Q (3S

Newsletter from the Lederberg Laboratory

 

December 2, 1992

Dear Friends of the Lederberg Lab:

On this occasion, our topic will be some issues in science policy. This is especially timely
during the period of presidential transition. The end of the cold war and the urgency of this
country’s economic recession will both enforce drastic changes in science and technology pol-
icy. Many people will wonder what’s happened to our "peace dividend". At the same time,
the phasing down of defense procurement and of the size of the armed forces is leaving many
people hunting for jobs. One of this country’s great strengths has been the vigor of its basic
science. Many people in Washington are asserting that science must be more effectively har-
nessed to technological applications, and in turn to economic productivity. This is a laudable
goal; but there will be much frustration and mischief if there is not a clearer understanding of
how basic science is done, and the ways it can be harnessed to a specific mission.

In biomedical science we have lived with these paradoxes for many years. Although, the im-
provement of human health is understood to be the guiding goal of our research, we have also
understood that all the easy paths, for example, for the cure of cancer or of AIDS, have been
heavily trodden long since. This is not like the more applied engineering project of space-
ships to the moon. There we already knew exactly where the moon was -- more importantly,
the laws governing its motion, celestial mechanics, had been well known for 200 years. We
also knew fairly precisely how to build rocket engines, and could estimate (within a factor of
five) how much it would cost and how long to build the necessary hardware. I don’t underes-
timate the engineering skill and imagination that was needed -- but the achievability of the
project was firmly grounded on existing scientific information.

Indeed, basic science is quite different. We start with tentative hypotheses, our models of
what the world is like, and then try to devise tests to probe those ideas. We are constantly
trying to overturn or to revise the “established truths”, or to fill in areas of perplexity and
darkness. Our greatest accomplishment (and joy) is when we have made an unanticipated
discovery, when we are confident that we can discard our preconceptions, our experiments
having shown that we were wrong, and we can revise our paths with greater confidence. This
relentless criticism is at the very root of science: and we don’t do it alone. The whole point
of publication is to expose our claims and data to the scrutiny of the worldwide community of
science. And they are more than happy to collaborate in this quest, by letting us know right
away when we were on the wrong track. Sometimes we will get the courtesy of a few hours
advance notice, but the entire process is highly public. And it is our obligation to be party to
this objective and unceasing scrutiny, to volunteer all we know that might bear on the conclu-
sion.
How different this is from, for example, trial procedure in our legal adversarial system. Here
each side is expected to garner whatever persuasive tools and evidence it can for presentation
to a third party jury. Neither will volunteer any more than it is obliged to, for that might be
damaging to its case. The jury is, hopefully, impartial -- but what it is allowed to hear is con-
strained by rules of evidence that often have little bearing on the contribution of some facts:
for example, evidence secured by unwarranted search. This, of course, is what to be expected
when there is a tradeoff between the investigative process qua process and other important
values like our constitutional guarantee against unlawful search and seizure.

In any case, a scientific finding will be judged by some objective test of its predictive power
-- it is an insight into the laws of nature. A legal finding usually is concerned about a single
historic event, and is necessarily validated by the community’s more subjective sense of fair-
ness and justice.

The detailed method of science is best illustrated by specific examples -- hopefully some of
our Lederberg Lab "get togethers" in the future will highlight that. The most important thing
to note for for now is that it is far from a linear process, marching predictably to rigorously
defined goals. I have illustrated that with the following diagram. The term “epicycles" refers
to the back loops on the circle -- what the ancient astronomers figured as the orbits of the
planets, not knowing that earth and Mars were both circling the sun. The diagram displays
how much of science is a backloop: hopefully 3 steps forward for every 2.9 in reverse; often
even the definition of the problem will be revised as we encounter the experimental facts and
derive new insights. Under the pressure of intense public interest in in the results of still
unfinished studies, conclusions may be adopted before they have been critically validated. We
see many examples of confusing about-faces in fields like human nutrition.

Political leaders may then be asking the impossible of basic science, if they seek a tangible
pathway of applications and products that will affect, for example, the balance of trade. We
badly need ever more innovative technology, but not at the expense of this substratum of
scientific inquiry. In the short run, economic competitiveness will mainly depend on capital
formation, efficiency in management, and better process technology, none of which, I admit,
have much to do with basic science. But science can provide two irreplaceable things for us:
* A discipline about facing reality, a nature that is not amenable to sweet talk or self delu-
sion

* The basis for long-term revolutions in technology that could not even have been foreseen at
the time of discovery. Examples of the latter abound: all of electronics stems from arcane
studies of subatomic particles. The laser from mathematical inquiries. Biotechnology from
the roots of DNA structure and the natural history of breeding in bacteria and viruses.
The priorities for science policy have more to do with "science for policy” than "policy for
science”. Our nation and the world face innumerable problems crying out for compassion,
resources, managerial and political skills, and a high level of scientific and technical expertise.
Some of these are quickly summarized in the enclosed editorial. Procedurally, the first priori-
ty is for Mr. Clinton’s transition team to come to grips with setting up the executive struc-
tures that will provide the highest quality of advice. In the technical arena that means ready
access to the skills and the critical reaction of the scientific community. The so-called Repub-
lic of Science is just as important as the brainpower of individuals in providing science with
its special brand of authenticity.

As a final note, I have enjoyed co-chairing the Carnegie Commission on Science, Technology
and Government during the past 3 years. It has generated a number of reports that may be of
special interest to some readers -- so take a look at the enclosed list.

With best wishes for the forthcoming holidays, peey

P.S. If you haven’t gotten to it yet, it’s not too late for ygur flu shots. The minor discomfort
is a very good bargain to avoid the flu, and to help keep it from being passed on to others.

Epicycles of Scientific Discovery

Education, motivation and socialization
of an investigator; includes perception
of demand: needs of society; anticipation of
reward (prestige support, profit, poetic satisfaction)

e
Aitract new students: a oY Theoretical Framework Discovering,

augment or establish *. Formulating Questions Conscious
new organization YK and unconscious immersion
Establish technical - ® Designing exptl. approach

applications

Community reaction: rejection,

be overlain with mis- or neo-
understanding

* \ (de novo or from literature)
mi on:! assembling support staff equip-
oblivion, or canonization. May ® e ment, reagents, SRMs; resources:
| (grant appl?)
Publish
° AE @ Conduct experiment

Draft ms. informal presentation

to colleagues OX. a analyze data
s

Draw conclusions, abduce theories, critical match to data