72 
1 but most useful type. 
2 The principle of the containment level matching 
3 the most potentially hazardous component should apply here, 
4 just as well as in the rest of the guidelines. No danger, 
5 no containment. The absence of this null category has 
6 in the past in part clouded public discussions, because 
7 despite the series of graded containment levels, recom- 
8 binant DNA researchers are always visualized as one 
g single thing. And if hazard was assumed for any of 
10 its parts — for example, forbidden experiments -- then 
11 the whole thing was hazardous. 
12 I think we just heard some indications of 
13 that kind of reasoning. 
14 The new duties of the IBC will make the guide- 
15 lines a bit more like guidelines and less like rules. As 
16 the guidelines note, one can't anticipate every kind of 
17 experiment. Hence, flexibility is needed. Basic research 
18 moves in erratic directions because its data base is 
19 continually changing. New information, ideas and techniques 
20 suddenly appear, and experimental approaches are modified, 
21 dropped, and new ones initiated. A distantly approved MUA 
22 tends to codify one approach. Were the guidelines really 
23 guidelines, most amended MUA's wouldn't be needed. New 
24 approaches generally fall into the category of previously 
25 approved. Under the new system it would require only a 
[164] 
