"Conditioned Aversive Signals
A timely conditioned positive reinforcer tells the recipient, "What you are doing now is good and will
gain you something, so do it some more." You can also establish a conditioned aversive, or punisher, which
communicates, "What you are doing now is not good, and something bad will happen unless you stop."
Conditioned aversive stimuli are more effective than threats. Some subjects—cats come to mind—are
unresponsive to shouts and scolding. However, a friend of mine quite accidentally cured her cat of clawing
the couch by establishing "No!" as a conditioned aversive stimulus. One day in the kitchen she happened to
drop a large brass tray, which fell right next to the cat. She cried "No!" as the tray fell, just before it landed
with a loud clatter. The cat, dreadfully startled, jumped into the air with all its fur on end.
The next time the cat clawed the couch, the owner exclaimed "No!" and the cat, looking horrified,
desisted immediately. Two more repetitions were enough to end the behavior permanently.
Reprimands are a necessary part of existence. Using positive reinforcement as your main teaching tool
does not mean you cannot use "No!" when you need to, for example when the baby pokes at the wall outlet.
However, some trainers use this real-life circumstance to justify their own general and abundant use of
"correction" in instruction. In doing so they make two mistakes. First, they view correction as if it were
equivalent in value to positive reinforcement, without taking into account the other effects it has on the
learner (see "Punishment," Chapter 4); and second, they use those reprimands and punishers without
establishing a warning signal, or conditioned aversive stimulus.
The trick to making "No!" effective is to establish it as a conditioned negative reinforcer. For example,
anyone who feels it necessary to use a choke chain on a dog should always say "no" as the dog does the
wrong thing, and then pause before yanking on the chain, giving the dog a chance to avoid the aversive by
changing its behavior. To just yank on the chain without a warning turns the yank into a simple punishment,
with no predictable effect on future behavior and a potential cumulative effect on the dog's willingness to
work at all. A third popular error, continuing to yank while the dog is back in position, simply punishes both
behaviors.
Failing to use a conditioned"