The positive reinforcement and clicker training thread

[smolmaus]

My comment about Ben was just based on a couple of clips I've seen so I may well be wrong. I also really like Connie Colfox, she doesn't have a slick SM presence but I like the output of her training and she had a great rep years ago when I lived down her way. She takes in quite a few competition horses with various issues and tries a different way with them.

I only know him from a brief stint listening to the podcast he does with Jenny Rudall and SM clips as well. He seems a decent sort of chap! Talks the talk on the welfare of his horses which is good.

I love that the first thing on Connie's training page is "does your horse feel safe".

 

[Exasperated]

i'm sorry for intruding on your thread. I thought all threads were for all posters. I am not demanding simply asking how your training would used and work on this sort of horse. Yes he is unruly and dangerous which is why I asked for the benefits of R+ on this sort of problem. It is clear that posters here dismiss the method shown in that video of pressure and release as being disliked/ unacceptable/unjustified or whatever else so I don't feel it unseasonable to ask about R+.

The method isn’t up for discussion on this thread, neither the theoretical somersaults nor the practical application to situations, any curiosity concerning the plusses and minuses is ‘judgemental’.
Leave it, just be another five minutes of your life you won’t get back.

 

[Sussexbythesea]

Part 1 to put my knowledge (or lack there of) into context.

I’ve had a long challenging journey with Oliver and I’ve learnt a lot even though I’d have preferred a smoother one.

I’m absolutely no expert in distinguishing the different types of training and definitely now use a mish-mash as (I feel) appropriate.

It’s been a long and stressful journey with Oliver. I got him in October 2023. He was fine at viewing and good to ride.

Because I didn’t have a saddle to start I thought no problem I’ll do groundwork in the meantime. I initially had what I suppose you’d call your more traditional BHSII trainer and senior coach that had also done some “natural horsemanship” trained with the TTT and Mary Wanless RWYM. I was also trying to follow some TRT that I’d started with my other more “normal” horse. By December due to his behaviour she actually said I should sell him. As it was December I didn’t think it was the right time to sell so decided to review in the new year.

I found a lot a styles of training are about dominating the horse and forcing compliance - and in many cases this works just fine but with a very dominant intelligent and sensitive horse it was imo a disaster.

This was my impression of JW who I had out in Jan 24. Interestingly there were some friends there observing who had sent him a horse previously and really rated him but they were not so keen after watching this session.

Now I only had one session when he was doing some local clinics and adding in various sessions. Maybe I got him on a bad day. He was borderline rude and didn’t really want to hear my experience or my issues but rushed us into the arena. My assumption is that he thought I was a silly middle-aged woman who was too soft.

He started the session with pretty much what you see but on a shorter line getting him to step back and move his feet etc.. But Oliver was determined not to move his feet so it took a lot of quite unappealing flicking him in the face with rope and halter. Then he asked me what else I wanted so I said put him out on a longer line as then he’ll show you what he’s capable of. I unfortunately didn’t record it but at that point he tanked off dragging JW down the arena (he ran behind him holding the rope and Oliver double-barrelled him because he was running directly on his backside with the rope going over his back - fortunately I think it must of been glancing because he managed to hold on)

At the end of the session his feedback was “you need help” yes really that’s why you are here! I didn’t see any change or acceptance in Oliver’s demeanour and whilst it was only one session I’d hoped I’d see some sort of change in his eye or body however minor. I certainly don’t have the strength that a man does to hang on to a tanking rhino so I couldn’t see that method of training working for me in this instance.

There was a big gap after this as I had him scoped and treated for ulcers and also had him tested to see if he was a rig and he wasn’t. Then the yard gave notice of closure and I decided to put him on sales livery. 3 days before going he got a displaced colon and had colic surgery so that put paid to that. After completing box and field rest we had to move. I also discovered that he couldn’t go out with my other horse as he kept going over and reversing into him and kicking him. Why?

After moving yards in July 24 I immediately came down with shingles. So it wasn’t until about September/ October 24 that I started looking into finding another trainer. Part 2 to come.

 

[NR88]

Do you think your horse thinks you are a horse? Do you think the lion thinks the zookeeper is a lion?

As Titchy Pony pointed out further up, food is a primary reinforcer, a biological imperative. The use of a bridge signal ensures the food is strongly linked to the behaviour. To cover off Exasperated's point about it not being clear that horses understand the rational, of all the training methods I use (and I use mostly p&r because I ride and handle my horses in a completely conventional way), R+ is the one where I see my horse understanding most clearly. I always see the marked behaviour repeated quickly and consistently.

I also have never had any issues with my horses getting stressed or rude about food in any way. The expectations are so clear.

Anyway, that's my last response around this. I was enjoying this thread as a way to compare our training experiences in R+ and that's what I'll be focusing on now.

As I said, I was of the same mind as Paddy's post. I did not see involving food as a good thing and did not see a correlation between equine behaviour and this method of training.

Your first two sentences have certainly given me pause for thought.

Thank you for taking the time, and energy, to make the reply.

 

[paddy555]

I agree with you on this. However I'm always interested in find out about other ways so that I can make an informed decision. I have no ill feeling towards R+ but, currently, it does not reconcile in my thoughts with how prey animals like horses learn and interact. This may change as I have changed my mind and ways many a time over the preceeding decades.

In regards to your question @paddy555 as to how I would have handled that cob. Firstly I cannot offer up an R+ option if that was the basis of your question.

I am very confident in my ability to train horses with behaviours such as this cob. I am no longer so willing to put myself into that position to help others these days because while training horses is rewarding, working with the people who created the issues is a very different challenge.

If I found myself in the shoes of SY I would have tried to find a well mannered horse to teach the owner how to handle horses correctly and safely. This would also involve hats and glove all round.

Secondly I would use a rope halter with a longer rope, but not a lunge line. Thirdly I would never work a horse on a concrete yard. Fourthly I would never put myself in such a dangerous and vulnerable position as to walk in front of an unknown, bargey horse with it directly behind me.

Ideally an arena of sorts, but a paddock would suffice, I'd put the horse on a circle and teach it to move away from pressure. I would use the spare length of rope in my non direct hand and swing/circle it to drive the horse forward and keep it away from my space. As I am typing I realise that I will not be able to describe this accurately enough as to avoid scrutiny. Basically I would teach the horse to yield from pressure, with the reward of no pressure. Using these methods I would train the horse to yield it's whole body, then isolate quarters, shoulders and ribs along with backing up. Then I have a soft horse who does not barge into my space.

Massive improvements can be made within one short session. But it does require consistency and correct positioning and timing by the handler.

Hopefully that answers your question of me as to what I would do and some examples as to why I do not class SY as a good Horseman. Having seen his dog training video, he is a not a good dog trainer either.

Sean Coleman, Jason Webb, Warwick Schiller, Joe Midgley, Michael Pearce and Richard Maxwell all have better training videos if anyone was looking for better guidance in horsemanship and handling of bargey (spoilt by their owner) horses.

I was interested in pure R+ but any discussion and ways is very helpful to discuss so thanks for replying.

It is difficult for the likes of SV or any other trainer if they are only given 3 hours, a stable, a very spoilt horse who is ruling the roost, a very weak owner and they have to produce results. SV fell foul when the problems started and was lucky not to get hurt. That would have been a step way too far for me on a first session but if he only had one session.

I would have been happy to deal with the horse but possibly on more the same lines as you. For me it would be making a suitable way from the field to stable to get him in and out safely and preferably slowly (loose) The concrete yard was dreadful as were the railings but I suppose you get what you are given. If he had lost the horse on the concrete the only way would have been food and I'm not sure I would have trusted the horse.

for me something like your para 6, lots of short training sessions and repeat, repeat. My safety would have been paramount and also the horse's and absolutely no food for me. I thought he had a lot of rope trailing on the ground. For me that was potential for him to trip if things went wrong. Then it would be working in a small area and gradually opening up larger areas over the days. I am lucky to have that choice but SV wasn't.

the video (which just came up by chance for me I had never heard of SV) did sadly show what happens with a young horse when it is not taught manners and given instruction early on.

 

[millitiger]

Interesting thread guys.

I think clicker might be something for me to look into.
I already do R+ with the boys and find it really effective but I think the clicker will help me pinpoint the right response much more clearly for them.

 

[smolmaus]

Taking this as a win for R+; Sadie has been clipped twice in her life and she stood like a rock for a little Irish clip this evening. Asked for a little reassurance by kissing my face (one of my favourite ways she communicates) and just needed a light hand on the headcollar when we got near her ears. Superstar pony!!!

She did run round the arena like a maniac after so there was definitely some stress there to get rid of!

 

[stangs]

So I was a bad clicker fan at the Horsemanship Showcase last week, and I didn't end up attending any of the clicker specific talks/demos going, but I did still watch Ben Atkinson and Emma Tytherleigh's demos in the main arena, both of whom use R+, so here are my tired ramblings for @Matafleur and anyone interested.

Ben Atkinson - the round pen challenge:
(Disclaimer being that I attended on day 2, so I'm not sure what the foundation of his work was). He used a mix of treats and R-, but he didn't have a mike on for most of it, so I couldn't tell if he was using a clicker/marker word or not, and it wasn't always clear to me from the distance when he was treating or not. He has a foam ball on a stick that he uses as a target. Again, I couldn't tell if he was initially rewarded her (the horse he was working on) for touching it with her nose, but that progressed to him touching her body with it, under the idea that "if she touches her nose to it, that’s her consenting to being touched with the object". He did also use the target-ball thing to tap on her lower leg to start training the lift of the leg/picking up feet. (As an aside, apparently, the snatchy reaction of the fore to being touched is also how he start bow training). I do wonder about whether it's wise to utilise a target in that way, but, again, I don't know if he did use R+ to train her around the target or not.

He talks a lot about "treat hygiene" - which is what it sounds like, effectively using treats in a way to encourage good manners around food. Last year, in relation to treat hygiene, he mentioned that he treats such that the horse turns their head away from him quite tightly to get the treat, but he didn't seem to do that with this mare. She also was very in his space, nose on his right hand and right pocket a lot, which makes sense, given the fact he seems to only treat out of that hand (and also carry a treat pouch on that side, though he didn't have a treat pouch with her). Admittedly, we're all biased to treat out of one side or another, but it's a good reminder that you want to vary how you treat if you don't want them getting fixated on the Hand.

He did also say that he uses fewer treats with a young horse than with an older horse, as part of treat hygiene.

And the context for this photo: by working in protected contact, and using the horse's inclination to follow him, he could bring her closer to the fence and to the crowd, neither of which she liked. He would also then treat at the end of the round pen closest to the audience so he could "reprogram" her ideas about them (one thing about Ben Atkinson is that he doesn’t seem to use much of your standard R+ terminology, but that could also be the type of crowd he was talking to.)

(I should also add that Sean Coleman also mentioned that he uses treats sometimes to train respect around feeding time but also to bring out seeking behaviour, under the principle that horses go fight/flight -> yielding -> seeking in terms of behaviour.)

Emma Tytherleigh and her trick trained/liberty trained Shetlands:

• From the bow, you can teach the lie down (just need to get the second leg under) and then from lie down, you can teach sit
• She puts a different noise on every single one of her tricks.
• To keep a horse bowing, reward on the floor and don’t let go of the treat until you’ve got the duration you want
• She uses Baileys Fibre Nuggets for treats, finds carrots to be too high value for her horses.
• To teach horse to pick something up, use a baseball cap or a chew toy. Hand something that they will mouth to them, and treat as soon as you get it back
• Her cue for a horse to break from a behaviour, like having two legs on a pedestal, is the vocal cue "break" and tapping twice on the hind

The thing that was most interesting for me from her demo was that, with liberty work, she never uses treats, and, with trick training, she only starts using treats once the behaviour is starting to be established. So she uses them not to capture the behaviour, but more to motivate the horse so they start working longer and harder as the trick develops, which is very different to the usual idea of starting with high reinforcement rates and then fading out food, but makes sense in its own way. It clearly works for her.

Ben Atkinson - Movie Madness Behind the Scenes

• Horses trained to come to a ‘mark’ for those “horse galloping over” scenes
• The mark is an old square of wood, big enough for the front feet to go on it easily, clear to see, and it makes a sound when hooves on it so the horse knows they’re on it
  • 1) Beforehand, you need the horse to be able to work at liberty (come, stay, stop, turn) to a degree that you don’t really need to concentrate on them
  • 2) Teach the horse a Spanish walk – or Spanish stomp in this horse’s case – and use it to ask them to walk onto target
    • (I mentioned this in my other post, but I'd be interested to know why you need the Spanish walk, when normal place training is just done with normal stepping onto something. My working theory is that it looks good in a movie to make it a behaviour chain of come to mark -> flashy step -> halt.)
  • 3) To get them off the target, he doesn’t use a whip cue (because it risks them learning to break position) – instead he physically moves them back off the platform, to give them a tactile ‘break’ cue
    • And if the horse breaks their position on the mark, he just sends them back and keeps walking over, only treating on the mark.
• He uses clicker for "small intricate behaviours" (which sounds to me like capturing?)
• “Destination addiction” – when a horse stops concentrating on what they’re doing now and starts concentrating on the behaviour they’ll do next. So he only releases the horse to the mark when they're settled and concentrating.
• Releasing the horse to the mark is done with an away signal where he places his hand on the horse’s cheek
• You need to train film horses to be good with tracking vehicles, which you do by:
  • Sit on back of truck, give them treats, try to get them to touch truck with nose
  • Then start asking them to follow truck on lead, as the truck moves, and then click and whistle for stop (the ordering of which doesn't quite make sense to me)
  • When the horse gets more confident, you can start riding them behind the truck, then let them creep up the side and drop back down, etc
  • Big disclaimer that you should only do this with a bluetooth headset in both trainer and driver's ears so they can communicate for safety reasons, and that the driver should never slam the brakes, because a car stops faster than a horse does.
• For the “jumping through sandbags” (this)
  • 1) Fill the sandbags with fluffed-up straw because foam doesn’t squish down enough, which makes it a trip hazard
  • 2) Build up slowly so the horse can start by just nudging it with his head to teach him that it’s soft
  • 3) When going for the jump, the rider maintains momentum/pace and the trainer clicks when horse goes through the jump. Horse comes to them for the reward, and goes straight out of the fence as an added benefit.

Need to get to bed now before I collapse but I hope that makes some sense.

 

[Matafleur]

So I was a bad clicker fan at the Horsemanship Showcase last week, and I didn't end up attending any of the clicker specific talks/demos going, but I did still watch Ben Atkinson and Emma Tytherleigh's demos in the main arena, both of whom use R+, so here are my tired ramblings for @Matafleur and anyone interested.

Ben Atkinson - the round pen challenge:
(Disclaimer being that I attended on day 2, so I'm not sure what the foundation of his work was). He used a mix of treats and R-, but he didn't have a mike on for most of it, so I couldn't tell if he was using a clicker/marker word or not, and it wasn't always clear to me from the distance when he was treating or not. He has a foam ball on a stick that he uses as a target. Again, I couldn't tell if he was initially rewarded her (the horse he was working on) for touching it with her nose, but that progressed to him touching her body with it, under the idea that "if she touches her nose to it, that’s her consenting to being touched with the object". He did also use the target-ball thing to tap on her lower leg to start training the lift of the leg/picking up feet. (As an aside, apparently, the snatchy reaction of the fore to being touched is also how he start bow training). I do wonder about whether it's wise to utilise a target in that way, but, again, I don't know if he did use R+ to train her around the target or not.

He talks a lot about "treat hygiene" - which is what it sounds like, effectively using treats in a way to encourage good manners around food. Last year, in relation to treat hygiene, he mentioned that he treats such that the horse turns their head away from him quite tightly to get the treat, but he didn't seem to do that with this mare. She also was very in his space, nose on his right hand and right pocket a lot, which makes sense, given the fact he seems to only treat out of that hand (and also carry a treat pouch on that side, though he didn't have a treat pouch with her). Admittedly, we're all biased to treat out of one side or another, but it's a good reminder that you want to vary how you treat if you don't want them getting fixated on the Hand.

He did also say that he uses fewer treats with a young horse than with an older horse, as part of treat hygiene.

And the context for this photo: by working in protected contact, and using the horse's inclination to follow him, he could bring her closer to the fence and to the crowd, neither of which she liked. He would also then treat at the end of the round pen closest to the audience so he could "reprogram" her ideas about them (one thing about Ben Atkinson is that he seem to use much of your standard R+ terminology, but that could also be the type of crowd he was talking to.)

(I should also add that Sean Coleman also mentioned that he uses treats sometimes to train respect around feeding time but also to bring out seeking behaviour, under the principle that horses go fight/flight -> yielding -> seeking in terms of behaviour.)

Emma Tytherleigh and her trick trained/liberty trained Shetlands:

• From the bow, you can teach the lie down (just need to get the second leg under) and then from lie down, you can teach sit
• She puts a different noise on every single one of her tricks.
• To keep a horse bowing, reward on the floor and don’t let go of the treat until you’ve got the duration you want
• She uses Baileys Fibre Nuggets for treats, finds carrots to be too high value for her horses.
• To teach horse to pick something up, use a baseball cap or a chew toy. Hand something that they will mouth to them, and treat as soon as you get it back
• Her cue for a horse to break from a behaviour, like having two legs on a pedestal, is the vocal cue "break" and tapping twice on the hind

The thing that was most interesting for me from her demo was that, with liberty work, she never uses treats, and, with trick training, she only starts using treats once the behaviour is starting to be established. So she uses them not to capture the behaviour, but more to motivate the horse so they start working longer and harder as the trick develops, which is very different to the usual idea of starting with high reinforcement rates and then fading out food, but makes sense in its own way. It clearly works for her.

Ben Atkinson - Movie Madness Behind the Scenes

• Horses trained to come to a ‘mark’ for those “horse galloping over” scenes
• The mark is an old square of wood, big enough for the front feet to go on it easily, clear to see, and it makes a sound when hooves on it so the horse knows they’re on it
  • 1) Beforehand, you need the horse to be able to work at liberty (come, stay, stop, turn) to a degree that you don’t really need to concentrate on them
  • 2) Teach the horse a Spanish walk – or Spanish stomp in this horse’s case – and use it to ask them to walk onto target
    • (I mentioned this in my other post, but I'd be interested to know why you need the Spanish walk, when normal place training is just done with normal stepping onto something. My working theory is that it looks good in a movie to make it a behaviour change of come to mark -> flashy step -> halt.)
  • 3) To get them off the target, he doesn’t use a whip cue (because it risks them learning to break position) – instead he physically moves them back off the platform, to give them a tactile ‘break’ cue
    • And if the horse breaks their position on the mark, he just sends them back and keeps walking over, only treating on the mark.
• He uses clicker for "small intricate behaviours" (which sounds to me like capturing?)
• “Destination addiction” – when a horse stops concentrating on what they’re doing now and starts concentrating on the behaviour they’ll do next. So he only releases the horse to the mark when they're settled and concentrating.
• Releasing the horse to the mark is done with an away signal where he places his hand on the horse’s cheek
• You need to train film horses to be good with tracking vehicles, which you do by:
  • Sit on back of truck, give them treats, try to get them to touch truck with nose
  • Then start asking them to follow truck on lead, as the truck moves, and then click and whistle for stop (the ordering of which doesn't quite make sense to me)
  • When the horse gets more confident, you can start riding them behind the truck, then let them creep up the side and drop back down, etc
  • Big disclaimer that you should only do this with a bluetooth headset in both trainer and driver's ears so they can communicate for safety reasons, and that the driver should never slam the brakes, because a car stops faster than a horse does.
• For the “jumping through sandbags” (this)
  • 1) Fill the sandbags with fluffed-up straw because foam doesn’t squish down enough, which makes it a trip hazard
  • 2) Build up slowly so the horse can start by just nudging it with his head to teach him that it’s soft
  • 3) When going for the jump, the rider maintains momentum/pace and the trainer clicks when horse goes through the jump. Horse comes to them for the reward, and goes straight out of the fence as an added benefit.

Need to get to bed now before I collapse but I hope that makes some sense.

Wow! That is a great write up, thank you so much for making the effort. My new horse arrived today (on my birthday!) so I am re-motivated again and did a few minutes of clicker with him today to give him the idea.

Really interesting to read the BAM write up, it sounds like he mixes methods a fair bit.

 

[smolmaus]

@stangs u legend! Fab write up!!

 

[Caol Ila]

Cheers for the write up. Film horse training always fascinates me. In another life, I wish I’d found a way into that.

 

[BBP2]

I feel like Ben Atkinson is a bit like me (well, nothing like me as he is actually competent!) in that he mixes methods and does what works for him, rather than being a purist about the science of any particular method of training. He’s gone through phases of using food rewards, then never using them other than as just a treat, and now is back to using them for training again. And because he’s competent, it does work for him, because he is probably quite consistent in how he applies things, even they aren’t done by the book. For more average mortals like myself it’s easy to get in a bit of a muddle by lacking consistency in the application when dabbling between different methods.

As a random aside, I was in a bit of a pickle with my young Connemara and working out how to try more R+ type work, as he is a bit like trying to feed treats to an elephant seal. His big old lips tend to engulf your whole hand, there is no delicacy to how he receives them, even if he is trying to be gentle. So I bought a silicon dog bowl with edges that curve round and over to create a big rim. I have attached that with some twine to my treat pouch full of chaff. So now I shove a big handful of chaff into it and most of that ends up under the rim. So each time I offer it to him I can shake a bit of chaff free and he can get that from the middle, and I don’t have to top it up every time as the rest stays under the edges. It’s working really well, and means I can offer low value reward without losing a finger so we both stay relaxed.

 

[soloequestrian]

(Caveat: I haven't read this whole thread so I'm sorry if this has already been discussed and happy to be directed to it if so!)
I was recently watching sea lion training. The animals are in a zoo and do performances once or twice a day. They use a clicker type system. The trainers were very clear that they only use R+ - the sea lions choose whether or not they want to perform and can leave at any point. The younger animals are still being habituated to the performance arena and they are basically left to explore and rewarded when they go further towards the arena than they have previously.
I didn't get a chance to discuss with the trainers so I'd like your view on this. I think that pure R+ is not functional in the 'real world'. The sea lions live in a totally sealed environment so if they choose to leave one area they can't actually just wander off. The trainers didn't say anything about what happens when a pup decides to rough house with a trainer or try to bite - this species are classed at the same 'danger level' as the other big carnivores. I think if we tried to use pure R+ with an otherwise untrained horse we would end up with a thug. There are some behaviours that are dangerous, and also there are some things that we ask horses to do that perhaps they would prefer not to at that exact moment but are not uncomfortable for them (for instance do work instead of stay in the field grazing) which seem fair to me as a form of the horse earning it's living in the same way we all have to. We wouldn't allow children do only ever do exactly what they wanted and I don't think it's any different for horses (or dogs, cats etc.).
I use clicker training with my mare and I really like that it focuses me on the R+ aspect of training, but I do use all the others (R-, P+, P-) as I feel the situation demands. She can still be slightly thuggish because I haven't got it totally right - I've had her since she was a baby. Other horses who I've had, mainly rehabs, all adult on arrival, have been much easier than her and I think part of that is that they have learned to be wary of humans and so are much more respectful. I could tell that this had happened via P+ because of their various reactions. I wonder what the best way is to get that same level of respect without the need to actually scare the horse...
Sorry, several discussion points there, views on all very welcome!

 

[smolmaus]

I wonder what the best way is to get that same level of respect without the need to actually scare the horse...
Sorry, several discussion points there, views on all very welcome!

Well, it isn't "respect" it's a conditioned response. Horses don't have the brain to understand what we think of as respect. Personally I'll very happily give up a portion of perfect behaviour to not have a horse that is wary of me. Nothing makes me sadder. I will also say tho a horse with met needs is going to be better behaved. When I turn up to a "naughty" horse, 90% of the time I actually have a stressed one, or a hungry one or a sore one. "Poor behaviour" is communication.

I think you can probably use "purely" R+ as a deliberate training system, as in that is what you use when you are intentionally teaching a behaviour but it isn't possible for a horse to live an R+ only life. The environment itself will provide punishments and pressures without our input. If a horse steps on your foot and you yell and shove them off it doesn't mean you're suddenly thrown out of the R+ club. The difference is if you turn around and whack them for a mistake, or if you just be more careful about where everyone's feet are in the future. There is some discussion about this on previous pages I think but getting purist about it gets a bit messy.

 

[HaplessHorse]

(Caveat: I haven't read this whole thread so I'm sorry if this has already been discussed and happy to be directed to it if so!)
I was recently watching sea lion training. The animals are in a zoo and do performances once or twice a day. They use a clicker type system. The trainers were very clear that they only use R+ - the sea lions choose whether or not they want to perform and can leave at any point. The younger animals are still being habituated to the performance arena and they are basically left to explore and rewarded when they go further towards the arena than they have previously.
I didn't get a chance to discuss with the trainers so I'd like your view on this. I think that pure R+ is not functional in the 'real world'. The sea lions live in a totally sealed environment so if they choose to leave one area they can't actually just wander off. The trainers didn't say anything about what happens when a pup decides to rough house with a trainer or try to bite - this species are classed at the same 'danger level' as the other big carnivores. I think if we tried to use pure R+ with an otherwise untrained horse we would end up with a thug. There are some behaviours that are dangerous, and also there are some things that we ask horses to do that perhaps they would prefer not to at that exact moment but are not uncomfortable for them (for instance do work instead of stay in the field grazing) which seem fair to me as a form of the horse earning it's living in the same way we all have to. We wouldn't allow children do only ever do exactly what they wanted and I don't think it's any different for horses (or dogs, cats etc.).
I use clicker training with my mare and I really like that it focuses me on the R+ aspect of training, but I do use all the others (R-, P+, P-) as I feel the situation demands. She can still be slightly thuggish because I haven't got it totally right - I've had her since she was a baby. Other horses who I've had, mainly rehabs, all adult on arrival, have been much easier than her and I think part of that is that they have learned to be wary of humans and so are much more respectful. I could tell that this had happened via P+ because of their various reactions. I wonder what the best way is to get that same level of respect without the need to actually scare the horse...
Sorry, several discussion points there, views on all very welcome!

Ooh, fun topic that I’m going to love delving into!

I think that +R has varying levels of usefulness in the real world. I use it all of the time in groundwork to teach tricks, bombproofing, and general good behaviour with Erin. That’s a controlled environment, where I can control variables, predict responses, and make it as easy as possible for Erin to find the “correct” response to avoid frustration. Which is why I generally use designated “groundwork sessions” in a controlled environment for teaching new behaviours. Until she’s entirely confident, I don’t try and take that outside the controlled environment.

Once she is confident in a cue-behaviour-reward path, I can apply it in a real-world setting. For example, an in-hand walk where we encounter a Very Suspicious Rock in a Puddle, I can diffuse the situation by asking her to reinback and refocus, which we’re 100% on. Then I can use our spook-busting behavioural pattern to introduce her to the rock, and then I can give her a big reward, a scratch, and a breather next to the rock. At future points, the rock becomes a rest stop, which turns the rock itself into a reward (or release, -R). Another point I’d like to make, is that one doesn’t always have treats in the real world, but that’s not necessarily a bad thing. That introduces an element of variable ratio reward, where the behaviour grants an uncertain reward, which makes it exciting and addictive, effectively controlled equine gambling .

I also think that earning the respect of your equine doesn’t always need force (I am not a total fluffy fluff bunny hugger, and will slap a horse on the belly if genuine dangerous behaviours are displayed, eg. biting to push a boundary*). When Erin arrived, she was extremely bolshy, and used to getting her way with food whenever she wanted. But I worked with +R and trained her to stand politely and walk politely. I’m not afraid of a harsh “no”, and a gentle tug on the leadrope, but the main route through which I garnered Erin’s respect was via +R and -R (polite standing = stop work, and bolshy behaviour = reinback). I also behave like a passively dominant herd leader, in that I will push her out of my space (initially by asking her to reinback, but I can now simply step purposefully towards her with a low head angle, and she will back off. This is very useful for handling Saus, who is the bottom of the herd, and Erin will respect her space when I’m near Saus).

These passive-dominant behaviours on my part are useful for giving Erin confidence as well, so not only does it build respect for my boundaries, it also helps her feel safe and confident in a strong (but kind, and sympathetic) leader. So it absolutely is possible to garner respect without physical force.

My trainer advocates for developing a “yes reflex” in horses, where the horse respects and trusts in you enough that they are willing to give things a try, even if they’re not sure. It’s also important to know when the no might be coming, to avoid even getting to that point, for example, if you see your horse getting bored, you can move onto something else. However, there are absolutely moments where I still get a “no”, in which case, Erin isn’t getting out of work, she’s doing something else of a similar level. For instance, an in-hand hack with groundwork exercises, instead of a ridden hack.

*horses absolutely do have the capacity to push boundaries, it is a natural herd behaviour to see if the leader is still in top form, and would’ve been advantageous in the wild. For example, trying to sneak over and take some of the leader’s food, or giving them a nip. The leader will retaliate in some way, either with a glare or a return nip. The important thing is to see the difference between accident/stress/pain and boundary pushing. With Erin, it’s quite easy to tell, as she has a clear stress face, and a clear “hmm, can I..?” face. If it’s accidental, she usually hasn’t noticed what’s going on around her, and is probably semi-asleep.

Well, it isn't "respect" it's a conditioned response. Horses don't have the brain to understand what we think of as respect.

I think that respect boils down to a “yes reflex” in the end, so it is conditioned, but it’s also conditioned in a natural herd setting. It’s a learned behaviour in any sense, for instance, horses that are deprived of social interaction at a young age with other equines are more likely to be “bolshy” with humans, as well as prickly in a herd, due to a lack of conditioning. So I think that learning “respect” within a herd is a natural process, not a purely human process, although we undoubtedly manipulate these processes in our training.

And I think that a horses’ “respect” for a human leader mimics their respect for a herd leader. For instance, if a herd leader has a tendency to bully a horse, that horse will become avoidant of the leader, and will do nothing to cross the leader. That’s the same as if a horse interacts with a nasty human, they will be avoidant, and they will not cross the human, which is interpreted as “respect” and “good behaviour”, even if it comes from a place of fear.

I believe that a mutual horse-human respect and communication of boundaries is far more beneficial to both parties and leads to better, safer, and happier long-term outcomes.

I’m doing an EPQ on this subject and find it really interesting, sorry about the huge essay! I think I’ll probably use aspects of these arguments in my EPQ

A custard cream for anyone who makes it this far!

(Edits: spelling and slight changes to wording)

 

[stangs]

Agree with the above but also (and I’ll preface by saying I’m not a zookeeper or anything - I just have a bit of a fixation on the history of orca training.)

I think that pure R+ is not functional in the 'real world'.

I think it’s perfectly possible to train a horse only with R+ if you only work with them in a secure space, like an arena or a field, which is technically the real world. It’s also perfectly possibly to use R+ to help set the horse up for the world outside that space, e.g., by helping them learn to cope with stress or using R+ to train behaviours that will you manage your horse safely when outside the arena.

But, realistically, if you’re taking them outside secure spaces, then at some point you may end up having to use P+ or R- to ensure the safety of the horse, yourself, or any passersby. Because ultimately, as much as we aim to set the horse up to succeed, we can’t predict what comes round the corner.

(I do also think there are a lot of ‘positive only’ trainers - especially in the dog world - that do not only use R+, but don’t realise how much R- or P- they’re using. But that’s a conversation for another day, and, unless you want to be very purist about your training, I don’t think it particularly matters.)

The trainers didn't say anything about what happens when a pup decides to rough house with a trainer or try to bite - this species are classed at the same 'danger level' as the other big carnivores.

Generally speaking, I would think they would use a least reinforcing scenario (LRS), which is effectively just the handler going still for a few seconds (with SeaWorld’s orcas, it involves eye contact as well I believe) so the animal doesn’t get any sort of reinforcement for their behaviour. It acts as a bit of a reset - a pause, and then you ask again, and give the animal the opportunity to receive reinforcement again.

For biting, there’s also the option of going reinforcement of incompatible or alternative behaviours.

But, realistically, a big part of working with any animal, but particularly a wild animal, is management. If the trainer doesn’t read the animal’s body language and tailor the training session accordingly, they set the animal up to fail. A good trainer manages the session such that they’re not putting the animal in the position where they might start rough housing or biting or whatever.

And realistically, by the time the animal’s behaviour is escalating, you’re definitely not going to use P+. Not just because you don’t want to damage the relationshp between you and a wild animal whose trust is much more easily lost than that of a pet’s, but also because using P+ may mean the animal’s behaviour becomes more dangerous.

If you read up accounts of trainer-directed orca aggression, once the behaviour has escalated, the trainer’s focus is solely on getting everyone to safety, e.g., by trying to recall the orca to a different side of the pool, throwing fish, or bringing another orca in. Once everyone’s safe, they study the session, figure out what went wrong, and plan what their next steps will be. What the absolutely don’t do is punish the animal for the behaviour - can you imagine how dangerous that would be?

I think if we tried to use pure R+ with an otherwise untrained horse we would end up with a thug.

Animals go where the reinforcer is. If you ensure treat placement is always away from your body, the animal never has a reason to get in your space.

If you make sure to manage training sessions such that frustration is kept to a minimum, why would a horse become a thug just because it’s trained with R+ ?

We wouldn't allow children do only ever do exactly what they wanted and I don't think it's any different for horses (or dogs, cats etc.).

But using R+ doesn’t mean that the animal is enjoying their work. Captive orca aggression is probably the best example of this, but I can put my hand up and say my boy’s not enjoyed all the training sessions we’ve done together. When I was working with him to be able to inspect his mouth, he let me do it at liberty, he took rewards, but the stress was very visible on his face. I kept sessions very short, and alternated with a lot of easy and highly reinforced work, to ensure he didn’t start shutting down. But he very easily could have - and again, that was all R+.

 

[smolmaus]

And I think that a horses’ “respect” for a human leader mimics their respect for a herd leader. For instance, if a herd leader has a tendency to bully a horse, that horse will become avoidant of the leader, and will do nothing to cross the leader. That’s the same as if a horse interacts with a nasty human, they will be avoidant, and they will not cross the human, which is interpreted as “respect” and “good behaviour”, even if it comes from a place of fear.

I believe that a mutual horse-human respect and communication of boundaries is far more beneficial to both parties and leads to better, safer, and happier long-term outcomes.

This is where it gets muddy, as the word "respect" is, even between humans, a subjective and contextually/culturally changing idea.

What I see when people talk about "respect" is compliance. Horse knows what to do, how to move, where to be, what to accept. Conversely, "disrespect" is usually stress behaviour or simple non-compliance. A "bolshy" horse is usually stressed, again in my experience. Having clearly communicated boundaries on where you (the horse) should be, how to move etc so as to keep the human happy will be easier for the horse and reduce stress in the long term, so of course it isn't a bad thing, but that doesn't make them "respectful". As you say it's learned behaviour. That goes for interactions between horses too. Actions produce consequences, this is how you act to have an easy life and access to the herd. Clear communication, clear boundaries, clear expectations.

I'm not saying you are doing this, I don't think you are, but talking about respect implies a mutually understood hierarchy that isn't there and starts to involve human ego. Being disrespectful to another human for example, is a deliberate choice with intent to make someone feel disrespected. A horse can't make that choice as they don't know what respect is. But if someone is thinking about respect and being disrespected, they attach a motive to the horse that isn't there, are more likely to decide it's "on purpose" and are then more likely to turn to punishment for the disrespect, rather than looking for where the misunderstanding came from.

Its a linguistic problem really, and as you can see, this is one of my own personal bug bears so thats MY essay. But I've seen the exact scenario play out many many times and the result is rarely fair to the horse.

 

[soloequestrian]

Lots of interesting points!
I've thought about orcas a lot over the years and I'm not sure it's fair to compare them to anything else in terms of training. I have a strong suspicion that any orca in a captive environment is psychologically damaged. Their life history absolutely does not fit with any ability we have to keep them in a high welfare state. They can also kill something the size of a human quickly and efficiently and they have the instinct to do that.
In terms of respect, I think this is semantics. Horses definitely have the ability to understand that they behave differently around different individuals, both other horses and humans. I use 'respect' as a way of saying 'dominance without (necessarily) force'. I want to be the equivalent of the horse that is just slightly above any of the others in the pecking order, otherwise things get dangerous. That means that some behaviours have to be met with P+, just like they would be by another horse (or a mummy sea lion). I don't think just pausing/ignoring unacceptable behaviours is a good way forwards - for instance biting, kicking, barging, all of which would be met with retaliation from a more dominant horse.
The 'thug' thing also fits with this. My mare is lovely to work with when I have her attention but she is also really inattentive - she finds all sorts of things about the environment far more interesting that me, even me with treats. If something exciting/anxiety inducing happens, she will pay attention to that and sometimes that means knocking into me - she briefly forgets I exist. This may well be a specific personality/upbringing thing but none of my previous horses were as extreme. All of them could manage to continue to respect my space even if something exciting was going on that they felt they had to respond to. This is what I mean by 'wary' - not that they weren't relaxed around me (and other people), just that they were always aware of me even when under stress. Perhaps that's more to do with the lack of an older mare when my horse was growing up though.

 

[smolmaus]

There is a case to be made that domesticated horses are also in a psychologically damaging captive environment. I think a lot about the "alpha wolf" theory of dominance which has been applied to wolves, dogs, adult humans, human children, cats, horses, all sorts of social groups when it never even applied to wolves in the first place. There is no Alpha Wolf and there is increasing evidence that there is no Boss Mare or strict hierarchical pecking order with horses either. I like this quick summary from Debbie Busby.

 

[soloequestrian]

For me the difference between most animals and orcas is that we just can't keep orcas in captivity in a way that allows them to have good welfare. We can, I believe, keep the majority of other species in a way that doesn't damage them, we just don't always do that. Horses need space, forage and company and many of them don't get that.
In terms of hierarchy, my interpretation of the new ideas is that they are the old ones with slight changes, not radically new. My own little herd has a very obvious pecking order but who is at the top changes according to motivation - that just means that if the older mare really wanted to she could come in first to be fed but she's not that bothered so the next in line always comes in first. Horses that know each other really well can look like good friends, and might well be, but there will always be one who is in charge. In horse human relationships the one in charge has to be the human.

 

[little_critter]

Here’s a training question for you.
My horse is good to be bridled, his quirk is that I can’t put the bit in in the usual way (cradled in hand and placing in mouth). If I try that he fusses and head tosses.
Our method is that he has to grad the bit, I hold everything else out of the way and hang the bit under his mouth, he picks it up no bother.
Our difficulty is he’s keen to please and will grab and bit of bridle that come near him, getting the reins over his head is a challenge because he intercepts them.
I don’t get annoyed at him, just do my best to guide him to the bit.

Any suggestions on how to encourage him to only aim for the bit and not just grab anything?

 

[DabDab]

I can't remember the last time I put a bit in a horse's mouth - all mine just take it when I dangle it in front of them. If they ever start preemptively grabbing leatherwork it is just a case of slowing it all down - put reins over head, take them back off, generally faff around with the bridle by them, if they grab something just gentle take it back out of their mouth and then only give a treat when they take the bit calmly when it is held up in front of them.

 

[smolmaus]

Here’s a training question for you.
My horse is good to be bridled, his quirk is that I can’t put the bit in in the usual way (cradled in hand and placing in mouth). If I try that he fusses and head tosses.
Our method is that he has to grad the bit, I hold everything else out of the way and hang the bit under his mouth, he picks it up no bother.
Our difficulty is he’s keen to please and will grab and bit of bridle that come near him, getting the reins over his head is a challenge because he intercepts them.
I don’t get annoyed at him, just do my best to guide him to the bit.

Any suggestions on how to encourage him to only aim for the bit and not just grab anything?

I have a cue for reins over head before we get to the bit. But that's more to get her head out of her haynet tbh. Could cue a "head down"? A wait cue is handy for everyday use but kinda hard to teach I think as you don't want it too context specific?

You could teach a target on the bit specifically? So nothing else in the mouth gets reinforcement. I suppose if he's already too keen to get the bit in his mouth you dont have to worry about making him too keen