The Art of Natural Dressage

Working with the Horse's Initiative
It is currently Thu Mar 28, 2024 9:10 am

All times are UTC+01:00




Post new topic  Reply to topic  [ 6 posts ] 
Author Message
PostPosted: Sat Oct 26, 2013 4:43 pm 
Site Admin
User avatar

Joined: Fri Aug 31, 2007 8:20 am
Posts: 6281
Location: Dresden, Germany
Today during a walk with Nora and Summy I had to do some soulsearching again. In my interaction with my horses, I try to adjust my body language according to the effects it produces in the horse. I have explained it elsewhere...

Romy wrote:
I try NOT to train my horses to react to a certain set of cues. It's more like if the horse perceives something to be a cue, then it becomes a cue or part of a cue for me - namely for the response the horse just gave. That's basically what I tried to explain in the R+ for humans video and text. So I do not have a certain, fixed sideways cue for example but a general idea of what could make the horse go sideways, which is defined and constantly redefined as our interaction changes, depending on what effects it produces. If in response to my current sideways cue the horse moves forward more than I want, I reward just as much but adjust my position the next time I ask. So I guess there is not very much need for the horse to filter things out.


The benefit of this is that the horses are teaching me a very precise body language so that I can learn to give exactly the right cues when asking for some movement. However, the downside is that if you do NOT give the right message with your body language (according to the horses' perception), they simply do not respond. This is no problem at all during our regular training because there is nothing we have to do. But outside the pasture it can make things very difficult for people who are not that much in control of their bodies yet. I witnessed this with Nora and Summy today, because it was one of these days when during our walk he just stopped moving all the time and then stood there as if he had fallen asleep. It's not that he wanted to go home - asking him to walk into that direction did not work either - and it was also not a day on which he preferred to be left alone, because as soon as I took over, he started trotting and jumping and being very communicative overall.

The reason it did not work was specific to the way Nora was moving. What would be needed in these situations is something like what Nelly shows here: First getting the horses attention, then (and only then) drawing the horse forward, and then giving immediate feedback. Compared to Nelly, I do this more with my body than by luring with treats, but the principle is the same. You need to establish a connection first and then bind your moves to those of the horse so that the connection stays intact, as if there was a thin elastic band between you that breaks if you make your own moves instead of coupling them to those of the horse. For some people this is easier than for others, but the problem was that Nora was doing her very best. She tried soooo hard to get it right, and still it did not work for her. We know that it CAN work, because we have had the same situation this spring and after lots of practising, she developed a body language that made Summy react to her just perfectly.

Anyway, this made me wonder to what degree I want horses whose reactions are so pure. It would be possible to establish a set of default cues that make them respond regardless of how well the human is doing. This could be done with slight pressure, but most likely also in a purely positive reinforcement based way. For example, Bianca does that in her interaction with her horses. She has a set of symbolic gestures (e.g. hands up or stretching her arm to one side) that have a clearly defined meaning and her horses have learned to respond to these signs in a particular way. This makes it possible for everyone to ask them for certain things like stopping or moving to one side, and people can do this no matter what the rest of their body is suggesting to the horse. In that way, I could make sure that everyone is able to handle my horses.

Well, I have come to the conclusion that I want to keep things the way there are. This might mean that I am sacrificing the "usability" ;) of my horses, because there are no shortcuts to eliciting a response from them. But what I really want is to help the children to become better at communicating in the long run, more than I want them to get certain things done. After all, I can always take over if it does not work, so there is no actual need to do it differently. And still that question keeps spinning through my head, especially after we have lost someone we were training with due to exactly that thing, that it felt like too much effort if you have to stay awake and be precise and adjust to the horse all the time instead of having things just work out easily. Therefore, I'd love to read about your opinions and experiences with this. :f:


Top
   
PostPosted: Sat Oct 26, 2013 7:30 pm 

Joined: Sat Dec 17, 2011 2:46 pm
Posts: 250
Location: Meadow Lake, Saskatchewan, Canada
Lovely topic Romy :f:

For me, as I am training a few horses for others, I HAVE to put in that "other language". For the horse's sake, there needs to be a set of "commands" that supersede anything else. I "install" these following Dr McLean's work - using negative reinforcement, we get extremely light, reliable responses to the foundation tasks of:
Go: - includes up transitions, speed up/lengthen stride
Stop: - includes down transitions, slow down/shorten stride and back up
Turn front end: includes turns, turns on the haunches/pirouettes, and weight shift from one front leg to the other
Turn hind end: includes any lateral movement with the haunches, turns on forehand

I am struggling with this sometimes too - how much do I want the horse to follow my energy/subtle (and ever changing) body language - vs touch based, clear, learned cues.... But it's getting better as I use both tactics....

This is such an important topic - I sometimes encounter folks who say they only use positive reinforcement etc. Yet when a situation arises when their horse does not want to participate - they turn to downright violence against their horses. And I have a problem with that... I am in no way saying that everyone needs to use dominance/negative reinforcement and so on - but I think it's best if people are aware that there will likely come a time when the horse chooses not to comply. And I have so much respect for the people on this forum, for they truly respect their horse's opinion and do not have the "need" to make them do something against their will.

I feel like I am slowly making my peace with combining these (seemingly opposite) approaches. :)


Top
   
PostPosted: Sun Oct 27, 2013 10:19 am 
Moderator
User avatar

Joined: Wed Oct 20, 2010 7:42 am
Posts: 2147
Location: Vienna, Austria
I've been wondering about 'pure responses' and communication in total and I think I would divide it into two parts. One would be like verbs and subjects: trot, stop, turn, step on that pedestal, and so on. The other like adjectives: fast, slow, collected, lively, ...
If there really is such a thing as an universal body language or a pure response than it would be based on a common understanding of the world, communicated by something like verbs, subjects and adjectives. The first two I find rather easy to agree upon. A tree is probably a tree for a horse as well as for a human. The same with trot or halt.
Where it gets fuzzy real quick though, is when it comes to adjectives. When exactly is a walk 'energetic'? Is it the same for one horse, as it is for another? Or for a human?

Where I am going with all this is that I imagined Nora with Summy and I wondered, if she can interact with Pia like I know she can, what is different with Summy? Did Summy really not understand that Nora wanted to go forward with him? Or was it more that the individual representations of how that walk should look like - the adjective description - did not correspond?
I am pretty sure that Pia's mental image of 'walking through the forest' is quite different to Summy's one ;).

So, what's the pure response of the horse then? I'd imagine that Pia's response to Nora would be to follow immediately, whereas Summy's would be to hesitate. Not because the question was all different, but because it was 'walk on' in Pia style. It was basically the same words, but coupled with a different adjective.

Now, I believe that a certain language of cues can be taught and they would even work in most circumstances, with most horses. But what really fills the gaps - and in the end makes the cue all the more reliable - is the adjective quality that is added to it and that is specific tuned to the individual horse. It's the colourful vibes that make the body say 'come on, walking on is the most fun thing in the world right now' to the unmotivated horse. Or to the insecure one: 'hey the way you walk on is the best I've seen so far'...

In the end, I think it boils down to a matter of motivation. If prerequisites are met and the goals of horse and humans coincide, then it's very easy to give a cue to 'walk on' for example. If that is not the case, it get's more complicated.
Either it's done with attention to smallest detail and rewarding every hint of compliance. Or with negative reinforcement and the promise of unpleasant consequences.
I prefer a derivation of the first approach. I try to use positive motivation by rewards and the use of attentive body language in combination of a simple cue. Over time I see that the cue I give beforehand gets so habitual, that it is sufficient in trivial situations.
This takes away the burden of having to be overly attentive all the time in normal circumstances - also for the horse I believe.
In difficult situations of course - severe lack of motivation or fearful stimuli - I have to resort to micromanagement again.

_________________
Volker

The horse owes us nothing.


Top
   
PostPosted: Sun Oct 27, 2013 4:57 pm 
Site Admin
User avatar

Joined: Fri Aug 31, 2007 8:20 am
Posts: 6281
Location: Dresden, Germany
Zu, thank you for your reply! :) For me these two things aren't mutually exclusive, either. My horses also know some symbolic emergency cues that they are responding to, like holding up the rope towards their hindquarters and perhaps waving it as a sign for "You must go forwards now, please!" However, I don't use these cues unless it's totally necessary, and I also don't show them to the children, because I don't want them to take the pressure shortcut. I can totally understand that it's different in your situation, but for us there aren't many things the horses have to do, because we are only interacting with them for their and our own pleasure. :smile:

And a lengthy response to Volker. First, thank you very much for making me think, this has made it much clearer for me! :kiss: The thing I meant is neither a matter of verbs nor adjectives, but rhetorics. If my horses want to walk forwards anyway, it totally does not matter how you are asking, because they will simply walk. No need whatsoever to communicate this in any particular way, and later during our walk Summy became so interested in walking and so eager to move on that it did not matter anymore at all what Nora and I were doing, as long as we did not block him. However, if the horse does not have this genuine interest in an activity but you still want him to do it, I think it is easier if you communicate this in a way that automatically triggers the response you want.

Do you remember our first cordeo walk with Lily, when I used the attention checks to make sure she stays with me instead of just going for the grass? Her very own personal interest was not to walk in synchrony with me and I am sure she would have been perfectly happy if I had just let her do her own thing (eating and walking), with me following passively. However, by doing these abrupt stops and variations with my body, it was possible to capture her attention and interest and make her look at me, which then I could reward. Instead, I could have stood there passively, said a verbal command for halt, raised my hand, slowed down gradually, or done any other thing that might also count as a cue for halt but is not an automatic trigger for it. Lily might have reacted to all these other signs if she had gone through lots training with them, but my movement cues worked at once. That's what I mean by a pure response – one that is automatically triggered by the human's movement.

You were referring to human verbal communication, so I'll explain my analogy for the thing I meant to express in my first post. If you want someone to do something, then you could also say that the only thing that should matter is that you use the words to describe the thing you want him to do and the fact that you want him to do it. It should be completely irrelevant how you say it, as long as the facts are present. But of course that's not what is happening. Instead, it makes all the difference if you are able to formulate it in a way that makes it sound logical and desirable, so that the other one immediately feels that he wants to do it. In that way, you can talk him into something he might not have done in the first place, and he might even feel that it was his own idea – simply due to the way you talk, whereas saying the same thing in a less persuasive way might get you no response or even resistance to your idea.

And that is the "pure response" I was talking about: the horse doing what he is being asked only if the question makes him feel that he wants to do it (or not even consciously feel anything, but just going along with it automatically) versus doing it because he knows that this particular question is what he has learned to be the appropriate response. It's the same with humans: You can train them to do what you were suggesting anyhow. For example, pupils often do what their teacher says no matter how unconvincing it is, because they have learned that this is the assignment and they should do it. In the same way, I could train my horses to respond to a forwards cue even if this cue itself does not make them want to go forwards, just because they have learned that the cue means forwards. So I guess my question was whether I want horses who react only if the cue itself is convincing, or who react because they know it's the signal for this or that.

Perhaps I seem like a very manipulative person when saying that I want the children to learn how to "talk the horse into something". And I guess I am. However, during some of our joint activities, it is not an option to just let the horse do whatever he wants without trying to influence that. This works when training in the pasture, because when Summy just stands there and does nothing (or eats or walks somewhere or does anything else that differs from what we want to do), we can simply let him do this and play with another horse instead, or leave or do something on our own. But on a walk we are stuck with the horses, because we cannot just set them free and leave. Therefore, there are basically two options: We can do what the horse offers (which in Summy's case might mean that sometimes we just stand in one place for half an hour), or we can persuade him to do what we want, and in that case I very much prefer it to make my idea his idea rather than simply forcing him or having him respond to a learned signal.

In response to your question why it is different for Nora and Pia: In one way it is, and in another way it isn't that different at all. I'll talk about the differences first. Pia has strong preferences for certain activities. For example, she loves chasing people, so you can reframe the activity itself (for example by turning something that is about fear into something that is about chasing Papa). With her, it matters much more what you do than how you do it, and Nora has a wonderful ability to work on the what (much better than me, I might add). The thing that isn't that different in Nora's and Pia's interaction becomes apparent when it doesn't work to change the meaning of the situation, because Pia wants or does not want something very strongly. In these situations, Nora cannot get her to do things, either, and then she just hangs on to the rope until she asks me to take over. Nora is brilliant at changing situations, but it is harder for her than for me (or Azhar and Nelly) to focus on the way she communicates. Again, it's similar to inter-human communications: Some people are just fun to be with because they have great ideas – that would be Nora – and then there are people who are very subtle in the way they talk and adjust this to the situation. This may or may not go along with having great ideas, but for me these are two fundamentally different modes of influencing your interactions with others. With the children, I am lucky to have experts for both modes, so we can all learn from each other. :f:


Top
   
PostPosted: Wed Oct 30, 2013 6:51 pm 
Moderator
User avatar

Joined: Wed Oct 20, 2010 7:42 am
Posts: 2147
Location: Vienna, Austria
I took a bit to respond, as it's a topic on which I haven't entirely made up my mind yet. Romy, I totally love the way you elicit what you call "pure responses". It's something I work hard to improve in my own interaction. Still, there are some points I like to hear your opinion about.

It might be just a matter of naming, but I guess what makes me hesitate is when a 'pure response' is described like a reflex reaction from the horse - a response that comes automatically as reaction to certain body languge for example. I don't know if you meant it like that, Romy, it's just how it presented itself to me.
While I think that there are some cues which elicit more unconscious reactions, I don't think that a response like moving in synchrony, is totally subconscious. The term 'pure' suggests it though - for me anyways ;).

I believe that certain cues - especially body language ones - are so much more familiar to the horse than any other learned cue that they are more easily understood. It's basically like speaking in their own language and thus the answer might come more readily and more accurately. The motivation to answer to an understandable question is a lot higher than to a obscure one, where a lot of guessing and thinking needs to be done and where the risk of failure is high.
What also might add to the motivation of following body cues versus abstract cues is that horses are herd animals and are very used to act in synchrony. Also, if I remember correctly, moving in synchrony fuels the dopamin/oxytocin system, which makes such movements highly motivating and rewarding.
So that would explain to me that body language vs. abstract cues are more are more effective in terms of responsiveness.

Another point I see that surely makes the way you apply these 'pure responses' very effective is described in this quote:
Romy wrote:
Do you remember our first cordeo walk with Lily, when I used the attention checks to make sure she stays with me instead of just going for the grass? Her very own personal interest was not to walk in synchrony with me and I am sure she would have been perfectly happy if I had just let her do her own thing (eating and walking), with me following passively. However, by doing these abrupt stops and variations with my body, it was possible to capture her attention and interest and make her look at me, which then I could reward.
Now, that's a very clever way of getting the horse to do what I want, but I aks myself if it is really related to something like a 'pure response'? Basically what you did was not cueing the wanted action directly (like "walk with me"). Instead you cued for little things, which you were sure the horse would react to, then used the resulting attention to form synchrony, which you then used to walk together in the direction you wanted. It was in fact not so much a result of a pure way of responding to the orignal question, but a result of cleverly rephrasing the interaction, so the fulfillment of the original goal was more like a byproduct of the little pieces of interaction.
I think that horses respond extremely well to this kind of interaction, because it takes the focus away from the overall task, which might be something abstractly boring as "go from A to B", or "train your body" and put it right where it is needed: in the here and now.

I'm not entirely sure where I want to go with all this - maybe it's really just the term 'pure response' that doesn't fit for me... :smile:

_________________
Volker

The horse owes us nothing.


Top
   
PostPosted: Thu Oct 31, 2013 8:22 pm 
Site Admin
User avatar

Joined: Fri Aug 31, 2007 8:20 am
Posts: 6281
Location: Dresden, Germany
Volker wrote:
It might be just a matter of naming, but I guess what makes me hesitate is when a 'pure response' is described like a reflex reaction from the horse - a response that comes automatically as reaction to certain body languge for example. I don't know if you meant it like that, Romy, it's just how it presented itself to me.
While I think that there are some cues which elicit more unconscious reactions, I don't think that a response like moving in synchrony, is totally subconscious. The term 'pure' suggests it though - for me anyways ;).


I don't think that being subconcious is mandatory for what I call pure responses, but I believe that conscious reflection is not necessary for them to occur. Of course I can be very aware of the fact that I am moving in synchrony with someone, but if i do not pay attention to that, I do it anyway.

For me it's a matter of similarity between the cue and the movement it elicits - or in more mechanistic terms, a question about the degree to which the cognitive codes for representing the cue and the movement overlap. Movements as well as perceptions are bundles of features. For example, an apple is not represented as a unitary concept in the brain (there is no "apple neuron" or "apple brain area") but as a combination of the firing of neurons that code its aspects, for example neurons that code a round shape, neurons that code a particular shade of red, neurons that code the shininess of its surface, and so on. If these neurons fire in synchrony, the mental representation of the apple gets active. Similarly, a forwards movement with a particular speed, impulsion and distance is represented by a pattern of synchronous firing of the neurons that code for these aspects (i.e. speed x, impulsion y, distance z) - and so is the perception of a movement with these parameters.

In consequence, when you perceive a body language cue with particular parameters, this automatically activates a movement pattern with the same parameters by means of spreading activation (from the perceptual codes to the corresponding motor codes) - so the mere perception of that cue can activate a movement that fits with it. This does not mean that you automatically have to perform that movement, but this seems to be a matter of inhibition more than a lack of activation (and indeed, patients with particular brain lesions that compromise the ability to inhibit tend to involuntarily imitate a movement when they see it). The term pure response might or might not fit for a response that gets directly activated by the perceived cue, but I am not particularly interested in conventions of naming things, so we can call it whatever you like. :smile:

What I just wrote about code overlap and spreading activation still does not explain why for example Summy follows me in a gradual movement but does not follow Nora when she just flips forward. After all, the features of her movement could activate the same feature in Summy as well and thus make him go along with that, automatically. However, I think this has to do with the threshold the activity of the feature codes needs to cross in order to trigger a behaviour. If, for example, Summy feels very sleepy, it is much harder to push the activations of his cognitive codes for fast movements above that thershold, because their baseline level of activation is so low. Accordingly, while these "fast features" do get some activation from perceiving her fast movement, it is not enough to trigger a response, so no movement emerges. In contrast, when it's very stormy for example and he is very much awake anyway, the basic activation of these same codes is higher already, so it takes less additional input to make them cross that threshold and make him start running.

So why do gradual increases work better to get him out of that resting state? I am not entirely sure, but I think it is because activation spreads to similar codes. That is, if I start off very slowly, the activation spreads from his codes for very slow movements to those for a tiny bit faster movements, so when I increase my speed afterwards, it's easier already to push them over that threshold, and at the same time the codes for movements that are still somewhat faster get a tad more activated as well, which then makes it easier to push them across the threshold, and so on. This might or might not be the way it works, but from what I know about the way mental codes for perception and action are activated, it does not seem unlikely. ;)

Volker wrote:
I believe that certain cues - especially body language ones - are so much more familiar to the horse than any other learned cue that they are more easily understood.


Yes, definitely. From a common coding perspective (i.e. the notion that perception and action are mentally represented in the same way and therefore have the ability to activate each other), this higher familarity goes along with the fact that throughout a horse's life, the perception and performance of such movements have always co-occured. For example, when performing a fast movement, the horse has always perceived (his own) fast movement as well. Therefore, the feature codes of these two events (performing and perceiving "fast") have become highly associated. As a result, the perception of any fast movement facilitates the performance of fast movements by means of spreading activation, and if that activation is high enough, the fast movement gets triggered.

By the way, this co-occurence explanation also implies that a symbolic cue might become a direct trigger for a movement as well if it goes along with that movement often enough. But then a few or a few hundred co-occurences of the symbolic cue and the movement stand in no relation to the millions of times the movement features that constitute the body language cue have co-occured with the movement features of the horse's behaviour, so I guess the body language cue will always be more direct in activating the movement.

I guess this was more fuzzy than what you wanted to hear and probably does not answer your question, does it? :blush:

Volker wrote:
The motivation to answer to an understandable question is a lot higher than to a obscure one, where a lot of guessing and thinking needs to be done and where the risk of failure is high.


Yes, I totally agree that motivation plays a huge role as well, as do emotions and preferences and personality, and a lot of other factors. But then that's the beauty of reasoning about behaviour and interactions between individuals, I think - no single theory can account for all of it, and there are so many different explanations on different levels. So we will never find the one and only truth and can keep thinking and discussing until we are old and grey. :green:

Volker wrote:
Now, that's a very clever way of getting the horse to do what I want, but I aks myself if it is really related to something like a 'pure response'? Basically what you did was not cueing the wanted action directly (like "walk with me"). Instead you cued for little things, which you were sure the horse would react to, then used the resulting attention to form synchrony, which you then used to walk together in the direction you wanted. It was in fact not so much a result of a pure way of responding to the orignal question, but a result of cleverly rephrasing the interaction, so the fulfillment of the original goal was more like a byproduct of the little pieces of interaction.


Not sure whether I understand your point here. Do you mean that it's just the "splitting up into small pieces" that got her to react? If so, do you think that I would have reached the same effect if I had given little arbitrary cues, such as raising my finger and rewarding if she had attended, then turning my hand upside down and rewarded her for slowing down, then turning my hand a little more and rewarded for slowing down more?

I think of it as a pure response because to me it did not feel like I had an "original question" for which I first got the attention and then rephrased it, but more like a sequence of hundreds of little component questions, each of them being "Will you adjust your way of moving to mine?" But perhaps it would indeed be better to understand your question first before writing even more weird stuff. :funny:


Top
   
Display posts from previous:  Sort by  
Post new topic  Reply to topic  [ 6 posts ] 

All times are UTC+01:00


Who is online

Users browsing this forum: No registered users and 5 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum

Search for:
Jump to:  
cron
Powered by phpBB® Forum Software © phpBB Limited Color scheme created with Colorize It.