Page 86 - Zoo Animal Learning and Training

58 4 Choosing the Right Method: Reinforcement vs Punishment

VetBooks.ir possible side effects that may develop from specific behaviours performed. This differ-
ence may seem semantic in nature but there
the use of aversive tools, but the most prob-
are key pragmatic differences. Two examples
lematic is aggression.
in the human world will illustrate the
difference:
4.5.3 The ‘Sneaky’ Animal
A child gets a bad report card. The parents
Another major challenge associated with choose to punish the child for the grades they
punishment is that the learner begins to received by not allowing them computer priv-
associate punishment as coming from the ileges for a week. The behaviour of getting
trainer. When this happens, the animal learns bad grades was not punished, that behaviour
to behave appropriately in the presence of occurred weeks if not months earlier. Instead
the trainer, because that is the only time that the behaviour that was punished was the
punishment occurs. But in the absence of the behaviour of showing the report card and the
trainer, the animal will continue to exhibit punishment is more likely to teach the child
the unwanted behaviour, creating the appear- to not show the report card or to forge their
ance of being ‘sneaky’, when in reality this parent’s signature on the card in the future.
happens because punishment is seldom a The child was certainly punished, but not for
consequence in the absence of the trainer. behaviour the parents were targeting!
A criminal commits a serious crime. He
eventually goes to jail but sentencing and jail
4.5.4 Emotional Responses
time happen many months (in some cases
Training is successful because of the proper years) after the crime was committed. A large
application of scientifically proven learning percentage of criminals go on to commit
principles. However, when training animals additional crimes once they are released
we can become quite emotionally attached (Flora 2004). Jail punishes the person (which
when an animal does something extremely may be a necessary aspect in modern society)
well or extremely poorly. These emotions but it does not punish the behaviour.
often cloud our judgement and make it When punishment or reinforcement is
difficult for us to reason through what is hap- directed at the learner instead of the behav-
pening or adhere as precisely to proven prin- iour, at its best the animal is receiving vague
ciples. It is often the frustration and anger guidance, at worst it is meaningless or
that comes when an animal is perceived to teaches the wrong thing. Only when training
misbehave that causes the use of punishment is focused on behaviour and uses good tim-
to get out of hand. Emotional responses can ing and consequences of appropriate value
often be less precise and in many cases too will training be successful and an animal
severe; there is a fine line between a well‐ learn desired behaviour.
timed punishment and inhumane abuse.

4.6 Making an Informed Choice
4.5.5 Misuse of Common Definitions
Too often when training an animal there is Choosing the appropriate method requires
too much reliance on common uses of the knowledge and skill. But most importantly it
terms ‘reinforce’ and ‘punish’. In our society requires setting goals and knowing what you
we tend to reinforce or punish people and ultimately want to accomplish. By having a
animals, as a result of what we consider good clear vision of the behaviour you want from
or bad actions. However, that is not how the an animal and a set of expectations for that
science of animal learning works. We should animal’s behaviour, you can then create a
never reinforce or punish people or animals. training plan to achieve those goals and vision
We should only aim to reinforce or punish (MacPhee 2008). Training new behaviour and