Walter of Henley’s scheme.

It is a strenuous and sanguine, if not an impossible, programme. When harvest time and the holy weeks are omitted, the plough is to ‘go’ every week-day throughout the year, despite frost and tempest. Obviously it is a programme that can only enter the head of an enthusiastic lord who has supernumerary oxen, and will know how to fill the place of a ploughman who is ill. We have little warrant for believing that what Walter hopes to do is being commonly done in his day, less for importing his projects into an earlier age. In order that he may keep his beasts up to their arduous toil, he proposes to feed them with oats during half the year[1335]. If we inferred that the Saxon invaders of England treated their oxen thus, we might be guilty of an anachronism differing only in degree from that which would furnish them with steam-ploughs. But, to come to much later days, the Domesday of S^t. Paul’s enables us to say with some certainty that the ordinary team of eight beasts accomplished no such feats as those of which Walter speaks. For example, at Thorpe in Essex the canons have about 180 acres of arable land in demesne. These, it is estimated, can be tilled by one team of ten heads together with the ploughing service that is due from the tenants, and these tenants have to plough at least 80 acres, to wit, 40 in winter and 40 in Lent[1336]. We must observe that to till even 120 acres according to Walter’s two-course plan would mean that a plough must ‘go’ 180 acres in every year, and that, even if it does its acre every day, more than half the week-days in the year must be devoted to ploughing. We may, however, seriously doubt whether a scheme which would plough the land thrice between every two crops had been generally prevalent[1337]. Nay, we may even doubt whether the practice of fallowing had been universal[1338]. Not unfrequently in our cartularies the villein is required to plough between Michaelmas and Christmas and again between Christmas and Lady Day, while nothing is said of his ploughing in the summer[1339]. We are only beginning to learn a little about medieval agriculture.

However, we have now said all that we had to say by way of preface to what we fear will be a dreary and inconclusive discussion of some of those abundant figures that Domesday Book supplies. A few we have endeavoured to collect in the tables which will meet the reader’s eye when he turns this page, and which will be explained on later pages.

§ 2. Domesday Statistics.

TABLE I. STATISTICS

Continued

TABLE II. AVERAGES.

Northern formulas.

We may begin our investigation with a formula common in Derbyshire.

In M [place name] habuit K [man’s name] a car[ucatas] terrae ad geldum. Terra b car[ucarum or carucis]. Ibi nunc in dominio d car[ucae] et ... villani et ... bordarii habent e car[ucas].

The Lincolnshire formula is perhaps yet plainer. Instead of saying ‘Terra b car[ucarum],’ it says, ‘Terra ad b car[ucas].’ Still more instructive is a formula used in Yorkshire.

In M habuit K a car[ucatas] terrae ad geldum ubi possunt esse b car[ucae]. Nunc habet ibi K d car[ucas] et ... villanos et ... bordarios cum e car[ucis].

As a variant on the phrase ‘ubi possunt esse b car[ucae],’ we have, ‘quas potest arare 1 car[uca],’ or ‘has possunt arare b car[ucae][1340].’

The teams on the demesne (d) and the teams of the tenants (e) are enumerated separately. The total number of the teams (d + e) we will call c.

Now occasionally we may find an entry concerning which the following equation will hold good: a = b = c: in other words, the same number will stand for the carucates at which the manor is taxed, the ‘teamlands’ that there are in it (or to put it another way the number of teams that ‘can be there,’ or the number of teams that ‘can plough it’[1341]) and also for the teams that are actually to be found there. Thus:—

Terra Roberti de Todeni.... In Ulestanestorp habuit Leuricus 4 car[ucatas] terrae ad geldum. Terra totidem car[ucis]. Ibi habet Robertus in dominio 1 car[ucam] et 6 villanos et 3 bordarios et 8 sochemannos habentes 3 car[ucas][1342].

Here a = b = c. But entries so neat as this are not very common. In the first place, the number (c) of teams often exceeds or falls short of the number (b) of ‘teamlands,’ or, which is the same thing, the number of teams that there ‘can be.’ An excess of ‘teamlands’ over teams is common. In some parts of Yorkshire and elsewhere instead of reading that there are so many teams, we read ‘modo vasta est’:—there are no oxen there at all. But the reverse of this case is not very uncommon. Thus we may be told that there are 3 carucates for geld, that ‘there can be there 2 teams’ and that there are 4 teams[1343]; we may find a manor that contains land for but 3 teams equipped with as many as 7[1344]. As to the relation between a and b, this is not fixed. On one and the same page we may find that a is equal to, greater and less than b. Thus in Lincolnshire[1345]:

Southern formulas.

Leaving now for a while the carucated part of England and postponing our visit to Kent, we find similar formulas. They tell us (A) that the manor contains a certain number of units of assessment, (B) that there is land for a certain number of teams, (C) that there are so many teams upon it. But we have a new set of units of assessment; instead of carucates and bovates, we have hides and virgates. The Huntingdonshire formula is particularly clear. It runs thus:

In M habet K a hidas ad geldum. Terra b car[ucarum or carucis]. Ibi nunc in dominio d car[ucae] et ... villani et ... bordarii habentes e car[ucas].

The number of hides that is put before us is the number of hides ‘for geld.’ So in Cheshire and Shropshire the number of hides that is put before us is the number of ‘hidae geld[antes].’ From this we easily pass to the formula that prevails in Wiltshire, Dorset, Somerset and Devon:

K tenet M. T[empore] R[egis] E[dwardi] geldabat pro a hidis. Terra est b car[ucarum]. In dominio sunt d car[ucae] et ... villani et ... bordarii cum e car[ucis].

A formula common in Sussex, Surrey and several other counties instead of telling us that this manor has a hides for geld, or has a gelding hides, or gelds for a hides, tells us—what seems exactly the same thing—that it ‘defends itself’ for a hides. Then we pass to counties such as Middlesex, Hertford, Buckingham and Oxford where the entry does not commonly use any words which explicitly refer to geld:—we are told that K holds M for so many hides (pro a hidis). Lastly, we may pass to counties, such as Warwickshire and Staffordshire where, at first sight, the entries may seem to us ambiguous. They run thus—‘K holds M. There are there a hides. There is land for b teams.’ Here for a moment it may seem to us that we have two different statements about the actual extent or capacity of the manor:—there are a hides there, but land for b teams. But comparing the formulas in use here with those in use in other counties, we can hardly doubt that they all come to one and the same thing:—a statement about b, the capacity of the manor, is preceded by a statement about its taxation, which statement may take the short form, ‘There are a hides there,’ instead of one of the longer forms, ‘It gelds, or defends itself, for a hides,’ or ‘He holds a gelding hides, or a hides for geld.’

Kentish formulas.

In Kent again, we have the three statements, though here the units of assessment are sulungs and yokes:—the land ‘defends itself’ for a sulungs; there is land there for b teams; there are d teams in demesne and the men have e teams.

Relation between the three statements.

In the hidated south, as in the carucated north, the relation between the three amounts is not invariable. We may find that a = b = c. It is common to find that c is less than b, but occasionally it is greater; on one and the same page we may find that c is equal to, is greater, is less than b. Then a is often equal to b, often it is less than b, but sometimes it is greater. We have therefore three statements about the manor, between which there is no necessary connexion of any very simple kind.

It may look pedantic, but will be convenient if, by means of the letters A, B and C, we try to keep distinctly before our minds ‘the A statement’ about the units of assessment, ‘the B statement’ about the ‘teamlands,’ or teams for which ‘there is land,’ and ‘the C statement’ about the existing teams. We shall find hereafter that there are certain counties in which we do not get all three statements, at least in any of their accustomed forms. In Gloucestershire, Worcestershire and Herefordshire we rarely get the B statement. As to Essex, Norfolk and Suffolk, we seem at first sight to obtain A and not B, or B and not A, while Leicestershire will require separate treatment.

Introduction of statistics.

Now if we are ever to understand these matters, it is necessary that we should look at the whole of England. Far be it from us to say that microscopic labour spent upon one county or one hundred is wasted; often it is of the highest value; but such work is apt to engender theories which break down the moment they are carried outside the district in which they had their origin. Well would it be if the broad features of Domesday Book could be set out before us in a series of statistical tables. The task would be gigantic and could hardly be performed except by a body of men who had plenteous leisure and who would work together harmoniously. However, rather to suggest what might and some day must be done, than to parade what has been done rapidly and badly, some figures have been set forth above in two tables[1346]. That they are extremely inaccurate can not be doubtful, for he who compiled them had other things to do and lacks many of the qualities which should be required of a good counter of hides. For unmethodical habits and faulty arithmetic no excuse is possible; but it will be remembered that, as matters now stand, two men not unskilled in Domesday might add up the number of hides in a county and arrive at very different results, because they would hold different opinions as to the meaning of certain formulas which are not uncommon. What is here set before the reader is intended to be no more than a distant approach towards the truth. It will serve its end if it states the sort of figures that would be obtained by careful and leisurely computers, and therefore the sort of problems that have to be solved[1347].

Explanation of statistics.

Sidenote: Acreage.

We must now explain our statistics. In Column I. we give the acreage of the modern counties[1348]. A warning bracket will remind the reader that in the cases of Yorkshire, Cheshire and Rutland the modern does not coincide even approximately with the ancient boundary. To Middlesex we give a figure larger than that given by our statisticians, for they know a county of London which has been formed at the expense of its neighbours[1349]. Many minor variations should be remembered by those who would use Domesday Book for delicate purposes; for example, they must call to mind the merger in circumambient shires of what were once detached pieces of other counties. But of such niceties we can here take no account[1350].

Population.

In Column II. we state the ‘recorded population’ as computed by Ellis. In the cases of Dorset and Somerset we also state, and we sign with the letter E, the result of Eyton’s labours. We must not forget that these figures give us rather the number of tenants or occupiers than the number of human beings. Our readers must multiply them by four, five or six, according to knowledge or taste, before the population of England will be attained.

Danegeld.

In Column III., for a reason that will become evident hereafter, we place the amount of danegeld charged against the counties—charged against them, not actually paid by them[1351]—in the middle of the twelfth century. The sources of these figures are the Pipe Rolls of 31 Henry I. and 2 and 8 Henry II. In these accounts the amount charged against a county is approximately constant. Some of the variations are probably due to a contemptuous treatment of small sums[1352]; but there are cases in which a sheriff seems to have been allowed to deduct £10 or so, without any recorded explanation[1353]. We choose the highest figures when there is any discord between our three rolls. The danegeld was being levied at the rate of two shillings on the hide, and therefore, if we would find the number of geldant hides, we have to multiply by ten the number of pounds that are set against the county.

Hides, carucates, sulungs.

Column IV. contains our estimate of A: in other words, of the number of hides, carucates or sulungs. As we are arguing for a large hide, we have thought right in doubtful cases to lean in favour of inclusion rather than of exclusion. We count all hides, except those ascribed to the shire’s boroughs[1354], even though we are told that they have ‘never’ gelded. Also, when a hide is mentioned, we count it, even though we have a strong suspicion that the same hide is mentioned again on some other page. Especially in Sussex, where the rapes have recently been rearranged, this may make our figures too high[1355]. Then, again, we have frankly begged important questions by assuming that in Domesday Book the following equations are correct.

In the counties with which we have dealt, except Norfolk and Essex (Suffolk we have left alone), acres are so rarely mentioned that the error, if any, introduced by our hypothesis as to their relation to hides and carucates will be almost infinitesimal, and, even if we are wrong in supposing that the virgate is the quarter of a hide and that the bovate is the eighth of a carucate, the vitiation of our results that will be due to this blunder will but rarely be considerable[1356].

Reduced hidage.

Almost everywhere we may find some hides (carucates, sulungs) that do not geld and many cases in which a tract now gelds for a smaller number of hides (carucates, sulungs) than that for which it formerly paid. In four counties, however, Sussex, Surrey, Hampshire and Berkshire, we see that since William’s advent there has, rightfully or wrongfully, been a large and generally distributed reduction in the tale of the gelding hides. In our Column V. we give a rough statement of the reduced number[1357]. In Cornwall we read of an assessment that prevailed in the Confessor’s day and of a heavier assessment. The figures which speak of this heavier assessment we place in our Column V[1358].

The teamlands.

We now pass from A to B. In Column VI. we set the number of teamlands, thus answering the question Quot carucarum [carucis] ibi est terra. We have assumed, but this rarely has an appreciable effect on our calculations, that the land of one ox is the eighth, the land of two oxen the fourth part of the land of one team. There are certain counties where we receive no statement about the teamlands, while in certain others the statement, though it seems to be expected, is often omitted[1359]. For this reason some blanks will be found in this column. In most of the other counties instances occur with more or less frequency in which nothing is said of the teamlands. In these cases we have thought it fair to assume that there were teamlands equal in number to the teams (B = C). The effect of this assumption will be to bring the number of teamlands (B) somewhat closer to the number of teams (C) than it would otherwise have been, but no very great harm will have thus been done to our rude statistics[1360].

The teams.

Column VII. gives the number of teams. Here we assume (we shall endeavour to prove hereafter) that the caruca of Domesday Book always means the same, namely, eight oxen[1361].

The values.

Lastly in Column VIII. we place the results attained by Pearson[1362] and Eyton in their endeavours to add together the various sums which the various estates in a shire are said to be worth (valet) or to render (reddit) in the Conqueror’s day, and to thus obtain a total valet for the shire. We need hardly say that these values are ‘annual values.’

The table of ratios.

The relations between our divers sets of figures are more important than the figures themselves, therefore we have worked the division sums the results of which are printed in the second Table, the first seven columns whereof are filled by quotients[1363]. The last column calls for more remark. The valets obtained for the various counties by Pearson and Eyton are somewhat precarious. They involve theories as to the relation between the values of gold and silver, as to the relation between the value of a pound reckoned by tale and a pound reckoned by weight, as to ‘blanched’ money and the cost of ‘a night’s farm.’ Also a good deal is included that can hardly be called the value of land, since it comprehends, not only the value of mills and the like, but also in some cases the revenue derived from courts. In order therefore that we might compare the values given to land in the various counties, we have taken at hazard a number of small estates in order that we might by addition and division obtain the value of a typical teamland with typical appurtenances. In general we have chosen ten estates each of which has one teamland, ten estates each of which has two teamlands and ten estates each of which has five teamlands, and then we have divided the sum of their values by eighty, the number of teamlands that they comprise. On the whole, the figures that we thus obtain and place in Column XVI. are not widely removed from those in Column XV., which represent the quotients arising from a division of Pearson’s ‘county values’ by the number of teamlands that are contained in the counties[1364].

An apology.

In order that not too much credence and yet just credence enough may be given to the figures that we have hastily put together, we will set beside those that we have stated for Gloucestershire the results of a minute analysis accomplished by Mr Charles Taylor[1365]. We have set down: Population, 8366 (from Ellis); Hides, 2388; Teams, 3768; Total Valet, £2827 6s. 8d. (from Pearson). Mr Taylor gives: Population, 8239[1366]; Hides, 2611 (or 2596); Teams, 3909; Total Valet, £3130 7s. 10d. Now these variations are wide and may in some sort be discreditable to those who differ from Mr Taylor[1367]. But they are not very substantial if we come to averages and ratios and a comparison of counties. For the purposes for which we shall use our figures, it is no great matter whether in this county there are 2·1 or 2·2 ‘recorded men’ to the plough-team[1368]. The broad features of Gloucestershire are that its hides fall far short of its teams, that its recorded population is sparse, that the average value of the land tilled by a team falls well below twenty shillings, that this shire differs markedly and in certain assignable respects from Wiltshire, where the hides exceed the teams, from Lincoln, where, despite the fen, the population is thick, from Kent, where the average value of land tilled by a team rises above thirty shillings[1369].

Constancy of ratios.

Our figures tell of wide variations; but we may be allowed to call attention to the stability of certain ratios, a stability which is gratifying to the diffident arithmetician. In twenty-one counties we can divide ‘the recorded population’ by the number of teamlands. The quotient never falls as low as 2 and only twice exceeds 4[1370]. For the same twenty-one counties we can divide the number of teamlands by the number of teams. Only twice will the quotient fall below 1 and only once will it touch 2. We must not, however, be led away into a general discussion of these figures. That task would require a wary and learned economist. We must keep our minds bent on what may be called the A B C of our subject[1371].

The team.

Now we may start with what seems to be the most objective of our three statements, that which gives us C, the number of teams. We know that in A there is an element of estimation, of assessment; we may fear that this is true of B also; but an ox or a team ought to be a fact and not a theory. At the outset, however, a troublesome question arises. We have assumed that whenever our record speaks of a caruca it means eight oxen. On the other hand, there are who maintain that whereas the carucae of the demesne consisted of eight, those ascribed to the villeins comprised but four oxen[1372], and others have thought that the strength of Domesday’s caruca varied from place to place with the varying practice of divers agriculturists.

Variability of the caruca.

But, in the first place, it is abundantly clear that the clerk who compiled the account of Cambridgeshire from the original verdicts held himself at liberty to substitute ‘half a team’ for ‘four oxen’ and ‘four oxen’ for ‘half a team[1373].’ In the second place, the theory of a variable caruca would in our eyes reduce to an absurdity the practice of stating the capacity of land in terms of the teams and the oxen that can plough it. We are carefully told about each estate that ‘there is land for b teams, or for b´ oxen, or for b teams and b´ oxen.’ Now if a ‘team’ has always the same meaning, we have here a valuable truth. If, on the other hand, a ‘team’ may mean eight or may mean four oxen, we are being told next to nothing. The apparently precise ‘there is land for 4 teams’ becomes the useless ‘there is land for 32 or 16 or for some number between 32 and 16 oxen.’ What could the statesmen, who were hoping to correct the assessment of the danegeld, make of so vague a statement? They propose to work sums in teams and teamlands. They spend immense pains in ascertaining that here there is ‘land for half a team’ or ‘land for half an ox.’ We are accusing them of laborious folly unless we suppose that they can at a moment’s notice convert teams into oxen.

The caruca a constant.

If it be allowed that in the statement (B) about the number of teamlands the term caruca has always the same meaning, we cannot stop there, but must believe that in the statement (C) about the number of teams this same meaning is retained. Often enough when there is equality between teamlands and teams (C = B), the entry takes the following form:—There is land for b teams and ‘they’ are there[1374]. What are there? The teams for which ‘there is land’: those teams which are serving as a measure for the capacity of land. Let us try the two modes of interpretation on the first lines that strike our eye. Here we have two successive entries, each of which tells us that ‘there is land for 6 teams[1375].’ If the caruca is a constant, we have learnt that in one particular there is equality between these estates. If the caruca is a variable, we have learnt nothing of the kind. Let us see what we can gain by reading further. In the one case there were 3 teams on the demesne and the villeins had 6¹⁄₂; in the other there were 2 teams on the demesne, the villeins had 2 and the sokemen 2. We want to know whether the second of these estates is under-teamed or over-teamed. There is land for 6 teams and there are 6 teams on it; but 2 of these teams belong to villeins and 2 to sokemen. If we give the villeins but 4 oxen to the team, how many shall we give the sokemen? Shall we say 6? If so, there are 36 oxen here. Is that too many or too few or just enough for the arable land that there is? That is an unanswerable question, for the king’s commissioners have been content with the statement that the number of oxen appropriate to this estate lies somewhere between 23 and 49

The villeins’ teams.

Surely when we are told that 8 sokemen have ‘2 teams and 6 oxen’ or that 9 sokemen and 5 bordiers have ‘3 teams and 7 oxen[1376],’ we are being told that the teams in question have no less than eight oxen apiece. Surely when we are told that there are 23 villeins and 5 bordiers with 2 teams and 5 oxen[1377], we are being told that the teams of these villeins are not teams of four. And what are we to say of cases in which a certain number of teams is ascribed to a number of persons who belong to various classes, as for example when 6 villeins and 7 bordiers and 2 sokemen are said to have 3 teams and 5 oxen[1378], or where 3 villeins, 2 bordiers, a priest and a huntsman are said to have one team and 6 oxen[1379], or where 19 radknights ‘with their men’ are said to have 48 teams[1380]? Even if we suppose that the officers of the exchequer have tables which tell them how many oxen a caruca implies when it is attributed to a Northamptonshire sokeman or a Gloucestershire radknight, we are still setting before them insoluble problems. The radknights of Berkeley ‘with their men’ have 48 teams:—this may cover less than 200 or more than 300 oxen. And yet the record that is guilty of this laxity will tell us how in Bedfordshire Terra est dimidio bovi, et ibi est semibos[1381].

The villeins’ oxen.

The main argument that has been urged in favour of a variable caruca is that which, basing itself on later documents, protests that a villein ought not to have more than two oxen[1382]. Now true it seems to be that if by the number of the teams belonging to the villani and bordarii of Domesday Book we divide the number of villani plus half the number of bordarii (and this would be a fair procedure), we shall obtain as our quotient a figure that will be much nearer to 2 than to 4. But it must be common ground to all who read our record that some villeins are much better supplied with oxen than are their neighbours, and that some villeins have whole teams, whatever a ‘team’ may mean. There is so much difference in this respect between manor and manor that we are not justified in talking of any particular number of oxen as the normal outfit of the villanus, and outside of Domesday Book we have far too little evidence to sanction the dogma that the average number must stand close to 2[1383]. Even the villein virgater on the monastic manors of the thirteenth century is often expected to have four oxen, and his having eight is a possibility that must be contemplated[1384].

Light and heavy ploughs.

That light as well as heavy ploughs were in use we have not denied. At a little later time we see teams of six beasts and teams of ten engaged in ploughing. But the compilers of Domesday Book are not concerned with the methods of husbandry; they are registering the number of oxen. If a man has one ox which is employed as a beast of the plough, they say of him: Arat cum uno bove[1385]. If he and another man have such an ox between them, they say: Ibi est semibos. If he has four oxen, they set this down as dimidia caruca. Instead of telling us that there are thirty-eight oxen, they speak of five teams less two oxen[1386]. Twelve pence make a shilling; and, at all events at the Exchequer, eight oxen make a team.

The team of Domesday and other documents.

Very lately an argument has been advanced in favour of a caruca, the strength of which varies from place to place. In many instances the Black Book of Peterborough in its description of the abbatial estates will give to the demesne of a particular manor exactly the same number of teams that are ascribed to it by Domesday Book, and, while in some cases the later of these documents will tell us that there are eight oxen to the team, in others it will speak of teams of six[1387]. That there is force in this argument we must admit; but many changes will take place in forty years, and we can not think that the correspondence between the two documents is sufficiently close to warrant the inference that the caruca of Domesday can have fewer beasts than eight. An exactly parallel argument would serve to prove that the hide of Domesday contains a variable number of fiscal ‘acres.’ Were it possible (but we shall see that it is not) for us to regard the teamland of Domesday as a fixed area, then we might afford to allow the strength of the team to vary; but if the teamland is no fixed area and the team has no fixed strength, then King William’s inquest ends in a collection of unknown quantities.

The teamland.

We turn from the team (C) to the teamland (B), and must face some perplexing questions. Reluctantly we have come to the opinion that this term ‘the land of (or for) one team’ does not in the first instance denote a fixed areal quantity of arable land. We have adopted this opinion reluctantly because we are differing from some of the best expositors of our record, and because it compels us to say that many of the statistical data with which that record provides us are not so useful as we hoped that they would be.

Fractional parts of the teamland.

In the first place, we must notice that if this term stands for a fixed quantity, a very rude use is being made of it. We see indeed that fractional parts of a teamland can be conceived. We often meet the land of (or for) half a team; we may come upon the land of or for two oxen, one ox, half an ox. But, except in a few counties, any mention of fractions smaller than the half of a team is rare, and even halves seldom occur. Now certainly the teamland was a large unit for such treatment as this. If, for instance, we suppose that it contained 120 acres, then we must infer that in some shires the jurors who had to describe a mass of 420 acres would have called it land for 3 or else land for 4 teams, and that in most shires an odd 80 acres would have been neglected or would have done duty as half a teamland. The hides or the carucates (A) have often been split into small fractions where the jurors distribute integral teamlands. One example of this common phenomenon shall be given. In Grantchester lie six estates[1388]: