This thread looks to question and evaluate the efficacy of positive reinforcement (training solely with reward – zero correction) when faced with an existing self-rewarding behaviour, something that is performed entirely for the pleasure of the performance in itself. Although tongue-in-cheek, the ‘orgasm’ analogy is used to empathetically illustrate the point – something which gives a Very powerful, pleasurable internal consequence for no apparent extrinsic gain. There is no reason whatsoever for an orgasm to exist, unless perhaps one considers the possibility that evolution has placed the human animal on a random reinforcement schedule in order to keep the associative behaviour (reproduction) strong; though this has been hotly contested. (There also exists the theory that the contractions lend a ‘helping hand’ to reproductive fluids however this too, is neither proven, nor relevant!)
For those readers unfamiliar with the concept of reinforcement schedules, the point i’m making is that an animal is far more likely to perform an action for which they often receive little or no reward, but sometimes get very well rewarded, than they are for an action which always earns an expected reward. In never knowing which performance will earn the desired reward, the animal tries harder, for longer. The same is true of the addicted gambler (excuse the cliched analogy)!
So rewarding is this behaviour that for many humans it has become more powerful than the intended consequence, the result; the child. In fact, the seeker might consider that consequence to be incredibly punitive; yet still they roll the dice. In doing so, it might be accurate to suggest that some might appear to be ‘chasing’ the orgasm.
Here’s the artistic bit (wink!) – What when the chase ‘becomes’ the orgasm?
Enter the dog!
When a dog engages in a self-reinforcing (intrinsically rewarding) behaviour, it does so for nothing more than the internal reward on offer for doing so. Examples include idiopathic barking (of unknown cause), excessive digging, ‘humping’ inanimate objects (also common in bitches), rolling in fox faeces or carrion, and chasing.
Yes, that’s right; chasing.
I’m sorry but you are obviously wrong. ‘Chasing’ forms part of the predatory pattern in dogs. It is a step within a chain of steps aimed at catching prey. Therefore it is performed in order to catch prey, not as an end in itself. The dog is trying to catch something and that is what drives the behaviour, the reward of the capture. Your synopsis therefore, is incorrect.
It is true, absolutely true that so far as knowledge permits, the chase is a step within a sequence of steps to take the dog from a state of hunger to satiated, whether for itself or conspecifics. However, the point to make is that there is a huge underestimation amongst owners, trainers and those who rant from their self-constructed platforms of moral superiority, regarding the intrinsic strength of this behaviour. Once released, once rehearsed, chasing is far beyond ‘fun’. It is certainly beyond a step, existing solely within the chain of prey acquisition.
It is much more than that.
For a great many dogs, chasing becomes their raison d’être; their everything. They spend their walks seeking out chase objects, targets (known as ‘triggers’ or ‘releasing mechanisms’ for Fixed Action Patterns), forever watchful; almost hyper vigilant. Sadly however, their ‘addiction’ carries ‘fallout’ which is detrimental to their welfare as they shift from pleasant companion to a nuisance, requiring a life of constant anxious supervision, environmental isolation, long lines and verbal corrections.
Is this in the interest of the animal? Is frustration ( through prevention ) more humane than education? Does restriction equal freedom, or do restriction and prevention enhance frustration, thereby contributing to a reduction in welfare?
Allow me to use a fictitious example of a dog which chases squirrels in a local park. The dog is 3 years old whatever-breed, and has enjoyed and rehearsed this behaviour over a two and a half year period. For the owner, it has mostly been non-problematic. The squirrels always escape and for the most part, the dog will recall successfully, having had a solid foundation in reward-based recall training. It’s only when she sees a squirrel or is engaged in chasing after them that the recall falls apart. She loves food, but isn’t too interested in it when she’s out and about, preferring instead to explore and enjoy her surroundings. She loves chasing balls and anything thrown, but she will invariably either take the ball with her as she tears after a squirrel, or ignore the ball altogether. Once she has finished her chasing of squirrels, the recall responsiveness returns to full strength and the ball-play is suddenly fun again.
Okay so far …
Recently, the owner reports that the dog has started heading off for a particular area of the park (associated with chasing squirrels) the moment the lead is unclipped, even though there are no squirrels in sight. The young dog takes off, shutting down to recalls until she has scoured the hunting grounds and satisfied herself that there are no squirrels present. Sometimes this can seem like an age, but in reality is probably only 5 minutes or so. Over the past 6 months or so, she (the dog), has also started chasing local cats along the fence line at home whilst barking franticly at them, rather embarrassingly this includes bedtime toilet trips and the neighbours have begun to mention her ‘guarding’ the garden during hours of darkness.
In these situations too, the recall goes completely out of the window.
In two and a half years and over the course of hundreds of chases, the dog has never caught either a squirrel or a cat.
Although fictitious, this story is played out for real thousands of times the world over.
When you actually look at it, it defies what many believe to be Universal laws of psychology and learning, that (all things being equal) any behaviour which results in reward or an improvement in the current state of affairs will likely be repeated under similar conditions. Conversely, any behaviour which results in a worsening of one’s state of affairs will incrementally reduce with confirmation of such a worsening following repeated experiences.
In a nutshell – don’t repeat behaviours which result in bad consequences but repeat those from which you benefit!
It is essentially this process which drives learning and therefore training, however there are exceptions; the orgasm being one and chasing being another. Although world’s apart, each is identical in the fact that neither results (nor needs to result) in measurable extrinsic reward, yet both are at the very pinnacle of the reward pyramid at that specific moment in time, in that specific context and for that specific individual. What I am essentially saying is that for both, at that specific moment, the environment is holding all the aces in terms of reward potential.
We are now free to dispense with the orgasm analogy, for it is a benign activity, and therefore negates any modification requirement, or further discussion! For a trainer following the ‘laws’ of learning in their entirety, this canine ‘problem’ too, is a non-problem …. How do we deal with a potentially hazardous behaviour for which the reward is not extrinsically acquisitive, but internally incomparable? We alter the internal, associative emotional state. Simple. That which felt good, no longer feels so. But that which formerly felt inadequate by comparison, has now gained a reward capacity beyond anything previously experienced!
What follows is a list of the most commonly cited ‘advice’ received by client owners of committed chase junkies from people or businesses who market themselves as ‘trainers’ or ‘behaviourists’ … These are NOT fictitious:
- Keep the dog on a lead – avoid
- Walk somewhere else – avoid
- Carry ‘higher value’ rewards* – impossible
- Go back to training in a sterile environment and expose your dog to the chase-provoking stimulus in a more controlled manner – avoid
- Re-home the dog – avoid
- Euthanise the dog – avoid
- Work on your recall more – avoid/blame
- Become more rewarding/animated for the dog – avoid/blame
- Say “uh uh” or “Bad dog” and withhold reward for failing to respond straight away – absurd
- Consider drugs (for the dog) – mask/avoid
- “Premack” the chase behaviour … Essentially, reward the dog for recalling by allowing her to return to chasing – GREAT for vehicles livestock and joggers ….. – absurd, wishful thinking
- Positively reinforce an incompatible behaviour … (Tell the dog to ‘sit’ when she see’s a squirrel and reward sitting with food/toys etcetera) …. And if they choose not to comply? – NEVER correct a failure – as per 11
- (This is a personal favourite of mine in terms of total b*** sh** clueless academic staples) … Put the behaviour (chasing) under ‘stimulus control’!!! … What this essentially means is, use reward to teach the dog only to chase when you command it to chase, but do NOT, EVER, use correction to achieve this aim .. – as per 11 and 12
- DRO …. Differential Reinforcement of Other behaviours …. Meaning – reward everything and anything other than chasing, but NEVER correct chasing …. as per 11, 12 and 13
- Have more lessons – avoid at my gain and your loss
- Book a block session – ditto
- Buy my book – speechless
* I HATE this stock-answer because there is no higher value reward!
What soon becomes glaringly obvious when reading this (absolutely true) list of advice, is that nowhere at all does anyone suggest correcting the chase in respect of certain triggers. Not one person is prepared to state the most blatantly obvious answer to the problem … Change the perceived consequence.
Once more, the ‘Laws of learning’ (Thorndike’s Law) …. “any behaviour which results in a worsening of one’s state of affairs will incrementally reduce with confirmation of such a worsening following repeated experiences.” [My interpretation – yet absolutely the same].
So why why why why why why why …. Are so many ‘professionals’ claiming the above 17 options as an answer to a problem, Yet not a single one can demonstrate the problem being cured with any of the above answers?
Because it is a lie. Because it simply is not possible to achieve such an aim with such a dog with such prescriptive criteria (non-corrective).
If you read this post and think “I disagree, I have achieved that aim (non-chasing) with that dog (committed chaser) using those procedures (correction-free), and I have the video footage to prove it” (and I sincerely hope that someone does), then I will close both my mouth and my business and seek employment with you.
I look forward to your replies.