Technology in Society 26 (2004) 385–390

www.elsevier.com/locate/techsoc

Engineering ethics and society

Wm. A. Wulf 
National Academy of Engineering, 2101 Constitution Ave. NW. Washington, DC 20418, USA

Abstract

The engineering profession has a strong tradition of ethical standards for its members.
However, the practice of engineering is changing in ways that raise ethical issues not pre-
viously considered. These ‘‘macro ethical’’ issues are ones for the profession as a whole
rather than for individual engineers. This paper lays out some of the issues as a way of
encouraging further discussion.
# 2004 Elsevier Ltd. All rights reserved.

Keywords: Engineering; Ethics; Compexity

At the annual meeting of the National Academy of Engineering (NAE), the

president is expected to present a brief lecture on a relevant engineering issue. For
the 1999 meeting, in anticipation of the new millennium, I thought it would be
appropriate to talk about the achievements of engineers in the twentieth century
and the challenges facing them in the twenty-first. Preparation of the first part of
the lecture was relatively easy. We had worked with the engineering professional
societies to create a list of the twenty greatest engineering achievements of the
twentieth century—achievements we had selected based on their impact on people’s
lives rather than for their technological ‘‘gee whiz’’. The resulting list1 is impressive
and includes electrification, the automobile, the airplane, clean water, electronics,
radio and television, agricultural mechanization, computers, the telephone, air con-
ditioning and refrigeration, and ten more. It is striking that our lives are dramati-
cally different and (mostly) better than those of our ancestors because of these
engineering innovations.
However, preparing the second half of the lecture was more difficult, at least at
first. As Niels Bohr said, making predictions is difficult, especially predictions


Tel.: +1-202-334-3201; fax: +1-202-334-1680.
E-mail address: [email protected] (W.A. Wulf).
1
The list can be seen by following the link on the lower left side of the NAE website: www.nae.edu.

0160-791X/$ - see front matter # 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.techsoc.2004.01.030
386 Wm.A. Wulf / Technology in Society 26 (2004) 385–390

about the future. Conventional wisdom about the near-term future of technology
suggests that we there is a wealth of technical opportunities and challenges that
will continue to transform our lives. But my crystal ball is no better than anyone
else’s, and it goes completely foggy after a dozen years or so. How could I antici-
pate the challenges for a full century?
When I looked back at the list of the twentieth-century achievements, I was
struck by two things. First, I was awed by how much the work of engineers mat-
ters. Arguably, engineers and their creations did more to shape our lives than any-
one or anything else in the last century! Just imagine how our lives would be
different if even one thing were missing from that list—electricity, for example.
Second, I realized that the immense social impact of most of our inventions was
not predicted by their inventors. As Norm Augustine said: ‘‘The bottom line is that
the things engineers do have consequences, both positive and negative, sometimes
unintended, often widespread, and occasionally irreversible’’ [1].
Because of the enormous impact of engineers on individuals and society, they
also bear deep moral and ethical responsibilities. That realization changed the nat-
ure of my quest, and in the end I decided I would pose only one challenge for the
twenty-first century—engineering ethics. Specifically, I am convinced that the nat-
ure of engineering is changing in ways that raise new kinds of ethical issues, ones
that engineers have not faced before and hence have not thought deeply about.
Those new issues, I believe, have ramifications not only for the behavior of individ-
ual engineers, but also for the profession as a whole and, indeed, for the process of
engineering itself. In this article I would like to elaborate on that theme and its
implications for both the engineering profession and society.
Before proceeding, let me be clear that I believe the vast majority of engineers
behave ethically; I do not believe we are facing some sort of crisis or moral decline.
The purpose of this paper is to explore the ethical implications of new circum-
stances and to help the profession and society prepare to deal with them, not to
suggest that there is a current problem. Nonetheless, the issues I raise are complex,
extremely important, and increasingly urgent, so we need to get on with thinking
them through.
Let me start by saying that I believe the new ethical issues are ones for the pro-
fession rather than ones for the individual. Issues for the profession are called
macro ethical questions in contrast to those for the individual, which are called
micro ethical questions. (This is not to suggest that the macro questions are more
important than the micro ones; they are not!) An example from medicine may help
explain the difference. In medicine the micro ethical issues are much like those in
engineering. The ‘‘First do no harm. . .’’ of the Hippocratic oath is much like the
‘‘Hold paramount the safety, health and welfare of the public. . .’’ of the NSPE
code.2 Both focus on the behavior of the individual practitioner.

2
The National Society of Professional Engineers (NSPE) code is the model for the ethics codes of
many other professional engineering societies.
 Wm.A. Wulf / Technology in Society 26 (2004) 385–390 387

Medicine has had to face other macro ethical problems, such as the allocation of
scarce organs, drugs, or physicians’ time. When there are fewer organs than
patients needing transplants, who gets them? Problems like allocation are not for
the individual doctor to decide; rather, the profession, or perhaps society guided by
the profession, needs to make these decisions. These are the macro ethical issues,
the kinds that changes in the nature of engineering practice (and our perception of
it) are raising.
Enough generalities! Let’s consider some specifics. A class of macro ethical ques-
tions arise because of our increasing inability to predict all the behaviors of the sys-
tems we build, and the near certainty that some unpredicted behaviors will be
deleterious and possibly catastrophic. How does one engineer such systems ethically?
The inability to predict all the behaviors of a system may arise from several
sources:
1. Complexity. Any sufficiently complex system will have emergent properties—
those that cannot be predicted a priori. Starting from quantum mechanics, one
would not predict the personality of Bill Wulf, for example; that personality is
an emergent property. There is an increasing number of engineered systems that
have reached the threshold of complexity where emergent properties occur, for
example, current proposals to ‘‘remediate’’ the effects of previous man-made
changes to the Everglades. The complexity of the Everglades ecosystem is such
that one cannot completely understand, much less predict, the outcome of all
the interactions of its components, and hence of an engineered change to them.
Similarly, many of the bugs reported in computer software are not errors or
mistakes in the way we usually use those words, but emergent properties that
are impractical to predict a priori.
2. Chaotic systems. Even when we have a nice, correct mathematical model of a
system, it may be chaotic, that is, the long-term behavior of the system may be
exquisitely sensitive to its initial conditions. The classic example of the flap of a
butterfly’s wings that just might cause a hurricane thousands of miles away sug-
gests that, although unlikely, the consequences of our engineering decisions can
indeed be catastrophic.
3. Discrete systems. For both of the above examples, because their mathematical
descriptions are continuous, we can at least reliably predict their near-term
behavior. That is, because of the old argument from our introductory calculus
class, ‘‘For every E there exists a d. . .’’, we know that for physical systems, small
changes in the input produce small changes in the output. Unfortunately, not so
for discrete (digital) systems. A one-bit change in the pattern of 0s and 1s in my
computer memory can completely change the interpretation of all the others.
Thus a small change in the ‘‘input’’ can cause an huge change in the ‘‘output.’’
Coupled with the enormous number of possible states (patterns of zeros and
ones),3 the lack of continuity in digital systems not only prohibits effective

3
There are on the order of 10100,000,000,000,000,000,000 states in the memory of my 64-megabyte laptop.
By comparison, there are something like 10120 atoms in the universe!
388 Wm.A. Wulf / Technology in Society 26 (2004) 385–390

prediction of their behavior but inhibits exhaustive testing processes as well. For
physical systems, one can test a representative sample of states and let conti-
nuity ensure than others will behave similarly—but not so for digital systems.

It is important to understand that in each of the cases above, it is impossible to

predict all of the behaviors of the systems under consideration. Not that it’s hard
to predict them, and if only we were more careful we could do it. No, it’s imposs-
ible! Because the systems we engineer are becoming ever more complex and are
increasingly critically dependent on digital systems, we are now building systems
with behaviors we cannot predict. How do we ethically engineer such systems?
The answer is not ‘‘don’t do that.’’ Consider the remediation of the Everglades,
for example. We know that past actions by developers and others have damaged
the ecosystem. So doing nothing is essentially a decision to perpetuate the damage.
Yet there is no way to know with absolute confidence what different and possibly
worse damage might result from any particular action. Moreover, we have plenty
of evidence of well-intentioned decisions—‘‘the best we knew at the time’’—turning
out to have unpredicted and negative consequences, for example, damming western
rivers to prevent flooding. What, then, is the ethical thing to do about the Ever-
glades, and how do we proceed in a manner that minimizes the possibility of
unpredicted damage? Even doing nothing is in fact doing something, and doing
anything may have an unpredicted negative outcome. So what is ethical behavior?
Another example of macro ethical questions concerns the combined effect of
many individual decisions. I don’t think anyone would suggest that individual engi-
neers working for automobile manufacturers are unethical simply because of their
employment. Yet 40,000 Americans die each year because of automobile accidents;
the emissions from those autos contribute to the increase of greenhouse gasses that
might (or might not) cause global climate change. I can’t quite bring myself to say
that collectively we engineers have no responsibility for the deleterious effects of
the cars we design. Nor can I argue that any particular individual engineer is
responsible.
In somewhat the same vein is the collective effect of many local decisions on our
global environment. Paving one parking lot and blocking the entrance to an aqui-
fer that is used hundreds of miles away will not destroy the environment, nor will
the diversion of one stream to provide irrigation for a farm field. But thousands or
millions of such decisions can. Moreover, the really important effect may not be
the direct one, but one resulting from the highly connected and interdependent web
of ecological relationships. As Brad Allenby has pointed out, the earth is already
an engineered artifact because of just such decisions, and it is high time for the
engineering profession to (1) take responsibility for the consequences, and (2)
devise a process of engineering that takes them into account when making local
decisions [2,3].
In a sense, these examples of macro ethical issues are not new. There have cer-
tainly been engineering decisions in the past with enormous societal impact, but it
does seem like we are crossing a qualitative threshold. For instance:
 Wm.A. Wulf / Technology in Society 26 (2004) 385–390 389

. The machines that my father designed could not achieve sufficient complexity to
exhibit emergent properties; today a single computer engineer or bio-engineer
can, and routinely does, build systems of that complexity.
. The collective effect of our use of fossil fuels is having a measurable effect on
greenhouse gas concentrations which, combined with recent evidence of rapid
climate change in the past, raise a worrisome specter.
. Because of their low cost, digital systems are being introduced into the control of
nearly every engineered product and process. Because of the inability to thor-
oughly test any but the most trivial of these systems, there will be catastrophic
failures.

The question, then, is what should we do? I have the sense that engineering will
change profoundly over the next few decades in response to these issues, although I
have no idea what the nature of the change will be. Certainly our concept of engin-
eering ethics must expand beyond the cozy codes of individual behavior we are
familiar with. We will have to establish a new understanding with society about
what constitutes ethical behavior on the part of the profession. But it is much lar-
ger than that. The very process of engineering design, which evolved in an environ-
ment where all behaviors of a system could be predicted, will have to change to
accommodate an environment where they cannot be predicted and hence one that
supports our new societal compact. Even the notion that one can completely spe-
cify the properties of a system before building it may need to be re-examined unless
some new approach allows us to bound our uncertainty.
If we cannot a priori completely specify the properties of a solution to an engin-
eering problem, then engineering will become a more interactive process with
society, and the culture of engineering will change as well. Engineering and tech-
nology have always been influenced by society as well as influencing it, but the time
constants have been large and, except for large infrastructure projects, mostly
indirect. Typically, engineers have been reluctant to participate in public policy and
political debates, even when there is a major technical component to the issue
under discussion. But that may be precisely the form of societal interaction needed
in the future.
The first step is to recognize that we have a problem, begin to discuss it, and
encourage scholarship on it. That was my purpose in this paper. The community of
engineers and ethicists must engage in a dialogue; the intersection of these com-
munities is tiny and separately they do not share a vocabulary, much less a schol-
arly value system. Until we enlarge the community that can communicate with
each other through such dialogue, I fear we cannot make progress on the substan-
tive issues we face. I hope to use the NAE to encourage this dialogue, but I would
like to urge the professional societies to do so as well in order to bring their disci-
plinary expertise into the discussion. It is my hope that from these discussions a
body of scholarship will emerge which reflects the ethical issues of the changing
nature of engineering, and from which we can begin to rethink the way we educate
engineers, the way we practice engineering, and the way we engage with society.
390 Wm.A. Wulf / Technology in Society 26 (2004) 385–390

References
[1] Augustine N. Ethics and the second law of thermodynamics. The Bridge, Fall 2002.
[2] Allenby BR. Earth systems engineering and management. Technology and Society 2000;19(4):10–24.
[3] Allenby BR. Observations on the philosophic implications of earth systems engineering and manage-
ment. Batten Institute Working Paper. Darden Graduate School of Business, University of Virginia,
Charlottesville, VA, 2002.

Wm. A. Wulf was elected President of the National Academy of Engineering (NAE) in 1997.He is cur-
rently on leave from the University of Virginia, where he is a University Professor. His research spans
computer architecture, computer security, programming languages, and optimizing compilers. In 1988–
90 Dr. Wulf was Assistant Director of the National Science Foundation. Prior to joining the
University of Virginia, Dr. Wulf founded a software company, Tartan Laboratories, based on research he
did while on the faculty at Carnegie-Mellon University. Dr. Wulf is a member of the National Academy
of Engineering, a Fellow of the American Academy of Arts and Sciences, a Corresponding Member of
the Academia Espanola De Ingeniera, and a Foreign Member of the Russian Academy of Sciences. He
is also a Fellow of four professional societies: the ACM, the IEEE, the AAAS, and AWIS. He has
authored over 100 papers and technical reports, written three books, holds two US patents, and has
supervised over 25 PhDs in Computer Science.