Is that dog calm or still?

I want you to try a mental exercise:

Think about a small pond somewhere out in the countryside, far away from the city or crowds. Is it "calm" or is it "still?"

Now think about a dog lying on the ground. Is she "calm" or "still?"

The difference between "calm" and "still" when it comes to something like a pond seems rather small, but when it comes to a dog the difference can be night and day. It's possible for an animal to be "still" without being "calm", right?

Researchers in Scotland compared how two behavioral consultants characterized greetings between two dogs with the observations of sixty other people. The "lay" observers were divided into four different categories: non-owners, dog-owners, veterinarians, and dog trainers.

The interactions were a set of videos of eight dogs interacting with a neutral dog. The neutral dog was the same dog for every trial. The observers were asked to describe the initial interaction by scoring eight different adjectives on a scale of 0 to 6. The adjectives were provided without an exact definition and were chosen to encompass a wide range of behavioral styles. They were then asked to chose one of the adjectives to characterize the behavior of the subject dog, and then finally asked to explain what it was about the dogs' body language that made them select that adjective.

Their conclusion, in part, was: "The results suggested that working and living with dogs, without any theoretical knowledge of canine behaviour, may not be determinant for allowing correct description of the majority of dog behaviours."

I don't want to delve too deep into the methodology, it had limitations and they are ultimately up front about it. But I do have a bit of an issue with the idea that their two experts were always "correct" and that any differing interpretations were therefore always incorrect. I would have liked to have at least seen more than two "experts" before they decided what the "correct" interpretation or at least more focus on the difference between observations instead of grading for right or wrong.

What I found very striking was the wide variances in interpretation of the interactions and the body language. For example, they highlight that submission and defensiveness were "properly" identified by about half of the observers. Casting aside "proper" and "improper," this means that 50% of the group disagreed on what submission and defensiveness look like! 

The only significant differences when comparing the "accuracy" of the four categories of observers was their ability to recognize "confidence" and "play solicitation:" non-owners had significantly more trouble recognizing them. Interestingly, across of the human subjects, age and sex of the observers had no significance - only the four categories of dog experience.

A common mistake made by observers was focusing too much on one specific aspect of the dogs' body postures. Piloerection ("raised hackles") was commonly misinterpreted as aggression rather than simply arousal. I'm pretty familiar with this mistake, but to see it have an impact in some areas of a formal study is pretty startling. Observers also missed eye contact (or the lack thereof) almost completely, while they seemed to rely on tail position and wagging most often.

But here is the money quote: "In our study, the four categories of observers were similar in their descriptions of dog behaviour. Nevertheless, due to the fact that the adjectives were not behaviourally defined, the results could not distinguish between 'labelling' and 'interpreting' behaviour."

If you are not using a common grammar, you can't agree on much. What does "confidence" mean when you are watching two dogs greet for the first time? I'm still not sure what the researchers meant by it. They say it was "incorrectly" identified as friendliness 43% of the time, but even they admit that "This result might simply indicate a different use of words by observers."

Without clear definitions you can't tell the difference between an observation and an interpretation. It is also difficult to accurately tell the difference between them without some training, and this may be the real difference between the veterinarians, the trainers, and the "experts" in the study.

"Still" is a lack of movement. It's what all three of my dogs are right now, since it is 2:00AM as I write this and they are asleep. It's probably safe to call them calm now too. (Neither of the border collies are sleep-running at the moment.)

Is a dog that is "still" after being restrained by a person or a piece of equipment also necessarily "calm?" No. How can you honestly tell if a dog is calm? You need to observe a variety of other factors, such as (and not limited to) breathing and heart rate and then you can try to accurately interpret the situation. (This might be difficult to do if the dog is on TV. Just sayin'.)

This was an interesting study, not because of the conclusions, but because of the questions it raises and ultimately fails to answer. But I'll wrap up with one more quote:

The limited reliability of dog-owners in describing some behaviours may also create misunderstandings in the context of behavioural consultations.

If you are a trainer that handles behavior cases you are (hopefully) thinking something along the lines of: "Duh!" right now. If you are a dog owner, this might help you understand why we ask so many questions along the lines of "What did your dog look like?" or "Why do you say she was calm just before she bit the mailman?"

Given a hammer, everything looks like a nail. We all have our own biases, and those biases are driven by our experiences, both recent and not-so-recent. It's good to keep them in mind while observing a situation and when it comes time to make a judgement on what is happening.

Based on the level of disagreement between different people reading dogs, I think it's safe to say that as a species, we're not very good at it. How good are dogs at reading us? Check out my on for information on how dogs read our emotions so well.