TOTAL RECALL UNDER FIRE ……. What follows is a direct lift of a discussion taken from my Linkdin page, please add any comments you may have!

TOTAL RECALL UNDER FIRE ……. Is it possible to effectively modify a heavily reinfornced (completed) predatory sequence using TOTALLY punishment-free procedures outside of the training hall?

27 days ago

 

18 comments

Edward Meakin • Jamie, Whilst I would not advocate pain as a punishment to my clients, I have to concede that totally punishment free methods to modify predatory sequencing are not entirely effective in the case of experienced sheep chasers/killers.

Strong instincts are extremely difficult to control even if the dog has never carried out a chase kill sequence. Instictive drift is a powerful phenonenom and whilst in this state the dog is not open to learning.!

Keeping a dog away from livestock, or keeping him on a line is not training, It only serves to manage the problem. This is not always possible for some dog owners. I have seen some dogs’ behaviour modified by use of a spray collar in the hands of an experienced trainer. This is punishment but not does not cause pain. True, it startles the dog and interrupts the unwanted behaviour, and you have to ask yourself whether this is justified. The alternative in the U.K. is for the dog to be put to sleep or shot by an irate sheep farmer.

This then begs the question is the use of an electric shock collar also justified.?Research suggests that pain, if used at the correct level, can be an effective tool in training. Often a single event shock can stop sheep chasing in its tracks. I ask the same question. If the alternative is euthanasia or the dog being shot, can it be justified?

I use motivational methods in all my training and do not use shock collars, but I have to honest and say that punishment can be effective.

Punishment does not have to be painful. Removal of something pleasant is also punishment, and is a useful technique. My main problem with pain as a punishment is that it is often used as retribution for a perceived misdeed. If ‘trainers’ have to keep using pain for the same training issue then that stops being punishment and moves into the realms of abuse. One of the more worrying side effects of pain used as punishment is aggression.

If a dog understands what it is required to avoid pain, it has control over the situation and forms part of operant conditioning. In the wild, dogs often learn from puishment e.g. the pungent spray from a skunk will ensure the dog will forever aviod further encounters with that particular animal, or if the dog bites into a toad and the poison makes him vomit and burns his throat he will do likewise.

I am open to any methods that would be equally effective to punitive ones in the case of livestock worrying.

Thank you for your time, Edd

26 days ago
0

David Thatcher • Anything is possible.

If the question is would it be easily done my answer would be “it depends”.

26 days ago

David Thatcher • Always possible. Elaporate a little.

26 days ago

Jamie penrith • Thanks for your input Edd! Hi David, we may have touched on this discussion (or similar) on your very thorough NRM discussion.
Okay, to further ‘fill the pie’ …. let me firstly give a very recent example of canine interaction that I witnessed some 45 minutes ago between my lab’s. One dog >24 months and one bitch; same age. Not siblings or in any way related, ownership of bitch since 8.5 weeks and dog rescued pre-centre appointment at 11 months. They live, sleep, eat and play together and have done so for 11 months. Today, SHE had a rawhide chew outside on the deck (escape/avoidance option open), HE, chose to approach and try to remove it from her without any form of threat display. SHE, immediately turned and fired into his flank with bared teeth, heavy vocalisation and determination. Her bite – inhibition is marvellous, a 5 second flurry of attack without a mark on either dog. TOTAL punishment. 5 minutes later, both dogs are laying, grooming one another before sleeping side by side on my sofa. No fear, no anxiety, no resentment, no avoidance; no confusion. She had communicated a consequence with pinpoint timing and delivered it a sufficient a level for it to be effective in her intention, without any ‘fallout’ (Sidman)!
As I watched this in awe (not of punishment, but of communicative timing and restraint), I thought of this post and whether it is truly possible (David – realistically achievable within a clients time-frame and budget) to achieve eradication of a full,determined and reinforced predatory chain, devoid of ANY punishment?
Here’s a link to an ‘expert’ (not confirmed, but her credentials seem sound) in sheep chasers/killers …. http://www.dogpartnership.com/livestock.htm.
You will see that HER method of choice as a total reward based, positive method, punishment denouncing trainer, is to have a sheep which will ‘stand its ground’ – quote:
” However, allowing a dog in a field with a sheep who will stand it’s ground can have a beneficial effect. In a field, the dog has the opportunity to make the right decision – TO LEAVE!” – unquote.
Now, why would a dog in these circumstances who is a confirmed chaser, choose to leave? It shall be argued that it is through R- no doubt. But why has something that has been such a R+ for the dog, acquired R- properties? For one reason alone. The fact that the sheep will “stand it’s ground”, implies that the sheep is to be used as a P+. There is nothing here to suggest any marking or rewarding of incompatibility or ignorance on the part of the dog. The inference to be drawn is that the sheep will “stand its ground” causing the dog “TO LEAVE”.
What I find hypocritical and dishonest in this example, is that the trainer no doubt truly believes that their hands are clean of any punitive interjection; the SHEEP did that. The trainer is there simply to reward correct responses; avoidance – acquired and maintained (initially) through fear (perhaps). Yes, we can then reward and reinforce to the extent that the avoidance behaviour is maintained through positive reinforcement, or so we believe! But the dog will never forget the initial training, so although we might like to believe that R+ principles are to credit, it is R- (sheep avoidance) that are at work.
So let us then propose that we choose not to take this route. Instead, we are going to shape an acceptable or incompatible response via mark-reward, quitting for the session whilst the dog remains sub-threshold for the chase sequence, whilst incrementally decreasing distance thereby shaping both distance and response right?
Okay, a few questions.
* How do we know what the threshold is? We need to establish it; yes? So we restrain the dog using a humane restraint and ascertain his/her reactivity point; yes? Okay, what has happened here …. we have introduced an appetitive stimulus thereby triggering the dogs ‘seeking’ behaviour – we’ve switched him on. What have we proceeded to do next?

continued ..

26 days ago
 

Jamie penrith • continuation …
We have removed either the sheep, or the dog. We have negatively punished the dog. Not only that, but with the introduction or wearing of the restraint of choice, we have introduced (without choice preference testing the dog) a positive punisher. Even if the dog is conditioned to readily accept a restraint under normal circumstances, at that particular moment, when his threshold is triggered and the ‘chase is on’, it is an item that the dog would choose to avoid/slip/escape – Without removing the restraint we are continuing to punish.
This, I believe (as mentioned by Edd), is where the training rift manifests itself. In a fundamental misunderstanding, misinterpretation and/or misrepresentation of the word PUNISHMENT. Punishment is not pain. Punishment is not abuse. Punishment is not necessarily a negative quadrant (in either form) of operant conditioning. HUMANE does not equal without punishment. WELFARE does not equal without punishment. Kindness (I don’t really use the word kindness as I associate it as meaning a combination of humane and fair? It is kind to give an old lady chocolates, but that makes the reward non-contingent and not really relevant or applicable in my opinion)? FAIR does not equal without punishment. FAIR implies choice, in both behaviour and consequence.
There are a couple of other issues to briefly raise before I step aside!
For R+ to be effective, there needs to be a gap; a need or desire. For there to be these things it is a requirement that there be a feeling of need. I can’t reward train a non-hungry, lethargic, socially withdrawn or highly aroused dog (very effectively). There are certain needs i must evoke for training to be successful. What am I doing by instilling a need? I am either removing or withholding, right? Which of these two is non-punitive? If I wish to ‘up criteria’ (and this is in a trained behaviour, not a modification protocol), I ‘up’ the requirement that earns reward. Equally, I withhold or remove the reward from a previously reward-earning behaviour. INITIALLY, I negatively punish the dog. It could be argued also that by eliciting frustration; I am positively punishing the dog could it not?
Back to sheep ..
Could it be, that counter-conditioning attempts whilst shaping alternative, acceptable behaviours around sheep, (even if successful) will never achieve real standalone/remote reliability? For example, if my dog has learned that the presence of a startled flock equals treats from dad for a sit/stay and focus – what when dad isn’t there? How can we (if at all) achieve fluency of response with no owner alongside, using R+ methods alone?
I have never seen any evidence whatsoever of a true chaser having their reinforced chase behaviour modified, through R+ methods alone. AND THIS IS OUR PROBLEM AS R+ TRAINERS, there are hundreds of evidenced examples on the web, (scientific and P+ trainer achieved) of punitive methods achieving said long-term (often lifelong) fluency.
PLEASE, link me to a before-and-after video or study ANYWHERE (not anecdotal) of this having been achieved within a realistic (sub ten lessons) time period? comments??

26 days ago
 

David Thatcher 

David Thatcher • Interesting Jamie,
There are somethings that raise my eye brows. I know it is impossible to be extremely detailed here in this forum so I am thinking this is where the issue lies. From what I’ve seen you write before I know you are well versed in behavior.

First the dog who decided to resource guard his bone. This is a normal dog behavior. There is no need to train anything. If a resource guarding dog seeks out to harm or had no bite inhibition then we train. This wasn’t the case so stand back and let them sort it out. Secondly, while the dog showed “no anxiety”, resentment and confusion aren’t behaviors. These are emotional states. So there were no conflict behaviors and no appeasement behaviors. I try to stay away from Anthropomorphism because it can get me in trouble when trying to diagnosis. 

My opinion is Unwanted Behavior should be defined in a few different categories. We have the “behavior modification” category which covers behavior that dog owners do not like and want changed in their dogs. We have be “reactive/ aggression” category for dogs who have conflict with other dogs, people, etc and are in need of a trainers assistance. Finally we have a “clinical” category for those dogs who need medical assistance coupled with a behaviorists written protocol for course of action to work through their issues. 

The reason I said “it depends” on your “Total recall under fire” was there wasn’t enough information to determine which category it went in and I needed information to determine a recovery protocol. Some examples of other information will include if the dog has other issues beside this one? Does the dog take treats while under threshold? Whether or not there are known triggers? Are there multiple triggers? Since all aggression is fear based, what is the dog fearful of? What are the known reinforcers? What is the given history of the behavior? As you can see there are more questions than there was information on the topic. I believe trying to post these types of examples are difficult to get all the information possible. So depending on some of the answers above is how I would work out a training protocol. I do believe R+  is not only effective in cases which involve herding dogs and sheep but just as effective as any other method depending on the answers to some other questions. The only route of punishment that possibly would be used in cases are dogs who do not take treats under threshold. Then using C.A.T. or B.A.T. protocols will work sufficiently. These protocols are P- and in fact very similar to the rest of your post on the topic.

Wags and Wiggles,

David

25 days ago

Jamie penrith • Hey David! Thanks for the reply!
I’m not looking to ‘prove a point’, rather i’m looking to see whether alternative arguments have legs!
It seems that the answer to the discussion might simply be – sometimes.
That’s fine, that’s honest.
It’s the polarised opinions of ethically charged individuals/groups that attempt (either way) to put a smokescreen over truth, which, (i feel) do more to support ego and doctrine, than address the issue at hand in the most appropriate way possible!
And on my final point reagrding evidence? Do you know of any?
Time/financial impact factors?
I enjoy discussing with you mate! And Terry and Edd!
Warmest wishes,
Jamie

25 days ago
 

Jamie penrith • Final point – promise!
Remote reliability – handler/trainer/owner absent. Replacement response reliability in this circumstance?
I’m thinking of trained animals for entertainment purposes having to have a trainer permanently present to cue-Mark-reward.
Can a R+ modification protocol for a chaser operate without the presence of the owner (non-contingent reliability)?
A P+/- trainer will argue that their methods can, and there is evidence to support.
Thoughts?
Sorry, I’m a tormented soul!!!

25 days ago

David Thatcher • Jamie, you have a dog and a sheep together and the dog is displaying unwanted behavior. First, if you remove the dog from the scene this is P- for the dog. Secondly, if you remove the sheep from the scene this is R- to the dog’s unwanted behavior.

As far as traditional trainers wanting to discuss whether or not R+ or P+ is more affective in any situation I’ll leave that discussion to them. I am fully confident in my protocols in both resolving issues and maintaining the behavior.

24 days ago

Jamie penrith • Thanks David, I feel you will not respond to this but it’s worth a crack!
You cannot seperate the ‘behaviour’ from the ‘dog’ according to training protocols – it’s professionally unethical and misleading. one cannot operate independantly of the other? Behaviour independant of a dog is non-existent, and a dog stripped of behaviour; is dead!
Removing the dog from the sheep is P- (agreed), the dog wants to remain, to chase.
Removing the sheep from the dog is ALSO P- though – it is what the dog wants at that particular moment in time. The whole procedure is punitive?! To say, ‘ah, but it is the ‘behaviour’ that is being punished’ is, to be honest, indicative of an unwillingness to recognise a glaring flaw in a R+ approach to an appetitive behaviour?
Confidence is reflected in ability and results, ability and results are evidenced.
I ask once more, can you show me a single scrap of non-anecdotal evidence to support positive training protocols acheiving results in a confirmed livestock chaser??
My unfortunate conclusion is that there is NOT a scrap of such evidence, the reason being that such a protocol is unattainable in practise – particularly in terms of remote reliability of a replacement behaviour.
I agree with the earlier comment, restriction removal and avoidance is not training.
I shall post a very detailed explanation, supporting this viewpoint shortly.

24 days ago

 

Jamie penrith • Just to be ABSOLUTELY clear … I personally, am as far removed from a traditionalist as you’ll find.
This doesn’t however, mean that I’ve closed my eyes ears and mind to either discussion, disagreement or factual evidence as it evolves or presents itself. In doing THAT, I am no more honerable or secure than a man with a big stick!

24 days ago

Jamie penrith • “Pavlov’s investigation appears to draw into question the viability of counterconditioning and sit-stay training as stand-alone procedures in the treatment of aggression problems …. theses procedures, as used in many contemporary, applied animal behaviour systems, should receive experimaental and controlled clinical investigation for efficiacy. Despite their widespread use and virtually unquestioned acceptance, neither procedure has been proven to reduce aggression. Although counterconditioning may palliate or antagonise the arousal elicited by aversive stimuli and facilitate the disconfirmation of previously acquired conditioned threat expectancies, preattentive biases in association with trait anger and anxiety may nevertheless persist or be rapidly reinstated despite the most conscientious counterconditioning efforts….Dogs expressing a rigid watchdog scrip may be genetically programmed to resist counterconditioning effects actively…..Despite the reduction in overt reactive behaviour, conditioned autonimic responses may persist or worsen over time via schizokinetic and autokinetic mechanisms…..As a result, even though a dog appears to be more relaxed, it may still remain under the influence of an unstable sympathovagal equilibrium that can unexpectedly shift and trigger aggressive behaviour”.
Lindsay, 2005, Handbook of Applied Dog Behaviour and Training, Iowa, Blackwell, p499.

(the chapter on ‘biobehavioural monitoring and electronic control of behaviour’ ought to be compulsory reading for all R+ trainers who wish to convincingly argue our position)!

24 days ago

 

David Thatcher • I’m sorry Jamie, I’ve done a grave disservice to you on my last post by not going into detail. I posted a shortened version thinking you would rethink your position of punishment. So in this post I will cover 3 items.
1) An analogy
2) What is aggression?
3) Why the removal of the sheep is R-?

I. Why do dogs bark at the postman? This question is asked in all my behavioral classes? When we look we see (1) the postman comes to the door and puts in the mail. (2) This startles the dog who in turn barks. (3) Then the postman leaves because he has delivered the mail. (4) The dog then believes he was successful in chasing away the postman with a barked. In other words the bark worked. The bark is reinforced by the postman leaving R- and when it happens daily the bark becomes a learned behavior.

II. I believe the problem area in our discussion is understanding aggression. The “chase” isn’t the unwanted behavior. It’s the dog being aggression/ reactive towards the other animal. Aggression is a fear based behavior. Again Aggression is a fear based behavior.

Being the owner of a herding dog, an Aussie, going to herding trials, etc. one of the behaviors I have grown very fond of is the different instinctive behaviors of each breed in the herding group. However the “chase” is a primary reinforcer for dogs. The same is true for dogs that chase tennis balls, etc. It’s not the ball, it is the chase. No one should want to take the chase out of a dog because it is a valuable reinforcer to create wanted behavior, R+.

The herding breeds DO NOT like chaos. Their instinctive behavior is to stop chaos and return order. This is what makes them so valuable in being a herding dog. We use this behavior to direct the flock where we need them, keeping the flock together, and other reasons. We give the herding dog the definition of what is order. Then they do the job as we directed. When the herding dog gets to the point where they DO NOT think ORDER is in place and they get over aroused because they “can’t” achieve this order then unwanted behavior appears; whether it is knocking the sheep to the ground or more critical consequences. They hurt or kill because of “fear of not completing the job of order” in a climate of chaos. When they hurt or kill while alone it is because they say what order in their own minds is and when they can not achieve it they go to the extreme. Once this is practiced it becomes a learned behavior.

III. So why is removing the sheep R-? Because the dog gets aroused when the sheep enters the area and since we have a learned behavior that this means chaos the dog gives us unwanted behaviors. The dog needs “order” and wants to create order. When the sheep leaves the area the dog is no longer aroused because order is “restored again” from the sheep leaving. So the behavior the dog was displaying when the sheep left was R- because the dog believes his behavior created order again

david

24 days ago

Jamie penrith • Wow, that’s thorough!
But I’m still missing any evidence to support the hypothesis, other than spoken or written word???
Please re-read my stance in relation to punishment David. I have no ‘position’ – that would anchor me to a particular set of protocols and beliefs, which in turn would narrow my perspective and open-mindedness. In practise, I reinforce; but in an advisory/consultatory capacity, I am obliged to speak honestly about the practicalities and limitations of any quadrant when utilised in a stimulus rich environment.
I find it difficult to accept an ‘order from chaos, therefore I am acting in the interest of the animal to remove chaotic stimuli’ theory. That is not to say that I reject the notion; just that I require actual evidence to support any such belief – can you offer any??
To say that all aggression is fear based is incorrect. There are not only neurological disorders and chemical imbalances as causatory factors, there is also the phenomenon termed idiopathic aggression right? (cause unknown).
Is birdsong or insect posturing either aggressive or fear based? Both serve to secure access to a resource or requirement, but aggressive? Aggression is a term which I consider to be as readily used and misdiagnosed as dominance. I agree with you, much if not most behaviour which serves an acquisitory end (including defence for safety) would indeed appear to be fear based, but none of us are in a position to categorically state ‘all’?
Correct me if I am misinterpreting your former post, but in laypersons terms, it would appear to be a roundabout way to justify avoidance of trigger stimuli – which, is simply to avoid the issue at hand? Is that correct?
A typical premack approach might involve control over the reward of which availability becomes contingent on incompatible or ‘appropriate’ behaviours right? How do you apply this principle to a chase-grab-kill sequence?
You mention herding breeds in your example, but the predatory sequence is truncated in such breeds pre-grab (save heelers) kill bite. Therefore it is acceptable to propose the ‘chaos hypothesis’. But what of the fighting/hunting breeds? A foxhound does not seek to restore order from chaos for psychological homeostasis, any more than a pit bull seeks order when chasing a cat? Terriers do not seek to round up rats or squirrels, they are acting within the instinctive parameters if their genetic predisposition. The behaviour is intrinsically rewarding, not motivated through fear/avoidance of a chaotic state?
I too, can only hypothesise however, as we all lack that one element which we would all love so dearly – to ‘talk dog’!
In closing, I ask again,
“do you know of ANY recorded or scientific evidence to support the effectiveness of a R+ approach in effectively treating sheep chasers with a reinforcement history?”
Without evidence, we have little more than an unconfirmed hypothesis, and who but the easily persuaded can be expected to accept that?!!
🙂
Jamie

24 days ago

David Thatcher • Jamie wrote:
“I ask once more, can you show me a single scrap of non-anecdotal evidence to support positive training protocols acheiving results in a confirmed livestock chaser?? 
My unfortunate conclusion is that there is NOT a scrap of such evidence, the reason being that such a protocol is unattainable in practise – particularly in terms of remote reliability of a replacement behaviour.”

Jamie just because you haven’t read or heard of a single case doesn’t prove anything. This position speaks more about you than any method. In fact this position isn’t science. It’s humorous. If memory serves me correctly you are the one who says there are NO ABSOLUTES. =)

There is evidence if you are only willing to look for it. With that said I’ll leave this discussion. I look forward to dialogue in other formats.

Cheers,

David

24 days ago

Jamie penrith • As suggested!! 🙂
Thanks for you thoughts and discussion David,
I feel disappointed that my ‘posiyion’ is attacked when quite evidently, I don’t anchor to one!!
Warmest wishes for a happy future
Jamie!

23 days ago

Jamie penrith • For the information of anyone else reading this post, here is my reasoning behind it:

I am a R+ (positive/reward-based trainer) who is extremely interested in the psychology and physiology of behaviour as it relates to human:dog dog;dog relations. I have never been an unconditional sheep! (excuse the pun)! In deed, perhaps; but in mind – no. I believe it to be neither wise nor honest (to myself or others) to accept ANYTHING on face value. I have spent the past >16 years listening to accounts and theories of events, and then seeking out the truth which might support or disprove them – in the form of EVIDENCE.
Without evidence, I am a theorist. I must be able to evidence these theories if I am to convince others of their potency; and to convince myself. If I wished to make a living from deviating or distracting from direct questions, I would have chosen to sell fake watches.
To declare that “there is evidence if only you are willing to look for it” is the sort of thing I tell my children in relation to garden fairies?!
If together, as R+ trainers we cannot direct and evidence our ethics and procedures to one another; we are in trouble!
As it stands, in the absence of any real evidence (in the presence of which I will absolutely detract everything proposed herein), we are just no more credible than a traditionalist when dealing with ‘chasing’.
Without factual evidence, our proof may just as well be having tea with the fairies and the goblins – THAT, unfortunately, is unscientific and humorous 😦
Someone ‘please’ direct me to a video or scientific study on the matter?

23 days ago
 

Jamie penrith • David Thatcher says:
” If memory serves me correctly you are the one who says there are NO ABSOLUTES. =)”

Close David! What I said was, “Nothing – taken in isolation – is absolute”.

This statement is what supports ‘my position’ on this discussion. Namely that R+ (as a singular course of action) CANNOT be absolute; it is being taken in isolation.

There must be a counterweight; balance. Having never experienced punishment, I cannot experience reinforcement.
cheers!
Jamie

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s