Smells like improbable cause
By Thoreau
You know how a cop can search any car/person/house/apartment/boat/plane/train/automobile/blimp/etc. he wants if he claims that a dog did something to suggest that it might smell drugs? Well, a couple of bloodhounds in Texas reacted to the smell of a deputy sheriff during a murder investigation, and the guy spent 5 months under suspicion before DNA evidence led to the arrest of another guy who eventually confessed.
I am not an expert on the forensic science of police dogs, but everything I hear is that the “cues” these dogs give are at best often open to interpretation, and at worst may be manipulated by cops who know that the dogs just want to please their handler. It may very well be that the Ideal Platonic Police Dog can sniff out evidence in some objective manner, but the Real Actual Police Dog seems somewhat less than reliable. I might also note that even if the dogs acted reliably in this case, if they got the victim’s scent and went to a neighbor’s house (in this case the deputy’s house) that doesn’t necessarily mean that the neighbor killed anybody. It could just mean that in the last day or two the victim visited the neighbor because they’re, you know, neighbors.
Now that the dog has crossed the Thin Blue Line, we can only hope that cops will give up on this junk science. I doubt it, though. This sort of tool is way too useful in our continuing War on Common Sense, Black Men, Supply and Demand, and Sick People (AKA War on Drugs).

Comment by David —
October 5, 2009 @ 11:51 am
Never underestimate the ability of law enforcement to compartmentalize about the value of these kinds of things when it suits their purpose.
Comment by Don K —
October 5, 2009 @ 12:23 pm
Perfect. If the suspect is one of theirs they can keep investigating until they find the right guy, and if not they can
frameconvict the first poor schnook the dog will sit down for.Comment by Captain Oblivious —
October 5, 2009 @ 1:48 pm
I think maybe you’re being a little harsh – yes, it’s an imperfect technique, but so are ALL techniques!
Should they be more careful, maybe require some independent evidence besides just a twitchy dog? Yeah, probably so…
Should they ignore a mechanism by which some crimes can be solved, just because that mechanism can’t solve all crimes perfectly? I’ll let you answer that one…
Comment by Jason —
October 5, 2009 @ 2:25 pm
Knowing how “what happens in Texas, stays in Texas”, The lawsuit is going to have to make it to the Federal system outside the State reach, before any justice will be served. Texas is a big state though. The system can swallow you up
Comment by dhex —
October 5, 2009 @ 2:28 pm
sue the dog!
[ow stop hitting me mr. henley]
Comment by Keith —
October 5, 2009 @ 2:53 pm
Never understood the logic behind police dogs. All you’ve done is teach them to play fetch with narcotics. And if there are no narcotics, well, hell what dog would pass up an opportunity to play fetch?
Comment by Jack Sonberg —
October 5, 2009 @ 4:21 pm
Isn’t it obvious that as the victim’s neighbor and friend, the dogs picked up his scent on her and on the path between their homes? They didn’t do anything wrong — it’s the assumption that this was the ONLY line of scent that is the error.
Comment by Daryn —
October 6, 2009 @ 1:02 am
“It’s left me with a pretty bad taste for law enforcement,” Buchanek said. “It’s pretty much ruined my life altogether.”
Oh boo hoo hoo! This cop has done the very same thing to so many people before, ruining families and people’s lives, but, now that it effects him, it’s unfair? Tough shit for him, I’m glad his life was ruined. How’s it feel, cop? Welcome to the real world!
Comment by Doug T —
October 6, 2009 @ 8:32 am
If used correctly, tracking dogs are great assets. They can smell narcotics and can sniff out explosives. And they do it far better (and much cheaper) than the chemical sensors that people have come up with so far. I’m sure that bad or unscrupulous handlers can manipulate their dog, but that’s not the same thing as saying their use is “junk science.”
Do you also think search and rescue dogs are worthless and their use should be discontinued because the whole idea is junk science?
Comment by Barry —
October 6, 2009 @ 10:16 am
Doug, perhaps you could rephrase your comment, so as to get a *passing* grade in high school logic?
Comment by Eric the .5b —
October 6, 2009 @ 12:13 pm
High school logic?
Comment by Doug T —
October 6, 2009 @ 1:11 pm
OK, I’ll try.
Dogs can detect drugs, bombs, and human smells, among other things. They do so better than any alternative detectors which are available. Although they are not perfect, they are still valuable. Saying something is imperfect is not the same thing as saying any use of it is “junk science” and should be discontinued.
If you do think police dogs are so unreliable as to be useless, do you feel the same way about other more or less identical uses of dogs, such as search and rescue? If not, why not? What is the difference that makes dogs responding to a trained smell cue in one case junk science, but in the other case a valuable asset?
Comment by dhex —
October 6, 2009 @ 1:14 pm
High school logic?
i’ll give it a shot:
“police dogs are gay.”
Comment by Thoreau —
October 6, 2009 @ 1:54 pm
I don’t object to any and all uses of police dogs. If a police dog can lead a cop to find evidence in some sort of open or public area, or an area where a person does not have a “reasonable expectation of privacy” or whatever the legal standard is, and that evidence is useful, great.
However, the mere fact that the dog allegedly responded to something should not be sufficient cause for, say, tearing your car apart.
The actual finding of the body or bomb or whatever in some sort of open area is objective: There is in fact a body, or a bomb, or whatever. However, the officer’s assertion that the dog was reacting to a scent should not be sufficient to do a search that would otherwise not be permitted. Too much subjectivity. We might as well allow searches any time a cop says “I had a hunch and wanted to check it out.”
Comment by Barry —
October 6, 2009 @ 2:04 pm
Doug T – better. However, you forgot that (a) the standard of proof for criminal cases is reasonably high; better than nothing is not good enough. For search and rescue, ‘better than nothing’ is what you use, if you have nothing else (e.g., nobody’s going to turn down a helicopter with FLIR, and rely only on dogs).
(b) As Thoreau said, when the dog’s alleged reactions are used as an excuse to take damaging actions, we’d like some good evidence that the dog’s reactions are reliable, rather than arbitrary.
Comment by Doug T —
October 6, 2009 @ 3:12 pm
Fair enough. I agree with you regarding searches. I think I over-keyed on the “junk science” toss-off wording, which I think sells dogs short.
On the flip side, the case described in the link seems different. The dog investigated a crime scene and his tracking gave a reason to regard a neighbor as a suspect. I think that’s perfectly legitimate. It’s bad for the neighbor, but with any crime, police use a variety of means to try and identify suspects, none of which are perfect. And sometimes innocent people get accused of crimes. But the best remedy there is in the courts weighing tracking evidence correctly, rather than removing it as a tool for police.
To put it another way, the dog tracking is the beginning of the process, rather than the end of it for the police and the courts. As it seemed to be in this case, where they eventually cleared the suspect.
Comment by Kelly —
October 7, 2009 @ 2:14 pm
http://en.wikipedia.org/wiki/Clever_Hans
Pingback by Rad Geek People’s Daily 2009-10-07 – Wednesday Lazy Linking —
October 13, 2009 @ 1:21 am
[...] Smells like improbable cause. Thoreau, Unqualified Offerings (2009-10-05). By Thoreau You know how a cop can search any car/person/house/apartment/boat/plane/train/automobile/blimp/etc. he wants if he claims that a dog did something to suggest that it might smell drugs? Well, a couple of bloodhounds in Texas reacted to the smell of a deputy sheriff during a murder investigation, and the guy… (Linked Wednesday 2009-10-07.) [...]