Article Text


Excavating the history of ancient veterinary practices
  1. Annelise Binois, DVM, PhD
  1. Université de Paris 1 Panthéon-Sorbonne
  1. e-mail: annelise.binois{at}


People have lived alongside animals for many centuries, and it can be assumed that they also treated them when they were sick. There are some texts that describe early veterinary care, but details are lacking and medical instruments are hard to come by in archaeological sites. However, as Annelise Binois explains, archaeological findings can provide a lot more information than might be thought, and can start to give a glimpse of the vast history of veterinary medicine.

Statistics from

YOU can almost picture the scene. A pasture a few kilometres away from the seaside, on the French coast of the North Sea; the springtime winds shake the frameposts of the lightly built stable. Inside, the cow has been calving for several hours already. The calf is badly positioned, with its right forelimb extended backwards, keeping it from entering the birth canal, but no-one in attendance knows how to deal with the situation. In an attempt to save his cow, the farmer pulls with all his might on the left forefoot of the now-dead calf, pulls until the cannon bone breaks, but only manages to wedge the calf further into the maternal pelvis. The following day, both animals are unfortunately dead. The dismayed farmer digs out a deep pit in the sandy ground just outside the stable wall and buries the carcases, but skins the cow beforehand, to save the hide and cut his losses.

To older veterinarians – and a few younger ones perhaps – this story may sound familiar, albeit without the skinning. And yet the scene actually took place in the late Middle Ages, in the 14th century to be more precise. No historical texts attest to this minor event of the probably uneventful life of an unknown farmer – a dead cow among so many others in that place and time – but most details included in the narrative above are nonetheless accurate and verified. Archaeologists carrying out rescue excavations on the site of a future housing estate in Téteghem in northern France unearthed the postholes of the ancient stable and the pit containing the two skeletons, frozen in a vivid picture of animal life and death on early farms (Fig 1a) (Lançon 2015). The careful excavation and recording of the bones' positions and an in-depth zooarchaeological analysis of the skeletons (Binois and Lançon 2014, Binois 2015) allowed researchers not only to diagnose the cause of the dystocia, but also to identify human intervention during and after the calving. The position of the calf's head, drawn forwards outside the pelvis, and the perimortem metacarpal fracture of the calf's left forelimb (Fig 1b), hint at a probable attempt at forced extraction, and thin cut marks on the cow's lower forelimbs testify to the skinning of the carcase.

FIG 1:

(a) Skeletons of a cow and a calf that died during a dystocic calving. The scattered remains of another, older calf can be observed on top of the two animals. (b) Illustration of the pelvic area of the cow in (a). The cow skeleton is in beige, the calf skull is in lavender, the calf forelimbs are in blue and the hindlimbs are in aqua. The skeletons were excavated at Téteghem ‘Carlines 3′, northern France

Photograph: V. Devred, INRAP

Paleopathology, archaeology and veterinary science

It is this fleeting glimpse of a Medieval farmer tending, however inadequately, to the delivery of his cow's offspring that shows the importance of this case. Dystocic births are accidental events that have existed independently of human societies for hundreds of thousands of years; obstetric deaths killed long before humans even existed, the oldest known example being perhaps a 248 million-year-old marine reptile fossilised in the course of giving birth (Motani and others 2014). There is, therefore, little to be learned in the fact that malpositions occurred in Medieval calvings as they do today. What is interesting is what this case teaches us of the obstetric knowledge, or lack thereof, of herdsmen in Medieval times, and of the relationships they had with their stock. Such cases help us fill in the almost blank maps of veterinary practices in ancient societies and of past attitudes towards animal health and disease.

About the author

Annelise Binois is a French veterinarian and zooarchaeologist. She graduated from the Ecole Nationale Vétérinaire d'Alfort, France, in 2003 and worked as a mixed-practice vet for seven years in Brittany and Normandy, before turning to archaeology and obtaining an MSc in environmental archaeology at the Université de Paris 1 Panthéon-Sorbonne in 2011.

She is currently a PhD candidate in zooarchaeology at the same institution, focusing her research on the identification and diagnosis of mass mortality events in archaeological livestock, especially in the Medieval period. Her research uses an interdisciplinary combination of archaeological, historical, epidemiological and paleomicrobiological approaches to the diagnosis of past epizootic disease, and aims to further the understanding of ancient animal plagues.

Embedded Image

Embedded Image

Her other research interests include animal paleopathology in general and the study of the complex interactions between humans and domesticated animals in the past.

The study of how past societies dealt with animal disease and with sick or injured animals based on archaeology could be termed veterinary paleopathology. Paleopathology is usually defined as the study of the diseases which can be demonstrated in animal or human remains of ancient times; these diseases, their nature, their antiquity, their origins, are in themselves the object of paleopathological studies. Using similar study materials, veterinary paleopathology focuses on the other hand on the interactions between past human societies and animal disease. The medical knowledge of these societies, the way illnesses were prevented, the treatments and interventions animals were eventually subjected to, the attitudes towards animal suffering and, overall, any archaeological testimony of human consideration of animal health or disease all fall within the scope of this novel field of research.

‘Archaeological animal remains can indeed provide objective evidence of practices undocumented in historical sources and demonstrate that written recommendations were actually put into practice’

Naturally, the way sick or injured animals were cared for is also documented in textual and iconographical sources, and several historians, both early (eg, Moulé 1891-1900) and more recent (eg, Dunlop and Williams 1995, Cam 2007), have focused their research on past veterinary practices. However, historical evidence is relatively uncommon for periods before the early modern era, and even more so when indications relating to species other than the horse – which has always held a singular status – are sought. In many ancient texts, though exceptions exist (such as a few extensive treatises on animal illnesses by antique Roman authors, for example, Columella or Vegetius), relevant mentions are short and ambiguous, and the true availability of the information to the layman and its actual application in everyday life are questionable. For those reasons, little is known of veterinary practices in the average Antique or Medieval household and the situation is even worse when it comes to proto- or prehistoric agricultural societies, for which nothing is known at all.

Archaeological artefacts are unfortunately of little help in this context. Medical instruments are very rare finds on archaeological excavations, and it is often very difficult to ascertain the exact nature and purpose of such objects. Furthermore, even when the surgical nature of an instrument is widely accepted, such as for certain pliers or scalpels, their universal function means they could have served either exclusively for people, exclusively for animals, or for mixed surgical purposes. A few Roman instruments specifically related to horse medicine or to farriery have nonetheless been identified. This is the case for a type of very large pincers of which a few exemplars are known and which have been varyingly interpreted as horse emasculators (Von den Driesch and Peters 2003) or as twitches (Heeren 2009), and for the well described and widely distributed hipposandal, which is increasingly interpreted as an equine therapeutic device meant to protect an injured foot, to keep a poultice on or to apply a pressure bandage on the sole (Dixon and Southern 1997, Von den Driesch and Peters 2003). Evidence is limited to the care of horses though, and no instances of instruments devoted to other species have been described. Similarly, to our knowledge, no other veterinary instruments are known either from earlier or from later periods in Europe. A recent conference specifically dedicated to the subject of veterinary instruments in Antique and Medieval times (La trousse du vétérinaire dans l'Antiquité et au Moyen Âge: instruments, médicaments, pratiques. June 10 to 12, 2014, Lyon, France) is liable to shed further light on the subject, but has yet to be published.

In this underdocumented field of past veterinary practices, paleopathology has a role to play. Archaeological animal remains can, in some cases, provide objective – though limited both in time and geography – evidence of practices undocumented in historical sources and demonstrate that written recommendations were indeed put into practice.

Veterinary paleopathology is an area that has until now seen very little investigation, if any. No single mention of the possibility of identifying therapeutic interventions in animal remains can be found in the seminal work on animal paleopathology by Baker and Brothwell (1980), and the latest book on the subject (Bartosiewicz and Gal 2013) only mentions it in passing among the concluding remarks. The main reason for this is that soft tissue is almost never preserved in archaeological remains, and bones are often the only material available for study. This means that most ancient medical practices are out of reach to us. Whether an animal had been given draughts and medicine, had had a wound stitched up or a tumour removed are questions that archaeologists are unable to answer. And even when specific interventions do affect the skeleton, in most cases, no self-evident proof of human involvement exists and the interventions must be inferred from very subtle indications. Sound knowledge, both of archaeology and of veterinary medicine, are therefore necessary to carry out reliable investigations in this field of research, which requires an extensive cooperation between two professions that rarely find the opportunity to collaborate. Unless researchers studying archaeological faunal material present a specific interest in the topic from the beginning, many discrete hints of past veterinary practices may go by undetected. The hints are there, however, when you search for them.

Routine surgical interventions and preventive medicine

All livestock breeders wish to have a healthy stock, and that aim is achieved both by the management of diseased animals and by keeping healthy animals from getting ill. A minor preventive surgical procedure that is routinely practised in sheep farms nowadays is the docking of female lambs to prevent the soiling of the vulvar area and to protect it from maggots. The earliest mention of tail-docking in sheep dates, as far as we are aware, to the mid-14th century, in a French agricultural treatise, which states that the ‘improvement’ of both male and female lambs should be done at the age of two months, ‘. . . and for the females it is useful to shorten the tail of three fingers’ width, and no more nor no less' (Jean de Brie, edited by Clévenot 1984). The author gives no explanation as to the reasons for this procedure, but several early modern sources do describe sheep being plagued by the presence of maggots under their tails (Fitzherbert 1523, edited by Skeat 1882, Markham 1623). However, pictorial evidence of Medieval sheep most often represents the animals with long, dangling tails, and some authors have argued Jean de Brie's precepts were not actually put into practice in Medieval flocks (Mane and Wilmart 2011).

But, in the same site on which the calving cow was buried, excavations also discovered a pit containing the complete skeletons of six ewes buried together (Binois 2015), which were carbon dated to the 13th century (Fig 2). All six animals had preserved articulated caudal vertebrae, but none had more than five, and no vertebra caudal to the fifth was found in the pit even though part of the sediment was sieved. Considering common European sheep breeds have between 16 and 22 caudal vertebrae, and that the less common Nordic short-tailed breeds still possess between eight and 11, it is safe to conclude the ewes in the pit were docked (Binois and Vallat 2014). The systematic docking of female lambs seems therefore to have been applied in some Medieval sheep flocks at least one century before the earliest written mentions of the practice.

FIG 2:

Thirteenth century burial pit containing the complete skeletons of six ewes excavated at Téteghem ‘Carlines 3’

Photograph: M. Lançon, INRAP

The other minor surgery routinely practised on sheep farms is the castration of male lambs to produce wethers. Contrary to tail amputation, it is an intervention amply documented in texts since the beginning of historical times. It is also one that has proven particularly difficult to evidence archaeologically. One of the rare collections in which the presence of wethers can be reasonably convincingly demonstrated is that from a 17th to 18th century pit from the Paris region containing the subcomplete skeletons of 18 sheep (Achères ‘La Petite Ferme’, INRAP excavation 2004, unpublished) (Fig 3). Fourteen of these animals presented identical dental eruption and wear stages, and could be estimated as having died between 18 and 21 months of age. Within this subsample of 14 individuals, 11 had preserved pelves allowing sex determination, of which five were female, five were male, and one presented ambiguous characteristics. It is known that ewes and intact males present similar ages for epiphyseal fusion and dental eruption, and that early castration delays epiphyseal plate closure, but does not impact dental eruption and wear (Popkin and others 2012). So, if some or all of the males included in the collection were indeed castrated, a discrepancy would be expected in the growth plate fusion stages observed between those individuals and the five females of the sample. We noted the fusion stage of the three growth plates expected to fuse between one and two years of age that were preserved in all 11 skeletons; the results are presented in Table 1. A very consistent discrepancy in fusion stages was observed between all animals identified as males and all those identified as females (Fig 4), which allows us to conclude that these males were castrated and raised as wethers. This is no surprise as most agronomic and husbandry literature from the relevant time period recommends castrating all male lambs, save one or two for reproduction (Hastfer 1756, Carlier 1770). However, our results show that demonstrating castration in archaeological collections is feasible, and that this approach could be applied to more ancient assemblages of similar type, allowing perhaps to answer the question of the antiquity of this practice.


Sex and epiphyseal fusion data for 11 sheep skeletons (Achères ‘La Petite Ferme’, INRAP excavation, 2004)

FIG 3:

Excavation of a 17 to 18th century burial pit containing the skeletons of 18 sheep at Achères ‘La Petite Ferme’

Photograph: INRAP
FIG 4:

Comparison of epiphyseal fusion stages between two sheep of an identical age from Achères. (left) Individual A, female, 18 to 21 months, (right) Individual H, male, 18 to 21 months

‘Bringing assistance to a domestic animal in distress, be it for economic or humane reasons, seems in our modern societies a natural behaviour, and historical texts giving remedies for a diversity of animal ailments indicate this was equally a concern of ancient elites, at least with regards to the more valuable animals'

Therapeutic interventions

The two examples above illustrate the fact that surgical procedures have long been used on livestock both to prevent disease and to control the animal population. It is not the prophylactic side of the trade, however, that springs to mind when thinking about veterinary practice, but the curative treatment of diseased or injured animals. Bringing assistance to a domestic animal in distress, be it for economic or humane reasons, seems in our modern societies to be a natural behaviour, and historical texts giving remedies for a diversity of animal ailments indicate this was equally a concern of ancient elites, at least with regards to the more valuable animals. It is notoriously difficult, however, to evaluate the type and level of assistance offered to diseased stock in lower status, general population households, and very little archaeological data are available on the topic; the difficult calving described previously is an exception.

Most often, unfortunately, therapeutic intervention does not leave any skeletal traces and cannot be identified in archaeological material; either the pathology cannot be detected, or a treated pathology cannot be distinguished from a non-treated one. This is especially the case for fractures. For some time, well aligned, well healed fractures in limb bones for which no adjacent bone could act as a natural splint were considered signs of the immobilisation and treatment of the fracture. However, recently, several examples of such healed fractures in indisputably wild animals have put that hypothesis into doubt, and no archaeological case of therapeutic intervention on an animal fracture can be unquestionably demonstrated (Udrescu and Van Neer 2005). In fact, very few curative surgical procedures are liable to leave identifiable skeletal traces in archaeological remains; one of the rare candidates would be trepanation. This procedure is documented in human remains at least from the Neolithic onwards (Arnott and others 2003), and the technique could well have been applied to animal patients. At least one post-Medieval 16th century text describes a trepanation of the skull as a treatment for coenurosis in sheep (Fitzherbert 1523, edited by Skeat 1882), a disease identified in Medieval texts since the 14th century (Jean de Brie, edited by Clévenot 1984), and sinus trepanation in horses with chronic sinusitis is documented in early modern horsemanship treatises (Von den Driesch and Peters 2003). Both interventions are likely to have been put into practice earlier, and may be identified in archaeological faunal material in the future.

Care, abuse, and mercy killing

Although the actual therapeutic intervention leading to an animal's recovery is most often impossible to identify in skeletal material, the fact that the animal survived life-threatening or seriously debilitating conditions can be in itself indicative of treatment and/or care. Several skeletons of old and sick domestic animals unlikely to have survived without specific attention are known. One example is the case of a small Roman-period ‘lap-dog’ from Yasmina, Tunisia (MacKinnon 2010) that lived into advanced old age with an accumulation of pathological conditions. The skeleton demonstrated antemortem loss of all maxillary teeth and of all but three mandibular teeth, with extensive bone remodelling of alveolar sockets, advanced abscessing and important calculus build-up on the remaining teeth, along with a long-standing luxation of the right hip with the formation of a false joint on the ilium and advanced osteoarthritis of the spine and of articulations of all four limbs. The lesions, especially the oral ones, are so extensive that it appears extremely unlikely that this animal would have survived for so long without specific feeding care in the form of soft, specially processed foods. This case is not an isolated one and other such examples, in particular from the Roman period, have been described (MacKinnon 2010).

However, this must not lead us to conclude that all ancient dogs were treated with tender loving care. Several studies of cohorts of archaeological dog skeletons and of disarticulated dog bones show particularly high prevalences of costal, vertebral and cranial fractures (Warren 2000, Murphy 2005, Teegen 2005, Bartosiewicz 2008), which could be indicative of frequent non-accidental trauma and animal abuse. Abuse can likewise be hypothesised in a young adult dog buried in the Medieval settlement of Guimps in France (Binois and others 2013). This animal presented five independent traumatic lesions that occurred in at least three different events (before the age of 10 months, in adulthood but some time before death, and shortly before death). The lesions included a nasal fracture, the fracture of four adjacent ribs, the fracture of one lumbar vertebra and the deformation of five others, a fracture of the fibula and a traumatic patellar dislocation, and a curvature of the radius subsequent to early trauma of the growth plate (Fig 5). Both the skeletal distribution and the chronological distribution of the trauma are highly consistent with patterns observed in abused dogs (Tong 2014), and animal abuse appears a very probable diagnosis for the accumulation of lesions. It is interesting to note that this animal was carefully buried alone in a disused storage pit and not thrown away on the junk heap as had been other dog carcases on site, indicating it was probably not a stray but a household member that had attracted interest from at least part of the community.

FIG 5:

Skeletal distribution of pathological lesions in the skeleton of a dog excavated at Guimps ‘le Poteau’ (light gray, normal bones; red, pathological bones; white, not recovered) and illustrations of forearm (left) and knee-joint (right) lesions compared to contralateral elements. Reproduced with permission from the International Journal of Paleopathology/Elsevier

These two canine examples illustrate the complex relationships that human communities entertain with their domestic animals, and how veterinary paleopathology can bring these attitudes to light. A further illustration of such attitudes can be found in another mass burial of sheep dated to the later Middle Ages and excavated in Louvres, in the Paris region of France (Gentili 2012). Nine sheep carcases were piled in a pit, probably following a mass mortality event; the uppermost skeleton demonstrated cut marks on the ventral side of the atlas that were not found in any of the other animals of the deposit (Fig 6). Cut marks of this type are commonly observed in archaeological fauna and can be related to two different activities – either the cutting of the animal's throat to put it to death, or the beheading of the carcase as part of the butchering process. As the carcase was deposited in complete articulation, it is clear that the animal had not been butchered or beheaded, and the only explanation remaining is that it was put to death. But, even though this sheep was deliberately killed, its carcase was apparently not exploited in any way. It is, therefore, probable that these cut marks indicate either a Medieval case of mercy killing, with the slaying of a terminally ill or injured sheep to end its suffering, or the pre-emptive culling of a diseased animal to protect the rest of the flock from infection. This last hypothesis is not as anachronic as it may seem, as antique sources already seem to mention preventive culling. Virgil writes in the third book of Georgics, in the 1st century BC, that if a diseased sheep is noticed in the flock, the keeper should ‘stop the mischief straightaway with his knife, before the terrible contagion spreads through the unwary herd’ (Saint-Denis and Pigeaud 1998). An early Medieval text similarly describes the slaying of diseased animals, without mentioning whether this should be done to end the animals' suffering or to prevent infection. Describing ‘an unspeakable pestilence of cattle, more fierce than any enemy, [which] killed the entire species’, the Poeta Saxo, an anonymous Saxon poet, mentions ‘those who, seeing that an animal was about to die, chose to lay it low with a sword’ (Newfield 2010). In any case, the eight other sheep carcases present in the pit demonstrate that the ewe from Louvres was put to death during an ovine mortality crisis, be it an epizootic outbreak or accidental event, and that whatever the reason for the kill, this act was carried out in response to a pathological condition.

FIG 6:

Cut marks on the inferior aspect of the atlas of a ewe excavated at Louvres ‘Bouteiller’

Research potential

The examples provided in this article all demonstrate that veterinary practices and the attitudes of past societies towards animal health and disease can, for some part, be detected in archaeological animal remains. As expected, animal disease was taken into account by ancient communities, prevented if possible and treated when necessary; some animals were provided long-term care allowing them to survive debilitating conditions, others were put down in the face of desperate situations. Less pleasing attitudes, such as incompetence and abuse, can also be identified in certain cases through zooarchaeological study.

Most cases illustrated here do not, however, provide us with much more than anecdotal evidence; it could even be argued that none of the interventions presented is truly novel, in the sense that all are documented in historical sources either pertaining to the period or earlier, and that, even if they were not, they all fall within the range of expected behaviour in ancient agricultural societies. It is true that demonstrating that early modern sheep had been castrated is hardly earth shattering. However, these examples illustrate the fact that the interventions described or advised in historical sources were indeed put into practice in daily life, which is, in itself, a novel result.

The significance of the cases presented here must also be examined in their context, that of a previously completely unexplored research area. They are merely initial observations, for the most part cases encountered by chance by a single researcher, myself, during the three years of a PhD research dedicated to another – mostly unrelated – topic. The number of instances of veterinary paleopathology that I have come across during my research demonstrates the potential of the field for significant discoveries.

This line of research is bound by so many limiting factors and is so underdeveloped it cannot of yet be considered an independent object of study. However, veterinary paleopathology has the potential, when included in an interdisciplinary discussion comprising historians, traditional archaeologists and veterinary scientists, to contribute significantly to the field of veterinary history and to our understanding of past human-animal relationships.


I would like to thank both the archaeologists who trusted me with the animal remains presented in this paper, Mathieu Lançon, Françoise Bostyn, François Gentili and Pierre Rio, and the zooarchaeologists who informed me of the existence of these deposits, Jean-Hervé Yvinec, Ginette Auxiette and Christophe Wardius.

Thanks are also due to my supervisors Anne Bridault and Christophe Petit for their helpful comments an earlier draft.

This research is part of a PhD project funded by the University of Paris 1 Panthéon-Sorbonne.


View Abstract


  • Fig 5 has been reprinted from the International Journal of Paleopathology, 3, Annelise Binois, Christophe Wardius, Pierre Rio, Anne Bridault, Christophe Petit, A dog's life: Multiple trauma and potential abuse in a medieval dog from Guimps (Charente, France), 39-47, Copyright (2015), with permission from Elsevier.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.