Following is a transcription of the original typed notes of a speech given by Daniel McGilvary Gillies. The speech was delivered in the early 1960's. He was research director at Linde, a division of Union Carbide Corp.
When I came to Niagara Falls last September I was asked to serve as a Seminar Director for the Joe Berg Foundation. I was assured that this would not involve much work, which was gratifying to hear, since I was looking forward to a very busy year, coming into a new job, building a new home, etc., etc. What I had been told turned out to be accurate. My name appeared on the announcements but no one asked me to do any work. The was fortunate and I was grateful for the situation. But, actually, my conscience began to hurt me, not seriously but perceptibly. Then, a couple of months ago, I was asked to be the principal speaker at this dinner. Now, my conscience no longer hurts and you, my friends, are constrained to compose yourselves for a time and listen to the sound of my voice.
I propose to talk briefly on the topic: "Depth Versus Breadth in our Approach to Science." The word versus implies a conflict or a dilemma. The nature of this dilemma is evident to anyone who has had any experience with scientific matters. There is built into the very fabric of Science a strong bias towards concentration, specialization, comparmentalization. (Disciplines) (Examples) The Scientist or Engineer is thus, very early in the game, influenced to pick a field and dig into it deeply and narrowly. On the other hand, there is an obvious unity to Science and obvious advantages to be
derived gained from a broad knowledge of Science. Unfortunately Science is big, life is short, and our intellectual capacity is limited. And so we are faced with a dilemma. Before trying to answer the question of what, if anything, can be done about this situation, I would like to elaborate a little on the nature of the dilemma and, especially, I wish to dwell on one important feature of it, namely, that as a problem it is increasing in severity very rapidly with time. It is, therefore, a much more serious problem for the young scientists in our audience than for us old scientists. And it is this feature of the thing wihch impelled me to select this topic for discussion before this audience.
Science is man's organized effort to understand nature and natural laws. Men pursue science because they derive intellectual satisfaction from discovering how nature works and because they have found it possible to use their knowledge of Science to control and manipulate nature. The organized effort to apply science for useful purposes is called Technology. We are constantly reminded that our whole civilization and way of life is increasingly dominated, for better or for worse, by influences stemming from Science and Technology.
When we come to grips at close quarters with Science or Technology, we very quickly encounter the necessity, mention previously, to concentrate our efforts and dig deeply and narrowly into the subject at hand. This is because, whereas natural laws are often of simple mathematical form, natural phenomena, in their detailed manifestations, are complex. (Example) It is only through penetrating these complexities of detail that we achieve the
deep satisfaction which comes with mastery of a difficult subject. And it is only through this type of mastery that we acquire the ability to control and manipulate nature in an important way. A cursory knowledge of biology does not lead to the development of a polio vaccine.
In contrast to the intellectual satisfaction which comes with mastery of a specialized subject, in all its
wonderful details, there is equally deep satisfaction to be devied from the ability to comprehend the broad relationships in the "big picture," the ability to see the forest in spite of the trees. Furthermore, the ability to transcend the more or less arbitrary boundaries between the various scientific and engineering disciplines is of ever increasing technological importance. To put a man on the moon requires contributions from essentially every discipline and skill of Science and Engineering. Finally, the history of Science records many cases where major advances in a given discipline are made possible only through the use of concepts and techiques from other disciplines.
An especially interesting recent example is to be found in the career of Georg von Békésy [diacritical marks handwritten], who received the 1961 Nobel Prize in Physiology. He received this prize for researches, covering a period of some thirty years, on the anatomy of the human ear. In fact, most of what we know about the detailed structure and operation of this amazingly complex and delicate instrument is due to his work. The interesting point is that this major contribution to the science of Anatomy was accomplished through experimentation and lines of conceptual reasoning which were derived from the disciplines of Mechanics and Electronics as much as from disiplines of Biology, and von Békésy [diacritical marks handwritten] was, in fact, educated, initially, as an electrical engineer and began his career as an employee of the Hungarian Telephone Company. This approach to science, in which one tries to bring to bear on the investigation of a specific phenomenon, techniques and concepts from any and all fields of science, is in the great tradition exemplified to Galileo, Newton, and Lord Rayleigh, and it is gratifying to see that it can still be practiced by individuals, even in this age of ever increasing specialization.
We now see the two horns of the dilemma in clear perspective. It poses many problems for all of us. I can assure you that it poses serious problems for me as a Director of Research in a large technologically oriented Corporation. It causes almost insuperable difficulties for the educator trying to formulate a reasonable scientific curriculum and for the student trying to squeeze himself into the curriculum and still leave room for a few electives. (Extension to more general premises.)
Let us now turn our thoughts for a moment to the question of how this dilemma behaves as a function of time. It seems self-evident that the severity of the problem at any given time is directly related to the quantity of scientific knowledge existing at that time. It was possible, I believe, for Aristotle to know everything of significance that was known in his world. By the time of Faraday, the approach to omniscience could not be made so closely, by a significant margin. Today, the thought is preposterous. Thus, to investigate the change in our dilemma with time, we must examine the change in the quantity of scientific knowledge with time. To do this I should like to digress for five minutes and present what might be called a "Simplified Mathematical View of Modern History."
Many terms are used to describe the times in which we live. We are told that we are living in the Modern Age, the Machine Age, the Age of Mass Production, the Age of Automation, the Automobile Age, the Jet Age, the Missile Age, the Age of Plastics and Synthetic Fibers, the Electronic Age, the Computer Age, the Atomic Age, the Nuclear Age, the Space Age. I suggest to you that one can in a very accurate and descriptive sense, encompass all of these terms and many more by saying that we live in the Exponential Age. By this I mean that most of the unique characteristics of our Age either result directly from or, at least, are closely interrelated with the fact that, between three and four hundred years ago, a number of quantities which are of great importance to human destiny began to grow at an exponential rate. For the moment, I will mention only four: the population of the world, the capacity of the world to produce manufactured goods, the number of books published, the quantity of scientific knowledge.
Before going further, let us examine [handwritten note: take a look] the elementary characteristics of exponential growth. The basic rule is familiar to all; it is, simply, growth according to the compound interest law. If I invest a hundred dollars at 5% compounded annually, at the end of a year I have my hundred dollars plus 5% of a hundred dollars or a total of a hundred and five dollars. At the end of two years I have my hundred and five dollars plus 5% of a hundred and five or a total of a hundred and ten dollars and twenty five cents. At the end of 400 years I have thirty billion dollars. This illustrates the most important feature of exponential growth: quantities which continue to grow in this way eventually get very, very large. The growth rate (interest rate) is, naturally of some importance and can, conveniently, be related to period of time required to double the quantity growing at that rate. (Illustrate Shape of Curve.)
|Growth Rate (Interest Rate), % Per Yr.||Doubling Period - Years|
According to this thesis Modern Times (Exponential Age) began about 1600 A.D. At about this time population and a number of other important quantities began to grow exponentially. World population in 1600 was about 100,000,000. Since then it has grown remarkably steadily at about 1% per year, doubling about every 75 years.
|Year, A.D.||World Population|
Backward extrapolation shows that population was not growing exponentially prior to 1600. Forward extrapolation is guess work. (Elaborate)
I will not try to give statistics going back to 1600 on growth of other quantities of interest to us. It is instructive, though, to examine more recent statistics on growth in number of scientists and number of scientific publications.
There are in the U.S. today about 1,100,000 active scientists and engineers, - 1.4% of a working population of 80,000,000. The addition rate is about 5% per year and the retirement rate about 2% per year, leaving 3% per year for growth. It is estimated that in 2000 A.D. there will be about 6,000,000 scientists and engineers constituting about 3% of a working population of 200,000,000. The world growth rate in number of scientists is estimated at about 10% per year, - doubling every 7 years.
Perhaps as good a measure as any of the rate of growth of scientific knowledge is the rate of publication of scientific papers. The number of papers abstracted in Chemical Abstracts was about 15,000 in 1910 and about 150,000 in 1960. The average growth rate is about 8% per year. I believe the overall growth rate of scientific knowledge is greater than this. Entire new disciplines emerge regularly: - Polymer Science, Aero-Thermochemistry, Cryo-Biology, Plasma Physics, etc., etc. (Illustrate with Library Bulletin.)
Let us now recapitulate our "Simplified Mathematical View of Modern History." About 1600 A.D. human population, human knowledge (at least scientific knowledge) and human manufacturing capacity, together with many related quantities began to grow exponentially. In the 100,000-odd years that man had been on earth prior to 1600, growth was laboriously slow and did not follow the exponential rate law. Speculation as to what initiated the so-called explosion is beyond the scope of this talk. However, it seems likely that growth in knowledge was the precursor and initiator, the fuse. The growth has not been steady in time or geographical distribution, when viewed in short time segments. Over the long haul it has been remarkably and inexorably steady. In the 17th and 18th centuries, the trend was not apparent to those living then because the numbers were still as yet relatively small, in an absolute sense, and the world was still relatively empty. But the seeds of the 20th century were being surely sown. We are now fairly far up the growth curve - the numbers and their consequences are getting big. It is hazardous to predict the future. It is hard to believe that this type of growth can go on very much longer. But when and how a limiting law will come into play is not easily foreseen. At the end of the last century most physicists thought that physics was an essentially closed subject. They could hardly have been further from the truth. It seems clear, though, that at least my children and my children's children will live in the Exponential Age, and that those of them who choose to be scientists or engineers will face, in increasing measure, the dilemma of depth versus breadth in their approach to Science. So let us return our attention to that quarter.
First, let me say that the last thing I want to do is to be a prophet of doom. Exponential Age or not, dilemma or no dilemma, Science and Technology provide opportunities for careers as honorable and as interesting as any. We need more and better professional scientists. And we need more understanding of science by the amateurs. With specific reference to the great dilemma, there is, of course, no pat answer. I would, however, like to make the following comments:
(1) Considering the extremes, infinitely deep and narrow versus infinitely broad and shallow, the former has finite value, the latter has little or no value. It is necessary to put down a fourdation somewhere by concentration. There is little room for the dilettante.
(2) The ideal goal for the educated man remains unchanged: to know everything about something and something about everything. Let us not give up this goal too easily. We are very prone to underestimate the capabilities of the human brain, especially the young brain. Students are generally capable of more than we think. As the sum of knowledge increases, we may have to lower our sights a little. Let us retreat slowly and grudgingly. Perhaps our goal should be: to know almost everything about something and somthing about almost everything.
(3) It has always been true that a formal education can only prepare a man to educate himself. In an exponential age this truism becomes increasingly true. The scientist or engineer who stops studying when he graduates is doomed to failure. A corollary of this proposition is that the educator should strive for a curriculum which tries to teach the student how to think and to learn, and which emphasizes the fundamental building blocks of science and engineering rather that a high content of factual knowledge. I advise the student to concentrate first on the fundamentals of chemistry, physics, biology, and mathematics. Especially mathematics, this is the universal tool and language of Science and facility in mathematics is probably the most important single skill that a scientist or engineer can have.
(4) It seems to me that the old approach of trying to get a liberal education first followed by specialization is, where possible, to be preferred. And in general, I recommend an education which is as liberal as can be managed. However, increasing numbers of scientists and engineers are forced to specialize early in order to enable them to begin earning a living in a professional capacity at a reasonably early age. This is not necessarily fatal. If one does this, one should recognize the need for trying, later, to broaden one's technical base. Conversely, the person who gets a liberal education must realize the need for later concentration and specialization.
(5) Two approaches toward broadening one's base of operations can be cultivated:
|(a)Set aside time for reading outside your field|
(b)When working on a problem be willing to learn new disciplines in order to solve it.
(Open Mind - Humility.)
(6) It is my belief that the need for breadth is actually greater in a career associated with Technology than in one associated with pure Science. This is particularly true in Technological (Industrial) Research. This in contrary to the generally accepted opinion.
[The following point was hand-written]
(7) Mechanical aids. Speed reading. Data retrieval. Computers.