|
Post by ghostrider on Dec 29, 2010 11:51:00 GMT 1
I'm trying to keep quiet, honest! Lol I've missed Laura's winking and smilies in general btw. May I just clarify something in my (strange admittedly) head? Desensitization (without food). The horse learns that the 'thing' isn't going to hurt him. CT. The horse replaces a feeling of fear with a pleasurable feeling in the presence of the scary 'thing'? Would that be a correct very basic difference in the way behaviour is changed in the two scenarios? Or am I way over thinking this? I'd agree with that definition - and both work perfectly well. No one has to use food to desensitize. However, if you only use desensitization to overcome a scary situation, there is a chance of what is known as 'spontaneous recovery' - at some point in the future something may possibly trigger the fear again in the horse - and then you are back to square one. It's also important that the horse is not made afraid during the desensitization process - eg, if you were desensitizing for foot handling say, but before you had reached the final goal you had to have the farrier and he ended up having a tussle with the horse - you'd probably be right back to square one again - infact things may be a bit worse. When food is added into desensitization it's called counter-conditioning - that's changing the animal's emotions about the scary thing - so not just "it won't hurt me" but "oh, good, it's (whatever the scary thing is). Adding counterconditioning into desensitization tends to make the behaviour 'stronger', whereas desensitization alone can be a bit more 'fragile' - more likely to break down again. Depending on how fearful the horse is initially, some people would go through the whole desensitization process first and then do counterconditioning afterwards when the horse is showing no fear of the scary situation. Other people might combine the two, if the fear was low level to begin with. The reason being that if you combine food training with something that is too scary, the horse associates the food with the feared situation and actually becomes afraid of the food-training. Eg - I have a dog that I had done a lot of CT with, and she was very good. One day we were on a course together and I was CT'ing her and there was shooting (She is sound sensitive). It only took a couple of minutes for her to shut down completely and refuse to work at all - and even when I took her home and tried to CT her again, she couldn't do it. So I gave her a few months off, did lots of fun stuff with her and then re-introduced CT but changed from a clicker to a word and then she could manage it and is fine now. But she had associated the CT, which she loved, with the shooting which she hated, and that was the end of that. So the postive has to be stronger than the aversive for it to be effective.
|
|
|
Post by Yann on Dec 29, 2010 12:04:48 GMT 1
Ghostrider, it may well be the same one in that case. That's a fascinating training schedule, I suspect you'd be hard pushed to find many mature domestic horses who would have tolerated all the same tasks That's my understanding of it. There is a difference though if you're dealing with a phobia rather than a fear which took me a while to get my head around, in that with a phobia you're aiming to avoid triggering the irrational fear reaction at any point. So you wouldn't use approach and retreat with a clicker as such (which is fine otherwise of course so long as you're not triggering flight) but would simply present the stimulus in increments and create new associations with it in the mind of the horse.
|
|
|
Post by mandal on Dec 29, 2010 12:05:18 GMT 1
Thanks for that Ghostrider. So using both has the strongest most lasting effect? Surely though, 'good' and careful desensitization changes the horses emotion about anything scary too. OK he may not associate great pleasure with it but he has learned that is no longer anything to fear. Yes, he might fear it again in the future if something happens to make him fearful of it. I just worry that pairing purely an emotion with scary stuff to change behaviour isn't true 'learning' it's more a conditioned resonse rather than an intellectual one. :-/I'm probably using the wrong words but does anyone understand what I'm getting at? Just to give a horse a new emotional association with something can go pear shaped as easily as purely desensitizing surely? Desenstitizing followed by CT seems to be the way I would go myself.
|
|
|
Post by mandal on Dec 29, 2010 12:09:01 GMT 1
I'm a fair way behind you Yann in understanding so I'll think about phobia later! I have to look up a definition of phobia first to remind me when something is considered a phobia.
|
|
|
Post by Yann on Dec 29, 2010 12:09:48 GMT 1
Cross posted with you both. Ghostrider has explained it far better than I ever could
|
|
|
Post by ghostrider on Dec 29, 2010 12:57:46 GMT 1
Thanks for that Ghostrider. So using both has the strongest most lasting effect? Yes, I think that's the general perception. Surely though, 'good' and careful desensitization changes the horses emotion about anything scary too. OK he may not associate great pleasure with it but he has learned that is no longer anything to fear. oh, absolutely - the whole point of desensitization is teaching the horse that there is nothing to fear, and it can work very well. Yes, he might fear it again in the future if something happens to make him fearful of it. I just worry that pairing purely an emotion with scary stuff to change behaviour isn't true 'learning' it's more a conditioned resonse rather than an intellectual one. :-/I'm probably using the wrong words but does anyone understand what I'm getting at? It isn't the aim to purely pair an emotion with a scary situation though - it's what I tried to say earlier about food affecting both operant and classical conditioning - but I wasn't very clear. At an operant level the horse is learning things like "it's OK to stand still for the clippers", so the horse is learning that being clipped doesn't hurt and he can change his behaviour accordingly to allow it to happen. The horse is not only learning "oh, it's the clippers but they don't hurt" (operant learning - through consequences - I stood still and it didn't hurt) he's also learning "oh, good, the clippers - that means we play the click and treat game" (so that's classical or associative learning) (this is a bit simplistic obviously, but you know what I'm getting at ) It is at the classical level that the emotions come into it - so now in the horse's mind the positive reinforcement and the clippers become paired or associated - so the horse is happy to see the clippers becuase he knows a) they don't hurt and b) they mean rewards so he looks forward to being clipped. Now, not being afraid of being clipped is great - but in my mind to look forward to it has got to be a step further along in the right direction. So, I don't think, anyone is only ever aiming to effect either emotions or behaviour - it's just always a mix of the two, and we hope to be affecting them in the right way Just to give a horse a new emotional association with something can go pear shaped as easily as purely desensitizing surely?I'm sure it would - but I don't think anyone aims (or even could) purely change emotions without the horse learning appropriate behaviours to go with those changed emotions - so the horse is learning "intellectually" - "this is how I respond to clippers, I know they don't hurt and I can stand still for them". Desenstitizing followed by CT seems to be the way I would go myself.Yes, I think it's a safer option - although I must confess if the fear is low level I'm happy to combine the two pretty much straight off because it's a lot quicker and sometimes people just need to see results, and there is no point keeping the horse in prolonged state of 'uncertainty' (is that the right word, not sure) when you're pretty confident that things will be OK when SD and CC are combined. It's just important to remember that the positive must always be stronger than the feared stimulus - so it's about diluting the feared object or situation to a level where the horse is not afraid of it before the training starts (that may be through proximity, or some other way). But even if we go down the route of SD first, get that solid and then combine CC - we are still always going to be influencing emotions as well as behaviours. By the way, if anyone is wondering why I don't SD and CC my noise sensitive dog - it's because I live in a farming area where there is a lot of shooting - so I would never get her through the programme without her hearing gun-shots and triggering her fear again and putting her back to square one.
|
|
|
Post by mandal on Dec 29, 2010 13:23:58 GMT 1
So, I don't think, anyone is only ever aiming to effect either emotions or behaviour - it's just always a mix of the two, and we hope to be affecting them in the right way Just to give a horse a new emotional association with something can go pear shaped as easily as purely desensitizing surely?I'm sure it would - but I don't think anyone aims (or even could) purely change emotions without the horse learning appropriate behaviours to go with those changed emotions - so the horse is learning "intellectually" - "this is how I respond to clippers, I know they don't hurt and I can stand still for them". Woo hoo! I think this is where my misunderstanding comes in. I was worried the "I know they don't hurt" bit could be bypassed but it seems that isn't possible. This is where my fear of training a robot comes from I think. ;D Btw. Thanks for your patience Ghostrider. lol
|
|
|
Post by ghostrider on Dec 29, 2010 14:38:06 GMT 1
you're welcome Mandal, and I'm glad you feel a bit more comfortable with that :-)
However, I think I'm going to muddy the waters a bit for you now, apologies, but please bear with me ...
I think your concerns about training 'robots' are absolutely justified and should be taken into account. ANY training system is only as, well, I don't like to use the word 'ethical', perhaps 'true to the nature of the horse' is a more apt description, as the person who uses it. Clicker training is used to teach horses and other animals to cope with aversive situations, of that there is no doubt - it is used to teach animals in zoos to tolerate husbandry procedures and our own horses to cope with things that they may find uncomfortable, or at least not enjoy. However, I think as long as it is done carefully and systematically (as Yann has described taking a long time to teach his mare to tolerate an injection) I think it is a good and acceptable way to teach these necessities. I do think that horses can learn at both an emotional and intellectual level in these circumstances, if the work is carried out with empathy and intuition, as well as regard for all the laws of learning that we understand, and awareness of the natural behaviour and instincts of the horse. There is no doubt in my mind that the horse is 'co-operating' (perhaps being 'coerced') for the sake of the treat in these instances - however, as long as CT and positive reinforcement is used for fun and enjoyable activities for the horse as well, it is, to my mind, an acceptable 'trade-off'. I think the presence of food rewards makes it more likely the horse will 'try' (standing for the clippers, having feet picked up, having an injection or whatever) and then of course they discover it is not as bad as they had feared - so they overcome their reservations. Breaking each task down into small, small steps and marking and rewarding the try is a kind and efficient way to teach horses to do what we want them to.
However, I think horses are being trained with clicker training in ways that may be demeaning or demaging for them. I don't mean horses taught 'tricks' - I have no problem with horses taught tricks, to my mind it is all mental stimulation for the horse - but I have seen horses coerced to do things for a click and a treat that I was uncomfortable with - for example watching a well-known trainer who used to be known for Parelli and has now switched to CT have his horse stand on a high and slippery pedestal in the rain - the horse was clearly uncomfortable, and as he stepped off his back leg slipped out behind him at a very awkward angle - that did not make for pleasant viewing when this approach is meant to be about the horse enjoying the training and being an active participant.
In clicker training I think there is a danger of horses being 'micro-managed' as much as for example in Parelli style work where the horse must always give the handler 'two eyes and two ears'. In clicker training, some people actually 'shape' their horses to point their ears forward - to me that is wrong - the horse's ears are a way of monitoring his environment and expressing his emotions - we have no right to insist that he puts them forward for a click and a treat. On the covers of some very well known clicker training books there are horses that have contorted themselves into all sorts of weird shapes in a parody of 'collection' and 'self-carriage' - some ridden and some in-hand - and these are being held as examples to which others should aspire to. Other people are simply adding in a click and treat whilst still using a heavy amount of negative reinforcement. It seems that the additional tools such as targetting and freeshaping that clicker training gives are ignored by many of these trainers - and yet these are the training tools that really encourage animals to think, use their brains and work it out for themselves (I have no problem with light -R being used with +R by the way and agree that it has to be used if we are to teach horses the usual range of activities we want them to do).
So I think the point I am trying to make is, that although I do think clicker training is a very good approach, there are plenty of pitfalls, plenty of ways to mis-use or abuse it, and nothing is ever black and white or clear cut. I think with ALL forms of training one has to be prepared to look beyond what is being said to what the horses are saying - be that in pictures, videos, or real-life.
|
|
|
Post by mandal on Dec 29, 2010 15:41:16 GMT 1
I must say I agree with almost all you have written in your latest post and is much the way I think. However, I think horses are being trained with clicker training in ways that may be demeaning or demaging for them. In clicker training I think there is a danger of horses being 'micro-managed' as much as for example in Parelli style work where the horse must always give the handler 'two eyes and two ears'. In clicker training, some people actually 'shape' their horses to point their ears forward - to me that is wrong - the horse's ears are a way of monitoring his environment and expressing his emotions - we have no right to insist that he puts them forward for a click and a treat. I especially agree with things like 'ears forward' and instances where horses are actually at risk such as the slippy podium example you give. Ethical is such a difficult word isn't it? Ethics are very personal to the person especially difficult when it comes to thinking of ethics on behalf of horses. We all have slightly different stuff we are or are not comfortable with especially when we get to 'unnatural' behaviours such as tricks etc. I'm not sure how ignoring free shaping and target training means necessarily erm, questional CT tbh. I thought target traing was the first step in teaching CT so don't see how it can be avoided. Free shaping could in my view still lead to a horse being asked/expected to stand on a slippy podium though and although I see some value in it I don't see it as a requirement for 'good' CT training. That may well be down to my inadequate understanding and thinking of free shaping always in terms of CT when in fact we all do use it to some extent or another. I suppose all training is actually down to the 'mind set' of the handler in the end and what that handler feels is abusive, unethical, reasonable, essential, and justified?
|
|
Caroline
Grand Prix Poster
Intermediate Poster
Posts: 2,277
|
Post by Caroline on Dec 29, 2010 15:54:24 GMT 1
Dear Kelly,
The research and the theory may or may not agree with my findings, but I truly did see a massive fear-overcoming change in my gelding, Zee, when we started doing clicker training.
I had mentioned him to you a few times I think. He had defied all attempts to tame him - for many years, about 7 years I think it was. He was just so wary, so jumpy, so not tame. We think it was caused by a traumatic gelding combined by living with a wild herd for the first 3 years of his life. I tried everything - Join Up (went V badly - he practically killed me), Parelli (even worse), traditional, desensitisation, habituation, advance/retreat, A/C.
But thanks to SarahW for introducing us to it - clicker changed him dramatically. In a matter of weeks, he was catchable. After about 6 months I think (timeline is a little sketchy!), I was able to file his feet. This was a massive step forward for him and one that I felt particularly important. We still have a way to go and I am not sure he will ever make a riding pony as the tendency to spook and bolt is still there, but at least we can make calm contact and do the maintenance stuff like feet and grooming.
Both he and TigerLily are easy to catch without headcollars now too. When they escape, I just call them, they generally come running (sometimes at a canter!) and will follow me home at liberty. There was always an element of that in our relationship, but it has become much stronger and more dramatic since we started clicker.
With anything, we can only speak as we find I think. I hadn't expected clicker to be so effective, but it really was. The results are undeniable. There are so many people that knew him before his clicker-induced personality transformation who now cannot believe the change in him. I don't put it 100% down to clicker. I think that growing older has settled him more and consistently gentle and calm daily care has gradually chipped away at his fears, but there was a definite rapid emotional growth spurt in him when we started clicker.
Sure, he is just one horse. But when he is your own horse, that's all the result you need!
|
|
|
Post by ghostrider on Dec 29, 2010 16:01:02 GMT 1
I especially agree with things like 'ears forward' and instances where horses are actually at risk such as the slippy podium example you give.quite - I even read one person say she had taught her horse to walk along with his ears sticking out sideways because it was so 'cute' Ethical is such a difficult word isn't it? Ethics are very personal to the person especially difficult when it comes to thinking of ethics on behalf of horses. We all have slightly different stuff we are or are not comfortable with especially when we get to 'unnatural' behaviours such as tricks etc. oh, yes, the route of ethics ... very tricky and a whole different subject - which I why I try to shy away from using it in relation to training now. What is natural about horsemanship - very little really, at the end of the day I'm not sure how ignoring free shaping and target training means necessarily erm, questional CT tbh. I thought target traing was the first step in teaching CT so don't see how it can be avoided.yes, it often is one of the first things taught - and then left while people move onto other stuff - but it has so much potential that most people don't explore - and it is a great way for the horse to do some training without any form of restraint (apart from the treat bag, of course ) Free shaping could in my view still lead to a horse being asked/expected to stand on a slippy podium though oh, absolutely, and just because we can it doesn't mean we should ... and although I see some value in it I don't see it as a requirement for 'good' CT training. I guess that's just something that is down to the individual perception of the trainer and the horse in question - free shaping can cause confusion and frustration if either the trainer isn't very good with their timing, or the horse is new, or indeed the horse is shut down and doesn't offer many behaviours. Personally, I love working with horses that will offer behaviours and then 'play the game' with me to figure out what I want them to do - I guess I think of it as a sort of brain exercise for the horse (and me ) I suppose all training is actually down to the 'mind set' of the handler in the end and what that handler feels is abusive, unethical, reasonable, essential, and justified? [/quote] yup! I'd agree with that.
|
|
|
Post by ghostrider on Dec 29, 2010 16:03:45 GMT 1
Great post, Caroline, lovely to hear of your experiences :-)
|
|
|
Post by mandal on Dec 29, 2010 16:10:16 GMT 1
I guess I think of it as a sort of brain exercise for the horse (and me ) ;D Using CT and free shaping is definitely much more of a brain exercise for me than the horses! ;D
|
|
|
Post by june on Dec 29, 2010 21:23:39 GMT 1
On the covers of some very well known clicker training books there are horses that have contorted themselves into all sorts of weird shapes in a parody of 'collection' and 'self-carriage' - some ridden and some in-hand - and these are being held as examples to which others should aspire to. Glad I'm not the only one who has seen those books and wondered what on earth was good about those pictures!
|
|
sixfootblonde
Grand Prix Poster
www.western-saddler.co.uk
Posts: 1,138
|
Post by sixfootblonde on Dec 29, 2010 22:35:18 GMT 1
I guess I think of it as a sort of brain exercise for the horse (and me ) ;D Using CT and free shaping is definitely much more of a brain exercise for me than the horses! ;D Isn't it just?? ;D Well explained Caroline and wonderful that you have had such great results. I do see what you were saying earlier about this thread being about someone new to CT. And as it is such a powerful learning tool, that can go either way. You were right about getting an expert to set you up, that way you can avoid many pitfalls.
|
|