Backward Chaining

Discussion in 'Labrador Training' started by Jen, May 14, 2016.

  1. MaccieD

    MaccieD Guest

    Worked for me in France on walks with the French dogs. They all learned those English words and worked out that it meant a treat :D
     
    Newbie Lab Owner likes this.
  2. MaccieD

    MaccieD Guest

    Whoop, whoop, book has arrived :happyfeet:, it's got 100+ pages :facepalm:, although a lot of pictures (I like books with pictures :chuckle:), so it might take me a couple of days to read, so watch this space :nod:
     
    Newbie Lab Owner and Jen like this.
  3. Jen

    Jen Registered Users

    Joined:
    Aug 30, 2013
    Messages:
    2,897
    Get reading Rosemary, get reading. ;)
     
  4. MaccieD

    MaccieD Guest

    Will do asap, need to walk Juno first. She's been a bit neglected today, bless her, as I've been struggling to sort out a planning application for the council whilst fighting technical problems due to BT ;)
     
  5. MaccieD

    MaccieD Guest

    @Jen just to let you know I finished the book late last night/early this morning. Brain is still filing info and got lots to do today but will be back with thoughts asap :)
     
    Newbie Lab Owner and Jen like this.
  6. Jen

    Jen Registered Users

    Joined:
    Aug 30, 2013
    Messages:
    2,897
    Thank you Rosemary I do appreciate it. Don't worry if your busy we can wait we aren't going anywhere. ;)
     
  7. MaccieD

    MaccieD Guest

    Ok, jobs done for now so time to get the brain cells fired up :)

    In the Keep in Mind Turid says
    "It is futile to reward a dog for being quiet right after having barked. There are two reasons for it.

    - Dogs learn well from backward chaining and will quickly learn to bark, then stop barking, then get a reward. Dogs learn lots of things exactly this way.


    What she is saying here is that dogs bark for a reason. The human then often speaks to the dog, to tell it stop or whatever. For the dog the speaking forms an interaction and therefore is rewarded. Dogs can then learn to bark and turn to look for the owner waiting for the interaction which of course is not what we intended when first speaking to them. Turid advises that rather than speaking to the dog you use a hand signal - hand raised with palm to the dog (which many people use any way as a wait type signal) and then move between the dog and the object that it is barking at next actions are dependent upon the source of the barking.

    - Dogs learn best when getting a reward for something physically they have done. Rewarding for absolutely nothing (i.e. not barking) will not lead to learning just frustration "

    Here, obviously, dogs learn well, as we know, when being rewarded for performing a cue; rewarding when they haven't done anything is confusing. She clarifies this by having conducted experiments with some of her student trainers in a classroom environment to highlight how a dog might feel. While they are busy working, or perhaps sitting quietly, just listening she goes up to a student, having ensured that the student was not engaged in any physical activity, and praises them. The students is bewildered and does not understand what the praise is for. They all say that they feel totally frustrated, and find the experience unpleasant.

    I had some doubts on this 'confusion' but turned it around by questioning as to why would I disturb my dog, who is laying quietly, to reward them. Perhaps a bit like the old saying 'Let sleeping dogs lie'. Other situations we normally give our dog a cue and reward them when they perform satisfactorily, they don't just get a reward for no reason so I can see where we could cause confusion


    Turid also does an 'angry' with her students having forewarned them that she was going to do it during the session. They never know what she is going to reprimand so she might just single out someone fiddling with a pencil, who scratches their nose or raises a hand to ask a question. They all report though that it was a very unpleasant experience and left them feeling uncertain for quite a while

    So what she says is that we should think more about what we actually say and do to our dog to understand the reactions we get. If people feel uncomfortable by things we say and do how do our dogs feel?

    Well I hope that makes some sense, and is my interpretation based upon the book.


    I'm lucky and have a dog that only rarely barks, but found the book very interesting and I like how she links some of the techniques from her other books with this one. An overall useful little book, but then I'm already a fan. @Jen if you would like to borrow the book to read, or anyone who is interested, please pm me as currently the book is just going on canine behaviour bookshelf :)

     
    charlie and Newbie Lab Owner like this.
  8. Beanwood

    Beanwood Registered Users

    Joined:
    Jan 28, 2014
    Messages:
    7,303
    Now this is interesting, mainly because we have had to work so hard on calm behaviours with Benson. It was notable when we realised "hanging about" simply wasn't in Benson's vocabulary. Evidenced by the kangeroo in gundog group class, and impossible settle whilst all of us were in the barn. Very embarrassing. So we went about teaching a settle. We put him in the context of where we wanted him to settle. So we sorted "his" mat, and worked a lot outside first, then moved to cafes and pubs. Eventually we took away his mat once he understood it was cool to be calm. We rewarded for nothing. Deliberately waiting for him to disengage, become quiet, and when he had been doing this for 10 seconds, quietly dropped a treat in between his paws. I avoided eye contact, or any indication I wanted him to engage. Sometimes quietly said "good boy" or maybe a relaxed smile. Calm behaviour begets calm behaviour, if I am waving, and jumping excitedly, this induces similar behaviour in Benson. Indeed most of the time I am asking for energy and enthusiasm, especially in agility for example. I think dogs are very much led by our body language. We know how easy tension can glide down a lead! By giving rewards we were encouraging a nice calm settle. It is (on a roll here... :)) so important to get the timing and context right though. The lovely thing is Bramble at only 5 months old has a good settle, she had been copying her big brother so I would treat her as well.

    I have to admit to confusion with the example of students being perplexed by praise whilst they were working/sitting quietly. If you look at shy children drawing at a table...I would imagine a quiet nod, or smile from an adult would be very reassuring. Think I will to read Turid Rugras's book, sounds interesting.:)
     
    Oberon and Newbie Lab Owner like this.
  9. bbrown

    bbrown Moderator Forum Supporter

    Joined:
    Jun 13, 2011
    Messages:
    9,435
    I might be wrong but a settle is a physical thing. It's on a mat, it's a lie down and it's usually a lie down with hips over rather than alert sphinx lie down. You reward in place to maintain the behaviour.

    Although it's very passive it's still a behaviour with several physical elements.

    That's different to rewarding the absence of something - to my brain at least :)
     
    Newbie Lab Owner and Beanwood like this.
  10. Beanwood

    Beanwood Registered Users

    Joined:
    Jan 28, 2014
    Messages:
    7,303
    Yes that is how I interpret the settle, I only reward for being in that position :)
     
    Newbie Lab Owner likes this.
  11. bbrown

    bbrown Moderator Forum Supporter

    Joined:
    Jun 13, 2011
    Messages:
    9,435
    I'm trying to teach Obi to settle bwahahaha!!!!
     
    drjs@5, Oberon and Newbie Lab Owner like this.
  12. Beanwood

    Beanwood Registered Users

    Joined:
    Jan 28, 2014
    Messages:
    7,303
    haha! Almost choked on my wine! :tail:
     
    Newbie Lab Owner likes this.
  13. bbrown

    bbrown Moderator Forum Supporter

    Joined:
    Jun 13, 2011
    Messages:
    9,435
    One day!!!! In a year or four LOL!
     
    Newbie Lab Owner likes this.
  14. Newbie Lab Owner

    Newbie Lab Owner Registered Users

    Joined:
    Aug 16, 2015
    Messages:
    1,447
    Location:
    UK
    This is really interesting as I tried treating Dexter when he was laying down nice and quiet of his own accord and thought it was really working although he did look a bit confused at the time. That was until after a few days, he would be laying quietly and if I ignored him, he would then get up, sit and stare at me, lay down and then pester me, nudge me, bark at me, anything to get my attention.

    I did start to think that I'd asked for no behaviour but had rewarded him. I can see now why he must have gotten frustrated with me, poor boy, no wonder he couldn't switch off.
     
    MaccieD likes this.
  15. Oberon

    Oberon Supporting Member Forum Supporter

    Joined:
    Mar 6, 2013
    Messages:
    14,194
    Location:
    Canberra, Australia
    When you've got that sorted you can train my Obi to do a settle too.
     
  16. bbrown

    bbrown Moderator Forum Supporter

    Joined:
    Jun 13, 2011
    Messages:
    9,435
    I'll be old and grey by then LOL!!!
     
    Oberon likes this.
  17. charlie

    charlie Registered Users

    Joined:
    Sep 29, 2012
    Messages:
    12,217
    Location:
    Hampshire, UK
    Thanks Rosemary and oh dear I have done it wrong, again :rolleyes: when Charlie barks at the doorbell or when he's in the yard and barks at people walking past I say "thank you" he comes in, mostly, looks at me and gets a treat which he really doesn't understand?. So wrong!! :oops: x
     
    Newbie Lab Owner and MaccieD like this.
  18. MaccieD

    MaccieD Guest

    No. it's not wrong and what many people have advised in the past. Turid just advocates a different method of training with the barking. At the end of the day Charlie barks at someone, you speak to him, he stops and comes in and gets a reward - job done. What would be more worrying is if he were to just start barking and then look for you ready for his reward. Might be interesting though if you started using the hand signal rather than speaking though .... but I'm not advocating that you start experimenting with him. Stick with what is working for you and Charlie :D
     
    Newbie Lab Owner likes this.
  19. MaccieD

    MaccieD Guest

    During the day in the kitchen Juno used to just curl up where she fancied for a sleep - in the crate, in the cats bed, by my feet and I used to just leave her. In the evenings in the lounge I used to pretty much ignore where she laid sometimes on her bed, sometimes on ceramic floor and sometimes by my feet. During the course of the evening she used to end up in her bed in the lounge and I used to say good girl as she curled up and give her a little pet. Nowadays when we are out and about Juno sits under/next to tables with no fuss and normally decides to lie down once she's a had a little look around and has found the best spot considering human legs and those of tables and chairs.
     
    Cath and Newbie Lab Owner like this.
  20. heidrun

    heidrun Supporting Member Forum Supporter

    Joined:
    Feb 10, 2012
    Messages:
    2,626
    I've got more chance of teaching his sister a double somersault backwards than a settle.
     
    Oberon, Beanwood, drjs@5 and 5 others like this.

Share This Page