. English translation by Girth, H. H. and

Mills, C. W., as “Class, Status, Party. Pages 180–195 in From Max Weber: Essays in

Sociology, Oxford Univ. Press, New York, 1941, 1958.