515 



anoing of information sources to cancel out bias is difficult. Those who 

 testify have many different classes of bias — personal, political, eco- 

 nomic, ethical, intellectual, institutional, social, associational, rela- 

 tional, etc. Those who select the witnesses also have biases. Since only 

 the most extreme forms of bias are obvious, it is not feasible to select 

 witnesses according to any precise calculus of bias balancing. Indeed, 

 a person selecting Witnesses is sometimes oblivious to the biases of a 

 candidate w*hose biases he happens to share. The only feasible answer 

 to the problem of witness bias is a plurality of witnesses with knovN-n 

 differences in views; when such witnesses confront one another in an 

 adversary proceeding they tend to ferret out each others' biases and 

 expose them to other observers. However, mere numbers of Vvitnesses 

 testifying as a panel offer no assurance per se of eliminating bias : when 

 all members of a group share a bias, the tendency is forthe group to 

 take a more biased position than would be taken by any single member 

 alone. Finally, those who evaluate the information also have biases. To 

 cancel out evaluative bias, it may be useful to circulate staff summaries 

 and reports of the results of information gathering to outside persons 

 with technical qualifications in the field concerned. Their reviews of 

 the reports can help to increase the objeotivity of the findings and 

 conclusions. 



TechnicaUy Sound 



Since all scientific and technological testimony deals in relativistic, 

 probabilistic subject matter, it is important that the tech]iical witness 

 both possess and communicate information in properly weighted and 

 evaluated form. The witness needs to be knowledgeable about quantita- 

 tive relationships within his discipline, familiar with the language of 

 his discipline, and skilled in interpreting the data of liis discipline. 

 The essential truth is presented in the form of approximations. 

 Orthodox positions need to be challenged : unorthodox positions m.ust 

 bear the burden of proof. "^^Hien technical witnesses wrth comparable 

 professional qualifications offer conflicting testimony, it is necessary to 

 analyze their testimony and supporting evidence, in order to determine 

 whether they are operating on the basis of different hypotheses. If so, 

 the testimony of additional witnesses with comparable qualifications 

 may be helpful to shed light on the relative validity of the conflicting 

 hypotheses. Basically, assuring technical somidness is an iterative 

 process of successive closer approximations. 



Arrangements To Assure Completeness of Technical Information 



The importance of dealing thoroughly with technical issues is para- 



m.ount. Failure to anticipate the hidden dangers in a new technological 



situation can be catasJtrophic. Plausible decisions can set in motion 



irreversible processes that cause irreparable damage. The concluding 



quotation in the report on the Pesticide Case warrants reproduction 



here : 



A well intentioned but poorly informed society is haphazardly deploying a 

 powerful, accelerating technology in a complex and somewhat fragile environ- 

 ment. The consequences are only vaguely discernible. 



Clearly, it is the function of the decisionmakers to satisfy them- 

 selves that all tlie essential questions have been asked and answered. 

 This does not imply that INIembers of Congress need to make them- 



