Monday, October 24, 2011

More on Lennox-and breeds in general


The latest news from Belfast indicates that Lennox has, for the moment, been given a reprieve with a new appeal.  Lennox, the dog under death sentence in Belfast, Northern Ireland for looking like a prohibited breed, has had me thinking more about the issue of breed more than usual.  And as I petted Parker, my Curly Coated Retriever, while he sat at my feet, it struck me: we are doing the whole thing backwards!

To set up the situation; Lennox is a dog that never showed bad behavior towards anyone-other than barking a bit at the invaders that came into his home, a behavior that is normal and expected from nearly any dog-he simply looks like what Northern Ireland considers a dangerous breed.  To verify this the dog wardens reportedly took out a measuring tape, checked Lennox’s head, body length, and leg length (which would have been nearly impossible if he had been at all aggressive), and pronounced him prohibited, based on comparisons to a written breed standard.  They then seized him and have ordered his death, claiming that his physical attributes, as measured by a tape, have determined that he is dangerous and a threat to society.  Oh, and that moment of barking.

The problem here is that the process of breed standard use is meant to go completely the other way around-including the issue of temperament and behavior.

If I go to a dog show with Parker, the first thing I must have is proof that he actually is a Curly Coated Retriever.  This identification is typically based on registration, with documentation of his parentage back many generations.  This gets us in the door.  Then, he is examined by a judge and that judge compares him to the written breed standard to see if, as a Curly Coated Retriever, he compares favorably (or not) to that breed standard.  He is also compared to the other Curly Coated Retrievers in the ring, who have also been compared to the written standard, to see if he is the best representation of the breed at that show on that day.  He may then be compared to similarly excellent members of other breeds to be judged as to which dog most represents the ideal example of their own breed to determine Best in Show.  All of these dogs are also observed for their behavior in the show setting, as aggressive or dangerous behavior is never permitted, no matter how beautiful or physically adhering to the standard the dog may be.

This particular scenario takes place in the rarefied atmosphere of the dog conformation show, but it directly applies to the real world too.  We meet a dog and we ask his owner what “kind” of dog he is. We ask about his personality, we watch his behavior, and then we use this information to make predictions about the dogs’ nature, suitability as a pet, and honestly whether we think he is a good example of that “kind” of dog.

But with Lennox, and so many others, we are going the wrong way.  We look at the physical form of the dog, apart from behavior and personality, and try and guess which breed the dog most resembles.  Then we make personality and behavior assumptions based on that guess.  In the case of Belfast, they took some measurements and then, without regard to the actual parentage of the dog, and without any regard for the dog’s individual behavior or observed temperament, made a guess that resulted in the seizure of the dog.

First off, this makes huge assumptions that the physical form of the dog (phenotype) must of necessity accurately reflect the breed (genotype), without allowing for gradations between perfect specimens.  This also assumes that physical form defines the dog’s behavior.  Have a largely white dog with black spots and short hair? Must be a Dalmation…unless of course it’s a badly bred English Setter.  Or a pale Catahoula.  Or a Jack Russell/Pointer mix.  You see how well that goes. 

In the real world, we may meet a friend with a new dog.  When we ask what “kind” of dog it is, they may tell us “Oh, this is my new Chihuahua, Peanut.”  We see that Peanut is a bit bigger then we usually find in Chihuahuas, and we may note that, but we don’t run out, get a tape measure and say “Oh no-Peanut is thirteen inches at the shoulders, and Beagles are thirteen inches tall, so Peanut is obviously a Beagle and will run after rabbits!”

This mistaken application of physical attributes to determine behavior, as ridiculous as it sounds, is exactly what is happening in the UK and other places with breed-based laws.  They are attempting to use a dog’s physical attributes to assign projected behavioral traits.  For instance; in this twisted world, a dog that a dressmakers’ tape says has a wide head, broad shoulders, and powerful musculature must be a “Pit Bull” or other forbidden breed.  Yet I doubt, with all due respect and regard, that the Kennel Master at Sandringham Kennels would tolerate a dog warden with a tape declaring that Her Majesty’s Labrador Retrievers were “Pit Bulls”, even though they are broad (and handsome) of head, muscular and fit, and are wide shouldered so they can swim and work efficiently and with grace and style.

Contrast the following parts of breed standards for Staffordshire Bull Terriers and Labrador Retrievers, as published (and copyright to) The Kennel Club of Great Britain:

“Strongly built, short-coupled, very active; broad in skull;”
“Short, deep though with broad skull.”

“Smooth-coated, well balanced, of great strength for his size. Muscular, active and agile.”
“Good-tempered, very agile…broad and deep through chest and ribs; broad and strong over loins and hindquarters.”

“Jaws strong, teeth large, with a perfect, regular and complete scissor bite, i.e. upper teeth closely overlapping lower teeth and set square to the jaws.”
“Jaws and teeth strong with a perfect, regular and complete scissor bite, i.e. upper teeth closely overlapping lower teeth and set square to the jaws.”

Ideal examples of each of these fine breeds will have the above traits.  Examples of dogs of these breeds should, according to the standards, possess the following behavioral temperaments”

“Intelligent, keen and biddable, with a strong will to please. Kindly nature, with no trace of aggression or undue shyness.”
“Highly intelligent and affectionate especially with children.”

Quickly now, which is which?  Which of these potentially may be labeled a “Pit Bull” type dog and banned, and which one is known around the world as one of the finest of the waterfowl retrievers?  Can’t tell from this?  Exactly.  That is my point.

Now let us depart from the company of Her Majesty’s finest and our Best in Show Staffordshire Bull Terrier and go down the breed scale towards home pets and general doggie companions.  These fine specimens of breeds may look clearly separate at the apex of their “type”, but down the genetic lines, even though they may still be clearly Labradors or Staffies, their looks may begin to approach one another; a bit coarser head here, just a bit oversized there, one family having a bit shorter snout than the other…and sooner or later we may have a serious problem distinguishing poor examples of either breed-even though they are related directly over time to the once stellar examples.  Where-and when-do we start drawing the line?  When does a badly bred Labrador become essentially indistinguishable from a badly bred Staffordshire Bull Terrier?

And to get back to our original concern, when does behavior become less Lab-like and more Staffie-like, or vice versa, as their appearances converge?  And more to the point, is there really a predictable difference anyway?

In my time training, working with problem dog behavior, hanging around dog shows, and investigating attacks, I have found Labradors that won’t retrieve, Pointers that hate birds, big brave German Shepherds afraid of the vacuum cleaner, and even French Bulldogs that don’t snore (well, not as loudly…).  Individual differences in dogs are as critical as individual differences in people.  That is why some dogs are a threat, just like some of the people that I used to arrest.  Individual behavior is affected by individual learning history, individual talents and preferences, and ultimately individual choices.

My advice to the officials in the UK, and wherever else breed bans and regulations are being proposed or enforced: Forget what the dog looks like, or is supposed to look like.  Life is not a dog show, and there are good and bad physical specimens of every breed and type out there.  Forget what the ideal good-or ultimate evil-example is supposed to act like.  The individual range across breeds is far greater than the commonalities within breeds, especially when you get away from the ideal example.  Not all Best in Show Labradors can hunt.  Instead, develop Dangerous Dog regulations that regulate and address specific, quantifiable behaviors regardless of appearance.  Bad Old English Sheepdogs should be held to the same standard as bad Anatolian Shepherds.  An evil little Jack Russell Terrier can be just as dangerous as a psychotic Tibetan Spaniel.  Address individual dog behavior, and more importantly, address the functioning of the responsible human.  Now there’s your dangerous breed.

Tuesday, October 11, 2011

"Evaluations" and the tragedy of Lennox

Over the last few months I have watched the case of Lennox, a dog seized for having the “wrong” looks, as it has unfolded in Belfast, Ireland.  Lennox was seized, not for behavior, but because he has a particular physical structure.  He looks like what Ireland terms a ‘restricted breed’-a Pit Bull.  He is neutered, has obedience training, is properly vaccinated and was legally licensed-yet he was summarily seized and has been condemned to die.  As I have watched Lennox’s case, and his impending death sentence, several things have sparked my attention. Not only does the issue of destroying this animal solely based on his looks appall me, but I am particularly concerned by the "evaluation" of Lennox that the Council and Court is depending on to make a determination of his level of threat to society.

To begin, Lennox has been held for over a year in a shelter facility.  He has been deprived of his normal social contacts-his family, has had limited exercise and interaction outside his kennel, and has even been, according to some reports, medicated with amitriptyline.

Two dog behaviorists have evaluated the dog to date.  I understand both have weighed in that Lennox is not a dangerous dog.  The videos and evaluations have shown Lennox to have substantial control of his behavior, that he is a sociable and pleasant animal despite his long isolation and confinement away from his home, and that he showed clear restraint when one evaluator pushed him into a trapped area in a threatening manner.  At that crisis point Lennox did the only thing that makes sense to a dog; he lunged, with no contact, in order to communicate clearly that he was frightened and felt threatened when he had no where else to retreat.  He did the equivalent to a human raising their voice when other means of communication fail.

This speaks volumes for this individual dog.  Despite everything that has happened to him he still shows restraint in his behavior and a desire for human social contact.  He still displays clear bite inhibition.  He still responds appropriately to social cues.  This is also despite the conduct of these evaluations in a restricted shelter environment.

Yet these two evaluations are being ignored by the Court.  Instead, a third evaluation is being used as the litmus test for Lennox.  This third evaluation was conducted not by a behaviorist, but by a police dog handler.  As a retired police Lieutenant I have known a number of canine handlers-and the trainers that prepare the dogs before police get them.  I have participated in the testing and evaluation of police dogs before their training.  And I can say this-police canine handlers and trainers are special, valued and talented persons-but they are not behaviorists.

A police dog is a special animal.  Only about ten percent of the candidates are chosen.  They need terrific drive, huge levels of trainability, and a great desire to work in tandem with a human handler.  They must be brave enough to go in where no person or animal reasonably should, yet must be able to instantly disengage when ordered to, despite inertia and provocation.  They must not be aggressive, as anger would interfere with the ability to disengage at need.  They must also be able to use nearly human levels of discrimination to understand when they must self-deploy to protect their handler, yet must recognize the difference between a violent suspect and the approach of an innocent child.  We ask so much of them-and they give it all willingly, sometimes to the death.

Police dog handlers and trainers must be highly skilled to get this level or performance.  But that skill is limited to the task at hand.  Police handlers do not address behavior problems of other animals-they are focused on the training, maintenance and development of their special charges.  These handlers conduct obedience work with their dogs as part of the control mechanism, but do not diagnose or treat problems that range from house training to nuisance barking.  They do not treat, or particularly evaluate, aggression issues.  If a dog exhibits aggression in training it is eliminated as unsuitable.  An aggressive or "mean" dog is a risk to the Department, the handler, and the public.

Even Animal Control Officers may be deficient when evaluating what is a "dangerous" dog.  They encounter animals that are usually not at their best, often threatened or injured, and frankly do not get the behavioral training necessary to make the decision between treatment of repairable behavior and that which is clearly dangerous.  They can say whether a dog's behavior, in a specific incident, meets the legal definition of "dangerous" in their jurisdiction, but often fall far short of being able to diagnose whether this was truly dangerous aggression or was a storm brought about by a collection of predictable, reasonable animal behavior added to human failing.  In the case of Lennox the dog warden's job was in some ways too easy; did Lennox look like one of the "usual suspects"?  He did, so the case was closed, even though Lennox never had a chance to speak.

Assessing dog aggression, and evaluating whether a dog is "dangerous", even when presented with clear criteria (which do not exist in this case) is a job best left to those familiar with more than just whether a dog is physically able to bite.  Any dog can bite-they have teeth.  A competent evaluator must understand the psychological issues behind the multiple behaviors we lump together as aggression.  Is the dog territorial?  Is the dog a resource guarder? Is the dog fearful? Can the dog adapt to novel and potentially scary situation while maintaining an acceptable level of composure? Is the dog responsive to human signals, and is the dog able to signal its own intentions clearly?  Does the dog have the inter-species social skills needed to peacefully coexist in a multi-species social environment?  Those are the questions that need to be asked before determining if a dog's behavior is "dangerous".

Having a police dog handler evaluate Lennox for his suitability as a patrol or detection dog would be appropriate; it would be having a skilled technician and trainer choosing whether Lennox would make the cut as a working dog.  We would not ask the police trainer to evaluated Fire Department equipment, even though he might like the red suspenders.  To have the police handler evaluating Lennox as a behaviorist is a disservice to the dog-and the handler. 

And the worst part of this?  The case is no longer about Lennox.  It is about rules, it is about discrimination, and finally about egos.  Problem is, the bruised egos will heal-but when Lennox is dead, he is de

Tuesday, October 4, 2011

A Death in India

Recently I was asked to render an opinion on an alleged dog attack fatality in Bangalore, India.  The basis of the case was that a child, a two and a half year old boy named Sandeep, was found dead on the side of a road, mutilated.  His throat showed what appeared to be bite marks, his left leg was amputated, and significant tissue was missing from the leg between the knee and what would have been his hip.  A horrid sight, and a horrid way for any child to die.

The reason I was consulted was that local interests had questions about the investigation, and the probable manner of poor Sandeep’s death.  The police ruled that the death was due to dog attack, but other parties felt that there were issues that should have been investigated more closely.

The dogs in question that inhabit the streets of India are termed Native Indian dogs or INDogs, formerly called pariah dogs.  They are the native feral dog breed that inhabits India, and have done so for centuries.  They range around 25 inches at the shoulder, and weigh an average of 30 to 45 pounds-about the size of a Border Collie.

The parties involved sent me the full file; crime scene photos, the police report, the autopsy report.  And as I sat down to dig through the information and digest the terrible scene, several things struck me as inconsistent.


Since the release of my report other opinions by alleged “experts” have appeared looking as far afield as blaming “imported Pit Bulls”.  Partly in response to these other opinions I have decided to open up the book on the investigation, and my concerns, and let you all have a look at the evidence.

To begin, the story related to the police was as follows:  Sandeep and his family, migrant workers from the Bihar province of India, had traveled for work to the city of Bangalore.  Sandeep’s mother was pregnant and due, and so she was a patient at a hospital in the Bangalore suburbs.  While she was a patient Sandeep, his father and his uncle were allowed to sleep in the secured basement of the hospital, a fairly common arrangement in India where many cannot afford a hotel room while they await the treatment of a relative. 

During the night Sandeep, his father and uncle went to sleep.  When the father awoke before dawn he found Sandeep was missing.  The father told police that he began to search the area for Sandeep, but was unsuccessful.  About two hours latter, after sunrise, Sandeep’s mangled body was discovered 700 meters (for US residents, that is almost half a mile) up a city street, piled at the edge of the road where garbage is commonly abandoned.

The police arrived and photographed the scene.  The child was taken to the Medical Examiner’s Office for autopsy.  A cursory autopsy was conducted and the finding was, “Yep-the dogs did it”.

If you have followed me for any time you already know my position on making broad assumptions without a detailed investigation.  Certainly, a quick look at the crime scene and the paperwork seemed to support the idea that the local roaming INDogs dogs had attacked and killed Sandeep.  This would not be the first case of that happening.  But as I looked at the collected information things started to bother me.  The facts presented just didn’t add up.

Although there were a number of concerns, all listed in my full report, I want to concentrate here on details of the specific injuries present, and how they relate to each other and the case overall.

The first injury was the clearly visible scalping of the child.  In dog attack cases removal of part or all of the victim’s scalp is relatively common.  Partial removal of a scalp has occurred in many cases of children and adults, including the case of Mary Bernal of Florida in 2006.  Dogs that attack the head and face of a victim can easily remove chunks of torn scalp.  A clear example of dog-torn scalp injury can be found in the classic reference work Bitemark Evidence, edited by Robert B.J. Dorion (Marcel Decker Publishing, New York, 2005), pages 316, et al. 

Torn scalp-that is the key here.  Dog attacks that remove chunks of scalp show torn, ragged edges.  In the photos of Sandeep, his scalp is cleanly cut in a straight line longitudinally.  No tears visible, no tooth marks, no ragged edged flesh.  All of those indicators are familiar in dog attacks.  The flesh of the scalp is thin (as anyone who has ever busted their head in a fall knows), but when assaulted by teeth it tears unevenly, not in a clean, straight line.  That is unlike Sandeep, whose scalp wound is clean and straight, more consistent with a sharp object such as a knife-or a vehicle fender. 

But I did not make a conclusion from a single cut.  The next inconsistency was the severance of Sandeep’s left leg.  His leg was detached completely from his body.  Again, limb severance is not unknown in dog attacks, especially in children.  But the leg was detached just below the head of the femur, the large bone in the thigh.  That was visible because the flesh between the top of the thigh and approximately half way to the knee had been cleanly and evenly removed in a neat circle.  There was no visible torn tissue, no ragged bits as one would expect from dogs ripping off flesh.  And an examination of the upper exposed surface of the femur shows an angular, straight break in the bone, similar to the type of mark that a physical tool, such as a sharp metal object would have caused.  This expose bone does not show any clear marks of gnawing or a ragged break. 

This wound is difficult on two grounds.  First, I have seen limbs stripped of muscle and tissue in dog attacks.  Several of them have involved the consumption of the flesh.  Yet these have all shown ragged, non-uniform removal of meat, not the clean, circular pattern shown on Sandeep.  For a simple illustration, look at how your pet dog cleans off a large bone.  The dog takes the easy, removable bits first and then, with time, goes back and eventually picks the bone clean.  The dog does not evenly, progressively, remove each bit in a careful circle and then slowly move on to the next bit.

This is slightly consistent with an injury termed “gloving” where flesh, for instance, on a finger, is pulled off sharply by a dog (a common dog bite) that encircles the finger with his teeth and pulls away, much as you would remove a glove from your hand.  Yet gloving by a dog here would require that the dog encircle the entire top of the leg (after amputation, of course), take it almost knee-deep into the dog’s throat, then cleanly pull directly backwards stripping a clean tube of muscle and skin backwards.  With no dangling bits.  The wound at the hip is also surprisingly clean and even, not what I would expect from extended tearing in the removal of the leg by dogs.

Further, when a dog breaks a bone as large as a femur (thigh) they typically crush and crack the bone in pieces.  Sandeep's thigh bone was cleanly cut.  The location of the break is also unusual-just below the top of the bone, immediately before the joint and the ball end of the femur.  If dogs had simply ripped off the leg the most likely result would have been to tear and shred the flesh around the joint until they could pull the leg free from the socket, including the head of the femur, not to try and break the strongest part of the bone.  This tearing off of the leg would have required a substantial force, and a strong grip on the lower leg, but deep, full dentition gripping injuries are glaringly missing on the visible surfaces of the lower leg, inconsistent with the leg being ripped off by dogs.

The lack of visible blood, both on the exposed bone and the surrounding intact tissue is another concern.  Ripping a limb off results in a lot of blood.  Yet there was minimal blood visible on the skin or the bone, or under the body, an indication that the majority of the bleeding happened other than where the body came to lay.

For other concerns, please read the full report.  To summarize, the injuries, physical evidence, irregularities in the autopsy, positioning of the body on the roadside, and other circumstances bring the simple verdict of “death by dog attack” into question.  As a retired police officer I have seen both deliberate homicides and traffic crashes that could have produced all of the injuries in this case that are inconsistent with canine predation, and that could have easily combined with scavenging to give the scene presented.

My conclusion in this case is that there are a lot of unanswered questions.  None of these issues alone eliminate dogs as the cause of death-nor do they prove it.  They collectively cast doubt on the initial assumption.  I don’t know exactly how Sandeep died.  I am sure that it was violent and that Sandeep deserved a better investigation than he got.  I am certain that there was dog scavenging involved, a behavior fully consistent with observed behavior of native and other dogs in India and other places where sanitary disposal of waste, including hospital waste, is common.  And I am fully convinced that the damage in this case done by dogs does not require the mysterious importation of “Pit Bulls” into India.  I strongly suspect that there was more to this case than “Yep, the dogs did it” as was assumed, probably involving human action that could range from a deliberate attack to a hit and run traffic crash.  But Sandeep’s family will never have the closure of knowing for sure what happened to their son.