IN Blakwater, thirty-eight people died out of a total of about a hundred and fifty; dose to a quarter of the population. In Preston Stautney things must have been worse; probably nearly half the villagers succumbed. Which of these villages was nearer the national average? Can, indeed, any national average be established? Did a higher proportion of the population die in England than, for instance, in France or Italy? And how large was the actual death roll? Did a million English die? Two million? Three?
To none of these questions is a categoric answer possible but, now that the geographical tour of Europe has been completed, it is at least possible to hazard a few guesses. The most ample material on which an estimate can be based is certainly to be found in England but even here the base is shaky and deductions hazardous. It is possible to arrive at a wide variety of conclusions by differing but reasonably valid lines of argument, and exceedingly hard to establish which, if any, is the best one.
The first and, in some ways, most perplexing problem is the size of the total population in the middle of the fourteenth century. The main difficulty is that no attempt at anything approaching a general census was made between Domesday year and the poll-tax returns of 1377. Nor did even these attempt to cover all the counties of England or all kinds of men. Nevertheless it is possible to hazard a reasonably confident guess that the population of England in 1086 was something near 1.25 million,1 and that, by 1377, this had risen to about 2.5 million. If it were permissible to assume a steady increase of population between these two points then it would, of course, be easy to arrive at the approximate size of the population at any given date. But this is very far from being the case. On the contrary it is now established with a fair degree of certainty that the population rose to a peak about 1300 and then stagnated or even declined in the first half of the fourteenth century.
Exactly what caused the economic decline between 1300 and 1348 and how far, if at all, it was reflected in a reduction of the population has been the subject of much debate. Dr Titow has cited evidence from the Winchester Account Rolls to show that the great famines of 1315 to 1317 were the turning point.2 Though, in some areas, the recession seems to have begun ten years or more before, in general the statement seems valid. The famines themselves cost many lives but, in the palmier days of the thirteenth century, this loss would quickly have been made good. In the fourteenth century no such recovery took place. In a highly important article Professor Postan has demonstrated that, while wages rose gradually and taxation did not decline, there was a fall in agricultural output and in exports.3 The explanation must be a smaller force of labourers to share the pay packet. There is evidence to the same effect to be found in the narrowing wage-differential between skilled and unskilled labour and the withdrawal from previously cultivated land. ‘The contemporaries obviously believed that they were living in an age of contracting settlement,’ commented Postan, ‘and there is no reason why we should not accept their belief at its face value.’
The population in 1348 was, therefore, certainly little greater and probably less than it had been in 1310. But this does not tell one how large it was. Seebohm was the first historian to grapple with the problem in anything approaching modern terms.4 He visualized 1348 as a peak, attributed the rapid rise in the preceding century largely to the immigration of fishermen and manufacturers of woollen cloth, and concluded that the population just before the Black Death was in the region of five million. Thorold Rogers promptly countered with the contention that England could not possibly have supported a population of five million.5 He analysed the farm accounts of 8,000 bailiffs and deduced from the production figures that the population of England and Wales together must have been somewhere between two and two and a half million. After a delay for cogitation Seebohm replied challenging Rogers’s figures for corn production.6 The wrangle was there allowed to rest. For the next seventy-five years population estimates varied between these two points, usually inclining towards the higher.
In 1948, Professor Russell for the first time brought highly sophisticated statistical techniques to bear on the problem. His conclusion was that the English population in 1348 was some 40 per cent larger than at the time of the poll-tax, in round figures about 3.7 million.7 His graphs and tables are awe-inspiring but behind his arcane statistical manoeuvres the validity of his conclusion rests to a great extent on the gratifyingly comprehensible assumption that the average medieval household contained only 3.5 members and not five as had previously been assumed. The significance of this figure lay in the ratio which it established between land-holders, whose deaths were recorded, and the rest who usually died unchronicled. If his index figure were to be raised by even half a person per household, the total population would certainly be increased to well over four million. An accurate index figure must therefore be fundamental to any calculation.
Professor Russell justified his somewhat dramatic departure from accepted theory by evidence drawn from inquests of enclosures, poll tax lists and other sources.8 This is far from being unchallenged. The counter-argument, in its simplest form, was that the Russell household unit contained only the nucleus of parents and children. But there is good reason to include other members, such as a retired father, unmarried brothers or sisters, servants and sometimes even sub-tenants.9 Roger Tyler’s household included seven people in addition to the tenant himself – a large unit, certainly, but by no means improbably so. Professor Russell’s calculations, it is claimed, were based on an extremely limited number of cases and his evidence drawn largely from the period which followed the plague. If the 3.5 index were applied to the figures established for 1311 then, Dr Titow has pointed out, it would ‘postulate a society in which male persons over twelve years of age constituted 59 per cent of the total population’.10 Undoubtedly there is considerable variation between one period and another but Dr Krause, who analysed Professor Russell’s calculations with thoughtful distaste, cannot accept that in the fourteenth century the index fluctuated more than between 4.3 in a period of low childbirth and 5.2 when the rate was high.11
Faced with statistical juggling of this kind the layman is apt to feel a sense of baffled helplessness, leading often to blind acceptance of the latest theory which happens to have been propounded. He would do well to remember Professor Elton’s expression of lapidary wisdom:
Those determined to put their faith in ‘sophisticated’ mathematical methods and to apply ‘general laws’ to the pitifully meagre and very uncertain detail that historical evidence often provides for the answering of just such interesting and important questions, are either to be pitied because they will be sinking in quicksand while believing themselves to be standing on solid earth, or to be combated because they darken counsel with their errors.12
Professor Russell is far from having accepted these strictures on his theory.13 But he is too serious a scholar to maintain categorically that his must be the correct solution. The question remains open. In so far as any consensus can be said to have evolved it would probably be that the total population could have been anywhere within a range of which Russell’s 3.7 million would be the lower point and 4.6 million or so the higher. A total of 4.2 million has no more precise justification than any other but is certainly no less plausible and is a convenient central point from which to work.
Of this 4.2 million, how many died. ‘Only one in ten survived’, says one chronicler; ‘three quarters perished’, says another; ‘four fifths’, a third. Few estimates fall as low or lower than a half. Such lurid speculations, of course, contain little of interest to the statistician; enough cases have been established where the estimates of the chroniclers were palpably impossible to dispel any lingering belief that the man on the spot knew best But to arrive at a more rational figure is not easy.
One much favoured method is to seek to calculate from the ecclesiastical records the number of the beneficed clergy who died, to establish this as a proportion of the total and then to apply the same or some related percentage to the lay population. It was the application of this technique which led Cardinal Gasquet to claim that fully 50 per cent of the population perished in the two years beginning in July 1348.14 The imperfections of the method have already been discussed.15 Cardinal Gasquet himself was led badly astray by such deficiencies and even the more evolved workings of Professor Hamilton Thompson and Dr Lunn leave certain pockets of uncertainty.
Nevertheless studies of this kind can produce interesting and highly relevant results. Hamilton Thompson16 and Lunn17 between them have established the mortality rate of beneficed clergy in ten of England’s dioceses. The figures are remarkably consistent, ranging between just under 39 per cent for York and 39.6 per cent for Lichfield to 47.6 per cent for Bath and Wells, 48.5 per cent for Ely and 48.8 per cent for Exeter, Winchester and Norwich. On this basis it is reasonable to assume that something close to 45 per cent of all parish priests died during the plague. Similar statistics based on twelve of the more important monasteries show a surprisingly similar rate among the monks, 44 per cent of whom perished.
But though these figures are undoubtedly relevant to the problem of the total casualties caused by the Black Death, exactly how they should be used is harder to establish. It is as certain as any medieval statistics can be that, for England as a whole, the mortality rate among the people was lower than 45 per cent. Applying the ratio between dead clergy and dead people referred to above18 one must conclude, on the other hand, that it cannot have been lower than 34 per cent or so – say a third to avoid any false impression of exactitude.
Professor Russell, who found the figures for clerical mortality difficult to reconcile with his own very low estimate for deaths among the whole population, tried to overcome his difficulty not, as might have been expected, by assuming the existence of a larger differential between the two categories but by suggesting that the former figures were incorrect. ‘With some reluctance’,19 he reached the conclusion that Professor Hamilton Thompson, that ‘careful scholar who knows ecclesiastical practice so well’, had nevertheless been guilty of some fairly elementary blunders. But since the blunders whose presence he suspected were specifically those which Hamilton Thompson had set out to eliminate from the earlier calculations of Cardinal Gasquet, since Lunn has subsequently confirmed Hamilton Thompson’s conclusions and since neither Russell nor anyone else has yet done any work which yields substantially different results, it would seem premature to discard the fruits of their researches. It would be reasonable to say that, if no evidence existed except that of the Ecclesiastical Register, an overall mortality rate among the people of England of at least a third might be expected.
But there is other evidence, and Professor Russell has summarized it faithfully. There is, for instance, the possibility of arriving at an answer through figures for the payment of frankpledge dues. The value of the calculation is limited since it rests on a narrow statistical base of eighty-four case-histories in Essex, but it is worth noting that it gives an overall mortality of 43 per cent. Court Rolls also provide some evidence, though the principal lesson to be learned from them is the wide variation between different areas. In the Farnham manors the loss between 29 September 1348 and September 1350 seems to have been more than 28 per cent but less than 38 per cent, depending on the index figure taken for the ratio between tenants and dependents.20 A study of the manor of Cuxham in Oxfordshire indicates a death roll of something over two thirds.21 Similar figures for three manors belonging to Crowland Abbey suggest a rate of 56 per cent.22 On the other hand, in her analysis based on the Winchester Pipe Roll of eleven widely scattered manors belonging to the Bishop of Winchester, Dr Levett, while venturing no exact figure, could find no evidence to suggest a death rate high enough to disrupt the working of the manors and, in the case of one very large manor, felt that the figure of a third must be over-pessimistic. ‘The general impression gained from an attempt to make any such calculations,’ she concluded dryly, ‘is that they are singularly useless.’23
Finally there are the figures derived from inquisitions post-mortem, to which Professor Russell attaches particular importance. Based upon some five hundred such inquisitions he assumes that some 27.3 per cent of the population died during the plague; a figure which would be reduced to 23.6 per cent if allowance were made for the higher mortality among the older people. While admitting the limitations to this approach he concludes ‘Nevertheless, it presents the best evidence available as yet upon the effects of the plague.’24 23.6 per cent is far lower than any percentage which can be deduced from the other methods of calculation already mentioned. It might, therefore, be expected that it would be the lowest point in the range of possible death rates. But in his final summing up, Russell puts forward a still lower figure of 20 per cent, ‘The reduction of the … loss to 20 per cent’, he comments, ‘proceeds from better calculation of plague losses, which could take into account age specific mortality … discounting of ordinary mortality of the three years …’25 Since the 23.6 per cent figure derived from Russell’s favoured inquisitions post-mortem is itself loaded to take account of age specific mortality this further reduction is hard to accept.
It will be obvious from all this that to draw any conclusion must be hazardous. Little more than an informed guess is possible. It can safely be assumed that a figure of 45 per cent, equivalent to the figure of mortality among the clergy, would be the highest point of any possible bracket. Pace Russell, it seems incredible that the figure could be less than the 23 per cent which he suggested as the adjusted total derived from inquisitions post-mortem. A half-way point between these two poles would suggest a death rate of roughly a third. This would accord reasonably well with the evidence of the ecclesiastical registers. It is conspicuously lower than the figure derived from the payment of frankpledge dues or than most of the manors for which precise statistics happen to be available but these latter totals are not adjusted to take account of natural death. It is anyhow not unreasonable to expect that dramatically bad news would be recorded more enthusiastically than the humdrum figures of the luckier manors.
As a rough and ready rule-of-thumb, therefore, the statement that a third of the population died of the Black Death should not be too misleading. The figure might quite easily be as high as 40 per cent or as low as 30 per cent; it could conceivably be as high as 45 per cent or as low as 23 per cent. But these are surely the outside limits. On this basis the approximate total for the dead in England would be 1.4 million. No figure above one million and below 1.8 million would be astonishing but the nearer that the actual figure approached the median, the more it would seem to accord with the existing evidence.
Finally there is the question whether more or less the same proportion holds good for continental Europe. Probably the most useful observation which one can make on this is that there is no obvious reason why it should not. The regional variations which are so evident in England must be, of course, immeasurably greater in a continent with its wide range of climate, landscape and racial types. There are certain areas, for instance Tuscany, where the joint efforts of contemporary chroniclers and modern scholars make it almost certain that the death rate was higher than in England. There are others, for instance Bohemia, where the incidence of the plague was clearly lower. For most of the Continent even the inadequate medieval records which exist in England are lacking or have not yet been exploited – the material for a serious comparison does not exist. But it can still be said that there are few grounds for drawing distinctions between England and the rest of Europe.
Such calculations as have been made for individual countries are not at variance with this somewhat negative assertion. In his careful study of the effects of the Black Death in France, Renouard concluded that the only rule which could safely be put forward was that the overall death rate varied between one-eighth and two-thirds according to the region.26 Doren has estimated that between 40 per cent and 60 per cent of city dwellers in Italy died but that, in the countryside, the proportion was considerably lower.27Neither of these estimates can be translated into a death rate valid for the whole country, a form of calculation which these leading authorities prudently eschewed. In so far as they point in any direction it is towards a figure greater than a third. Certainly they are hard to reconcile with anything substantially lower. But such speculation is unprofitable. To maintain that one European in three died during the period of the Black Death can never be proved but, equally, cannot be wildly far from the truth. Further than that, in the present state of knowledge, one cannot go.
1 Though see J. C. Russell, British Mediaeval Population, op. cit., P.54.
2 J. Z. Titow, ‘Some Evidence of the 13th Century Population Increase’, Econ. Hist. Rev., 2nd Ser., Vol. XIV, No. 2, 1961, p.220.
3 M. Postan, ‘Some Economic Evidence of Declining Population in the Later Middle Ages’, Econ. Hist. Rev., 2nd Ser., Vol. II, 1950, p.221.
4 F. Seebohm, ‘The Black Death and its Place in English History’, Fortnightly Review, Vol. II, 1865, pp. 149–60 and 268–79.
5 J. E. T. Rogers, ‘England Before and After the Black Death’, Fortnightly Review, Vol. III, 1865, pp. 191–6.
6 F. Seebohm, ‘The Population of England before the Black Death’, Fortnightly Review, Vol. IV, 1866, pp.87–9.
7 J. C. Russell, op. cit. p.246.
8 ibid., pp.22–33.
9 G. C. Homans, English Villagers of the Thirteenth Century, op. cit., pp.209–12.
10 J. Z. Titow, op. cit., p.222.
11 J. Krause, ‘The Mediaeval Household: Large or Small’, Econ. Hist. Rev., 2nd Ser., Vol. IX, 1957, p.432.
12 G. R. Elton, The Practice of History, Sydney, 1967, p.34.
13 J. C. Russell, ‘Recent Advances in Mediaeval Demography’, Speculum, Vol. XL, No. 1, 1965, p.84.
14 Black Death, op. cit., p.225.
15 p. 131 above.
16 ‘Registers of the Bishop of Lincoln’ and ‘Pestilences of the 14th Century in the Diocese of York’, op. cit.
17 PhD. thesis, op. cit., p.126, n.22, above.
18 p. 132 above.
19 op. cit., p.221.
20 E. Robo, ‘The Black Death in the Hundred of Farnham’, Eng. Hist. Rev., Vol. XLIV, 1929, p.560.
21 P. D. A. Harvey, A Mediaeval Oxfordshire Village: Cuxham, op. cit.,p.135.
22 F. M. Page, The Estates of Cropland Abbey, Cambridge, 1934, p.125.
23 A. E. Levett, ‘The Black Death on the Estates of the See of Winchester’, Oxford Studies in Social and Legal History, Vol. V, Oxford, 1916, pp.80–81.
24 op. cit., p.216.
25 ibid., p.367.
26 Y. Renouard, ‘Conséquences et intérêt démographique de la Peste Noire de 1348’, Population, III, 1948, p.459.
27 A. Doren, Storia Economica dell’ Italia nel Medio Evo, Padua, 1937. P.579.