The Good, The Bad And The Ugly
Conditioned Aversive Stimuli Avoidance conditioning; a warning signal
In the canine reinforcement analysis we looked at all the things that the dog liked and would work for. On the other side to positive events we have negative events. Things the dog will avoid and escape from. We could list these negative events and pair them with a new special signal. A timely conditioned positive reinforcer tells the dog, “that we liked that version of the behaviour and here is something for your trouble, so do it some more” . You can also establish a conditioned aversive stimulus, which communicates ” I don’t like that version of your behaviour, it is not good and something you don’t like will happen unless you stop”.
Using positive reinforcement as your main teaching tool does not mean that you cannot use a “NO” signal when you need to. If the dog is doing something that is NOT SAFE TO DO, a timely “stop that” signal is what we need.
In avoidance conditioning, if behaviour can prevent a negative stimulus from occurring, the behaviour increases in frequency. In order for this to happen the dog must learn that the negative stimulus is coming and must know what he can do to avoid the negative stimulus. This procedure requires conditioning, just like click (positive) means something nice is coming, a new (negative) sound means something you wish to avoid is coming.
The late John Fisher’s Dog Training Discs have become very popular around the world, as the use of these brass discs makes a sound which is unique to the dog and one that has been conditioned with the removal of something desirable (NEGATIVE PUNISHMENT), in this case, food! Any new stimulus (sound) can be used to condition the dog that something desirable will be taken away.
There has been a great debate about using a non-reinforcement marker signal. It has become popular among dog trainers. The use of words, such as “wrong”, “oops”, “try again” “too bad” are often spoken in a neutral tone. The idea is that we can give the dog some extra information, by using a signal that signifies, “that will not be reinforced”. B.F.Skinner’s definition of punishment – taking away something-desirable -means the “wrong” signal is a conditioned punisher, since it means that reinforcers are not available. For experienced trainers who have a dog with a vast repertoire of fully shaped behaviours and cues, a dog that keeps throwing variable behaviour at you establishing a “wrong” signal could be a useful tool. They can condition the dog to understand the sound as meaning simply “stop what you are doing as you are going down a path that leads to a dead end”. Or put another way, “no reinforcement is available for that version of the behaviour”. When giving this “wrong” signal it has to be communicated to the dog “immediately and with certainty” when the undesirable behaviour occurs. Any delay will punish the wrong behaviour.
Punishment, the debate goes on
When the dog is in a situation in which reinforcers are readily available and he performs the undesirable target behaviour, you remove the dog from that reinforcing situation. The dog should not be put in a time out for a long time. Usually several minutes is most effective. When dogs are given a time out and there is no supervision they should be put into a safe, secure area such as a crate. Or put into an environment where no harm can be done to the dog or the surroundings. This is known as a ‘time out’ and is used to reduce the frequency of target behaviour by removing the dog from a reinforcing situation. Time out, when used effectively, can be a punisher. If the crate is over used for time outs the crate itself will become a punisher. If this happens the crate will not be a safe, secure den for your dog. For the time out to be effective, the original environment must be more reinforcing than the time out environment.
I suppose the crux of the matter is to know what is a reinforcing environment and what is punishing environment. The definition for punishment is the procedure of providing consequences for a behaviour that decreases the frequency of that behaviour. If a dog likes lots of attention, time outs are likely to succeed. If your dog likes lot of rest and sleep, then time out may not be an effective punisher.
Using punishment does not teach the dog what is safe to do, it only teaches the dog what not to do. One side effect of punishment is that it produces undesirable emotional reactions, particularly fear and anger. Fear may take the form of efforts to escape or avoid anything associated with the punishment. This is undesirable, since escaping or avoiding an aversive event is going to be negatively reinforced! Murray Sidman PhD., a behaviour analyst who is an expert on punishment- and one of its severest critics, likes to say that people who use shocks become shocks. You become a conditioned punisher because you are paired with the punisher, and if your dog cannot get away from you he may attack you. Even small puppies can paw, bite and struggle with the owner as he attempts to punish, and if the owner backs off, the owner is very likely to reinforce this aggressive behaviour. You must also be aware that using anything as a conditioned punisher may easily have other consequences. I heard a story recently that a six-year-old dog had bitten a city gent on the thigh while being walked down the street. The dog suddenly attacked for no good reason. It was later said that the dog had apparently attacked unprovoked! This dog had never shown any signs of aggression and after a health check was found to be quite a friendly dog. On further investigation the man in the suit was holding a rolled up newspaper and was hailing a cab by the side of the road. This dog’s early learning included being threatened with a rolled up newspaper.
So if you use punishment you may find that there are toxic side effects, as you cannot expect those animals whose behaviour you punish to be glad to see you coming, particularly if you use some form of physical punishment.