|
Training involves the
development of desirable responses and the effective deletion of
unwanted responses. It reflects the principles of
learning theory
which describe the way associations develop between events and the
way to influence relationships between stimuli and responses.
The ways animals learn
to behave to remain comfortable and supplied with resources that
meet their behavioural needs have been systematically evaluated in
experimental studies. Admittedly, the rat is the only species that
has been studied comprehensively but it provides a very useful model
for learning in other species. Whether or not they appreciate it,
the best trainers apply critical elements of learning theory and are
also full-time students of animal behaviour. By studying their
animals’ responses they can predict responses before they occur.
This allows them to improve the timing of cues and rewards.
Top trainers know when
to reward their subjects. This may sound simple but it is a critical
skill since giving rewards too frequently stops animals from
developing improved responses whereas being too stingy causes them
to lose interest and motivation. Similarly, removing rewards is
pivotal in any extinction programme designed to eliminate unwanted
responses. We can all see how to stop giving titbits that are linked
to unwelcome behaviour (such as begging) but the best trainers can
work out how to remove rewards that are being delivered
unintentionally for responses that were never specifically trained
(such as attention-seeking behaviours).
Trainers must also
know how to reward their subjects. To do so they need to be
aware of the current priorities of their animals and possibly how to
increase their drive for a given resource. For example, sniffer dogs
trained to find explosive for food rewards are fed their rewards
close to the spot where they smelt the explosives. This maintains a
strong link between the learned stimulus (odour of explosive) and
reinforcement (food). Additionally they are given no food unless
they find explosive. So, hunger maintains the motivation to find the
odour of explosive.
The label used to
describe all the resources for which an animal has evolved to work
is primary reinforcers. So, a tasty fish given to a dolphin that has
offered a desirable response is an example of a primary reinforcer.
Training or circumstance can link such innately rewarding resources
with novel stimuli that become what we call secondary reinforcers.
An example of a secondary reinforcer is the clicker used in dolphin
training to tell the animal to expect a fish for the desirable
response it just made. So, the presentation of the secondary
reinforcer is a way of rewarding the animal before a true reward can
be delivered to it. This system allows animals to be trained at a
distance and in the absence of apparent primary reinforcers which
sometimes have a distracting effect on their eventual recipients.
Even if you are only
training a dog to shake hands, we recommend that you become familiar
with basic learning theory. In the meantime, here is a list
of some of the characteristics that distinguish top trainers:
· Train one response at
a time.
· Train one response for
one stimulus.
· Consistency means
using uniform cues for responses and not blurring one's signals.
·
Timing is the critical
factor when developing associations between responses and rewards
and responses and their intended cues. So, poor timing can make
reward-based training ineffective. In contrast, poor timing in
traditional negative reinforcement training can amount to abuse.
·
Shaping relies on the
reserving of reinforcement until an improved response appears.
· Secondary reinforcers
are most effectively established when presented before or up until
the presentation of a primary reinforcer.
·
Classical and
instrumental conditioning sometimes conflict but good trainers often
combine them with excellent results.
·
Punishment can be
effective when the punishing agent is only mildly aversive. Having
said that, aversive stimuli should be used in training only with
great care since they can rapidly cause a decrease in motivation and
creativity while effectively undermining the human-animal bond.
·
Once an animal has
learned to fear a particular stimulus, it will also show fear of
other similar stimuli.
·
Fear often disrupts
conditioned responses.
For
a detailed discussion of the techniques used in animal training we
recommend that you look out for a forthcoming book called Carrots
and Sticks.
The overall goal of this book is to take
the mystery out of training by unpicking the various processes by
which animals acquire novel behaviours. It
explains the science underlying the principles of learning theory in a
straightforward manner in and, in doing so, exposes the most
effective, gentle, cunning, insightful and cruel ways to make an
animal perform. It also contain a range of case histories to
illustrate how the basic principles have been put into practice by
trainers of companion animals,
exotic animals (used mainly for human entertainment) and working
animals. At the very end of the book we have provided a glossary of terms that
helps to explain technical terms from the worlds of psychology and
animal training.
go to Learning
Theory |
