Let's start with something we all know and which has become painfully clear in recent months: police lie routinely and actually teach lying at the police academy. Lying is considered a core competency of the job.
In short, police are lying liars.
And guess what? So too are their dogs.
Most police dogs and their handlers are so poorly trained, and bad bites so common and payouts so large, that increasing numbers of police departments are reconsidering canine units altogether.
Where the world of lying, dogs, and police really comes together is in the arena of drug sniffing dogs which are often trained to throw false positives in order to gin up probable cause to search a person, locker, or vehicle.
In a ruling earlier this week, a federal appeals court upheld the conviction of a man caught with cocaine in his car, but said that Lex, the "drug dog" deployed by the police was so incompetent he should not be used to establish probable cause for searches. As the Associated Press story notes:
Tuesday's ruling pointed to records showing Lex nearly always signals drugs are present — 93 percent of the time. And it cited other figures that indicated he is frequently wrong — more than 40 percent of the time.
"Lex's overall accuracy rate ... is not much better than a coin flip," the ruling says.
The court upheld Bentley's conviction in part because other indications at the traffic stop -- including contradictory statements made by the defendant -- may have separately justified a search.
But the dog? That dog was simply being used as a bogus pretextual search machine. As the court said: "not much better than a coin flip."
Of course, that makes him the perfect police dog. No doubt the Canine Training Institute of Emden, Ill, which takes credit for training Lex, will now be flooded with requests from police departments looking for dogs with the same set of "skills" and training.
The only way to fight back is the obvious one: If you are ever on a jury, assume the police officer may be a liar and that his or her dog may be a lying liar as well.
The chance that either one is telling the truth appears to be no better than a coin toss.
2 comments:
Does the dog know that it's lying?
Possibly -- what the dog knows is that it's well rewarded for making finds, so finds it makes. What people *should* do is reward as heavily and well for not finding anything when there's nothing to be found, but being human, the tendency is to reward the positive, not the correct. Very few programs actually do this. http://leerburg.com/scentid.htm (one notable exception -- there's also some interesting critical literature on the programme out there) And then you add in the human being and his biases and ego...
There was a nice run down of the mess it generally adds up to here: http://www.science20.com/gerhard_adam/how_reliable_are_sniffing_dogs-95956
Bottom line: dogs may have good noses, but they're desperately motivated to please us.
Post a Comment