#31 from R&D
Innovator Volume 2, Number 4
Three Silly Notions About
Technology Transfer: And
One That's Not
Berry is professor of chemistry at The University of Chicago.
He was awarded a MacArthur Fellowship in 1983.
Dr. Berry was a cofounder of the Telluride Summer Research
Center, which holds annual multidisciplinary workshops.
What are the roles of basic and applied research?
What are the roles of industries and universities?
Industries need innovation just to stay in business, let
alone compete successfully in the global economy.
But too many "solutions" to questions of
innovation rest on some misconceptions of the relationship between
science and technology, as well as between academia and industry.
Here are the most
Basic science precedes and leads to applied technology.
The findings of academic science usually lead directly to
Academic and industrial science have a natural
collaborative relationship because the results of the former solve
problems of the latter.
these notions seem like common
sense, but scrutiny shows that they are foolish. When well-meaning believers use these ideas as their basis
for orchestrating relationships among university, government, and
industry, all parties can suffer.
In fact, there is
a sound basis for a relationship among these players—but before
we discuss a healthy R&D enterprise based on the realities and
strengths of existing institutions, we must understand why these
three notions are misleading.
belief—that the applied flows from the abstract—is not wrong
so much as dangerously incomplete.
Much of 20th century technology (radio, television, lasers)
did indeed arise from the basic ideas of quantum mechanics and
relativity, just as much of 19th century technology stemmed from
the fundamental scientific discoveries of Faraday, Maxwell and
On the other
hand, many basic theories have emerged in response to practical questions. Probably the most striking illustration of this "reverse
flow" is thermodynamics, which evolved when scientists faced
an utterly practical problem:
What was the minimum coal required to pump water from
characterizes the motivations for productive "reverse
well-intentioned, would-be reformers of the R&D enterprise
suppose that by channeling funds to academic scientists—not to
be used for basic or curiosity-driven research, but rather for
goal-directed applied research—the academics, with their prior
basic knowledge, will find solutions to well-posed, exogenous
problems, on command. This
notion has no foundation in reality.
The genius of successful "reverse flow"
scientists has been their capacity to respond to challenges of
applied problems with the same spontaneity, curiosity, and
originality that characterizes the quality of their work in basic
science. One cannot
legislate the unforeseen, even by manipulating funding
danger of believing the incomplete notion that the applied always
flows from the basic is that inadequacies in technology and
applied science will be blamed on the upstream partner's
"failure" to do basic research that leads directly to
Good Are Universities?
This brings us to
the second misconception: That
the findings of academic science normally lead to marketable
products. In the
United States and elsewhere, basic science—one of the most
innovative and productive of human enterprises—has become
unattractive to legislators, government agencies, and industry.
This has occurred despite the fact that basic science has
been performing well in the United States, Europe and Japan in its
primary function of generating fundamental ideas and producing
Basic science is
a high-risk, high-return activity.
Its costs are low and its return on investment, overall,
seems to be somewhere between 25 and 40 percent, according to
different economic analyses.
There is no reason to suppose that the output of basic
science in the 1990's is any less applicable than it was in the 1960's and '70's.
Investors in basic science realize that a long-term
research program is an investment portfolio, that the likelihood
is small that any single project will bring a large return but the
probability is high that from the overall portfolio, high-yield
results will emerge. History
has fully justified this concept and rewarded those investors.
argued that innovative applications of new science have not been
forthcoming, because basic scientists have not solved our social
and economic problems!
has problems that basic science has not solved.
Basic science may even be relevant to some of those
The question is,
what would facilitate more effective application of results of
basic science to societal problems?
Who has the incentive to make those innovative
not consumers—they’re at the end of the chain.
The bottleneck, if it is real, lies somewhere between the
basic scientists and the marketers.
Marketable, innovative technologies—it seems to
me—should spring more freely from the minds of people close to
markets and manufacturing than from people close to advances in
Could there be a
lack of incentive for creative people in industrial settings to
advance innovative ideas? Could
it be that industry is simply not very hospitable to innovation?
Those of us who
train young scientists and engineers for industry answer that
question with a loud, vigorous "YES."
We watch firm after firm cut back its investment in R&D
during a period of falling interest rates; this is deeply
disturbing. All the
doctrine of traditional economics argues that the value of
investment in long-term research should go up
when interest rates fall. Firms
that do the opposite are bent on technological suicide.
With bitterness, we watch managers sacrifice the long-term
health of their firms to competitors who understand time horizons
and the role of R&D in global markets.
The third silly
idea—that basic university science has a natural affinity for
technology in industry—rests on two misconceptions.
First, that the principal university product for industry
is its steady flow of scientific results.
Second, that the modes of work in business and academia are
notions are terribly wrong. The
first, because universities' cardinal product for industry is manpower, well-trained, versatile scientists able to attack a
succession of ever-changing problems.
information may be relevant, even useful, to industrial
problems—but the reason industrial scientists use information
generated in universities is that university scientists publish
their results. If an industrial scientist notices something relevant in the
literature, there is a reasonable possibility that the new
information will be put to use.
This is a matter of serendipity far more than a natural
course of affairs.
misconception—that the modes of scientific work in industrial
and academic settings are naturally compatible—is the result of
naive wishful thinking. The
best way a university scientist “markets” the primary
“product,” a new
graduate or postdoctoral fellow, is by showing off all the strong
qualities of that young person in the standard scientific venues,
journals and meetings. Credentials
depend on the quality, novelty and impact of the person's work, so
it must be published quickly.
R&D group, on the other hand, succeeds by translating its
ideas into new, profitable products before competitors can produce
better or cheaper products. Hence
ideas developed in industry are secret until they have been
exploited. Furthermore, once an idea has been published, industrial
scientists have little motivation to build on that idea unless
they can contribute enough to give them a competitive edge.
For all these
reasons, it’s difficult for academic and industrial scientists
to work together, except in the relatively uncommon situations in
which publishable and proprietary parts of the problem can be
separated. Furthermore, there are few incentives for either academic or
industrial scientists to look hard for such opportunities.
the real relationship between basic science and applied
industrial history gives us a few hints.
Although in some industries, such as auto manufacturing,
fundamental science contributed little compared with the
adaptation of known technologies; in other cases, such as solid
state electronics, basic science provided the concept and first
proof of principle. After
that, inventiveness led to profitable devices.
The one hallmark
of an effective relationship between basic and applied research is
an easy flow of ideas.
Human inventiveness itself, as well as an environment that
encourages and rewards that inventiveness, are two factors that
make ideas flow freely—and in both directions.
isn’t at fault in the current thrust toward technological
suicide—rather, it’s the failure of industry, industrial
managers and stockholders to care about the long-term health of
their firms. The
market system fails to encourage pension funds to invest in firms
on the basis of their performance, or to make investors prefer
long-term gains over short-term returns, or to motivate an
oligarchical executive officer to care about the performance of a
firm after retirement. These
factors, in an irresponsible trading game whose players have no
obligations to the future, have transformed many large industrial
firms from self-restoring, self-strengthening enterprises into
To restore health
to industry, we need incentives favoring long-term capital gains
and high-risk, high-payoff portfolios of basic research
investment. We need
option and retirement plans that motivate business executives to
care how their firms perform years
after they have left their managerial positions.
We need an incentive system that makes shareholders
concerned about the performance and long-term sustainability of
the firms in which they invest, and discourages
"churning" of stocks.
We need to invent
a bi-directional enterprise to scan the products of basic research
and to inform basic researchers of the problems and opportunities
that challenge industrial scientists.
We would gain much from a system of allowing industrial
scientists sabbaticals for working with their counterparts in
basic research, and vice versa. And we need
incentives to exploit the innovations that will result from such a
productive interaction between basic and applied science.
We need not—nor
cannot—know in advance what innovations are possible—but we
must have reasons to make those possibilities happen.