Posting on Tintagel

Dog training - essay about an engram

June 15, 2017 | 7 Minute Read

Here’s Iria. An Australian Shepherd. My dog, friend and family member. Iria since her puppy days up till now - an adult dog (8 years) - have been training with reward based methods / leadership without force techniques, where the reward is used as positive reinforcement.[1]
Below, I’m sharing one of my animal studies dissertations.

“Traditional” dog training is based on using an aversive stimulus to teach a choosen behavior. According to definition the aversive stimulus should cause pain or other unpleasant feelings, both physical and mental. As a result, the dog treated with such stimuli tries to avoid it. /The question is, does the dog have enough information - if it has any of them at all - on how to achieve it./
Same as its adversary the positive dog training[2], it is based on conditioning as a learning method. Positive punishment[3] and negative reinforcement[4] are used during training. The whole learning process is carried out on the rule of negative reinforcement, thanks to the potential increase in likelihood of the expected behavior in exchange for avoiding the correction (punishment). /Question, what connection - if any - exists between convincing the dog to avoid the correction and teaching the selected behavior./

What type of memories does a dog entering the training ground have, on which he was previously scolded? How does it impact the training process? What effects related to memories may then appear?
Let’s consider a hypothetical situation:
On the training ground a dog is trained to heel. Under the eye of the trainer, it is tugged by the leash by its handler when it moves away from the handler’s leg. Other course students are observing that situation.

Engram is a memory trace created by a certain experience with the use of senses. Remaining in the brain it can be reached from the memory at any time.
An engram of the training ground from that last training may contain following components:

  • visual -> the dog nearby, its handler, the trainer, own handler behaving in an unusual way, rainy aura, fenced terrain
  • sonic -> growl of the dog nearby, the verbal command ‘heel’ from the handler’s mouth, the voice of the trainer, the sound of the leash being pulled on
  • saporific -> the smell of the dog sitting nearby, its handler, the trainer, own handler, the wet grass
  • haptic -> tug of the leash, wet lawn as the tract.

The pool of possibilities is obviously not exhausted here, a trace (image) in the memory may contain other details.
Such an engram composed of sensory experiences may be forgotten
-> Leash correction during training have been omitted in dog memories, it means the dog learned nothing.
or moved to the short term memory.
-> Leash correction during training was engraved in the dog’s memories as something unpleasant, so it can be again forgotten if it wasn’t consolidated.
The natural consequence of the mentioned consolidation of behavior via repetition of action in a set moment (-> leash correction each time the dog moves away from the leg) will be remembered durably (-> unpleasant feeling (leash correction)) still without any guarantee that the memory of it will be related to any other circumstances (-> leash correction, because the distance from the leg was too far).
It’s also unknown if a dog is able to reach in memory to a consolidated engram when heeling - takes place in different conditions, with a different family member, in presence of other dogs, without the trainer, in better weather conditions, on another ground, on a different leash, with prong collar or the muzzle.
Engrams of course are not an invariant - one can be replaced with another (aversive training can be replaced with reward based one, where opposite methods are used) if the former did not turn to be “special” for the dog or the dog didn’t happen to be exceptionally susceptible for choosen methods. A lot depends on the dog’s age, gender, race (or race type) and also on its prior curriculum vitae.

Positive emotions enable and negative ones hinder - the ability to remember and finally to learn.

Let’s assume that the atmosphere on that training ground was tense, due to which the dog’s handler, hinted by the trainer, “corrected” the dog in a slightly harsher way than intended.
The handler, as can be seen, unloaded his emotions:
Annoyed (-> by the sharp notice from the trainer, multiplied if others were witness to it happening) felt uneasy (-> wrath, anger, shame) didn’t control himself and directed his emotions on the item held in hand (-> the leash). Unfortunately, on the receiving end was the dog which not only couldn’t anticipate the increase in stimuli but also not granted the privilege of unloading its emotions - too busy making sure to not go too far away from the heeling position to avoid the correction.
It’s obligatory at this point to mention that the accumulation of emotions in the dog can be a reason of aggression (including a redirected one) or conditioned helplessness. Such behavior should stop the training session for that day, to avoid straining the stress spiral and engraving the negative memory.

Why?

Let’s say that the anger of the dog’s handler transferred to the dog which disoriented tried to avoid a random punishment and his alarming behavior, due to which he refused further training (body lowered, ears back, slow movements, moving away from the humans leg), for what was scolded not only with a pull on the leash but also by a verbal correction (loud command repeated), additionally with the guide leaning over the dog with arm movement revealing an intent to hit.
Here’s the effect:
The memory related to the fear created on the training ground during dog training along with a full gamma of threatening signals sent by the handler as a consequence of the stress spiral straining.
Bonus which the dog didn’t deserve at all - a chain of negative connotations which hinder learning. Those memories are best remembered - potentially for hastening the future reaction in similar danger conditions.
From this point we are not far away from reasoning that traditional training favors development of fear including the fear towards the dog’s own handler.

Episodic memory is a collection of personally experienced events along with their mutual relations. Those memories ‘fade’ in the light of new upcoming, better ones. And yet it happens that the old ones hinder the ability to remember the new because they are deeply engraved in memory.
The fear is the example - the most important, the primal emotion related to survival. Fear memory is very durable, if not permanent. Memory trace containing fear is often narrowed down to stimulus responsible for fear. That’s the reason why it’s so difficult to learn something, even if it’s rewarded basing on positive reinforcement methods, when fear memory is still ingrained.

Keep it in mind before your decision on the training method is made.

[1] Positive reinforcement is a form of increase - it escalates the probability that desired behavior will be repeated, by presence of stimulus (something pleasure, reward for example).
[2] Positive dog training is a type of training opposite to the traditional one, based on modern knowledge about animal cognition. Reward is used as positive reinforcement during training (and the lack of a reward as negative punishment).
[3] Positive punishment is a form of decrease - it reduces the probability that unwanted behavior will be repeated.
[4] Negative reinforcement is a form of increase - it escalates the probability that unwanted behavior won’t be repeated, by absence of stimulus (something unpleasure, punishment for example).