The widely accepted narrative that fire investigation has improved and that the myths of the past have been rejected is very appealing but I tend to think over exaggerated.

While some clearly unreliable methodologies (crazed glass, concrete spalling, witness statements of a “particularly hot fire,” etc.) are no longer popular, a couple of the “old wife’s tales” still have traction. Specifically, I still see fire investigators pointing to “pour patterns” or a hole burned into a floor as a valid method for confirming the presence of an ignitable liquid. Dog handler testimony to unconfirmed ADC alerts (still admissible in most courts) is also an ongoing problem.

Even more concerning, other equally unmeasured, unreliable and unscientific methodologies have taken their place, changing the process but maintaining the result.

As I point out in this op-ed (http://www.thecrimereport.org/viewpoints/2016-04-the-misleading-science-of-arson-investigations#.VwPXRXEzvI4.mailto), fire investigators work in a forensic discipline free of any measurement of accuracy or error. And although the courts usually view NFPA 921 as a standard of care, fire investigators generally regard it as a guideline, not a standard. The title of the document supports this interpretation, allowing the investigator to follow it or dismiss it as needed.

As a result, even practices specifically rejected by NFPA 921, such as Negative Corpus, continue to be routinely used and testified to in court.

The on-going retrial of Daniel Dougherty in Philadelphia (http://articles.philly.com/2016-03-30/news/71903170_1_arson-fire-science-fire-scene) is an excellent example.

Here, a fire investigator for the state testified to multiple points of origin in a post-flashover/full-room-involvement fire. His testimony did not involve “pour patterns,” “crazed glass,” or an unconfirmed dog alert. He simply used the most imprecise measuring tool available (his own mind), analyzed all of the factors (both relevant and irrelevant, scientific and circumstantial), and came up with a conclusion that matched the prosecution’s case.

That this form of testimony was acceptable at Dougherty’s original trial in 2000 is outrageous enough; that is being used in 2016 is mind boggling.

In the end, I tend to think that ideas that fire investigation has “come a long way” towards greater reliability and validity are kind of like “pour patterns,” “crazed glass,” and “alligator charring” to show the presence of an ignitable liquid – a figment of our collective imagination.

Paul Bieber
Director - The Arson Research Project
www.TheArsonProject.org