We had set ourselves a fairly rigorous target of at two trails a week to build experience and confidence through the autumn and winter months, and although Venn continued to perform well when the trail was known to me, we continued to experience occasional "flights of fancy" where we would go a long distance in the wrong direction when the trail was blind to me . Sadly my contribution to the partnership still seemed to be a bit of a negative one. In particular I was still missing the subtle changes in behaviour when Venn ran out of scent.
Some background and terminology: Scent doesn’t just sit where a person walked, but is also carried by wind and thermal air currents to be deposited elsewhere. If you imagine a 4 way junction: the person would have arrived along one option and departed along another, but depending on the prevailing weather conditions, scent could easily have been spread down one or both of the other two options too. In the case of the latter, the concentration of scent will decrease the further away from the junction you get, whereas with the correct option it will remain comparatively constant. So, all options at the junction might initially be positive for scent, and the dog might reasonably try to follow the pattern down any one, but on the “negatives” this should eventually peter-out until the dog can no longer follow it. This is what we mean by “running-out.” The dog is not technically “wrong”(though I do use the term to indicate an option the target person did not follow) in these instances as she is still reacting to the target scent. Confusing! When the dog does not immediately identify the strongest scent and “trail-through” a junction, then we go through a process of systematically testing the options, imaginatively termed “working the junction.” The process starts with the up-wind option; the thinking being that scent will not have been able to be moved in this direction by the wind, so a positive response by the dog is likely to mean the person walked this way. It’s a good theory, and it is a starting point, but by no means an absolute truth; on a trail laid hours ago, the weather can have changed completely between setting and following, and in some environments, such as urban streets and forestry roads, the wind can swing 180 degrees while you are still working the junction.
Anyway, back to the tale. It was clear that we would not be ready for the spring assessments. This was a great disappointment, as I was keen to get Venn into service, but I really didn’t want to be going into assessment with doubt in my own mind. I figured that there would inevitably be a drop in my performance due to assessment nerves, so I needed to be sufficiently above the standard that it could stand the hit. I certainly didn't want to be going in with my fingers crossed.
In a way, this realisation that we would not be doing assessment took a pressure away, and we continued to build experience with both known and blind trails, and in particular looking at Venn's behaviour when she went wrong and ran out of scent. As previously noted, it is of course a self-limiting process: the better our performance gets, the fewer opportunities there are to see what happens when it goes wrong.
Whilst trying to create opportunities for looking at Venn's negative response, we went back to swab searches, but started in a "clean" area where there was no target (looking at negative behaviour), then moving on to the target area to finish off with a positive. What quickly became apparent is that Venn could pick up the small amount of scent coming off a small cotton-wool swab in a perforated tin from a surprising distance. If my "clean" area was adjacent to the target area, then she would indicate, then follow the scent for more than 50 m to find and indicate on the target. In the context of blown scent on a trail junction, it is not hard to see how she might overshoot a turn by some considerable distance, and yet still be detecting the target scent. This apparent super-sensitivity certainly gave me a better understanding of why we were getting false-positives when working a junction to determine direction of travel. A potentially more problematic side effect of this ability to follow very faint scent sources, was her evident ability to short-cut loops in trails by joining sections several hundred meters apart.
For most of our training, in our peer-group of one, we have been way behind the other qualified and unqualified trailing dogs, and so have always had comparatively slick performances as worked examples from the other trailing teams. However, the problem with slick performances is that you don't always notice the subtleties of what is actually going on. In the spring we hosted a training weekend with SARDA Ireland, and a couple of dog-teams at very much our level came over to work. This was hugely useful for me.; for once I was seeing people get it wrong (sometimes in the same way we did, and sometimes in ways we never did). The discussions about how to address these issues, and general handler-craft were just what I needed. I was also impressed by one of the trainee-handler's ability to remain unrattled by confusing responses from the dog, and her ability to stick with it and work through the evidence methodically. Her ethos centred on the familiar principle that getting stressed rarely made her situation better and often made it worse. "What would Edelle do?" has become part of my internal monologue when things go astray out on the trail.**
So, in summary, we are inching forwards with a mixture of known and blind trails, and an awful lot of pouring over video footage. It still feels possible that we might get there this year.
**For her part, Edelle denies all charges, and maintains she was just having a good day. I'm not so sure....
Komentarze