My stance on working dogs in general is that they impress me. A dog's nose is more sensitive than our best scientific equipment. That will change. One day, probably in the not too distant future, we will have a way of confirming a dog alert and then all these arguments will become obsolete.
I tend towards the belief that a dog will alert to what it was trained to alert to. You've given us an example of a dog that alerted 45 times in 24 locations in which it was believed a body had been, only later to discover that the 'murderer' was lying.
As far as I can see the only real explanations for that is that:
A, the dog was somehow being unconsciously cued by the handler
B, the murderer was actually telling the truth the first time around and lying on his recantation.
C, By coincidence those locations were the scene of a crime and cadaver was left there.
C is too much of a coincidence for me. I am (to my detriment) unable to believe six impossible things before breakfast, so C is off my list.
B is possible, but from the article it would appear to be highly unlikely. More likely than C, but still a very outside chance, so I'm left with A.
Does believing that handler bias is the most likely cause in this case mean that all working dogs should immediately be retired? That we should dismiss the experiments that have proved their accuracy? that the bodies, bombs, drugs etc that have been discovered by these dogs should be put back and dismissed as coincidence?
Again, that isn't a logical step to take. Accepting that mistakes can happen doesn't mean that mistakes always happen.
All the alerts in the Zampo case were uncorroborated apparently. That is not the case for Eddie and Keela. There were human materials found in locations that were alerted to. That could be a coincidence, but so could Zampo happening across cadaver from other crimes, so without further information, it's probably best to assume coincidence is unlikely.
I would love more information about Zampo. From the bit of research I did yesterday, I found that his trainer is still working and that Zampo successfully alerted to 2000 year old remains (links supplied in last nights post) I would really like to hear what his trainer has to say about it all.
Edited to add, all IMNEO
I agree with most of that, Cariad. However, I think that there is another possibility. A subset of A, possibly.
In operation, as opposed to research blind tests, the handler would presumably be aware of the search area in question and may - or may not - be already aware of why the canine team is being asked to search in a particular area.
A dog may pick up inadvertent cues from the handler and concentrate to the point where some of the substances it's trained to react to are present, even if not in totality. I or someone else pointed out that such dogs can react in areas with peat bogs (I think, from memory), as some of the compounds would naturally be present. The dogs aren't necessarily wrong, but it's the human interpretation of alerts that can be mistaken (as well as possible mistakes from the dogs themselves).
As with any aspect of a case, innocent explanations should be ruled out.
ETA Both by forensic evidence as, after all, that is what is being sought, and by ruling out irrelevant/innocent other reasons for alerts or any forensic evidence that may be found.
