Panjab Castes: 09-The organisation of the book ‘Panjab Castes’
This article is an extract from PANJAB CASTES SIR DENZIL CHARLES JELF IBBETSON, K.C. S.I. Being a reprint of the chapter on Lahore : Printed by the Superintendent, Government Printing, Punjab, 1916. |
Arrangement and contents of the caste-chapter
The rough classification adopted in Abstract No. 64 on the opposite page'^^' will serve as a clue to the arrangement of the detailed description of f he various castes. A complete index of castes and tribes will be found at the end of the volume. I shall close this part of the chapter by discussing the system adopted for the record of castes and tribes and their sub-division at the present Census, and the nature of the results obtained. The matter is one of considerable moment, and the system followed has been the subject of adverse criticism both within and without the Province. The tribal constitution of the popu lation possesses much more political and administrative importance in the
Panjab than in most other parts of Northern India, and indeed it may be said that the statistics which display it arc almost the most valuable results of a Panjab Census. The remaining- parts of the chapter will he devoted to an examination of the figures for each caste, and a description of the caste so far as my knowledge enables me to describe it. The crudeness and imper fection of this portion of the work are to me a source of great regret. It is not only that our knowledge is as nothing compared with our ignorance of the subject ; that is unavoidable. But 1 have to feel that of the information that I have collected only a portion has been utilised^ while even that portion has been hastily put on record without any attempt to arrange or digest the material. I had intended to make some attempt at classification of the various castes based in some measure upon what appeared to be their ethnic affinities, and to examine carefully the question of the probable origin of each with the help of the whole of my material ; and indeed I have carried out this intention to some extent with regard to the Biloch and Pathan tribes, the sections on which were written before orders regarding the early com pletion of the report were received. But as regards the remaining castes and tribes the time allowed me was too short to permit of any such treatment of the subject ; and I was compelled to arrange the castes roughly in classes and to content myself with stating the leading facts regarding each. The chap<er has been written backwards, beginning from the end, and I have not been able even to read over again what I had written before sending it to press. As I proceeded with the work faults in the classification became only too apparent, new lights were thrown upon what had gone before, and new facts were brought to light. There was no time to re-write what had once been written, and all that I could do was to add the new to the old. Thus I shall often be found to repeat myself, the sequence of ideas will often appear to be broken and irregular, and even conflicting statements may have escaped my notice. But the present chapter must be taken as only a rough preli minary outline of the subject. Detailed tables of tribes and clans are now in course of preparation which will embody all the sub-divisions of castes entered in the schedules of the present Census. Maps showing the distribution of the landowning castes and tribes have been prepared for each district and state and though it would have been impossible without great delay and expense to reproduce them with the present Report, I hope that the material thus collected will be more fully utilised on some future occasion. One apparent omission in my treatment of the subject calls for a word of explanation. I had prepared tables comparing the caste figures of the present with those of the last Census. But I found that the classification followed in 1868 had so evidently varied from district to district that the figures were devoid of any determinate meaning, and it would have been sheer waste of time to attempt any such comparison. To take one instance only, I find that in the Census of 1868, of 205,000 Musalman Jats returned for the Multan division, I. 188] 159,000 are in Muzaffargarh, 29,000 in Montgomery, 17,000 in Jhang, and only 63 in Multan. In Dera Ismail Khan and Shahpur this column is actually blank.
Scheme adopted for the record of castes and tribes
Unless I have utterly failed to express the facts, a perusal of the foregoing paragraphs will have made it clear that we have three main units of social and ethnic classification to deal with in the Panjab ; the caste or race, the tribe proper, and what I have for want of a better word called the section of the caste. Now these three units are of very different value in different parts of the Province and among various classes of the_ community. In the east caste is of primary importance; among the landowning com munities of the west it is little more than a tradition of ancient origin. Among the agricultural classes the tribe is most important, and in the west it is the one great fact to be ascertained ; among the priestly and mercantile classes il is almost meaningless, and what we want is the section of the caste. What we did was to attempt to record all three facts, where they existed, intending afterwards to select our figures. If we had asked for two only we should have run the risk of getting one we did not want and missing one that we did want. Of two Khatri brothers one would have returned himself as Khatri Kapur and the other as Khatri Charzati ; of two Brahman brothers one would have appeared as Brahman Sarsut and the other as Brahman Gautama ; of two Biloch brothers one would have been recorded as Biloch Rind and the other as Biloch Laghari ; tabulation would have given us wholly meaningless and imperfect figures. We therefore divided our caste column into three sub-columns headed ■' original caste or tribe, clan, and •' got ??? or sept???. Now the first difficulty we encountered was the translation of these headings. In the east qaum is used for religion and zat for caste ; in the west qaum for caste, zat for tribe or clan. In the east qot? is the universal word for tribe among the peasantry, insomuch that the na.iputs call their royal races not kul but gots ; everywhere it is used by Brahmans, Banyas and the like for the Brahminical gotra ; in the west it is unknown save in the latter sense. As for the local term for smaller tribes or clans they vary almost from district to district and from caste to caste. After consisting Commissioners we translated our headings ' asl? qaum,' ' zat ya firqah,got ya ' shdJch.' ?? [shakh?]The instructions issued for filling up these columns will be found in general letter C, Appendix D., section 5, at section 13 of the enclosed instructions to enumerators and at section 25 of the enclosed instructions to supervisors. Their general tenour was that the caste or race such as Rajput or Pathan was to be shown in the first, its principal section such as Bind, Gaur, Agarwal in the second, and its secondary sub-section such as Chauhan, Ghatwal, Bharadwaij in the third column ; that the .^'o^ if there was any was always to go into the third column ; and that where there was only one division the second column was to he left empty. The staff was warned against the loose use of the terms .Jat and Gujar as names of occupations, and it was explained that the * original caste ' column was intended to contain, not the caste of traditional origin, but the actual caste to which the people were recognized as now belonging. To these instructions was appended a sample schedule filled up by May of example.
Errors in the record of castes and tribes
I should explain that when I drafted these instructions I knew nothing of any portion of the Panjab except the Jamna? districts, and had no conception how utterly different the divisions of the population and the relations between tribe and caste were in the west of the Province. For my sample schedule I procured specimens filled up by District and Settlement Officers from all parts of the Province, and consulted many natives of different castes, yet there were several mistakes in the schedule ; in fact I believe it would be impossible to frame a set of entries which should not contain errors if judged by the varying standards current in different parts of the Panjab. More than this, there were errors in the very examples given in the instructions ; for I had not properly apprehended the nature of what I have called sections, and I did not rightly estimate the relation between the Rajput tribes of the Panjab and the great huh [kul?] or royal races. But the worst mistake of all was the use of the word asl or original with caste, and the use of the word got. The addition of asl induced many of the tribes of the western districts and Salt-range Tract to return, not their caste, or tribe as it now stands, but the Mughal, Kureshi, or other stock from which they are so fond of claiming descent ; and it doubtless tempted many undoubted Jats to record their Rajput origin. And the use of the word got set people to find out what was the Brahminical gotra of the person under enumeration. In the eastern districts the word was perfectly understood. But in tine hills and in the Western Plains it is only used in the sense of gotra. It did not matter that I had asked for got or shakh. The latter word is not commonly used in connection with family or tribe ; the former is ; and every enumerator insisted upon each person having a got. In Plach ? Ml. Anderson found a village all tutored as of one of gotra, and that an uncommon one. On inquiry from the people themselves they said they really did not know what was their got, but that some one in the village had consulted the Brahmans at Nirmand?, who told him he was of the Pethinesi ? got, and the whole village followed him. The headman of the village when asked of what got he was, could not even pronounce the word. The better and more intelligent classes know their /7o/ .9?, and others did not wish to be behind them. Now all this trouble was obviously caused by asking for i\\G? gotra. What I wanted, and what I said I wanted plainly enough in the instructions, was the tribe or sub-division of the caste ; and that the people could probably have given readily enough. What was needed was to substitute the local term, whatever it might have been, for got or shakh ; but the people knew what a got was, even if they did not know what was their got, and hence the confusion. Another great cause of error was the insistence with which the Census Staff demanded that all three columns should he filled up for each person. I had said that I only wanted two entries where there was no second sub-division, as is the case in a very large number of cases, but that did not mattter ; the columns were there with separate headings, and one after another the District Officers in their reports point out the diffi culty of getting entries for all three, the reason being that in many cases there were only facts enough for two. The result is that many of the Jats entered as the third bending the name of the Rajput tribe from which they claim to have sprung. And another most fertile cause of error must have been the efforts that were made to attain uniformity, In many districts committees were
held and n scheme of entries decided upon and prescribed for the guidance of all enumerators. have discussed the danger of all such attempts in my section on Difficulties and Suggestions in Chapter XUI under the head ' Discretion to be allowed in enumeration. ' Educated natives are almost more apt than we ourselves to go wrong in such matters for we at least are free from prejudice and are ready to admit our ignorance ; and a committee composed of the Tahsildars and Extra Assistants of a district with power to decide upon the entries of castes and tribes, would ensure with absolute certainty the ruin of a caste Census as an independent means of acquiring information.
Inherent difficulties of a record of caste
But even supposing that I had not made any mistakes in my instructions and examples, and supposing that they had been rigidly followed according to their intention, the difficulties inherent in the case are still so enormous that a really accurate record which should be correct in all its details would have been quite beyond hope of attainment. I have attempted to show in the preceding pages that it is almost impossible to define a caste and difficult to define a tribe, and that it is often impossible to draw a clearly marked line between two castes of similar standing. In fact the tribe proper is afar more definite and permanent unit than the caste. Mr. Steedman, who has criticised the scheme more severely and at greater length than any other officer, sets forth the difficulties so ably and completely that I quote the passage in full : —
With the exception of the three columns relating to caste no difficulty was found in filling the schedules up. It will be understood that my remarks regarding these three columns are solely applicable to the Western Panjab. 1 have had no experience in the Panjab east of the Ravi. Having spent three years in Gujrat, 3J in Jhang, and 2 in Dera Ismail Khan, I think that my remarks will apply to the Mahomcdan population of most districts west of the Chenab.
These three columns assume, as Mr. Finlay very truly wrote, that the zemindars know far more about their ancestry and tribal divisions that they actually do. I do not deny that the three columns could be filled up correctly for each caste by an intelligent enumerator who understood exactly what was wanted, and who was acquainted with the tribes whose members he had to enumerate ; but the Census economy prohibited the employment of men of this stamp, There are a considerable number of Mahomedan Rajputs in the Western Panjab, known as Syals or Chaddhars in Jhang, Janjuhas, Bhakhrals, Budhals, Satis, Dhunds, Alpials?, Jodras, &c., &c., in the Rawalpindi Division. Now any member of these tribes if asked what his ' kaum ' was would reply Bhakhral or Sati, &c., as the case might be. Or he might very probably give the sub-division to which he belonged. A Syal would be sure to answer thus. You would in nine cases out of ten have to put some distinctly leading question before you ascertained whether he claimed to be a Rajput or not. The result is that sometimes Rajput the ' asl kaum, ' some times ' Syal ' the clan, and sometimes Chachkana the sept or family, is entered in the first of the three sub-divisions of column 7 : I noticed many entries of this description. In fact most of the Rajputs of this district would give Rajput as their ' got, ' placing their tribe as the ' asl kaum.' Entries of this description naturally depreciate the tabulation results considerably.
Similar errors crept into the entries of the village artisans. A man may ply the trade of a weaver, oil-presser, or shoe-maker without being a weaver, oil-presser, or shoe-maker, by caste. In Jhang weaving had been taken to as a livelihood by many persons who were not of the weaver tribe. Yet many of these I have no doubt will be put down as weavers in the * asl kaum ' column. Again men of these low castes are very fond of claiming relationship with the higher tribes, especially those of Rajput origin. I saw many entries such as these — 'asl kaum' Mochi * zat ' Janjuha, Bhatti. Awan, &c. Now Janjuhas and Bhattis are Rajputs. If the Mochi was a Janjuha originally his ' asl kaum ' is Rajput, his zat Janjuha, and shoe-making is his trade. If he is a Janjuha by fiction then Janjuha must be put down as he states. Shekhs, i.e. converted HIndus, or men of low caste who have risen in the world, also advance most ungrounded claims in the way of descent Apparently there is no escape from these difficulties in the case of village artisans, Shekhs, and other similar tribes ; but in the case of agriculturists I think more definite instructions would havo left the tabulation entries much more trustworthy.
I now venture to criticise some of the specimen entries attached to the enumerator's in structions. The entries opposite the name of Mahomed Ibrahim are 1, Rajput ; 2, Syal ; 3, Panwar.^ I can confidently assert that not one man in a hundred of the Syals is aware that he is a Panwar Rajput. I wonder if there are ten men who have heard they are descended from this got of the Rajput tribe. I know exactly what answers an enumerator would get from a representative Syal zamindar. Question. — What is your tribe (A;aM»I ?) ? Answer. — Bharwana -. •' Question. — What is your clan (zat) ? Answer. — Syal. Question. — What is your family {got or Shake Answer.— QtoHi ?only knows. He will inevitably give his sub-division as his asl kaum
- This is one of the mistakes I have already referred to. The entry should have been ^' Mdj' ?
p tit?—Funwdr ?— Sid I.? and his clan as his zat. Nothing less than a direct question as to whether he is a Rajput or a Jat will elicit from him the fact that he is a Rajput. As for ' got ' he probably has never heard the word. The truth is that the present Mahomedan tribes of the Western Pan.-jah, though immigrants from Hindustan, have forgotten their 'gots' entirely and very often their' asl ' kaum.' In some few instances only is the name of the ' got ' preserved, and then the tribesmen are quit« unaware that their tribal name is that of their old ' got.'
The next question is. What arc the asl kaums in each district ? I notice that in one of the specimen entries Gujar is so entered. There are various theories as to whether the Gujar is a separate tribe of Tartar or Hindu origin or whether it is an offshoot of the great Jat tribe. In Jhang and D era Ismail Khan and Shahpur the Mahomedan agriculturists are usually divided into Rajputs and Jat? in local parlance. I mean that if a Rajput is asked whether he is a Jat he will at once deny it, while a Jat admits that he is a member of the tribe. I do not mean to assert that, excluding Rajputs and other tribes who have migrated from the other side of the Indus, all other agriculturists must be Jats ; but if they are not I ask who are the numerous tribes who reside in the Chach and Sind Sagar Doabs and along the left hank of the Chanab ? What is their asl Kaum'^ Their Hindu origin is undoubted. They are not Rajputs. If' they were they would claim their relationship. I have not room here to go fully into this question. I have noticed it in the Final Report of the Jhang Settlement. But my object is I think attained, and that is to indicate how vcry necessary it is that instructions should be given separately for each district as to « hat tribes are to be considered' Asl kaum.' Take the Khokhars. They are an influential tribe in Jhelam, Shahpur, and Gujrat. Are they converted Rajputs as many claim, or descendants of the son-in-law of the prophet as the Shahpur Khokhars state, or mere Jats as their enemies allege. In the second case only can they be an asl kaum. If in the tabulation of different districts the tribe is sometimes entered as an ' asl kaum ' and at others as a branch of the Rajput and Jat tribes, the results are likely to be misleading. Then again there are tribes who are admittedly of ancient standing and yet have no traditions. Who are these ? It is not unlikely that they were the original inhabitants before the immigration of the Hindu settlers. As far as my limited experience goes I think it would be an easy matter to settle this point beforehand for all the main tribes of each district, and also to give a few general instructions as to how doubtful tribes were to be treated. The question Are you a Rajput or a Jat ? would clear up most cases of doubt where the tribe was originally Hindu, the enumerator being warned of the custom of calling all agriculturists Jats. Then all tribes who came from the other side of zxc the Indus would also be ' a«l kaum, ' the Pathans, Biloches, Muglials, &c. The village Kamins would al=o be included in the same list. Here the enumerators would he warned to ask the individual whether he was a Kamin by trade only or both by trade and tribe.' I would arbitrarily class all agriculturists who admitted that they were not Rajpvits and who were of undoubted Hindu origin, as Jat^. This clas'^ification is perhaps not ethnologically accurate, but every Patwari and most zamindars would understand what is meant. I think too for the Mahomedan population two columns would have been enough. It seern'^ unnecessary to ascer tain the numbers of each sub-division. We want to know the total Syal, Ghakkar, and Awan population. I do not think much is gained by working out returns showing the total popnlation of the Bharwana, Chvichkana, Adnial, Firozal, and Bugdial families. There are no restrictions on iotermarriage between members of the different families.
I have already explained the reason why three columns were taken instead of two. We wanted two facts only ; but we wanted to make sure of getting them in the many cases where three facts were available and one was not wanted, by recording all three and rejecting for ourselves the useless one ; otherwise if we had had two columns only, one of them might have been wasted on the useless fact. As it was, one of our three C'^himns was commoidy occupied by the name of some wholly unimportant sept or family. And I do not agree with Mr. Steedman in his proposal to issue detailed instructions concerning the agricultural tribes of each district. Who is to issue them ; and how is it to be ensured that the same tribe is classed similarly in two (Cerent districts ?
Reasons why the scheme did not work
I think that on the whole the scheme was the be st that could have been adopted ; and if it had been possible to carry it out to the end as it had been intended to do when the instructions were framed, I believe that results of very con siderable accuracy would have been obtained. What was intended was this — to record every thing;, to tabulate all the entries, and /^f-u to classify them throughout and produce the results as the final caste table. Thus, supposing one man bad entered himself as Jat Bhatti and another as Rdjput Bhatti, or one man as Qureshi Khattar, another as Awan Kbattar, and a third as Qutbshahi Khattar, we sho uld have tabulated them all separately, and then classed them as might be decided upon after cons; lertitiun and inquiry. It was not expected that the material would be properly arranged in tl c schedules ; but we hoped that it would all be recorded
1 Would not this suggest to the artisan the setting up for himself of a mythical origin from some aaste of glorious renown ? there, to be arranged afterwards. But when we came to examine the schedules we found that the separate entries in the caste column alone were numbered by thousands, while the sub divisions were nnmlicred by tens of thousands. I certaudy had not, and I do not believe that any body else concerned had, the very faintest conception of how numerous the entries would be. At any rate it was obviously quite out of the ((uestion to tabulate and examine them all before com pilation ; and what was done was to deal with the entries in the first or caste column only, so far as the compilation of the final Census Table VUI was concerned. Even those entries I was com jielled, for reasons given in the Chapter on Tabulation, to allow the Divisional Officers to class'fy for themselves where there appeared to be no reasonable doubt as to the classification. With the headings for which they returned separate figures I dealt as is described in the Cliapter on Com pilation. The figures for the sub-divisional entries were tabulated in detail ; but only certain selected entries were taken out to bo used in the Census Report, the principles on which the selection was made being explained in the Cliapter on Compilation.
Nature and degree of error in the flnal figures
Thus the figures as now given in the abstracts and appendices of this report are liable to error in several ways. In the first place many members of a caste or tribe entered as their caste some race to which they are pleased to refer their origin in remote antiquity. For instance, some Gakkhars returned themselves as Gakkhar and others as Mughal, and are shown under those headings respectively in the final tables, which there fore do not give the total numbcr of Gakkhars in the Panjab. So some low caste men returned their caste as Eajput or Mughal or Quiesh ' out of joke ' a-i several Deputy Commissioners note. On the other hand some men of good caste, such as Sial, Khokhar, or Mughal, who were following the trade of weaver or carpenter, returned their caste as Paoli or Tarkhan, though the adoption of that hereditary occupation had been in many cases too recent to have brought about a change of caste. This last error was for the most part confined to the Western Plains. Again, persons who belonged to the same tribe and returned that tribe as their caste will have been duSerently classed in different divisional offices, or classed under one heading in one division and returned separately and then classed by myself under another heading in another division. Thus the Bhattis will have been classed as Jats by the Dcrajat and as Rajputs by the Eawalpindi office. So the Langahs were classed as Jats in Multan, while the Derajat returned them separately and I classed them as Pathans. These errors however affect only those cases where the tribe was returned and not the caste. Where a man returned himself as Jat, Rajput, Pathan and so forth, he was treated as such although the tribe he gave might raise suspicion as to the correctness of the returns. Moreover the errors, if they must he so called, do represent actual facts. The IShatti is a Rajput in Rawalpindi because there Rajputs are recognised. In the Dcrajat he is a Jat, because there no distinction is drawn between Jat and Rajput. And it must be remembered that though the cases in which the errors detailed above occurred are numerous^ the total figures affected are seldom large. There were certainly hundreds, I believe there were thousands of so-called castes returned in the Multan division which only included ten or fifteen people in the whole division. The great mass of each caste returned themselves rightly and are shown correctly in our tables : the items that are wrongly clashed are wholly insignificant in their total amount as compared with the items that are rightly classed. But there are exceptions to this statement. The distinction between Jat and Eajput is so irdefinite and so variable that it can hardly be called a mistake to class a tribe as Jat in one place and Rajput in another. This however has been done. But I have picked out the figures in each case and put them side by side in the ah-tracts contained in the Fcction on these two castes, and I think the error which has not been corrected may be taken as exceedingly fmall. It is now in each man's power to transfer the figures for any tribe from Jat to Rajput or vice versd, according to individual taste. The other chief exceptions are in the case of Mughals and Shekhs. For Shekhs I was prepared. I knew that all sorts of low caste men, recent converts to Islam, would return themselves as Shekh ; and I had the figures examined with a view to separate these, and the details will be found in the text of this chapter. But I did not know that in some parts of the western Panjab Mughal was as favourite a supposititious origin as Shekh is m other parts of the Province, and I have not had the details worked out so carefully. Still almost all the large numbers have been separated from these two entries. So with Pathims. Many people, such as Dilazak, have returned them selves as Path<ans who do not really belong to the race ; but their claim to the name is often admitted, and they have become in a way affiliated to the nation. Thus the considerable errors in the caste tables, as corrected in this chapter, anu unt to this ; that there is a confusion between Jat and Raj put and between Pathdn and certain allied races, which exists in actual fact fully as much as in the figures ; that some tribes or castes have been wrongly shown as Mughal and Shekh ; and that some ot the artisan castes have been shown as belonging to the higher castes, A while some of the higher castes have been included in the artisan castes merely because they followed their occupation. Taking the Province as a whole the errors are probably insignificant, and hardly affect the general distribution of the population by caste. They are probably greatest in the cis-Indus Salt range tract, where the tendency to claim Mughal origin m strongest.
Error in the figures for tribes and sub-divisions
The figures for tribes and Bub-divisions given in this chapter are professedly only rough approximations. The manner in
which they were tahulated and the final figures compiled will he explained in Book 1 1 under the heads Tabulation and Compilation. The whole process was intended to he merely a rough one. The detailed tabulation is now in progress, and I hope wu.hin the next few years to bring out detailed tables of tribes and clans for the whole Panjab. Rut beudes inaccuracies that will have crept into the wcrk of tabulation, there are several causes of error inherent in the material. In the first place the spelling of local names of tribes, a rendered by the enumerating staff, varied extraordinarily. Some were evidently mere variations, as Dhariwul, Dhaliwal and Dbaniwal ; some I knew ito represent separate tribes, as Sidhu and Sindhu, Chhrna and Clu'ma ; some I an\ still in doubt about, as Buta and Bhutta, Sara and Sarai. In working with a staff not always acquainted with the names of the clans, figures referring to two different tribes must often have been .-joined together, and other figures wrongly omitted because of some variation in the spelling. Another source of error doubtles-: was the uncertainty regarding the woman's clans discussed in section 354. On a future occa-iou I would tabiJate sub divisions of castes for males only. Again many of the people are presented twice over in two columns. Thus the Sial are Puuwar Eajputs by origin. Suppose that 1,000 Sials returned themselves as Rajput Punwar Sial, auother thousand as Sial Puuwar, another 1,000 as Eajput Sial, and a fourth 1,000 as Rajput Punwar. All the 4,000 people would be shown in Table' VUI as Rtbput ; but in the details of tribes we should have 3,000 Sial and 3,000 Punwar or 6,000 in all. This was quite unavoidable so long as only one tribal division was tabulated ; but as a fact the cases in which this happened were few, or at least the numbers affected small. I had all cases in which the same people were entered twice over shown in a separate memorandum attached to the tribes table, and wherever the numbers were at all con siderable I have mentioned the fact of their double inclusion in the text. This double entry occurred most often with the Jat tribes, who, in order to fill up their three columns, entered the Rajput tribe from which they claimed origin as well as their own Jat tribe, so that we had peDple returning themselves ai Jat Sidhu Bhatti, and sucli people appear among the Jat tribes both as Sidhu and as Bhatti.
Proposals for next Census
What then is best to be done at next Census ? It will be seen that many of the difficulties are due to the intrinsic difficulty of the question and to the varying nature of caste in the Pan jab. So far as this is the case no scheme will help us. In one respect, however, I hope that the task will be made much easier by next Census. I hope by then to have brought out classified lists of all the tribes and clans returned in the present Census. The way in which they will facilitate the treatment of the subject is explained in the section on Tabulation. If I had had such a classified list my task on this occasion would have been easy enough ; and it is I think one of the most valuable results of the present Census that it has given us materials for the preparation of such a list. With such a list the three columns of the schedule of 1881 are almost perfect in theory. But 1 do not think they worked as well in practice. I believe that the three columns which they erroneously thought they were hound to fill up, puzzled both people and staff, and caused a good many of our difficulties. Thus in future I would have but two columns, and would head them Quam and Shakh. I would not care whether caste or tribe was entered in the first column, as the classified list would show the tabulator how to class the tribe J and I would hope that the second column at any rate would generally give tribe. In very many cases it would not. There would be entries like Biloch Rind instead of Biloch Laghari, Brahman Bashisht instead of Brahman Sarsut, Banya Kasih instead of Banya Agarwal, and so forth. But on the whole I think it would be better to accept the fact that the entries must he incomplete, whatever scheme be adopted ; and would prefer the certainty of error of the two columns, rather tb an the confusion and perplexity which the three columns cause to those concerned in the enumeration. Above all things I would avoid the words asl and oot. I would let the patwaris, who should make the preliminary record, exercise their discretion about entering high castes for menials or artisans, directing them to show the caste by which the people were commonly known in the village. I would tabulate both males and females for tribes and clans, and arrange them in order of numbers ; and 1 would have the Deputy Superintendent personally examine the tribal tables for all above say 500, before compiling his final caste tables. Such an examination would do an immense deal towards increasing the accuracy of the caste figures ; but it was impossible in the present Census owing to the double sub-division. I would show in njy tribal tables the figures for males only, though those for females must he tabulated in the first instance in order to allow of transfer of entries from one caste heading to another,
Bibliography
The most detailed and accurate information available in print regard ing certain, and those the most important from an administrative point of view, of the Panjab castes is to be found in the numerous Settlement Reports, and more esjiecially in those of recent years. Unfortunately they deal almost exclusively with the landowning and cultivating castes. Sir H. Elliott's Races of the N. W. P., edited by Mr Beames, is, so far as it goes, a mine of information regarding the castes of the eastern districts. Sherring's Ilindu Castes contains much information of a sort, the first volume being really valuable, but the second and third being infinitely less so ; while the whole is rendered much less useful than it might be by the absence of any index save one that maddens the anxious inquirer. On the ancient form of the institution of Caste, Wilson's treatise on Indian Caste, and Vol. I of Muir's Sanskrit Texts are the authorities. The second volume of General Cunningham's Archcological Reports has a dis sertation on Punjab Ethnology by way of introduction, and there are many small pamphlets which contain useful information. But on the whole it is wonderful how little has been published regarding the specially Panjab castes, or indeed regarding any of the menial and out cast classes. Sir Geo. Campbell's Indian Ethnology I have not seen ; but it should be instruc tive. At the head of the section on Pathaus and Biloches I have noticed the books which may be most usefully consulted. In the case of the other castes I know of no works that deal with any one particidar, or indeed with our Paujab casi^es in general save those specilied above.
38
panjab castes.
