Cross-Species Lineups and "Forensic Facial ID Experts"

Australia's high court yesterday denied the final appeal of a man charged with murder and attempted kidnapping in the Australian outback, where one piece of government evidence was the identification of the accused's dog from a lineup procedure that the defense alleges was unduly suggestive. Given the serious problems with mere cross-racial IDs, it's tough to imagine that even a carefully crafted "dog lineup" could have the slightest hope of resulting in a reliable identification. That the Australian court gave this evidence its stamp of reliability and allowed it in front of a jury is inexplicable.

And the government in this case wasn't satisfied with a cross-species ID. In addition, they managed to introduce the testimony of a "forensic facial ID expert" -- not to be confused with an actual scientist -- who apparently viewed some video footage from a truck stop and testified, bolstered by the court-sanctioned semblance of "expertise," that the man in the video was "definitely" the defendant. I'd like to see the validation studies and error rates on that particular trick.

Also central to the government's case was an identification of the defendant by the woman who witnessed her husband's murder, but who only reported the defendant to the police after seeing his photo on an internet news site. The suggestive and corrupting effect of that sort of exposure prior to an identification procedure is also well documented.

It appears as though there may have been other strong evidence in this case, but the existence of corroborating evidence should not serve as justification to throw basic standards of reliability out the window. The reliability and admissibility of any piece of evidence should be assessed on its own strength. There is nothing in the rules of evidence to suggest that a purported DNA match should serve to diminish the admissibility requirements for all other evidence in a case, making way for the evidentiary free-for-all that this case appears to have been.