Before he learned to read the messages written in blood, Lt. Rod Englert tended to ignore crime scene carnage.
"It takes very little blood to create a huge, bloody scene," Englert said. "I used to put the blood on my `pay it no mind' list because I didn't understand it."That changed about 10 years ago when Englert, a homicide division commander with the Multnomah County, Ore., sheriff's office, learned to trust the arcane forensic science of blood-pattern analysis.
A key prosecution witness in the preliminary hearing on murder charges against Sam Kastanis, who is accused of stabbing and bludgeoning to death his wife and three children, Englert agreed to explain blood-pattern evidence with the understanding the court case would not be discussed.
Englert said he has been in law enforcement for 30 years, and for 20 years has been a homicide investigator. Questions raised during classes he taught on solving unresolved murders in the early 1970s prompted his curiosity about the ghastly evidence he once disregarded.
Now, he is one of a handful of investigators in the United States considered knowledgeable enough in blood-pattern analysis to be called on as expert trial witnesses.
Englert said that blood-pattern analysis - also known as blood-spatter analysis - began in Germany in the late 1930s, where scientists fired real bullets into living animals and then studied the patterns the gore created.
A New York researcher further advanced the science in the late 1960s. Englert said there are now about 200 members of a group called the International Association of Blood Pattern Analysts, which he once headed.
Though investigators no longer use live animals, their technique remains necessarily gruesome. Spatter analysis can be explained through metaphor - think of flinging water off your arm in the bathtub, for example - but blood, literally, is thicker than water.
So Englert and his colleagues re-create a crime scene in a laboratory setting, where they may stab, bludgeon or shoot specially made, blood-filled mannequins to test their hypotheses.
Most blood patterns can be traced back to an assailant's hands.
"That simplifies it a lot," Englert said. "What did the hands do?"
From there, an investigator classifies the patterns into three main categories.
Low-velocity patterns include drops of blood, handprints or smears. Medium-velocity patterns show evidence of blunt-force trauma - for example, if someone bloodies another's nose, there won't necessarily be blood flying. If an assailant strikes again, however, spatters will tell the story.
High-velocity patterns usually - but not always - indicate a bullet wound. A bullet creates mists of blood, atomizing the blood into particles so small they can't be seen with the naked eye.
Englert said he examines his every assumption, even those gleaned from reading other forensic scientists. He prefers to go to the scene himself - but can work from photographs - and insists on recreating the entire scene as faithfully as possible in the laboratory so as not to interpret a blood pattern out of context.
As always, mistakes are possible, but Englert said he doesn't take mistakes to the courtroom. "I have a lot of confidence. But I'm still willing to admit error," he said.
If someone starts looking guilty, he'll redouble his efforts on that suspect's behalf.
"The best thing is to work as an advocate for those who are suspects," he said. "When you can't salvage them, they are in serious trouble."