Guest post about dog training and science,
By Julie Nutter, dog blogger, Feb 2011
Introduction
The bloggospere’s wittiest dog writer, Julie Nutter, kindly accepted to write a guest post on Canis bonus.
Boy did she deliver! Read on to discover Julie’s fresh and thought-provoking style.ert line here
insert line here
No Science About It
And by “it,” I mean dog training.
If I’ve managed to irritate the living daylights out of you with that statement, I’ve succeeded in one third of my goals for this article – the other two being to intrigue and to delight you.
(I … might have to work at those two.)
With that, let me introduce you to the grand world of scientific dog training.
It’s one part science, one part dog training, two parts nominal fallacy.
“Scientific Dog Training” should really be titled “Theoretical Dog Training.”
Why?
Well, here’s the challenge of the day: Come up with an irrefutable argument that proves that dog training works.
Despite the fact that we all know dog training and behavior modification works, and despite the fact that we’ve all seen it work – have all worked it’s magic – there isn’t a single one of us who can actually prove it.
Irritated yet?
Most people, I’ve come to realize, probably have a hard time not hitting me when they’re brave enough to get into a science-based discussion with me. I can prove to you – and anyone listening – that you cannot prove anything.
Especially that dog training works.
Here’s the short answer: You cannot devise a way to come up with a controlled experiment.
In fact, whenever the people who do experiments with dogs refer to “the dogs”
…
They are NOT referring to YOUR dog or to MY dog; they are referring to the dogs who were part of their experiment.
Don’t confuse that.
-
The dogs they worked with may have had one reaction, but your experience may be different.
-
The few dogs they worked with are not – and have no right to be – the speakers for all dogs everywhere.
If a scientist picked a group of “random” humans and proceeded to do behavioral tests with them, and came up with the conclusion that “the humans” were addicted to alcohol and cigarettes, and therefore all humans drank and smoked, would you be offended?
Does this describe you and everyone you know?
Let’s have a little fun and set up an experiment.
This experiment, of course, has to be objective and set up according to the guidelines science provides us:
– There must be only one variable, and
– The rest of the experiment must be controlled.
Sounds easy, right?
Pick a group of dogs and prove that not only can they learn, but that it’s the training that gets them from Point A to Point B.
Here’s where I get to giggle.
This. Isn’t. Possible.
Why not?
1.) As a human, it’s impossible to be objective. By our very nature, we are judgmental, assuming creatures; it’s part of our charm.
For the proposed experiment, this simply means that whatever we observe will be clouded and interpreted based on the observer’s unique perception on the world…. Which means that the observations and conclusion recorded cannot be objective. Period.
Fun note: If you see the word “objective” anywhere, what any scientist really means (but won’t tell you) is that they’re being as close as they can be, will the full understanding that they can’t be truly objective.
We need this closeness to objectivity to keep the sanctity of the experiment. Without that, we’re just throwing stones with our eyes closed and proclaiming to the world that we know where they’ll land.
2.) The experiment is further tainted by the way it is set up. There’s a sort of experimental bias going on here.
So, let’s say we run this little experiment of ours.
We need dogs and we need people to train them.
– Well, what is the age, the breed, and the gender of the dogs?
– What part of the country or world would they hail from?
– Are they from breeders, shelters, or a combination of the two?
– What kind of dogs get in? Are some fearful or aggressive?
– Are the handlers trainers? What methods do they use?
– Where would the experiment be done?
– Do the dogs have prior training?
I could list criterion for days and get nowhere, and without answers to these questions, we can’t set up a proper experiment.
With answers to these questions, we’ve just nullified our own results. As you go through and answer those questions, you’re introducing a bias to an experiment that needs to be pure. Each answer you provide taints the experiment, and thus taints the results.
*Note here that if your answer is that certain things should be “random” know now that there is no such thing. Also know that if you say “random” for everything, you don’t even have an experiment because science dictates that there can be only one variable.
*Second note, and very wise point, is that according to my logic there isn’t an existing experiment that hasn’t been tainted by the observer and by the way it is set up.
Welcome to dog training….there are thousands of variables.
And that’s kind of the point.
Say you have a puppy with a potty training problem, and your trainer tells you to put the dog on a potty schedule. Voilà! The dog is now house-trained.
There’s a sort of pseudo-paradox in this that fascinates me.
Here’s the deal: No one can ever prove that putting the dog on a schedule is what cured the dog’s potty problem. There are too many variables in the dog’s life – perhaps:
– a bladder issue was cleared up by the body,
– the dog started to grasp the concept of impulse control at the water bowl,
– a week or two was all the dog needed to gain a little more control over his bladder.
Point is: There is no way to tell.
Where does this pseudo-paradox of mine come in?
Well, by that same logic, it can never be proven that it wasn’t the potty schedule that solved the dog’s bladder problem.
With that, let’s introduce point #3.
3.) It’s all inductive reasoning.
Research and personal experience with dogs has taught me a lot about them. Both research and personal experience give me a few examples of specific dogs with specific problems, and then go on to generalize that this training method or that will work with every dog.
Making a general conclusion from a few specific examples is called inductive reasoning… exactly like the above statement.
And we’re back to assuming all humans are alcoholics who smoke.
Those few dogs I have worked with do not represent every member of either their breed or their species. Any dog can make a lie out of any generalization. Assuming that your experience has taught you that something will work is just that – assuming.
It’s unfair – and unreasonable…and irresponsible.
And that’s what the dogs we work with, the dogs the scientists work with, and the dogs that the clients have worked with are. They are a few specific examples.
I hate inductive reasoning.
Inductive reasoning takes the future for granted, assuming that because things have happened a certain way in the past, they will continue to do so. You can’t prove the future by saying it will resemble the past because it has always resembled the past.
(It hasn’t.)
Dog training is largely based on this kind of reasoning.
If I haven’t managed to irritate you, here’s one last thing to think about.
Associative learning works with most animals – humans, dogs, cats, rabbits, etc.
You take the general context of that and apply it specifically to dog training.
(This is called Deductive Reasoning.)
Voilà.
Learning by association works.
i.e. : Dog training – because it is learning by association – works.
What you can’t prove is that it was learning the association that changed the behavior.
If, by chance, you’re wondering where I get this stuff, feel free to check it out for yourself!
Try Alexandra Horowitz’s book: Inside of a Dog
Or check out Scottish philosopher David Hume’s argument on the problem of inductive reasoning.
Contact Details
- Read Julie’s blog: The Koda Diaries
- Follow Julie on Twitter: Jaennutter
Any comments?
Julie and I really value your comments, so don’t be shy and share your thoughts. I would particularly like to hear from you if you…:
- are, to some degree, involved in dog training and science;
- have experiences, positive and negative, with dog trainers presenting their work as scientific (guilty as charged here); and
- share Julie’s perspective, or have a contrasting view.
Further reading
Dogs and science
- How to respond to fallacies in dog training – Post about logical fallacies in dog training
- The D-word – Article about what is really known about dogs and dominance: nothing.
- Dog training: traditional versus modern theories – Academic essay comparing traditional and modern dog training theories
Dog blogging and writing
- Dino Dogan, dog blogger extraordinaire – Interview with influential dog blogger, Dino Dogan
- The case of the learned dog – Interview with published scientific author on pets, Linda Case
10 Comments
HAHA, right on. Exactly what I believe about science in general LOL
Yes, that’s what I thought when I read it too. It’s not science and dog training that is based on extrapolating from an isolated, specific situation, it’s science as a whole.
Woo. Totally second that, even though I love science to death. Actually, that is what I love about science… the uncertainty. =] TY for the comment, and sorry for the delayed reply! I’m such a slacker….
Yep, but science being the last resort for the rational, touching on how reliable (or not) it is, is a hot potato, isn’t it?
I so thoroughly enjoyed this!! It had me grinning from start to finish.
I can always count on this blog to satisfy the geek in me.
Oh bless you, Ashley. Yep, Julie certainly has a way with words!
Ha! That’s awesome. I love geeky people =B And I’m not sure I have a way with words, but I’ll take that compliment!
he he he, before you know it, you’ll have your own geek army!
Finally! An understanding of the difficulty in convincing people that what they KNOW as a scientific fact is not knowing at all though it MAY be based in science. Thanks for the reminder – I can’t hear it often enough!
Oh absolutely, Elizabeth. It’s a difficult thing, isn’t it? Everyone is so busy trying to back up their point with whatever means they have.
It can be:
scientific ‘evidence’ that is misunderstood and biased to fit one’s purpose, or
emotional ‘common sense’: these things that feel so true, they simply must be, or
tradition: we’ve always done it, it must be true…
etc. etc. etc.
I am more sensitive to scientific ‘fact’ than other backing arguments, as I see science as the last bastion of objectivity. But that requires the science in question to be well-designed, quoted honestly and well understood, which hardly ever happens. Scientific ‘fact’ is being twisted as a political tool left, right and centre, but that’s all we have.
One Trackback
[…] This post was mentioned on Twitter by jana rade and Moritz, Laure-Anne Viselé. Laure-Anne Viselé said: Julie Nutter's guest post on Canis bonus: No Science About It, scientific dog training http://tinyurl.com/64a2nad […]