Messenger Dogs: Soldiers in the Great War

An Australian officer saw one for the first time on the Aisne in June 1918:

“He was coming from the front-line trenches—a little Welsh terrier. The ground was in a terrible condition and absolutely waterlogged. The little creature was running for all he was worth, hopping, jumping, plunging, all with the most obvious concentration of purpose. I could not imagine what he was doing until he came near and I saw the bulging collar. I had heard about the message collars. As the dog sped past I could not help but notice the terribly earnest expression on his face.”

Captain von Stephanitz said that dogs could cover five kilometers in a quarter of an hour, and described one dog near Peronne that lost a hind leg while carrying a message but completed the assignment on three legs before dropping dead.  

World War I

Electronics have largely made messenger dogs obsolete, so it takes some imagination to realize the importance of this assignment for war dogs in prior times. During the First World War, the messenger dog was the most elite canine servant along every front, and those armies that used them significantly limited their human casualties.

Although using dogs as message carriers was known from antiquity, Edwin H. Richardson exaggerates only slightly in saying that messenger dogs were particularly a product of the Great War, and Richardson devotes almost half of the nearly 300 pages of his book, British War Dogs, to describing the training and use of messenger dogs by the English, French, and German armies.

The typical messenger dog was taught to return to his handler, whom Richardson calls a “keeper.” Training was based on what Richardson called the dog’s homing instinct. If it were desired to have a message delivered from a forward station to battalion headquarters in the rear, the handler and the dog would go to battalion headquarters, then the dog would be taken to a station near the front by another soldier unfamiliar to the dog. When the message for headquarters was ready, it would be put in the dog’s collar and the dog released to return to his keeper at headquarters.

That was the most common type of messenger dog. A liaison dog, which took about twice as long to train, recognized two handlers, and could be directed by either to return to the other. These dogs could be sent back and forth continually between the front to battalion headquarters. Although Richardson trained some liaison dogs, he noted that this “more difficult system” entailed “a considerably greater wastage of life, both among the men themselves and the dogs, as the position of the keeper in the front line is fraught with risk, and the dogs are also required to run a double journey over the danger area.” The Germans, having more dogs trained for messenger work, and having devoted more time to training them going back to the period before the war, made greater use of liaison dogs.

Edwin H. Richardson

Lieutenant Colonel Edwin H. Richardson was not a shy man and portrays himself as fighting for years against institutional ignorance about the value of war dogs. Even before the war, he states that “I was the sole person who took any interest in trained dogs for the army and police, and the outflow from my kennels constituted the only source of supply.“

When World War I began, Richardson tried to persuade the powers that be to let him train sentry dogs, already known to be in use by the Germans. His idea having been rejected, he began training ambulance dogs, but this came to little because the Germans did not respect “the inviolability of the sacred symbol of the Red Cross, whether on man or beast, hospital or ship.” Also, the nature of the Great War was that it involved little movement, often only glacial change in the positions of the front lines, which did not suit ambulance dogs that could be used primarily in finding the wounded after a battlefield was controlled. Richardson concedes that the “only ambulance dogs that were used with any success were those with the German army when the Russians were retreating on the Eastern front.”

British Messenger Dogs

Although the Germans had been using messenger dogs since the opening of the war, it was not until 1916 that Richardson was asked if he could train dogs to carry messages between an “outpost and the battery, during heavy bombardment, when telephones are rendered useless, and the risk to runners is enormous.”

In his book about dogs in World War I, Richardson expresses his enthusiasm for the idea of using dogs as messengers, but in his earlier book, War, Police and Watch Dogs, published in 1910, he had been far more skeptical:

"There is a system of canine service used to a certain extent in the German Army, but personally I am not greatly in favour of it. The idea is to train the dogs to run to and fro with dispatches carried in a small case on the neck; but my experience tends to show that too much is expected of the dog, and although dogs are found that are sufficiently intelligent to discriminate direction under very difficult circumstancdes, still the result is too uncertain, and the teaching partakes too much of the trick-training to be of practical value."

Despite such earlier reservations, Richardson began teaching two Airedales, his favorite breed, Wolf and Prince, teaching them to carry messages homeward for two miles. Sent to the Royal Field Artillery, 56th Brigade, the commanding officer soon sent an exuberant report describing how the dogs had run from the front during an attack to brigade headquarters, “travelling a distance as the crow flies of 4,000 yards over ground they had never seen before and over an exceptionally difficult terrain.” The dogs, released two hours apart, each took under an hour to carry their messages over the two and a half miles of the battlefield.

More dogs were immediately requisitioned. Richardson and his wife opened a dog school at Shoeburyness and began training men and dogs in courses lasting five weeks, after which they were sent to war zones. Richardson goes into considerable detail about the training system he developed. He found that the selection of the handler (keeper) was even more important than the dog. Prior experience was not necessarily useful because the training of messenger dogs was “so different from every other kind of dog work that practically anything that a man has learned before about dogs has to be forgotten before he is qualified to be trained himself, and to train others.”

Richardson considered patience an essential quality in a trainer, and dismissed any trainer or handler who was incapable of working gently with the dogs:

“Coercion is of no avail, for of what use would this be when the dog is two or three miles away from its keeper? In fact, it may be said that the whole training is based on appeal. To this end the dog is gently taught to associate everything pleasant with its working hours. Under no circumstances whatever must it be roughly handled or roughly spoken to. If it makes a mistake, or is slack in its work when being trained, it is never chastised, but is merely shown how to do it over again. If any of the men under instruction are observed to display roughness or lack of sympathy with the dogs, they should be instantly dismissed, as a promising young dog could easily be thrown back in his training, or even spoiled altogether, by sharp handling.”

Distances and Distractions

Dogs began by being separated from the handler for short distances and for short periods of time, which were built up until a dog might not be released for 12 hours to cover a distance of three and a half to four miles. More than 12 hours was not considered advisable to separate a dog from his handler. Dogs were familiarized with trenches so that they could remain in them for extended periods, as shown in the second picture.

Dogs were first taught to return to one location, then to different locations:

“First of all, they are trained from a fixed base, and then from a movable base. It was always interesting to observe how the different dogs thought out the return journey, when under training. Sometimes they would be taken out by a roundabout route by road, and certain dogs would always return by the shortest way, which sometimes took them straight across country. Others, on the other hand, quite as clever, chose to take the longer route back on which they had been brought outwards, because they thought they could run quicker on the smooth road unmolested by obstacles of any sort. I noticed that the most cunning and elderly dogs generally did this. In the field, it was much the same. The most experienced dogs generally took what they judged was the easiest route homewards, even though it might be somewhat longer.”

Dogs had to be taught to cross quite varied terrain:

“They must be taught to travel along high roads, amongst lorry and other traffic, through villages, and past every sort of camp and cook-house temptation. They must be taught not to be afraid of water, or of any inequalities in the ground. To aid the dogs in overcoming all these difficulties, all sorts of artificial obstacles are introduced into the route of the dog's journey over and above those he would meet in the ordinary way. Barbed wire entanglements, palings, fences, water dykes, smoke clouds, made by harmless means, etc., should intercept its homeward journey, and it must be induced at all costs, one way or another, to surmount these difficulties by going over, through, or under. It is left to the dog to choose, but come he must.”

Specifically as to water, Richardson says that the dog has to cross a stream “either by jumping or swimming, or wading.” One picture shows a dog training to jump over a canal. Another shows messengers jumping over barbed wire.

Gunfire and Artillery Fire

Battlefields are noisy places, and a dog had to be capable of working with noise and even while under fire (as is still required of American military dogs). Richardson describes how the tolerance to gunfire is built up:

“The first training each day is the firing drill. The entire parade of dogs, excluding the new arrivals, are led to a large shed, where a certain number of keepers are drawn up with rifles loaded with blank cartridges. Several rounds are fired, and many of the haughty spirits that have been making such a display on the parade ground are now inclined to put up a sorry show. Much gentleness is, however, extended at this lesson, and any dog that shows timidity is taken further off until it gets accustomed to the firing. This they very soon do, and the old hands proudly stand right under the firing keepers. After this there is a system of bomb-firing, which is a further call upon nerve force, and has often to be carried on for some time. The dogs are also trained to run among the keepers who are firing their rifles from a recumbent position across the road by which the dogs have to come. They are also taken to the batteries and accustomed gradually at varying distances to the sound of 18-pounders up to the 12-in. guns.”

Richardson says that it “is a good plan to feed the dog with tit-bits during the firing. It is remarkable how soon most dogs get accustomed to the heaviest firing.” To teach dogs to deal with a smoke barrage, bundles of straw or hay are set afire and the dogs must be able to run through it.

Theo F. Jager, writing on the training of war dogs for World War I, mentions the importance of training stealth, which distinguishes the training of war dogs from that of police dogs, particularly messenger dogs:

"While the police dog is always ready to announce a culprit through loud and incessant barking, the army dog should remain silent. He must be taught that barking is a nuisance. His future duties as a scout or message bearer require that he travel silently and screen his whereabouts. In the open field work he should be taught to avoid the higher land, to go over open spaces on a run and to advance just as a soldier would, always well covered. If the trainer will keep this forever in mind and avoid to lead or even exercise with the more advanced dog in exposed areas, the dog will soon realize, that to gain his approval, he must hunt up the lower or protected and covered spots. The wide open fields that sometimes have to be crossed, should always be covered with trainer and dog at a run. The dog will do this as a matter of habit later on when the plain is fireswept and he carries a message of importance."

Pictures here show dogs learning to pass a barricade during rifle fire, and a messenger dog running past an explosion in a field.

People and Other Dogs as Distractions

One of the most serious obstacles could be a village and the other dogs in it. Richardson describes the draconian measures that were taken to protect the messenger dogs from such interference:

“The villages present the greatest temptation to the dogs on account of the ash-heaps, foodshops and also the allurement of pleasant chats with local canine busybodies, who thoroughly delight in holding up a messenger dog, which may be conscientiously endeavouring to do its duty. The difficulty of the village dog at the front was one, which had to be taken seriously into consideration, and it would be better in future to face this situation in a more practical manner. In France, there were such large numbers of stray dogs in the devastated areas, that their presence was sometimes a serious menace to the successful working of the messenger dogs. In order to remedy this state of affairs, large numbers of these dogs were ordered to be destroyed by the G.O.C. in the various districts. This order was, from many points of view, considered a stern necessity at the time, but was a pity, as if it had been realized at first, how valuable the services of dogs would become to the Army, and how unequal would be the available supply to the demand, every one of these dogs would have been of use in some form or other, either to the British or the French Army.”

Richardson considered that situations such as this made some level of shyness useful in a messenger dog:

“There are some dogs which, while they are very plucky in every other direction, have a strain of timidity in their natures in relation to strange people, and these dogs often avoid villages if they can, and put themselves to great trouble to go round them instead. This disposition in the messenger dog is of great value, as it is, therefore, saved much temptation.”

Richardson credits his wife with having kept him from rejecting a particularly timid dog name Tweed that eventually proved a particularly effective messenger dog. She is shown with some of the dogs she trained amid the kennels the dogs lived in.

Livestock, however, would seldom be encountered near the front lines, and Richardson took this into account during training:

“An important question arises when selecting a suitable training site, in connection with live stock. No time must be wasted by breaking the dogs to sheep, cattle, etc., for the obvious reason, that they do not have temptations of this sort on the battlefield, all herds and flocks having been cleared by the fleeing inhabitants or by the enemy. Therefore, the immediate vicinity of the training ground must not be too much of a pasturage ; thus the young dog, when it commences its training, is not distracted by any temptation to chase.”

One possible distraction was found not to be a factor. Dogs almost never crossed the lines and went into enemy-controlled territory. Richardson attributes this to the fact that neither the land nor the people would smell familiar. Jager, writing as World War I raged, however, insists that some dogs did cross lines:

"On several occasions the German messenger dogs have run into the French trenches. At first the French soldiers did not realize they were carrying messages, and they were allowed to pass unmolested. Since the real use was discovered the dogs are now immediately shot."

Specialized Gear

Messenger dogs were given a leather collar on which were their registered numbers and a message carrier made of two parts that were held together by a piece of string.

“The message carrier (which is always to be kept on the collar) is made in two parts which fit one into the other ; the two parts are held together by a piece of string. When a message is to be placed in the carrier, the dog's collar should be pulled round so that the carrier is uppermost, the two halves of the carrier pulled apart, the message inserted, the two halves pushed well home and tied together and the collar slipped round so that the carrier is under the dog's neck. The dog should then be released without his chain [leash].”

The last picture below shows a message being put on a dog’s collar at the front.

The German Experience

Richardson cannot keep himself from the occasional reference to the Huns, but credits the Germans with foresight concerning war dogs, seeing “the large organization of police dogs” as “a camoulflaged system of service, always available in the event of war.” The breed clubs in Germany were, he admits, more advanced in their thinking than the clubs of England:

“The Germans have, however, always made a feature of training dogs in connection with their clubs. In our own country, we have our gun dog and sheep dog trials, but this particular aspect of the education of dogs is carried on in Germany on a much larger scale. During the Herrero War—the German West African Colony—sixty trained dogs were sent out with the troops from Germany, and many experiments were made with them. These tests must have been considered to be of a promising nature, as after that time, the whole question of dog training in Germany was taken up much more seriously, and under Royal and official patronage.

“All these aforementioned dog clubs were now amalgamated, and formed one huge Association, which called itself ‘Der Verein für Deutsche Schaferhunde.’ The Crown Prince was president, and the affairs of the organization were directed by military officers. Branches were inaugurated all over Germany, and in Austria as well, and shows were constantly held to display the training standards attained. The association kept and issued annually a list of dogs under training, and against each entry was inscribed notes relating to that particular animal's capabilities. This book was open for all to purchase, who wished to do so, but the fact was that the whole association was a war organization, and each of the branches was ordered to keep a secret register of those dogs considered especially suitable for war purposes.”

Thus, when the war broke out, dogs already in the army were sent straight to the front, and dogs on the lists of clubs were mobilized immediately. Almost 6,000 dogs were deployed in the first months of the war, with orders directly from General Erich Ludendorff. Training establishments were opened in various towns, with qualified trainers already available, who served under officers who did not need to learn the value of the dogs, unlike the situation with the English army. Dogs captured in the advance into Belgium and France were sent back to Germany to be trained for war. German owners of suitable dogs could be paid up to 50 marks for a dog. As messenger dogs, Germans preferred German shepherds, Dobermans, Airedales, and Rottweilers.

The basic training of the German messenger dog was that of the liaison dog, unlike the British program which only used a small portion of liaison dogs. As did Richardson in training British messenger dogs, the Germans focused on teaching dogs to deal with a large number of distractions. German dogs were to be kept busy: “A dog must run its message route two or three times a day. This must be done for practice, even when there is no occasion for sending information.”

Captain von Stephanitz describes the German use of a relay system:

"In cases where the distances were too long for one dog, stages were arranged, between which dogs ran to and fro.  If any of these despatch stages had to be moved, the dogs sought out the new place by tracing out the track of the leader or the deputy leader.  When a despatch stage was organised in a permanent way, the dogs would run over the accustomed route by sight only; they made short cuts and regular 'passes'. This system fo dog posts worked so well and with such certainty that, Regiments outside the fighting area availed themselves of its possibilities to save despatch riders and cyclists."

The Germans were also concerned with interference with messenger dogs and made intentional troubling of dogs by troops a punishable offense.

Precise rules specified where dogs were to be kept awaiting work:

"Immediately the dog reaches the attendant, or his assistant, its dispatch collar should be taken off and not put on again until the moment that the animal is sent back. The putting on of the collar will thus be a sign to the dog that its journey is beginning. Should the attendant have to take the message brought by the dog a little further (e.g., from the telephone exchange to battle headquarters), the dog should not be taken with him, but tied up and left behind at the terminal point of the route, in order that this point may be retained by the dog as a fixed datum for its return.”

Every German army headquarters on the Western Front had a Messenger Dog Section (Meldehundstaffeln). Such sections had to serve army groups on the Eastern and South-eastern fronts. An infantry regiment was to have a maximum of 12 dogs allotted to it, and an independent battalion, six. Richardson praised this wide distribution of messenger dogs, and stated that the British system of confining messenger dogs to the signal service (Royal Engineers) was a mistake. The French also distributed dogs to infantry units.

When not working, according to Captain von Stephanitz, the dogs were often use to capture the rats in the trenches, assisting the snappers and terriers that were supplied for rat catching. They were also used to carry ammunition, rations, and carrier pigeons, and sometimes to unroll cables and wire.

Americans

The Americans did not join the war until it was half over, and viewed the British canine units with envy. One sergeant reported that “I had great trouble to prevent the Americans from appropriating the dogs, as they had no dogs of their own, and they all spoke of the great success of them everywhere."

Heroism

Richardson was not at the front but includes a great many dispatches he obtained from the War Office. He recounts how dogs could continue to work after being wounded:

“A black retriever dog called 'Dick' had a wonderful record.... While carrying a message in the Villers-Bretonneux sector he was wounded very severely in the back and shoulder. The dog completed his run in good spirit, and was ultimately sent to the section kennel for treatment from the veterinary officer. As no foreign bodies could be located in the wounds they were stitched up, and he was soon healed up and at work again. He did his runs in the line as well as before, and seemed all right when we moved to the 8th Corps. A day or two after moving the dog was seen to be suffering, and the attention of the vet. was called to his state. After a few days' observation, the veterinary officer concluded there was some foreign body in the wound, and so, as poor Dick was on the point of death, he was ordered to be destroyed. At the post-mortem examination it was discovered that a rifle bullet was resting between the shoulder and body, while near the small of the back a piece of shrapnel was found lodged close to the spine. Through all his sufferings the dog carried out his duties cheerfully and most faithfully until he was overtaken by death."

Another handler describes how hard the dogs worked:

“After we had got the kennels fixed up, orders came for 8 men and 24 dogs to go up the line. I was amongst them and next morning went to Gentilles Wood, where I was attached to the 47th London Div. The dogs did splendid work there, and were working day and night. I lost one dog there, an old collie—he had carried 5 messages that day and was on his way back with his 6th one, when he was killed by shrapnel.”

Arranging a Temporary Truce

One dog was credited with bringing about a temporary truce. A wire haired terrier was seen crossing the no-man’s land between British and German troupes, coming to the British side across shell holes, water, wire, and the noise of bombardment. The British soldiers were so impressed that they fed the dog, only to see him return across the no-man’s land to the German side. This began to be a regular occasion.

One day, when the dog was on the British side, a British soldier was wounded in the no-man’s land. The commander had an idea. After feeding the terrier, he attached a message to the dog’s collar: “Will you allow us to bring our man in?” In a few minutes the dog returned with an answer: “Will give you five minutes.” Two soldiers went out and returned with their wounded comrade. As they came back, the British gave a cheer to the Germans by way of thanks.

Conclusion

Although Lemish reports in his book on war dogs that the Americans used some messenger dogs in the Pacific Theater in World War II, Richardson is no doubt correct that World War I was the war of the messenger dog. The Americans apparently did not pick up the value of positive reinforcement in training war dogs, as demonstrated in their failure to get dogs to work effectively as mine detectors in the Second World War. The British, continuing Richardson’s emphasis on gentle and positive training, were much more successful.

The rigorous training standards of the breed clubs of Germany do much to explain the development of the war dog training programs of Germany in the First World War. One of these clubs, the German Shepherd Club of Germany was also responsible for the idea of using dogs as guides for the blind after the war.

Sources: E.H. Richardson (1920). British War Dogs: Their Training and Psychology. Skeffington & Son, Ltd. London; (1910) War, Police and Watch Dogs. William Blackwood & Sons. Edinburgh and London; M.G. Lemish (1996). War Dogs: Canines in Combat (1996). Brassey’s Washington DC (describing the French use of messenger dogs in WWI); A Biography of E.H. Richardson, website posting of K-9 History: The Dogs of War; A. Sloan and A. Farquhar (1925). Dog and Man: The Story of a Friendship. Hutchinson & Co. London;; T.F. Jager (1917). Scout, Red Cross and Army Dogs (Arrow Printing Co. New York) (at 23 discussing the risks of messenger dogs); v. Stephanitz (1923). The German Shepherd Dog in Word and Picture. Anton Kampfe, Jena, Germany. 

Although not strictly a source, I should acknowledge that I would not have thought of this title had I not long ago read Mark Helprin's brilliant novel, A Soldier of the Great War (1991).

The opening quotation is adapted with considerable freedom, but hopefully not with any change of critical fact, from a passage in Richardson’s book.

Additional Note. Ingrid Bahlenberg informs me that modern messenger dogs are still being trained in Scandinavia, and that messenger dog trials are a canine sport in Finland and Sweden. Dogs compete in relays between different stations that are up to two kilometers apart.

Did Kings Kissing Greyhounds Give Us Hepatitis C? Perhaps, but It Doesn’t Mean Service and Therapy Dogs Should Be Kept Out of Hospitals

When the popular media reported that hepatitis C may have originally come from dogs, I got several emails asking if this was going to have an effect on whether therapy dogs should be admitted to hospital settings. I read the scientific paper, and tracked down some related research regarding viruses, and am happy to say that this kind of interspecies transfer should not affect our ability to take dogs into hospitals. Since this is something I do on a regular basis (at Benedictine Hospital in Kingston, NY), I am personally relieved as well.

At least that is the simple answer. The transfer of the virus from dogs to humans, if that is what happened, probably occurred only once, at least 500 years ago, after which the virus that made the transfer, probably itself mutated from the standard virus in dogs, established a foothold and remained a human problem. There are viruses, however, such as a parvovirus that infects both cats and dogs, that can transfer more often between species, and there is no guarantee that this could not happen with other viruses that affect dogs and humans. Nevertheless, such events will remain rare, and the discovery of the strong similarity between human and canine hepatitis C should not alter hospital practices regarding the admission of therapy and service dogs.

Canine and Human Hepatitis. The hepatitis C virus (HCV) was discovered in 1989 (Alter, Choo), and it is now known that 200 million people are chronically infected with this virus, putting them at risk for liver fibrosis, cirrhosis, and hepatocellular carcinoma. Efforts to find homologs of HCV in non-human primates have so far been unsuccessful.

While looking at respiratory viruses of domestic dogs, a team of scientists from American and English universities and research facilities (Kapoor et al.) found a virus genetically similar to HCV, which they have tentatively named canine hepacivirus (CHV). The virus belongs in a group of probably four genera of viruses in a family of viruses called the Flaviviridae. The virus is found in respiratory samples, as well as in livers, of domestic dogs.

The team’s results “indicate that hepaciviruses are not restricted to primates and suggest the possibility that HCV may have been introduced in the human population through contact with canines or other nonprimate species.” As with hepatitis C in humans, an analysis of the chemical structure of CHV predicts that infections from it “may be persistent in its natural hosts.” The disease patterns in humans and dogs are different, however. CHV-infected dogs, for instance, have high levels of CHV in respiratory samples, but HCV-infected humans do not. The researchers note that a “significant difference in life span of humans and canines can also affect the disease pattern caused by genetically related viruses.”

The divergence of the human C virus and the canine virus is about the same as between a virus found in humans (GBV-C) and one found in chimpanzees and New World primates (GBV-A), a virus not known to cause disease in humans but which may affect the immune systems of patients with HIV.

A statistical calculation of the time to most recent ancestor (TMRCA) led the researchers to “estimate that the shared common ancestor between CHV and HCV genotypes probably existed between 500 and 1,000 ybp [years before the present].” Reaching this estimate requires some assumptions, and the researchers caution that there are complexities in determining rates of divergence. If the time frame is correct, the common ancestor could have passed from one species to the other during the Middle Ages, a time when kings were kissing greyhounds and paupers slept in kennels.

The researchers note that other non-human primates should be tested for HCV-like viruses, which has not so far been done “because of a primate focus in screening paradigms.” It is possible, however, that “hepaciviruses are primarily canine viruses and HCV in humans arose zoonotically from contact from dogs or other related members of carnivore mammalian order that harbor these types of viruses.” The researchers note that such a zoonotic origin “would explain its high degree of pathogenicity in humans” and the “apparent absence of HCV homolog in nonhuman primates.”

In the stages described by Wolfe, Dunavan, and Diamond by which a pathogen transforms from exclusively infecting animals into a pathogen exclusively infecting humans, HCV has apparently reached the fifth stage, only transmitting between humans.

Canine Parvovirus. A much more recent species jump involves the parvovirus, which moved from cats to dogs (perhaps from a single cat to a single dog) in the late 1960s or early 1970s. Parrish and Kawaoka note that animal viruses may be generalists, infecting a variety of hosts, or specialists, infecting only one or a few host species. Various changes are required for a virus to become a successful epidemic pathogen in a new host. The transfer to a new host likely starts with only a single virus.

The parvovirus found in cats (FPV) since the early part of the twentieth century has been recognized since the 1940s as a natural pathogen of minks, foxes, and raccoons. Canine parvovirus (CPV) was first observed in May 1978, when two new diseases were reported: (1) myocarditis leading to sudden death in neonatal pups, and (2) entiritis accompanied by diarrhea in dogs older than two months. The diseases spread to wild and domestic dogs throughout the world. It remains an endemic pathogen with dogs everywhere.

The FPV and CPV viruses are more than 99% identical in DNA sequence. Serological studies showed that CPV had a low prevalence in European dogs between 1974 and 1976, but spread worldwide in 1978. CPV-like viruses were all derived from the introduction of a single common ancestor (a single mutant virus) from cats into dogs in the late 1960s or early 1970s. A variation of CPV (CPV-2a) occurred in 1979 and replaced the earlier strain (CPV-2) throughout the world.

Dog Flu. Another crossing of species lines has been the transmission of equine influenza (EIV) to dogs (CIV), first recognized as a pathogen of dogs in 2004 when found in greyhounds in Florida, though serological analysis suggested an initial transfer around 2000. CIV is now common in certain regions of the U.S. A recent study (Hoelzer et al. 2010) found that many mutations occurred in infected dogs, but the researchers also argued that “mutations that facilitate adaptation to a new host species might occur transiently in the donor host … and provide a transient reservoir of pre-adapted mutations.” This flu virus seems to constantly throw off mutations, with the chance that some of them will survive and adapt, even in a new species.

Conclusion. Most viruses that infect wildlife and domestic animals do not infect humans, though sustained contact between species increases the likelihood that a virus will adapt and jump across the species barrier. Research in this area is in its infancy, and a good deal more will be learned in the coming years. Many viruses cause few or no symptoms in long-time hosts, and it is sometimes difficult to say where a virus jumped a species barrier. The more animals that are investigated, however, the more zoonotic researchers will be asked if their findings should limit human-animal contact, such as with patients who have diminished immune responses. This is not an argument, however, for keeping dogs from humans, any more than the parvovirus transfer is an argument for keeping cats from dogs, or the flu transfer from keeping horses from dogs. Of course, a sick dog should not be taken to visit patients in any case.

When I take Chloe to a hospital, it remains true that I am much more likely than she is to bring an infection in with me.

Sources: A. Kapoor, P. Simmonds, G. Gerold, N. Qaisar, K. Jain, J.A. Henriquez, C. Firth, D. Hirschberg, C.M.Rice, S. Shields, and W.I. Lipkin (2011). Characterization of a Canine Homolog of Hepatitis C. Virus. Proceedings of the National Academy of Sciences. Approved April 28, 2011. H.J. Alter (1989.) Discovery of the non-A, non-B hepatitis virus: The end of the beginning or the beginning of the end. Transfusion Medicine Reviews, 3, 77–81; Q.L. Choo, G. Kuo, A.J. Weiner, L.R. Overby, D.W. Bradley, and M. Houghton (1989). Isolation of a cDNA clone derived from a blood-borne non-A, non-B viral hepatitis genome. Science, 244, 359–362; C.R. Parrish and Y. Kawaoka (2005). The Origins of New Pandemic Viruses: The Acquisition of New Host Ranges by Canine Parvovirus and Influenza A Viruses. Annual Review of Microbiology, 59, 553-586; C.R. Parrish, C.F. Aquadro, M.L. Strassheim, J.F. Evermann, J.-Y. Sgro, and H.O. Mohammed (1991) Rapid Antigenic-Type Replacement and DNA Sequence Evolution of Canine Parvovirus, Journal of Virology, 65(12), 6544-6552 (description of mutation in CPV becoming prevalent after 1986); S.-F. Chang, J.-Y. Sgro, and C.R. Parrish (1992). Multiple Amino Acids in the Capsid Structure of Canine Parvovirus Coordinately Determine the Canine Host Range and Specific Antigenic and Hemagglutination Properties. Journal of Virology, 66(12), 6858-6867; C.R. Parrish, E.C. Holmes, D.M. Morens, E.-C. Park, D.S. Burke, C.H. Calisher, C.A. Lauglin, L.J. Saif, and LP. Daszak. Cross-Species Virus Transmission and the Emergence of New Epidemic Diseases. Microbiology and Molecular Biology Reviews, 72(3), 457-470; K. Hoelzer, P.R. Murcia, G.J. Baillie, J.L.N. Wood, S.M. Metzger, M. Osterrieder, E.J. Dobovi, E.C. Holmes, and C.R. Parrish (2010). Intrahost Evolutionary Dynamics of Canine Influenza Virus in Naïve and Partially Immune Dogs. Journal of Virology, 84(10), 5329-5335; N.D. Wolfe, C.P. Dunavan, and J. Diamond, Origins of Major Human Infectious Diseases. Nature, 447, 29-93.

Thanks to Richard Hawkins and Joann Lindenmayer for bringing additional sources to my attention. My apologies to them and other scientists if I have failed to detect some of the more subtle nuances of the papers referred to. Although at Berkeley in the 1960s I took Chem 8 from Melvin Calvin and attended Gunther Stent's Friday afternoon mol bi seminars, some of the organic chemistry was a bit beyond my reach, and still is.

Legislators Focus on Autism Service Dogs, Service Animal Trainers, and Service Dogs for Veterans

Some recent legislative efforts will, if enacted, make changes to service animal law, particularly with regard to autism service dogs, service dog trainers, veterans using service animals, and emergency rescues where pets and service animals need to be taken care of. Most of these efforts are federal, but a significant statutory change is expected to be signed into law in Alabama. Other proposals would require or authorize funding so their chances in the current climate are not great. Nevertheless, it is time to take stock of what our elected representatives are doing, or attempting to do, for skilled dogs these days.

Alabama House Bill 502 (revising Code 21-7-4). The Bill, which is awaiting the Governor’s signature, would revise Alabama’s statute regarding service animals by defining a service animal as “any dog that is individually trained to do work or perform tasks for the benefit of an individual with a disability.” The revision provides that “a person who is totally or partially blind, hearing-impaired, or diagnosed on the autism spectrum shall have the right to be accompanied by a service animal in any public place, including a public or private school, and [other places of public accommodation].”

Further: “A person training a service animal shall be entitled to the same privileges granted to a person with a disability….” The places where a person with a disability can take a dog include not only places of public accommodation and transportation facilities, but also lodging.

Then comes the most interesting part of the new law:

“In the case of a disabled child, including a child diagnosed on the autism spectrum, any aide assigned to assist the child shall be trained with the service animal in basic commands in order to assist the child as a team.”

This is a significant recognition that a school that admits an autism service dog will not be able to avoid the situation by saying that the child and the dog can come but the family must provide the handler. This is a common sense recognition that if the aide is supposed to work with the child, the aide must learn to work with a trained animal that helps the child. This is a correct result, but in my opinion, and as I discussed in a prior blog, by no means certain under federal law.

H.R. 1878. To require that the same access to transportation and public accommodations that is afforded to individuals with disabilities who use service animals under the Americans with Disabilities Act be afforded to certified trainers of service animals. Representative Larry Kissell of North Carolina (Democrat) introduced this Bill to provide access to transportation and public accommodations to certified trainers of service animals. The text of the legislation says that the same access is to be provided to a “licensed or certified trainer of a service animal, or a handler of such animal that has credentials issued by an accredited school for training service animals,” as is provided to individuals with disabilities under titles II and III of the Americans with Disabilities Act.

The Secretary of Transportation and the Attorney General are given the responsibility of issuing regulations to carry out the purposes of the section. Since those agencies have not always agreed on rules applicable to service animals, presumably it is not expected that they would have to do so here.

Though intended for the Departments of Justice and Transportation, the proposed legislation does not adopt a definition from either agency, but rather says that a service animal is “a guide dog, signal dog, or other animal individually trained or being trained to provide assistance to an individual with a disability and distinguished by wearing a collar, leash, harness, or cape that identifies the animal as a service animal in training, regardless of whether such animal has been licensed or certified by a State or local government.” Thus, it appears that the animal does not need to be licensed or certified, even though the trainer does.

The legislation has a laudable objective, one not squarely faced by federal regulators, in that it is desirable that trainers should be able to take dogs in training into environments where the animals will ultimately be expected to work. North Carolina, where Kissell is from, specifies that an “animal in training to become a service animal may be taken into [places of public accommodation and transportation] for the purpose of training when the animal is accompanied by a person who is training the service animal and the animal wears a collar and leash, harness, or cape that identifies the animal as a service animal in training….”(North Carolina Disabilities Code 168-4.2) Thus, it appears that Kissell has taken the North Carolina statute and hoped to federalize it.

While the Department of Justice requires that a service animal be under the handler's control, generally with "a harness, leash, or other tether," these items are not identifying, but rather a requirement for access. (28 CFR 36.302(4)) The Department of Transportation allows that harnesses and tags can help verify that an animal is, in fact, a service animal, but such gear is not required. (See, e.g., 49 CFR 39.91(d))

The Department of Justice has “never imposed any type of formal training requirements or certification process.” (73 Fed. Reg. 24516) Most states, on the other hand, provide access rights to trainers, though many laws require that trainers be employees of recognized service dog training facilities and further require that trainers be prepared to produce documentation on their connection with the school, as well as proof that the dog is being trained to be a service animal. Since there is no such federal requirement, H.R. 1878 is not going to get very far before someone points out to the Congressman’s staff that the Bill, as drafted, would require a complete rethinking of the federal regulatory approach to service animals. It would be relatively easy to rewrite the proposal to fit within the current federal system and it is to be hoped that someone will do so. The Bill has only garnered four sponsors so far, and will probably die in the House Subcommittee on Highways and Transit.

For a detailed discussion on access rights of trainers and temporary handlers, see Service and Therapy Dogs in American Society, Chapter 15.

HR 198. Veterans Dog Training Therapy Act. Representative Michael Grimm (Republican, New York) introduced this Bill to assess “the effectiveness of addressing post-deployment mental health and post-traumatic stress disorder symptoms through a therapeutic medium of training service dogs for veterans with disabilities.” The pilot program is to be carried out at three to five VA medical centers during a five-year period. Each program site is to have a certified service dog training instructor, with preference being given to “veterans who have successfully graduated from post-traumatic stress disorder or other residential treatment programs and who have received adequate certification in service dog training.” Selecting dogs from shelters is to be considered an option, and dogs are to live “at the pilot program site or a volunteer foster home in the vicinity of such site while receiving training.” Data is to be collected from participating veterans to determine if the program is helpful in:

(1) reducing stigma associated with post-traumatic stress disorder or other post-deployment mental health condition;
(2) improving emotional regulation;
(3) improving patience;
(4) instilling or re-establishing a sense of purpose;
(5) providing an opportunity to help fellow veterans;
(6) reintegrating into the community;
(7) exposing the dog to new environments and in doing so, helping the veteran reduce social isolation and withdrawal;
(8) building relationship skills, including parenting skills;
(9) relaxing the hyper-vigilant survival state;
(10) improving sleep patterns; and
(11) enabling veterans to decrease the use of pain medication.

Further, reports to Congress are to assess the effects of participating in the program on:

(A) symptoms of post-traumatic stress disorder and post-deployment adjustment difficulties, including depression, maintenance of sobriety, suicidal ideations, and homelessness;
(B) potentially relevant physiological markers that possibly relate to the interactions with the service dogs;
(C) family dynamics;
(D) insomnia and pain management; and
(E) overall well-being.

It would seem that some professor in a public health program would be interested in researching these matters, and could assign the data gathering to eager graduate students and senior undergrads. Whether federal funding is really necessary is open to question.

The Bill gathered 71 sponsors by mid-June 2011. It has been referred to the House Committee on Veterans’ Affairs, which referred it to the Subcommittee on Health on February 18. It does not seem to have much traction at the moment.

S.769. VETS Dogs Act of 2011 (see also HR 1154). Senator Tom Harkin (Democrat, Iowa) introduced this Bill amending 38 U.S.C. 1714 to add a new subsection (e), which would provide that the “Secretary [of Veterans Affairs] may not prohibit the use of service dogs provided under subsection (c) in any facility or on any property of the Department or in any facility or on any property that receives funding from the Secretary.”

The previously existing subsection (c) provides that service dogs may be trained for hearing-impaired veterans, for persons with “spinal cord injury or dysfunction or other chronic impairment that substantially limits mobility to veterans with such injury, dysfunction, or impairment...,” and for “persons with mental illnesses, including post-traumatic stress disorder….” Patients granted service animal access must be enrolled in the VA system, which may meant that visitors with service animals will still have a problem. Under 38 U.S.C. 1714(d), the VA may pay for the veteran’s travel expenses to the training facility for the dog and “for expenses incurred in becoming adjusted to the dog.”

Enactment would effectively modify an existing regulation, 38 CFR 1.218, which provides that that dogs "and other animals, except seeing-eye dogs, shall not be brought upon [VA] property except as authorized by the head of the facility or designee." Some VA hospitals have apparently been restricting access to non-guide dog service animals, though most have not.

Hearings were held on the Bill in the Senate Committee on Veterans’ Affairs. The House version of the proposal, introduced by Rep. John Carter of Texas, has 58 sponsors and was referred by the Veterans’ Affairs Committee to the Subcommittee on Health. Since the bill should come with no price tag or a very small one, it may have a shot at enactment.

HR 57. Disaster Recovery Improvement Act. This proposal, introduced by Representative Steve Scalise of Louisiana (Republican) in January, would amend the Robert T. Stafford Disaster Relief and Emergency Assistance Act (42 U.S.C. 5192(a)) to state that, in an emergency, the President may assist state and local governments by providing assistance for “rescue, care, shelter, and essential needs (A) to individuals with household pets and service animals; and (B) to such pets and animals.” Thus, the President will have the ability to forward emergency funds for such purposes.

The proposal was referred to the House Subcommittee on Economic Development, Public Buildings and Emergency Management, where it will probably die, given that there are no cosponsors. The Republicans are apparently worried that President Obama is too much of a dog lover. The Democrats too.

Thanks to Kristina Chew for a blog that alerted me to the Alabama legislation. Thanks to Joan Esnayra for comments on several pieces of legislation, and to Patty Dobbs Gross for describing autism service dog issues that are developing in school systems.

War Dogs in Ancient Military Strategy

Dogs accompanied the armies of antiquity. If you ever get to Thessalonica, look at the top panel of the Arch of Galerius. On the left is a dog looking up at the horse and rider above it. Perhaps no more than a camp follower, but perhaps a dog that guarded the camp of his master once the army stopped for the night. In an Attic Black Figure vase from 540 to 520 BC, a dog stands amidst the warriors departing. Saying goodbye? Just as likely going with them. (Louvre F19) Then as now, dogs can be taught to take risks. Above is the side of a vase made about 500 BC showing a dog running beside, almost under, the horses pulling a racing chariot. (Tampa 86.35)

What did they do if they went to war?

Military functions of dogs have their origins in the uses dogs have had since the beginning of domestication. Aeneas, known as the Tactician, writing in the fourth century BC, describes using dogs for a range of purposes. On nights when a city is under siege, he recommends tethering them outside the wall (άριστον δ'εν τοιαυταις νυξιν έξω του τειχεος κυνας προσδεδέσθαι νυκτερευοντας) so they can detect spies approaching and deserters trying to flee, and so that their barking will wake any guards who have fallen asleep. (Aeneas 22.14) The guard dog below in a mosaic from Herculaneum that is tied to a hoop high on the wall has generally been assumed to be guarding a house, on the belief that the Vesuvians and Herculaneans were only interested in themselves, but the military symbols might argue that the dog was outside the city or perhaps guarding a camp.

Polyaenus, describing a siege in 385 BC, tells how dogs stopped traitors from aiding the enemy:

“While Agesipolis was besieging Mantineia, the Lacedaemonians were joined by their allies, who were sympathetic towards the Mantineians, but were obliged to help the Lacedaemonians because they were at that time the leading power in Greece. Agesipolis was informed that the allies were secretly supplying the defenders with whatever they might need. To prevent this happening in future, he let loose a number of dogs around the camp, and particularly around the part which faced towards the city. This stopped the communications with the defenders; because no one ventured to cross between the camp and the city, for fear of being discovered by the barking of the dogs.” (Polyaenus, Stratagema, 2.25)

Dogs could accompany human guards on their rounds outside a city wall, as Aeneas says they did before the Battle of Naxos in 376 BC. (Aeneas 22.20; see also Vegetius, IV.26) During defensive maneuvers inside, Aeneas advises keeping the dogs chained up (δεσμειν) lest they attack and harass the troops. (Aeneas 38.2-3)

Dogs might have to be silenced for certain maneuvers. Aeneas says that for making secret sorties against an enemy encamped outside the walls, the dogs would have to be kept from barking and the cocks from crowing. He says they are to be rendered temporarily mute by cauterizing some part of their mouths (επικυσαντα τι του σωματος). (Aeneas 23.1-2) Hopefully the guard dogs would be brought inside and locked up to minimize the need for this. Something of the reverse of this situation was a usage Darius made of dogs to deceive the Scythians. Darius left his camp, but tied up the dogs so that their barking and baying convinced the Scythians he was still in the camp. (Sextus Julius Frontinus, Strategemata, 1.5.25)

The numbers of dogs used for guard work were not small. Plutarch, in his Life of Aratos (24.1), says that, in 243 BC, the Achaeans guarded Acrocorinthos with 400 soldiers and 50 dogs and as many handlers (πεντηκοντα κυσι και κυνηγοις ισοις). Dogs on the column of Marcus Aurelius probably reflect guard dogs that were brought by the Roman Army in the Danubian campaign. The dogs are shown in the drawing, by Pietro Bellori from a publication of 1704, coming ashore from ships during the invasion.

Soldiers might be accompanied by dogs in battle. Aelian in his Miscellaneous History (XIV.46) states regarding a battle from the eighth century BC:

“The Magnesians, who border the Mæander, warred against the Ephesians and every member of the cavalry brought a dog and a slave. When the enemy came near, the dogs fell upon them first, then the slaves with bows and arrows, and finally the masters.”

This sort of coordinated use of dogs to attack an enemy in a group suggests some level of training, and specifically of attack training.

A painted chest in the tomb of Tutankhamun shows two dogs in battle below the pharoah's chariot, evidencing the use of dogs in battle before 1300 BC. In the section of the chest shown above the dogs bite the heads of their victims. They look more like hunting dogs than war dogs, and it is possible that the artist has put the greyhound-like animals in an environment where they would never have been. The second dog looks out from the painting directly at the viewer, allowing us to meet his gaze for eternity. Perhaps the pharoah's own dogs were immortalized by an artist anxious to please his benefactor.

In a passage in his History of Animals (VII.38), Aelian describes dogs soldiering with their masters (συνεστρατεύοντο), mentioning one that fought with his master at Marathon and was portrayed on the Painted Colonnade in Athens with his master. The poet Valerius Flaccus describes the Caspians streaming forth, their pack of dogs dashing no less swiftly at the trumpet’s blare, joining their masters in battle. (Argonautica VI.107-113) Pliny describes the king of the Garamantes as being escorted back from exile by 200 dogs that did battle for him. (Natural History, VIII, 61) Herodotus says that Xerxes brought Indian hounds for his invasion of Greece (VII.187)

Strabo refers to the Celts using dogs bred for the chase in war as well. (IV.5.2) This likely means they were used in battle. The Alexander Sarcophagus in the Istanbul Archaeology Museum shows a dog under a horse at full gallop, the dog oblivious to the danger of running between the horse's legs as it breaks to the right.

Pliny says the Colophonii and the Castabalenses had faithful cohorts of dogs that fought in the front rank and never refused battle. (Natural History, VIII, 61) Polyaenus writes that when Alyattes, king of Lydia, fought the Cimmerians in the sixth century B.C., he brought fierce dogs that fell upon the barbarians and injured many of them (Stratagema 7.2; see also 7.9, using dogs on the front line defensively). The great frieze of the Temple of Zeus at Pergamum shows a battle against the giants in which dogs aided the gods. The Molossian in the section shown below might be mistaken for a lion were one not to look closely at the animal’s tail.

Pliny describes a dog that would not leave his master’s side after he died in battle, driving away birds and beasts of prey. Another starved to death in mourning, and yet another threw himself on his master’s funeral pyre. Though not in a war, Pliny tells of a dog that, when thrown food, took it to the mouth of his dead master, and when his master’s body was thrown into the Tiber, the dog jumped in the rushing water and tried to keep him afloat, something a great crowd watched. (Natural History, VIII.61)

Dogs were used as messengers in antiquity, and were used for this as recently as World War I, electronics having now relieved them of this function. Aeneas says that messenger dogs were used in Epirus, ancient home of the Molossian. He describes how this was done:

“In Epeiros, dogs were widely used as follows. A dog would be taken away from home on leash and a collar with a letter sewn inside put around its neck (περιεθηκαν περι τον αυχενα ιμαντα, εν ώ επιστολη ενερραπτο); then it would be released, at night or during the day, and would inevitably return to its owner. The Thessalians did this too.” (Aeneas 31.31-2)

Polyaenus describes Philip of Macedon using hunting dogs (θηρευτικας κύνας) to pursue barbarians fleeing into the Balkan mountains, finding most of them. (Stratagema, iv.2.16) This is a very early instance of using dogs to track humans rather than game. Ionnes Zonaras, a Byzantine chronicler, describes using dogs from Italy to track men and cattle in 231 BC (κύνας εκ της Ιταλίας μετεπεμψατο ευρινας, και δι’ εκείνων την στίβον και των ανθρωπων και των βοσκηματων ευρων πολλα απετέμετο), something bloodhounds were doing on England's border with Scotland in the time of King James. (VIII.18) E.S. Foster, in referring to the passage by Polyaenus, describes the dogs as bloodhounds, but hunting dogs is the most that can be made of the Greek. In neither of these instances is it certain that the dogs were following a specific track. Both passages seem more like area searches, as when modern police dogs look for a suspect who has fled a crime scene, going on the attack once they have found someone hiding.

Forster says that the only function ancient war dogs did not perform was Red Cross work.This is less true now. Although Forster was writing decades after Otto Kalischer discovered that dogs could be trained to recognize specific chemicals, he was writing well before this was put to practical use to create bomb, narcotics, and arson dogs. Still, much of what they do for us now they were doing for us then, and often paying the ultimate price for this service.

Sources: J.M.C. Toynbee (1973). Animals in Roman Life and Art. Cornell U. Press. Ithaca; D. Whitehead (translator and commentator) (1990). Aeneias the Tactician. Oxford U. Press, Oxford; M. Lemish (1996). War Dogs: Canines in Combat. Brassey’s London; S. Menache (1998). Dogs and Human Beings: A Story of Friendship. Society & Animals 6(1). E.S. Forster (1941). Dogs in Ancient Warfare, Greece & Rome, 10(3), 114-7; G.B.A. Fletcher (1941). Word on Dogs in Ancient Warfare, Greece & Rome 11(31), 34; Aeneas Tacticus (Loeb Classical Library 1923); Corpus Scriptorum Historiae Byzantinae: Ioannis Zonarae, Annales (Bonnae Impensis ed. Weberi 1844).

For a description of Kalischer's work and the evolution of scent detection, see Police and Military Dogs (Taylor & Francis/CRC Press, forthcoming October 2011). Thanks to Richard Hawkins, Brian Duggan, L.E. Papet, and John Grubbs for many helpful suggestions. Thanks to Michael Kranzler for citing this blog (and for quoting Polyaenus) in his article in the University of Miami National Security & Armed Conflict Law Review, 4, 268: Don't Let Slip the Dogs of War: An Argument for Reclassifying Military Working Dogs as "Canine Members of the Armed Forces."

Cueing and Probable Cause

An article on handler cueing of bomb and drug dogs that appeared in the scientific journal, Animal Cognition, has caused quite a stir in the world of law enforcement canine handlers, as well as among criminal lawyers. The article, Handler Beliefs Affect Scent Detection Dog Outcomes, was written by three researchers at the University of California at Davis, Lisa Lit, Julie B. Schweitzer, and Anita M. Oberbauer. L.E. Papet and I have written a paper discussing this research, as well as the law that has developed on cueing issues in criminal prosecutions. The article, Cueing and Probable Cause, is posted on the website of the Michigan State University College of Law.

Also, see our article, How to Prevent Cueing Arguments from Getting Canine Evidence Thrown Out in Court, which appeared in Deputy and Court Officer, 3(2),36-38 (2011).

The Good Life of the Medieval Hunting Dog

Dogs of the nobles fond of hunting in the late Middle Ages and early Renaissance had a fairly comfortable life.

The Oxford English Dictionary derives "kennel" first from the Old North French, kenil, becoming French, chenil, perhaps from a combination of the Latin for dog, canis, and a suffix meaning a fold, a place of keeping. The earliest usage identified by the OED is from a lexicon written about 1440 (Promptorium parvulorum sive clericorum). The lexicon gives canicularium as a Latin equivalent, a word that did not make Lewis & Short's Latin Dictionary, unless as a genitive plural of canicularis (of or pertaining to the Dog Star). Perhaps the term was only a theoretical construct of the lexicon. D.B. Hull concludes that there were kennels at least by Byzantine times, citing a passage in the Geoponica compiled in 10th century AD Constantinople. This work does not give a word for the house of the dogs, but refers to puppies scratching at their doors, presumably the doors of kennels.

By late medieval times, the kennel could be rather ornate. The miniature above is from Livre de Chasse, by Gaston Phoebus. John Cummins, in his wonderful work on medieval hunting, says that the kennel was about 60 by 30 feet, with the hounds living on the ground floor on beds of oak. The miniature shows the ground floor open all around, but Cummins suspects that this is artistic license so that we can see the pages scattering the straw bedding.

Edward, Second Duke of York, who died at Agincourt on October 25, 1415, translated, with considerable freedom, Phoebus into English as The Master of Game, saying the kennel is to be cleaned every morning, with fresh water put before the hounds twice a day. A thick layer of straw should be put down each day. The straw should be on boards a foot above the ground because lying on dirt could make dogs sick. A translation into modern English from 1909 renders York's description of the kennel as follows:

"The hounds' kennel should be ten fathoms in length and five in breadth, if there be many hounds. And there should be one door in front and one behind, and a fair green, where the sun shineth all day from morning till eve, and that green should be closed about with a paling or with a wall of earth or of stone of the same length and breadth as the hounds' kennel is. And the hinder door of the kennel should always be open so that the hounds may go out to play when they like, for it is a great liking to the hounds when they may go in and out at their pleasure, for the mange comes to them later. In the kennel should be pitched small stones wrapped about with straw of the hounds' litter, unto the number of six stones, that the hounds might piss against them. Also a kennel should have a gutter or two whereby all the piss of the hounds and all the other water may run out that none remains in the kennel. The kennel should also be in a low house, and not in a solere (an upper chamber), but there should be a loft above, so that it might be warmer in winter and cooler in summer, and always by night and by day I would that some child lie or be in the kennel with the hounds to keep them from fighting. Also in the kennel should be a chimney to warm the hounds when they are cold or when they are wet with rain or from passing and swimming over rivers."

Cummins writes that "pages had to spend their days and nights with the hounds, partly to keep them from fighting." Sometimes poor people were allowed to sleep with the hounds as a kind of charity. It would have kept them warm but they might have had to put up with more than a few fleas. He refers to the straw being changed completely once every three days.

Note that there are no greyhounds in the kennel. Greyhounds were favored above all other breeds and had the run of the castles. This may also explain how the lines of greyhounds were kept pure by the nobility. The second plate is from Tres Riches Heures of Jean, Duke of Berry (1340-1416), now in the Musee Conde. The Duke is the man in blue with the brown hat. The wedding scene in the third plate, from the Histoire de Renaud de Montauban, dates from about 1470 and shows a greyhound, or conceivably a Saluki, since part of the Renaud romance involves a crusade. Other dogs might be kept by hunting staff in their quarters in the castle, but would probably not have had the same freedom as greyhounds. For pictures of collars worn by greyhounds from late medieval times onward, see the website of the Leeds Castle Dog Collar Museum.

Greyhounds may have been separated from other dogs not just for their comfort but to keep the lines pure. Phoebus gives a recipe for aborting an unwanted litter. A sort of canine chastity belt was known even in the ancient world (Anderson, p. 46).

A decoration in the margin of a medieval manuscript (Bodleian 264, dated 1338-1344) shows dogs being fed from baskets. Since bread crumbs and grains were standard in canine diets then (as now), this would be an appropriate vessel. 

Sources: W.A. and F. Baillie-Grohman (eds.) Edward, Second Duke of York, The Master of Game (Chatto & Windus, London 1909); J. Cummins, The Art of Medieval Hunting: The Hound and the Hawk (Weidenfeld & Nicolson, 1988); J.K. Anderson, Hunting in the Ancient World (University of California Press 1984); D.B. Hull, Hounds and Hunting in Ancient Greece (University of Chicago Press 1964). 

Thanks to Brian Duggan and Richard Hawkins for keeping me focused on the unique status of sighthounds through much of history.