Hi everyone, I just want to share the links to some of Pippa's newest training articles. These ones live over on our sister site, Dogsnet.com - they're all about modern training methods, how they work, and why we love them. What Matters Most When Choosing A Dog Trainer The Changing Face Of Dog Training Operant Conditioning In A Nutshell Clicker Training: A Promise To Your Dog We hope you enjoy them. Let us know if they raise any questions, or if you have any suggestions for new article topics! Sammie
Just skimmed through the trainer article. It's so important to choose the right person to help with training. Had a bad experience once with a trainer who used harsh methods - never again! These guidelines are spot on.
Hi I've read through the piece on operant conditioning. I think the use of the definite article in sentences such as "Operant conditioning is the process by which animals, including dogs, learn to change and modify their behavior" (my emphasis) colours the discussion in a biased way. Dogs learn and change their behaviour as a result of classical conditioning. And non-associative learning can change behaviour too. Second, while negative reinforcement can be something to avoid such as in the case of force fetch training, it need not draconian measures to be of value. Mild pressure on a harness, which is released when the dog stops pulling constitutes negative reinforcement.
Hi again I would add to the piece on clicker training, that a click also ends a behaviour--the sound of the click also acts as a terminal stimulus. That aspect of a click is used to great advantage in clicker retrieve.
Hi Michael - thanks for sharing your thoughts. Perhaps it would be clearer to say operant conditioning is the process by which they learn to change voluntary behaviour - which is where the bulk of our focus as trainers lies We completely agree that negative reinforcement doesn't have to be extreme, or painful, to be effective in changing behavior. However, it does have to be unpleasant - and the use of aversives of any kind, including no pull harnesses, has been associated with poorer outcomes for dogs - including an increase in aggression, and a decrease in willingness to offer new behaviors. So we advocate for positive only training approaches, wherever possible.
This is an interesting point! The advantages and disadvantages of using a clicker to end a behavior with duration are really worthy of a whole article of their own - I'll pop that on the end of Pippa's giant List Of Article Ideas!
Thanks Sammie for the interesting comment. I think I know why you have emphasized voluntary choice in your definition of operant conditioning. Under classical conditioning, and forms of non-associative learning the notion of voluntary choice on the part of the dog does not seem to be apt. Up until recently I would have been inclined to agree with your novel definition of operant conditioning. I am now not so sure. Robert Sapolsky's book Determined casts serious doubt on the notion of free will (voluntary choice). I would be inclined to define operant conditioning in terms of its emphasis on the consequences the animal gets from various behaviours, which is how I understand the Skinnerians to define the term. I agree with you that positive reinforcement trumps negative reinforcement: If one can achieve a result with R+, then it is preferred to all the other quadrants. Can you point to any evidence that harnesses result in the dangers you speak of associated with R-? I'd be interested in reading the papers.
Hmmm. It's a good idea. But I don't think I was suggesting that Pippa needs to write another article, which could be seen as an aversive!!!!!
LOL! This series (links to each part lower down this page) https://www.companionanimalpsychology.com/2012/07/positive-reinforcement-and-dog-training.html links to a range of studies on punishment in training, including some that had leash pulling in the aversives studied.
I use a no-pull harness, but I had no idea it was aversive. If I hook the leash to her collar, we don't get anywhere, and I'm not wanting to choke her to death to get her to walk or stop her from running full speed into the road. She's 6 1/2 months old and definitely "stops" walking by sitting or laying down, often times in the middle of the path. The front clip harness stops her from pulling against the leash (cause it tends to turn her around if she pulls on it) and allows for loose leash walking without much hassle. She's still a stubborn dork, but it seems MUCH less aversive for me to use that harness rather than just clipping to her collar. Perhaps I need to go back to basics. I'm relatively new to this.
Hi there, We definitely recommend using a harness, rather than a collar. As you say, we don't want her to do actual harm to her throat. However, because the no pull harness is a form of aversive (in that it works by making pulling unpleasant) re-training a loose leash walk from scratch so that you can move away from the no pull harness in the longer term is definitely a great plan. If you need to use the no pull harness for safety reasons in the meantime, that's fair enough - safety has to come first! Do you receive Pippa's Training Tips emails? If so it's worth searching your inbox for one called 'Ditch The Destination' - it's a great summary of how to get back to basics with a loose leash walk
I mean it's safer than hooked to her collar, but I didn't ever see it as aversive. Is it just because it makes pulling pointless? She really just isn't triggered to pull in the first place unless she REALLY wants to eat a rock or something LOL.
Bless her! But yes, anything that stops a behaviour by making it unpleasant, is an aversive. No pull harnesses stop dogs pulling by making pulling feel unpleasant! Or... to think of it another way - if pulling while wearing a no pull harness felt just the same as normal walking on a normal leash..... the no pull harness wouldn't work To give a slightly silly example, but maybe one that helps demonstrate a little.... Imagine if you were walking along holding someone's hand, and every time you got ahead of them they tugged your hand sideways and pulled you off balance/made you face the wrong way. It would start to get pretty stressful after a bit. You'd walk 'nicely' next to them to try to stop it happening. You would change your behaviour to avoid a negative consequence. An aversive. That's basically how no-pull harnesses work. So we advocate for using them as a short term solution where needed to keep you or your dog safe, but alongside a training plan, that aims to build a nice loose leash walk in the long term, using positive reinforcement This can be a slow process, and takes real commitment - but you can do it! And we're here to help with any questions
I'm not sure why it works because she has never been "flipped around" literally. It just puts the resistance to the side and not the back. She didn't pull on it ... ever. Clip that leash to her collar and she will strangle herself on it. The harness she doesn't even try. She still shows the desire to move in a direction or whatever, but she doesn't ever act like she's pulling a sled. I'm still a little confused on the adversive thing because there was no "tugged sideways" moment. Anything that "frustrates" her is adversive? Any time I stop her from eating a dead cat, that is frustrating her. Training her is teaching her to frustrate herself and be okay with it. Drop it, leave it, even sit is a redirection from "what she wants" to "what I want." I am genuinely confused now about what I should and shouldn't be doing. I would never punish her or do anything that hurts her, but I feel like that bar is now raised so far that nothing can get over it. Do I need to train her in her collar? Is walking in a harness inherently bad and must be stopped? We train 24/7. That's all we do besides sleep, eat and play/sniff things. But, I have been doing things like having her sit rather than run into the bathroom and grab the towels. If she does stop at the door, mark/reward. If she doesn't, she grabs a towel and runs off ... so I ask her to drop it and mark/reward. But if I wanted a towel, and someone kept tricking me into giving it up, it would be stressful. Maybe I'm having trouble generalizing what you have said into specific actions.
It can be difficult to pick apart what's an aversive, and what's 'redirection' in a non aversive way, I know. I'll try to help! In learning theory we divide consequences into 'things get worse', 'things stay the same' and 'things get better'. Aversives fall under 'things get worse' - for example discomfort when pulling - they were comfortable when they didn't pull, they pulled and became uncomfortable. Gently preventing your pup from reaching a dead animal falls under 'things stay the same' - they had no cat to eat before, they still have no cat to eat . And the towel situation falls under 'things get better'. This is because you're actually setting up a behaviour chain for your pup - a sequence of actions - that end in a reward (food). The chain is: grab towel -> run off -> get told to drop it -> get reward From you dog's POV this is very much 'things got better'! In fact the reward at the end actually makes grabbing the towel more likely to happen in future, too. So your decision to try to prevent the towel grabbing in the first place is a great one. Our aim as trainers is to have 'making the correct choice' always culminate in 'things get better', and to use 'things stay the same' (eg redirecting when there is a dead cat) to reduce unwanted behaviours. Please don't feel you have to always be perfect at this - the occasional accidental small aversive will not cause big problems in a happy healthy pup! But you're likely to find training goes more smoothly and is more fun overall, if you try to follow positive principles as much as possible I hope that helps!
