Jump to content

corvus

Community Members
  • Content count

    7,328
  • Joined

  • Last visited

  • Days Won

    1

About corvus

  • Rank
    Resisting hysteria
  • Birthday 07/03/1983

Contact Methods

  • Website URL
    http://www.creatureteacher.com.au

Profile Information

  • Gender
    Female

Extra Info

  • Location
    NSW

Recent Profile Visitors

5,113 profile views
  1. Japanese Spitz Breeds

    As a total sucker for grown-up spitz breeds of which several originated in Japan, I appreciate the appeal. However, I would not get one of these breeds for a suburban home, and doubly not if my lifestyle involved a lot of being out and about in public with the dog. Reason why is that many of these dogs are reserved. They are not unfriendly exactly, but they don't want to be friends with everyone, and out and about in suburbia, that is really hard to manage. I know, because I have two dogs that are not really ideal for suburbia. They take work. Will they hike with me? Absolutely. It's their favourite thing. They like the beach. But dog parks... not so much. The neighbourhood's friendly labradors and staffies? Nope. Kids running around screaming in the park? Er... They will need some distance to handle that. My Finnish lapphund, though, is an ideal suburban dog. As long as you don't mind the coat. He will certainly come hiking with us, for hours. Not much for running or swimming, but he loves everyone and he's really hard to unsettle. Friends with any stranger. He's fine off leash. It takes a bit of work to keep him responsive with strong recalls, but not as hard as it is with my podengo. I'd also go for a Samoyed, and I'd consider a Japanese spitz. I'd be careful with Finnish spitz, though. That is a hunting breed. I have only met one and she was not an easy dog. Buhund would be worth checking out if there is a litter coming up. I have a vallhund, and I don't think they are an easy suburban dog. Quite doable, but you have to appreciate they are a cattle driving breed first and foremost. Shikoku are gorgeous, but they are serious, grown-up dogs. I wouldn't consider it unless I lived in a quiet area. The more time I spend in suburbia with highly strung or not massively sociable dogs, the more I promise myself my next dog will be hyper-sociable. It comes with its own challenges (not everyone wants a dog on their head), but much easier on the dog day-to-day, and so much easier to manage.
  2. Who are you quoting this time? Not me at me to counter a point you think I am making, I can assure you of that at least. I think only one person can win that argument, by the way. Tortora's safety training was good training, in that he built associations and response strength away from the problem stimulus first, and then introduced signals to allow dogs to escape an aversive consequence, which, funnily enough, was what I was proposing for snake avoidance and you thought was overly complicated. It's pretty heavy-handed, though, and we can do it without the e-collar or any tool that delivers an aversive stimulus. There are lots of people doing similar things to treat fear-related aggression with only two quadrants (R+ and R-) and seeing success (myself included), and they are using stimuli already prompting problem behaviour in negative reinforcement. Didn't quote that part of my paper, did you. Maybe you don't like to read that kind of thing. ;) But, it's not terribly relevant to livestock chasing, which is not usually motivated by avoidance. Huge, huge difference. The results are limited in that it's hard to separate what was ultimately responsible for the improvements in behaviour. He did punish unwanted behaviour, at times, severely if I interpreted the paper correctly. And he was aiming for pro-social behaviours, but I don't think we would really necessarily consider obedience behaviours to be "pro-social" in a dog-dog context. I don't at any rate. We have no clue what other changes in behaviour might have occurred with these dogs. He did not record indicators of conflict or avoidance, for example. We care about that kind of stuff these days. First time I've ever had someone describe the revelation that the scientific literature they are quoting at me to reveal my apparent dislike of some scientific literature was actually written by me in the first place as "a coincidence". But then, it's the first time someone has quoted me at me in apparent ignorance, so I don't know how people deal with these awkward moments. I'm still laughing, myself. But I am glad you did your homework. Now write out 50 times "Steel men are stronger than straw men".
  3. Well, IME, there are differences in priorities and goals. I have had e-collar trainers sneer at how obsessed my dogs are with training with me in exciting environments. Like, are you kidding? They think I'm fun to train with! That's my whole thing. I have had e-collar trainers declare my dogs do not listen to me, yet in my mind, they listen to all the things I want them to listen to, and respond promptly. I do talk to them a lot, though, and there are some things I don't care if they do or not when cued. I have had e-collar trainers hold me to standards of training they don't hold themselves to. And I have had e-collar trainers write me off because I have spent no time at all training something they think is critically important. And conversely, I have seen dogs trained with e-collars recalling and I'm like "mine do it faster and with more joy". And I have seen e-collar-trained dogs that have cautions mine don't, and I have specifically set out to banish those cautions. At the end of the day, my experience is one side is always looking to discredit the other, and my dogs are the only ones I live with, and their behaviour reflects what is important to me. Differing importance does not mean differing trainer skill or method limits. Necessarily. I would say 90% of the time at least, if you wanted to go aversive-free, you could, and if you were skilled enough and motivated enough, you would get whatever standard of performance you were after. It is astonishing to me what people can do with positive reinforcement alone, and I keep being astonished, so I'm obviously still learning what the limits are. My one caveat is the dog matters. For things like high level competition, it just helps to have the right dog in the first place, regardless of your methods of choice. And some dogs are just more challenging than others. I'm super proud of my tiny hound's off leash responsiveness because it's been so much harder in her to get than it is in even my spitz breed. There are some breeds I'm not sure I want to try. I don't feel that invalidates aversive-free training. You only have to look at what others are doing with it to see that the limits of what I'm prepared to try don't mean much.
  4. People often seem to think scientists don't know many dogs.
  5. Ah, our old friend Tortora. IMO, a bit of a visionary for his time, but there is a limited amount you can get from this study, really. If you look my published work up, you would find my thoughts on it. I can't remember the citation off the top of my head, but it's a paper about safety signals in animal training, published in JVB, I think Henshall is the lead author. Tortora is not really relevant to the livestock chasing discussion now going on. I wasn't exactly going for a comprehensive literature review, here, but the study I did link to has a nice discussion of the literature. ETA LOL, I've just realised you linked to the exact study I was talking about! I'm on the author list. I actually wrote that paragraph myself! Too funny.
  6. OMG. Are you serious? I haven't even used the words "evidence" or "proof". I didn't express an attitude. Go away. And write out 50 times "Survey data is data". And read the freaking study while you're at it, because the authors have some interesting things to say about the studies on sheep chasing and kiwi avoidance and what they show about e-collar efficacy.
  7. I am guessing that this is the gist of what you were trying to say? I don't think that I have ever suggested anything otherwise? In response to the letter, I remember Cooper saying at Canine Science Forum in 2012 that when they first started looking at how people were training with e-collars, it was so terrible that they had to change the study design or no one would take the research seriously. He went on to say that those people were clearly already breaking current prevention of cruelty to animals laws, and we didn't necessarily need to ban e-collars to stop that. Just enforce the law that existed. I'm not pro or anti e-collar, but it's really a moot point seeing as I live in NSW. I am not gonna use a tool that is illegal, I am CERTAINLY not going to sell or recommend one to a client, and I do pretty well without it. Do people in NSW use them illegally? Yes. I know where they buy them from. I know who the professional trainers are that recommend and/or sell them. I consider that highly unethical. Some of these people that were told by a professional trainer to buy one did not know they were illegal, or assumed that because they were receiving professional advice to use them that they were wrong about them being illegal. To recommend and sell clients an illegal item and not even tell them it is illegal blows my mind. That is not professional behaviour in any way shape or form. Can we achieve what we want without an e-collar? There are always lots of factors to consider, but I would say often we can. If we cared to tackle it seriously and come up with a good plan. That's what I did for my swallow-chaser. I know someone you could ask for help with that. ;) Why should we try? Because we like giving our dogs stuff they like, obviously. Because dogs that don't ever fear negative consequences are often easier to shape and less prone to conflict. Because e-collars are really expensive (and illegal). Because unwanted positive associations are a fair bit less troublesome than unwanted negative associations. Because when you make a mistake with rewards, you kick yourself, but when you make a mistake with aversive stimuli, you kick the dog.
  8. In fact, the authors of the study say this in their conclusions: "More owners using reward based methods for recall / chasing report a successful outcome of training than those using e-collars." It is what it is, and I did not misrepresent it. They are scientific results, and they are not clean results, but we are already dealing with owner reports in this very thread, not experimental data. So, if people in this thread are claiming that e-collars are the best tool for livestock chasing, by owner report, one assumes, then this graph suggests otherwise using a comparable measure, and at least has a published N. Surveys are science, too, people.
  9. Some people on this thread may find this interesting. It is a graphical representation of owner reports on the effectiveness of different training methods for solving recall and chasing stock problems in the UK. For those claiming that e-collars are the best tool for stock chasing, this suggests otherwise. Source here: https://bmcvetres.biomedcentral.com/articles/10.1186/1746-6148-8-93
  10. Apology accepted. Not in my experience. There are certainly dogs who are very comfortable with going to considerable lengths to take the fight to the threatening stimulus wherever it happens to be. It's a valid strategy, so it does get used. Especially when the individual is highly aroused (interested, threatened, loving life), and especially when less overt strategies have been tried and they failed. *shrug* I've seen said fallouts. Trust me, it's not theory. Thing is, a lot of trainers are not looking for what I'm looking for, and that holds regardless of their methods of choice. I could count the trainers I know that I would trust to detect their training method going awry. I have a PhD in dog behaviour. Frankly, I am smart. I am experienced. I see some pretty kooky dogs. I have read a lot of scientific studies. I see the world differently to most dog trainers. My mistrust in aversive training methods is of course well placed as far as I'm concerned, or I wouldn't have it. So, you trust yourself, or you trust me. It's no skin off my nose. I do get annoyed when I have to fix something another freaking trainer broke, though. That is really annoying. I consider part of my job to be to communicate the risks to clients. I would train using aversive stimuli if I felt it might do the job better, and I have offered to more than once. I just tell clients all their options and my professional advice on the risks involved and let them choose. No one has chosen the aversive methods yet. I've always got alternatives, and people choose them.
  11. Oh, please. Now we can't take credit for our observation skills, our ability to predict and test our predictions, and our skill in applying our findings to new contexts to achieve novel goals? But, Mother Nature made us so good at it. Phew. It's okay, you have done a neat job of justifying yourself there. You are FIXING what Mother Nature, who is the queen of animal training, left broken. Don't take credit for it, though. You're just her humble servant. I'll just sit here while you continue to justify your decisions, okay? I'm pleased I at least provoked you to think about it, though. That's my goal attained.
  12. My point was it may or may not be effective at protecting a dog from snakes, but you know what would be pretty effective? Snake proof fencing. Naturally, I can't take my dog trail running with me in a snake exclusion zone, so I have to rely on some kind of training or I don't take her. It doesn't matter what training I rely on, it won't be as safe as snake exclusion. As soon as you throw two animals together in a dynamic environment, you open up a huge number of possible ways that could play out. Training is a game of probabilities, and a probability of 1.0 becomes increasingly unlikely the more dynamic the situation. So, I live with knowing my dog is not 100% safe from snake bite while I am out running trails with her. Bob Bailey was asked about snake avoidance training when he was here last year. He said he had seen dogs that had been trained in snake avoidance walk right by snakes without reacting to them, because once the dog is busy following other scents etc., then they don't notice snakes. That is not necessarily going to keep them safe. It's also not necessarily a reason not to train any kind of response to snakes that might enhance the dog's safety, but Bob said he would absolutely not rely on snake avoidance training, and I wouldn't, either. I very much would like to do a study on this. I finish up my current contract in July, and am toying with the idea of crowd funding a snake avoidance study.
  13. Dude, you responded to MY post about RECALL by saying there is still a problem of livestock chasing. No there isn't. Not for me. You can't talk about OC and not consider CC. As they say, Pavlov is always on your shoulder. I would argue a strong negative association is a dangerous thing to create. It could result in any number of behaviours, and some of them worse than what you started with. For example, I have a cattle driving breed. He responds to things that he doesn't like by trying to drive them away. Say I come in hard with an aversive stimulus in attempting to train snake avoidance with the cattle driving dog above. This could go two ways. The dog might go "wowsers, best to stay away from that thing" or the dog might go "that thing needs to get away from me and I'm pretty sure I know how to make it". Obviously, the latter would be very bad, and it puts the trainer in a position where they can only increase the aversiveness of the consequence and hope the dog doesn't escalate their proactive stress coping strategy. And then we have to hope that the next time the dog sees a snake, he's going to remember he lost this argument last time and he should not do what his genetics tell him to do. If we had to use a highly aversive stimulus to get him to back off, then we have also probably incited a very aroused response to snakes, which makes it more likely that he will revert to his hardwired driving behaviours. In contrast, if we start with low level aversive stimuli AWAY from snakes and train him to turn off the merely unpleasant stimulus by backing up, then first of all, we don't have to risk him going after the snake in training. We can teach him the appropriate response in a safe environment with signals that we can afford to ditch if something goes wrong in training. When we have a dog that is very confident in how to turn that noxious stimulus off, and we can also train a signal that tells him now he should back off to avoid the noxious stimulus all together. Then we introduce the snake, and transfer the signal for backing up to the snake. Because he already knows this scenario and how it works, the snake becomes the new signal to back up to avoid the noxious stimulus. And because he has learned this whole thing at moderate arousal instead of high arousal, he has a nice clear head around snakes, so he is not liable to panic and revert to the kinds of alarmingly forward things he does when he is in high arousal, and nor is he likely to decide a variety of other stimuli present while he's very aroused and trying to escape are also harbingers of doom. I'm not proposing a paradigm shift, here. If you must train avoidance, go ahead. I'm suggesting a low risk approach to it. I also happen to think it's more humane and more useful, but that is my opinion. I think you have been lucky so far. A lot of snake aversion training overseas is questionable IMO. How do we know it works? There are no data on efficacy, and that goes for e-collar methods and so-called "positive snake avoidance" (a misnomer if I ever heard one). The only data we have is on kiwi avoidance, and in that case, some dogs did not retain the aversion training. We have data on coyotes, who generally overcome the aversion training sooner or later. Cooper et al. reported equal efficacy for livestock chasing and recalls between e-collar training and positive reinforcement training in domestic dogs. IME as a canine scientist, domestic dogs as a group have astonishingly low resilience, so I would expect aversion training to hold better for dogs than coyotes, but the fact remains, there are no data to speak of. You don't seem to recognise the shortcomings in your approach, though. The people that have had e-collar training go wrong don't usually go back to the people that trained it. They go to other trainers to try to fix it. The same can be said for positive reinforcement-based training, of course. So, we both need to acknowledge the limitations in training.
  14. How could you think you know what it's called when I didn't actually tell you specifically what I did? I used several tools. Premack was one of them - i.e. Yes you can chase the swallows, if you can give me a hand target first. I mean, once she's off leash, she has a lot of control over what she spends her time doing. The nice thing about swallows is they are pretty contained to good foraging ground, and easy to find. She can safely chase them to satiation in some places, which she did at times. A wee dog can only sprint around a field for so long before she needs a breather. Guess what? She's very likely to recall when she's too tired to keep chasing swallows. Build from there to recalling at full stretch after them and barking. Build from the other end with Premack. Use your reinforcement history as leverage to get check-ins. There are some things I would not rely on training for at all. Would I let my dog run off leash next to a highway? Nope. Just not a risk we need to take, so I wouldn't take it. A couple of my dogs would almost certainly be responsive and reliable enough to be safe, but my ego as a trainer is not so large that I want to test it. I know people that have had dogs blow off e-collar-trained recalls to run in front of cars. Training is not as safe as management, and it doesn't matter how you train or how good you are at it or what dog you have. E-collar training is always trotted out as somehow the gold standard for creating a dog that can be safe off leash, like no other training method could match it. There is just no evidence for that. There are certainly challenges in training some behaviours that leave you with some tough choices about what you can achieve and how, and I wish that all trainers would acknowledge that.
  15. What problem? I trained my dog to recall off swallows, not to leave livestock alone when I'm not around. You don't build up the CONSEQUENCE. You build up the STRENGTH OF RESPONSE. This is utterly critical to understand. Because relying on stronger consequences means you train hard instead of smart. I mean, sure, bring out your big rewards for recalls always, because it's a tough thing to train at the best of times, but if you hammer the recall with the best rewards away from the most challenging scenarios, then you can work on strengthening and strengthening it until you can start working it into those challenging scenarios with success, and that's where good environmental control becomes really useful, whether you are working with positive reinforcement or negative reinforcement. It's not an anti-compulsion argument. It's a pro-Matching Law argument. The better your control of the environment is, the better your chances of getting the behaviour you want and not the behaviour you don't want, whether you are reinforcing the behaviour you want or attempting to suppress the behaviour you don't want. If you come in hard with an aversive stimulus, you are not training avoidance. You are punishing a behaviour and therefore trying to suppress it. If I'm training avoidance, I don't want a dog that is desperately trying to escape a stimulus because they know it's going to be horrible. I want a dog that is thinking about how they can avoid a stimulus. The risk of unwanted associations is smaller, the risk of aggressive retaliation is much much lower, and you get a dog with a much more sensible level of arousal.
×