UNIVERSITY OF CALIFORNIA. SANTA CRUZ 
BCM£LKY • DAVU • mviNE • LOS ANGELES • MVEASIDE • SAN DIECO • SAN FKANCISCO 
SANTA BARBARA • SANTA CRUZ 
omcx or the chancellor 
SANTA CRUZ. CALIFORNIA 950A4 
January 11, 1982 
Director, Office of Recombinant DMA Activities 
Building 31, Room 4A52 
National Institute of Health 
Bethesda, MD 20205 
Re: Proposed Revisions of NIH Guidelines for Research 
Involving Recombinant ONA Molecules 
Dear Sir: 
The case for the proposed revlslon(s) of the Guidelines rests upon the 
arguments presented In the "Evaluation of the Risks Associated With 
Recombinant DNA Research" prepared by the Working Group In Revision of 
the Guidelines. This Is a thoughtful document which attempts to set 
forth -- and to refute as to "extremely unlikely" — the various putative 
scenarios of hazard from recombinant DNA experiments. The net result of 
this presentation Is to dismiss the possibility of hazard from all 
recombinant DNA experiments except those Involving known , very pathogenic 
agents (highly lethal toxins or highly dangerous pathogens) , I.e., to 
dismiss the possibility of the creation of a novel pathogen, or of a 
more virulent form or a new mode of dissemination of an existing pathogen. 
The arguments are plausible -- but In no Instance are they quantitative. 
In no Instance do they take Into account the possibility of accident, of 
multiple and non-linear Interactions, of evolutionary processes producing 
unexpected consequences In unplanned directions over a period of years. 
I regard both proposed revisions (Federal Register, Dec. 4, Dec. 7) as 
too extreme, although of the two, the latter (presented by Dr. Susan 
Gottesman) Is the preferrable. 
I would suggest that the minimum provisions of a revised set of Guidelines 
should Include that: 
1) The Guidelines be mandatory, not voluntary; 
2) Accordingly, there be penalties for non-compliance; 
3) Institution Biohazard Committees of specified composition be 
retained with functions essentially as described In the 
Gottesman revision; and 
16491 
