Introduction
In medieval Europe, deer hunting played an important role in defining social separation and discrimination in access to resources. It was mainly conducted to the advantage of the wealthy and powerful, who used the practice as a potent status symbol, but it had important implications for the less advantaged too. Though the hunting of most animals was restricted to the aristocracy, the chasing of large game had an especially high profile. Deer are, along with wild boar, the most widespread large wild mammals in Europe and it is therefore unsurprising that they played a major role in the characterization of hunting.
Deer could be hunted in the forest, in a context that approached pure wilderness, or in the much tamer context of deer parks, generally associated with castles and manor houses. Though venison would provide a substantial contribution to the diet of the privileged (Birrell, Reference Birrell1992), the greatest significance of hunting was social, according to a process that followed strict, even ritualized, rules. This ritual did not just affect the chase and the kill, but also the treatment of the carcass and the distribution of its parts, which were often considered the climax of the hunt. Such a ritual is quite well known from historical and archaeological records, but there are still many uncertainties about its origin, spread, and cultural connotations.
This article explores the archaeological and, comparatively, historical evidence to address the following questions:
• What is the origin of the ritualized butchery of the deer carcass and the formalized distribution of its parts?
• How widespread was this phenomenon in time and space?
• Did it apply to one or more deer species?
• What can this formalized butchery tell us about cultural connections between different people in medieval Europe?
Material and Methods
This article concerns mainly three study areas: England, northern France, and southern Italy (Figure 1). These regions have been chosen because they were partly or entirely occupied by the Normans during the Middle Ages and therefore potentially shared some cultural traits. Evidence from adjacent areas that were not directly occupied by the Normans, such as Wales, central France, and northern Italy, is mentioned comparatively. The current research covers the period ranging from the eleventh to the seventeenth century, i.e. the (later) medieval to early modern periods.
Figure 1. Distribution of sites mentioned, with areas of Norman occupation highlighted. Triangle: solid evidence; circle: tentative evidence; cross: no evidence or no information. The rectangles include multiple sites in Paris and Palermo (map template downloaded from www.d-maps.com).
Since we consider ‘unmaking’ rather than, more generically, deer hunting, we have only included archaeological sites that have yielded a sufficient number of deer bones to be informative about the relative occurrence of body parts. Mostly, these are castle sites.
Large-scale analysis of body part distributions is notoriously hazardous, as this evidence is highly dependent on local taphonomic factors and quantification strategies adopted by individual researchers. Obtaining quantitative information concerning the occurrence of a certain anatomical pattern in a region or period therefore runs the risk of being unreliable. Consequently, the present study focuses on individual cases, which have provided information pertinent to our research questions. Although this means that most of the information presented here is largely qualitative, some comparison between areas and periods is also undertaken. The quantitative information must, however, be treated cautiously because the number of informative sites is inevitably limited.
Extracting the anatomical element frequencies from individual reports was complicated by the variability of the quantification systems used by different researchers and the way the data were presented. Hence different decisions (explained in Table 1) had to be taken in the treatment of the data from each report. Since there are limitations in the comparability of data from individual sites, the results presented must be used exclusively to identify broad trends in the distribution of the data.
Table 1. Red and fallow deer. Comparison of fore and hindlimb frequencies for the sites mentioned in the text. The following elements are included in the forelimb: humerus, radius, metacarpal; in the hindlimb: femur, tibia, metatarsal. If in bold, the numbers represent the minimum number of individuals (MNI) by anatomical element (also known as MNE or MAU, according to the way they were calculated and alternative terminologies). Numbers of identified specimens (NISP) are not in bold. Note that for Okehampton only distal ends with fusion evidence are counted, apart from the calcaneum (astragalus not counted). The data from Vatteville-la-Rue may include a small quantity of roe and fallow deer as the three species are undifferentiated in the original publication (but red deer is largely predominant). The data from Monte Iato combine identifications of ‘Fallow Deer’ with ‘Fallow/Red Deer’. There was no positive identification of red deer at this site, making it likely that the uncertain identifications belong to fallow deer. For Andone, only distal ends were considered, to enhance comparability. For Mehun-sur-Yèvre only percentages of individual body parts were provided, but the total NISP for red deer is 45.
Deer Hunting and Butchery: Background
Five deer species have historically lived in Europe, and these are discussed in the online Supplementary Material. The three species that are relevant to this article are red deer (Cervus elaphus), fallow deer (Dama dama), and, to a lesser extent, roe deer (Capreolus capreolus) (Figures 2–4). The two medieval hunting methods used to hunt deer are also described in the Supplementary Material. Par force (by strength) is the strategy that includes the unmaking process.
Figure 2. Red deer (https://depositphotos.com/ - royalty free stock photos).
Figure 3. Fallow deer (https://depositphotos.com/ - royalty free stock photos).
Figure 4. Roe deer (courtesy of Bjørn Reidar Olsson, University Museum, Bergen, Norway).
The dismemberment or, to use the terminology used by most historical sources, ‘unmaking’ of the deer carcass was an essential phase of the deer hunt in the Middle Ages, with important social and symbolic connotations. The practice is described in several hunting treatises and other medieval literature; numerous publications emphasize the implications for historical (Cummins, Reference Cummins1988; Almond, Reference Almond2011; MacGregor, Reference MacGregor2012), archaeological (Sykes, Reference Sykes and Pluskowski2007a; Thomas, Reference Thomas and Pluskowski2007), and anthropological (Pratt, Reference Pratt2013) interpretations. To stress the importance of this stage of the hunt, it is worth quoting Cummins (Reference Cummins1988: 41): ‘There was a recognised way of doing everything: formulaic cries, commands and horn-calls; ritualised ceremonies. The most striking imposition of ceremonial and activities […] came after the death, in the flaying and butchering (“unmaking” […]) of the animal.’ Cummins’ account of this unmaking is based on textual and iconographic sources, the latter relying especially on the pictorial representations included in Le livre de chasse de Gaston Phébus (late fourteenth century, south-western France). This includes key images of the hart held on its back while butchered and skinned (BNF, MS. fr 616, fol. 85). A similar image is also represented in the approximately contemporary Livre du roy Modus et de la royne Ratio tentatively attributed to Henri de Ferrières (KBR, MSS. 10218–19). Le livre de chasse also illustrates the curée, the stage in which the leftovers of the deer carcass are given to the hounds, while the stag's head (recognizable from its well-developed antlers) is carried away (BNF, MS. fr. 616, fol. 72).
Key to archaeological interpretation is that different parts of the deer carcass were destined for different purposes. Briefly, the pelvis (os courbin or corbyn bone) was for the crows and/or ravens, the left shoulder for the hunter or the ‘unmaker’, the right shoulder for the forester, and the haunches for the lord (Thomas, Reference Thomas and Pluskowski2007: 128). There was, however, a degree of variation, depending on the instructions provided by different hunting manuals. The butchery process that led to the separation of these different cuts generally took place on the spot, but in some cases the carcass was carried whole to the hunting lodge before dismemberment (Cummins, Reference Cummins1988: 42). It is somewhat surprising that the pelvis was thrown away as it potentially carries a substantial amount of flesh (contra Cummins, Reference Cummins1988: 42) but, presumably, it was detached from the body in such a way that only a limited amount of meat was wasted—unless feeding the corvids represented an important part of the ritual.
The earliest source—from which all later literature takes inspiration—to report the practice of ritualized unmaking dates to the early thirteenth century (c. 1210). It is found in the romance Tristan by Gottfried von Strassburg, written in Middle High German. Tristan (who is French) explains the practice of unmaking to his Cornish hosts, which shows the foreign (at least to the Cornish) origin of the ritualized dismemberment of the deer (cf. Cummins, Reference Cummins1988: 43; Almond, Reference Almond2011: 75). The origins of the practice described by Tristan are unknown (Almond, Reference Almond2011: 76). The poem from Picardy La chase dou cerf is only slightly later (c. 1250); it also describes the formulaic breaking-up of the deer carcass (Almond, Reference Almond2011: 64). English sources are later and tend to plagiarize earlier French sources such Phébus. The Art of Hunting (1327) by William Twiti (MacGregor, Reference MacGregor2012: 114) predates Phébus but was originally written in Norman French, thus betraying its geographic area of inspiration. Other well-known English hunting treatises, i.e. the Master of Game, the Tretyse off Huntyng, the Boke of St Albans, and the Noble Art of Venerie, date from the fifteenth and sixteenth centuries and do not provide much original information concerning the practice of unmaking (cf. Cummins, Reference Cummins1988: 41–43; Almond, Reference Almond2011: 75–76; MacGregor, Reference MacGregor2012: 114–15). There are also French literary sources that are later than Phébus and Henri de Ferrières, such as Jacques de Brézé's La chasse, which adds variations to the instructions in Le livre de chasse (Cummins, Reference Cummins1988: 41). Brézé was the sénéchal (bailiff) of Normandy, indicating that at the very latest by the late fifteenth century (but probably a lot earlier) the unmaking of deer was a well-established practice in that region.
The focus of the butchery instructions is the hart, but the male fallow deer (buck) was subjected to similar treatment (Cummins, Reference Cummins1988: 87). Historical evidence is contradictory concerning the handling of the roe deer carcass, with some sources suggesting that it was similar to that of the hart and others stating the opposite (Cummins, Reference Cummins1988: 91).
There are no Italian written sources we are aware of that mention unmaking. Frederick II's De arte venandi cum avibus (mid-thirteenth century) could have been a source, but there is no mention, though it must be considered that the book is largely concerned with falconry as its title suggests. Cortonesi's (Reference Cortonesi1995) and Arrigoni Martelli's (Reference Arrigoni Martelli2015) historical reviews of late medieval hunting in Italy make no mention of deer unmaking, which indicates the silence of the sources on the issue. The historical and iconographic review of medieval hunting in Veneto (north-eastern Italy) undertaken by the Centro di Documentazione per la Storia della Valpolicella (1990) also does not refer to any form of ritualized deer hunting. Although it is perilous to build an argument on negative evidence, it seems unlikely that such an overt activity would have gone unnoticed by the observers and reporters of the time. Note, however, that most of the literature mentioned above does not focus on areas of Norman influence.
Deer Butchery: The Archaeological Evidence
The key archaeological evidence for identifying the potential unmaking of the deer carcass is represented by the bone frequency of different body parts. This article focuses on high-status sites, where such evidence is more frequent and the most abundant assemblages are found. If the advice of the hunting manuals discussed earlier was followed, then we should expect the bones present in the hindlimbs to predominate at these sites. With the pelvis given to the corvids, these would have been represented by some or all of the remaining hindlimb bones: femur, tibia, and possibly also tarsals, metatarsals, and phalanges. The destination of the head varies according to the source, and therefore we must be prepared to account for its potential presence as well as absence.
Although not all anatomical elements have the same rate of survival in archaeological contexts, there is no reason to think that, overall, uneven representation of fore and hindlimb bones can be caused by taphonomic processes. Nevertheless, the methods used to define different body parts are not always explained in zooarchaeological reports, adding to the uncertainty of what is counted and making us wary of relying on small differences in the representation of anatomical elements.
Although skull fragments tend to be poorly preserved, teeth are highly durable, therefore making the head unlikely to be under-represented through preservation bias.
Overall, it is reasonable to assume that when there is clear unevenness in the representation of fore and hindlimb bones, human behaviour should be regarded as the cause. However, marginal differences may be the consequence of taphonomic and counting biases or the vagaries of small sample size.
Here, the discussion of the representation of anatomical elements of deer focuses mainly on red and fallow deer, the two cervid species for which we have more abundant archaeological and historical evidence, with brief references to roe deer. The evidence is presented separately for the three main geographic areas. The raw data are given in Table 1.
England
England is discussed first as it has the clearest evidence, which generated the original research questions subsequently examined in other geographic areas. This does not imply that the practice of unmaking originated in England.
The first archaeological review of the unevenness of deer body part representation at medieval sites is by Albarella & Davis (Reference Albarella and Davis1996: 33–34), who, having identified such a bias in fallow deer at Launceston Castle (Cornwall), compared it with parallel evidence from other high-status sites. Four other castles showed a predominance of hindlimb bones: Okehampton Castle (Devon; Maltby, Reference Maltby and Higham1982), Sandal Castle (West Yorkshire; Griffith et al., Reference Griffith, Halstead, MacLean, Rowley-Conwy, Mayes and Butler1983), Barnard Castle (Durham; Jones et al., Reference Jones, Sly, Simpson, Rackham and Locker1985), and Prudhoe Castle (Northumberland; Davis, Reference Davis1987). At most of these sites the evidence concerns fallow deer; but, at Barnard Castle and Prudhoe Castle, it is mainly observed for red deer. Invariably, hindlimb bones are better represented than other parts of the body, confirming the high-status preference for haunches as suggested by literary sources. For Launceston, a ratio of eight to ten haunches to a complete fallow deer carcass was proposed. Perhaps the locally hunted deer were fully processed on-site while the more numerous animals caught further afield were butchered off-site and only selected parts brought back.
An updated review of the evidence was published by Thomas (Reference Thomas and Pluskowski2007), prompted by his work on the animal bones from Dudley Castle (West Midlands; Thomas, Reference Thomas2005) that revealed a similar pattern. To the sites mentioned above, Thomas could add the newly recovered evidence from Pontefract Castle (West Yorkshire; Richardson, Reference Richardson and Roberts2002) and Faccombe Netherton (Hampshire; Sadler, Reference Sadler and Fairbrother1990), the only manor house where evidence of unmaking has been detected. At these sites, the predominance of hindlimbs mainly characterizes fallow deer but also red deer in the later phase at Dudley (fourteenth–sixteenth century).
The higher representation of hindlimb bones at castle sites should logically be complemented by a bias towards the other anatomical elements (forelimbs, possibly heads) at other sites, but the sample size of deer bones is small on lower-status sites, preventing us from identifying a well-defined pattern in body part distribution. Town sites also tend to have a small proportion of deer bones (Albarella & Davis, Reference Albarella and Davis1996: fig. 41). This is partly compensated by the large size of some of the animal bone assemblages, but the precise context (i.e. who consumed what?) is often hard to establish in urban milieus. Working with such limited evidence, Sykes (Reference Sykes and Pluskowski2007a: figs. 11.4 and 11.5) has nevertheless established that forelimb bones predominated at keepers’ residences and cranial elements on some low-status rural sites. She comments that it would be wrong to interpret venison as only affecting the social life and status of aristocratic people.
Hindlimb predominance is thus observed in high-status sites across England, including the south-west (e.g. Launceston), the centre (e.g. Dudley), and the north (e.g. Prudhoe). It is mainly confined to secular sites, though it has been tentatively identified for fallow deer at Austin Friars in Leicester (Thawley, Reference Thawley, Mellor and Pearce1981; Albarella, Reference Albarella2019: 224). Chronologically, the evidence ranges mainly from the twelfth to the fifteenth century, with the earliest case probably represented by the twelfth-century phase at Barnard Castle. It is unclear whether the practice was introduced immediately after the Norman Conquest, but it seems to have been well-established at the very latest a century later. At Launceston Castle, the hindlimb predominance in fallow deer can also be observed in the post-medieval phases (Albarella & Davis, Reference Albarella and Davis1996: tab. 11), but more tentatively given the smaller sample size (the overall frequency of deer drops in the later phases as the site declines in status). The evidence of a post-medieval unmaking is, however, quite convincing at Barnard Castle and Sandal Castle.
Maltby & Hambleton (Reference Maltby, Hambleton, Baker, Carden and Madgwick2015: 193–94) have pointed out that a prevalence of deer haunches was not limited to England, being observed at the southern Welsh site of Laugharne Castle and, more tentatively, at Dryslwyn Castle (south Wales; Gidney, Reference Gidney and Caple2007) and Hen Domen (central Wales; Browne, Reference Browne, Higham and Barker2000). The relevant species in these cases is red deer as fallow deer is uncommon at Welsh medieval sites. The unmaking evidence at Laugharne is present from the early twelfth century to the end of the Middle Ages, thus making it one of the earliest sites characterized by this pattern.
France
Tristan's story suggests that the practice of unmaking is likely to have originated in France, but the archaeological evidence for France is neither as abundant nor as clear as for England.
As in England, fallow deer was not native and starts appearing only in the eleventh–twelfth centuries, in different parts of the country, inside and outside Normandy (Binois-Roman et al., Reference Binois-Roman, Ther and Borvon2022). It is, however, not as common as in England and provides no clear evidence of unmaking (but see below). There is, conversely, a clear predominance of red deer hindlimb bones at the castle of Mehun-sur-Yèvre in central France (Cher; Jouanin, Reference Jouanin and Bon2011); at that site, for roe deer, there is a predominance of limb extremities, while fallow deer is absent. Two aspects concerning the evidence for unmaking at this site are worth noting: first, its location outside the main area of direct Norman influence, and second, its rather early date (eleventh century) predating any English evidence.
Further evidence of unmaking in France is rather sparse, but at Logis Royaux (Indre et Loire, central-western France) in the eleventh–twelfth century-phase, red deer hindlimb bones dominate the cervid assemblage (Duval, Reference Duval, Papin and Gaultier2020), though the sample is small. In addition to the bones counted in our Table 1, there are also three scapho-cuboids (hindlimb) and no carpals. The roe deer sample is too small to provide a reliable body part distribution pattern and fallow deer is absent.
Although less chronologically tight (twelfth–fifteenth century), at the castle of Vatteville-la-Rue (Normandy; Sykes, Reference Sykes2007b: 20–22) the combined total of the three deer species indicates a predominance of hindlimbs, though not especially pronounced (the hind to forelimb ratio is 3:1 at most; Sykes, Reference Sykes2007b: fig. 17). Although the three deer species were left undifferentiated in this analysis, the pattern is bound to be largely determined by the distribution of red deer as this species is by far more common than roe and fallow deer (Sykes, Reference Sykes2007b: tab. 6).
Other tentative evidence derives from the residence of the Counts of Anjou in Tours (central-western France; eleventh–twelfth century). Here too there is no fallow deer, but the small bone assemblage shows a slight predominance of hindlimbs for both red and roe deer (Genies, Reference Genies2011). If any unmaking was practised, this only affected a small proportion of the hunted animals, and/or there was a mixture of bones deriving from consumption by people of variable social status.
At the eleventh-century high-status site of Andone (Villejoubert, Charente, south-western France), red deer is abundant and the anatomical element distribution is uneven but not in the way we might have expected, as radius and tibia are the most common bones (Rodet-Belarbi, Reference Rodet-Belarbi and Bourgeaois2009; Bourgeois, Reference Bourgeois2011). Hindlimb bones are slightly more abundant, but there is no clear evidence of unmaking. Either it was not undertaken, or different activities became mixed in the archaeological assemblage. Roe deer is rare and, once again, there is no fallow deer.
It is notable that other high-status sites, such as the Château Ganne at La Pommeraye (Calvados in Normandy, eleventh–fourteenth century; Borvon & Flambard Héricher, Reference Borvon, Flambard Héricher, Cocula and Combet2014), Château Boves (Picardy, twelfth century), Le Louvre-Cour Carrée (Paris, thirteenth–sixteenth century), and the Château de Courtrai (Lille, fourteenth century) (all mentioned by Clavel, Reference Clavel2001) have limited evidence of deer hunting and no evidence of unmaking. The same is true for ecclesiastic sites such as the monastery of La Charité-sur-Loire (central France; eleventh–twelfth century; Audouin-Rouzeau, Reference Audouin-Rouzeau1986; see Clavel, Reference Clavel2001 for other ecclesiastic sites in northern France). This is not surprising, as in England too some castle sites have not yielded evidence that is compatible with typical high-status meat choices. This may be due to a variety of taphonomic and contextual factors and should not necessarily be interpreted as evidence of absence.
At the later medieval castle of Suscinio (Brittany, fourteenth century; Vincent et al. Reference Vincent, Dubois, Borvon, Brunie, Daré and Fray2017), fallow deer dominates over red deer (Borvon, Reference Borvon, Vincent and Dubois2017), thus mirroring the increasing occurrence of the latter species over time in England (Holmes, Reference Holmes2017: 88; Albarella, Reference Albarella2019: 203). Roe deer is also present. The fallow deer sample size is too small to provide a reliable analysis of the body part distribution, but, overall, there is no evident bias in the distribution of body parts (Borvon, Reference Borvon, Vincent and Dubois2017: 341 and annexe 2a). More recent work on the later kitchen context at Suscinio (late fifteenth–early sixteenth century) confirms the predominance of fallow deer, with hindlimb bones more than twice as common as those of the forelimbs (Aurélia Borvon, pers. comm. 11 November 2023), hinting at the possibility that some unmaking was taking place.
At the even later castle of Vincennes (Val-de-Marne) just outside Paris, red deer is absent, but fallow deer is present with just one specimen in the late fifteenth century phase. In the much larger assemblage dated to the sixteenth century, however, there are more than 100 fallow deer remains (Clavel, Reference Clavel2001: 17–18), confirming that this species became more common in later times. Unfortunately, the body part analysis is not available for this site, precluding any comments on the potential practice of unmaking.
In summary, the French evidence for unmaking is not widespread but slightly predates the English data (eleventh versus twelfth century) and occurs in different parts of the country. So far, unlike in England where it also affects fallow deer, the evidence is limited to the processing of red deer carcasses; the data from Suscinio, and potentially Vincennes, however, show that this is worth exploring further.
Italy
A relatively recent survey of medieval zooarchaeological data from the whole of Italy makes no mention of any evidence of deer unmaking, despite including an extensive section on deer hunting (Salvadori, Reference Salvadori2015). Here, we focus primarily on the southern Italian regions that were more directly influenced by Norman culture.
The tightest chronology comes from the Angevin castle of Lagopesole (Avigliano, Basilicata; Fiorillo, Reference Fiorillo2005) where the clearest and most abundant evidence is from the late thirteenth century. Both red and fallow deer are present and relatively common, while roe deer is not. The most frequent bones are the tibia, metatarsal, and femur in both species, therefore indicating a clear preference for the hindlimb. Cranial and forelimb elements are far less common but sufficiently represented to indicate that, if unmaking was taking place, some of the less prestigious body parts also found their way to the castle, perhaps because full carcasses were processed on site.
The castle of Calathamet in north-western Sicily (Di Patti et al., Reference Di Patti Lupo, Di Salvo, Di Trapani, Schimenti, Lesnes, Poisson and de Bresc2013) has archaeological deposits that are less well dated than those of Lagopesole (twelfth–fourteenth century) but nevertheless attributable to the later medieval period. The tibia is the most common element for red deer, but it is for fallow deer that the hindlimb bones are better represented (though foot bones, i.e. metapodials and phalanges are absent); for forelimbs, the radius is quite often present. It seems that several parts of the body were present, but hindlimbs were probably predominant. The two species are roughly equally represented, but roe deer is absent.
The site of Monte Iato (also in north-western Sicily; Kistler et al., Reference Kistler, Öhlinger, Dauth, Mölk, Irovec, Wimmer and Forstenpointner2018), thought to be a late Muslim outpost, has a fairly abundant assemblage of fallow deer remains in a thirteenth-century context (Benjamin Wimmer, pers. comm. 16 January 2023). There was no positive identification of either red or roe deer. Although hindlimb bones are clearly more common than those of the forelimbs, most body parts are documented, with mandibles particularly well-represented. If any unmaking was taking place, this is partly masked by the contemporary use of different butchery practices.
The rural site of Brucato, again in north-western Sicily (Beck-Bossard, Reference Beck-Bossard1981; Bossard-Beck, Reference Bossard-Beck and Pesez1984; Bossard-Beck & Maccari-Poisson, Reference Bossard-Beck, Maccari-Poisson and Pesez1984), yielded an assemblage of animal bones starting in the eleventh–twelfth century but whose main component is dated to the fourteenth century. Since it is a village, there was no expectation of finding evidence of high-status food consumption. Yet large game hunting was common and the three deer species (red, fallow, and roe) are all present in fair quantities. Of the three, fallow deer is the most frequent (three times as abundant as the other two species); and, perhaps surprisingly, it shows a prevalence of hindlimbs. This is clear in the contrast between the femur and humerus as well as the metatarsal and metacarpal, but only marginal in the comparison of the tibia and radius. Plainly, this is not a straightforward unmaking pattern, but the unevenness of the anatomical element representation is an indication that different body parts were subject to different treatments. People of different status are likely to have lived in the village.
The high-status Norman Palace in Palermo (Sicily, twelfth century; Aniceti, Reference Aniceti2022) has yielded a small assemblage but it is worth noting the presence of both red and fallow deer in a context that is earlier than the previously mentioned cases (although fallow deer was already present at Brucato in its eleventh–twelfth century phase). Only three cervid remains could be identified (two fallow and one red), all from the hindlimb.
As in England and France, some high-status sites have provided almost no evidence of large game hunting, let alone unmaking. These include sites such as the castle at Fiumenidisi (Messina, Sicily; Villari, Reference Villari1988), the Palazzo Steri (Palermo, Sicily; Di Patti & Lupo, Reference Di Patti and Lupo2009), and the castle at Canne della Battaglia (northern Apulia; De Venuto, Reference De Venuto2013). To understand the reasons for this pattern, contexts and site formation processes must be analysed in detail, which is beyond the scope of this article. We must, however, acknowledge that not all high-status sites provide evidence of conspicuous consumption.
Conclusions
Although reasonably well investigated in England, the archaeology of deer unmaking has largely been neglected in France and Italy. Yet the historical evidence suggests it was a widespread phenomenon of the European later Middle Ages, and our case studies show that some geographically distant elite sites provide archaeological evidence of the practice. Although many gaps in our knowledge of this practice exist, there is enough evidence to develop new interpretations.
While we have seen that some lower-status people contributed to the practice of unmaking (Sykes, Reference Sykes and Pluskowski2007a) and, in some cases, would benefit from it, the evidence largely relies on high-status sites, where it is more easily detectable. This is partly because they tend to provide large animal bone assemblages and partly because the sites themselves are more prominent and visible, and therefore more frequently excavated.
Following the earliest currently known documentary (Tristan) and archaeological (Mehun-sur-Yèvre) evidence, the origins of the unmaking phenomenon should probably be sought in France around the eleventh century. The language of the deer chase was typically French and the best-known manual describing the unmaking as well as all other stages of ritual hunting (Phébus) also comes from France. This book was widely plagiarized by later authors, giving the impression that the documentary evidence is richer than it really is. The earliest evidence from England (twelfth-century Brandon Castle) and Wales (twelfth-century Laugharne Castle) is not much later, indicating that it did not take long for the practice to be exported north. In Italy it occurs in the high Middle Ages (in late thirteenth-century Lagopesole). We must, however, consider that the phenomenon has been largely overlooked by Italian archaeologists and that our interpretations rely on our analysis of the datasets rather than observations by the original authors.
Although the starting point in our investigation of unmaking has been the Norman world, this geographical constraint seems unnecessarily limiting. Since there is no evidence of the practice in Anglo-Saxon England, it seems reasonable to assume that it was introduced by the Normans following the eleventh-century Norman Conquest, but in France the ritualized hunting of deer is not confined to Normandy. The author of the earliest documentary evidence for the unmaking practice, Gottfried von Strassburg, was from Alsace. Gaston Phébus—unlike Tristan, a real historical character—was from southern France. While the latter's treatise dates to the late fourteenth century, thirteenth-century evidence exists in Picardy (La chase dou cerf) as well as fourteenth-century Normandy (Jacques de Brézé's La chasse). The archaeological evidence is geographically widespread, with examples from central, central-western, and south-western France as well as Normandy and, tentatively, Brittany (Figure 1). The Normans, therefore, may have been responsible for spreading the practice of unmaking far and wide, but they were not the only people to practice it.
The Italian case is more problematic as it relies solely on the archaeological evidence, the written sources being apparently silent on the issue. The animal bone data are also not as straightforward to interpret as those from England; nevertheless, two sites suggest that the existence of the practice in one form or another is plausible. The lack of evidence in Italy outside the area of Norman influence would suggest that the Normans were responsible for its introduction, as in Britain. Future work should clarify the geographic and cultural limits of the ritualized art of hunting in Italy.
Historic and archaeological data are consistent in indicating that unmaking was applied to both red and fallow deer (Figure 5), despite the two species being hunted in different places and using different strategies. Red deer was largely a beast of the forest, while most fallow deer were hunted in deer parks (e.g. Cummins, Reference Cummins1988). The case for roe deer is more problematic, with historical and archaeological evidence for the unmaking of this animal rather scant. Archaeologically, the issue is compounded by the fact that roe deer tends to be less common than the other two deer species, thus producing sample sizes that are often inadequate for a reliable body part analysis.
Figure 5. Number of sites reported in this article with evidence of ‘unmaking’ for red deer and fallow deer from medieval England and Wales, France, and Italy. Black: solid evidence; grey: tentative evidence.
At early sites, such as Mehun-sur-Yèvre and Brandon Castle, the evidence exclusively or largely concerns red deer, because fallow deer would only become more common in the later Middle Ages. The distribution of anatomical elements shows that there is no consistent indication that the carcasses of the two species were treated differentially. In France, our archaeological evidence is almost entirely based on the treatment of red deer carcasses because, with a few exceptions, this was the most common species hunted throughout the Middle Ages and most of the known cases date to the early second millennium ad, when fallow deer was absent or rare. In Italy, conversely, the emphasis appears to be clearer for fallow deer but this is partly due to a chronological trend, as the Italian sites are later in date.
Since the fallow deer is an eastern Mediterranean species, and since it was fully introduced to the European countryside approximately at the same time as the practice of unmaking was established, it is tempting to suggest that the two phenomena were associated. However, there is no evidence that this was the case. We have seen that when fallow deer was still relatively uncommon, unmaking was well-established in the treatment of red deer carcasses. The written sources focus far more on the red than the fallow deer, which suggests that the introduction of fallow deer contributed to the expansion of the practice but not to its establishment.
In terms of our historical perspective of the medieval world, the practice of unmaking adds to our understanding of the cultural connections between the Normans of the Mediterranean and northern Europe, as it shows that they shared ritual and procurement practices as well as a common geographic origin.
Although the practice underscores the inequality of medieval society, it also brought together people of different social ranks, whose lives were variously affected by the rituality of the hunt. Unsurprisingly, the peasantry was segregated from such activity and had to resort to the great risk of poaching to obtain its share of wild game. People of higher social rank would, however, also hunt illegally from time to time (Manning, Reference Manning1993). On poaching sites, no ritualized butchery should be expected (see Holmes, Reference Holmes, Baker, Carden and Madgwick2015).
While this article provides new and illuminating evidence on the practice of unmaking in medieval Europe, there is still much to be explored. Zooarchaeologists must apply sound knowledge of taphonomy and quantification to assemblages of an appropriate sample size, but this does not require sophisticated equipment to be undertaken. Our method should be extended to other sites and areas to reconstruct a full geography of deer unmaking in medieval Europe and understand how widespread the practice was beyond the Norman world. Traditional zooarchaeology continually provides new opportunities to understand many different and still unknown facets of our past. For the medieval world, it is, however, necessary to combine archaeological and historical sources, in an open and constructive dialogue between the different disciplines. We have shown that the ritual of unmaking comprises a significant practice within medieval society in several areas of Europe, but without the integration of the zooarchaeological evidence with that of the written sources our understanding would be far more muddled and incomplete.
Supplementary Material
To view supplementary material for this article, please visit https://doi.org/10.1017/eaa.2024.11.
Acknowledgements
Umberto Albarella could work on this article thanks to a Leverhulme Major Research Fellowship. We would like to thank Annelise Binois, Aurélia Borvon, Colin Duval, and Benjamin Wimmer for allowing us to refer to their unpublished work and for advice, and Isabelle Rodet-Belarbi for help with references. Maeve Moorcroft and Stephanie Baron provided valuable comments on a first draft, while Catherine Frieman and Madeleine Hummler helped substantially with the final editing. We are grateful to two anonymous reviewers for their comments. Florence Gilham helped us with the formatting of the images.
Open access
