Working With Fearful Dogs

[NaturallyWild]

Hi all,

i'm putting forward several questions out there more for the experienced trainers, but happy to hear from anyone with personal experiences.

Following each question is either some of my thoughts or further theory of the question asked or examples in regards to a dogs fear (eg. strangers, but could be any type of training situation, not just fearful).

1: Which type of learning theory is more efficient and gives a stronger result

a- classical condidtioning eg. on appearance of a stranger treats are fed by owner, when stranger leaves treats stop - the dog has a passive role

b- operant conditioning eg. when dog chooses to move towards stranger they are rewarded - possibly use target stick, etc to lure towards stranger

2: Are the following true or actually misconceptions

a- praise and petting reinforce fear or fearful behaviours

Why would doing something nice cause them to become more afraid? Isn't true fear not an active behaviour (eg choosing to do it), but rather an automatic response. Why is it ok to do classical conditioning in fearful situation eg rewarding, but not ok to praise/pet (which for most dogs is less rewarding than food, toy etc)?

eg shaking/whimper on appearance of trigger - praise/pet to stop reaction - does this cause shake/whimper to occur in different situations or for it to become stronger in the next situation? is the dog thinking if i cry and shake/harder this time when i'm afraid i'm more likely to get a pat.

b- punishment is contraindicated (eg increases fear, or hides the reactions)

Is the association the dog is creating being paired with his reaction or with the trigger. Are we changing the emotional state in the right direction? Can punishment be used eg. if dog reacts to trigger then something good gets taken away ie owner, but dog must stay in presence of trigger.

3: Do the following techniques for either classical conditioning (a) or operant conditioning (b), have the down sides of proceeding comments.

a- Positive conditioned emotional response - potentially hardest to get right, due to timing and being persistant (need to always provide reward on appearance of trigger) - can often be harder to control the trigger eg if red bus is trigger it is not that controllable.

b- Luring - false readings if the lure is so much more enticing than what the fear issue is, causing dog to not notice the fearful thing - association is not created

Cajoling/happy talk - may mask fearful symptoms if dog is extremely social or attached (similar to above) eg if afraid of stairs, will bolt up to get to someone, but won't go near of own choice.

Counter-commanding - have a requirement for this command to deal with situation - does not necessarily learn to deal directly with situation (needs to be done as trigger appears, or as a response to stopping reaction, not as a response to starting reaction)

4: Working dogs in regards to behavioural thresholds - should we be working above or below thresholds, in what situations and what are thresholds eg which is above the threshold - stare, growling, lunging, exploding. If working above threshold (because very likely to happen in the real world with the dog and owner and they need to know how to deal with situation) do we try to get something positive out of it whilst still in the situation, or just move away and regroup/retry.

I would like to hear some of peoples opinions/thoughts, experiences, etc.

I personally believe that each situation must be dealt with differently (with an individual program that will often change, particularly as you progress through the fear). I feel that if the dog can take an active role (operant conditioning) then you are likely to get a stronger result, because the dog has "chosen" to make that positive response . There are however the potential draw backs as mentioned which can make it harder to work the problem and have the owners work the problem - eg masking the problem.

Talk away

Tim

[haven]

Great questions Tim, a topic I'd dearly love to discuss in great depth and detail, but unfotunately I'm noo busy at work and don't have email at home

I'll try to post my response when I get a chance, can't wait to see the other replies and ensuing discussions!

[Vickie]

Very interesting post! I'll look forward to reading the comments/discussion.

One thing I think is often an important factor when working with fearful dogs is leadership. I think there are many dogs with fear issues that can be reduced significantly if they trust you, the leader, to be in control. I try to take that approach as much as possible. I have one fearful dog, but I like to think that he knows he doesn't have to worry too much these days as he knows I can deal with anything that might worry him.

[NaturallyWild]

Absolutely Vickie, leadership plays a very important role in working through situations (as long as it is a natural leadership, not a forced/physical leadership). Its all about free choice in my opinion but controlling the situation so that choice goes in the direction you want.

Need to be careful that not reacting to that fear is due to leadership alone - eg what happens in the absence of the leader further down the track. The dog needs to be given tools or new associations to be able to deal with situations or not feel they are in a situation at all.

Edited December 1, 2006 by NaturallyWild

[Nekhbet]

Depends what causes the fear. If it is a habitual fear then over reaction from the owner reinfoces the behaviour. The dog can show fearful physical behaviour to get its own way but not neccesarily be truely scared:

eg I saw a pomeranian once that would scream (and i mean like it was being tortured) if the owner brought it into situations it didnt like. Automatic reaction was to pick it up, cuddle and sooth it. THe dog didnt look geniunely scared to me but the owner thought it had all manner of mental problems. When distracted the dog could be brought closer to the stimulus that caused it previous tantrums.

true blinding fear is an autamated response. Weak nerved or manic dogs I dont beleive in pushing the issue. Sometimes medication may be a solution as part of a lifestyle change for the dog as chacgin pack structure can be stressful enough.

I have a working Malinois. We worked through her fears and fear periods with a lot of socialisation. She would shake and whine in new situations but I persisted by just sitting next to her and ignoring her and the whole situation. EG heavy traffic. I would sit with her near the noises and just stare into the distance. I ignored her whining and shaking, but when she showed interest and alertness I would calmly pat her and quietly praise her. She learned to take my physical cues to react to a situation - if I didnt make a big deal about she shouldnt either.

Though in her training we use nervousness, stress etc. It can sometimes be that little bit that pushes her from prey to defence when her brain realises that this is no longer play time. You cant train a working security dog with no stress - real life situation the dog will find that prey doesnt deter the attacker and it baulks or runs off. The trick is to show leadership - that the dog looks to you 'mum I'm not sure! what do I do" and I'll give the command, she takes the bite and realises it releases her fear and makes the bad thing go away.

I beleive that socialisation and desensitisation plays a big part, especially early on. A common mistake for some people is to allow a fearful dog to get away with a lot of things so they dont stress. Absense of leadership then confuses this dog further and there are more problems are new behaviours to then deal with.

As for a dog with stranger issues I dont beleive in getting the stranger to offer food straight off the bat. I would rather let the dog have 1) and escape route (like and open door out of teh room) and the owner present. If the dog wants to approach then thats fine, but the stranger makes no attempt to look, touch or talk to the animal. This is good at assessing the dogs behaviour too. If it eyeballs the stranger, walks round but wont leave the room with the option there you have to wonder is it truely terrified? Is there enough curiosity in there for the owner to then harness to get the dog over that hurdle. Then progress to leaving food near the stranger and see if the dog has enough confidence to take it. Will it take it a meter away? Heaps of verbal praise then no more engaging the dog. Will it take it a foot away or will it pace nervously?

If the dog walks into an open door room and attacks a stranger that has made no attempt to touch, talk or even look in the dogs direction then I dont class that as fear.

Like you say though NaturallyWild all cases should be assessed on an individual basis, but I would always reccomend a full health check first and use of professional guidence during the dogs rehabilitation.

Edited December 1, 2006 by Nekhbet

[m-j]

I personally believe that each situation must be dealt with differently (with an individual program that will often change, particularly as you progress through the fear). I feel that if the dog can take an active role (operant conditioning) then you are likely to get a stronger result, because the dog has "chosen" to make that positive response . There are however the potential draw backs as mentioned which can make it harder to work the problem and have the owners work the problem - eg masking the problem.

Hi Tim

I agree, as basically every dog handler team is an indivual team, with many variables that go into making up that team, as well as there being different reasons for the fear and degrees of fear.

cheers

M-J

[kamuzz]

2: Are the following true or actually misconceptions

a- praise and petting reinforce fear or fearful behaviours

Why would doing something nice cause them to become more afraid? Isn't true fear not an active behaviour (eg choosing to do it), but rather an automatic response. Why is it ok to do classical conditioning in fearful situation eg rewarding, but not ok to praise/pet (which for most dogs is less rewarding than food, toy etc)?

eg shaking/whimper on appearance of trigger - praise/pet to stop reaction - does this cause shake/whimper to occur in different situations or for it to become stronger in the next situation? is the dog thinking if i cry and shake/harder this time when i'm afraid i'm more likely to get a pat.

We recently had a session with K9 force with our GSD, and Steve was trying to explain this very point to us. Hopefully he'll come along and post more eloquently than I can.

I was walking down the drive when Steve's dogs barked from inside the garage. Coda stopped dead in his tracks. Without thinking, I bend over to pat him, made reassuring noises and then got him to continue walking.

Once we were inside Steve went upstairs. When he came near to the stairs you could hear his footfalls where we were waiting. Coda "started" at the noise, and backed away from the stairs. Again, I bent over to reassure him.

Steve explained it has a lot to do with body language. By bending over, you make yourself look smaller. Which then seems to the dog as though YOU are afraid too. Then the dog thinks, well if my human is scared too, this must be really scarey. Better to stand up tall and walk on with a confident "lets go". That way you are not giving the dog any false vibes.

Guess thisis one of the things the other posters meant when they were talking about leadership, and how important it is?

[Rom]

Wow. Thanks for asking these questions, Tim. I've used them as a test for what I know (or what I think I know might be more to the point)! Have typed up 2 pages of answers and am waiting for the pro's to come on before I mark myself! Sometimes school never goes out!

[NaturallyWild]

We recently had a session with K9 force with our GSD, and Steve was trying to explain this very point to us. Hopefully he'll come along and post more eloquently than I can.

I was walking down the drive when Steve's dogs barked from inside the garage. Coda stopped dead in his tracks. Without thinking, I bend over to pat him, made reassuring noises and then got him to continue walking.

Once we were inside Steve went upstairs. When he came near to the stairs you could hear his footfalls where we were waiting. Coda "started" at the noise, and backed away from the stairs. Again, I bent over to reassure him.

Steve explained it has a lot to do with body language. By bending over, you make yourself look smaller. Which then seems to the dog as though YOU are afraid too. Then the dog thinks, well if my human is scared too, this must be really scarey. Better to stand up tall and walk on with a confident "lets go". That way you are not giving the dog any false vibes.

Guess thisis one of the things the other posters meant when they were talking about leadership, and how important it is?

I wasn't there so don't know the exact situation, but just throwing around a few thoughts.

In this instance do you think that the patting or reassuring noises increased the nervousness, or instead just a response to your body language? What if you had started giving him heaps of treats at the same time as bending over, would his attitude have changed, would he think that you where afraid but happy?

When Coda was at the stairs do you think that he reacted more because of your first response or just that he was at a point in his life of being uncertain or not confident of new/unknown things (he reacted in a nervous way in the first instance before your response)? Theoretically, if there were issues of leadership with Coda (ie you weren't considered leader in his eyes) do you think he would be looking to you for leadership or just looking out for himself rather than worrying abour your body language?

Isn't bending over a dog a "dominating" posture, could Coda have felt nervous about the situation but also being dominated by you causing further nerves?

I think there can be a lot of assumptions made about what a dog thinks and associates rather than what is just happening.

Rom - post away. I want to hear everything anyone has to say. Don't feel that you have to wait for pro's. I have so many views that conflict with a lot of people, but you don't learn whether you believe your views are more correct, less correct, or should be slightly adjusted unless someone discusses them with you.

Edited December 3, 2006 by NaturallyWild

[MaxSpots]

2: Are the following true or actually misconceptions

a- praise and petting reinforce fear or fearful behaviours

Why would doing something nice cause them to become more afraid? Isn't true fear not an active behaviour (eg choosing to do it), but rather an automatic response. Why is it ok to do classical conditioning in fearful situation eg rewarding, but not ok to praise/pet (which for most dogs is less rewarding than food, toy etc)?

eg shaking/whimper on appearance of trigger - praise/pet to stop reaction - does this cause shake/whimper to occur in different situations or for it to become stronger in the next situation? is the dog thinking if i cry and shake/harder this time when i'm afraid i'm more likely to get a pat.

I don't know if this helps, but something that stuck in my head concerning praise when a dog is fearful

Is that your praise (or when you comfort them) is rewarding them for that thought... the fear thought of that moment. And thats how it gets reinforced

Where as distraction, or praise for doing something else (like sit, down, heel, give them a job to do), and ignoring the fear episode, gets their mind off it.

Sorry I'm not explaining it well

eg shaking/whimper on appearance of trigger - praise/pet to stop reaction - does this cause shake/whimper to occur in different situations or for it to become stronger in the next situation? is the dog thinking if i cry and shake/harder this time when i'm afraid i'm more likely to get a pat.

They aren't thinking they'll get a pat.....They'll be thinking, " gee I was right to be scared and shake and whimper because my owner agrees with me" , By patting you are saying Yes, you were right to be scared, good dog.......... Just my understanding of it.

Edited December 3, 2006 by MaxSpots

[Rusky]

1: Which type of learning theory is more efficient and gives a stronger result

a- classical condidtioning eg. on appearance of a stranger treats are fed by owner, when stranger leaves treats stop - the dog has a passive role

b- operant conditioning eg. when dog chooses to move towards stranger they are rewarded - possibly use target stick, etc to lure towards stranger

It would depend on the level of fear, if the dog were very fearful then ignoring the dog completely by both the stranger and the owner give the dog nothing to fear as nothing is expected so I would do neither initially. If the dog then approaches the stranger with body language suggesting timidity and actually sniffs or sits next to the stranger then I would reward with verbal praise. Also I am assuming the dog is known well to the owner and that the scene is engineered by inviting a stranger to your home. A wild dog would not have a fuss made of it unless it were an important member.

2: Are the following true or actually misconceptions

a- praise and petting reinforce fear or fearful behaviours

Why would doing something nice cause them to become more afraid? Isn't true fear not an active behaviour (eg choosing to do it), but rather an automatic response. Why is it ok to do classical conditioning in fearful situation eg rewarding, but not ok to praise/pet (which for most dogs is less rewarding than food, toy etc)?

eg shaking/whimper on appearance of trigger - praise/pet to stop reaction - does this cause shake/whimper to occur in different situations or for it to become stronger in the next situation? is the dog thinking if i cry and shake/harder this time when i'm afraid i'm more likely to get a pat.

b- punishment is contraindicated (eg increases fear, or hides the reactions)

Is the association the dog is creating being paired with his reaction or with the trigger. Are we changing the emotional state in the right direction? Can punishment be used eg. if dog reacts to trigger then something good gets taken away ie owner, but dog must stay in presence of trigger.

yes praise and petting do actually reinforce fearful behaviour.

punishment should never figure at all. I have no idea why people punish dogs and have yet to read, hear or see anything which will make me think to the contrary.

I don't believe dogs think as we do, they react because they are afraid and their fear has been reinforced

4: Working dogs in regards to behavioural thresholds - should we be working above or below thresholds, in what situations and what are thresholds eg which is above the threshold - stare, growling, lunging, exploding. If working above threshold (because very likely to happen in the real world with the dog and owner and they need to know how to deal with situation) do we try to get something positive out of it whilst still in the situation, or just move away and regroup/retry.

sorry I don't understand the question are you meaning police, army rescue or similar? I doubt these dogs are fearful, they generally are most affable and confident when not working. I thought if the dog were fearful it would not even be considered for such work.

I am sure a mixture of both classical and operant conditioning is used by all dog trainers and handlers, intentionally or otherwise. My dog drools when I move towards the food bin at feed time but she doesn't approach.

[NaturallyWild]

Depends what causes the fear. If it is a habitual fear then over reaction from the owner reinfoces the behaviour. The dog can show fearful physical behaviour to get its own way but not neccesarily be truely scared:

eg I saw a pomeranian once that would scream (and i mean like it was being tortured) if the owner brought it into situations it didnt like. Automatic reaction was to pick it up, cuddle and sooth it. THe dog didnt look geniunely scared to me but the owner thought it had all manner of mental problems. When distracted the dog could be brought closer to the stimulus that caused it previous tantrums.

true blinding fear is an autamated response. Weak nerved or manic dogs I dont beleive in pushing the issue. Sometimes medication may be a solution as part of a lifestyle change for the dog as chacgin pack structure can be stressful enough.

Like you say though NaturallyWild all cases should be assessed on an individual basis, but I would always reccomend a full health check first and use of professional guidence during the dogs rehabilitation.

Good stuff Nekhbet

By habitual fear, do you mean that the owners response is part of a habitual response to a fearful situation (eg the dog screaming does so as it knows the owner will pick it up and take it away). Is this reinforcing (strenghtening the behaviour), or just stopping the behaviour from being broken ie not providing an oppertunity for the dog to learn the situation is actually not bad. Once again i don't see the petting or praising as being a reinforcer, but its the actions of us taking them away or not dealing with the situation as a leader should that could reinforce the behaviours.

True blinding fear to me, particularly if it is a disproportionate response to the provocation and it persists well after the trigger is removed, is an indication of neurological or some other health problem, and as you said its always best to get a health check.

[NaturallyWild]

yes praise and petting do actually reinforce fearful behaviour.

punishment should never figure at all. I have no idea why people punish dogs and have yet to read, hear or see anything which will make me think to the contrary.

I don't believe dogs think as we do, they react because they are afraid and their fear has been reinforced

What is it that is actually reinforcing in a fearful situation? What are the dogs focused on in a fearful situation and what is the thing that they would like changed? Like i said earlier how can praise and petting be reinforcing if providing treats in a treatment plan is not? eg if a dog was afraid of a stranger and every time they came into the room the dog was given lots of food treats, does this not create a positive association of the stranger to the dog - could not the same thing be done with petting/praise. Is it not the timing of the reward that reinforces the response.

Rusky, what about negative punishment as by the example i gave eg owner leaves the dog as a response to the wrong behaviour?

4: Working dogs in regards to behavioural thresholds - should we be working above or below thresholds, in what situations and what are thresholds eg which is above the threshold - stare, growling, lunging, exploding. If working above threshold (because very likely to happen in the real world with the dog and owner and they need to know how to deal with situation) do we try to get something positive out of it whilst still in the situation, or just move away and regroup/retry.

sorry I don't understand the question are you meaning police, army rescue or similar? I doubt these dogs are fearful, they generally are most affable and confident when not working. I thought if the dog were fearful it would not even be considered for such work.

Part of my question was what are thresholds and do people consider them in treating fearful cases. What i mean is there are levels of reaction that a dog will give in response to a fearful situation. Does anyone think that it is useful in working with the dog whilst it is in its most reactive state (or high up there) - ie as i said it is likely that the owners in the real world will be in this situation and it would be beneficial that they have learned skills to deal with the situation in a positive manner that either helps with the problem or at the least minimises or has nil negative effects

Edited December 3, 2006 by NaturallyWild

[Rom]

OK, here goes. But I'm gonna copy and paste to save time, so apologies if I repeat something that has already been said.

1: Which type of learning theory is more efficient and gives a stronger result

a- classical condidtioning eg. on appearance of a stranger treats are fed by owner, when stranger leaves treats stop - the dog has a passive role

b- operant conditioning eg. when dog chooses to move towards stranger they are rewarded - possibly use target stick, etc to lure towards stranger

{/quote]

I think that classical and operant conditioning are basically both on the same learning line and I wonder if a dog was left to its own devices to learn, whether you’d get one without the other? The good thing about the clicker training movement is that it demonstrates that learning begins with classical conditioning ie the priming or charging of the clicker which is showing a relationship between two stimuli. Consider the dog that learns to chase a car. First it hears a car coming (audio stimulus) and then it sees the car zooming by (visual stimulus). It draws a relationship between the two stimuli and from there it moves into operant conditioning, which is basically drawing a relationship between an action and a consequence and in this instance the action is the chase and the consequence is the satisfaction of a drive. I don’t think its an either/or situation. I think that you need to start with classical conditioning then move onto operant conditioning because at some stage you want the dog to become active in building its confidence so that it can be confident independent of you.

2: Are the following true or actually misconceptions

a- praise and petting reinforce fear or fearful behaviours

Why would doing something nice cause them to become more afraid? Isn't true fear not an active behaviour (eg choosing to do it), but rather an automatic response. Why is it ok to do classical conditioning in fearful situation eg rewarding, but not ok to praise/pet (which for most dogs is less rewarding than food, toy etc)?

eg shaking/whimper on appearance of trigger - praise/pet to stop reaction - does this cause shake/whimper to occur in different situations or for it to become stronger in the next situation? is the dog thinking if i cry and shake/harder this time when i'm afraid i'm more likely to get a pat.

Fear may start as an automatic response, but dogs are innovative learners and they can get to a point where they can generalise. So they can and do learn that if I want attention, I need to be afraid, or I know that mum won’t make me do x if I’m afraid. I once heard about a dog that had learned to vomit on cue. Owner thought the dog was really sick because it kept vomiting but extensive tests could not reveal a cause. A behaviourist observed both the owner and the dog over a period of a couple of hours, noted that the owner was unintentionally cueing the behaviour, dog vomitted and got rewarded with attention and pats. On the human level we know that a yawn is an automatic response to depleted oxygen levels in the blood stream, but if we pretend to yawn, we can generally ellicit genuine yawns in response, not only from ourselves, but those around us (including our dogs!). This is where anthropomorhism is so dangerous to dogs, just as our pretend yawns increase our blood oxygen levels, pretending to be fearful does increase the chemicals related with stressful responses in the dogs blood stream and the dog can develop stress related illnesses. Classical and operant conditioning can be occuring without us realising it!

I believe that its best to totally ignore fearful behaviour...its bl@@dy hard because everything inside me is screaming 'I just want to protect this dog and reassure it', but that is a human emotion...not a dog one. The most I can do is remove the dog from the situation until I can set up conditioning exercises.

b- punishment is contraindicated (eg increases fear, or hides the reactions)

Is the association the dog is creating being paired with his reaction or with the trigger. Are we changing the emotional state in the right direction? Can punishment be used eg. if dog reacts to trigger then something good gets taken away ie owner, but dog must stay in presence of trigger.

As for punishment in this situation? That would depend on whether the behaviour is genetic or learned...thats something that only a pro could assess. If the behaviour is genetic, then I believe that punishment would be counter productive, but a learned behaviour can be unlearned and punishment/correction may be appropriate in this instance, especially if the behaviour is risking the health of the dog. In the example of the dobe above, a pop on the check chain was used to interrupt the dog the instant it started heaving. I once watched a demo where a show poodle was turning to jelly on the table when the judge approached to examine it. To figure out whether the behaviour was learned one the environment was changed a little ie, handler other than owner was used and poodle was taken off the table. In this situation, the dog didn't go to jelly. For some reason it had learned to fear being approached by a stranger on a table. Punishment/correction wasn't to fix this problem, rather a pressure/release method where the pressure was the judge taking one step toward the table from a distance, if the poodle didn't react the release=reward of the judge moving away and the handler rewarding for confident behaviour. Judge slowly increased the number of steps he took towards the table before providing the release. I think this is operant conditioning with a twist! Sometimes a reward can be removing something from the environment!

As for punishment/correction the timing would have to be exquisitly fine if it were deemed necessary, and would have to be used with the intention of shocking the dog out of one drive so that you can reward in another ie shocking out of defence drive and returning the dog to pack drive. For my level of knowledge at this time…this question is probably best left to the professionals! In the instance of the dobe, if the dog were genuinely ill I believe that it would have just learnt ‘Its not OK for me to vomit in the presence of people, therefore when I need to vomit, I’ll hide.

3: Do the following techniques for either classical conditioning (a) or operant conditioning (b), have the down sides of proceeding comments.

a- Positive conditioned emotional response - potentially hardest to get right, due to timing and being persistant (need to always provide reward on appearance of trigger) - can often be harder to control the trigger eg if red bus is trigger it is not that controllable.

As with all training its best to start in a low distraction environment, so the conditioning should start away from the fear trigger where the dog is not showing signs of stress. That way in the initial learning stages it is easier for the dog to learn exactly what it is being rewarded for and also easier for us to be rewarding the correct behaviour.

When you’re dealing with fearful dogs there is another variable to consider and that is critical distance. There is an invisible line that you can cross and when you pass this point the dog can not only lose all ability to learn, but also you can undo the training that you’ve done to this point. So its important that you know what the dogs critical distance is and work to keep the an attitude of confidence in the dog. The critical distance from the fear trigger can be decreased when you know that you can get and keep the dogs focus for short periods of time and then you can gradually increase the time that you spend on the other side of the critical distance line. By holding the dogs focus, then crossing the line briefly at first, you show the dog that it could get closer to the trigger with nothing bad happening.

So in short, yes, if you’re trying to condition a response in the presence of the trigger. But if start conditioning the response away from the trigger in a distraction free environment then work with the critical distance issues, no. Basically you need to raise the dogs threshold to the trigger in a gradual manner. This we can control.

b- Luring - false readings if the lure is so much more enticing than what the fear issue is, causing dog to not notice the fearful thing - association is not created

Association is still created, its just in a very diffused form because you have prevented the dog from entering one drive by keeping it in another. The trigger is still there. You’ve basically broken the problem down and are delivering it in chunks that the dog can deal with and maintain a confident attitude. You’re teaching the dog that it can be in the presence of the trigger without bad things happening. This builds confidence.

Cajoling/happy talk - may mask fearful symptoms if dog is extremely social or attached (similar to above) eg if afraid of stairs, will bolt up to get to someone, but won't go near of own choice.

Counter-commanding - have a requirement for this command to deal with situation - does not necessarily learn to deal directly with situation (needs to be done as trigger appears, or as a response to stopping reaction, not as a response to starting reaction)

The picture in my mind I have of the above is not one of a dog that is masking symptoms, but rather one that is switching between two different drives depending on both the trigger and the value of the percieved reward. After all, its not yet going up the stairs confidently. Confidence comes with time and repetition. Some beginner agility dogs have the same issues with the contacts. The problem is just broken down for them and given to them in steps that they can confidently handle.

4: Working dogs in regards to behavioural thresholds - should we be working above or below thresholds, in what situations and what are thresholds eg which is above the threshold - stare, growling, lunging, exploding. If working above threshold (because very likely to happen in the real world with the dog and owner and they need to know how to deal with situation) do we try to get something positive out of it whilst still in the situation, or just move away and regroup/retry.

I think that the training should start below the behavioural thresholds with consideration given to low distraction and critical distance and a gradual elevation on the pressure given by these two criteria with the aim of raising the dogs threshold to the trigger. Also I think a closer attention to the finer behaviours like those identified in the work with calming signals rather than waiting until the gross behaviours like stare, growl, lunge and explode happen. By the time the lunge and explode have happened, the learning opportunity has been missed.

[Rusky]

Rusky, what about negative punishment as by the example i gave eg owner leaves the dog as a response to the wrong behaviour

the owner leaving the dog is not my understanding of punishment also wrong behaviour for whom?

If a dog is fearful it is a learned response to a situation. When a dog is afraid and cornered it may bite, if it can run it will run. Fight or flight for a fearful situation, if there is nothing to fear because the dog is ignored there will not be a response in a fearful dog. A dog needs to feel safe and secure combined with trust of their leader.

Witholding treats for example when training whilst shaping is not my understanding of punishment. Giving a dog a safe place to go, to cave like a crate and then leaving is not punishment. I know that a year or so ago there was an extensive thread about positive and negative punishment.

A few of my personal examples of punishment are

slapping

kicking

alpha rolls

scruff shakes

throwing outside

neck jerks etc.

Rewarding desirable behaviour and shaping desirable behaviour provides exact information. What part of the behaviour are you punishing/ rewarding.

Stranger comes in room, dog fearful, gets treat, still fearful but likes treats and can gobble the treat with one eye on the stranger.

Dog works out that strangers mean treats but still afraid of stranger as you are actually reinforcing the fear.

Stranger comes in room, you both ignore dog which has hidden under the table barking, you don't look at or speak to the dog. Dog thinks

'hmmm waste of time this, no one is bothering me, I will lie down now' Lies down quietly for a few seconds

owner says

'good boy'

Like i said earlier how can praise and petting be reinforcing if providing treats in a treatment plan is not? eg if a dog was afraid of a stranger and every time they came into the room the dog was given lots of food treats, does this not create a positive association of the stranger to the dog - could not the same thing be done with petting/praise. Is it not the timing of the reward that reinforces the response.

It may be that the dog feels that his fearful behaviour is rewarded. Most definitely timing is critical but when the dog is relaxed and comfortable not when he is hyped and distressed or the treat is associated with the hype and distress, I would reward for calmness and acceptance, usually not with treats but as I said before verbal praise. treats work well for shaping, many dogs depending on personality will not actually take food when they are hyped. Depends on the personality.

[MolassesLass]

What is it that is actually reinforcing in a fearful situation? What are the dogs focused on in a fearful situation and what is the thing that they would like changed? Like i said earlier how can praise and petting be reinforcing if providing treats in a treatment plan is not? eg if a dog was afraid of a stranger and every time they came into the room the dog was given lots of food treats, does this not create a positive association of the stranger to the dog - could not the same thing be done with petting/praise. Is it not the timing of the reward that reinforces the response.

I think the point of a treatment plan such as you've mentioned is to treat the dog when it's NOT showing fear and thus you are rewarding the behaviour you want. You wouldn't take a human fearful dog, plant a stranger right next to them and then try to get it to eat treats. You'd have the stranger across the yard, at a point where they dog is not fearful yet and reward nice calm behaviour and then work closer, always rewarding non-fearful behaviour and stepping back once fear is shown.

Praise tells a dog that they did what you wanted and to get more praise do more action. Thus, what humans think of as consoling/calming/quientening a fearful dog is actually telling the dog, they did exactly what you wanted them to.

Well, that's what I think anyway.

[Rusky]

Well, that's what I think anyway

so do I

[kamuzz]

When Coda was at the stairs do you think that he reacted more because of your first response or just that he was at a point in his life of being uncertain or not confident of new/unknown things (he reacted in a nervous way in the first instance before your response)?

think this may have been what Steve was trying to explain.

Coda's first reaction is to be nervous, at this time of his life.

Therefore I should be careful not to do anything that confirms he should be nervous.

Things that may confirm he should be nervous could include:

- fussing over him

- showing anxiety myself

[Nekhbet]

Exactly kamuzz. like I said when my belgian went through her fear periods it was up to me to guide her and shape how she would come out the other side. If I didnt make a fuss, then she shouldnt.

Also handy for a working animal, if you show nerves or fear then she has to pick up on that and react accordingly. If i dont make a fuss of a person or situation then I dont want her taking it upon herself to launch at someone for no good reason. Getting a young dog to take cues from your behaviour is very important in the dog/handler relationship and especially for working animals.