Help me recalibrate my brain: what are R+ actions/thoughts I should use where classic P+ responses are ingrained?

lorrainekat

New member
A follow-up to my other thread about introducing our new 11-month old pit bull rescue to cats, more generally I am concerned about developing a consistent training methodology. For the dogs I grew up with in my parents' house, where none of us had any knowledge whatsoever about training, there are so many situations I can think of where an immediate P+ response was natural for all of us.

We want to focus on R+ training to build the best relationship with our dog. What sorts of things should we do instead where our natural reaction is going to be to quickly and loudly say "No!" or "Uh-uh!"?
  • Dog jumps up on furniture we don't want it to?
  • Dog pulls strongly at the end of the leash when seeing other dogs/people from a distance outside (I am trying to avoid the use of a prong/pinch collar)?
  • Dog starts chewing on a non-toy object like a remote control?
  • Dog is begging for table food?
  • Dog jumps up on people who come over to visit?
  • Dog is whining outside of a door or inside of his crate?
I just want to avoid constantly having to tell the dog "No", as from what I am reading they aren't going to understand that well anyway, and learn about what R+ training can substitute (probably done in advance, prophylactically, to avoid ever getting into the situation where you have to say no, I'm guessing, but this can be challenging to think about how to approach). I've got a couple of books, "The Puppy Primer" and "How to Behave so Your Dog Behaves", but he'll be arriving at the house tomorrow and I want to do things right from the start.
 
@lorrainekat Pretty much anything you might instinctively react to with P+ can be reframed with the following mindset: instead of trying to reduce the frequency of the undesired behavior by making it unpleasant (P+) remove the reward for the unwanted behavior and capture or teach an incompatible behavior (R+).

Here's a version of that mindset for each of your examples. I don't go into huge detail each time, but there are plenty of resources out there (some of them written by me, lol) on how to do each of these things with R+ (and typically a dash of P-).

Dog jumps up on furniture we don't want it to?
Teach the dog "off" with R+. Set clear criteria to make it easy for the dog to learn the difference between when you do and don't want him on the furniture. Immediately ask the dog for "off" every time he violates the criteria. For example, if the rule you want is the dog has to be invited, tell him "off" every single time he comes up uninvited. Or if the rule is he can come up but not put his head on your laptop keyboard, he has to go "off" every time he does that.

Dog pulls strongly at the end of the leash when seeing other dogs/people from a distance outside (I am trying to avoid the use of a prong/pinch collar)?
Teach the dog how you want him to walk instead of telling him he's doing it wrong. Show him there's a loose leash zone by your hip where he gets treats and other rewards if he volunteers to walk there. Use P- to show him that pulling is ineffective. If he pulls (ideally right before he even gets to the end of the leash), change direction. Never follow a pulling dog towards what he wants or you will show him that pulling works. Look for opportunities to walk towards what he wants while he's behaving nicely, and immediately walk away from it (or make a perpendicular turn) just as he starts to pull.

Dog starts chewing on a non-toy object like a remote control?
Teach the dog "leave it" with R+. Use it every time he puts his mouth on an non-toy object. Add extra rewards when the dog chooses an appropriate object for chewing.

Dog is begging for table food?
Teach him what you want him to do at mealtimes instead of telling him what not to do. Teach the dog where to sit and what behaviors to engage in when people eat. For example, put a mat someplace you wouldn't mind your dog lying down when you're eating. Teach him "go to your mat" with R+, and then use it when you eat. Capture any non-staring behaviors and reward them.

Dog jumps up on people who come over to visit?
Teach him what you want instead of a jump rather than telling him not to jump. Use +R to create an alternative behavior, like "go to your mat" or even "sit." Use -P when jumps actually happen by training with "guests" who go silent, still, and turn away in response to jumping. If you can't train the guest in advance, use a leash during greetings (in the short term) and step on it with just enough slack so the dog can volunteer a sit, but not enough that he can start a jump.

Dog is whining outside of a door or inside of his crate?
Capture any silence you get, increasing the requirement of duration of silence to trigger a reward. Teach "hush" as a cue with +R.
 
@lorrainekat
  1. Management. First, think of all those typical situations that are likely to arise (your list is a great start). Find ways to prevent them by default. Examples could be using an x-pen to keep the dog away from off-limits furniture, giving him a freshly stuffed kong to chew on away from the table during dinner, putting him on leash before visitors come in. Exercise him at home before walks and use a front-clip harness to get enough physical control to walk safely.Take preventative measures by default to totally prevent the issues without training.
  2. Differential reinforcement of incompatible behaviors. Choose some behaviors that are incompatible with those unwanted bad behaviors. For example, laying on the dog bed is incompatible with laying on the couch. Sitting for greetings is incompatible with jumping. A down-stay away from the table is incompatible with begging. Walking next to you, or checking in and making eye contact with you, or bumping your hand with his nose, are incompatible with pulling ahead on the leash. Start teaching each of those behaviors (focus heavily on sit, down, and a bed/mat settle for a great start). Teach them like tricks, using treats, separately from important times like visitors entering or mealtime. Get them really solid with no distractions. Then, start asking for those behaviors in gradually more realistic situations to set up behavior you like. Ask for eye contact outside on walks. Put him in a down-stay on his bed before meals, before sitting on the couch. Practice having him sit before greeting him when you come home, then have other folks do the same.
  3. Avoid reinforcing the unwanted behavior. Pay attention to what he wants in any given situation, and don't give that to him if he's doing things you don't like. Begging at the table--he wants food. Don't give him any food at the table, ever. Jumping on visitors--he wants to greet, smell them, get attention from them. Don't let him carry out greetings with jumping, ever. The only way he gets visitors to look at him, pet him, speak to him, even acknowledge his existence is to sit politely.
With practice and consistency, the management in 1 can gradually fade out as you rely more and more on his training to perform good behaviors in various situations.

To answer your question more directly, if your dog does something you don't like, a good response in the moment would be to cue an incompatible behavior and reward it. For example, if my dog tried to beg during dinner, I'd tell him to "go to bed", praise him once he was there, and then continue eating. When he's on his bed, he might get an occasional surprise treat (more treats were used in the beginning--now they're not really needed).

However, most of the work involved in force-free R+ oriented training is proactive (1,2,3 above). The more the dog is doing bad things you feel you need to respond to, the more there's probably lacking in management and proactive training.
 
@lorrainekat We had dogs when I was a kid that we trained using force and punishments too. The biggest change in mindset for me was to teach my dogs what TO do, not teach them what NOT to do.
I would highly suggest checking out clicker training and using it to mark your dog's good behaviors. Every day (once you've "loaded" the clicker) click and treat 50 different positive behaviors. Puppies offer good behaviors all the time, but they often only get attention for the bad ones. Mark the moments when he's on the leash and looks back at you. Mark the moments when he is chewing on an appropriate toy. Mark the moments when he is sitting calmly (kikopup has a great video on encouraging calm behaviors).
Ignore or redirect from the negative behaviors. Use management tools such as a crate or ex pen to set him up for success.
 
@lorrainekat I'm assuming you're using shorthand for positive/negative reinforcement and punishment.

You seem to have a pretty good grasp on things, you're right, dogs don't understand "no" innately. Dogs do understand stuff kind of the way humans do in the sense of body language, fluctuation and timing of voice, etc. Dogs also rely heavily on smell to put meaning to situations (as far as emotional context goes).

I'm no expert by any means (having gone to school for addiction studies and psychology gives me a great understanding of classical and operant conditioning, but I'm definitely not a dog psychologist...), but from everything I'm reading on this subreddit it sounds like the most important thing is consistency and using these tools we're given to increase the likelihood of good behavior occurring again.

As far as using punishment as a tool. Punishment doesn't necessarily mean hitting your dog. Something as simple as not giving attention to bad attention seeking behavior will suffice.

Also know about negative reinforcement. It's where a lot of trainers mess up. Think of positive reinforcement like drinking a beer for its euphoric effects. It feels good so you're more likely to do it again. Negative reinforcement is like drinking a beer to remove a hangover (never do this). You're removing something painful or returning to normality, and are likely to repeat the behavior. In the context of dog training, paying attention to a dog who is misbehaving, or outright allowing a dog to avoid something entirely that is stressful (avoidance is negative reinforcement) are some immediate examples.

I hope that helps, good luck!
 
@godloveme1
As far as using punishment as a tool. Punishment doesn't necessarily mean hitting your dog. Something as simple as not giving attention to bad attention seeking behavior will suffice.

Yup. Negative punishment. We advocate for R+ and P- here, but not R- or P+.
 

Similar threads

Back
Top