APPENDIX Q 
WHEN MAY RESEARCH BE STOPPED? — COHt.N 
uct) presents risks (cither of miohrc or of accident) so 
great as clearly to outweigh the benefits reasonably 
anticipated.’’ A misfire might be, for example, the cre- 
ation, through DNA recombination, of an organism 
with unintended pathogenic capacity against which 
ordinary antibiotics proved ineffective. An accident 
might be, for example, the undetected escape, from a 
laboratory thought to be sealed, of a micro-organism 
giving rise to contagion. 
This principle also is sound and applicable. Bring- 
ing it to bear upon DXA research, however, is com- 
plicated, and the outcome of that application is un- 
certain. Just here lie the major technical problems 
that have been and remain the focus of m*uch scientif- 
ic debate. The problem of containing the recombined 
DNA, either by physical retention within the labora- 
tory or by weakening tlie host organism so as to ren- 
der it not viable outside the laboratory, has been the 
major topic in controversy over what is unreasonably 
risky and what is not. Only in the light of the present 
state of effectiveness of containment mieasures can 
risks be rationally estimated. Hence the emergence of 
guidelines (laid down by the National Institutes of 
Health) for containment, and for permissible risks giv- 
en known levels of effective containment. Hence, too, 
the need to reassess what is reasonably safe to do in 
the light of existing technical capacities to contain — 
especially since the capacity to contain biologically, 
by “disarming the bug,” is being steadily improved. 
R.ecombined DNA molecules do create special dan- 
gers, which do, rightly, require special attention to the 
conditions and precautions under which specific re- 
search activities are carried on. 
Still, the principle accepted here, that risks must be 
minimized, and never allowed to exceed the reason- 
ably anticipated benefits, is one of general applica- 
tion. It applies to all research in medicine, in physics, 
in biology, in aeronautics, and so on. It has a bearing 
upon DNA research, to be sure — but only to the ex- 
tent that the risks encountered in a specific experi- 
mental project within that domain appear to equal or 
to outweigh the anticipated benefits of that project. 
Critical here are the risks to persons other than 
those involved in the research (6b): the people in the 
street who may be endangered by accident or misfire. 
Risks to experimental subjects (6al) must be screened 
by human-subjects review committees [described 
above under (5)] designed precisely for that purpose. 
Risks to researchers themselves or to others formally 
involved in the research project (6a2) may be grouped 
for present purposes with risks to outsiders. It is for 
this combined group that the key question arises. 
That key question, now heatedly argued, is this: 
Are the risks of recombined DNA, whether of misfire 
or of accident, of such enormity and such probability 
as to justify prohibition of further research in that 
sphere? 
The first thing to notice about this question is that 
although here framed in the singular, it must in fact 
be asked about a host of very different research pro- 
posals. Estimating risk-benefit balance for some pro- 
posed investigation is often a vexing task. But it 
should be empliasized that the decisions called for in 
this family of cases are not, in principle, different from 
those we are commonly obliged to make when data 
are incomplete, the time-frame long, the object risky 
but promising. In this family of cases, as elsewhere, 
we will do the best we can. 
Some ccriiend that in this sphere of research, un- 
like others, the general sum of anticipated risks, tak- 
ing into consideraiion both their degree of serious- 
ness and their probability, outw'eighs the general sum 
of anticipated benefits, their value and likelihood sim- 
ilarly weighed. Therefore (they conclude) the proper 
estimate of risk-benefit balance calls for cessation now 
of all further research in this sphere. 
This argument is unsound. Consider: 
First, the key premise is false; it makes a stronger 
claim than the evidence supports. Granting that the 
present stale of knowledge is short, all indications are 
that, if one weighed, as on a balance scale, the ex- 
pected goods in view, multiplied by the likelihood of 
their probability, the result would be very different 
from that supposed by this argument. Everyone will 
allow that there are dangers, some not yet fully 
known, but the most careful and sophisticated dis- 
cussions of these dangers, taking severity and proba- 
bility into account, do not begin to show that they out- 
weigh, or even approach, the sum of advantages like- 
ly to accrue from such research over the long term.* 
The fact remains that some future proposals for in- 
vestigation in molecular genetics, because of the spe- 
cial risks entailed in the process (and with full consid- 
eration of the nature and probability of the benefits in 
view) may, after thoughtful deliberation, then be re- 
jected by the research institution. Clearly, such rejec- 
tions, if they transpire, will require the continuing ac- 
tivity of an institutional review board, on the model of 
a human-subjects review committee, but with a dif- 
fering focus. Rejection on the grounds of excessive risk 
is a step that must not be taken lightly, but it proba- 
bly will be taken in some cases. 
Since general prohibition is not in order, and ap- 
proval for individual proposals must be given on a 
case-by-case basis, it is appropriate that such contin- 
uing review bodies set the conditions for the permissi- 
ble pursuit of the risky inquiry proposed. Here lies the 
•A number of bodies, with members both in the sciences and in the hu- 
manities. have deliberated long and carefully upon the likely balance of risk 
and benefit, over long term and short, of recombinant DNA research. Per- 
haps the two most thorough and probing studies are the Report of the 
Working Party on the Experimental Manipulation of the Genetic Composi- 
tion of Micro-Organisms. Presented to Parliament by the Secretary of State 
for Education and Science, London, January, 1975 (this document is widely 
known as the Ashby Report, after the Chairman of the Working Party, Lord 
Eric Ashby), and the report of the LIniversity Committee to Recommend 
Policy for the Molecular Genetics and Oncology Program, presented to the 
vice-president for research at the University of Michigan, Ann .Arbor, 
March, 1976 (attached to this document, widely known as the Report of 
Committee B, is a dissent and a reply to the dissent). Subsequently appeared 
a critique of the Report of Committee B, a response to that critique by Com- 
mittee B, and a separate endorsement of the Committee's original report, all 
presented to the Regents of the University of Michigan. 
Appendix Q--7 
