Pramāṇa · Search · About · API

Growth of Muslim Population in Medieval India, A.D. 1000–1800 c6: II. Some Methodological Problems of Estimating Population

Growth of Muslim Population in Medieval India, A.D. 1000–1800 c6

Language
English

Contents

II. Some Methodological Problems of Estimating Population

c6.2

Our source materials discussed in Section I are many and varied in nature. However their sifting and utilization for a demographic study of medieval India are beset with many difficulties. Of course it would not be fair to expect from medieval chroniclers definite population statistics; these are not found in any historical works in any country in the pre-census times. But while our sources are deficient in figures and data, they are rich in facts providing historical background and circumstantial evidence for demographic behaviour. We shall discuss their merits and faults in the same sequence as adopted in Section I.

Arabic and Persian Works

c6.s1.1

The weaknesses of medieval chronicles are well-known. Their style is by and large turgid and ornamental; their narrative is often exaggerated. And this applies as much (if not more) to figures as to facts. A few no doubt are trustworthy but many of them are extremely faulty with regard to figures and statistics; and almost all of them let their imagination and their pen run riot. Consequently, even when they are not quite reticent on demographic matters, they are neither very informative nor always reliable.

c6.s1.2

Thus Al Masudi (writing about A.D. 941-42) says that there are 120,000 towns and villages in Sind.¹ Rashiduddin, in *Jami-ut-Tawarikh* (completed 1310) writes that "it is said that Guzerat comprises 80,000 flourishing cities, villages, and hamlets".² He also says that Sawalik (Haryana) contains 125,000 cities and villages, while Malwa had 1,893,000 villages and towns.³ al-Qalqashindi's *Subh-al-Asha* (completed 1411) says that the villages of Multan are 126,000.⁴ Similarly, Shihabuddin Ahmad also grossly exaggerates when he says that the Delhi (of Sultan Muhammad Tughlaq) contains "a thousand colleges.... In it there are about seventy hospitals... the chapels and hermitages amount to 2,000... The army consists of 900,000 horsemen... At the court there are 1,200 physicians, 10,000 falconers, 1,200 musicians...and 1,000 poets... In the royal kitchen the animals that were killed daily were 2,500 oxen, 2,000 sheep, without taking into account fatted horses and birds of all descriptions."⁵

c6.s1.3

¹ Al Masudi, *Muruj-ul-Zuhab*, trs E and D, *op cit.* (I,n. 2) I, p. 23.

c6.s1.4

² Rashid-ud-din, *Jami-ut-Tawarikh*, trs. E and D, I, p. 67

c6.s1.5

³ *Ibid.*, p. 68

c6.s1.6

⁴ al-Qalqashindi, trs. by Otto Spies (Aligarh, n.d.), p.24. It may be noted that in 1931 it was estimated that the whole of India (including Pakistan) had about 685,665 villages. *Indian Year Book*, 1931, p. 10.

c6.s1.7

⁵ Shihab-ud-din Abul Abbas Ahmad, *Masalik-ul-Absar Fi Mamalik-ul-Amsar*, trs. E and D, op.cit.(f.n.2) III, pp. 576-78.

c6.s1.8

Only a few of these figures appear to be correct. For example, seventy dispensaries, large and small, Government controlled or privately owned, of Unani and Vedic systems of medicine, would not be too many for the then Delhi. Even a thousand poets may have been there because we know that Mahmud of Ghazni had 400 at his court.⁶ Again since Muhammad Tughlaq could send an army of 370,000 on a single expedition, a total of 900,000 for the whole of his dominions may also be conceded. Else, the testimony of Shihabuddin as well as of others cited is not wholly reliable. But it is not entirely useless. As Professor Karl F. Helleiner points out, the ancients spoke in terms of more and less, large and small, high and low. "They cannot often have mistaken prosperity for misery, progress for decline, demographic growth for depopulation."⁷ Therefore, what these writers say is true in spirit if not in letter. They were writing (as Shihabuddin says) about "a most important country with which no other country in the world can be compared.

c6.s1.9

⁶ Muhammad Qasim Hindu Shah Ferishtah, *Gulshan-i-Ibrahimi*, also known as *Tarikh-i-Ferishtah*, Persian Text, Nawal Kishore Press (Lucknow, 1865), I, p. 19.

c6.s1.10

⁷ *The Cambridge History of Europe*, IV, edited by E.E. Rich and C.H. Wilson (Cambridge, 1967) p. 3.

c6.s1.11

in respect of extent, riches, the number of armies. . ."⁸ and to impress this idea they gave inflated figures. But though their data may not be correct, their demographic image of India is correct. To quote Helleiner again, "the demographic history of that 'proto-statistical' age requires that quantitative information is supplemented by circumstantial evidence, as well as recorded opinions of contemporaries."⁹ That far the testimony of the above writers has a value.

c6.s1.12

⁸ *Masalik-ul Absar*, *op cit*, (n 5) p 574.

c6.s1.13

⁹ *Cam. Econ. History*. *op. cit* (n 7) p 2

c6.s1.14

Now large figures or testimony of 'foreigners' in themselves need not always create misgivings. Ibn-ul-Asir did not come to India, but most of what he writes about its geography and history is by and large correct. For example when he says that the population of Banaras was "said to amount to a million"¹⁰ in the 30's of the eleventh century, we may not reject his statement just because today Banaras does not have a million inhabitants. We know on the authority of Alberuni that, besides being a centre of Hindu learning and pilgrimage, Banaras had received large number of refugees after Mahmud of Ghazni's invasion of the Punjab.

c6.s1.15

¹⁰ Ibn-ul-Asir, *Kamil-ut-Tawarikh*, trs. E and D, II, p. 251.

c6.s1.16

Indian chroniclers share the weaknesses of foreigners. They also often exaggerate, particularly when writing about casualties in wars. Of the dozens of medieval Persian chroniclers, it cannot be said that some are truthful while others are less reliable. The same writer may give some figures correctly, in others he may commit mistakes and even wilfully. Therefore, every figure and fact given by medieval chroniclers needs to be put under the minutest scrutiny. We have to cross-check the figures given by one writer with those of others, and only when this opportunity is denied, can the data be accepted only if supported by historical antecedants and situations. A few examples would show how careful one has to be when dealing with medieval writers.

c6.s1.17

Ziyauddin Barani is an eye-witness historian, but his figures and data are not always precise. When sometimes he means to convey 'a very large number', he gives the figure of 100,000. That in 1297-98 the Mongol leader Kadar invaded India with 100,000 horse,¹¹ may be acceptable, but when he says that the Mewatis, during Balban's reign, killed 100,000 royal troops¹² he definitely exaggerates, for neither the Mewatis were such well-organized fighters nor Balban so weak a king as to have let them destroy a hundred thousand of his troops. It may be said that Barani's '100,000' only conveys an idea of the immensity of numbers, and he is usually casual with his figures. For example, at one place he says that the army of Qutlugh Khwaja, the Mongol leader who invaded India in 1299, was 200,000 (*ba bist tuman*), while at another that it was only 100,000 (*yak lak sawar darand*).¹³ Similarly, during Targhi's invasion (1303), according to Barani, the Mongol army was 120,000 horse (*dawazdah tuman*), but a little later, on the same page of his narrative, it becomes 30 to 40 thousand (*ba si chahal hazar*).¹⁴ But this very chronicler, in many other cases, is quite moderate in his estimates and is reliable. The whole story of Alauddin Khalji's Market Control as given by him together with the price lists of foodgrains, cloths, horses, slaves, etc. appears to be true.

c6.s1.18

¹¹ Barani, Ziyauddin, *Tarikh-i-Firoz Shahi* (Calcutta, 1862) p. 250.

c6.s1.19

¹² *Ibid*, p. 57.

c6.s1.20

¹³ *Ibid.*, pp. 254, 256.

c6.s1.21

¹⁴ *Ibid.*, p. 300.

c6.s1.22

Let us take another fourteenth century historian, Shams Siraj Afif. If Barani has a weakness for 100,000, Afif is very fond of '180,000', so that the slaves of Firoz Tughlaq numbered 180,000, the revenue from his 1200 gardens was 180,000 *tankahs*, and in his war with Shamsuddin Ilyas Shah (A.D 1353), he killed 180,000 men in Bengal.¹⁵ The immensity and coincidence of the numbers create misgivings. On checking up it appears that while the figure in the case of slaves and revenue from gardens seems to be more or less correct, the numbers of people killed in Bengal are exaggerated.¹⁶ To make us believe him, Afif gives detailed description of the war in Bengal. The army of Ilyas consisted of 10,000 horse and 200,000 foot. He was almost lured into an ambush. The slaughter of Bengalis went on for a whole day and in an area of seven *kos* around. After the massacre the Sultan ordered his soldiers to collect the heads of the slaughtered, promising to give one silver *tankah* for each head. The troops therefore took to the task in all earnestness. The heads were counted in the Sultan's presence and were found to be 180,000.¹⁷ The chronicler has left no stone unturned to convince us of his veracity. Still the figure seems to be unduly. large for Firoz Tughlaq; it could have fitted in with the narrative of massacres of Chingiz or Timur. And then there is the authority of the *Sirat-i-Firoz Shahi*, which says that only 60,000 were killed.¹⁸

c6.s1.23

¹⁵ Afif, Shams Siraj, *Tarikh-i-Firoz Shahi* (Calcutta, 1890) pp. 270, 296 and 121 respectively.

c6.s1.24

¹⁶ According to Afif the slaves were distributed all over the empire. About 40,000 were in the King's personal service, 12,000 worked in the Royal Workshops, some were employed on reciting the Qruan, others were in the Royal Suites (Alim Khan, Ghariyal Khani etc); and still others were sent to Shiqqas, Iqatdars and muqanddars in such far-off places as Multan, Dipalpur and Gujrat and some were sent even to Mecca (Afif, *op cit* (n.15) pp. 267-72). With such details on hand it would not be fair to suspect, much less to reject, this figure.

c6.s1.25

¹⁷ Afif, *op. cit.* (n 15) pp. 121.

c6.s1.26

¹⁸ The Mss of the National Library, Calcutta, and the Oriental Public Library, Bankipur, of *Sirat-i Firoz Shahi* have 60,000. See R.C. Jauhari, *Firoz Tughluq* (Agra, 1968), p. 51. The Allahabad University copy of the Ms. contains a fantastic estimate of 600,000 which is obviously an error of the copyist. See J.M. Banerjee, *History of Firoz Shah Tughluq* (Delhi, 1967), pp. 32 and 51-52 n. Yahıya, a later historian, merely says "The Bengalis were defeated, and their casualties were large." Yahiya Sarhindi, *Tarikh-i-Mubarak Shahi*, (Calcutta, 1931), p. 125.

c6.s1.27

Only one more case may be examined. Muhammad Aslam's *Farhat-un Nazirin*, written in 1770-71, is a contemporary work for the 1761 Battle of Panipat But when the author says that "The dead lay strewn shoulder to shoulder from the plain of Panipat to Delhi. About ninety thousand persons, male and female, were taken prisoners, and obtained eternal happiness by embracing the Muhammadan faith. Indeed, never was such splendid victory achieved from the time of Amir Mahmud Subuktigin...",¹⁹ he appears to be trying to convince us about the large number converted, but fails to impress. No other contemporary work (and there are quite a few for this Battle) mentions proselytization on such a large scale. The author exaggerates, but how much discount could be given to the number of the converted is difficult to determine. Were this the figure for conversions and deportations during *all* the campaigns of Abdali in the Punjab, it could be believed.

c6.s1.28

¹⁹ Muhammad Aslam, *Farhat-un-Nazirin*, trs. E and D, *op. cit.* (I; n 2) VIII, p. 171.

c6.s1.29

It must be emphasised, however, that large figures by themselves, though sometimes frightening, may not always be incorrect. In this connection let us examine the statement of Professor H. K. Sherwani, who, commenting on the details given by medieval chronicles on the Bahmani-Vijayanagar war says: "If the number of the enemies of the Bahmanis said to have been killed or massacred during the various campaigns undertaken during the hundred and fifty years of Bahmani rule were to be added together, there would hardly be a human soul left south of the Tungbhadra or in the Telingana."²⁰ It is true that the figures and statistics of contemporary (and later) writers are imposing, but there are reasons for their being so. The number of combat troops in the Deccan was undoubtedly large—so large that "it certainly demands a large strain on our credulity."²¹ Leaving aside the testimony of Indian writers in this regard, Nuniz, Barros, and Conti all place the army of Vijayanagar above a million. Abdur Razzaq (1442), who stayed at the Vijayanagar court for about six months and whose statement therefore can hardly be doubted, puts the number at 1,100,000 with 1,000 elephants. The Bahmanis too had an immensely large army. Nikitin states that in the 1462 campaign the Gulbarga forces amounted to 190,000 horse, 900,000 foot, and 575 elephants. The sultan himself, independently of his nobles, took the field with 300,000 men. Many other facts and data are available,²² but the above mentioned should suffice to prove that killings in the Bhamani-Vijayanagar war were on an extensive scale.

c6.s1.30

²⁰ Sherwani, Haroon Khan, *Mahmud Gawan*, Kitabistan (Allahabad, 1941), pp., 24 n. 50.

c6.s1.31

²¹ Sewell, Robert, *A Forgotten Empire*, Publications Division, Govt. of India (Delhi, 1962), p. 147.

c6.s1.32

²² *Ibid.*, pp. 142-145.

c6.s1.33

In South "the Hindus fought almost nude."²³ Such troops formed the bulk of the armies on both sides, and in any battle these 'naked' soldiers would have been killed in large numbers. Civilian population was always ruthlessly massacred, so the accounts about the South say. We come across wars originating in an after-dinner jest and ending with such slaughter that 'victims on the Hindu side alone were no less than half a million" (A.D. 1336).²⁴ In 1417-18 when Firoz Shah Bahmani invaded Telingana, the Hindus made a general massacre of the Musalmans and erected a platform of their heads.²⁵ The tables were turned in 1423 with the invasion of the Vijayanagar dominions by Ahmad Bahmani, who so overran the country that he used to halt for rest for three days only after he had massacred an instalment of 20,000 men, women and children.²⁶ When such details of the strength of forces deployed and killings among soldiers and civilians are given by contemporary writers (on whose narratives the accounts of Sewell and Wolseley Haig are based), it would not be possible to disbelieve the great losses in war. It is in a decisive battle that casualties are limited because the killing is over in a day or two. But the war between the two kingdoms was indecisive; it was a continual affair for more than a century and a half, and killings and massacres, often on a large scale, were its special feature. That people survived in Telingana and south of the Tungbhadra was due more to the demographic recuperation through natural procreation than to the smallness of the numbers killed. A slight exaggeration here and there does not minimise the enormity of the loss, and details of the campaigns lend support to the data of contemporary writers.

c6.s1.34

²³ *Ibid.*, p. 143.

c6.s1.35

²⁴ *Ibid.*, pp. 31-39.

c6.s1.36

²⁵ *Cambridge History of India*, III, Ed. Sir Wolseley Haig (Delhi Reprint, 1958, p. 329. Also Sewell, p. 65.

c6.s1.37

²⁶ C.H.I., III, pp. 397-98 ; Sewell, pp. 67-68.

c6.s1.38

In this context it may be necessary to say a word about how the information and data about casulties in war were obtained by medieval writers. Killing the ‘enemy’ and counting the dead in *field have been a pleasure and a pastime of the strong through the ages.* In primitive societies the victors used to cut the thumbs of those whom they killed, make garlands of these, and wear them as trophies. Necklaces of skulls are also not unknown. Similarly in ancient Egypt, “the military secretaries, immediately after the conclusion of a battle reckoned the number of the slain in the presence of the king. This counting was done by cutting off the hands......”²⁷ In medieval times we come across repeated references to the raising of “towers of skulls” after battles, or to the actual sending of heads of the killed to the court,²⁸ which would have helped in computing the numbers of the killed.

c6.s1.39

²⁷ Sir Flinders Petrie in *Story of the Nations*, p. 33.

c6.s1.40

²⁸ Amir Khusrau, *Deval Rani*, (Aligarh, 1917) p. 61, Barani, *op. cit.*, (n. 11) p. 321. The fact is so well-known and the references so frequent that they hardly need be cited.

c6.s1.41

Reference has already been made to Firoz Tughlaq’s counting the killed in Bengal. In the Assam campaign “khani-khanan ordered that the prisoners should have the heads of the slain tied round them, and be thus exposed to the derision of the camp...and afterwards put to death”.²⁹ Similarly the sultans sometimes gave cash awards for ‘hunting’ the heads of the enemy, and the amount disbursed would have provided a sure way of finding the number of casualties.³⁰ To gloat over the success of their patrons or just to please them by crediting them with extraordinary achievements, medieval chroniclers sometimes might have over-estimated the numbers killed by them, but they always gave figures and data on the basis of some source, official or private, which they knew to be reliable.

c6.s1.42

²⁹ Khafi Khan, *op. cit.* (I, n. 3). Trs. in E and D, VII, p. 268.

c6.s1.43

³⁰ *Deval Rani*, *op. cit.* (n. 28). p. 61; *Afif*, *op. cit.* (n. 15), p. 121. Also see Rizqullah, *Waqiat in Mushtaqi* fols 38(b)-39(a) who mentions a case of collection of shoes of the enemy who lost their lives in action.

c6.s1.44

Some wrote on the basis of information lent to them by their father or grandfather or some other old and experienced people who were eye-witnesses. Others had friends in government service and narrated events of campaigns in which they themselves were participants and information so transmitted was collected by these chroniclers.³¹ Such sources could not be unimpeachable and have sometimes landed the writers into errors. But most of the chroniclers of medieval India were official historians who wrote at the behest of the king or nobles or under their patronage. These writers had official records at their disposal. “There were no secrets in Mogul administration” says Vincent Smith,³² and when even foreigners could consult official records,³³ Indian court and private historians certainly did. One wonders if there was any document in Akbar’s archives (in the Imperial Capital, provincial capitals or chief cities) to which Abul Fazl was denied access. Similarly to Ziyauddin Barani was available Kabir-ud-din’s *Fateh Namah* in which every day occurrences of Alauddin Khalji’s reign were recorded. Such examples need not be multiplied. Official records were many; most of these have been lost by now, but they were easily available to medieval historians.

c6.s1.45

³¹ E.G. Barani, *op. cit.* (n. 11; pp. 67, 114, 229, 370. Isami, *Futuh-us-Salatin* (Agra, 1938). pp. 199, 340, 341.

c6.s1.46

³² Smith, *Akbar the Great Mogul* (Delhi Reprint, 1962), p. 2.

c6.s1.47

³³ *Loc. cit.*

c6.s1.48

In conclusion it may be said that on a study of the medieval chronicles it appears that with regard to the figures of the casualties of the ‘enemy’ in war our chroniclers, probably to glorify their heroes—conquerors and rulers—sometimes give inflated figures. Their estimates of the population of a city or region are also sometimes liable to be optimistic; their one major drawback is that they ignore the rural scene altogether. But in many other details, which are helpful in demographic study, they are fairly trustworthy. For instance, many chroniclers give reliable data about the prices of foodgrains and other commodities pertaining to their times. About the same period the amount of revenue, the rate of revenue, the price-schedules on which the revenue demand was settled, are also given. All this information helps in calculating the total amount of produce on which the then population subsisted and estimating, on the basis of per capita consumption of foodgrains and other edibles, the then density of population. Thus, in spite of their few weaknesses, the facts and figures supplied by medieval chroniclers are of great value in our study.

c6.s1.49

Writings of Foreigners

c6.s1.50

So far as population studies about medieval India are concerned, the itineraries of foreign visitors to this country like travellers and traders are even more important. They have written about famines, pestilences and wars and about towns and cities thinly or thickly populated. They often compare Indian cities with those of their own countries. On a few occasions they give definite and fairly reliable population statistics. But their notices too are not free of weaknesses. Muslim travellers, coming from countries of small population densities, are invariably impressed with the teeming millions in India and, therefore, sometimes give exaggerated estimates of Indian population. The Europeans are more sober. But in fixing the proportion of Muslim numbers in the overall population structure, the testimony of European writers has to be accepted with great caution. Many of them were either not capable or not careful about distinguishing between peoples of various religious faiths. Vasco da Gama had landed near Calicut on 17 May, 1498. But even after spending three months in Calicut in a Hindu country, he was hardly aware of the existence of the Hindu Religion or Hindu Society, and 'he mistook a Hindu temple for a chapel'³⁴... entered the temple and prayed there. As Denison Ross pertinently remarks, "It remains a mystery why they (the Portuguese) failed to discover that the Zamorin was neither Christian nor Muslim, seeing that they were for so long in daily intercourse with him."³⁵

c6.s1.51

³⁴ R.P. Rao, *Portuguese Rule in Goa*, (Bombay, 1963), pp. 24-25.

c6.s1.52

³⁵ *Cambridge History of India*, Ed. H.H. Dodwell, V (Delhi Reprint, 1963), p. 5.

c6.s1.53

Even in northern India, where Muslims were in good numbers, there is great doubt if Europeans could always distinguish a Hindu from a Muslim. Just as to Indians, all Europeans of various nationalities look alike, similarly many Europeans in all probability could not always differentiate between a Hindu and a Muslim in medieval times. After all the majority of Muslims were Hindu converts and their physiognomy did not change with conversion to Islam. Besides, Hindus and Muslims usually dressed alike. As early as the tenth century Ibn Hauqal was struck by the fact that "the Muslims and infidels in this tract (Kambaya to Saimur) wear the same dresses, and let their beards grow in the same fashion. They use fine muslin garments on account of the extreme heat. The men of Multan dress in the same way."³⁶ While it can only be assumed that the high classes of both the communities in the Sultanate period dressed alike, they are actually found to be doing so in the Mughal period. Hindu converts to Islam, especially of the lower classes, did not change their way of dressing with change of faith. That Hindus and Muslims looked alike is also proved by the fact that many rulers of the Sultanate and Mughal times enforced regulations requiring Hindus to wear distinguishing marks on their dresses so that they might not be mistaken for Muslims.³⁷ Qazvini says that Shahjahan had ordered that Hindus be not allowed to dress like Muslims.³⁸ The *Fatawa-i-Alamgiri* also recommended that the Hindus should not be allowed to look like Muslims.³⁹

c6.s1.54

³⁶ Ibn Hauqal, *Ashkalal-ul-Bilad*, trs. E and D, *op. cit.* (I, n. 2) I, p. 39.

c6.s1.55

³⁷ Sri Ram Sharma, *The Religious Policy of the Mughal Emperors* (Asia Publishing House, Bombay, 1962), p. 5.

c6.s1.56

³⁸ Qazvini, *Badshahnamah*, p. 445, cited in Sharma, *op. cit.*, p. 93.

c6.s1.57

³⁹ *Fatawa-i-Alamgiri*, Urdu trs. (Newal Kishore Press, Lucknow), III, pp. 442-45.

c6.s1.58

Many local officers also issued similar orders in their assignments.⁴⁰ But all these regulations could not have been strictly enforced. The paintings of the Mughal period as well as those of the Rajput and Pahari *qalam* clearly point to similar dresses for the Hindu and Muslim elite. Manucci says that many Rajputs wore gold armlets, "so that, if killed, those finding the body may see to its cremation." Otherwise, except that the Rajputs usually shaved, all put on a 'cabaya' (qaba) and more or less looked alike.⁴² Of course, the Indians themselves could not have mistaken a Hindu for a Muslim and *vice versa*. But it would have been difficult for a foreigner to distinguish between the two unless his stay in the country was of long duration and his curiosity and interest very keen. But, as we shall see later on, even keen observers like Pires and Bernier make mistakes in estimating the proportion of Muslim numbers. Sartorial patterns are determined more by court etiquette, current fashions, and climatic conditions, than by mere religious affiliations, and the mistake of the Europeans was unconscious. But in many cases, and because of this difficulty, the accounts of these foreigners are not quite helpful in fixing the ratio between the Hindus and Muslims in the population of Medieval India.

c6.s1.59

⁴⁰ Husain Khan, Jagirdar of Patiali and Shamsabad, had ordered that the Hindus of his jagir should wear a piece of cloth on their dress as a mark of distinction from the Muslims. Badaoni, *Muntakhab-ut-Tawarikh* (Calcutta, 1868-69), II, p. 223.

c6.s1.60

⁴² *Ibid*, II, pp 122-23.

c6.s1.61

Else, the notices of Europeans are very informative. Coming from abroad, they were curious to know, and their observations are often more informative and reliable than those of indigenous writers.

Modern Works

c6.s2.1

In modern works, the monumental work of Kingsley Devis on the population of India and Pakistan gives very reliable estimates. Edward Thornton and W. H. Moreland have added to our knowledge, the former by collecting a plethora of reliable data about the total population and Hindu-Muslim ratio in cities, districts and regions, and the latter by doing valuable pioneering work in the field of medieval Indian demography. But from Thornton's Gazetteer itself it is evident how sometimes contemporary writers and even surveyors could make mistakes in counting. There may not be many such instances, but one is indeed staggering. Writing about the city of Thatta in Sindh, Thornton says:

c6.s2.2

⁴¹ Manucci, *Storia do Mogor*, English trs by W. Irvine, 4 vols., (London, 1906), II, p. 437.

c6.s2.3

"The present population is variously estimated—at below 2,000 (Kennedy, I, 81), at less than 10,000 (Burnes, *Report on Commerce of Thatta*, 8), at 12,000 (Outram, 18), 15,000 (Burnes, *Bokhara*, III, 30,227), 18,000 (J.R.A.S., 1834, 234), 20,000 (Pottinger, 352), and 40,000 (James Burnes, *Sinde*)".⁴³ About Moreland's estimate Carr Saunders in *World Population* has commented thus: "Moreland's figure has been quoted with favour in the census reports of India, no better estimate is available, but its factual basis is of the most slender kind".⁴⁴ Even so, modern works on population studies are results of patient scientific research and are very valuable.

c6.s2.4

⁴³ Thornton's Gazetteer, *op. cit.* (I, n. 7), IV, pp. 663-64.

c6.s2.5

⁴⁴ Cited by J. M. Datta in "A Re-examination of Moreland's Estimate...at the Death of Akbar", *op. cit.*, (I, n. 10), p. 165.

c6.s2.6

The Indian Census Reports are authentic and provide a detailed analysis of population trends from 1881 onwards. The demographic patterns marked out by them help in guaging the situation prior to 1881 and indeed for the medieval times. But these reports have only a limited value for us. If it were not so, our task would have become very easy. On the basis of the known statistics, say of the Muslim population between 1881 and 1941, we could have calculated regressively to find the Muslim numbers in the year 1781, 1681, 1581, 1500, 1400 or, for that matter, for any single year in the medieval period. But when this was done on the electronic computer, the results obtained were fantastic. (Calculation in regression on the electronic computer showed that 94 million Muslims of 1941 or 49 million of 1881, would have been 510 in A.D. 1401 and only 10 in A.D. 1241—an absolutely wrong result.) Moreover, Indian census figures have not always been unimpeachably reliable. Uneducated and even educated but superstitious parents have sometimes been found reluctant to truly disclose the number of their children as a protection against the 'evil eye of the jealous'. In the 1941 census population figures were inflated on account of communal rivalry; and the Pakistan census report of 1951 estimates that the inflation of Muhammadans in 1941 was of the order of 10 per cent.

Conclusions

c6.s3.1

It is thus clear that contemporary and later source materials, which are the bases of our study, are far from being 'perfect' for research on medieval Indian demography. But while direct information on medieval population is not available, some facts and data left by contemporary chroniclers have a direct bearing on our area of study. Some information can also be gleaned from the references of foreign visitors, mostly travellers and traders, especially about cities and regions thickly or sparsely populated. But contemporary writers themselves might have made mistakes about their figures, and the validity of our conclusions based on their evidence may be open to question. Our data, assumptions, and conclusions, therefore, cannot be unquestionably correct. We can only estimate possibilities and seek limits within which truth may lie.⁴⁵

Cites (4)