Myths and Legends in Animal Behavior: From the Past and Present
�Karen L. Overall, MA, VMD, PhD, DACVB, ABS Certified Applied Animal Behaviorist
What it means to "take a history"
There are two classes of errors that are commonly made when taking a behavioral history: (1) thinking that you knew something going into the history that you did not, and (2) thinking that because you were told something there was information worth having in what you were told. One of the most valuable skills that people who are interested in behavior and behavioral medicine can cultivate is the ability to recognize and acknowledge what they don't know. To do this you have to be willing to let go of labels, to watch the dog / client dyad with the unbiased eyes of the true ethologist or anthropologist, and to listen both to what is said, and what is not said. Concrete examples will make sense of these concepts.
Class 1 errors: The first class of errors - thinking you knew something when you didn't - often involves images conjured by labels. Think of the mental pictures generated by the following labels "rottweiler", "intact male dog", "herding dog", "docile breed", "sharp dog", "attack dog", "Chippendale", "dumb blonde". If you're not cringing, laughing, or saying, "uh huh" by now you're missing the point. All of these descriptions are labels whose repeatability and reliability have not been tested. In other words, I don't know if what I am calling a "herding dog" conjures up the same image that it does for the person whose pet I am treating. They may be telling me that they have a "herding dog" and describing some smallish fuzzy home-made crossbred who is just excellent at rounding up the family's chickens and in my limited mind's eye I see an Australian shepherd grabbing a cow. Unless I let go of or suspend my image of the phrase or label "herding dog" and ask the client what they mean by it, I cannot help them because I'm discussing a label - not the patient. This example seems trivial, but if a miscommunication can occur because of phrase that has no intrinsic value, think of the problems that can occur with some of the more judgmental labels in the string above. I periodically exasperate my friends who show dogs by calling most herding breeds "working" dogs. I know that they are not in the AKC's working dog group, but I am using the words to describe what they do, not their label.
Let's examine another label in the string: rottweiler. I like rottweilers and know a lot of very gentle, smart rotties. When I think of a rottweiler, I think of a large, well muscled dog with excellent hips, and a well defined brown facial mask in an otherwise glossy black coat. My rottie is cocking her head, and her tongue is slightly out. The lawyers who call me weekly see Cujo when they think of rottweilers, and to judge by the ones seen by Penn Vet's orthopedics department, you'd be hard-pressed to blame the residents if they believed rotties had some of the worst joints on the planet.
What about the fairly neutral descriptor: "intact male". To me this just means a dog that still has his testicles. For me it says absolutely nothing about his behavior. But if he cocks his leg and urinates over another dog's urine, people always sort of look away as they shrug and say "intact male". The problem here, of course, is that cocking your leg has more to do with the social situation that it does with testicles, and many dogs - not just intact males - will mark with urine. No one has ever done the study to answer the question: do intact males mark more often and with more urine that any other class of dog? I suspect that once the social context is explored, the noise will swamp any conclusions.
Labels are convenient ways to disguise our preconceived notions and prejudices. When someone tells me that they have seen the "Chippendales" I think of scantily clad - or unclad, but that part is fantasy - tall, muscular, really stacked guys in their 20s, with Madison Avenue chiseled faces. My hunk is always white. I'm white, so I make an assumption that has no basis in fact - really stacked, sexy guys who are Chippendales come in all colors.
What about "dumb blonde"? How did a hair color trait get coupled with intelligence? Here's one case where we all make fun of the stereotype, but in this day and age, such comments - or acting on the thoughts generated by such comments - are litigable.
Simply, what you call something matters. If the words or labels you choose affect your thinking, you'd better choose your words very carefully. Most people don't. That's why people think that breeds are aggressive, neutering makes animals stupid, and if you fall within a certain age group you are restricted to certain classes of problems (Or - conversely regarding age - the client does not have a dog with a true behavioral pathology because the dog is just at that "difficult age" and "letting him know who's boss" will "fix" him. If he is truly pathological, "letting him know whose boss" is likely to be abusive and will result in the client getting bitten. Did she deserve this? No. Did she earn it? Possibly. It all hinges on your definitions).
All we can ever say about "the signalment" (age, breed, sex, reproductive status) is that certain groups may be over- or under-represented in some populations. Facets of the signalment are virtually never the "cause" of behavioral problems, but can be confounding or contributing factors for certain situations. For example, most dogs with a true diagnosis of dominance aggression in the Behavior Clinic at VHUP are male. Being male does not "cause" the problem (there are plenty of lovely male dogs who aren't dominantly aggressive - if the Y chromosome caused dominance aggression, these dogs would have the diagnosis). However, if the dog is intact / entire / has his testicles, he may react more quickly or to a higher level when he does react in the contexts associated with the display of his dominance aggression. Testosterone is a behavioral modulator that facilitates greater reactivity.
Notice that I specified the population of dogs for which I was making the comparison of males being more commonly represented that females within this diagnosis class. The reason for this is that it is important that we back up our perceptions with actual data. I have established criteria by which a dog is decided to have or not to have a diagnosis of dominance aggression, every dog coming to the clinic is screened for the problem in the same way using the same repeatable, reliable, and verifiable history forms, questions, and techniques, and the decision about a diagnosis class is made on the basis of the dog's behavior. Once I have done all of this, and because of the large sample size, I can both count the number of males and females with the diagnosis and ask the question: "Are males more common in the diagnosis class of dominance aggression than females for this Clinic population?" I can then statistically verify that - in fact - males are more common, and it is not just my opinion. If I have to ask you to take it on faith - it's religion. If using the same data set I use you can generate the same answer, it's science. Way too much behavior, behavioral medicine, and dog training are religion.
Class 2 errors: Errors in this logical group involve thinking that because you were told something, that there was information worth having in what you were told. This is HUGE problem in all of medicine, but it is particularly pernicious in behavioral medicine where so many of the signs are "soft". Before we move on to behavior, think of something more tangible, but still largely illusory, like pain. If you have an autoimmune condition or you have ever just felt punky, doctors ask you "Does it hurt here?". They get more (or minimally as much) information from what you do as what you say, and if there is a conflict, they may attribute what you do - your flinching, pupillary dilation, intake of breath, muscle tensing - to pain, and your verbal dismissal as either stoicism or just your standard. Then they have to calibrate your changes with medication or other treatment to your individual response. If you break your leg everyone knows it hurts and if you'll let the doctor physically fix it the pain will subside. This type of pain, while no less illusory in terms of quantification and calibration, somehow seems more real or legitimate than they other because we have a true, non-arbitrary, non-debatable physical marker: the bone is either in pieces or it's not - there's not a lot of room for opinion in the diagnosis. Behavior is to orthopedics as fish are to bicycles.
Now let's talk about a person's description of perceptions they have about their dog's behaviors. By now everyone should see where this discussion is really going. "Fluffy is aggressive." A simple declarative statement, except, I have no idea what it means. Recently, for one client, it meant that the year old dog was - and had been since puppyhood - tethered to a tie-down 23 / 24 hours and tried everything - including grabbing the owner and leading her to the ball - to solicit play during her one hour of freedom. The action with the teeth seemed "aggressive" to the client and her desire to play with the dog decreased as the size of the dog and the desperation with which the dog sought play grew.
There is only one way out of this problem: collect data on the actual behaviors. Watch the dog when you are talking to the client, ask the client to interact with the dog the way they normally do, videotape the dog while you are doing this so that you can review the tape leisurely and learn something (I cannot simultaneously talk to the client, watch the dog, take notes, and make sure no one is being mauled - and I do this for a living), ask the client to videotape the dog at home doing its normal routine, ask about specific behaviors. If you are unsure of something, put the dog in a different situation. If the client tells you that the dog barks when he sees other dogs on the street, ask if it's a "happy", multi-toned, high-pitched, greeting bark, or a sharp, repetitive bark often associated with distress, or a low, decrescendo, menacing bark. If the client says, "I don't know", thank them for their honesty and take the dog for a walk. Don't make it the client's responsibility to be omniscient. Let them know early on that - not only is it okay to not know something - but that it's important for you to know when they are unsure so that you can collect the data. You don't have to go to the client's house to do this, and given that you automatically change the social environment when you do so, you might want to save home visits for situations needing further clarification. You can often do just as well and sometimes better with videotape of ordinary activities.
It's the difference between a judgement or interpretation and an event that is so critical in behavior and behavioral medicine. This is why I feel that people involved in these fields must actually see the dogs. I am willing to do fax consultations with veterinarians, but I require that their clients complete our objective questionnaires and, preferably, send us a video. If they don't send us a video, even with objective questionnaires, I am somewhat limited by the client's perceptions. In such cases I couch my diagnosis in terms that flatly state that I could be totally wrong. It is not an uncommon occurrence for the vet students and I to review the behavioral and medical history in a record, draw tentative diagnostic conclusions, and then totally change our minds when we see the dog. It's good science to do so since in behavior and behavioral medicine the data are the behaviors, and there is no excuse for not collecting the data to the extent possible. This is one reason why I have been so vocal in my opposition to the trend in the veterinary literature to publish reports of behavioral phenomena from researchers who never saw the patients. Part of doing science is to evaluate the quality of your data and ask if they are suitable for answering the question you asked. So when someone tells you that the dog is "shy", "fearful", "aggressive", "confident", "dominant", "submissive", "bored", "angry", et cetera, you should always, always ask yourself if what you were told contained information worth having. You should always confirm your assessment by asking: "Exactly what did he do? What his tail up or down? Where were his ears? Was he making any sounds? What behaviors does he exhibit that makes you think he is "bored" (or "angry" or "shy".....)? If he is in XYZ circumstance, what - exactly - does he do and how often does he do it?" If the dog's person cannot answer these questions, well, you're at an excellent place to start: you know exactly which data they need to go home and collect.
Early intervention and prevention of behavior problems: routine screening for behavioral problems
Most veterinarians obtain a routine history about physical complaints and concerns when they examine any dog or cat regardless of whether that pet is newly adopted or has been a beloved family member for a decade. We are all getting better about screening for geriatric health problems as we learn more about keeping pets happy and healthy for longer times. But the single biggest "health" problem faced by pet dogs and cats is still associated with behavioral pathologies or unmet behavioral expectations. Modern veterinary care should include routine screening about specific behavioral complaints in addition to routine questions that alert veterinarians to potential somatic medical problems. If we can ask "Any vomiting, diarrhea, changes in appetite?", we can also ask "Any inappropriate or undesirable chewing, any growling, any odd behaviors?". If we do this we accomplish the following goals:
(1) We initiate a dialog with the clients about behavior. This lets them know that not only is behavior important, but it is central to good veterinary care. Clients will then feel comfortable asking their veterinarian about behavioral issues. Such dialog represents our best chance for learning of a client's behavioral concerns before these concerns threaten the pet's life.
(2) We establish a baseline of the particular pet's behaviors. Such a behavioral profile will allow us learn "normal" for that pet so that we have a context in which to evaluate behavioral change or client complaints about behavior. This is exactly what we are doing when we recommend routine laboratory evaluation for healthy pets.........if we never knew when the pet last had a creatinine within the reference range it is difficult to know how long the creatinine has been elevated. Length of dysfunction - whether the dysfunction involves a "medical" or a "behavioral" complaint - can affect prognosis, and the extent to which this is true in behavioral medicine is profound.
Basic questionnaires for dogs that can be completed at each visit are found in K.L. Overall, Clinical Behavioral Medicine for Small Animals, Mosby, St. Louis, 1997; revised for 2nd Edition, 2001(working draft) & Overall, K.L. Handbook of Clinical Behavioral Medicine, Mosby, St. Louis, 2001.
Finally, it's important to remember that the clients may not know what normal behavior is, or that they may be uncomfortable with a behavior, but not know how to ask if it is abnormal. These questionnaires will give clients the vocabulary and opportunity to discuss their pet's behaviors with their veterinarian in an efficient, consistent, and meaningful way.
1. Derr M. Dog's Best Friend: Annals of the Dog-Human Relationship. Henry Holt and Company, New York. 1997.
2. Stur I. Genetic aspects of temperament and behavior in dogs. J Sm Anim Pract 1987;28:957-964.
ACVC Contact Information