5. Negative .

Please don’t distribute this material without my consent.

Chapter 1.

In the Humane Hierarchy, we are biased to looking for behavioural solutions on the Bright Side. However, understanding the Dark Side (negative reinforcement, extinction, negative and positive punishment) is equally important. Perhaps more important. Chapter 2.

The term negative reinforcement (R-) implies that behavior is reinforced, and that something is removed from the environment because of that behavior.

As the animal figures out which behavior works, exposures to the aversive stimulus tends to be shortened. If there’s a warning signal preceding the presentation of the aversive stimulus, the animal soon learns to avoid it altogether.

For instance, you put a hand on your dog’s bum. When he starts sitting, you remove the pressure. Gradually, he’ll respond sooner and sooner, and finally when you raise your hand in that particular way that he’s learned to associate with the pressure.

Intentional R- procedures look very different from intentional P+ procedures. The aversive stimulus is applied regardless of what the animal is doing (as long as it’s not the desired response), and is initially very low-intensity so as not to scare the animal, and very soon preceded by a warning signal. Some trainers maintain the low-intensity aversive throughout, others slowly escalate if the animal doesn’t show the desired response. Any approximation of the desired response terminates the aversive – which is an antecedent.

Intentional P+ procedures, in contrast, occur contingent on undesired behaviour, are highly aversive and unannounced (if not, the aversive stimulus, which is a consequence, is not an effective punisher). The aversive stimulus is applied until the animal does any behavior but the undesired one.

Any time an aversive stimulus occurs, behaviour may be positively punished, negatively reinforced, both, or neither. Chapter 3.

For instance, say that Dooby Scoo chases a rabbit. Somer Himpson starts yelling. Dooby cowers, and Somer stops yelling.

Somer’s intention may have been to punish the rabbit-chasing behaviour, but Dooby’s perhaps more likely to have learned that cowering is an effective way to stop the yelling.

Cowering has been negatively reinforced, because that behaviour terminates an aversive stimulus.

A vicious circle, since Dooby may well keep chasing rabbits. And Somer will keep yelling.

Chapter 4.

Negative reinforcement is the most misunderstood concept… perhaps in all of .

In a group of dog trainers, only a little over 1/3 could define it properly.

Perhaps this confusion, along with the connotations of the word ‘negative’, explains why it’s so stigmatized.

Chapter 5.

Since negatively reinforced behaviour is all about removing an aversive stimulus, animals (and people!) only perform well enough to satisfy the minimal requirements (the Have to do curve below).

In contrast, when using positive reinforcement, the sky is the limit and performance may be substantially better – the discretionary effort gives an altogether different outcome (the Want to do curve).

Chapter 6.

Negative reinforcement may include Escape responses (which is not running away but terminating an aversive stimulus), or, once the animal has learned to anticipate the aversive stimulus by paying attention to predictors such as commands, the behaviour is called an Avoidance response.

The problem with Avoidance is that we no longer see the aversive stimulus, so it can be hard to determine how a specific response was once learned. However, given enough time, we may see the Avoidance Paradox whereby the command no longer is a reliable predictor of aversives, and therefore undergoes respondent extinction.

Chapter 7.

Adding a correction or a negative reinforcer to a positively reinforced procedure changes the dynamics of the discriminative stimulus, or SD.

Cues (SDs learned through positive reinforcement) are promises, or opportunities.

Commands (SDs learned through negative reinforcement) are threats.

What does that make the SD learned through a combined reinforcement procedure?

Such SDs are called poisoned cues.

Poisoned cues have been shown to impair behaviour and performance in several ways compared to behaviour trained with pure cues.

However, there’s no data on how poisoned cues compare to behaviour trained with pure commands.

Chapter 8.

Animals habituate to low-level aversives and stop responding to them – a phenomenon called punishment callous.

This calls into question whether we should bring in positive reinforcement into a negatively reinforced procedure – this would enable us to keep the level of the aversive stimulus low, thus avoiding punishment callous and the avoidance paradox – and even perhaps the poisoned-cue effect.

Such a procedure could be construed as counter conditioning, which works best if the aversive stimulus is very mild.

Chapter 9.

Negative reinforcement may be conceptualized to move animals diagonally to the bottom right in Core Affect Space (into Quadrant 2). How animals perceive the negative reinforcer will depend on its intensity, predictability and controllability – and whether the animal is performing an Escape or Avoidance response. Emotional responses may range from fear to relief.

Chapter 10.

The use of negative reinforcement may lead to lack of discretionary effort, punishment callous, the avoidance paradox, poisoned cues, superstitious avoidance, passivity, depression, learning impairment, redirected aggression, phobias, chain breakdown, peer pressure, sublimination, reaction formation, combat fatigue, regression and multiple personality disorders.

Chapter 11.

Some of the misuses of negative reinforcement involve bad timing, big guns, not using shaping, not using positive reinforcement, and not providing warning signals (not allowing the animal to Avoid rather than Escape the aversive stimulus).

A miss-use of negative reinforcement is not to use it when helping animals overcome fear. Since well- performed negative reinforcement implies that the animal controls both the start and the end of the aversive stimulation, this has the potential of vastly improving the animal’s confidence in scary situations.

We might consider handing the control to the animals rather than keep the animals under our control.

Chapter 12.

Aversive stimuli range from excruciatingly painful and terrifying to annoying to barely noticeable.

You may not even notice a single fly, but horse-flies are another matter – especially when there are many!

Aversive stimulation may sneak in, resulting in inadvertent negative reinforcement processes. Animals learn to show aggressive behaviour to keep people or other animals at a distance.

In fact, many problem behaviours are established and maintained by negative reinforcement – and their solution often also is found in negative reinforcement procedures: teaching the animal an acceptable replacement behaviour to get the desired outcome, typically distance.

Training is a conversation. Some trainers are exploring new ways of utilizing such low-level aversive stimulation that the associated SD can hardly be called a command. Rather than telling the animal what to do, those trainers are asking.