Besides the obvious benchmarks of Domesday Book, the 1377 Poll Tax record and the start of Wrigley & Schofield's 330-year demographic reconstruction in 1541, the work follows from (though to some extent retreats from) Campbell's skilful demolition of the absurdly high peak 14th-century estimate of six million or more in vogue since it was suggested (initially as a speculative aside) by Michael Postan nearly half a century ago.
As far back as 1996 Campbell and Mark Overton demonstrated that English agriculture at its medieval peak c.1300 was likely to have fed only some 4-4.5 million people (and probably nearer the lower end of the range), revised by Campbell four years later to 4.38m: this time the authors have settled on a less rigorous 4.81m, which they admit is close to the maximum attainable. The concession derives from their needless timidity in dispatching the weakest of Postan's propositions:
Whereas Russell assumed that children under the age of 15 accounted for 33.3 percent of the population, Postan suggested that the ratio may have been as high as 40 to 45 per cent. For the period after 1541, when reliable data become available, the percentage of under-15s in the population never rose above 40 percent, which surely represents the upper limit for 1377. As Blanchard points out, such a high ratio tended to occur in periods of rapid population growth driven by high fertility. Since population was declining in the aftermath of the Black Death, a ratio as high as 40 to 45 per cent in the 1370s is improbable and a lower ratio more likely.
In fact Postan's 40-45% claim is quite preposterous and underlines the fragility of his whole speculation: there's plenty of evidence from populations with similar mortality and growth rates to indicate that anything much outside the 30-35% range is most unlikely for the under-14's (not under-15's) excluded from the tax's coverage: indeed given what we know of medieval mortality, to find 40-45% so aged we'd have to be looking at 2% annual growth, about six times the average for the 12th-13th centuries. The much-maligned Russell was spot-on on this count, for all his strange notions of 14th-century household size.
Since our estimates for the decades before the Black Death are founded on projecting back from 1377 with an allowance for the extraordinary intervening mortality (traditionally taken to be in the region of 40%, but raised in the Postan conjecture to 50% with no apparent allowance for the likely elevated birth rate between epidemics), the higher 1377 total means more people in 1348 than might otherwise be the case. Correcting the under-14 allowance to Russell's third offers a better fit with the agricultural carrying-capacity approach.
Postan was on stronger ground with his suggestion that omissions were far higher in 1377 than the 5% allowed for by Russell (indeed few now take Russell's rate seriously, comparing favourably as it does to many modern censuses). The authors err however in adopting his proposed 25% as a likely maximum for omissions in an earlier and quite different enumeration:
Although Harvey did not present any underlying calculations, she claimed that the Domesday population could well have approached 2 million. Rather than arguing for a higher household multiplier, Harvey argued for a much greater scale of omissions than the 5 per cent allowance made by Darby (1977), on the grounds that Domesday Book was more concerned with the landed wealth of the tenants-in-chief and their head tenants, and hence tended to under-record or omit the small-holding and landless elements. The final column of Table 1 presents an estimate of the English population in 1086 in the spirit of Harvey’s assumptions. This involves increasing the rate of omissions from 5 per cent to 25 per cent — the maximal scale of omissions claimed by Postan for the Poll Tax of 1377 — which results in a population of 1.87 million. Note that for the population to exceed 2 million, which Harvey claims should not be ruled out, would require an omissions rate of the order of 40 per cent.
The bulk of the omissions suggested by Harvey are, however, of a quite different character to those postulated for 1377. In 1086 the issue is not merely of oversights and evasions but of whole strata of the population being left out of the count because they were (quite correctly) deemed irrelevant to its purpose and therefore not worth recording. The number unrecorded will never be known precisely because they were left out, but the proportion could quite possibly have reached the higher level with no implication of similarly massive understatement in 1377.
So what of the final series? The figures are realistic, though by erring toward the mainstream, they may understate the level at the start (1086) and more likely overstate it in the middle (1290-1377). Interestingly, the authors find for 1348 on the great 14th-century peak population question, with the rival contender 1315 mustering only 4.69m, already down from 4.75m in 1290. Their estimate of a 12% fall in 1315-25 sheds new light on the disastrous results of the great famine of the early part of the decade and the harvest & livestock losses of later years.
The county estimates and maps toward the end of the paper offer further insight, even if they derive from already published distributions adjusted for the new totals: the striking growth of Lancashire's share in the period before 1600 should alone gladden the heart of those who see continuity where others see abrupt breaks.
All in all, it's a rewarding paper: apart from their over-hasty dismissal of Harvey's Domesday speculation, the authors may take sufficient confidence in their gut instincts to adopt a slightly more bare-knuckle approach in the inevitable revisions.
No comments:
Post a Comment