The Tech - Online EditionMIT's oldest and largest
newspaper & the first
newspaper published
on the web
Boston Weather: 45.0°F | Light Rain

Text of NSF Director Massey's address

(The following is a transcript of the speech by National Science Foundation Director Walter E. Massey to the graduates and guests at Commencement on Monday, June 3, as recorded by the MIT News Office.)

Good morning. I am pleased and honored to have been asked to deliver the commencement address here at MIT today. People sometimes say that a graduate's greatest achievement is getting through the commencement exercises. That is clearly not the case here today. Each of you has proven already that you have the intelligence and wherewithal to master one of the most rigorous academic programs in the nation.

I recently stumbled across an example of that rigor. When I expressed surprise that grade point averages at MIT are carried to a third decimal place, I was informed that only MIT and God are able to distinguish excellence with such precision.

When you leave here, you will be joining an elite corps of graduates from the nation's premiere research universities. For decades, MIT has been a leading site for advances in basic understanding and innovation in science and engineering, advances that are the foundation for continuing prosperity and an improved quality of life. As such, MIT is a precious resource and valued institution to more than its students, graduates and those directly involved in its operation. MIT and the other top US research universities play a critical role in setting and sustaining the highest standards of achievement in research and technology.

As MIT graduates, you have assumed a duty to uphold this tradition of excellence in all your pursuits. What does this mean? How do you go about it?

Excellence is a quality that is recognized by comparison. While it is judged generally from without, excellence begins within. Individuals achieve excellence through the choices and decisions they make regarding the conduct of their lives.

This morning I would like to talk about a specific arena of excellence, that is, the area of basic research in science and engineering. Many people think of research as a cut-and-dried process. And it is true that there are some clear "rules of the game." The object of research is, to paraphrase my good friend Nobel laureate Leon Cooper, discovering how the world works -- separating the truth about the way things are from conceptions of the way they might be. To accomplish this task, good researchers are skeptical; they evaluate claims empirically and logically, not on the basis of authority. Good researchers are open, sharing their hypotheses, methodologies and results, and making their primary data available to others. They do this so that results can be reproduced and findings confirmed. In this way, the research community protects its interest in the truth.

Good science and engineering research is the uncompromising pursuit of truth. As such, it represents the highest achievement of human intelligence and provides a constant source of enrichment to mankind's existence intellectually, spiritually, and materially. "In science there can be no perfect crime, no permanently unsolved murder," as chemist Carl Djerassi notes in his novel about prize-winning research, Cantor's Dilemma. If a finding is important, sooner or later the experiment will be repeated and the results subjected to independent verification. The rules of research keep science and engineering truthful.

Simply put, the whole edifice of science and engineering research is built upon honesty. More than in any other endeavor, individuals conducting fundamental research depend upon the veracity of the accumulated insights and accomplishments of others. Sir Isaac Newton expressed it best when he said, "If I have seen further, it is by standing upon the shoulders of giants."

But not even giants can see clearly if their feet are on shaky ground. Few things are more damaging to the research enterprise than falsehoods -- be they the result of error, self-deception, sloppiness and haste, or, in the worst case, dishonesty.

It is the paradox of research that reliance on truth is both the source of modern science and engineering's enduring resilience and its intrinsic fragility.

So, where do young researchers first learn the rules of the game that protect the integrity of their endeavors?

For generations, the community has relied on the unique mentor-apprentice relationship that develops during the process of doctoral and postdoctoral research to teach these important lessons. When all goes well, a bond of trust develops between the professor and his or her students which is grounded in intellectual curiosity, a desire to discover new knowledge, and a common commitment to truth. What is conveyed and what is learned is more than simply a body of facts; it is an approach to understanding, an appreciation of standards, and a set of values -- in effect, an ethic.

I was very fortunate in my own graduate training to have had a research director from whom I learned a great deal of physics and much more besides. Professor Eugene Feenberg (now deceased) was a prominent physicist, well-respected among his peers, a man of integrity and unyielding honesty. All of us who were his students saw these traits in him and profited from working with him.

To come to class ill-prepared or not take time to listen to his students and their problems; to publish a result prematurely simply to gain priority; or to put his name on a paper by one of his students when he had not shared in the work himself -- all these things were anathema to Gene Feenberg. Those of us who were his students realize how lucky we are to have had an almost ideal mentor-apprentice relationship.

Most mentor-student relationships, however, are not ideal. Too often, important lessons in scientific integrity remain unlearned. Fundamental changes in the way research is conducted in many fields, not just physics, increase this possibility.

We have moved away from small tightly knit research communities, in which misconduct is easily observed, toward large anonymous research enterprises, in which tasks are fragmented and accountability is hard to ascribe. This change imperils traditional scientific ethics.

Consider the experimental team that did the work resulting in the 1984 Nobel Prize in physics. It consisted of over 150 people. What is it like to be a graduate student or a postdoc in a group of 150? Exciting and important definitely, but unlikely to provide the same opportunities for personal interactions with a mentor that one has in a smaller group. In such circumstances, some important values may go unlearned, and a key safeguard against scientific misconduct left undeveloped.

Experimental replication, another major safeguard against research improprieties, is also more difficult. The size, complexity, and cost of modern research projects make it less likely that they will be reproduced -- even in fields where it is possible. Replication of experiments always has been more of a problem in the social, biological, and clinical sciences where there are often too many uncontrollable variables.

In addition, the vast expansion in the amount of research carried out has been accompanied by an explosion in scientific publications. As a result, too much of what is published goes unchallenged.

I recently heard of a Czech article published in 1887 entitled "O Uplavici," [pronounced u pla vee chee] literally, "On Dysentery." When the title was translated by an abstracter it became the name of the author, O. Uplavici. For many years it was cited as such.

This was just a thoughtless error, one that probably had little bad result. But indifference to errors can be very damaging. Errors lead researchers down blind alleys; published errors, if not corrected, can set back an entire research community.

Scrupulous attention is especially important now. Growing competition for funds, tenure and acclaim; increasing chances of financial conflicts of interest among researchers; even the scope of intellectual ferment -- with disciplinary boundaries breaking down, and new ideas and techniques challenging traditional paradigms -- all these conditions make science and engineering more vulnerable to falsehoods.

Under such circumstances, it is essential that the community of researchers -- as individuals and through the institutions that represent them -- uphold the highest standards of integrity. Scientists and engineers are the only people who can redress misconduct. Universities, as the primary locus of basic research, have a special responsibility in this regard. In cases of alleged wrongdoing, they have a duty to analyze the facts fairly, determine accountability, protect the rights of all involved, and see that any falsehood is corrected.

Like Caesar's wife, universities must be above reproach in all conduct relating to research -- be it ensuring scientific integrity or allocating indirect costs. As the costs of academic research increase, so too does the universities' dependence upon federal support. As the federal investment grows, so too does the public's scrutiny of the research universities. Misconduct of any sort imperils public sponsorship of research.

It is a high tribute to the integrity of our research universities that all who have investigated the incidence of scientific misconduct find very few cases of bona fide fraud. There are, however, more common errors in fact and in judgment that also have damaging effects.

Errors are inevitable because researchers live in a world of uncertainty. How, then, can the young scientist or engineer prepare for such a life today, with neither the certain guidance of a close mentor nor a firm sense of the "rules of the game?"

There are no easy "correct" answers to this question. Careful attention to treatment of data, choice of methods, and evaluation of one's hypotheses can help guard against the most common pitfalls -- mistakes and self-deception.

Publication in a peer-reviewed journal provides another important safeguard on the integrity of research. Critical review by peers can detect errors and omissions invisible to the untutored eye. For this reason, it is the most acceptable means for disseminating research results.

There are also some "incorrect" behaviors to be avoided. Deliberately bypassing the peer-review process is one. Such action can short-circuit the self-correcting mechanisms of science and engineering and also damage public trust. Researchers who release their findings directly to the public risk adverse reactions later if their results are shown to be mistaken or misinterpreted by the media.

I doubt that "cold fusion" research benefited from the fact that much of its scientific review was carried out in the mass media. In the same vein, the credibility of the entire university research enterprise has been jeopardized by the action of individual schools attempting to bypass peer review to obtain earmarked federal funds for their research facilities.

The practice of honorary authorship also deserves scrutiny. Renaissance painters trained their apprentices by allowing them to work on canvases to which the master then signed his name. This tradition gave way, over time, to fairer recognition of an individual artist's contributions. In some fields of science and engineering, it is traditional to place the senior researcher's name on all work done by the group. Not every tradition is good. Honorary authorship diffuses accountability and can lead to irresponsible research.

Failure to bring wrongdoing to the attention of those responsible for the research is also incorrect behavior. Any assault on the integrity of science and engineering damages all researchers. Misconduct erodes research standards and norms, and leaves a bad impression of science and engineering in the minds of the public. Researchers have a professional and ethical duty to protect the integrity of the enterprise and sanction misconduct promptly.

The conduct of research is a complex and demanding task. Why, then, pursue it? The answer is straightforward. The possibility of observing or understanding what no one has ever observed or understood before can be irresistible.

Ever since Archimedes streaked through the streets of pre-Christian Syracuse shouting "eureka," scientists and engineers have found in the moments of discovery or innovation one of the most exhilarating experiences in their lives. Even the process of research itself can be deeply satisfying -- putting the pieces of a puzzle in place and making sense out of a mystery.

Research results in knowledge that is as certain and reliable as anything we know. Science and technology are among humanity's greatest intellectual achievements, having transformed not only the material condition of our lives but also the way we see the world. A career in research offers an opportunity to join the pantheon of scientists and engineers who have changed the condition of life on earth and brought the universe to our doorstep.

T. S. Eliot once said: "We shall not cease from exploration/ And the end of our exploring/ Will be to arrive where we started/ And know the place for the first time."

We have now arrived where we started. You have survived the commencement exercises -- almost. Soon, you will be certified well-educated. And each of you will become standard-bearer for one of the greatest institutions of higher learning in the world. I know you will carry out this responsibility to MIT wisely, and with integrity, in whatever career you choose.

I am glad to have had the opportunity to share your commencement, and I thank you.