Tuesday, April 29, 2008

Maternal Ancestry of Malayalis - II

I have discussed about some of the maternal lineages of Malayalis before. At that time I didn't have any idea about my subclade. Recently, I came across a Jewish matrilineage data from yet to be published Doron M Behar's new study. I found my near match in one of the samples. That probably is a Malayali Jew lineage.

I have tested for HVRI and HVRII sequences and I have exact match in the following sequence.

EF556193 Haplogroup M
A73G T152C A263G 309.1C 315.1C G316A T489C A750G A1438G A4769G
T5201C C7028T A8701G T8843C A8860G A9180G T9540C G9947A A10398G C10400T
G10685A T10873C G11719A G12007A C12705T A13105G C14766T T14783C G15043A A15196G G15301A A15326G G15355A T15862C T15968C C16223T T16263C T16519C C16527T

My HVR sequence:

HVR1 : 16223T,16263C,16519C,16527T
HVR2 : 73G,152C,263G,309.1C,315.1C,316A,489C

I most likely belong to mtDNA haplogroup M4* as the complete sequence has mutation at location 12007.

M Derived: (HVR1 sequence)
M50* : 223, 263, 519 (*)

*Update 1-may-2008:
Behar et al. paper is out. I belong to mtDNA M50* subclade of macro haplogroup M4'30.

Update 10-May-2008:
Ibra has pointed out that the designation M50 has already been used in the study "Austro-Asiatic Tribes of Northeast India Provide Hitherto Missing Genetic Link between South and Southeast Asia", by Reddy et al. (2007). I appears Behar et al. are oblivious of this paper as they declare the designation is a new one. We can expect M50 getting a new name.

Sunday, April 27, 2008

The Origins of Indians: Version 8.1

Coastal Migration Theory and I:
As I have mentioned before I am skeptical about Coastal Migration Theory. This theory proposes a set of Homo Sapiens after leaving Africa rapidly colonized South Asia, SE Asia and Australia moving rapidly along the coastal regions. Y-Haplogroup C is generally considered as the marker of that coastal migration.

One of the C's in India belong to subclade C5 is distributed across India. However, there are C* in South India that might be the legacy of coastal migration. I compared the tribes and castes that carry this lineage and I observed a strange pattern in its association with other haplogroups.

I make the following assumption before coming to my conclusions.

If a haplogroup is observed in more than one endogamous tribe/caste then its uncharacteristic presence in any group cannot be explained with genetic drift or bottleneck or founder effect.

That haplogroup in the below table is R1a1. As far as I can see C* is observed in conjunction with R1a1.

Tribe/Caste: Observed Haplogroups - Region
Koya Dora/Konda Reddy - F, H, O2a - South - East
Kurumba - F, H, R2, L1 - South - West
Kamar - F, H, L1, O2a - South - East
Yerava - F, H, C*, R1a1 - South - West
Koraga - H, R2, YAP - South - West
Toda - L1, J2a, R1a1, C* - South - West
Pallar - F, H, R1a1, C*, J2b- South - West
L1, R2

Based on the above data I propose the following pre-neolithic migrations.
1. Africa-Middle East-South Asia : F, H
2. ? -> East India : O2a
3. Central Asia - south west coastal India : R1a1, C* (one of the first migrations of R1a1)
4. ? -> south-west India: L1
5. ? -> south-west India: R2

I don't see the possibility of coastal migration with this scenario.

1. Cordaux et al. 2004
2. Sengupta et al. 2006

Correlation between major Indian Y-chromosome markers

Using 31 different population clusters from the Indian subcontinent I wanted to see if there are significant correlations (positive or negative) between the frequencies of any two haplogroups. Two positively correlated haplogroups may indicate a shared history and expansion, whereas a negative correlation may indicate opposite histories of those haplogroups. I considered the 5 major Indian haplogroups for the analysis, which are, R1A1, H, R2, J2 and L.


Dravidian castes:

  • -R1A1 increases then L increases

  • -R2 increases then H decreases

  • -R2 increase then L decreases

Indo-European castes (IE):

  • -R2 increases then R1A1 decreases

Dravidian + IE castes

  • R1A1 increases then H decreases
  • R2 increases then R1A1 decreases


Some members of H were replaced by a Holocene population consisting of R1A1 and L. Some time later (possibility during the Neolithic), members for R2 expanded over an R1A1/L/H background population.

The full analysis can be downloaded here:


Saturday, April 26, 2008

Guest Blogger: Ibra

I would like to welcome Ibra as a guest blogger. His core field is mathematics but has a varied interest in Anthropology and Linguistics. He is a Canadian of Caribbean Indian ancestry. He belongs to Y-Haplogroup R1a1 clan.

Saturday, April 19, 2008

The Origins of Indians: Version 9.2

Formation of Dravidian Society:

The castes in the erstwhile South Indian society might not have indulged in an exclusive occupation. However, I would consider the dominant stereotypical view of their occupation as the defining identifiers of their group. The discussion is completely restricted to caste and tribal population for obvious reasons.

The biggest groups in Dravidian societies are cultivators. They are either numerically the biggest or the second biggest in all the states. They may form monolithic group as in Karnataka and Kerala or diverse groups as in Tamil Nadu and Andhra Pradesh. Generally, chieftains came from this groups. They may form 15-30% of total population in any state.

Toddy tappers:
In Malayalam and Tulu region they are the biggest group. In Tamil Nadu and Andhra Pradesh they make up sizable part of the population. I believe in Kannada region their proportion in the population is not significant compared to other Dravidian regions(I may be wrong). Though originally made up of diverse groups now overall a monolithic group.

I don't think in any of the states they make up even 10% of total population. Mostly, a single caste but there are exceptions, I suppose.

I am restricting to only sea fishermen. May not form more than 10% in any state population. May or may not be single caste depending upon the region.

Except weavers numerically insignificant. Even weavers' distribution is uneven across the states and in no state form 10% of the population. Weavers are diverse groups but other artisans are generally a single caste group. Though it could be argued that all these smiths(goldsmith, blacksmith, carpenter etc...) were originally a single endogamous unit that turned into multiple castes in the course of time. Tanners are relatively big in Andhra region I suppose. However, in other regions their numbers are insignificant.

Agricultural Labourers:
Agricultural labourers may form 10-15% of the population.

Other Occupational Groups:
Among hundreds of other numerically insignificant occupational groups priests may form the biggest group at around 3-4% of population followed by merchants.

Tribes may form 3-4% the South Indian population. They are unevenly distributed with the highest concentration in Andhra Pradesh(6%) and the lowest in Kerala/Tamilnadu (1%).

Phase 1: Sea fishermen
In the first phase few people moved away from tribal life and became fishermen. It is possible that tribes and sea fishermen might have come to South India independently. The land fishermen were part of tribes. At the end of first phase we have;
1. nomadic land tribes 2. sea fishermen

Phase 2: Broken tribals Around : 3500 BCE -3000 BCE
This is sedentary civilizational phase for a section of population. A section of population took up farming. These are independent farmers cultivating small strip of land. A section within this took up toddy tapping because of unique South Indian geogrpahy. These are interdependent groups but cut off from organized tribal society. The civilized society of sedentary life did not form fully due to uncertainty of their occupations because of unique South Indian conditions. These are broken tribals without any organization. At the end of second phase we have;
1. nomadic land tribes 2. fishermen 3. farmers 4. toddy tappers

Phase 3: Feudal farmers Around : 2000 BCE- 1500 BCE
In this phase farmers from IVC started moving to South India. These highly organized advanced farmers found easy serfs in unorganized independent farmers. For protection these feudal farmers employed the service of nomadic but tightly knit land tribes. This phase also saw sedentary life of many land tribes. The contact with IVC farmers allowed them to appreciate the sedentary life of farming. This phase saw development of village organization with tribal chiefs as the heads.

This is a significant phase in South Indian society as the society experienced master-slave divisions. The serfs were mainly from independent farmers and to a lesser extent from toddy tappers.

I have discussed about the origin of the name Holeya previously. I need to make certain corrections here. In an earlier post I had predicted that erstwhile bonded labourers in Andhra region, Mala, might have got their name because of purity-pollution concepts. My argument was that if Holeya is from a Dravidian root meaning impure then 'Mala' may also follow the same logic. Then I found Stephen Fuchs' "At the bottom of Indian society" deriving the name in the same fashion.

But I could feel my logic for these groups might have worked in other way round too. I have noticed that only Brahmins in Kannada region derive the name from the root 'impure'. All other people derive it from the root 'field'(the one who works in a field). Again, Mala may have other roots. In Karnataka it is considered as a name of a tribe. Also, Tamil equivalent Pallar were supposedly called 'Mallar' (few centuries ago) showing their affinity with Mala in Andhra region. The place where 'Mala' meaning impure is attested is an old Telugu dictionary. That probably means identification of these two names with 'impure' was the handiwork of brahmins. Since other communities associate Holeya to field (hola in Kannada) I believe they were the earliest broken tribes who ventured into sedentary farming.

At the end of phase 3 we have;
1. feudal farmers from IVC and from land tribes(chiefs) 2. tribal soldiers 3. slave farmers 4. fishermen 5. toddy tappers 6. land tribes

Phase 4:
Entry of IE herders Around: 1000 BCE
This period saw IE herders moving to South India. These herders assimilated many local tribes. However, many local tribes independently took up herding.
At the end of this phase we have;
1. feudal farmers 2. tribal soldiers 3. slave farmers 4. fishermen 5. toddy tappers 6. herders 7. land tribes

Phase 5: Entry IE guilds Around: 500 BCE - 0
By this time rudimentary civilization was taking shape in South India. This allowed entry of craftsmen and merchants to this region. Many tribals joined them or started their own guilds. The society until now only class based added one more divisive philosophy, the caste system, to it.

End of this period also saw entry of priests from eastern regions.
At the end of this phase we have;
1. feudal farmers 2. tribal soldiers 3. slave farmers 4. fishermen 5. toddy tappers 6. herders 7. craftsmen 8. priests 9. merchants 10. other occupational groups

Phase 6: 0 - 1900 CE
This period saw reorganizations of various groups according to caste rules. The tribal soldiers became part of various groups. The ruling classes saw north-Indian and non-Indian additions to their ranks.

At the end of this phase we have;
1. feudal farmers 2. independent farmers 2. slave farmers 3. fishermen 4. toddy tappers 5. herders 6. craftsmen 7. priests 8. merchants 9. other occupational castes

Sunday, April 13, 2008

The Origins of Indians: Version 9.1

Dravidian languages:

The legacy of pseudo-scientific anthropometric studies of the past is still deep rooted in India (at least with people who are interested in history, anthropology...others don't count anyway). Many physical anthropologists had wild and free run classifying Indians. I strongly feel the whole field is absurd.

First of all, the stature and skull shape of Homo Sapiens have never been constant. It changes because of changing food habits and changing geographical conditions. Also, if you compare the present haplogroup clans with those 'racial groups' you can clearly appreciate the futility of whole exercise.

The popular understanding is simple. India was initially inhabited by short and dark Australoid. Most of them were massacred by more robustly built also dark Mediterranean Dravidians. These Dravidians were later subjugated by light skinned Aryans from Europe.

I don't want to go on discussing about these self-sustaining myths (or science at a particular point of time). I would like to make few observations for my group, Dravidians.

If you observe Central India then it must be evident that the Dravidian tribes have in fact recognized the Austro-Asiatic tribes' religious supremacy in the regions where those tribes are dominant. The priests who officiate rituals of Dravidian tribes are from Austro-Asiatic tribes. That probably shows higher level of society belongs to Austro-Asiatics than Dravidians.

Second, there is no Dravidian marker if you go by genetic studies. In fact, there is no uniform distribution of dominant male genetic lineages among three biggest Dravidian groups like Telugus, Tamils and Kannadigas. J2b, H , R2 and L all dominate among different groups. Compare this wih Indo-Aryan speaking population and Austro-Asiatic population. There you can clearly associate R1a1 with IE and O2a with Austro-Asiatics. But not in the case of Dravidians.

This is the precise reason I believe the Dravidian languages are part of South India since the beginning. In north India you can find isolated languages like, Nahali, Burushaski and probably now extinct Bhil language. But not so in the case of South India. Therefore, in my opinion, Dravidian languages have been sustained by Dravidian females through multiple male migrations. Any words related to West Asian languages may be just loans from later arrived males.

Ravi Mundkur blogs about a new study that talks about distinct Koraga(a tribe in coastal Karnataka region) language. But by my understanding it is impossible that south Indian tribes have ever lost their languages. It should be noted here that mainstream Dravidans(along with majority of Indo-Aryans in the north) have maternal lineages that they share with these tribes and not only that nearly 30% male lineages(Haplogroup H) too.

If Dravidian languages have been imposed by the newly arrived males (probably J2b who make up 15-20% of the population and who qualify as "Mediterraneans") then I find it difficult to explain distinct Dravidian language branches among central Indian tribes. Some of these tribes(also J2b is hardly observed among Dravidian tribes) are so remote(Bison Horn Maria) and until recent times were totally cut off from mainstream.

I would propose Dravidian languages are matrilineal.

Saturday, April 12, 2008

The origins of Indians: Version 6.2.4

Varnas in RgVeda book 10:
I have come across a saying like 'the caste system only appears in RgVeda Book 10'. I don't know what they really mean by that. Does that mean Varnasrama was a late development and was not part of initial Indian Vedic society? Does that mean book 10 shows onset of entry of Vedic people into the subcontinent? Anyway, I find this is one of the frogowellian arguments that people make without proper data and analysis.

I have already discussed Norse mythology and classes of PIE society.

I'll copy the Wikipedia article that deals with Norse classes equivalent to Varnasrama.
The Rígsthula tells how Ríg happened upon a farm-hut which was owned by Ái 'great-grandfather' and Edda 'great-grandmother'. They offered Ríg shelter and poor, rough food for a meal. That night Ríg slept between the pair in their bed and then departed. Nine months later Edda gave birth to a son who was svartan (dark/black in color). They named him Thræl (thrall, serf, slave). Thræl grew up strong but ugly. He married a woman named Thír (slave girl, bondswoman) and they had twelve sons and nine daughters with names mostly suggesting ugliness and squatness. They became the race of serfs.

Travelling further, Ríg came across a nice house where lived a farmer/craftsman, Afi "grandfather" with his wife Amma "grandmother". The food was good and this couple also let Ríg sleep between them. Nine months later, a son, Karl (churl, freeman) was born whose face and hair was red. Karl married a woman named Snör (daughter-in-law) and they had twelve sons and ten daughters with names mostly suggesting a neat appearance or being of good quality. One of the names is smiðr (smith). These become the ancestors of the lesser farmers and herdsmen.

Travelling further, Ríg came to a mansion inhabited by Faðir (Father) and Móðir (Mother). They gave him excellent food served splendidly and, nine months later, Móðir gave birth to a beautiful baby named Jarl (earl, noble) whose hair was blond and who was bleikr (bright white in color). When Jarl grew up and began to handle weapons and to use hawks, hounds, and horses, Ríg reappeared, claimed Jarl as his son, gave Jarl his own name of Ríg, made him his heir, taught him runes, and advised him to seek lordship.

It should be noted here that though stories documented during medieval period in northern Europe they nevertheless had a long oral tradition. Therefore, the comparison between Norse mythology and RgVeda hymns of greater antiquity is still valid.

What should catch our eye is the primitiveness of Norse myth about development of three classes compared to rather abstract nature of four Varnas in RgVeda.

From this point of view I would argue it is inconsequential that Varnas appear only in RgVeda book 10.

I also argue the primitiveness that prevailed in Norse society was mainly because the literate priestly class was a West Asian phenomenon. Though it is possible that bards of IE society merged with Aryan priests in Iranian lands.

If anything its elevation from folk story to sacred hymn shows meeting of IE classes (not endogamous) and purity-pollution rules of West Asia.

The concept of single great man giving rise to multiple clans/division/clans probably was part of PIE folklore. In Kerala, the legend of twelve children of Parayi (from erstwhile outcast) and Vararuchi(a brahmin) might have been built upon this concept. The twelve children in this Kerala story and in Norse mythology may be a coincidence or remembrance of the old story. But twelve children(most probably representation of people from different walks of life who became well known in Kerala) becoming part of twelve different castes(one of them is muslim) resembling Norse myth is hard to ignore.

Monday, April 07, 2008

Absurdities of Caste Genetic Studies - Notes

Another absurdity that I come across routinely in the caste genetic studies is arbitrary distribution of upper/middle and lower caste categories. This random assignment of caste position adversely influences the haplogroup distribution and messes up our effort to understand the migratory patterns of various clans.

Consider the paper, "High Resolution Phylogeographic Map of Y-Chromosomes
Reveal the Genetic Signatures of Pleistocene Origin of Indian Populations", by Trivedi et al.(2007). I can summarize their method of caste categorization thus;

1. The states where Brahmins are not part of the samples, upper caste position is taken by middle castes.
eg. Kerala
-> Namboodiri Brahmins are not part of the study.

2. The states where Dalits are not part of the samples, lower caste position is taken by some of the middle castes.
eg. Andhra Pradesh
-> Mala and Madiga groups are not part of the study.

The situation is similar for Karnataka samples where Gowdas(equal to Nair/Reddy/Kammas who occupy middle caste position in the social scale) have been grouped under lower castes. Ironically, if we check Tamil Nadu samples Gowndar (Tamil equivalent of Kannada Gowda) are part of upper castes (as Tamil Brahmins are not part of the

With such a strange assignment of caste groups authors go on and determine distribution of various haplogroups in the pooled upper/middle and lower caste groups!

Now let's see about the consistency between two studies. In Thanseem et al.(2006) study, Kammas are part of upper castes and in Trivedi et al.(2007) study they are part of lower castes! Are any of the Indian genetic studies peer reviewed?

Once in a while Indian geneticists working with European counter parts take it upon themselves to teach a thing or two about Indian castes to their peers. So, we have Choubey et al.(2006) study with an information box on Scheduled castes and Tribes;

Scheduled Castes and Scheduled Tribes are communities in India that are given a special status by the Constitution of India. Scheduled castes were considered as ‘Sudra’ in the classical ‘Chaturvarna’ caste system while Scheduled tribes were considered ‘outcastes’ and were not a part of the Indian caste system.

There is so much cynicism in that para that I dread to touch it. I must only point out that tribes as long as they didn't join the mainstream their ritual purity was higher than that of communities that are classified now as Scheduled castes (Probably Choubey needs to read Ambedkar's essays first).

Buddhism and Jainism in South India - 6

The religious identity as we define today is not a good model to understand the South Indian society 1000-1500 years ago. I am more and more convinced of the fact that there did not exist Buddhist, Jain and Vedik (pre-cursor to present day Hindu or caste) identities in the context of South India. From historical times the caste identity has been fossilized to such an extent in south India it remained the fundamental identity though the person might have prescribed to Buddhist or Jain philosophies. Of course, this is true only in the case of brahmins, rulers and merchants. Identities of the rest of the population is rather irrelevant. Also, there have been multiple reorganizations, in terms of exogamy or endogamy, of caste identities of other population(cultivators, toddy tappers,artisans etc...) throughout the history.

Though this model is invariably true in the case of Buddhism, there are exceptions in the case of Jainism. It appears Jainism was able to develop a religious identity in South India which to a certain extent was able to obscure caste identity. A small native Jain population is still found in Karnataka and Tamil Nadu.

An exclusive Jain or Buddhist identity would require an existing strong opposite collective view. But in this case, I would think only Brahmins offered that opposing view initially*. Other castes identities did not have a literate dogma. In such a situation individual caste identities acted as present day religious identities. Some of the dominant individual castes followed Sivaism, Visnuism, Jinaism or Buddhism.

Even before the implementation of the caste system by the royal sanction of the purity-pollution rules introduced by brahmins, there could have been a proto-type of caste identity because of the migration of guilds from East India to South India. The replication of this proto-type by the tribes moving to mainstream cannot be ruled out. The varNAzrama, after the development of guilds, became a self-sustaining phenomenon, which is called caste system, throughout India. My model for development of the caste system in South India;

First phase: Formation of caste groups because of the migrations of guilds from East India
Second phase: Ritual sanction by the purity-pollution rules introduced by brahmins with the help of rulers

I would argue the entry of Buddhism and Jainism must have happened only after the first phase.

I don't deny that there were Bouddhas and Jainas in South India. It is even possible that many of them from non-brahmin, non-merchant or non-ruling families. But they were invariably monks hence most likely did not leave any progeny to carry their new identity. Also, the caste identity of a family that sent him or her did not change because of this.

*That probably means Brahmin identity in those times was akin to Buddhist or Jain identity not just a caste identity of present times.

Daily Life In Ancient India (from 200 BC to 700 AD), Author: Jeannine Auboyer
(Read 100 pages till now).

Saturday, April 05, 2008

Origin of my name

My name has been a source of humiliation (natha in its corrupted form ...nAta ... means foul smell in Kannada), source of mistaken sexual identity (manju is invariably a female name outside my native state Karnataka), source of uncomfortable closeness(I have observed some nick names, like manju, are difficult to take for stiff upper lip kind of people...including me->yes, that is preemptive strategy of image development). Nevertheless, there is no denying that it has been a source of much amusement for me and for people around me(along with my alliterator second name-> should it be the first name?). But my greatest amusement comes from the etymology of this word.

Sivaite, Buddhist or Ghostistic?
The first etymology that I heard was, the guardian (nAtha) of ice (maMju) mountains (Himalayalas, or Mount Kailasa) or Siva. The name became popular because of a famous temple in my region. Until very recent I would explain this etymology to people. It must be noted here that though the name sounds so much north Indian only in my region Siva is called Manjunatha. I found it bit strange. But as my understanding grew I found this name in fact has a Buddhist beginning.

The first Manjunatha temple was built at a place called Kadri in Mangalore. However, as I have already discussed before this was in fact a Buddhist monastery which was converted to Saiva temple. This event deals with assimilation of Tantrik Buddhism with Tantrik Sivaism. One of the Buddhist deities in that monastery was Manjusri. In my opinion, since Manjusri is an important Bodhisattva for Tantrik Buddhism the name Manjunatha might have been conceived from Manjusri. This I felt is the most plausible scenario. This was until the blogger Ravi Mundkur brought one more angle to the name.

In his series of posts on Kadri and Natha cult he proposes the native beginning for this name from ghostistic tradition. Ravi has developed his view based on legends surrounding Natha cult figure Matsyendranatha. Manju was supposedly one of his sons and a spirit worship was inititated after his death. This spirit Manju became deity Manju Natha in Saiva tradition. I am not yet convinced of this angle as Buddhist deities getting Saiva make over has been observed in the history. And legends can not be authentic accounts.

From Wikipedia on Avalokitesvara;
according to recent research, the original form was Avalokitasvara with the ending svara ("sound, noise"), which means "sound perceiver", literally "he who has perceived sound" (i.e. perceiver of the cries of the sentient beings who need his help). This is the exact equivalent of the Chinese translation Guan Yin. This name was later supplanted by the form containing the ending -īśvara, which does not occur in Sanskrit before the seventh century. The original form Avalokitasvara already appears in Sanskrit fragments of the fifth century.[5]

The original meaning of the name fits the Buddhist understanding of the role of a bodhisattva. The reinterpretation presenting him as an īśvara shows a strong influence of Shaivism, as the term īśvara was usually connected to the Hindu notion of a creator god and ruler of the world. Attributes of such a god were transmitted to the bodhisattva

When Avalokitasvara can become Avalokitesvara, Manjusri turning into Manjunatha isn't all that strange. But what came out of Ravi's posts is my ignorance about the word 'manju'.

Until Ravi mentioned about a Sanskrit word "manju" meaning beautiful, charming, pleasant etc..., I believed the word manju in Manjunatha is in fact a Sanskrit word meaning ice, coldness, dew, fog etc...! It turns out that manju that I know is in fact a Dravidian word. I didn't know. Had I known before I would have grown into Indic manju than Dravidian manju.

Therefore, the common derivation of Manjunatha for Siva is a mix of Dravidian word manju with Indic word natha. However, ice formations are not observed in South India therefore 'dew' manju might have been stretched to incorporate 'ice' in recent times.

May be Buddhist Manjusri and Kannada word manju together might have given the idea of Manjunatha.

Thursday, April 03, 2008

The first tribe of Eurasia?

Cordaux et al. 2004

Around 90% of Eurasian male lineages are descendants of Y-Haplogroup F. Y-Haplogroup-CF line is supposedly migrated out of Africa and moved to India. If at all India is the original homeland that spanned rest of Eurasian clans then I believe the Yerava tribe could qualify for being the first tribe of Eurasian clans with such high frequencies F and C. The Yerava tribe is mostly found in southern Karnataka and northern Kerala belt and speak a language called Ravula.

However, if it turns out that all C's in India are in fact C5's then they might have migrated as part of multiple Indo-Turk migrations from Central Asia along with R1a1. Probably C5 and R1a1 were part of the first of Indo-Turk migrations to South Asia.