example, one of the reasons for requiring a high level of con- 
tainment for experiments with primate DNA is that we know so little 
about such recombinations. On the other hand, the reason for 
prohibiting experiments which would transfer antibiotic resistance 
to non-resistent species is that we know such tremsfers will 
certainly impair our ability to cure human disease. Similarly, 
special provision is made in the Guidelines for experiments 
using E. Coli host-vector systems because we know that E. Coli 
colonizes in the human intestine. But other prokaryotic host- 
vectors, which do not colonize in humans, may prove equally 
harmful to the environment in ways we cannot now foresee. 
The elaborate classification system of the current 
Guidelines should be abandoned in favor of a much simplier 
system. NRDC suggests that the Guidelines at most have three 
classifications for experiments: prohibited, very hazardous 
and hazardous. Such a classification system would provide some 
significant protection from all recombinant DNA experiments 
and thereby protect us from our ignorance about the consequences 
of this research. Using no more than three categories also frankly 
recognizes the imprecision of our judgments about risk. 
24 
Appendix K — 150 
