#52 from R&D Innovator Volume 2, Number 8        August 1993

Discovery:  The Eureka! Moment Revisited
by Robert P. Crease, Ph.D.

Dr. Crease is a professor of philosophy at the State University of New York, Stony Brook.  This article is based on his upcoming book, The Play of Nature: Experimentation as Performance (Indiana University Press, Bloomington, 1993).

The nature of the discovery process is still one of the most mysterious and controversial subjects in the philosophy of science.  Some scholars seek to create a “logic of discovery” based on the formation of hypotheses.  Others, like Arthur Koestler, describe the psychological conditions that favor the sudden insights which they call “Eureka” moments.  (Archimedes, you recall, supposedly cried "Eureka!" when he suddenly realized that water displacement would reveal the true gold content of a crown.)

I believe discovery is better understood as an instance of recognition—and that the writings of ancient authors on this subject can prove surprisingly useful in helping us understand the discovery process.

The Relevance of Recognition

In his discussion of tragedy, Aristotle defined recognition as a “passage from ignorance to knowledge,” which can either come about as a complete surprise, or as the fulfillment of dawning suspicions.  Given the context, he naturally focused on recognition between persons, although he said one can also recognize objects, events and rules.  His favorite example of tragic recognition occurs in Sophocles’ Oedipus Rex, who was abandoned by his parents shortly after his birth because an oracle warned that he would kill his father and marry his mother. 

Recognition, Aristotle wrote, has three characteristics.  First, it always concerns something familiar (Oedipus discovered his true identity in the tragedy).  Second, it is the outcome of a quest in which one's knowledge grows progressively deeper; it’s not introspection but worldly interaction that prepares one for the eventual discovery (Oedipus reached his fateful recognition by relentless questioning.)  Finally, recognition has unanticipated consequences (the discovery that he was his father’s murderer and his mother's husband was so devastating that Oedipus blinded himself in despair).

Surely it seems odd to turn to Aristotle, of all people, and to his Poetics, of all places, to help understand the scientific discovery process!  But research, too, involves a passage from ignorance to knowledge.  And here, too, the passage can come as a surprise, or as the fulfillment of one’s suspicions.  Moreover, the scientific discovery process also shares the three other structures of recognition just mentioned. 

We Recognize the Familiar

How can this be, when we are speaking about novel structures of nature?  In reality, we never switch on a machine and have it eject a discovery.  By the time we discover something, our previous experiments or “putterings around” have given us a backlog of relevant experiences, so when we eventually become confident about the discovery, it is a recognition of something with which we are already familiar and with which we already have experience.

When Wilhelm Konrad Roentgen discovered X-rays in 1895, he was working alone with a cathode-ray tube and noticed a fluorescent screen on the same table commence to glow for no conceivable reason.  He turned off the switch, and the glow ceased—evidently the cause of the glow was something emitted from the cathode-ray tube; something that could penetrate paper, wood, human flesh and even thin metal; and something that did not behave like the cathode rays he had been examining.  Roentgen was so astonished that, as he told his wife, he had to repeat his procedures over and over to be sure that he was not hallucinating.  Only after eight weeks of experiments did he finally convince himself that the effect was real.

This was one of the simplest, clearest, and most momentous scientific discoveries.  Roentgen had no hypothesis about X rays before he observed them—hence his utter astonishment.  Indeed, he was not even sure what a “ray” actually was.  His hypothesis in his first paper turned out, as he suspected, to be completely wrong, yet he was confident he had discovered something.  One might imagine him slowly acquiring the confidence that he was observing a real phenomenon, and one might consider that achievement of confidence (which evidently took place over time) to be the decisive act of discovery.  What gave him the confidence was the exploratory puttering which acquainted him with the new phenomenon.  In other words, he grew to recognize something with which he was, by the end, already familiar.

Recent literature on recognition has stressed that one has to be prepared for recognition, that recognition may be dim  (incomplete, a vague feeling of knowing) or dawning (the movement from dim recognition to the explicit achievement of recognition, as happened to Röntgen), and that no rules can be written for infallibly recognizing the novel.

Ability to Recognize Depends on the Technological Context

In the quest for discovery that takes place in research, one must be technologically prepared to recognize.  Just as one cannot recognize an individual at a distance of half a mile without binoculars, so one cannot recognize scientific phenomena without technology that puts us “near enough” to it.  The better the technology, the more one is able to “see.” 

Why are inventions or discoveries sometimes made by different individuals at virtually the same time?  Because the context of the community has reached a level at which a recognition is finally possible.  Imagine a group of individuals watching a person emerge from a dense fog; while at one moment no one can recognize the person, when the person approaches just a few feet closer everyone can do so more or less simultaneously. 

Sociologist Robert Merton wrote of “unanticipated consequences of purposive social action,” something we often see in governmental policies.  Merton's observation could equally apply to scientific research:  Investigators frequently discover more than they wanted or expected.  And what is transformed by such recognition is not merely phenomena in the scientific world, but also the landscape of scientific knowledge itself.   Röntgen’s discovery made a major impact on fields such as medicine, chemistry and physics.

The Consequences of Treating Discovery as Recognition

Suppose I am right, and recognition is an important part of the research process—of what relevance is this to the researcher?

For one thing, managers might be more tolerant of “puttering around.”   This, after all, allows researchers to acquire the backlog of experiences they need to facilitate recognition of new phenomena.

Second, it underscores the significance of using the latest technology, the increasingly sensitive methods that give researchers the ability to discern novel features of the world.  This is one reason historian of science Thomas Kuhn wrote that “transformations in the established techniques of scientific practice prove even more important than the incremental knowledge provided by the discovery itself.”  In other words, a small increase in the sensitivity of a scientific measurement can lead to what Kuhn called a “paradigm shift” in scientific theory.

Finally, we again learn that the impact and significance of a new discovery or technique extends far beyond its obvious application.  This point might seem obvious:  The history of science is full of unsuspected applications of discoveries.  Still, it is a point repeatedly overlooked by those who emphasize the value of applied research over basic science.  In fact, the value and necessity of basic research may be the cardinal lesson of this clarification of the nature of the discovery process.