In the last instalment we had just made a step-change in reading Venn's body language and behaviour, the lines of communication were wide open, and all manner of trails were falling before us.
Alas, it was never going to continue to be that simple.
Having made this sudden progress in communication, we jumped into a series of complex and blind trails, and on the whole things went very well, but every now and then we went wrong, and when we did, we went wrong in a very big way. If everything went well, we were fine, but I was failing to read the situation on the occasions that we went wrong and Venn seemed quite happy to go for the walk up the wrong path and keep telling me that everything was fine.
Imagine meeting someone who speaks a completely unfamiliar language. After a bit of miming and chest-patting you might get to knowing each other's name and the lines of communication are open. Clearly, if the next phase in this dialogue is your new friend trying to explain the theory of relativity, there are going to be some misunderstandings. Much as this starting point is exciting and encouraging, there needs to be a bigger vocabulary available before anything complex can happen between you.
That's how it went with Venn. As soon as I started responding to what she was telling me, the message shifted and got more complex. Where silent working was a good sign, this became more and more replaced by frustration-barking, making the barking behaviour less useful as a negative indication.
Clearly we needed to go back to language school for lesson two
In order to get a better picture of negative-scent behaviour (what she does when she can't find the target smell), I started working every junction: getting Venn to test every option (some of which would be negative for scent) before asking her to trail to the next decision point. Doing this without knowing the answer myself forced me to be objective about her body language, but risked reinforcing incorrect behaviour if we went wrong. Of course, if I did not know which was which myself, then looking for subtleties in behaviour in Venn when on or off-scent was also a lot more difficult. It was tricky to know what to do for the best. In the end I opted to do the trails blind, but with immediate feedback as to whether we were correct in our decision making; hopefully getting the experience of running blind for me, but allowing us to catch any errors before they ran too far.
The prospect of being ready for the March assessment weekend started to slide further out of reach as it became clear that I had no real confidence in being able to spot an error. This was a big disappointment as we seemed to have turned a corner with our trailing work and things were mostly going so well. At the January SARDA training weekend, we got to try some urban trails and to have some fresh eyes on Venn's behaviour. Our first trail did not go so well, though we did eventually get around it with a couple of nudges. Urban is hard! Hard, dry surfaces do not hold scent well, and what little there is can be spread around by traffic and unusual air-flow around buildings. Any dog is likely to find it harder, and although Venn was picking up scent, she seemed to show no confidence in any particular direction.
My testing of every junction was also noted. The suggestion was made that a potential side effect of this approach was that in stopping Venn's flow through junctions, I was wresting control back from her, undermining her confidence and increasing frustration. In short, It seemed likely that I was the one spoiling the game.
In our second trail, I was asked to just let her get on with it, and only work junctions if she stalled. She did really nicely, particularly on grass tracks, but also much better in the urban environment. Seemingly, I am the weakest link.
In terms of body language, the Trainers also spotted a couple of subtle shifts in behaviour that might help reinforce our understanding of a negative response or a lost trail. I just need to keep an eye out and see how consistent these are.
So we have a conflicting set of requirements. In order to better identify the situation, I need to see more examples where Venn goes off trail and loses scent, or when I ask her to find scent where none is present. However, I also need to let her crack-on and flow through the trail when she is able to do so. As she gets better at the latter, there are fewer instances of the former.
The answer, in the short term at least, has been to revert to known but complex trails with as many decision points as possible so that sooner or later we will overshoot a turn or go the wrong way. They don't reproduce the stress, for me, of running a trail blind, but that same stress should be less once I can read Venn more effectively.
Comments