The Crisis of the Engineering Algorothm

by Kenneth Keniston

Massachusetts Institute of Technology

Talk for Institute for Advanced Studies in the Humanities

(Politecnico of Turin/Institute for Scientific Interchange)

October 17, 1996

(see also the Italian translation)

Algorithm: “Any special method for solving a certain kind of problem”.

I am honored by the invitation to address you on the subject of the humanities, the sciences, and their relationship to engineering education. The entire world admires the extraordinary accomplishments of northern Italy in the last decades, accomplishments importantly centered in this city, and in this famous school of engineering. It is a great honor to be asked to join in your discussion about how to adapt your educational programs to modern conditions.

Your invitation is also a source of discomfort, however, for I am not familiar with Italian engineering education, so I cannot comment specifically about Your situation. Having studied engineering education in the United States, France, and Great Britain, I am certain that in Italy as in every other country, national culture and local institutions also determine -and should determine- much of what happens in a school of engineering. I therefore warn you in advance that my comments will largely be based on my experience in other countries, and may not be relevant to the situation in Turin and in Italy.

The Crisis of Engineering Education

I want to speak very generally, and to argue that in America, as to a certain degree in France and Great Britain, we are witnessing a crisis in engineering and in engineering education. This crisis is superficially manifest in the schools of engineering by questioning of the traditional curriculum, by complains that engineers do not occupy sufficiently important positions in industry and society, by fears that change will mean the destruction of the traditional virtues of engineering. In France, schools like the Ecole Polytechnique once had primacy not only in technological innovation but, perhaps even more, in the public and private industrial sectors. Today, they must make place for the National School of Administration (ENA), and the grandes écoles of management, in particular HEC and ESSEC. Officials of the Ecole Polytechnique are explicitly searching for new ways to redefine the mission of the schools, so as to reassert and broaden their mission and influence.

In America, especially at a place like MIT, similar curricular questioning is widespread. On the one hand, MIT probably attracts the most scientifically talented group of students of any university in the United State. But on the other hand, it is noted even in an age when high technology becomes ever more central, the leadership of American industry is in the hands not of the graduates of MIT or of other technically-trained individuals, but rather of financiers and lawyers -graduates of the Harvard Business School and the Stanford Law School. MIT graduates, if they attain managerial positions at all, most often do so as technical vice presidents, not as the leaders of the enterprise.

The question inevitably arises, then, whether or how an MIT education contributes to the “failure” of MIT graduates to attain positions of the highest leadership.

Another symptom of the unrest vis-a-vis engineering education is the constant but inconclusive discussion of curricular reform that goes on today in American schools of engineering. The air is filled with proposals and counter proposals for reform. These proposals are usually based on two fears. First is the fear that, given the enormous explosion of knowledge, the present four-year curriculum is too brief to allow students to reach the frontiers of knowledge. Second is the fear that something about the content of engineering education disqualifies (or fails to qualify) students intellectually for positions of industrial, social, political, -perhaps even intellectual- leadership.

Behind these fears lie broad historical changes that have affected public attitudes toward engineering, engineering education, and indeed the entire enterprise of modern technology. Of these, the most critical historically is the growing and by now universal awareness that technological innovations can have -often do have- negative effects. Starting with Hiroshima, and continuing with increasing momentum since, critics both outside and inside the scientific and technical community have become aware that humankind for the first time possesses the technological capacity to alter and even to destroy the entire world. The list of current problems and dangers is long and familiar: nuclear warfare, nuclear winter, acid rain, ozone depletion, the destruction of species, toxic chemicals, depleted earth, polluted air, unsafe water, the proliferation of hazardous wastes, global warming, and so on. All these perils are attributed to the unintended effects of the vast extension of human powers made possible by modern technologies.

Such fears, widely diffused in the intellectual community, in the general public, and among politicians in America, have helped to undermine the 19th century American “heroic” view of the engineer as he who conquered space and mastered nature.

It is clear that from Hiroshima to Chernobyl there has been a marked decline in public enthusiasm for technology, and consequently, for the engineers who are the principal embodiments of the best in technology. I obviously do not mean that most students and faculty at places like the Politecnico, MIT, the Technion in Israel, the Technicke Hochschule, the Ecole Polytechnique, or Imperial College are themselves filled with doubts and anxieties. But I do mean that the public support for engineering -the cultural image of engineering- has changed because of awareness that powerful modern technologies almost always have, in addition to their good and intended consequences, unintended and undesirable consequences.

But in the last analysis, the fact that technologies have a “down side” is, for creative engineers, a practical problem rather than an intellectual or conceptual problem. As one president of MIT put it, “The answer to bad technologies is not no technologies, but good technologies”. For example, if present methods of energy production contribute to global warming, we need to discover, through science and technology, improved techniques that will not contribute (or not contribute so much) to global warming. Doing so is not a philosophical problem, but rather an interesting technical problem, a problem which should yield, given hard work and adequate resources, to the efforts of creative engineers armed with the weapons of science, mathematics, and cost-benefit analysis.

However important is the problem of declining public faith in technology, I believe there is a second problem at the heart of the crisis of engineering education, a problem that is deeper, more intractable, more conceptual -indeed a philosophical problem. It strikes at the fundamental intellectual assumptions upon which modern engineering is based, and it threatens to undermine the conceptual foundations on which this immensely creative and transformative enterprise was originally founded.

I will call this second problem the erosion of faith in the adequacy of the engineering algorithm -of the paradigm, the method of solving problems that lies at the heart of modern engineering and thus of modern technology. This crisis of the engineering algorithm -vigorously denied by some, ignored by most, and confronted by a few- amounts to a fundamental conceptual crisis in engineering.

An Historical Excursion

To explain this crisis, I must ask you to permit me a superficial history of engineering in America that will stress the originality and productiveness of the algorithm behind modern engineering.

It is now widely accepted that what historians call the “First Industrial Revolution” -that revolution in the production of goods that began in the 17th and 18th centuries and that took off in the 19th with the extraordinary development of textiles, transportation, metallurgy and other industries- was not the creation of science-based engineering, but rather of inspired amateurs, gifted craftsmen, and inventive dabblers. No doubt the concurrent development of science, and especially of scientific rationalism and the idea of progress, created a fertile climate in which the first industrial revolution could take place. But the first canals, factories, railroads, steel mills, spinning jennies, and machine-driven looms were not the result of the systematic application by engineers of scientific principles. Rather they resulted from the inventiveness and imagination of men whose training in science was usually rudimentary. As the historian of technology Elting Morison makes clear in his account of the first American canals, they were built not by the equivalents of modern day hydraulic engineers, grounded in basic science, but by inspired dabblers, ingenious craftsmen, and hard-driving entrepreneurs. Most of the canals leaked horribly; one 19th century canal was so badly surveyed that, when the two ends of the canal being constructed finally met, one end was two meters higher than the other!

Basic science, when it existed, did not exist in the minds of engineers, but somewhere else -mostly in urban academies and a few universities, largely a pastime of the aristocratic and the rich.

It was only toward the end of the 19th century, when the momentum of the First Industrial Revolution had become irresistible, that the idea of the heart of modern engineering gradually took hold, and the Second Industrial Revolution began. It was in the 1870s, ‘80s, and ‘90s that industrialists in search of new products and of cheaper ways of making old ones, first began to realize that the new sciences could be used -directly applied- to solve technological problems, to invent new products, to improve old ones, to multiply materials, and to lessen the costs of production. First in the German chemical and dye industry, in the nascent electrical industry in Great Britain, Italy, and the United States, in the fields of communications and metallurgy, a strange new professional, the precursor of the modern engineer, began to appear.

This new engineer, although he was trained in the science of his day, did not share the scientist’s primary goal of extending knowledge of the natural world. Rather, he aimed to use science to create new products and processes. His employer was not the university but his clients and later the industrial firm; his assignments were dictated not by his own curiosity but by the practical needs of his employers. The term “engineer”, once applied diffusely to gifted mechanics, drivers of machine engines, and educated soldiers from West Point or the Ecole Polytechnique, began to acquire a new specificity.

Increasingly, “engineers” were members of a new profession -those who had studied the basic mathematical and scientific principles essential to the applied and practical sciences, men (for this was an entirely masculine profession) who specialized in practical utility, and whose work as bridge and road builders, metallurgists, chemical engineers, electrical engineers, or mechanical engineers was premised on a grounding in the best and most useful science available at the time. With the creation of this new profession, and premised on the radical idea that underlay it, the momentum of the First Industrial Revolution fed that of a Second Industrial Revolution whose impetus has driven the transformation of the known world until our own times.

Behind the creation of the new professional engineer lay an idea -an idea of immense importance, radical simplicity and unprecedented productivity. It was an idea whose power had been anticipated by scientific optimists for centuries, but that had never before implemented on a large scale. It was the idea that the basic principles of science, until then a branch of philosophy aimed largely at understanding the natural world for its own sake, could be systematically and deliberately applied to transform the natural world to achieve human and industrial purposes.

This idea took practical shape only gradually, and went hand in hand with the astonishing advances of science in the 19th and 20th centuries. At first, the curricula of the new or revitalized engineering schools had a large component of learning by doing -imitating the best practice of those who had gone before. But science and mathematics began to play a newly important role, and in the 20th century, especially in the period between the two World Wars, acquired the centrality in the education of engineers as the “fundamental knowledge” on which the rest was built. It was then, at places like MIT, that a major, important, and autonomous School of Science was created to provide a fertile culture, a fundamental core, of an institution dedicated to the education of engineers. For in the 20th century, the interdependence of science and technology became increasingly clear: a new synergy arose between the engineer who created new instruments of measurement and discovery, and the scientist whose discoveries fueled further advances in engineering.

This transformation of engineering and, as a direct result, of the entire world, can be characterized in many ways. It was the product of powerful industrial, intellectual, economic and political forces working largely in concert; it produced new professions and vast redistributions of economic, intellectual and political power; it created new understandings of the world. It generated a new human type and a new culture hero, the pioneering, innovating engineer-entrepreneur whose efforts spanned continents, defied time and space, and brought the luxuries of the wealthy to the reach of Everyman.

Here, however, I want to examine the intellectual foundations of this transformation. As I noted, the idea that science could help man master nature for his purposes long antedates the engineering revolution. But it took the engineering revolution to demonstrate that it could be done, and to create the institutions of learning and practice wherein it occurred.

The Engineering Algorithm

The core idea behind the engineering revolution is that the relevant world can be defined as a set of problems, each of which can be solved through the application of scientific theorems and mathematical principles. Around this first principle are clustered a series of corollary ideas.

1. This first principle implies a metaphysical division of the world into two realms. The first is the realm of “problems” that can be “solved”. But of course, we all know that not every difficulty in human life is a “problem” as defined above. Thus, there is a second realm -variously called “the rest of life”, “values”, or “society”- that does not lend itself to a statement in “problem” terms and is therefore not relevant to the engineer qua engineer.

2. As for the “problems” worthy of engineering work, they are usually complex in nature. This means that they must be broken down -or analyzed- into component and simpler subproblems, each of which can be solved separately by the application of scientific principles and mathematical ideas. By correctly solving all of the component subproblems, and then by integrating the subsolutions, the engineer reaches the solution to the larger, complex problem.

3. The application of appropriate scientific principles to a genuine problem results in a correct answer. Solutions can be therefore classified as either correct or erroneous. Incorrect solutions can stem from a variety of errors: e.g. introducing the wrong values for the variables in an equation; inadequate knowledge of the scientific formulas required; sloppy calculations; application of the wrong scientific principle to the problem. Learning to avoid such errors requires much practice, great care, and extreme meticulousness.

Especially difficult is learning which scientific principle to apply to the solution of a problem.

4. Since the language of science is largely a quantitative language, genuine problems must be susceptible of translation into and solution in quantitative terms. Engineering education is necessarily grounded in mathematics. Questions or difficulties which cannot be quantified are therefore by definition not genuine problems. Qualities that cannot be measured (e.g. beauty, social justice, grace, peace, elegance) are to be excluded from engineering calculations, even though they may be desirable on other grounds.

5. Not everyone can become a good engineer. The necessary human qualities include mathematical and scientific ability, capacity to break large, complex problems into a set of small, simple problems, concern for accuracy, and ability to identify the right scientific principles to apply to each problem.

6. Given these human abilities, however, engineering can be taught. It is taught by first immersing students in the basic principles of mathematics and science, and then by teaching them the specific problem-solving principles of a particular field of engineering. The most difficult task in engineering education is teaching students how to select the correct scientific principle to apply to any particular problem. That task is achieved through constant practice. The student is given a problem -to begin with, a simple one- and is asked to choose from among all scientific principles the particular one needed to produce a correct solution. In English, this method is called the method of “problem sets” and it is the dominant mode of instruction in engineering education in the United States.

7. Finally, it is universally admitted that engineers will, in the course of their lives, run into situations which are important but are not “problems” as defined above. But there is little that formal engineering education can or should do to prepare students for such situations, apart, perhaps, from indicating that they will exist. Most often mentioned in engineering schools are what are called “social constraints”, i.e. psychological, political, ethical environmental, economic, cultural, organizational, and other factors that may “constrain” the engineer -limiting his capacity to implement the (correct) solution which his training has enable him to find. By implication, too, in a more perfect world, where such constraints did not exist, engineering solutions could prevail. But engineers qua engineers have no special aptitude or training in dealing with these “constrains”, and should, insofar as possible, steer clear of them.

This engineering algorithm -this idea of deliberately applying science to solve certain practical problems, and the correlated methodology that tells the engineer how to do so- has proven one of the world’s most revolutionary and creative ideas. Embodied in the pedagogy of engineering schools and in the practice of modern engineering, it has been a driving force behind the progressive growth in the dominion over nature by humankind, transforming life on this planet in the last century more than in all of previous human history.

The Good Old Days

In describing the engineering algorithm, I said that it defines as irrelevant, as therefore outside the education of engineers, all of those other issues, dilemmas and situations that are not “problems” so defined. This assumption is of course deliberately limiting, and to those who operate primarily in the messy and complex world of ambiguous human feelings, organizational confusions, political conflicts, ironic contradictions, and unsolvable dilemmas, it may seem unreasonable. But until the last generation or so, the algorithm was a workable simplification within which engineers were able, with minimum tension and self-deception, to practice their vocations in a creative way. Given that what I have termed “the rest of the world” has always existed for the engineer, it may be useful to consider how the engineer of the past dealt with that fact.

To begin with, earlier engineers benefitted from an honorable, and at times even heroic, position in the social firmament. In America, engineers were those who planned the canals and designed the railroads that united the continent, who built bridges, power stations, steel mills and factories, who later designed automobiles, skyscrapers and airplanes, who brought Better Living through Modern Chemistry and Conquered Nature. Although American engineers never had as high a social status as engineers in France, to say nothing of the Soviet Union (where until Gorbachev they constituted more than 90% of the Politburo) their position in the popular imagination and public esteem was secure. Until recently, only a few dissident intellectuals -Carlisle in England, Thoreau in America, romantic poets on the Continent- questioned the net worth of Technology or of engineers as its agents.

A heroic image in the public imagination obviously helps one to live with the contradictions of one’s profession. But even more important was the fact that, in their practical work, engineers were usually able simply to ignore “the rest of the world”. Dealing with “constraints”, however omnipresent they may have been, was defined as someone else’s job. Bankers, speculators, and venture capitalists found the money. The marketing people worked on selling the product. Advertisers and public relations people dealt with public opinion, which was largely positive in any case. If there was pollution downwind, that was price of Progress, and local politicians, ever sensitive to the need for jobs, were unlikely to jeopardize local employment by threatening production. As for the contemporary problem of tradeoffs between incommensurable factors, it rarely confronted the engineer head on. His task was to design something that did the job it was supposed to do and that could be fixed if it broke. And in any case, simple rules of thumb usually sufficed to deal with what today would require a complex cost-benefit analysis: engineers tried not to use expensive materials; they tried to design things that would last a reasonable amount of time.

Furthermore, large numbers of American engineers a century ago were either self employed or worked for small enterprise. As a rule, their assignments could be completed with minimal advice and assistance from others: or if there were others involved, they were either clients or employees who did the detailed drafting of the engineer’s designs or got the measurements he needed. To be sure, a small but growing number of engineers began to be employed by large firms in what were later to become the R and D branches of these firms. But their numbers were initially small, and their experience not yet typical. The illusion of working in isolation from “social factors” was therefore easier to maintain.

But starting two or three generations ago, and with accelerating rapidity ever since, the life of engineers began to change. First, as noted earlier, the public image of engineers and of “Technology” has suffered in recent decades. Once a creative conquistador of Nature, the engineer today is more often seen as a cultural illiterate. Even more important, technological innovation, once defined simply as a road to a better life, is now seen as a major cause of environmental degradation so extreme that, in the eyes of some, it marks the “death of Nature”. Even at a place like MIT, where technological enthusiasm still largely prevails, these changes are felt.

At the same time, the working life of engineers has become more troubled as the products they design have become increasingly complex. The simple bridge becomes part of an Interstate Highway System; the box camera is transformed into the auto-everything, self-focusing, red-eye-reducing, vibration-minimizing, etc., camera with four internal computer chips; the dynamo becomes a component in a nuclear reactor complex linked to an international power grid; the Wright brothers’ airplane, a glorified bicycle, becomes the Boeing 747 with a million interconnected parts. As a result, the solo engineer who designed or invented a single product from beginning to end has become a rarity; he is replaced by the interacting, coordinated team of engineering specialists working on a complex design of a component of a complex socio-technical system. “Society”, once something out there, has entered the workplace; indeed it is the workplace.

As a result, another change has occurred. As technological systems became more complex and their components more closely linked, the question of tradeoffs between incommensurable factors, once at the periphery of engineering, has moved towards the center. In the design of a modern airplane, the designer considers safety versus speed versus reliability versus cost versus capacity, along with playback, market acceptability and the plans of foreign competitors. To be sure, any one of these factors, standing alone, might be transformed into a “problem” solvable by the engineering algorithm. But nothing in that algorithm enables the engineer to balance desiderata that cannot be measured against each other on a single yardstick. Even cost-risk-benefit analysis, an effort to extend the engineering algorithm to complex tradeoffs by monetizing qualitative variables, usually collapses when it comes to maximizing simultaneously both apples and oranges.

At the same time, all those “constraints” that the engineer could once conveniently disregard have now reentered his work. Environmental problems are classic examples of what economists call “externalities” and of what engineers in the past had rarely to consider. But when safety and emissions standards determined in Rome or Washington are imposed upon automobile manufacturers in Turin or Detroit, safety and pollution become the automotive engineer’s “problems”. Finding the best location for an electric power plant used to be something for financiers, real estate agents, and politicians to worry about; and in any event, early electric power plants were sensibly located in the center of big cities, near the people who needed the power. Today, however, for the nuclear engineer, siting is part of his problem. By some standards, for example, downtown Genoa or Manhattan would make excellent sites for nuclear reactors, with ample water for cooling, good transportation facilities, and proximity to users. But it would be absurd today to consider the potential dangers or environmental impacts of nuclear reactors merely “constraints” that by implication prevent engineers from doing what is correct. Today, what used to be dealt with as externalities have become “internalities”; what used to be seen as “constraints” have become an integral part of engineering design. Yet precisely because, by definition, these constraints cannot be reduced to problems to be solved by the application of science, the engineering algorithm cannot deal with them.

In the past, too, the impact of the engineer’s work used to be local: the bridge, the power station, the steam engine or the biplane could be designed with attention, if any, only to the sensitivities of those in the immediate, local, environment. If there were undesirable side effects (a bridge that destroyed a view, a noisy airport near a residential neighborhood, a mill outflow that polluted a river, smoke that killed downwind vegetation), those who were injured lived within a limited area. They could therefore be placated, bought off, or yielded to in strictly local terms. Today, in contrast, technological systems and as a result the work of engineers have increasingly global effects. The Japanese fear another Chernobyl in Russia or the Ukraine both because of radioactive fallout over Japan and because of political fallout: another major nuclear accident in Russia could jeopardize Japan’s nuclear plans by further antagonizing already sensitized Japanese public opinion. The burning of coal in Ruhr power plants destroys trees and lakes in Czechoslovakia. Reducing the Amazonian rain forest may lessen the Earth’s capacity to absorb greenhouse gases. Commitment to coal-burning power plants in China over the next generation could increase global atmospheric warming. Tests of nuclear weapons in the South Pacific and Siberia increased levels of strontium 90 and radioactive iodine in milk throughout the world.

If I have drawn an exaggerated contrast between the engineer of earlier days and the engineer of today, it is to make a simple point. The point is that there has been a shift in the work of engineers away from simplicity to complexity, from solo work to teamwork, from single devices to complex systems, from ignoring tradeoffs (or dealing with them through rules of thumb) to comparing them directly, from letting someone else deal with externalities to having it intrude mercilessly into the center of one’s work life. Increasingly, Marconi and Edison have been supplanted by the large anonymous design teams at Microsoft, each of which is developing only one part of the next version of Windows NT. What is most important about these transformation is that they strike at the heart of the engineering algorithm. Still useful for many tasks, still at the intellectual heart of the engineering curriculum, that algorithm is explicitly irrelevant to these new realities of the engineers’ work. That fact constitutes the intellectual core of the current crisis in engineering.

Responses and Possible Solution

Stated in quasi-Marxist terms, I have argued that we now have a contradiction between the base and the superstructure in engineering. The base is the actual material conditions of work and the productive life of the great majority of contemporary engineers - of your students at the Politecnico and mine at MIT. The superstructure is the intellectual training that they bring to their work, a training that has its roots in late 19th century and early 20th century practice, but that no longer is adequate to the realities of present-day engineering practice in the advanced industrial societies.

We need to ask two questions of this contradiction. First, how do engineers who teach engineers - the faculties of engineering schools - respond to this problem?

Second, are there solutions that would adequately confront the new realities of the practice of engineering in modern industrial or post-industrial societies?

First, what are the responses of engineering faculty to the crisis. At MIT, they can be simplified into three categories: I will call them the Good Old Days, the Business As Usual, and the Policy responses.

The Good Old Days reaction is precisely that, a reaction against what its proponents view as undesirable changes in the world that threaten to undercut the integrity of engineering and the excellence of engineering schools. At MIT, faculty members in this camp worry about the erosion of standards, downgrade the importance of work outside of science and engineering, and argue that if MIT is to avoid becoming a “second rate Harvard”, it must renew its commitment to science and technology, leaving “values”, “humanities”, and “society” for MIT students to deal with at some later stage of life, if at all. Politically associated with a conservative or technocratic position, these faculty and alumni are also agreed in criticizing “sensation-seeking journalists” and “self-interested environmentalists” for “giving Technology a bad name”.

At MIT, the second response, which is by far the most common, is “Business-As-Usual”. For faculty as well as for students, MIT is an exceedingly demanding institution. Engineering faculty are expected to do original research, to teach undergraduates and graduate students, to direct laboratories, and often to raise a part of their salaries through government or industrial research contracts. In addition, they have lives, wives, children, and hobbies, act as consultants to industry and government, and must sleep at night. The result is that only the most pressing academic crises can receive attention. The MIT curriculum is more or less set, and few faculty members spend much time on the committees that debate curriculum reform. Few faculty have the time or energy (even if they had the interest) to worry about the shape of engineering education. For most, the “crisis in engineering education” is experienced (if at all) as something that worries administrators and researchers past their prime.

The third reaction involves an awareness that something is wrong whit engineering education. The precise nature of what is identified as wrong differs from person to person. Some, especially administrators, argue that the extremely able graduates of an institution like MIT too rarely take their places as captains of industry and leaders of the nation, and too often end up working in subordinate positions for the graduates of the Yale Law School and the Harvard Business School. With some justice, they note that these other graduates are educated to deal with the complexities of the “rest of the world” in a way that MIT students are not, and they call for educational reforms that equip MIT students somehow to deal with the “real world”.

Others, often from fields like civil engineering where dealing with social “constraints” has long been recognized as part of the engineers’ work, focus on “policy” as the antidote to an excessively narrow engineering education. But proponents of “policy” sometimes tend to define policy as an extension of the engineering algorithm into the socioeconomic realms: e.g., acquiring tools and principles that will enable one to “solve socio-technic problems” like productivity, energy, global atmospheric change, or a decaying infrastructure. Training for policy is then seen as requiring a kind of technocratic education in such fields as operations research, cost-benefit analysis, risk management, computer simulation, or management science.

“Broadening” the Curriculum

Despite this variety of responses, almost everyone involved in engineering education admits, at least in principle, that it would be desirable for engineering students to have a broad knowledge of the extra-scientific world -precisely the kind of knowledge taught by humanists and by social scientists. But before we simply advocate adding this to the curriculum, we must confront the single most important fact of engineering education today, namely the explosion of scientific knowledge and the problem of how to teach students in a few brief years all they need to know in order to reach the frontiers of contemporary engineering- and to stay there.

One example from MIT illustrates this problem. Several years ago there was an important educational initiative intended to broaden students’ understanding of what were called the “contexts” of engineering and scientific practice: new elective courses were developed on the history and sociology of engineering and science, on the relationship of technological development to industrial development, and so on. University-wide committees were formed to reexamine the required curriculum of MIT. But the “context” courses were not mandatory and, in the end, very few students had the time to take them. Eventually, the committee on the curriculum recommended -not the introduction of courses on “context”- but a new requirement that all students must take an advanced course in molecular biology!

This point is probably obvious to anyone who teaches engineering education. It is manifest in the fact that the major change underway in MIT engineering education at the moment is the gradual introduction of a new longer, five-year degree called the Master of Engineering, which, it seems, will be selected by most of our most gifted students. This degree is not a research degree, but it will give the departments an additional year to introduce students to advanced concepts and methods in fields like electrical engineering, chemical engineering, and computer science. The most intense educational pressure (or at least the pressure to which my engineering school is responding) is thus the pressure to remain at the forefront by lengthening traditional engineering components of engineering education.

But this response, however understandable , does not address the crisis of the engineering algorithm that I have tried to outline. On the contrary, it conflicts with steps to “broaden” engineering education so as to enable students to deal more adequately with the new realities of engineering as a profession and practice. What can one imagine as possible solutions?

One ideal solution is what, at MIT, is called “dual literacy”. The concept draws on C.P. Snow’s distinction between the two cultures, and argues that there are two modes of thought that characterize the two broad major intellectual cultures of our time. The first is that which I have described in talking about the engineering algorithm; the second is the contextual, approximative, non-reductive, and integrative mode of thought in fields like philosophy, literature, history, the arts, and some o the social sciences. In these fields, disciplined thought entails confrontation with ambiguity and the recognition of the perspectival nature of truth. It entails the development of a disciplined ability to deal intellectually with what Thomas Hughes, the historian of technology, calls the “messy complexity” of the actual world.

There is an MIT joke that illustrates the problem of dual literacy. Halfway between Harvard and MIT there was a large supermarket. In the supermarket was one checkout counter with a huge sign saying in large letters: “Rapid Checkout! No More than Six Items!”. A university student approached the “Rapid Checkout” counter with his basket brimming over with many dozens of groceries. As he reached the counter, the woman behind the counter asked, “Are you from MIT and can’t read, or from Harvard and can’t add?”. Restated in her concise terms, dual literacy is the ability both to add and to read.

But many years of teaching at a school of engineering makes me realize how difficult this goal is to achieve. Simply learning to add is enormously demanding, especially today when science-based engineering has advanced so deeply into the mysteries of the natural world. There are, of course, a few exceptional students for whom dual literacy is possible in the four or five years of a university education. And we cannot build an entire educational program around the hope of transforming all of our students into students who are, so to speak, a genetic rarity.

As I noted, efforts to introduce more “breadth” -that is, more courses that deal rigorously with the messy complexity of the real world- into the four -of five- year undergraduate curriculum at MIT have so far failed, paradoxically leading to an intensification of technical demands of the curriculum. To be sure, all students at MIT are required to do the equivalent of a year’s work in the humanities and social sciences. But there fields constitute a smaller proportion of the work in the new five-year curricula than hey did in the four-year curriculum. Having myself struggled to increase the amount of “breadth” within the MIT curriculum for almost two decades, I reluctantly conclude that this effort runs against the most powerful pressures in American engineering education, namely the demand for more time for the technical fields.

What is to be Done?

If my analysis of the contradiction between the real lives of today’s and tomorrow’s engineers and the engineering algorithm is correct, we need to think about how to respond to this disjuncture in the education of engineers.

In the future, I expect and hope that at places like MIT, and perhaps elsewhere as well, we will end up producing essentially three kinds of engineers.

The first will be the traditional engineers whose work remains well within the classical paradigm -the technical problem-solvers who devote their professional lives to using the engineering algorithm for which classical engineering education has prepared them. Such individuals will of course rarely become the presidents or chief executive officers of their firms. They will be a technical intelligentsia, largely doing the work assigned them by others in public and private enterprises. But we should remember that they will also be those who develop the next generation of integrated chips, the innovative new computer programs, the fail-safe nuclear reactors, the medicines that may cure today’s incurable diseases, and the bio-engineered plants that will revolutionize -for the second time- the production of food. The former President of MIT was correct when he said that the answer -or at least one answer- to bad technologies was good technologies. We need the accomplishments of gifted engineers, trained to work at the frontiers of their fields; we should praise their creative inventions and seek to train them well.

I believe we will also see a second type, whom I have called the dually literate. These are students who have, and who generally arrive at our institutes with, an extraordinary competence both for what the checkout counter lady called adding and reading. They possess as much gift for Dante, Goethe, and Shakespeare as they do for differential equations and thermodynamics. Many of them, at least at MIT, go on into fields other than engineering: medicine, legal studies, public administration, and business leadership. We should seek to provide these students in a school of engineering with enough exposure to the human sciences so that their talents in this area do not atrophy, and so that they learn that disciplined inquiry and intellectual excellence is a possible and necessary in the humanities as it is in engineering.

Finally, I believe and hope that we will see, in increasing numbers, a new kind of engineer who has discovered, often after a few years of work in the “real world”, that the engineering algorithm is not enough for hi or her professional life. My colleague at MIT, Professor Benson Snyder, studied a group of MIT engineers twenty years after they graduated. He found that a substantial number, who as undergraduates operated only within the limits of the engineering algorithm, had later discovered that they needed what they called “another way of thinking” for effective work in the real world. MIT’s engineering education had not prepared them intellectually for the needs of team research, of balancing trade-offs, of global consequences, of internalizing externalities, of ambiguous problems and solutions, of policy and politics in business life, medicine, or government. In order to deal effectively and intelligently with that messy complexity, they needed another conceptual framework. Many of them had in fact returned to higher education to acquire that framework in disciplined context. In America, the most common way of doing this is to attend a business school, and some had done precisely that. But the study of business, however useful, is perhaps not the best way for engineers to learn to deal with complexity. For well-trained engineers too often see in the advanced study of management simply an extension of the engineering algorithm to a set of equations designed to maximize profit, rate of return, or market share. What we need even more, I believe, are new graduate programs in which technological excellence, creativity, elegance, and accuracy are valued even as students learn to deal more systematically and intelligently with the non-engineering aspects of the world.

In this connection, K should note that at my own institute of technology, we are developing a series of programs intended precisely for such students. One, for example, the Leaders in Manufacturing Program, has been developed jointly with American corporations to train young executives, mostly engineers, in the advanced understanding of systems of production and manufacturing. The focus is not on engineering alone, but rather on the management of technology. A second program is the Technology and Policy Program, which mostly recruits young engineers with several years of work experience who return to a graduate program that provides them with the tools for analyzing technology policy. Some then go on to further research in technology policy; some return to corporate life as technology or environmental experts; others go to work for public agencies. There is also a new program being developed in the Management of Complex Socio-Technical System, which focuses on the relationship, including the historical relationship, between people, organizations, and technologies in modern industry. All of these programs are intended for people, organizations, and technologies in modern industry. All of these programs are intended for people who have, after graduation from MIT, discovered the need for what Benson Snyder called “another way of thinking”. All are intended to give engineers tools to help them think beyond the engineering paradigm, to deal effectively with the messy complexity of the world in which they will work.

It is obviously inappropriate for a foreigner, unfamiliar with Italian engineering education, to offer proposals for you who are on the frontlines of engineering education in Turin. But in conclusion two points seem worth considering. First, it seems important that engineering students be exposed, from the start of their higher education, to the fact that there are indeed disciplined ways of thinking in addition to the engineering algorithm, and that they learn early on that the fields of science and engineering are themselves products of extraordinarily complex historical, human, cultural, political, and social traditions. It is useful, for example, to ask students to understand something about the history of science and technology, something about the policy implications of their future work, something of how they will be interwoven with the rest of society. In other words, it is important that there be a systematic opportunity and perhaps a requirement, even during the early years of engineering study, also to be exposed to what the French call les sciences humaines.

Second, I think it will become increasingly important to provide young engineers after a few years’ experience in the field, with the opportunity to return to institutes of engineering for additional training. Some of that training will be technical; it will help them to remain on the frontiers of knowledge in fields that undergo revolutions each decade. But another important part will be intended to equip them to deal with the ambiguities and complexities of the socio-technical systems within which they work, and which they may, one day, lead. In America, we should not assign this task only to schools of business and management, for they too often neglect the very virtues of technological excellence and competence which are at the heart of institutes of engineering. What we need even more, at least at places like MIT, are post-graduate programs in which disciplined understanding of the messy complexity of the socio-technical world is combined with a continuing respect for technological elegance, accuracy, and creativity.

So my broad conclusion can be summarized in a few sentences. A century ago, the world made an extraordinary discovery: that practical problems could be solved by the systematic application of science. I have called this discovery the engineering algorithm, and argued that modern engineering education is grounded on this algorithm. This discovery has transformed the world beyond any imagining of those who lived a century ago. In the beginning, the engineering education that grew from this algorithm corresponded adequately with the actual conditions of work of the typical engineer -the superstructure was in harmony with the base. Today, however, a contradiction has opened between the algorithm and the realities of productive life of the engineers we educate. Immensely productive and creative, not to be rejected, this engineering algorithm nonetheless faces a crisis: it does not and cannot prepare engineers for the real contexts within which most of them will work. The fundamental question for engineering education, then, is how to devise methods of preparing graduates more adequately for that complex and messy world, but without losing the dedication to the brilliant problem solving algorithm of engineering. This is no easy question, and we at MIT join you at the Politecnico in seeking answers.

Thank You.