Positive Reinforcement

G

GinnyS

New User
Joined
29 August 2012
Messages
4
Visit site
I have a college question that I would be grateful for any views/experiences/opinions on please: 'Can you think of any ways that positive reinforcement rather than negative reinforcement could be used when training a young horse for riding?'
I understand that clicker training is based on PR but I've never seen it used for starting horses or used for ridden work so I have no experience to draw on with this one! Thanks very much in advance.
 
Positive Reinforcement is the basis of all the training I do I rarely need to use negative .
Right from when I start with them I vioce train them that good boy or good girl means they have have it right , I also use the pat( or stroke ) on the neck usually with the inside hand , a period of hard work will be rewarded with a the pat or stroking and a period of free walk with the voice .
So when young if use the leg and they respond I stop the "aid " say good boy that's positive reinforcement in my book.
 
You need to be careful with the definitions for a start! PR and NR mean different things to different people -- apparently!

Isn't simply letting an animal know that it has "done the right thing" PR? I'm sure there are plenty of similar examples. A slice of carrot for a reward? A pat? Scratch?

In which case, you'd first need to condition the animal to gain pleasure from your praise of "Good boy!"

But "Good boy!" means nothing to a horse trained by, say, a Dutchman speaking his native language. That horse would have to be conditioning into knowing what the Dutch equivalent of the praise "Good boy!" means.

Does that make sense? I am sure you will get some interesting replies to this question as I am still guessing.
 
I think teaching a horse to stand at the mounting block can be very well done by positive reinforcement and it's very useful when training a youngster. It makes the mounting block a good place to be.
 
A few years ago when my mare was on box rest I introduced clicker training for fun just to keep her occupied and she took to it like a duck to water. I also think a scratch and a soft tone of your voice is PR and when I am riding I eg say I ask for halt from trot and she does this, then I give with the reins PR. I'm sure someone will be able to describe it better.
 
Pale Rider said:
Just remeber that Positive does't mean good or Negative bad in this context.
Click to expand...
Indeed. The alternatives "additive" (for "good") and "subtractive" (for "bad) are arguably more descriptive and less emotive. Unfortunately they aren't standard terminology (yet!).

Similarly, it has been suggested that "anti-reinforcement" would be preferable to the emotive "punishment".
 
fburton said:
Indeed. The alternatives "additive" (for "good") and "subtractive" (for "bad) are arguably more descriptive and less emotive. Unfortunately they aren't standard terminology (yet!).

Similarly, it has been suggested that "anti-reinforcement" would be preferable to the emotive "punishment".
Click to expand...
Wont change the arguments and vendettas though sadly, which are a huge block to learning imo.
 
Except where clicker training is used, most training of the young horse is based on negative (i.e. subtractive) reinforcement. Pressure is applied and then released (subtracted) when the horse does what is wanted. True rewards may be given too, of course, but usually in a non-specific way e.g. after a lesson or part of lesson rather than to reinforce specific behaviours.
 
Positive reinforcement is basically training with a positive reward. True positive reinforcement is waiting for a horse to show a behaviour you want, then rewarding with food or praise or other positive stimulus. So you might stand with a horse and say nothing, but when it stands still for instance, you click or give a treat and repeat.
Most horse training is done with negative reinforcement which is NOT the same as punishment. The negative wording just means rewarding by REMOVING a stimulis. So you put your leg on, when the horse moves, you take your leg off so you are rewarding the horse by removing your leg, not by giving a treat etc. This is the same if you ask/ tell/aid/restrain etc. You stop your intervention when the hoses does as you ask. This is negative reinforcement, not the same as punishmnet.
 
amandap said:
Wont change the arguments and vendettas though sadly, which are a huge block to learning imo.
Click to expand...
Maybe not in the short term, but my impression is that a lot of the bad feeling comes from misunderstanding or miscommunication, and that the different sides in fact agree about more than they disagree about.
 
Frisbee123 said:
True positive reinforcement is waiting for a horse to show a behaviour you want, then rewarding with food or praise or other positive stimulus.
Click to expand...
I would like to add that it is often possible to get to the behaviour you want using previously learned cues or with 'lures', so one doesn't always have to do nothing while waiting.
 
The key to either positive or negative training is consistency.

Personally, I love clicker as it is an easy way for the 'human' to integrate positive training effectively, but its no use at all unless there is a lot of consistency in the training.

With my horse, I find using positive and negative together can be best in some circumstances. For example, my horse has just come back from a riding school, where she has got very dead to the leg - I would like her to go off a light commend. So, when I give the aid if she goes off nicely, she gets a click, if she doesn't, a couple of taps with the whip - but the timing of either is key!!!
 
As above make sure you are clear about your understanding of positive/negative reinforcement. From a psychology perspective the term 'reinforcement' means something that will increase the likelihood of a particular behaviour occuring and 'punishment' is something that reduces the likelihood of a behaviour occurring. These can both be either positive or negative depending on whether something (usually a positive experience) is added (positive)or taken away (something adversive) (negative). E.g. If you give your horse a carrot when they come over when they call you this is positive reinforcement as you added a carrot and the likelihood of them coming over again in the future is increased. The original experiments were done on rats in boxes and for negative reinforcement an example would be if the rats were given shocks that stopped when they pushed a lever, therefore the action of pushing a lever would be increased (i.e. reinforcement) and the adversive stimuli (electric shock) taken away (negative). For punishment a positive punishment could be giving the horse a carrot when it stand still at the mounting block as this reduces the behaviour of moving around by giving a 'reward', negative punishment would be punishment as we would normally understand it e.g smacking a horse if it bites us to reduce that behaviour happening again.

Hope this isn't too confusing, but could be v important assuming your tutors understand the definitions correctly themselves!!
 
Interesting thinking about this when I backed horses they where voice trained first on the lunge then when I backed them I just combined the voice and leg aids then gradually reduced the use of the voice .
When training driving horse we train them to stand loose on the yard we give a voice command and leave them when they move we put them back to where we left them and repeat until they get it not sure what I am doing PR or NR , staggers out of room confused .
 
NikNak1 said:
These can both be either positive or negative depending on whether something (usually a positive experience) is added (positive)or taken away (something adversive) (negative).
Click to expand...
In my understanding (ok somewhat limited lol) this is not correct and you are adding emotive phrasing to something that doesn't (shouldn't) contain emotion ie. learning theory.
Hitting a horse for biting you for eg (and providing the timing is correct) is positive punishment. The hit is added/applied.

ps. I am referring to the bit I put in bold in your context.
 
Oh Lordy! :eek:

Sorry to chip in (I swore I wouldn't) but NikNak in trying to be helpful has unfortunately given a completely wrong definition! I know it's complex ( I teach this along with the rest of undergraduate psychology, and this is the part, along with statistics, that students find hardest to grasp).

OK - you have two things to start with - things the horse want, and things the horse doesn't want. They will try to get the things they want, and they will try to get away from the things they don't. Hope that bit is clear to begin with :)

Right - in any given situation, you can add and subtract things - hence positive (add) and negative (subtract). So you can add something the horse wants, and you can add something the horse doesn't want. Example - you can add carrots :) or you can add annoying tapping with your heels on his side.

You can also subtract stuff. You have carrots, the horse knows you have carrots, but you take them out of reach. Or... you have spurs, you are tapping the horse's sides - but you stop.

Now let's look at what that does to behaviour :)

The horse comes and stands quietly next to the mounting block. You give them a carrot (positive (additive) reinforcement). Next time you approach the mounting block, the horse will be keen to do exactly the same thing in the hope of another carrot - you have reinforced standing quietly next to the block.

OR... the horse comes and stands a bit away from the mounting block. You tap the far side their hindquarters annoyingly with the whip, until they step away from the annoyance to line up with the block - at which point you stop the annoying tapping (you have removed/subtracted something the horse doesn't like). You have reinforced standing next to the block (but the horse might be a bit less keen to approach the block next time).

These are examples of positive and negative reinforcement.

Now punishment - reinforcement makes behaviours more likely to happen again, punishment makes them less likely.

Yet again, we can add and subtract the punishers.
So - we have the horse's dinner in our hand, but he is banging the door with his hoof creating a big racket. We walk away, with the dinner. We just subtracted something the horse wanted, and (we hope) punished the banging (although, warning - this subtractive or negative punishment can cause increased frustration.

We are also standing next to the horse with our whip when he starts to kick the door - we wallop the leg he's kicking with. Yay! He's stopped kicking! Except, darn, now he's just turned his back on us and is threatening to double barrel us and the door (we added something (positive punishment) that the horse doesn't like to reduce a behaviour - but we learned that punishment comes with additional bad stuff - and the horse remembers next time and is suddenly very difficult to catch and headcollar: punishment has "fallout".

In terms of training - asking the horse to do something in a schooling session, then allowing them to stop while praising them is technically negative reinforcement of a behaviour - you added a pressure, until they did what you wanted, then you released the pressure. The praise is simply a signal to the horse that there won't be any pressure for a minute.

Hope that is a little helpful, and hope it gives some training examples :)
 
Thanks so much for all your comments, you've all basically summed up what I've been trying to get straight in my head! I wasn't sure if it was possible to start a horse for riding totally using PR in its true definition and to be honest I'm still not quite there yet!
This is the last question in my current module of my equine psychology course and its definately proving a tricky one to get just right!
 
GinnyS said:
Thanks so much for all your comments, you've all basically summed up what I've been trying to get straight in my head! I wasn't sure if it was possible to start a horse for riding totally using PR in its true definition and to be honest I'm still not quite there yet!
This is the last question in my current module of my equine psychology course and its definately proving a tricky one to get just right!
Click to expand...

That's a clear question and one I've asked before. So far I haven't been able to find anyone who has started a horse for riding TOTALLY using PR in it's true definition. As Fburton says, most young horse training is based on NR. Some people reward NR with clicker or other treats, but it isn't totally PR then.

Edited to say - personally I don't have a problem with that.
 
Last edited:
That's a clear question and one I've asked before. So far I haven't been able to find anyone who has started a horse for riding TOTALLY using PR in it's true definition. As Fburton says, most young horse training is based on NR. Some people reward NR with clicker or other treats, but it isn't totally PR then.
Click to expand...

Life's like that :)

You want to wear your wonderful new shoes to the office Christmas party :) You anticipate lots of astonished compliments... "They're so shiny!", "They make your legs look great!", "What great taste in shoes you have". You will feel positively reinforced for wearing them, and so likely to wear them again :)

But after 2 hours of standing around in them drinking wine, you find that even the wine is not dulling the ache starting in your toes. You wish you could throw the darn shoes in the bin, and really, you would like never to see them again. You are being positively punished for wearing the shoes :D

Eventually, when nobody is looking, you slip them off. What a relief! The horrible pressure on your toes is gone (although they are still throbbing a bit). You have now negatively reinforced the action of taking off the shoes ;)

But, oh no! You have to go home, and this means walking in the wet to your train, so you need shoes on. Putting them back on is agony as your feet have swelled. You hobble to the train, get home, and kick the shoes off, thinking you would be quite happy never to see the darn things again.

But a few weeks later, there is another party - and you remember that everybody complimented you on your shoes (but that they hurt).

Would anybody like to describe the potential conflict behaviours that are likely to arise due to this combination of positive and negative reinforcement and punishment associated with party shoes? :D (btw, I suspect the "taking off the shoes" behaviour is likely to occur more often and faster in the future...)

We don't live (or train horses) in a vacuum. In life, there is a constant pull towards things we like (and the same for our horses) - but at the same time, there is a constant drive to avoid things we don't. Good training acknowledges the fact that we can only ever control a small part of the training environment, but still builds upon it :)
 
Brightbay said:
Life's like that :)

You want to wear your wonderful new shoes to the office Christmas party :) You anticipate lots of astonished compliments... "They're so shiny!", "They make your legs look great!", "What great taste in shoes you have". You will feel positively reinforced for wearing them, and so likely to wear them again :)

But after 2 hours of standing around in them drinking wine, you find that even the wine is not dulling the ache starting in your toes. You wish you could throw the darn shoes in the bin, and really, you would like never to see them again. You are being positively punished for wearing the shoes :D

Eventually, when nobody is looking, you slip them off. What a relief! The horrible pressure on your toes is gone (although they are still throbbing a bit). You have now negatively reinforced the action of taking off the shoes ;)

But, oh no! You have to go home, and this means walking in the wet to your train, so you need shoes on. Putting them back on is agony as your feet have swelled. You hobble to the train, get home, and kick the shoes off, thinking you would be quite happy never to see the darn things again.

But a few weeks later, there is another party - and you remember that everybody complimented you on your shoes (but that they hurt).

Would anybody like to describe the potential conflict behaviours that are likely to arise due to this combination of positive and negative reinforcement and punishment associated with party shoes? :D (btw, I suspect the "taking off the shoes" behaviour is likely to occur more often and faster in the future...)

We don't live (or train horses) in a vacuum. In life, there is a constant pull towards things we like (and the same for our horses) - but at the same time, there is a constant drive to avoid things we don't. Good training acknowledges the fact that we can only ever control a small part of the training environment, but still builds upon it :)
Click to expand...

Cracking post, :D
 
GinnyS said:
Thanks so much for all your comments, you've all basically summed up what I've been trying to get straight in my head! I wasn't sure if it was possible to start a horse for riding totally using PR in its true definition and to be honest I'm still not quite there yet!
This is the last question in my current module of my equine psychology course and its definately proving a tricky one to get just right!
Click to expand...

I think the trouble is Ginny that it isn't possible to teach a horse NOT to do something except by negative reinforcement. You can use positive reinforcement to strengthen an alternative behaviour that you hope the horse will choose over the behaviour that he wants to do.

But you can't actually dissuade him from doing the thing that you don't want him to do without a negative reinforcement.
 
Brightbay said:
Life's like that :)

You want to wear your wonderful new shoes to the office Christmas party :) You anticipate lots of astonished compliments... "They're so shiny!", "They make your legs look great!", "What great taste in shoes you have". You will feel positively reinforced for wearing them, and so likely to wear them again :)

But after 2 hours of standing around in them drinking wine, you find that even the wine is not dulling the ache starting in your toes. You wish you could throw the darn shoes in the bin, and really, you would like never to see them again. You are being positively punished for wearing the shoes :D

Eventually, when nobody is looking, you slip them off. What a relief! The horrible pressure on your toes is gone (although they are still throbbing a bit). You have now negatively reinforced the action of taking off the shoes ;)

But, oh no! You have to go home, and this means walking in the wet to your train, so you need shoes on. Putting them back on is agony as your feet have swelled. You hobble to the train, get home, and kick the shoes off, thinking you would be quite happy never to see the darn things again.

But a few weeks later, there is another party - and you remember that everybody complimented you on your shoes (but that they hurt).

Would anybody like to describe the potential conflict behaviours that are likely to arise due to this combination of positive and negative reinforcement and punishment associated with party shoes? :D (btw, I suspect the "taking off the shoes" behaviour is likely to occur more often and faster in the future...)

We don't live (or train horses) in a vacuum. In life, there is a constant pull towards things we like (and the same for our horses) - but at the same time, there is a constant drive to avoid things we don't. Good training acknowledges the fact that we can only ever control a small part of the training environment, but still builds upon it :)
Click to expand...

I would go to the chemist and by some Party Feet.
 
ChristmasPTrees said:
I think the trouble is Ginny that it isn't possible to teach a horse NOT to do something except by negative reinforcement. You can use positive reinforcement to strengthen an alternative behaviour that you hope the horse will choose over the behaviour that he wants to do.

But you can't actually dissuade him from doing the thing that you don't want him to do without a negative reinforcement.
Click to expand...
Don't you mean punishment? Reinforcement, whether positive or negative, is always about making a behaviour more not less likely.

Assuming you are talking about punishment ;), you make a very good point about alternatives to giving the horse something it doesn't like (aka aversives) in order to teach it to not do something. By teaching a different behaviour incompatible with the one you want the horse to stop doing, you achieve your goal without having to deliver anything unpleasant, such as a smack, tug or reprimand.

Unwanted behaviours can also go away 'by themselves' (aka extinction) if they are consistently not rewarded. This is how door kicking may peter out if the horse receives no attention (e.g. shouting) or other reward like food consistently over a period of time. If it doesn't stop eventually, it means the horse must be getting some kind of reward from doing it.
 
You see? Everyone is getting bogged down with the definitions and the horse is still not trained! I knew this was going to happen!:(

Does it really matter what we call it? "A rose by any other name..." etc.

Would it not be be more productive to look at specific training problems and the various way we solve them?
 
You must log in or register to reply here.
Top