AMICHIEN BONDING

"The Dog Listener", by Jan Fennell

I was having some problems training my collie Suki, so I contacted my favorite rescue for advice. They sent me to a particular trainer; when I contacted him, he told me that he was a big promoter of Amichien Bonding, a technique based on dominance theory. Before I considered hiring him I wanted to know more about it to make sure I was hiring someone who would make the problem better, not worse.

As soon as I heard the word, "Dominance" I became wary. There are a lot of trainers out there who physically coerce dogs to establish what they call dominance. I have a book on mushing training, for example, that details the correct way to use a whip to train a lead dog. There has been a good deal of research since the early 1960s strongly suggesting that positive punishment based training protocols do not produce the best results. But I try to be an honest skeptic, and if this trainer - who came so highly recommended - supported Amichien Bonding, I figured I should at least read enough to know what it was. He recommended the book, "The Dog Listener". This is actually two books in a single binding, and he suggested I skip the second book and concentrate on the first.

Several trainers I greatly respect warned me that Amichien Bonding was not based on good science. I frankly was expecting to read a book heavily leaning towards positive punishment (P+). My fears were not justified.

The book explicitly rejects use of physical coercion as being undesirable.That is in line with the research that has been done since the 1960s. But there is another problem with the book.

The book bases its theory of dog training on the behavior of wolf packs. This has a great deal of common-sense appeal: dogs are likely derived from wolves, and still share a significantly common amount of genetic material. They can interbreed as can - for example - dogs and coyotes. They look quite a bit alike. There are, however, two flaws with this approach:
  1. The wolf pack interactions described in the book are apparently based on studies done prior to 1960 of captive (zoo) wolf packs, which were comprised of unrelated individuals in a highly artificial setting. I say "apparently" because in the entire book there is not a single citation. Since the early 1960s (or maybe the late 1950s), there have been rather a large number of studies of wolves in the wild that have discarded the conclusions of the earlier studies. For example, wild wolf packs are not run as Stalinist dictatorships, with a single wolf (or single mating pair of wolves) dictating everything about pack life. Wolf packs are generally comprised of parents and their progeny, and violent interactions between members are quite rare. Pack members do not generally wait to begin eating until the Alpha is done, a key claim of Ms. Fennell's. Furthermore, humans lack the morphology to communicate as a wolf, so it is not clear what a wolf would think of a human trying to "speak" wolf.
  2. Humans and dogs have been co-habiting for something like 100,000 years during which time humans have selected for and against certain traits. Modern dogs simply do not act like wolves. Both wolves and dogs, for example, have a critical window in puppyhood to develop skills enabling them to survive. However, these windows are significantly different in timing and in content. Wolves left to their own devices will pack: packs are quite stable. Feral dogs, on the other hand, do not pack, as members come and go. Wolves are generally capable of surviving without human assistance because of their great hunting skills. Dogs, on the other hand, generally do not hunt well enough to feed themselves. Wolves make really, seriously bad house pets.
As you might guess from the above, the book is generally stunningly unscholarly. From where does the author's knowledge of wolf behavior derive? She makes no claims to have studied them first hand. The only observations she claims to have made is from watching films, presumably some unnamed documentaries; or perhaps she watched Disney's "The Jungle Book". There is no way to challenge the validity of her source material because she does not identify it. In the roughly 184 pages she does not provide a single citation.

Contrast this, for example, with Barry Eaton's book "Dominance in Dogs: Fact or Fiction?" The text is 64 pages long. There are citations on virtually every page, and the bibliography is 3 pages long. The book is written for consumption by the general public.  If Eaton has his facts wrong, I know how to challenge them. If he is misrepresenting a study I have a chance to say so. His book, by the way, makes most of the points I've made above. I read Ms. Fennell's book first, and was aware of most of the shortcomings from my previous reading. It is published by Dogwise (www.dogwise.com) and has an ISBN of 978-1-929242-80-1. I highly recommend it.

Ms. Fennell's book follows a formula very common for training books, especially "training breakthrough" books. They are laid out in the following sections:
1) My history
2) How I found this new method
3) Explanation of the method
4) How to apply the method
5) Success stories (which generally supply the bulk of the reading material)

Both good and bad training books seem to follow this formula, and I have no objection to it. Parts 2 and especially part 3 should be interlaced with  references to peer-reviewed studies, or books based on peer reviewed studies, if a book is to be considered credible. Absent that information the author is trying to convince me something works just because they say it works.

Now having said all of the above, I will state that it is likely Ms. Fennell's methods work.

The techniques she promotes fall in to two categories:
1) Daft suggestions based on an inaccurate understanding of wolf culture. These suggestions are harmless. I doubt that eating a biscuit or cracker before feeding your dog is going to make much of an impression on it one way or the other, beyond noticing (albeit possibly with great energy) that its feeding protocol has changed. Unlike other books that base training on an inaccurate understanding of wolf culture, this book does not advocate committing violence against a dog.
2) Classical operant conditioning P-, R+, and E . There is a good deal of research showing that using negative punishment, positive reinforcement, and extinguishment allows for a highly effective training regimen. Most of her suggestions focus on this. Her insistence on ignoring a dog for five minutes is (depending on the situation) either P- or E. She hammers on the importance of consistency, which is certainly a well accepted axiom. To her credit she briefly - very briefly - mentions Skinner and operant conditioning, which she is certainly not claiming to have created.

So the book is not really a bad book. This is not another example of a book making unsubstantiated outrageous claims. My objection to her book is that the portion of her book that claims to be a breakthrough is based on inaccurate information of wolf behavior that is likely irrelevant. To quote the prosecutor on the old Perry Mason shows, they are incompetent, irrelevant, and immaterial.

I give the book a grade of "C".