So Your Dog Understands "Sit"
Dogs are dogs and they learn our language very differently from the way a child might. When we teach a child to Sit on request on a chair in the kitchen, the meaning of the request is quickly generalized to chairs and other pieces of furniture elsewhere and anywhere. However, when you teach a dog to Sit on cue in the kitchen, for example, you might have a good kitchen-sitting dog, especially around mealtimes. However, your dog might not respond as reliably with other family members, or for you in other rooms, outdoors, in a dog park, on-leash on the sidewalk, with heavy distractions, or when you are out of sight.
As I explained in Barking Up the Right Tree, the Sit Test proves that the most common reason for non-compliance and misbehavior is that the dog does not fully comprehend the instruction, or instructions, in that context, or scenario.
For example, instruct your dog to Down-Stay in the living room and then, lie on your back with your arms folded across your chest, so that both of you are in a straight line, with your dog in a Sphinx prone-down, you in a coffin supine-down with just two inches between the top of your head and your dog's nose and then, instruct, "Rover, Sit", and film.
Now many of you won't even be able to get into position without your dog breaking the Down-Stay, several times, as you try to lie down on the floor. Equally as important, when you say, "Sit" — the easiest doggy command on this planet, very few dogs will sit. Some hyperactivate, take to the air, and come down on your chest to lick your lips and eyeballs, whereas others will creep and cuddle up beside you, rest their chin on your chest, close their eyes, sigh, and snooze. Influenced by the context (you on the floor), your dog reinterprets your instruction to mean either 'Playtime' or 'Sleepytime'.
Just last month, right after Billie scored 99% with verbal requests over three reps. of the C-S-D-S-St-D-St sequence, off-leash in the living room, we left through the front door to the sidewalk, where he barely scored 8% on-leash on the same test.
I find it entirely fascinating and illuminating testing response-reliability percentages in different scenarios and with different people, for example, between different family members, or between the trainer and the owner after board and train.
Once we understand what dogs understand from our instructions, the findings radically change how we teach and proof response-reliability to verbal requests. The two best bang-for-your-training-buck and enjoyable exercises to teach dogs to generalize the meaning of verbal instructions to all situations and scenarios are: to test the sequence every 30 seconds when your dog is off-leash and playing with other dogs, or sniffing; and to test every 25 yards when walking your dog on-leash. For example, just a three-mile walk would amass 211 mini training interludes, and each one in a completely different scenario.
I find testing distance commands to be the most revealing. With an instruction as simple as “Sit”, response-reliability is highest when the dog is toe to toe right in front of you, but then drops off precipitously with distance. Seldom do dogs score 100% on a protracted test, even when close and focused. Seldom do dogs score 0%, since they glean so much information from 'reading' your body language, facial expressions, tone of voice, and context. Instead, their response-reliability percentages range somewhere between 0% and 100%. This made me think of the notion of Progressive Comprehension. In this example, the closer you are, the better your dog understands your instruction.
I observed similar findings with the AutoTrainer — that emitted a cascade of 24 tones of increasing pitch and decreasing interval, culminating in a longer tone immediately preceding food delivery. At first, dogs ignored the machine until food delivery, then, they began to respond to the food-delivery tone, then to the one before that, then to the antepenultimate tone, etc. Eventually, some sleeping dogs responded with an ear twitch some 12-14 tones and many minutes prior to food-delivery.
Such lengthy, cascades of anticipatory stimuli are common in day-to-day life. A dog comes to the kitchen when dinner is put on the floor. The next day when food is put in the bowl on the counter. Then in following days, when the food is taken out of the fridge, when the fridge is opened, when the owner enters the kitchen etc. Each predictive stimulus increases in salience (comprehension) the closer the cascade approaches what the dog wants — dinner or a walk.
I knew one dog that activated when the ending theme tune of Mrs. Dale's Diary played on the radio in the late afternoon. The dog slept through the identical opening theme tune and for the duration of the program but would wake up, stand up, and move to the kitchen with the theme's second playing, knowing that the predictive cascade had begun. The dog became more excited with every progressive stimulus, starting with the owner yawning, and then stretching, rising from their armchair, going to the toilet, the toilet flushing, putting on the kettle for a cup of tea and then, starting preparation for the dog's dinner.
Lure-Reward Training and proofing dogs' cued-behaviors to 95% Response-Reliability removes and eventually eliminates most misbehavior and non-compliance and hence, the perceived need for aversive punishment.