An Inteseting Process for the Construction of Knowledge

Mud volcano Lusi has been erupting for two years,
leaving 30,000 people homeless.

Source: "Mud eruption 'caused by drilling'," by James Morgan, BBC News, November 1, 2008.

The American Association of Petroleum Geologists held a session recently to evaluate the evidence that a mud volcano eruption in Indonesia was triggered by drilling for oil. The Lusi mud volcano has been spewing out mud since 2006 and some 30,000 people have been displaced. Indonesian police are investigating the causes of the eruption in order to establish who should indemnify the refugees. The results of the deliberation of the geologists should inform that process.

At issue is whether the eruption was caused by the drilling 150 meters away from the eruption, by a magnitude 6.3 earthquake centered 280 km away two days earlier, or by a combination of both factors. The oil company released the measurements made in the days prior to the start of the eruption for analysis for the meeting.

Two professional geologists made presentations to an audience of some 80 other geologists attributing the eruption to the drilling and two made presentations attributing it to the earthquake. The session was moderated by a professor of geology who also is a professional football referee.

Following the presentations, the assembled geologists voted their conclusions. The majority supported the view that the drilling caused the eruption, with smaller numbers holding

• that both factors helped cause the event or
  
• that the data was inconclusive.

Only three voted that the earthquake alone had caused the eruption.

The article did not state how the results of the meeting would be used in the Indonesian legal process.

Comment: I found the case to be interesting because it involves a professional association seeking to construe the meaning of observations where that meaning has a public policy implication. The evidence was apparently such that not only could it not have been validly interpreted by lay people, but even professional geologists did not agree on its meaning. In the normal court case, the experts present evidence to a jury of lay people, but in this case it was presented to an audience of experts.

Obviously, volcanology is not an experimental science so the interpretation of the evidence in this case was epistemologically limited. (Of course, science is full of such ambiguity.) The interpretation of the data has already resulted in one peer reviewed journal article, and now that the data is out will no doubt be subjected to further review by professional geologists. Thus the scientific process of construing the data and determining the likely cause of the eruption is not finished.

I think this work by the association of petroleum geologists is a useful precedent for forensic science; perhaps experts should be employed not just as advocates but also as a panel of judges when the interpretation of scientific observations is both difficult and (financially) important. JAD