Indian Territorial Divisions. their Numerical Suffixes, and Populations

civfanatic

Ad Honorem
Oct 2012
3,318
Des Moines, Iowa
Many inscriptions from early medieval and medieval India record territorial divisions with numerical suffixes, such as "Gangavadi-96000", "Nolambavadi-32000", "Banavasi-12000", "Vengi-16000", "Kammanadu-16000", "Kollipake-7000" etc. It is generally agreed that these numerical suffixes represent the number of villages that were included under the territorial unit. However, I am confused as to how many people actually resided in each of these villages. An average population of 100 per village would seem to be too much, because this would mean that the Gangavadi region (comprising the Mysore plateau) would have a very high population of 9.6 million. To put this into perspective, the population of Mysore kingdom in 1941 was only 7.3 million, and the population of the whole Karnataka in 1951 (which would include the old Nolambavadi and Banavasi divisions, among others) was only 19.4 million. However, other divisions like Vengi, Kammanadu, and Kollipake would seem to have reasonable medieval populations if we apply the formula of 100 inhabitants per village. This would give a population of perhaps 5 million for coastal Andhra and about 1 million for Telangana (the Kollipake-7000 division included most of modern-day Telangana).

The population of the entire South India in medieval times is estimated to be around 25 million only. How do we reconcile this with the large numbers of villages recorded in the numerical suffixes of old Indian territorial units?
 
Jun 2014
4,516
India
Many inscriptions from early medieval and medieval India record territorial divisions with numerical suffixes, such as "Gangavadi-96000", "Nolambavadi-32000", "Banavasi-12000", "Vengi-16000", "Kammanadu-16000", "Kollipake-7000" etc. It is generally agreed that these numerical suffixes represent the number of villages that were included under the territorial unit. However, I am confused as to how many people actually resided in each of these villages. An average population of 100 per village would seem to be too much, because this would mean that the Gangavadi region (comprising the Mysore plateau) would have a very high population of 9.6 million. To put this into perspective, the population of Mysore kingdom in 1941 was only 7.3 million, and the population of the whole Karnataka in 1951 (which would include the old Nolambavadi and Banavasi divisions, among others) was only 19.4 million. However, other divisions like Vengi, Kammanadu, and Kollipake would seem to have reasonable medieval populations if we apply the formula of 100 inhabitants per village. This would give a population of perhaps 5 million for coastal Andhra and about 1 million for Telangana (the Kollipake-7000 division included most of modern-day Telangana).

The population of the entire South India in medieval times is estimated to be around 25 million only. How do we reconcile this with the large numbers of villages recorded in the numerical suffixes of old Indian territorial units?

Thanks that you brought this up, in my view even 100 persons per village is quite low figure and hence in reality, chances of these suffixes indicating number of villages are almost impossible unless we believe that population declined heavily which is in direct contradiction with what we know.
In my view,the idea that they represent only number of villages is no sound, it is also possible that they may represent some unit of revenue or number of soldiers manytimes.

Do we have any reliable account for number of armymen and especially cavalry from muslim sources about petty kingdoms like Reddy at Kondavidu, by comparing with contemporary societies in Eurasia we can take out population and then have some estimate about same for 800-1300 AD.
 

civfanatic

Ad Honorem
Oct 2012
3,318
Des Moines, Iowa
Thanks that you brought this up, in my view even 100 persons per village is quite low figure and hence in reality, chances of these suffixes indicating number of villages are almost impossible unless we believe that population declined heavily which is in direct contradiction with what we know.
In my view,the idea that they represent only number of villages is no sound, it is also possible that they may represent some unit of revenue or number of soldiers manytimes.
What is really striking to me is the huge disparity between the numerical suffixes. We have divisions with really small suffixes like Honavatti-12, Simhigrama-12, Sabbi-30, Jiddulige-60, Cheraku-70, etc., and at the same divisions with really large suffixes like Nolambavadi-32000 and Gangavadi-96000. I find it difficult to believe that both the really small and really large divisions represent the same thing. However, scholars are nearly unanimous that all of these numbers represent the number of villages, even if only the smaller numbers make real sense.

One of the largest divisions I have encountered is the seven-and-a-half lakh Rattapadi country, or Rattapadi-750000. This term was used to describe the lands ruled by Western Chalukyas, after they had seized power from the Rashtrakutas (Rattas). The term "seven-and-half-lakh" cannot possibly refer to the actual number of villages even if we consider the whole Deccan and South India, yet there is a passage from the Kumarikakhanda in Skandapurana which clearly says, "graamaanaam saptalakshancha Rattaraaje prakirtitam". The only explanation is that these numbers were deliberately exaggerated, but the reason for this exaggeration is beyond my understanding.


Do we have any reliable account for number of armymen and especially cavalry from muslim sources about petty kingdoms like Reddy at Kondavidu, by comparing with contemporary societies in Eurasia we can take out population and then have some estimate about same for 800-1300 AD.
I don't think such numbers are available, but I will need to consult some books and see.
 
Last edited:

tornada

Ad Honoris
Mar 2013
15,398
India
maybe it literally refers to territorial extent. Some form of land measurement?
 
Jun 2014
4,516
India
What is really striking to me is the huge disparity between the numerical suffixes. We have divisions with really small suffixes like Honavatti-12, Simhigrama-12, Sabbi-30, Jiddulige-60, Cheraku-70, etc., and at the same divisions with really large suffixes like Nolambavadi-32000 and Gangavadi-96000. I find it difficult to believe that both the really small and really large divisions represent the same thing. However, scholars are nearly unanimous that all of these numbers represent the number of villages, even if only the smaller numbers make real sense.

One of the largest divisions I have encountered is the seven-and-a-half lakh Rattapadi country, or Rattapadi-750000. This term was used to describe the lands ruled by Western Chalukyas, after they had seized power from the Rashtrakutas (Rattas). The term "seven-and-half-lakh" cannot possibly refer to the actual number of villages even if we consider the whole Deccan and South India, yet there is a passage from the Kumarikakhanda in Skandapurana which clearly says, "graamaanaam saptalakshancha Rattaraaje prakirtitam". The only explanation is that these numbers were deliberately exaggerated, but the reason for this exaggeration is beyond my understanding.




I don't think such numbers are available, but I will need to consult some books and see.


There is a rule of thumb here, whenever I feel numbers being exaggeration, I remove one zero and they start making sense. This is how I think that Chalukya inscription mentioning 9 lakh Chola soldiers should be read as 90,000, Kalhana's figures of same for army of Kashmir should be read and in similar manner, Arab description of 3 million soldiers of Pratiharas should be read as 3 lakh soldiers in reality. The authors in those times lacked good knowledge about numbers ( a trait still seen in India with people always claiming thousands of lives being lost in petty riots or floods when in reality some few scores are killed) and hence wrote his carelessly.
 

civfanatic

Ad Honorem
Oct 2012
3,318
Des Moines, Iowa
There is a rule of thumb here, whenever I feel numbers being exaggeration, I remove one zero and they start making sense. This is how I think that Chalukya inscription mentioning 9 lakh Chola soldiers should be read as 90,000, Kalhana's figures of same for army of Kashmir should be read and in similar manner, Arab description of 3 million soldiers of Pratiharas should be read as 3 lakh soldiers in reality. The authors in those times lacked good knowledge about numbers ( a trait still seen in India with people always claiming thousands of lives being lost in petty riots or floods when in reality some few scores are killed) and hence wrote his carelessly.
Yes, I have noticed the same thing. I think this was also done by ancient Greek writers when they wrote about India. The figure of 700,000 troops for Mauryans becomes more believable if we read it as 70,000.

If we assume that the figure seven and half lakhs is exaggerated ten-fold, then this means that Rattapadi division had around 75,000 villages. A figure of 100 persons per village seems too low as you have noted, while 500 per village still gives too high figures. If we use a middle value of 250-300 persons per village, this would give a population of about 20 million for the Rattapadi division (assuming 75,000 villages). If we accept that Rattapadi means the lands once ruled by Rashtrakutas, then this would include most of Deccan and would be a believable figure.
 

Jinit

Ad Honorem
Jun 2012
5,274
India
Probably because it represents the administrative system with a village as its basic unit. All the subsequent units/tiers might have been represented by the multiplication of the number of the villages. And if we take all those zeros into account then probably it might have been based on the multiplication of 10. So for instance if a village is basic unit, 10 villages make the second unit B, 10 unit B comprised the Unit C consisting of 100 villages, 10 unit C comprised the unit D consisting of 1000 villages and so on. Of course, in reality this theoretical norms might not have been followed strictly or might have been abandoned later on, however the territorial divisions might still have been represented by the the theoretical number (of villages) affiliated to the particular level of administrative unit rather than the actual number of villages and hence the large units ended up with exaggerated numbers of villages. For eg Banavasi - 12000 might have consisted 12 Mandal*. Although in reality each Mandal might not have consisted of 1000 villages but since theoretically Mandal supposed to be made of 1000 villages, they just called it Banavasi 12000 based on the number of the Mandals, rather than counting the actual number of villages in the entire territory. (*the use of the term Mandal here is just an example. It isn't supposed to be historically accurate territorial unit)

Although this is just a speculation and not supported by any proper source but it seems possible if we take a similar case of western India into account. There are mentions of territorial units such as Vaguri - 36000, Virata - 80000, Sapadlaksha (?) etc. In the same area the system of territorial division became famous from 7th century onwards where each subsequent division was represented in the multiplication of 6 villages. The most well known being the unit of 84 villages (such landholders with authority on 84 villages later on got the "Chaurasia" surname). Although in reality this theoretical division wasn't followed strictly. May be same might be the case in Deccan.

Btw I am not sure but I have read somewhere that according to R.S. Sharma the large numbers like 96000 actually represented the number of the revenue paying families or landholders into the entire territory and not the number of villages.
 
Last edited:
Mar 2014
185
Battlefield
Many inscriptions from early medieval and medieval India record territorial divisions with numerical suffixes, such as "Gangavadi-96000", "Nolambavadi-32000", "Banavasi-12000", "Vengi-16000", "Kammanadu-16000", "Kollipake-7000" etc. It is generally agreed that these numerical suffixes represent the number of villages that were included under the territorial unit. However, I am confused as to how many people actually resided in each of these villages. An average population of 100 per village would seem to be too much, because this would mean that the Gangavadi region (comprising the Mysore plateau) would have a very high population of 9.6 million. To put this into perspective, the population of Mysore kingdom in 1941 was only 7.3 million, and the population of the whole Karnataka in 1951 (which would include the old Nolambavadi and Banavasi divisions, among others) was only 19.4 million. However, other divisions like Vengi, Kammanadu, and Kollipake would seem to have reasonable medieval populations if we apply the formula of 100 inhabitants per village. This would give a population of perhaps 5 million for coastal Andhra and about 1 million for Telangana (the Kollipake-7000 division included most of modern-day Telangana).

The population of the entire South India in medieval times is estimated to be around 25 million only. How do we reconcile this with the large numbers of villages recorded in the numerical suffixes of old Indian territorial units?

As per the OECD study done by Angus Maddison:
The World Economy Volume 1: A Millennial Perspective and Volume 2: Historical Statistics

In 1000 AD, total world population was about 230M with Indian population at 75M(3 times of your estimate), China was second with around 59M. Asia had a total of about 171M.

So these numbers aren't unrealistic especially if you consider that many of these villages with population <100. Its true even in current India. Villages at that time were more like extended families especially true considering several separate 'villages' on the basis of caste living in a geographical area equivalent 1 modern village.
 
Last edited:
Jun 2014
4,516
India
As per the OECD study done by Angus Maddison:
The World Economy Volume 1: A Millennial Perspective and Volume 2: Historical Statistics

In 1000 AD, total world population was about 230M with Indian population at 75M(3 times of your estimate), China was second with around 59M. Asia had a total of about 171M.

So these numbers aren't unrealistic especially if you consider that many of these villages with population <100. Its true even in current India. Villages at that time were more like extended families especially true considering several separate 'villages' on the basis of caste living in a geographical area equivalent 1 modern village.

First, there is zero basis for Maddison's figure, we can say that Indian population was 30 million and we can also say it as 100 million as we have zero data . However, civafanatic is talking about population of south India and not entire India and by no means Gangavadi was so densely populated as figure of 96000 villages would demand. Even 50 persons per village would be giving a figure of 3.8 million which in my view was population of entire Karnataka leave alone Mysore region.
 
Jun 2014
4,516
India
So these numbers aren't unrealistic especially if you consider that many of these villages with population <100. Its true even in current India. Villages at that time were more like extended families especially true considering several separate 'villages' on the basis of caste living in a geographical area equivalent 1 modern village.
This is not correct, whatever we know tells us that within a village, there were localities for each castes so you had leather workers, potters, brahmins and martial classes all living in same village. Every definition of grama includes atleast 25 families to call an inhabited place as a village in pre modern Indian texts.