Manorial sources

Roughly speaking, and for purely technical points of our discussion, manorial documents can be divided into three types: manorial accounts; manorial court rolls; and different types of extents (including surveys, rentals, terriers, and, most relevant for our discussion, Inquisitions Post Mortem, henceforth, IPMs).Footnote ¹⁵

Manorial accounts – also referred to as demesne accounts or demesne compotus rolls – are agricultural and financial reports compiled by an official managing the demesne (the lord’s lands within the manor) – typically a bailiff, sergeant, or reeve. The records were audited and documented by a demesne clerk. The accounts followed a ‘charge–discharge’ format, with revenue received (incoming funds) detailed in one section before listing expenditures (outgoing funds or expenses). Generally, these accounts were produced annually, covering the period from Michaelmas (29 September) of the previous year to the same date of the current year. Occasionally, they spanned shorter durations – four months or half a year. Typically, they comprised two main sections: financial records of income and expenditures on the front of the roll and the agricultural account on the reverse or dorse side, which is relevant for our discussion. The agricultural section of the accounts is divided into three parts: the account of crops (compotus grangii), livestock (compotus stauri), and (in some instances) labour (opera). The crop section details the volumes of different crops harvested in the previous financial year and sown during the current financial year, as well as how the harvest was managed, primarily through sales and consumption. In some cases, crop yields appear in the account’s margins, though determining exact figures often requires comparing two consecutive accounts. The livestock section lists different species of demesne animals: horses, bovids (oxen and cattle), sheep, swine, and poultry, specifying, for each category, the annual patterns of gain and loss (including births, deaths, transfers, sales, and purchases), as well as the sex and age groups of the animals. Some – but not all – accounts also include the labour section, recording labour productivity rates of different works produced by local agricultural labourers.Footnote ¹⁶ As Figure 2 shows, there are about 2,700 demesnes with surviving accounts in England from the period c.1208–1529; some demesnes are blessed with very long and fairly complete chronological runs; some others – with only a single surviving account. The heyday of manorial accounts was the period c.1270–1370, after which their annual numbers were progressively declining, because of the seigniorial proclivity to lease out their demesnes to better-off tenants and thus to do away with manorialism. Taken together, it is possible to estimate that the total number of surviving accounts available for research is in the area of 22,000 rolls.Footnote ¹⁷

Figure 2. The County-Level Distribution of English Manorial Accounts, 1208–1529.

Source: Slavin, Manorial Accounts Database (tabulated data on annual crop yields, crop acreage and livestock heads deriving from manorial accounts and extents from over 2,600 English demesnes for the period c.1208–1529); Manorial Documents Register (https://discovery.nationalarchives.gov.uk/manor-search).

Manorial court rolls are records of local courts (curiae) that addressed and regulated various legal issues within the manor. While early rolls primarily documented land transactions among tenants, they later expanded to cover broader aspects of manorial life, involving both free and unfree peasants. These included criminal offences such as violent conduct, theft, harbouring criminals, poor work performance, baking, brewing, and selling contrary to the Assize of Bread and Ale regulations; land dealings, including inheritance, sales, purchases, exchanges, and disputes; matters of credit and debt; customary fines and dues; and the election of manorial officials.Footnote ¹⁸ Essentially for our purposes, one important feature of the court rolls is entries dealing with inheritance: heriots – inheritance dues payable by unfree adult tenants to their lords upon entering upon their deceased relatives’ parcels of land (and in some instances, fines paid on transfers of land), and deathbed transfers – a custom, whereby a dying person (described in manorial court rolls as languens in extremis) would instruct a manorial officer or a witness from among the customary tenants on a manor to ensure transferring his/her land plot to his/her beneficiary at the next court sitting. The court rolls survive in very large numbers – in hundreds of thousands – from 1237 on, going beyond the end of manorialism and persevering into the nineteenth century.Footnote ¹⁹

Different types of extents summarise the values of the various land components of the manor – arable, grassland (pasture and meadow) and woodland.Footnote ²⁰ While rentals and terriers, recording land held by all kinds of tenants (be it serfs or local lords), could be produced for any manor held by any type of lord, Inquisitions Post Mortem (IPMs) were produced only for manors held by lay tenants in chief. The IPMs, compiled and issued upon or shortly after the tenant’s death, tend to be much more detailed than other types of extents. The process began with the king sending a writ to his escheators in the counties where the tenant’s manors were located. An inquiry (or inquisition) was then conducted at each manor, by gathering detailed information through local sworn juries and with local clerks compiling the extents and forwarding them to the royal chancery. The IPMs provide details on manor layouts, listing components such as arable land, pasture, meadow, and woodland, measured in acreage and value. They also document minor land uses, including fisheries, warrens, dovecotes, buildings, mills, granaries, turbaries, and orchards, as well as dues, customs, and revenues associated with each manor.Footnote ²¹ Between 1236 and 1509, some 20,000 IPMs have been created.Footnote ²²

There are additional types of sources, sometimes embedded into manorial accounts and court rolls, and sometimes produced as separate documents. These include tithe receipts (roughly, one-tenth of agricultural produce paid by tenants to manorial lords in those instances, when the latter were rural rectors owing the tithes), pannagium rolls (payments by tenants to let their pigs into lord’s woodland), and estimaciones bladorum (a source unique to the estates of Worcester Cathedral Priory, estimating the annual demesne produce and tithe receipts, stored separately).Footnote ²³ Because of the textual limits of the present paper, they could not be included into the present discussion.

Crop yields and palaeoclimatology

One of many important advantages of manorial accounts is that they contain very detailed sections dealing with annual crop yields and the disposal of annual harvests on the demesne. Much work has been done on crop yields, both on a regional and a ‘national’ scale. The single largest contribution is Bruce Campbell’s database of crop yields for the period 1211–1495, based on over 30,000 individual yield observations from about 230 manors from most English regions (Figure 3). As Campbell’s dataset shows, year-to-year yields were highly volatile, with some years exhibiting ‘average’ yields, with some years being blessed with abundant harvests, and with some being marked by harvest failures. Given a very detailed nature of manorial accounts in general and the harvest sections in particular, it is very tempting to accept their contents at face value.

Figure 3. National, aggregated net crop yields in England, indexed on 1270–1429 (1270–1429 = 100 = average crop yield).

Source: Bruce M.S. Campbell, Three Centuries of English Crop Yields (https://www.bahs.org.uk/crop-yields-database) (accessed June 2025), supplemented by Manorial accounts database for the period 1311–1320.

Note: The figures are indexed on average net yields (inclusive of tithes, but exclusive of seed) for the period 1270–1429 (1270–1429 =100 = average crop yield). These are aggregated per-seed yields of all grains weighted together, according to each grain’s relative contribution to the total ‘national’ demesne acreage, and calculated as the ratio between a previous year’s seed and the current year’s harvest minus seed corn (annual harvest share invested in seeding).

While manorial accounts may indeed provide accurate and reliable information on annual crop yields in ‘average’ and good years, some caution must be exercised in ‘bad’ years. As suggested elsewhere, in some instances it appears as if manorial reeves, accountable for demesne crops and livestock, may have reported notoriously low yields, sometimes ascribing these to a bad weather (appearing in marginal glosses of some accounts – importantly, these entries were not being recorded as standard and their inclusion was reliant on local scribal or managerial practice), to avoid the imposition of financial penalties by their lords. Manorial officials occasionally manipulated their accounts, including grain sale prices, harvest yields, livestock disposal, and wages paid in cash or kind. To prevent fraud, manorial lords employed auditors, who sometimes detected these falsified figures.Footnote ²⁴

Obviously, the accounts do not state in which instances the low figures of crop yields are reliable and in which they are deliberately driven down by manorial officials. One potential way to approximate that – and, by doing so, to establish if crop yields reported by manorial officials and recorded by local scribes correlate with annual weather patterns – is to juxtapose the accounts against annual dendrochronological data. On the most basic level, lower-altitude trees are sensitive to precipitation levels, while higher-altitude trees are responsive to temperature variations. In the former case, wide rings reflect wet weather (and narrow rings indicate dry weather); in the latter case, wide rings record warm temperature (and narrow ones reflect cold temperature). Although such an exercise has much promise, it is not without drawbacks, for the following reasons. Firstly, England (along with other parts of the British Isles) experiences regional climate variability, causing trees in some areas to be less sensitive to rainfall. As a result, they record conditions as drier than they actually are – unlike those on the Continent. Secondly, seasonality plays a crucial role. English and other British oaks primarily register precipitation between March and July, but during late summer and early autumn – when wet weather posed risks to harvesting – they largely become unresponsive. In other words, they do not record rainfall outside these months. Lastly, trees across much of the British Isles are generally poor at tracking precipitation. This is mainly due to the region’s temperate climate, which prevents trees from responding distinctly to climatic stress, unlike those in other regions (e.g., trees in the arid Americas show a strong reaction to rainfall, while those in Scandinavia’s high latitudes respond primarily to temperature).Footnote ²⁵

With all these caveats, such exercise should still be undertaken – especially on a regional/local level. As of present, there are only two published English dendrological series for the late-medieval period – one from East Anglia (covering 21 sites, including three in Cambridgeshire, six in Suffolk, and twelve in Norfolk) and the other one from western and south-western/south-central counties (covering 15 sites, one in Shropshire, two in Warwickshire, seven in Oxfordshire, one in Gloucestershire, two in Wiltshire, one in Somerset, and one in Hampshire).Footnote ²⁶ The East Midlands and the North are, by contrast, yet to receive their attention. More recently, summer precipitation levels for England and Wales have been inferred from stable oxygen isotopes in oak tree rings for the period 1201–2000.Footnote ²⁷ The dendrochronological record can in theory be supplemented by precipitation record reconstructed from isotope compositions of lipids (arising from degraded fats) from annually dated sheep parchments (a method recently developed and proven by Matthew Collins and his collaborators) and thus fill many important spatial gaps in the existing palaeo-climatic record.Footnote ²⁸ The potential of lipid-derived analysis and reconstruction methods will be discussed in greater detail later in the paper.

One methodological way forward is to identify manors with good chronological runs that were situated in a proximity to one of the tree-ring sites and to analyse a degree of annual synchronicity between the three sets of data – manorial accounts (with annual crop yields and occasional weather-related glosses), lipid-derived precipitation levels and dendrochronologically derived precipitation levels. To achieve that, it would be necessary to see (1) to what degree poor yields are correlated with excessively dry or wet weather (at least in March-July, recorded by tree rings) and (2) if occasional weather references in harvest entries and marginal glosses of the manorial accounts are reflected in the local dendrochronological and lipid record. Although one would expect to see much harmony between the three datasets in many instances (especially in good years, but also in some crisis years, as demonstrated in a study of the Great Famine of 1315–1317, a major subsistence crisis initiated by the torrential rain of 1314–1316),Footnote ²⁹ there will undoubtedly be still some discrepancies between the series. The next step will be to determine in which instances such discrepancies occur. For instance, there are many references to extreme summer weather in the Pipe Rolls of Winchester Bishopric manors (skewed towards the southern, south-western and western counties), but only less than half of these match the dendrochronological records from the same regions.Footnote ³⁰ In order to understand what accounts for these discrepancies, it will be necessary to consider each such instance individually, by analysing, in a meticulous manner, various environmental and managerial aspects that can be gleaned from manorial accounts of the same year, as well as some preceding ones. Were fields under-manured before seeding (some accounts provide, in a systematic manner, valuable information about manuring rates)? Were there some issues with agricultural labour force? Or could it merely be a reeve’s miscalculation/scribal error? Such scenarios should all be taken into account. Apart from warning historians about potential pitfalls in relying on these weather references in manorial accounts, these discrepancies may reveal some ‘dishonest’ strategies of manorial authorities and managers to ‘cover up’ for poor performance. This, in turn, is related to a much wider question of the efficiency and transparency of manorial management in late-medieval England – a topic that has already attracted some scholarly attention in the past.Footnote ³¹ Juxtaposing manorial account glosses against palaeo-climatic record from different regions can be a methodological innovation with a potential to reveal some important aspects of this question.

Inheritance lists and aDNA

Another example of a potential marriage between manorial sources and palaeo-science is related to plague research. In analysing the demographic impact of plague on rural population, historians have used inheritance entries listed in manorial court rolls – heriots and deathbed transfers. Heriots were inheritance dues owed by unfree adult tenants to their lords upon entry on their deceased relatives’ plots of land, and in some instances, dues paid on transfers of land inter vivos – that is, among living individuals engaged in land transactions. Obviously, for the purpose of a proposed exercise, only post mortem heriots should be considered. Deathbed transfers were a custom, whereby a dying person would instruct either a manorial officer or a witness from among the customary tenants on a manor to ensure the transfer of his/her land plot to his/her beneficiary at the next court sitting.Footnote ³²

Despite a piecemeal erosion of serfdom and manorialism in the aftermath of the Black Death, the custom of heriot survived well into the fifteenth and sixteenth centuries on all sorts of manors all over the country, with death dues continuing being robustly collected by manorial authorities from local tenants – unfree, free, and freed alike – during court sessions. Likewise, deathbed transfers remained an integral part of court sessions in the post-Black Death era. Thus, heriot payments and deathbed transfers survived both on manors, whose ‘conservative’ lords were reluctant to give away, for time-being, seigniorial control, and on those, where local lords freed their servile tenants shortly after the Black Death.Footnote ³³

One of the most hotly debated questions in the field of plague studies is that of sex selectivity in plague mortality: in other words, were men or women more prone to this disease? Although it appears that in many instances, younger women were at a higher mortality risk than other demographic groups, there is still much discrepancy, and it appears that situation varied across spatio-temporal contexts.Footnote ³⁴ One particularly interesting case is the pestis secunda ravaging the British Isles in 1361–1362. According to numerous contemporary and later chronicles, that wave targeted proportionally more men than women, and most recently, this allegation has been confirmed in one study of heriot payments from 184 English manors, revealing that the proportion of female inheritance pronouncedly increased during the pestis secunda outbreak.Footnote ³⁵ But the same heriot entries, even though revealing this fascinating (and unusual) trend, do not explain it.

It has been hypothesised that the peculiar mortality patterns of the pestis secunda may be linked to unique molecular features of its associated Yersinia pestis strains, belonging to the so-called Branch 1B. As of Spring 2025, of some 104 sequenced post-Black Death genomes from the Second Plague Pandemic (deriving from various central, eastern and north European contexts from the period c.1357–c.1772), 95 are associated with Branch 1A and its multiple sub-branches, and only remaining nine belong to Branch 1B (Figure 4). All Branch 1B genomes, without exception, are associated with pestis secunda outbreaks, while all the later genomes, coming from subsequent waves, belong to Branch 1A – indicating that Branch 1B strains may have migrated out of Europe in the course of the pestis secunda wave.Footnote ³⁶ Could the peculiar molecular signature of 1B branch and its strains be responsible for the preponderance of male mortality during the pestis secunda? One potential venue to confirm or disprove this hypothesis would be to conduct an in-depth molecular analysis of Branch 1B genomes’ strains to see if there are any genetic features or loci that may have had a potential impact on oestrogen or testosterone. There is one sequenced genome from London St Mary Graces (SMG6330), associated with the pestis secunda outbreak of 1361 – in addition to eight more genomes from other European contexts (five from Krakauer Berg, near Halle in Germany; two from Bergen-op-Zoom in the Netherlands and one from Bolghar in Tatarstan, Russian Federation).Footnote ³⁷ To appreciate if the 1B strains had indeed some peculiar molecular signature potentially influencing sex selectivity in pestis secunda mortality, it is essential to compare their genetic features to those of other Yersinia pestis strains, associated with other Second Plague Pandemic waves.

Figure 4. Schematic Phylogenetic Tree of post-Black Death Genomes Associated with so-called Branches 1A and 1B.

Sources: Redrawn from Keller et al., ‘Refined Phylochronology’, SI Fig. 26; Katherine Eaton, et al. ‘Emergence, Continuity, and Evolution of Yersinia Pestis Throughout Medieval and Early Modern Denmark’. Current Biology 33 (2023): Fig, 3; Joanna Bonczarowska, et al. ‘Ancient Yersinia Pestis Genomes Lack the Virulence-Associated Ypfφ Prophage Present in Modern Pandemic Strains’. Proceedings of the Royal Society B 290 (2023): Fig. 2.

In the later fourteenth century alone, there was a series of both ‘national’ and ‘regional’ post-Black Death plague waves in England – including the pestis secunda of 1361–1362, the pestis tertia of 1368–1370, the 1374–1375 wave, the 1379–1380 wave in the north, the 1382–1383 wave in the south, as well as the 1389–1391, 1393–1394, and 1399–1401 regional waves. So far, sex selectivity in plague mortality in late-medieval England has been inferred only for the Black Death and pestis secunda – but not for subsequent waves. The same methodology establishing sex selectivity on the basis of heriot lists and deathbed transfers (inferred from changing proportions of land inheritance by men and women) may be applied to post-pestis secunda outbreaks. The results can then be supplemented by an in-depth molecular analysis of strains associated with each outbreak in question, to see if there is any correlation between gender-specific mortality patterns in each wave and peculiar genetic markers of their associated strains. Conjoining the two sets of data has a huge promising potential to improve our understanding of biological factors triggering different mortality patterns in plague and potentially other bacterial epidemic diseases.

In practice, however, achieving that can be not as straightforward, for several reasons. Firstly, it is very rare that court rolls specify ‘plague’ as a cause of death, in conjunction with inheritance entries. Secondly, as noted above, some late-fourteenth-century plague outbreaks were regional, rather national in their character, devastating only certain counties. Thirdly, there was, in some instances, a declining frequency of court sessions (widening the chronological gaps between two sequential courts), rendering the disaggregation of plague mortality from non-plague mortality challenging. Fourthly, the proportion of women holding unfree land has always been low, falling further down in the post-Black Death period – thus, heriot lists and deathbed transfers do not, in many instances, reflect the true rates of female mortality, but only in relative and approximate terms. Finally, unfortunately, virtually all of sequenced Yersinia pestis genomes come from undated burials and associating undated genomes with particular waves can be only approximate. All these pitfalls, however, may be at least partially overcome. As far as the first and second shortcomings are concerned, in order to look for and identify excess mortality associated with plague, it is necessary to know, on the basis of external information deriving from other textual sources, about the chronology of plague outbreaks in a specific region, in which a given manor was situated. The third challenge can be addressed either by analysing only those courts that continued reconvening with the pre-Black Death frequency, or by including court sessions with lowered frequencies, as long as these clearly indicate excess mortality during plague years compared to non-plague years, as reflected in inheritance entries. The fourth drawback can be addressed by comparing the relative shares of female heirs during the known plague years with those in the preceding five or so years. By doing this, we establish relative patterns of inheritance by gender over a six or so years and thus reveal whether the situation during the plague years conforms to or differs from those in the preceding years. This final pitfall can be overcome with the help of a novel methodology, most recently developed by Keller et al., associating post-pestis secunda Branch 1A genomes, whose precise chronological context cannot be established, with particular plague waves, on the basis of so-called ‘Phylogenetically Informed Radiocarbon Modelling’ (PhIRM), combining phylogenetic positioning, radiocarbon dating and historical context data.Footnote ³⁸ This approach allows to narrow down undated genomes to about 2–4 possible historical plague waves – which, in turn, would be a decent historical approximation for the purpose of such study.

Agrarian productivity and palaeo/archaeobotany

Another big controversy is the relationship between agricultural productivity and the total population. In a nutshell, for some historians, agricultural output is regarded as the main trigger to sustain population growth, while population trends are conditioned by food production levels – at least, in the pre-Black Death era, marked by overpopulation, rural congestion and low living standards.Footnote ³⁹ The estimates of total English population levels on the eve of the Black Death vary considerably. Thus, Bruce Campbell, a leading authority in late-medieval agricultural history, estimated in 2000 that the total agricultural output was unable to sustain more than 4–4.25 million people around 1300.Footnote ⁴⁰ More recently, Stephen Broadberry, Campbell and their colleagues, drove the figures higher to about 4.75 million c.1300 and 4.81 million c.1348 (before the arrival of the plague in June that year).Footnote ⁴¹ Higher figures of, respectively, 5.25, six, and 6.2 million c.1300 were proposed by Philip Slavin, Richard Smith, and Hugo La Poutré.Footnote ⁴² Most recently, Philip Slavin argued, on the basis of projecting figures back from the 1377 Poll Tax return via estimated losses in the 1374–1375 plague wave, the pestis tertia of 1368–1370, the pestis secunda of 1361–1362 and the Black Death of 1348–1350, that the total English population c.1348 (and c.1300) could not have been lower than 5.3 million people – and it could have been higher, anywhere within the range of 5.3–6.6 people.Footnote ⁴³ These estimates are at odds with Campbell’s and Broadberry et al.’s figures, based on agricultural output.

To account for this discrepancy, Slavin suggested that the previous studies (1) under-estimated the total levels of agricultural output (with peasantry sector yields being higher than those within the demesne one) and (2) may have under-estimated the calorific value (by weight) of medieval crops which could have been higher than these we are consuming today. The author hypothesised that pre–Green Revolution grains may have contained more protein, as crops with lower ploidy tend to have higher protein levels. When grains expand significantly in size, it is carbohydrates that primarily contribute to the increase, not other nutrients. Although protein and carbohydrates have nearly the same calorific value by weight, the key factor is that protein requires more energy for digestion than carbohydrates, making it a more efficient energy source per unit of weight.Footnote ⁴⁴

Without palaeobotanic analysis, this hypothesis remains purely conjectural. But there is more than enough potential material available for such analysis, in the form of macro-remains of food and agricultural products recovered from hundreds of late-medieval agricultural sites in England (and indeed other parts of the British Isles and beyond). Recent palaeo-/archaeobotanic studies deployed various biochemical, chromatographic, and spectroscopic techniques, as well as stable isotope analysis (a technique identifying isotopic markers of certain foods in human bone and teeth), to study macronutrients, including carbohydrates and proteins as well as calorific values in historical cereal diet, but these studies are focused on ‘pre-history’, rather than more recent periods.Footnote ⁴⁵ These advances in archaeobotanical studies in general would benefit the field of medieval agricultural history a great deal – especially when engaged in a collaborative framework with agricultural historians. Marrying the evidence from manorial accounts with archaeobotanical analysis of macro-remains of food and agricultural products, available in abundance, would advance our knowledge about calorific value and food consumption in late-medieval England and, with new data and knowledge, it would help resolving the decades-old controversy about productivity levels of late-medieval English agriculture and approximate size of population it was capable of sustaining. As noted above, any scholarly estimate about calorific productivity levels within late-medieval arable sector has been so far based on modern nutritional data. Conjoining manorial account evidence with new palaeo-/archaeobotanic analyses would allow, for the first time, to establish whether one bushel of wheat would produce as many calories as today. By extension, it would enable us to come up with more secure estimates about annual arable productivity levels on the manorial, regional (conjoining data from multiple manors within the same region) and ‘national’ (conjoining data from all the regions together) levels. This, in turn, can have some dramatic implications for our understanding of late-medieval agricultural productivity levels and capacity on the one hand, and potentially improve our estimates about the total English population levels before and after the Black Death, on the other. As noted, these are indeed some of the most hotly debated topics in the field of late-medieval economic and social history of England, largely unresolved until today.Footnote ⁴⁶

Commoners’ diet and stable isotope analysis

Stable isotope analysis can be similarly deployed in conjunction with another topic in late-medieval economic history: changes in living standards of English commoners, as reflected in food consumption. In a nutshell, the current consensus is that the widespread poverty and low living standards before the Black Death meant a heavy crop-based diet, while the rise in living standards after the Black Death implied a shift to a more diversified diet portfolio, with increased fish and meat shares.Footnote ⁴⁷ So far, the only quantitative study of long-term changes remains the 1988 article of Christopher Dyer dealing with food consumption by harvest famuli (hired agricultural labourers), on the basis of harvest expenditure section of manorial accounts from Sedgeford (Norfolk) between 1256 and 1431, detailing the annual allocation of food among famuli.Footnote ⁴⁸ Sedgeford has a very good chronological coverage, allowing such exercise; however, it is by no means a unique manor in this regard. There is a long list of other monastic, episcopal and collegial estates, where such exercise is possible; some of these have a particularly good chronological coverage – including the manors of Barnhorn and Alciston (both in Sussex and both belonging to Battle Abbey), running, with very few gaps, from, respectively, 1325–1326 and 1336–1337 until the 1490s, and some of Bishop of Winchester manors, running with virtually no gaps, from 1208 to 1209 until the 1470s.Footnote ⁴⁹ Sampling manors from different regions (say, several manors from each county or several manors from regions characterised by different topography and landscape) would also allow to identify regional differences. Would, say, fish consumption levels rise more pronouncedly in coastal than inland regions? Or, would tenants on higher-altitude manors, focusing on pastoral husbandry, increase the consumption of meat?

Manorial accounts are just one source capable of shedding light and answering these questions. But they reflect only one segment within the late-medieval society: hired agricultural workers, who often were better-off than ‘self-employed’ tenants and farmers. But even if the changes in food consumption patterns of the same famuli may be reflective of peasants from different walks of life in general, they may still not be reflective of those of urban population. To determine this point, manorial accounts can be juxtaposed against evidence of food consumption deriving from stable isotope analysis, identifying isotopic markers of certain foods in human bone and teeth, from nearby contexts – rural and urban. Ideally, such analysis should focus on contexts, for which we have excavated skeletal data spanning several centuries. Such urban sites included the cemeteries of St. Gregory’s Priory, Canterbury; St. Helen-on-the-Walls, York; St James Priory, Bristol and St Swithun’s Winchester.Footnote ⁵⁰ As far as rural sites go, these are much fewer, but St. Peter’s, Barton-upon-Humber (Lincolnshire) and Wharram Percy (Yorkshire) can fill the gap.Footnote ⁵¹ Importantly, stable isotope analysis has been successfully deployed in a series of studies reconstructing food consumption in late-medieval England (and the Continent),Footnote ⁵² – albeit, in isolation from manorial documents or other textual evidence and, in most instances, without considering secular changes.Footnote ⁵³

Land use and palaeoecology

Another potential venue for collaborative research involving manorial documents and palaeo-science is changes in rural land use over time. In a nutshell, the main contours of landscape change in late-medieval England can be summarised as follows – at least, as far as evidence from the demesne sector can tell. In the course of the thirteenth century, driven by population growth and demand for food, a relative proportion of arable acreage grew at the expense of pasturage and woodland, parts of which were being converted into ploughland. By the late thirteenth century, England may have reached the peak of its arable land capacity, as it is evidenced in, more or less, the stable size of arable acreage until the Black Death. After c.1350, given the changes in land-to-labour ratios brought about by population decline and stagnation in the aftermath of the Black Death and subsequent plague waves, both the relative and absolute size of arable land shrank, while that of grassland expanded.Footnote ⁵⁴ This was driven not only by demographic changes, but also by new economic opportunities: with fewer mouths to feed and fewer hands to plough, both lords and tenants shifted their focus to pastoral economy in general, and sheep farming in particular, boosted by international demand for English wool.Footnote ⁵⁵

In recent years, palynology (study of pollen remains) has been recognised as a powerful tool in appreciating post-Black Death landscape changes, as demonstrated in the 2022 study of Adam Izdebski et al.Footnote ⁵⁶ Unfortunately, the only region where good-quality pollen data was available for that study is Cornwall – and (unlike some regions in Scandinavia, Germany, France, and Italy) that county shows no evidence of increase in cereal pollen indicators in the aftermath of the Black Death. Obviously, Cornwall has always been a comparatively sparsely populated and more of a ‘pastoral’ county, focusing on dairy farming and hence, it is not the best candidate to deploy palynological analysis to study landscape changes.Footnote ⁵⁷

To get a fuller and clearer picture – and to see if these changes gleaned from manorial sources are reflected in the natural record – it is necessary to (1) gather pollen data from as many sites in different regions as possible and (2) juxtapose them against manorial sources. Here, manorial accounts provide some good indication of the changes in size of demesne arable acreage. However, this should be approached with caution: although they often indicate that parcels of land were leased out to local tenants, they do not indicate how much land was converted into seignorial pasturage. Hence, manorial accounts need to be supplemented with Inquisitions Post Mortem (IPMs). The IPMs make clear distinction between arable land, pasturage, and woodland within the demesne and record the physical extent of each land type and their respective values.Footnote ⁵⁸ In contrast with manorial accounts, rendered on an annual basis, the IPMs were only compiled upon death of a tenant-in-chief. In other words, an ‘average’ manor held by a tenant-in-chief would be surveyed once every, say, 30–40 years. Still, even such frequency allows us to discern changes in land use over time. To complete the picture, manorial accounts and IPMs can be supplemented with Royal Forest Proceedings (not manorial sources) dealing with royal forest use and management, and revealing absolute and relative changes in woodland acreage, in relation to other types of land. This important source has not been properly utilised since Ambrose Raftis’ 1974 study – despite its huge potential.Footnote ⁵⁹ Thus, conjoining manorial sources and Royal Forest Proceedings with palynological data can provide some important insights into landscape changes before and after the Black Death. These changes should, in turn, be connected to wider demographic and economic processes and shifts characterising late-medieval and early-modern England.

The early fourteenth-century crisis and palaeo-epigenetics

The topic of food crises – and particularly the Great Famine of 1315–1317 – and their impact on late-medieval society has attracted some scholarly attention in the last 15 or so years.Footnote ⁶⁰ The Great Famine has been regarded as one of the harshest subsistence crises in European history, caused by the unfortunate combination of a short-term climatic shock (torrential rain depressing crop yields), demographic pressure and, most importantly, a series of institutional/anthropogenic factors. In England, about 15–20 per cent of the population died of starvation-related diseases – a remarkable figure, even by famine standards.Footnote ⁶¹ To make things worse, however, the Great Famine was followed by another major eco-biological disaster: the Great Bovine Pestilence, most likely caused by rinderpest panzootic, killing about 63 per cent of English and Welsh bovids in the course of 1319–1320.Footnote ⁶² Because manorial authorities prioritised oxen restocking, the replenishment of dairy cattle stocks was slow (it would take about 13 years to restock cow stocks to their pre-panzootic levels), and, as a result, an entire generation of commoners experienced a temporary deprivation of some vital nutrients, including protein, calcium and B12.Footnote ⁶³

When studying famine or a similar subsistence crisis or food deprivation, it is essential to focus not only on victims but also on survivors – in order to assess the impact of the disaster on local communities and their socio-economic and biological standards – on both the intra- and trans-generational scale. In particular, starvation can have some long-term consequences on both physical and mental health of the survivors – and their progeny, potentially over two subsequent generations. Thus, it has been established that prenatal malnutrition during the Dutch Hongerwinter Famine (1944–1945) and the Chinese Great Leap Forward Famine (1958–1961) produced genetic changes triggering schizophrenia- and depression-related alterations in the epigenome. Additionally, these same epigenetic modifications weaken the immune system in foetuses and children, making them and their progeny more susceptible to various infections and diseases, as well as shorter stature.Footnote ⁶⁴ Likewise, several studies have argued that the Great Irish Famine of 1845–1852 had a long-term trans-generational impact on the mental health of the Irish – both the famine survivors and their offspring.Footnote ⁶⁵

Recent advances in palaeo-epigenetics, based on ever-progressing novel aDNA analysis methods, have unlocked a huge potential to study epigenetic changes in past populations and already yielded some fascinating results in recent years.Footnote ⁶⁶ These studies, however, tend to focus on human populations of the distant past, tackling ‘big’ evolutionary questions, rather than more specific cases from more recent centuries. Thus, deploying palaeo-epigenetic research in the study of late-medieval England (or indeed any other medieval context) would be a conceptual and methodological novelty in itself. One exciting possibility would be to study the impact of the Great Famine and the Great Bovine Pestilence on the cohort of survivors and their progeny, by conjoining manorial documents with aDNA-based epigenetic analysis. The most obvious and straightforward way would be to utilise specimens from Black Death context burials. There are several excavated Black Death-associated cemeteries in England, including East Smithfield, West Smithfield (aka, the Charterhouse) in London, Hereford Cathedral, Thornton Abbey (Lincolnshire), the Hospital of St John the Evangelist and the parish cemeteries of All Saints and St Bene’t’s (all in Cambridge).Footnote ⁶⁷ Most of the excavated skeletons, now archived and curated at different repositories, have been sexed and aged. This can allow researchers to differentiate between pre-Great Famine/pre-Bovine Pestilence cohorts and those born during and after the two disasters, in order to see if the latter cohorts exhibit distinctive epigenetic markers, impacted by stress and deprivation. As far as the ‘famine’ cohort is concerned, it would be intriguing to establish if the epigenetic impact was the same or different among (1) those growing up during the famine; (2) being in utero during the famine; (3) born into the famine; and (4) born after the famine.Footnote ⁶⁸ Importantly, the most critical windows for environmentally induced epigenetic changes occur in utero (particularly in the first two trimesters) and during the slow growth phase (SGP) between ages 8 and 12.Footnote ⁶⁹

The potential impact of starvation during the Great Famine and dairy shortage after the Great Bovine Pestilence on health of English commoners has been studied in 2013 by Dewitte and Slavin, on the basis of manorial accounts and skeletal evidence from the East Smithfield cemetery. The main findings of that study were that (1) the Great Famine weeded out frailer individuals; (2) those born during or after the Great Famine and the Great Bovine Pestilence exhibited various pathological signs, including shorter stature, cribra orbitalia (lesions on orbital roofs), and porotic hyperostosis (lesions on cranial vault bones); (3) the same frail individuals were more likely to die during the Black Death than healthier individuals.Footnote ⁷⁰ Although this skeletal analysis yielded some important results, it could not go much beyond these ‘external’ and ‘visible’ pathologies and stress markers. aDNA-based epigenetic analysis of the same individuals would take our appreciation of the trans-generational impact of Great Famine starvation and ensuing dairy deficiency, as well as our understanding of the connection between the three disasters (the Great Famine, the Great Bovine Pestilence and the Black Death) to new heights altogether.

The torrential rain of 1314–1316 precipitating the subsistence crisis of 1315–1317 impacted not only humans but also their livestock, by (1) keeping animals exposed to cold and damp weather and (2) rendering abundant fodder resources rotten and infested with parasitic fungi, bacteria, worms, and gastropods.Footnote ⁷¹ All this appears to have impacted animal health and caused several ovine and bovine diseases, culminating in elevated mortality rates in the period 1314–1317, as recorded in manorial accounts.Footnote ⁷² Although mortality rates in bovine stocks were lower than those within sheep flocks, it was the former that paid the price in a longer term. Most of the cattle perishing in the 1319–1320 panzootic were the survivors of the 1314–1317 crisis, born or/and growing up in its course. What explains these high mortality rates? One cannot help but wondering if by the time the panzootic entered England in April 1319, local bovids were already frail and susceptible to various diseases – possibly due to epigenetic alterations brought about by malnourishment and stress of the preceding years.

The only way to take this argument out of its conjectural framework is to conduct a palaeo-epigenetic analysis deriving from aDNA of fourteenth-century cattle remains – similar to that outlined above in reference to human survivors of the Great Famine. This, however, can be trickier. Unlike some late-medieval mass human burials, firmly associated with the Black Death, all animal bones come from very approximately dated contexts – usually, merely designated as ‘Anglo-Saxon’, ‘high medieval’, or ‘later medieval’. Also, there is, at this stage, very little evidence of mass cattle burials (or, rather dump pits), but the ongoing archaeological work (in particular that of Annelise Binois) will hopefully reveal more of these.Footnote ⁷³ Nevertheless, archaeological publications (including some ‘grey’ literature) may occasionally contain some ‘hidden’ references to potential victims of animal epizootics. For instance, during the 1979 excavations at Bierton (Buckinghamshire), a whole carcass of an ox/bull was recovered from a thirteenth or fourteenth-century ditch – a very unusual feature, likely indicating that the animal was culled and not slaughtered/consumed because it was diseased.Footnote ⁷⁴ Identifying similar examples would be vitally important for the proposed palaeo-epigenetic studies.Footnote ⁷⁵

In other words, palaeo-epigenetics can, in theory, reveal the impact of the nutritional crises of the early fourteenth century – affecting both humans and their livestock, – by identifying particular (epi)genetic markers of stress and malnourishment in human and animal remains. While manorial documents (court rolls and accounts) do not contain any hints to such ‘hidden’ information, their importance is that they provide the factual basis and historical context for the same crises, in which epigenetic changes may have occurred.

Parchment material and molecular analysis

The rich corpus of manorial sources contains potential not only in its contents but also in its physical make-up – sheep parchment sheets.Footnote ⁷⁶ Most recently, Matthew Collins and his collaborators developed a novel technique of extracting and analysing the molecular and isotope compositions of animal lipids (arising from degraded animal fats) from sheep parchments to infer the animal’s diet, health, climate and environment – known as ‘Minimally Invasive Vacuum-Aided Extraction Technique’.Footnote ⁷⁷ As a proof of concept, Collin’s team applied this method to nine sacrificial membranes from 1765 to 1825 and membranes from the Chancery Parliament Rolls (1814–1820).Footnote ⁷⁸ This novel method has a huge potential, when applied to manorial documents, for several reasons.

Firstly, the vast majority of manorial documents (apart from damaged rolls and detached lists of manorial dues and payments) are precisely dated to a specific year. It is even possible to infer the seasonality of their composition: thus, manorial accounts would have been compiled during the annual auditing, taking place around or shortly after the end period of each account (usually Michaelmas, 29 September), while manorial court rolls would have been written during court sessions (meeting several times a year, with dates varying across manors). In some instances, surviving demesne accounts and court rolls are ‘clean’ copies of the original drafts; the same clean copies would be made short time after the draft version had been compiled.Footnote ⁷⁹ In some instances, draft court and account rolls survive alongside with their ‘clean’ versions. IPMs would be put together shortly after the death of a tenant-in-chief. We can assume that sheep membranes were made (of locally bred and slaughtered animals) not long before they were used as written materials – most likely in the same or previous year. Hence, the temporal gap between the date of an animal’s slaughter and that of the composition of a manorial document upon its tanned skin must not have been too big.

Secondly, as mentioned above, in many instances, we have series of excellent chronological runs, with very few gaps. As noted, Winchester Bishopric Pipe Rolls, containing grouped accounts for several dozens of manors, run from 1208 to 1209 until the 1470s (even though most of the demesnes were leased out to local tenants by the early 1450s), while Battle Abbey manors of Barnhorn and Alciston (both in Sussex) are blessed with accounts running from, respectively, the 1320s and 1330s until the 1490s.Footnote ⁸⁰ On some manors, court sessions would meet as often as once every month or even once every three weeks, and this can potentially shed some light on seasonal variations in diet and weather. Some examples of such manors with frequent sessions and very good or excellent surviving runs include the manors of Thornbury (Gloucestershire), Kirton in Lindsey (Lincolnshire), Halesowen (on the Shropshire-Worcestershire border), Alrewas (Staffordshire), Knaresborough, Methley, Wakefield (all in Yorkshire), and the manorial complex of Dyffryn Clwyd (Denbighshire).Footnote ⁸¹ Finally, the fact that manorialism was present, albeit with various degrees of prevalence and intensity, in most regions of England and Wales, implies that court rolls and manorial accounts survive in every county. This fact allows to survey and analyse parchments from all regions, by sampling several manors in each county.

The suggested comprehensive approach can create much new fine-tuned data that will fill up many gaps in our knowledge and potentially help solving some ongoing debates – including some outlined above, such as the impact of weather on annual crop yields and the question of synchronicity between low crop yields reported by manorial officials and weather anomalies inferred from the natural record. Indeed, annual dendrochronological record can potentially be supplemented by annual lipid record, to improve the quality of fine-tuned climatic record – rainfall levels in particular. In theory, lipid-based precipitation reconstruction has a strong potential to become an important supplement for dendrochronological data, and particularly to make up for its numerous drawbacks, discussed above. Likewise, because lipids and proteins can also be used to infer different aspects of natural environment, in which sheep were reared, this novel technique can be deployed in order to supplement palaeo-ecological record (such as pollen remains) with a much finer-tuned (annual) series of lipid record.Footnote ⁸²

One aspect of particular interest is the change in the quality and quantity of grassland and its potential impact on wool yields. Manorial accounts provide ample evidence that per-sheep yields have gradually declined in the course of the fifteenth century (Figure 5). Was this phenomenon caused by endogenous factors – for instance, consistently failed sheep management?Footnote ⁸³ Or was it triggered by some exogenous aspects – thus, the cold and dry weather with the onset of the so-called Little Ice Age in the fifteenth century, and in particular during the combination of the so-called ‘Spörer Minimum’ of c.1420–1470 and the ‘Renaissance Drought’ of c.1437–1480, which may have shortened the grazing season and reduced the available biomass for livestock?Footnote ⁸⁴ Or could we blame some epigenetic and potentially anatomic alterations, whereby English sheep became smaller and frailer than their thirteenth- and fourteenth-century ancestors? Or, was it, in fact, a much more complex phenomenon, caused by a combination of all the above? Currently, there is contradictory evidence about the change in sheep height in the late-medieval era. On the one hand, there is some zoometric indication that fifteenth-century sheep heights and bones were shorter than those in the thirteenth and fourteenth centuries.Footnote ⁸⁵ On the other hand, there is some evidence (from Dudley Castle and London) indicating a long-term ‘improvement’ in sheep size from the thirteenth century into the modern era, with fifteenth-century skeletons being larger than those in the previous centuries.Footnote ⁸⁶ A more comprehensive analysis, requiring interdisciplinary cooperation among historians, archaeologists, and microbiologists, is necessary to draw clearer and more definitive conclusions about the potential effects of climate change on epigenetic modifications and anatomical changes in sheep during the late-medieval period. Such integrative research would, among other benefits, help clarify the causes of the mysterious decline in wool yields beginning in the 1430s.

Figure 5. English Wool Yields, 1211–1495 (in Quinquennial Means): (a) in Lbs of Fleece per Adult Sheep on Winchester Bishopric Demesnes, 1211–1454 and (b) in Lbs of Fleece per Lambs on Alciston Demesne (Sussex), 1336–1495. Both Series are Indexed on 1336–1370 (1336–1370 = 100 = 1.54 lbs for Adult Sheep Fleece on Winchester Bishopric Demesnes and 0.50 lbs for Lamb Fleece on Alciston Demesne).

Source: Slavin, ‘Merchants and Mites’; Manorial Accounts Database; M. J. Stephenson, ‘Wool Yields in the Medieval Economy’. The Economic History Review 41, no. 3 (1988): 368–91; East Sussex and Brighton and Hove Record Office (the Keep), SAS-G/644/91–139 (Alciston manorial accounts). Notes: The post-1454 period trends are reconstructed solely on the basis of Alciston wool yields, as no comparative material from other demesne was available.