LSA Committee on Ethics of Disclosure

[RonButters at AOL.COM]

I received the following response from Larry Solan, President of the 
International Association of Forensic Linguists. Larry also indicated to me in another 
private e-mail messsage that he would be willing to serve on an LSA commitee 
to investigate the posiblity of formulating a policy on the ethics of 
disclosure. 

I draw your attention in particular to Larry's last paragraph.

---------------------------------

 I was very encouraged to read Ron Butters’ inquiry concerning any obligation 
to disclose that a scholar has participated in a case on behalf of one of the 
parties before writing or speaking about it.  It is especially noteworthy 
because Ron publicly questioned his own practice.  I can think of no better use 
of this list [the Forensic Linguistics listserve] than to raise such questions.
  
 I favor disclosure as the better practice, whether or not it is considered 
an ethical imperative, as Geoff Nunberg suggests.  None of us is immune from “
observer bias.”  That is why the testing of new drugs requires double-blind 
trials.  Not only does the patient not know what is being administered, but the 
doctor administering the drug does not know either. 
  
 A recent article in the California Law Review discusses this phenomenon in 
forensic contexts: 
  
 Risinger, D. Michael, Michael J. Saks, William C. Thompson, and Robert 
Rosenthal (2002). “The Daubert/Kumho Implications of Observer Effects in Forensic 
Science: Hidden Problems of Expectation and Suggestion,” California Law Review 
90: 1-56.
  
 The article provides a rich discussion of observer bias.  The authors argue 
that the problem is especially prevalent in police labs, where it is clear in 
advance which findings will be “helpful,” even when everyone in the lab is a 
decent, honest person.
  
 While the authors do not discuss linguistic cases, there is no reason to 
believe that linguists are any more immune from observer bias than are 
researchers in other fields.  Linguists who report on work that was sponsored by a party 
with an interest in the result give their audiences the opportunity to take 
possible observer bias into account when they disclose their involvement at the 
outset.  In general, that strikes me as the preferred way to proceed.
  
 Of course, such rules immediately lead to difficult cases:  For example, 
what happens when working on a case triggers interest in a phenomenon, which the 
author then writes about generally, perhaps without even mentioning the 
original case?  It is worth discussing such line-drawing issues, but I do not think 
that they should swallow up the core cases in which disclosure is a simpler 
matter.