Does anyone train positively without the treats?

Just curious! Back when I got my puppy all the books recommended not using treats during training, so I just used praise (and sometimes verbal corrections. Mostly praise, though.) Often I would give her a treat after the training session was over, but usually not during. She ended up being a splendidly trained dog with good manners. But then she was very intelligent and trainable to begin with. Is it possible to train all breeds without treats, or only the ones who are easy to train and eager to please (like the retriever breeds and such)?

Behaviors which you want your dog to repeat must be reinforced. A reinforcement is something your dog loves, something which causes the behavior to be repeated. All rewards are not necessarily reinforcers. A reward which you think is a good enough thing may not be to your dog. Most dogs love food because it is an intrinsic, basic, primal need and so most modern trainers use food rewards and sometimes special toys that the dog goes ape over. Praise is used too.

Praise alone is not always sufficient for all dogs. It's nice, but they get that and attention, love all the time anyhow. Dogs who are trained with a lot of corrections like praise because it is a predictor that punishment is unlikely to occur right now because praise and punishment rarely happen at the same time. So, it appears that the dog likes praise. But it is not always a true reinforcer....sometimes yes.

Dogs are only willing to please because there is something in it for them, directly or indirectly. They are not able to know or think about what is going on in your internal state of mind.

I use treats for every correct repitition in the beginning of learning a new behavior. When the dog has learned the behavior, I go to an intermittent reinforcement schedule and can phase out treats, but do reinforce sometimes to keep the behavior strong. Intermittent reinforcement actually strengths behavior, once the behavior has been learned.

I use to not train with treats too. My Lab, my GSDs. And they did fine too. But now that I use these principles, I see a big difference in response rate, eagerness and the dog seems to have a lot more fun training. Everything happens quicker; recalls, sits, downs, fronts.....all snappier and with a lot more gusto.

I also see it as a paycheck. The dog does a job for me and he gets something he highly values. Why not? I wouldn't go to work for just a pat on the back. I want a paycheck, sometime I value. It would make me work a lot better and harder. I'd actually show up for work. LOL.

Waiting till the end of a session doesn't mean much to a dog. The timing of reinforcers is vital. A reinforcer (whatever that may be) must come within 3 seconds of a behavior to increase the probability of it recurring again.

So, I don't know why some people have a problem with giving treats for a job well done. I don't know why someone wouldn't want to give their dog something he reeeeeeeeeeeelllllly loves. Variable reward schedules prevent the frequent complaint of: "My dog will only work for treats." This is a fallacy. The treats can be phased out for the most part later on, once the behavior is solid.

IT's certainly POSSIBLE to train lots of breeds with no treats. My question is why would you? In competitive obedience circles we are now seeing so many breeds doing so well that were thought to be "untrainable" and "too stubborn" ie Huskies, shebas, various hound breeds, terriers etc but really what they are is very difficult to train using punishment (collar corections and verbal corrections). Punishment didn't make them very willing or motivated to work. You also see really soft dogs people previously said were just "too soft" to do anything with doing really well because correction based training made them shut down under pressure is some shelties, aussies, bc's , etc I was at a big trial in August and was just thrilled to see a husy with a CDX working towards a UD. That dog had some of the best attention and heeling I've ever seen and looked like he was just LOVING working.

There are so many myths about using food rewards.

There are also people who "use food" but do all the wrong things with it and then come to the conclusion that training with food 'makes the dog too hyper' or "the dog will only work if the food is present" etc That's like giving an amateur a choke chain and saying 'go have at it'. We've all seen those dogs who just learn to tune out collar correctins and need increasingly large aversives to get a response. Just the food or just the choke chain is not what makes training. You need to know HOW to use food (or any tool), the difference between a lure and a reward, how to use a variable rate, how to thin out the reward ratio etc

Reward based training done well gives you a dog who knows how to be calm around food instead of hyper and obnoxious.
- using reward based traing gives you a dog who will keep trying, keep offering behaviors because he's not afriad to make a mistake because he won't be corrected for being wrong. That's a beautiful thing to see- really.

- it more precisely marks a behavior making it easier for the dog to figure out just exactly what it was the earned him the reward so learning is faster and fun which motivates both thte handler and the dog.

-It is easier for the dog to lean than using corrections which only tell the dog he made a mistake not what he should do instead.

-Food is also a huge motivator compared to praise which is actually quite weak for most dogs. You can get lots and lots of reps in using food which is very helpful in a long training session like an hour long class.

-Food helps dogs dogs overcome stress- that's actually a physiological response to having a filling tummy - very helpful for a dog who is stressed at training class or that is learning to perform a skill that scares him, or you are at a really big, caotic place like a trial (or dinner at someones house ).

-Over time, the dog associates the food rewrds with the work so the work itself becomes rewarding (I really see that in my own dogs who are just dying to work no matter what "the work" is)

- The principles and benefits of using food rewrads have been well established in scientific literature.

-Dogs need a job to do, why not use their food? They need to eat-- let their food work for you- you'll both love it!

Also it must be said that food alone should not be the motivation. It should be one of many things including verbal and physical praise, toys, games etc

Victoria Lamont wrote a really great article entitled "Food fight - fact and fiction in the food training debate" in Dogsport magazine. I wish I knew how to post it here. It's in the Nov/Dec 2003 issue in case anyone wants it.

- Good post Sam. Yes, I read some studies done on dogs while they're eating or being given treats and their seratonin levels rise, their stress hormones lower. It definitely has an internal physical and psychological effect.

I would love to read that article.

You have yourself proven dogs can be perfectly trained with little or no use of food rewards. They can also be trained without the use of harsh "punishments", and only minimal corrections and redirections instead.

Detection dogs are chosen for their work due to their high drive (almost an obsession) to work for their favorite ball or toy as a reward.

In a sport such as agility however, I see a huge value in the use of treats to lure and also reward with the addition of the clicker. I doubt you would find a better way to train those dogs and achieve the same levels of enthusiasm and results without the use of positive reinforcement click-and-treat methods.

I use treats in the initial learning process, but use social interaction when dealing with social and behavioral issues.

It does very much depend both on the dog and what you are trying to achieve with the dog at that particular moment.

In dealing with social and behavioral problems, I rarely use treats or corrections, and switch to the use of direct communication as the dog's leader instead.

It's a wonderful article. Pm me your e-mail addy if you like and I'll scan it and send it to you.

Ooooo thanks! I'll p.m. you.

I'd like this article too, Sam. Sounds like good reading. There's been a big debate about food use, the past twenty years. Clicker training has had a big influence on previously established methods of training. When I first started training for a living, I had many students ask why I asked on registration forms that they bring treats. If they were in a certain age bracket, they'd been taught that you should not make your dog dependant on food reward as you won't always have that type of reward on you. Back then, choke chains and no food, was the prefered method. Barbara Woodhouse, her assistant at one time- Brain Killcomon, Koehler, Mathew Margolis, were the popular trainers. Old habits die hard. Your post (#3) states very clearly why the use of food is valuable in all areas of training, not just competition work.

Another great thing you do using food or reward that can't be nearly as well achieved using other methods is get a darned near perfect recall (and other distance work, teaching "go outs " for obedience or 'go to matt' etc) .
To me this is hugely important seeing as having a solid recall is often a lifesaver for a dog and also it allows him to go along with his owner to places he wouldn't be able to go without it.

Using Premack's principle which I like to think of as "having your cake and eating it too" is absolutely invaluable.

ie with recalls-- I train my dog to come and sit in front of me for a food reward as soon as he sits and takes he treats I immediately release him to go back to whatever he was doing ie playing with another dog at the park, swimming on the lake etc. You set it up in training so that your dog is never choosing between what he wants and what I want. Coming to you is always a good thing. If the dog does as I ask - drops whatever he is doing, turns on a dime and runs to me and sits in front-- he is released to go back to playing AND have a treat. If he chooses not to come I walk to wear he is and the fun ends.
What you see with dogs who aren't trained with high value reward is they get in many situations where they would rather do what they are doing so all your obedience work falls apart when you aren't near enough to give the dog a correction or give them a look or a verbal correction or otherwise intimidate them into complying.

The reward doesn't have to be food of course but food often tops a dog's list and for the handler it's very easy and handy to use.

We use the same principle in sheep herding but there's no food-- the dogs LOVE to chase the sheep it's about the biggest distraction there is for a herding dog. It's incredibly difficult to get the dogs to stop herding the sheep but it's also 100% necessary for obvious reasons. I was thrilled when I started herding to see how well the same training principles work-- You ask the dog to DOWN and the INSTANT they comply you release (with your release word) them back to the sheep.
The first time we went herding and I finally got my aussie to down, I was so exhausted from running to get between her and the sheep to force her to down, that when the herding trainer told me to rlease her I was like " NO! I need to catch my breath! I can't breath! LOL
It's very hard to get those first few downs even with a super well trained dog, they are just so high and having a blast doing something so rewarding on an instinctive level so it's totally it's amazing to see those dogs drop like a rock when they realise the key to getting back to chasing the sheep is to lie down.
I'm not sure my description does it justice, but I'm trying to illustrate the point of having, willing, cheerful total compliance from your dog in even the most distracting environments is easy to accomplish if you understand the principles of reward based training.

I think it depends on the dog. Our pug will not do anything unless there is a treat involved. Gunnar on the other hand, doesn't care about treats but if you have a ball or tug, that's his motivation.

It depends on how much the dog values praise. My Border Collie will work his little tail off for a hug and a belly rub. Occasionally he actually prefers affection to food. My Papillon considers physical contact to be some odd human form of torture.

Sam, that sounds like an interesting article. If you have the time, could you send it my way too?

Sam, that is exactly how I taught my dogs and continue to practice on our hikes. They come (either a formal come) or a "let's go." They get a treat and immediately get relased to go do what they were doing; exploring, sniffing, running and generally having a wee of a time. They have come several times in mid chase of deer or squirrels. So, they get a double reward, a treat and a life reward (returning to what they want to do)

Just a couple of days ago it happened again. A deer was in my yard and Lyric was the only dog outside. He wasn't where I could see him and then he appeared, saw the deer and was off like a shot. (didn't catch it soon enough to prevent) I called out, "Lyric, come." And he turned and came to me. He downs from a distance and when he's been in mid chase of a dog, he obeyed and I was probably a couple hundred feet away. He knows that it is not the end of the world if he leaves what he's doing and comes, that he'll get to go play again. Of course, I can't let him go chase deer but with this practice, it has worked to give me a reliable recall. It's never 100% a sure thing but it's close. And lots of times I don't have any treats on me. It doesn't matter anymore as long as I reinforce with a treat, my dogs' favorite motivator occasionally to keep the behavior going.

That's interesting about the sheep herding. Yes, I can see how and why that is absolutely necessary in training them. How in the world would someone give a collar yank to a dog who is hundreds of feet away from them? And what kind of motivation would that give a dog? "Oh yeah....this sheep herding is really fun. Ouch! another collar correction. I hate sheep!" LOL.

My dogs are far more toy motivated than food motivated. They would prefer to tug or retrieve for their reward too, but they have had to learn to work for whatever reward I'm offering at the time. For certain things I need to use food and for certain things I can't use food. I purposely train to have a dog who loves to work for either.
I find most people don't know how to USE a toy as reward, they know how to throw a toy but they don't have the timing down and use of it to reward something down at all. It seems like once they get the mechanics down of training using food, they can switch things up and use various other forms of reinforcement. I'm not sure if other people find that.

I'm probably not real good with the toy timing either, but he loves it as a reward. Example, when he retrieves the newspaper in the morning, he brings it in, hands it to me, and runs for his tug and brings it over. He knows that's his reward for doing his job, and we'll play tug for 15 seconds or so. When I train him outside, the ball is the motivator. I can get him to do a series of commands then reward him with some ball playing. At class, I keep his small tug in my pocket, and don't use it much- if he's especially good, like he does a perfect heel for the entire length of the course we are following or something like that, he'll get a few seconds of tug. When he runs thru the agility course, even if he misses something, I always reward him with the tug, though I think on that stuff, getting back in line to go again is more of a reward!

In my admittedly limited experience, I think the reverse is true. Independent dogs are usually already focused and alert, just not neccessarily on you or what you're trying to train them. I think the more independent a dog is, the more they respond to a narrative form of training, not a pure act/reward method. ie, they learn a new behavior as part of a larger activitiy. I think the easy-going, eager-to-please dogs respond very well to food because they're already just as happy to hang out around their person and oblige them - the food gives them the deeper focus they need.

This reminds me of the few dog's I've seen learning how to herd. Some trainers start their dogs off on ducks use a rake as a correction. Dog gets too close, SMACK!, dog tries to eat duck, WHACK!

I took one of my Collies for a herding test one day and the trainer brought out the rake just as he was going after the ducks, she smacked my dog directly in the face! Needless to say, he was out of there in no time flat, but I was suprised to see he didn't even give her another thought.

I watched the other dogs after him and likewise, they all got smacked a few times for getting too close and none of them seemed bothered by it.

I guess it just goes to show that if the drive is strong enough a correction, even that severe barely even registers. The dogs were still totally focused on the ducks.

Still...my dog never went back in that ring! I won't have people smacking him in face with a rake!

Back on topic....of course it's possible to train a dog without food. It's done all the time. Some because the trainer chooses not too, and some because the dog has no interest in food. Food rewards just make things easier and speed up the learning process. However. It's not a required element for training.

Thanks for all the replies, this is very interesting!

I guess my concern with food training would be that the dog would be unmotivated to sit once it was weaned off the treat reward for sitting. Is there a way to prevent it? Or is that even an issue? In my cat book, it explains how to teach your cat to come, but says you must ALWAYS give the cat a treat, every time, or it will stop coming--as cats have no natural inclination to obey and are purely in it for the treat. Is this not usually a concern with dogs because they're more of a social animal? Or for some other reason? If it did happen, how would you train the dog out of it? (I know someone with a toy dog who won't obey commands unless a treat is involved.)

Ebony generally wasn't very interested in food. (On the bright side, no food guarding!) She wasn't interested in toys either . . . Well, she enjoyed playing with them, but she wasn't interested enough for them to be a good reward. She got plenty of praise and attention around the house, but she still was thrilled with praise and a back scritch. I think she just liked pleasing people.

Actually, she had a couple weird behaviors related to food . . .

She wouldn't eat treats if she was on a walk. Instead she'd stare politely into the distance and pretend not to see the treat. If you were really persistant, she'd take the treat from you, mouth it, and drop it. But if you gave her the same treat at home, she'd gobble it up.

Also, she wouldn't eat unless we were eating. She was free-fed, but she only ate after we started eating breakfast or dinner. She would literally hover over her bowl, watching us, waiting for us to take the first bite. (No problems with free-feeding for her, BTW, she was a very slender lab-sized dog who weighed exactly 50 lbs.)

I wonder if anyone has any insights into why she did these things?

Oh wow . . . hitting a dog with a rake, yikes! I wouldn't have gone back either!

It can be an issue if you never move passed the teaching phase. Once your dog has shown that she understands the command (by responding right away multiple times and around low level distractions) then you can start to fade away the treats and only reward for speedy responses.

Personally, if the dog doesn't sit for whatever reason after I'm fading the treats out, I simply place them into a sit. Not a big deal. This tells the dog that if they sit, they might get a reward, and if they don't, I'm just going to place them anyway. Placement is NOT a correction. It's just helping the dog succeed.

I use to train without treats in the olden days...just used praise. My dogs did everything right. What I notice with using treats and toys, something out of the ordinary, something they particularily love, (not what I think they should love) is that they are much more exuberant, much snappier with their responses, more engaged with me, quicker to pick things up (because their behaviors are marked more precisely) and they have learned better how to learn. It's more fun for me too. My dogs do everything well that they have already learned without using treats all the time. But we started out with treats for all correct responses in the beginning of learning a new skill and then proceeding to intermittent reward, shaping for quicker or prettier sits and dropping the reinforcer for sloppy ones etc. Once the behavior is learned, it has become a habit to want to comply. There's hardly a thought in their heads other than doing what they've gotten use to doing.....responding to a cue because it's been reinforced amply in the past. And occasional reinforcements keep their skills fresh.

Guess what happens if a dog gets too close to a horse or some other animal that can kick them. They get HIT HARD. A true herding dog would learn to move the hell out the way, instead of totally avoiding animals for the rest of their lives. I'm sure alot of animals get hit by bigger animals all the time, but they learn to move faster next time. (I'm not justifying smacking a dog with a rake, before you guys decide to say I'm an abusive rake wacker, )