July/August Issue

by:  Jean Donaldson

Dear Jean,

I just got a new nine week-old Rottie puppy. He’s stunning, smart and generally friendly but growls and snaps if I go near him while he’s eating. He also does this to my adult Rotties. I’ve never seen this in a puppy so young. Is he some sort of lemon? Is he a dominant dog? Is there anything I can do? Help!

It is indeed alarming for most people to see frank aggression in puppies. In the case of resource guarding – food, bone, bed etc. possessiveness – there is good news and bad news. The good news is you can start addressing it in a young, hopefully plastic, spongy puppy with weak jaws. The bad news is that there is some sentiment out there among trainers that aggression in puppies is an insidious sign of the problem having Deep Genetic Roots and therefore fruitless to tackle. I’m going to explore the whole nature-nurture debate later but for now will simply say that there doesn’t seem to be any overwhelmingly tidy correlation between behavior problems that are thought to have a strong genetic component and their susceptibility (or lack thereof) to behavior modification.

I recently had a similar case, in my own foster puppy. Buffy, a stray six week-old Chow, presented with object and food guarding against people and dogs. I elected to not touch the dog-dog issues, which is a relatively common approach. Her socialization and play skills were coming along nicely and she was developing good acquired bite inhibition. The guarding against people, however, needed to be actively resolved. The following is a summary of Buffy’s food guarding exercise regime. Incidentally, Buffy also presented with socialization deficits and severe body handling problems, which were also addressed, as was her object guarding. The key to good hierarchy design is small enough incremental steps that at no point do you see the original guarding problem. In the case of a puppy, such as this, there may actually be more aggressive increment jumps. I did a few other things in the can’t-hurt-might-help category. These included impulse control (stay, off and wait) and extra soft-mouth training.


When approached while eating from her dish, Buffy would freeze and, if approach continued, growl briefly and then lunge and snap. If touched while eating, she would growl simultaneous to whirling and biting. Due to the independent body-handling problem, this had to be partly resolved prior to combining it with food bowl exercises. Buffy did not guard an empty dish.


Step 1 (day 1): Installment feeding of canned food. I sat on the floor next to Buffy’s dish and spooned in one mouthful. Once she had swallowed, I spooned the next mouthful into her dish. By the end of the second meal, she demonstrated a clear happy anticipatory orientation to my spoon hand after each swallow.

Step 2 (day 1-2): Overlap. This was essentially the same as Step 1 except that I added the next spoonful to her dish while she was still consuming, always a much dicier proposition. We did this for three meals without evidence of guarding seen.

Step 3 (day 2-3): Approach overlap. I was now standing. I spooned larger installments, withdrew two paces, re-approached and added the next spoonful while Buffy was still consuming. So, this combined approach with the overlap exercise. We stuck with this for three meals, at end of which time a Conditioned Emotional Response (CER) had become evident – Buffy wagged and looked up on approach. We then repeated the exercise for one more day (5 small meals) with larger withdrawal distances and intervals.

Step 4 (day 4): Trumping. Now I spooned her entire puppy-sized ration into her bowl. I withdrew five paces, paused 15 seconds, approached and added a (hidden) marble-sized dollop of goat cheese. I had pre-auditioned the goat cheese out of context and ascertained it to be in Buffy’s Top Five All Time Foods. I withdrew to six paces and waited for Buffy to continue to consume – this was not immediate (typical of trumping – dog orients to handler rather than back to dish) – then repeated. On the third trial I got a clear CER– withdrawal from bowl on approach, orientation to me and tail wag. Clever little thing.

Step 5 (day 4-6): Covering High Value Base. To up the ante, I tried some approaches while she was consuming a top food (bowl of treats), rather than normal meal ration level food. I trumped it with higher value stuff (gorgonzola). In two trials, I once again saw her happy anticipatory CER, a very rapid curve indeed.

Step 6 (day 4 onward): Cold Trials. To better simulate real life, I initiated random trumping. At least once per meal, from a random direction, at a random time and with one of Buffy’s top foods, I approached and added the bonus. Better than 80% of the time, I got an evident “yippee” CER. At no point did she guard.

Step 7 (day 8 onward): Generalization. I recruited my husband, colleagues in my office and a neighbor to do some random trumps, with careful monitoring for any evidence of regression, including the absence of “yippee” CERs to their approach. Had this been an adult dog, the hierarchy – and, notably, a much more gradual one too – would have been recommenced at the beginning by each new recruit, with likely accelerated progress rate for each successive person.

Step 8 (day 15 onward): Body Handling. It was only here that I commenced patting, grabbing or pushing her around while she was eating. In most cases this would come earlier (prior to cold trails), however with Buffy it took me this long to get the independent body-handling problem up to speed. The handling during eating exercise consisted of the body touch (later handling) followed by a trumping addition, repeated until the body touch/handling elicited the “yippee” CER. Buffy’s CER consisted of a wag as well as orientation to my hand. If I stored the bonus in my other hand behind my back or my pocket and reached with a blank hand, she would wag and orient to my face.

Buffy is now on maintenance with a cold trumping or body handling trial usually once per meal and use of other people whenever an opportunity presents itself. I ended up adopting her.

You can throw in bowl removals if you like, rather than sticking with approaches and body handling. The principles are the same. Good luck with your Rottie!

© Jean Donaldson, all rights reserved

Jean Donaldson is a native of Montreal, Canada. A graduate of McGill, Jean holds degrees in Music and Comparative Psychology.

In 1996 James & Kenneth Publishers published Jean’s first book, The Culture Clash, which has won numerous awards, including The Dog Writer’s Association of America’s Maxwell Award for the best training and behavior book of the year. Since its publication, The Culture Clash has been the number one recommendation for dog trainers of The Association of Pet Dog Trainers (APDT) each year it has had a recommended reading list. Her other books include the multiple awardwinning MINE! A Guide to Resource Guarding in Dogs, Fight! A Guide to Dog-Dog Aggression, and the newly released Oh Behave! Dogs From Pavlov to Premack to Pinker. She has also recently authored a DVD on basic obedience, Perfect Paws in Five Days.

In 1999 she founded The Academy for Dog Trainers at The San Francisco SPCA, which has gained a reputation as the Harvard for dog trainers. Jean is currently pursuing a doctoral degree in Evolutionary Biology. She lives in the San Francisco Bay area with her Chow, Buffy, adopted from The San Francisco SPCA in 2002, and currently the only Chow registered with the North American Flyball Association.