Dogs Notice Everything

There is a purple bathmat on the floor with a small piece of kibble next to it. We can see the head and shoulders of a sand-colored dog, who is looking warily at the treat with her body language pulling backwards.
Conflicting Cues: Clara can’t figure out how to get past the treat on the floor to get to her mat.

I am fascinated by how dogs perceive the world.

Dog experts and ethologists have been telling us for a while that dogs discriminate beautifully, but generalize poorly.

What this means in our training lives is that dogs notice all the little things that we, as trainers, tend to do the same every time. Humans fall into patterns easily and we randomize with difficulty.  Dogs notice the patterns. When I teach my dog to run a few feet to go around something and return, I usually use a lamp pole or fire extinguisher. So when I get out one of those items, it’s obvious to her what we are going to practice. It is _so_ obvious in fact, that uttering the verbal cue as she performs the behavior is probably just so much background noise to her. At the beginning stages of learning the behavior, this works to our advantage. I can get her to do the behavior just from context. But later, as we teach the verbal cue, it’s our job as trainers to vary everything else possible.

I can vary a lot of things with the “go around” behavior. These include the following.

  • the object
  • where I stand
  • the direction of travel
  • the distance to the object
  • the room we practice in
  • the presence of my other dogs

And you know what? I can vary all that, and even then she probably doesn’t know the cue. I left out a big factor. She has been watching my body language the whole time. Just try to teach a dog to go around something without clueing them into that with your body. (OK you herding people, I know you all have to do it all the time.) Sometimes we have to take the extreme measure of getting out of sight of our dogs to know whether they know the cue.

The dog starts to understand that it is all about that little sound we were making after we winnow out the contextual cues. Some breeds are better than this than others, and experienced dogs learn it faster. But none of us trainers can afford to skip generalization.

There has been quite a bit written on reasons a dog might not respond to a cue. Here are three nice blog posts about it:

Three Reasons Why Your Dog Isn’t Responding  by Eric Brad at Life as a Human.

The Disappearing Sit by Kevin Myers at DogLovers Digest.

“He Blew Me Off!” by Nicole Wilde at Wilde About Dogs.

Sue Ailsby in her Training Levels books approaches generalization in the best way I have seen in any book or system. She includes explicit instructions on generalizing in every behavior in the Levels. (Full disclosure: I helped edit the book.)

Sue writes in her colorful way about asking her service dog Stitch for a favorite behavior (spin to the right) in a new context:

I was THREE FEET from where I always ask her for this behaviour, holding a dish which was empty instead of full, and I was facing north instead of east. She wasn’t “blowing me off” or “giving me the paw.” She truly had no idea what I was asking her for. Those 3 little tiny differences changed what she saw so much that the behaviour seemed completely different to her.

— Training Levels: Steps to Success, Vol. 1, p. 226

Once you are in on it, the difference between dogs’ and humans’ perceptions is fun. But this disconnect can be bad news for some unlucky dogs. Force-based trainers seem to glory in the idea that when this communication breakdown happens, their dog is indeed “giving them the paw.” I have heard this expression uttered more than once in complete earnest.

Burch and Bailey wrote in “How Dogs Learn,”:

Well-intended owners sign up for classes at their local obedience school, only to get instruction on heeling and figure-8s…….Obedience instructors who run classes designed around formal exercises think their training will ultimately result in a well-behaved dog at home. They firmly believe the behaviors taught in class will generalize to the home. But the majority of obedience class dropouts in a 1991 study told us they quit obedience classes because they saw no changes in their dog’s behavior at home. This suggests that training is not generalizing the way some trainers think it is.

— How Dogs Learn, Mary Burch and Jon Bailey, 1999, p 78-79

The world of dog training schools and classes has doubtless improved since 1991. But at every obedience trial I have ever attended, I have seen handlers in states of rage or at least confusion at their dogs’ surprise inability to perform. Even if you attribute the dogs’ problems to “stress,” where did the stress come from? Changes in the dog’s usual training environment. Changes that involve strange dogs, strange people, new noises, a road trip, in addition to differences in the visual environment.

I have seen the furious trainer phenomenon once too many times. So I made a series of videos showing my dogs confounded by small changes in the environment, the props, and in one case, the effect of a previous reinforcement history.

In other words, I set them up to fail.

I admit it; I experiment on my dogs. I push the envelope at times. But just so you don’t think I am a complete meanie: in the videos, they have already succeeded and been rewarded several times. After they fail, I give them an opportunity to perform an alternative behavior and get rewarded again. So from their point of view, this is a normal training session with an imperfect trainer where one time they fail to perform the behavior and fail to get the reward. The only difference is that this time, for once, I actually had a clue it was coming.

The Missed Cue

In the Missed Cue video, I move my dogs farther and farther from their mat in a boring hallway and cue them to go to it. And then we see it. Here they respond confidently to my verbal cue. A few inches away, they look at me in utter bewilderment.  Some viewers have pointed out that besides the difference in distance, the dogs fail when the starting place is near the end of the hall with an open door next to them. So yes, it may have been more than just the inches. But think how different that is from how our minds work. The same word doesn’t compute when suddenly there is a familiar bedroom door to our left?

Missed Cue: Paw Touch

In the paw touch video, I let Summer practice her paw whack, a favorite behavior, on several objects, including a little basket lying upside down. Then I turn the basket over. Summer, who learned to fetch in 17 shaping sessions using that same little basket, is helpless in the face of that history. She fetches the basket proudly and prances around with it. Her discrimination is so fine that she reacts differently to the same small object depending on whether it is right side up or upside down.

The Missed Cue: Generalization

The generalization video shows Zani going around some objects. I flummox her by substituting a short plastic box for the pole lamp we had been using. She interacts with the box with a variety of behaviors, then checks back with me for further instruction. I then substitute a fire extinguisher, which is a vertical object like the pole lamp. This time she figures it out. (Nowadays I would have handled that a little differently. Stay tuned for the next blog entry for more about that.)

By the way, I know there are times when dogs understand our cues and do something else. But I believe that happens a lot less frequently than many trainers think. And when it does happen: that’s just a different training challenge.

I would love to hear from you readers about times your dogs surprised you by not understanding a cue. I hope to get some replies down below. Also, if you have a video that would be great. Submit it on YouTube as a response to one of mine. (Send me an email if you don’t know how to do this.) Wouldn’t it be educational to have a whole string of these?

Discussions coming soon:

Thanks for reading!

Copyright 2013 Eileen Anderson

Addendum, 8/22/12

There is a new Missed Cue video in the series. Attack of the Zen Field.

24 thoughts on “Dogs Notice Everything

  1. Great! Great! Great! I was always taught that dogs don’t generalize well but I thought that only applied to surfaces and rooms. My dog can sit on grass but not concrete, or my dog will sit in the living room but not the bedroom. I had no idea that it also had to do with our body movements, the orientation of objects, or the presence of small distractions. Thank you for this insight….maybe I will be more tolerant and patient with my genius dog. 😀

    1. Hi Jamie,
      We should all aspire to be as smart as our dogs! I have the surface problem bigtime with Zani, who because of her short hair and bare belly (I think) believes that she should always sit on concrete and down on rugs. Of course I followed her preference at first to get the behavior (knocking head against wall) so we’ve had to work very hard and are still not wholly successful on the verbals for those. My next post will be about how I fixed the “distance to mat” problem and some different (probably better) ways to generalize to objects in the generalization video. Thanks for writing!

  2. So very interesting! I have added this to my list of Things I Thought I Knew But Now Understand better. The videos are great – thanks for sharing.

  3. Nice post and examples! I learned a similar lesson as your first one when teaching go to a mark. Sienna loved, loved, loved going to the mark – until it was about 10 feet away from us. She’s a very handler-oriented dog, and is used to working within that 10 foot radius of ME. At that point the behavior broke.

    The way I fixed it was to work in a “round the clock” pattern. The mark was in the center, and I started with Sienna at 12 o’clock and myself at 6 o’clock. I started with her about 5 feet from the mark and myself a few steps back from it. Once she got the idea of stopping on the mark, I backed up a few steps at a time until it was halfway between us.

    Then I moved us so Sienna was at 1:00 and I was at 7:00, and did several more reps in that orientation. Lather rinse, repeat. Over a few training sessions we worked our way around the clock. The next step was to leave Sienna in position and move myself around the clock. Eventually I worked my way around until I was next to her.

    After that, I could gradually increase the distance. But each time I increased it, I did round-the-clock again at the new distance. Now I can send her to a mark in another room with no problems.

    I look forward to hearing how you worked through the similar go to mat generalization problem! I think with some dogs, they don’t want to go so far away from the handler on targeting behaviros, and need to be eased into it slowly.

    Oh, and at first when she arrived at the mark, she downed. So for a while I had a dog who would race to the mat, throw herself into a down and bark. Had to start over with a new mark to fix that one! I have a video around here somewhere, I’ll upload it so you can see my crazy downing mark dog. 🙂

    1. What an interesting solution Shannon! That sounds like it would result in a wonderfully strong mat behavior. I bet it is nice preparation for a drop on recall as well, with the mat between you part of the time. I hope you do put up a video. Would love to see it. Isn’t it cool how many dogs love these distance behaviors?

  4. Hi! I am commenting on an older post because something that happened this morning is *so cool* and relevant to this post that I want to share it here.

    I’ve had my dog for about five months, and I have never put much work into teaching her to discriminate between “go to your mat” and “go to bed” (her crate). Nevertheless, she usually uses context cues and gets it right (and I used to help her by pointing at her mat, since I’ve not really done formal sessions with that cue).

    I know she’s probably not responding to the verbals, but I also always figured I knew what the salient details were for her–the human has a bowl and her pajamas on, go to my crate. The human has nice clothes on and no treat, go to crate and get a kong. The human is carrying a big weird purple thing (a Kyjen slo-bowl), go to my mat.

    However, this morning, we were out of coconut oil, which I always add to my dog’s breakfast. So I added camelina oil, which goes in her dinner, to her bowl instead, and cued her to go to her crate. She turned and ran confidently to her mat.

    Cool, right? My dog doesn’t care what bowl I am holding, what words I say, or whether I point. Her salient cue is olfactory–coconut oil means crate. Plant oil means mat. Dogs are *so fascinating.*

      1. I am glad you agree it’s cool! 😀

        Nope, I am blogless, sadly, which is strange, because I love talking about my very cool dog.

        1. Well, if I write on that topic again I’ll ask you if I can use that story. It is such a **great** example!

Comments are closed.

Copyright 2021 Eileen Anderson All Rights Reserved By accessing this site you agree to the Terms of Service.
Terms of Service: You may view and link to this content. You may share it by posting the URL. Scraping and/or copying and pasting content from this site on other sites or publications without written permission is forbidden.
%d bloggers like this: