aversive stimulus – something unpleasant that the dog will work to avoid;
- it can be unavoidable (rain, street noise) or inflicted (shouting)
behaviour – an action (it’s impossible to not behave, behaviour occurs constantly)
- sit, watch, spin; but also: chewing, growling, jumping, lying down
behaviour modification – behaviour modification goes deeper than training, it’s more like therapy that aims to replace an undesirable behaviour by a desirable one by changing the underlying emotions
- if your dog barks and launches hysterically at cyclists it is probably rather a case for behaviour modification than simple training
capturing – animals exhibit many behaviours on their own. if we want our dog to learn to perform a particular action on a cue, we ‘capture’ it by using a marker
- observe situations in which your dog sits on their own and click/treat them the moment their bum touches the floor. after a while you can add the cue (for example “sit”), repeat for another couple of times and see if your dog will sit in a different context after they hear the cue
classic conditioning – a learning process: a new stimulus is paired with a ‘natural’ stimulus that elicits an involuntary response
- Pavlov was collecting dog saliva. His dogs drooled when they knew food was coming. Then Pavlov started ringing a bell immediately before food was served. The dogs learned to associate the bell (the new stimulus) with food (the ‘natural’ stimulus) and then they started drooling (the involuntary response) when they heard the bell – even before they could smell the food.
clicker – a small hand-held device where a piece of tin makes the clicking sound; see: event marker
complex behaviour – a behaviour consisting of multiple behaviours
- fetching is a simple example of a complex behaviour: the dog needs to run, catch the ball, come back to us and release the ball
counter-conditioning – changing a conditioned reaction to a different one, usually a “negative” one to a desired one;
- the dog might bark at the mailman. with counter-conditioning we can make teach them that the mailman is a cue for going to their bed.
cue – a sign (usually a word or a sound but it can also be a gesture or an object) that requests a behaviour
- saying “stay” or “heel”; pointing to the floor (a popular sign for “down”); squatting (very clearly says “come to me” to most dogs)
desensitisation – means making a reaction to a stimulus less severe; it is an intentional process;
- instead of hiding under the bed when the vacuum cleaner is on you want your dog to stay calmly a room away: you can start with a recording of the vacuum cleaner sound. when you first play it, it should be quiet enough to not make your dog react to it negatively and simultaneously you feed your dog treats, play with them etc. = create a positive association; then move the volume up a little bit and repeat
event marker – usually a sound, it is a sign that tells the dog what exactly it did right and that the reward is coming; the event marker must be paired with a primary reinforcer;
- the “click” sound of the clicker, a marker word (preferably a short one like “yes!” “yip!” “good”)
extinction – when a particular behaviour stops yielding rewards (reinforcement) 1
- If your dog is countersurfing, the simplest solution here is extinction – stop leaving food around. The countersurfing will stop when it is no longer reinforced.
habituation – getting used to something; it can only work if a reaction to a given stimuli is not severe;
- the more you jerk the leash, the more your dog gets used to it and you have to apply more and more pressure to achieve the results you want.
- your dog might not like the street noise but if they aren’t afraid of it and it doesn’t disturb them too much, they will get used to it (like we will)
lure/luring – using a ‘lure’ (a bait) to bring the dog into a desired position; a lure can be anything that the dog wants, it’s usually food;
- you take a piece of kibble in front of your dog’s nose and move it in a circle – your dog will spin as they follow the kibble. congratulations – you just lured your dog into doing the trick “spin” (it is a long way from there to spinning on cue though)
operant conditioning – a learning process: by using punishment or reinforcement voluntary behaviours can be taught;
- sit, recall, stay… these are all voluntary behaviours that can be taught using operant conditioning;
- let’s go back to Pavlov and his dogs (see the example at classic conditioning): he made them drool at the right moment (involuntary behaviour) but he probably also taught them to sit and wait for food (voluntary behaviour). the dogs could refuse to sit but they could not refuse to drool
punishment – aims at decreasing the occurrence of behaviours
- positive punishment (P+) – adding something unpleasant (aversive)
→ the dog barks and you shout at them: the unpleasantness in this situation is the shouting;
- negative punishment (P-) – taking away something nice (positive stimulus)
→ the dog barks to make you throw the ball and you put the ball away: here the ball/the game of fetch is the positive stimulus
!!! in this context “positive” means that something is added to the situation, not that punishment is a good teaching method !!!
reinforcement – aims at increasing the occurrence of behaviours
- positive reinforcement (R+) – something pleasant is added
→ the dog lies down and gets praised; the dog drops the piece of dry, mouldy bread and is rewarded with lots of treats and a little party
- negative reinforcement (R-) – something unpleasant is removed
→ e-collars: you 1. switch the unpleasant stimulation – usually electric shock (!!!) – on, 2. give a command and 3. switch it off once the dog obeys; a truly horrible method
!!! punishment and reinforcement are subjective.
- if you shout at your dog for barking and the barking doesn’t stop/gets worse, than shouting is a reinforcer in that situation.
- if you pet your dog after recall and their recall does not become more reliable, then petting them is not a reinforcer in that situation. !!!
reinforcer – usually a pleasant stimulus (food, play, affection). The dog determines what is and isn’t a reinforcer in a certain situation. If the dog doesn’t want what you are offering, it’s not a reinforcer.
- primary reinforcers can be defined as stuff that is valuable to your dog on its own (food, comfort, control, for some dogs play)
- secondary, tertiary (etc.) reinforcers are things that your dog learned to like
- applause is a secondary reinforcer for humans. Applause itself has no value to us but we learned to associate applause with good things (awards, appreciation, prestige) and it has become a reinforcer.
- in clicker training we use a whole chain of reinforcers: the treat (primary reinforcer) is preceded by the clicker (secondary reinforcer) which is preceded by the cue (tertiary reinforcer); cues can be reinforcing for the dog if they associate them with pleasant interactions
shaping – perfecting or modifying a behaviour
- you teach your dog to heel so first you reward them for being 1m from you, then 45cm, then 30cm… you get the idea
- in free shaping you start “from nothing”, for example you will put a mat on the floor and teach your dog to liedown on it by rewarding behaviors that move in that direction; free shaping requires the trainer to have clear criteria and the rate of reinforcement should be high, otherwise you risk frustration in your learner.