What’s Learning Theory? I’m Glad You Asked

I have recently been speaking to a lot of different people in the horse world. What seems to be a common factor is lack of understanding when it comes to learning theory. And no wonder! The information is out there, but it’s still not very easy to find. Especially when one is not looking for it! That’s why I decided to write a blog post about it.
The thing that is important to note, is that this is all science based. This is not something I make up, or twist around to suit my idea of what horse training should look like. This is thoroughly researched and tested, so in this point in time, there’s no arguing against it. It’s hard fact.

Now! What is learning theory? Learning theory is the knowledge of how animals learn. When I say animals, I mean ALL animals. And also certain (if not all) insects. Humans, dogs, fish, cats, kangaroos, butterflies, elefants, crocodiles and – you guessed it – horses!

When I say learn, I mean learning to act in a specific way to achieve or escape something. This can be a bit confusing – you’re probably not used to think of behavior in such a specific way. But I promise it will get easier to understand when I have explained the ways in which the animal learn.

There are four ways an animal can learn (in addition to one sneaky one). Period. There’s no other possible ways for an animal to learn something except these. We call them collectively the learning quadrant.

What the quadrants have in common, is that they are all dependent on motivation. You simply do not get behavior without motivation (you can also call it reason). Imagine if you’re on the couch. If you have no motivation to get up, you won’t get up.

Here are the four quadrants (I will explain them in detail further down):

  1. Positive Reinforcement (+R)
  2. Negative Reinforcement (-R)
  3. Positive Punishment (+P)
  4. Negative Punishment (-P)

That’s it.

Reinforcement strengthens behavior, punishment weakens behavior. And then there’s the positive or negative:

One should not think of positive and negative in a “good/bad” sort of way, but in an “add/subtract” sort of way.
Positive means adding something, negative means subtracting something.

This means that the quadrants can be explained thus:

  1. Adding something to strengthen behavior (+R)
  2. Subtracting something to strengthen behavior (-R)
  3. Adding something to weaken behavior (+P)
  4. Subtracting something to weaken behavior (-P)

 

Now that the quadrants are (hopefully) at least a bit more understandable, I’ll go into detail:

  • Positive Reinforcement:
    “Adding something to strengthen behavior (+R)”
    Positive reinforcement is the obtainment of something the animal values when the animal behaves in a certain way. For example:
    – A horse kicks a tree, an apple falls down, the horse gets the apple, and is more likely to kick the tree in the future.
    – I hold an object in my hand – the horse puts her nose on the object – I give the horse food – the horse is more likely to put her nose on the object in the future.
    – I wait for the horse to take a step forward – horse takes a step forward – horse gets some food – the horse is more likely to take a step forward in that situation in the future.
    – ETCThere are many ways to get behavior using this quadrant, but I won’t go into detail about them in this blog post.

    The good things about +R is that this triggers the horses brain to release happy hormones (endorphins). It also ensures that the horse associate both the training, environment and human with this feeling of happiness.
    Since the horse has these associations, she will be a VERY willing participant in training. She will often heighten her own criteria, do more than you asked for – because she loves the training!

  • Negative Reinforcement:
    “Subtracting something to strengthen behavior (-R)”
    When training with negative reinforcement, it’s important to note that in order for it to work, one has to first add an aversive stimulus (a stimulus the horse finds irritating/painful/uncomfortable etc). It is the removal of said stimulus that strengthens the behavior. So you add an aversive stimulus to the horse to prompt the horse to behave in a certain way, and when she does, you take the stimulus away. This is the most used quadrant in horse training today. For example:
    – A fly has landed on the horse – the horse finds the feeling of the fly on her skin irritating (a form for aversiveness) – the horse shakes the fly off – the horse is more likely to shake that body part when she feels a fly.
    – A rider adds pressure to the sides of the horse – the horse takes a sted forwards – the rider removes the pressure – the horse is more likely to move forwards when the rider adds pressure to her sides.
    – A rider pulls on the reins, which adds aversive pressure in the mouth/nose – the horse stops – the rider stops pulling – the horse is more likely to stop when the rider applies said pressure.What many seem to struggle with, is the fact that in order to prompt behavior this way, the stimulus HAS to be uncomfortable in some way. If it wasn’t, the horse would either ignore the stimulus (if it is neutral to her) or try to obtain the stimulus (if it is pleasant to her).
    What’s more, is that the horse will only do exactly what is necessary to escape the stimulus. To get any extra effort, you’d need to up the stimulus (make it more unpleasant) so the horse will react bigger/faster/etc.

    Since no one likes having unpleasant stimulus applied to them (horses no more than humans), this way of training won’t really release any happy hormones. Quite the contrary. It will actually trigger the horses brain to release stress hormones (adrenaline). Not necessarily a lot, that depends on the strength of the stimulus, and how aversive the horse finds it.

  • Positive Punishment:
    “Adding something to weaken behavior (+P)”
    This is the most common form of weakening behavior in horse training.
    The horse does something you don’t like, you apply some aversive stimulus, and the horse stops the behavior (maybe).
    What’s important to remember with this quadrant, is that the horse doesn’t learn that the behavior is bad. What the horse learns, is that the behavior is not safe in the presence of you. It also doesn’t tell the horse what to do, only what NOT to do in your presence.
    Also, the use of punishment ignores the reason behind the unwanted behavior. It doesn’t do anything to weaken the motivation behind the behavior. To put it simply: all it does is to mask the symptoms. Behaviors are symptoms/realizations of motivation.
    For example:
    – The horse threatens to kick – you hit it with a whip – the horse will be less likely to threaten to kick you in the future.
    – The horse starts running off with you – you yank on the reins to get it to stop – the horse is less likely to run off with you.
    Note that the motivation behind these unwanted behaviors is not addressed. The horse will still want to kick you, and run off with you. What tends to happen is this:
    The horse has been punished for threatening to kick. So instead of threatening, she jumps right to kicking.
    The horse has tried turning for home when out hacking, and has been punished for it. So instead of gently turning around and go home, she feels that her only choice left is to spin around and run. So you won’t get the chance to stop her.There is just no way of using punishment correctly and without side effects.
  • Negative Punishment:
    “Subtracting something to weaken behavior (-R)”
    Negative punishment is not used a lot in horse training. It is, however used in human training (with children etc) and dog training. Basically, the animal has something available to them that they find valuable. Attention, food, toys, company, the possibility of something good happening later (with children). Then the animal does something you don’t like, so you take that away from them.
    For example:
    – A dog is happy to see you, so it jumps up on you – you immediately turn around and walk away – the dog lost your attention when he jumped, so he’s less likely to jump up on you in the future.
    – A child does something you don’t like – you say “now we won’t go and get ice cream after all” – the child is less likely to do what resulted in this outcome in the future.
    – You are feeding your horse food – the horse does something you don’t like – you walk away and take the food with you – the horse is less likely to repeat the behavior that resulted in the food disappearing.As with positive punishment, this kind of punishment doesn’t deal with the reason behind the behavior, and doesn’t tell the animal what to do (only what not to do).
    In addition to this, using negative punishment will trigger a lot of frustration in the animal. Losing a valued resource is very frustrating!
  • The last way an animal can learn, is through extinction. This happens when a behavior no longer gets reinforced. If the horse kicks the tree, and an apple doesn’t fall down, the horse will first go through something called an extinction burst. This means that before the behavior gets extinct (stops happening), it will get stronger. The horse will kick the tree harder, more, on other spots etc. If an apple was to fall down now, it would reinforce this new, stronger version of the behavior. Next time the horse wants an apple, she would jump right to this. If, however, after enough effort the apple does not fall down, the behavior would go away completely.
    Same with negative reinforcement – if you apply pressure with a whip for instance, and keep applying the pressure no matter what the horse does, the horse will (in the end) stop reacting to the pressure. This is called flooding, but that’s another matter.

 

I think that’s enough brain feed for now, please contact me if you have any questions or would like to learn more. And check out this website for a lot more interesting articles!

 

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s