chemical engineering ethics HW

Hastings Center Report, May-June 1996

Ethics as Design Doing Justice to Moral Problems

by Caroline Wbitbeck

Solving actual moral problems is not simply a matter of choosing the "best" of several possible responses. It is also a matter of devising possible responses. Design practice in en- gineering affords important lessons about addressing practi- cal problems.

uppose I face a moral problem, how ought I go about figuring out what to do? The question is not

simply how I should evaluate pro- posed courses of action, but how I go about devising such courses of action, a subject on which, as Stuart Hamp- shire observed in 1949 and again in 1989, ethics has had little to say.'

Ethical judgments are important in devising responses to moral prob- lems, of course. These judgments come in many forms, from "What is being proposed is morally wrong" to "This safety factor (or margin) is suf- ficient for the circumstances in which

this object or process will operate." Yet people confronted with ethical problems must do more than simply make judgments. They must figure out what to do. This is the reason for

calling them "agents." Scholars and popular writers alike

often confine themselves to the judge's perspective, for example, when phi- losophers working in professional ethics take the making of moral judg- ments or criteria for praising and blaming to be the whole of their sub- ject matter, or when the press, report- ing on some accident or miscarriage

Caroline Whitbeck, "Ethics as Design: Doing Justice to Moral Problems," Hastings Center Report 26, no. 3 (1996): 9-16.

of science or engineering, takes the main question to be "Who is to blame?" In these cases the restriction of per- spective is fairly explicit. However, as I have discussed elsewhere, it is also implicit in the representation of mor- al problems as dilemmas to which the only solutions are those given with the problem itself, so that the only task is to judge which of the proposed solu- tions is the best (or least bad).2 It is not enough to be able to evaluate well-defined actions, motives, etc., be- cause actual moral problems are not multiple-choice problems. One must devise possible courses of action as well as evaluate them.

Suppose my supervisor tells me to dispose of some regulated toxic sub- stance by dumping it down the drain. In this case part of my problem is that I have been ordered to do something that is potentially injurious to human health and, furthermore, illegal. As- suming that my supervisor knows, as I do, that the substance is a regulated toxic substance (an assumption that I should verify), then my supervisor's order is unethical and illegal. This is an example of a moral judgment that I make in describing the situation.

In the case I have just described the question is what can and should I do? It is not enough to say that I should not dump the waste down the drain. My problem is not the simple choice

of answering yes or no to the question of whether I should follow the order.

I need to figure out what to do about the supervisor's order. Shall I ignore it? Refuse it? Report it to someone? To someone else in the company? To the Environmental Protection Agency? Should I do something else alto- gether? Is there any place I can go for advice about my options in a situ- ation like this? What are the likely consequences of using those channels (if they exist)? Where could I find out those consequences? Also, what do I do with that toxic waste, at least for the present? These are questions with important implications for fairness to others, including people in my or- ganization, and for the health and safety of the public, as well as for my relationship with my supervisor and for my own position within the com- pany. Answering the question of what to do will depend on a variety of fac- tors. Learning what factors to con- sider and how to assess them are com-

ponents of responsible professional behavior.

The importance of finding good ways of acting (and not merely the ability to come up with the right an- swer to a "whether" question) may be brought home by reflecting on when you or I last poured paint solvents, petroleum wastes, acetone (nail pol- ish remover), motor oil, garden pes- ticides, or other household hazardous waste down the drain (or put spent batteries in the trash). Was it only be- fore we were in a position to know that these were environmental haz-

ards? That is, was it only before we could answer the "whether" question correctly? Or was there a time when we knew it was not a good idea to pour it down the drain but did so be- cause we did not know what else to do?

The need for a response is what makes moral problems practical prob- lems. The similarities between moral

problems and another class of practi- cal problems, design problems, are in- structive for thinking about the reso- lution of moral problems and correct- ing some common fallacies about them.

Practical problems may or may not have solutions. Of those that are mor- al problems, some call for coping

rather than for solving. The perennial

9

M

This content downloaded from 128.193.42.7 on Thu, 13 Oct 2016 19:45:42 UTC All use subject to http://about.jstor.org/terms

Hastings Center Report, May-June 1996

problems of human vulnerability, suf- fering, and mortality are such. Ethical problems that call both for solving and for coping have their counterpart in design problems, although good ways of coping count as "solutions" in the case of design problems. For example, design of a system of drain-

to them. Analysis is important but it is not sufficient to devise responses.

Design Problems

Engineers recognize the ability to analyze the designs of others (that is, being an astute judge of designs) as

Denying that there is a uniquely correct solution goes against some common ways of speaking about ethics, such as "doing the right thing" in a situation.

age ditches to cope with (that is, to prevent damage from) periodic flood- ing of a nearby river counts as solving the problem of how to cope with pe- riodic flooding, although the drainage ditches do not keep the river from flooding.

Design problems are problems of making (or repairing) things and processes to satisfy wants and needs. The analogy with moral problems holds for a variety of design problems, from designing or repairing a book- shelf to devising a rotating work schedule, to designing or redesigning an experiment. The analogy between moral problems and problems of en- gineering design is especially instruc- tive, however. Like ethics, design is a

subject in the university curriculum. Therefore, much has been articulated about the design process in engineer- ing. Furthermore, engineering design stands out among college subjects in giving sustained attention to the syn- thetic reasoning necessary to con- struct good responses to practical problems. Because engineering rec- ognizes the importance of engineer- ing design as well as engineering the- ory, it appreciates the importance of practical as well as theoretical prob- lems and of synthetic as well as ana- lytic reasoning. Devising a good re- sponse requires synthetic reasoning. Ethics has paid more attention to ana- lytic reasoning and the analysis of ethical problems and possible answers

a useful skill for designers to possess, but not sufficient to make a person a good designer. For this reason, most engineering schools offer courses in engineering design that are markedly different from the engineering the- ory courses that teach students to un- derstand theory and how to apply it to solving problems with mathemati- cally exact and usually uniquely cor- rect solutions.

The products of design may be sin- gle objects (for example, a bridge at a given site) or a type of object (for example, a new type of toaster) or process (for example, a cost effective way of making newsprint from recy- cled newspapers or a process for making weather-resistant paint). This characterization applies to many types of design outside of engineering and science, but engineering design (and experimental design) are especially instructive for present purposes not only because the design process is well studied in engineering, but be- cause engineering design problems are typically highly constrained, as are challenging ethical problems. The de- sign process, especially in the ways in which it differs from merely analyzing the designs of others, highlights the very aspects of the agent's response to moral problems that philosophy and applied ethics have had difficulty illuminating.

To develop a good response to a moral problem I must typically take

account of a variety of considerations. In situations like the one just de- scribed where there is a question of either negligence or intentional wrong-doing, one prominent consid- eration is how to be fair to everyone. There may be some tension or con- flict between the moral demands or values associated with some pairs of these considerations, but it is often possible at least partially to satisfy most or all of these demands simul- taneously. Indeed, it is a mark of wis- dom to be able to do so. This seem- ingly commonsense observation about ethical problems has been ob- scured in recent years by a preoccu- pation on the part of philosophers with construing ethical problems as irresolvable conflicts between oppos- ing principles or obligations. Al- though such conflicts are occasionally irresolvable, to assume so at the out- set is misguided because it defeats any attempt to do what design engineers often do so well, namely, to satisfy po- tentially conflicting considerations si- multaneously.

The Design Analogy

To illustrate the characteristics of a design problem, consider the design of a mechanically simple object: a child seat to fit on the top of the suit- cases with wheels designed to be wheeled on board an airplane and stored under the seat or in the over- head bin. When removed from the suitcase, the seat must double as a comfortable child seat that will strap into a vacant airline seat, if one is available, and the seat must also fit easily into the overhead compart- ment. Several manufacturers make such suitcases. Most have similar fea- tures, making it possible to design a child seat that fits most of the ones

in use. This is a design project in which I directed three Massachusetts Institute of Technology mechanical engineering students in spring of 1994. One student, Colleen, focused her work on investigating what the potential user would require in such a device-such as ease of cleaning, having a place in the seat to carry a bottle, a pacifier, and similar para- phernalia. Two other students, Lisa and Kimberly, investigated standards and safety requirements and built

10

This content downloaded from 128.193.42.7 on Thu, 13 Oct 2016 19:45:42 UTC All use subject to http://about.jstor.org/terms

Hastings Center Report, May-June 1996

rough prototypes. These prototypes demonstrated that there are solu-

tions to the design problem and de- veloped some of the features of such solutions.

Lisa's and Kimberly's designs are significantly different solutions to the suitcase child seat design problem. For example, in Kimberly's design the long suitcase handle snaps into a clip at the back of the child seat, and a handle that is part of the child seat is used to pull the seat and suitcase. In Lisa's design the long handle on the suitcase continues to be used to pull the bag with the seat attached. The horizontal crossbar that holds the

child in place pivots around its per- manent attachment to the end of the

right armrest and secures into the end of the other armrest. In Kim-

berly's, the crossbar and armrest form a single U-shaped piece that lifts over- head like an old-fashioned high chair tray, pivoting from two attachments to the back of the child seat. (Both designs have the advantage that they do not detach from the rest of the

chair, so they cannot be lost.) Kim- berly's design is larger in dimensions. It would lead to a larger seat that might better suit a heavier child, but would be more expensive to manu- facture. Lisa's seat would accommo-

date most children under two years old, the age at which infants fly free with an adult.

The first point about design prob- lems that is important for moral problems is that:

For interesting or substantive engi- neering design problems, there is rarely, if ever, a uniquely correct so- lution or response, or indeed, any predetermined number of correct responses.

There may be no solution, how- ever-no way of making a thing that answers a given set of specifications. For example, it is not clear that there is any design of the child seat that would both be small enough to make a reasonable suitcase seat and strong enough to satisfy the additional speci- fications for a child's automobile safe-

ty seat. However, if there is one solu- tion to a design problem, there are usually several.

Both problems of engineering de- sign and moral problems may be triv-

ial in that the specification of the

problem leaves little leeway in an ac- ceptable solution. The question of what to do about a promise that one has freely made, in circumstances where no morally compelling coun- terclaims exist, is trivial in this sense: one should keep it. So is the design of a bolt to fasten the housing of the radar for a large commercial aircraft. In both cases devising an appropriate response is not demanding, so the principal moral question is whether one is sufficiently conscientious in carrying out the action to accomplish the goal.

It is for nontrivial moral problems that the analogy with problems of engineering design is most impor- tant. It may not be a great surprise that if there is one course of action

that provides an ethically responsi- ble resolution of a moral problem, a somewhat different one may also be acceptable. However, denying that there is a uniquely correct solution does at least go against some com- mon ways of speaking about ethics, such as "doing the right thing" in a situation.

The initial problem about the toxic waste is an interesting ethical prob- lem with several acceptable responses. It may be possible to change the su- pervisor's mind, perhaps by detailing the potential health effects or the le- gal liability to the company, or by simply stating that I cannot in con- science dump the waste. If the super- visor is adamant, it may be possible to get others in the company-the ethics or environmental office, if any; the legal department, if any-to countermand his order. The charac-

ter of my organization makes a differ- ence to my response, too. Although some organizations have a strict chain of command, others, including most universities, make a point of having "multiple channels" for work- ing through problems. There may be several ways of getting the waste dis- posed of properly while not embar- rassing the company or coworkers more than necessary.

This brings me to the second point about design problems:

Although there is not a uniquely correct solution, nonetheless some possible responses are clearly unac- ceptable-there are wrong answers

even if there is no unique right an-

swer-and some solutions are better than others.

A child seat that could not recline when in the airline seat would be more of an irritation than a comfort

to the child and accompanying adult. A design that lacked any safeguard should the handle slip out of the adult's hand, and so fall, hitting the back of the seat (and the back of youngster's head) on the floor, would be prohibitively dangerous. These are examples of clear criteria for ade- quate designs.

It may seem obvious that there are wrong answers to ethical problems. (Dumping the waste down the drain is a wrong answer, as is dumping the waste under the supervisor's spruce trees.) I draw attention to the exist- ence of wrong answers only because people often say, "There are no right or wrong answers" to ethical prob- lems. Theirs may be simply an at- tempt to acknowledge that there are no uniquely correct solutions to ethi- cal problems or they may be espous- ing an extreme relativism in ethics.

This leads me to a refinement on

the first point above: Although for interesting or sub-

stantive engineering design problems there is rarely, if ever, a uniquely cor- rect solution, two solutions may each have advantages of different sorts, so it is not necessarily true that for any two candidate solutions one must be incontrovertibly better than the other.

In the case of Lisa's and Kimberly's designs, one is not clearly better than the other, although some features of one are clearly better than the corre- sponding features of the other. If no design feature were constrained by the design of some other feature, it might be possible to collect together the best features into one best design. However, -some features are so con- strained; for example, the design of the security strap that fits between the child's legs and runs between the crossbar and the seat depends on the design of the crossbar. Furthermore, even a given feature may be better in some respects (easier to keep clean, more comfortable for the youngster, less expensive to manufacture) and worse in others (more cumbersome for the adult to operate, more likely

to break). Such a feature is likely to

11

This content downloaded from 128.193.42.7 on Thu, 13 Oct 2016 19:45:42 UTC All use subject to http://about.jstor.org/terms

Hastings Center Report, May-June 1996

be an overall advantage for some us- ers and a disadvantage for others.

For ethical problems, too, different courses of action that all satisfy basic constraints may have different advan- tages. Suppose, for example, in the case of the disposal of the toxic waste, my supervisor is acting on habits es- tablished in the 1970s and '80s when

such dumping of waste was far more prevalent. My supervisor may then be open to arguments that things have changed, especially if they come from his boss or from our environmental

or legal department. Before I go to those sources, I might tell him that I think we should get their view and that I intend to. That forewarning may prevent him from feeling under- cut when I do so, and it may even convince him to seek their counsel.

Suppose, instead, I take the approach of saying that I cannot in conscience dump the waste. This has the advan- tage that it does not raise the specter of my continually going over his head-an effect that may be more of a danger if I am new to the job-but it leaves less opportunity for him to find out that things have changed. Suppose that in response to my con- scientious objection he says, "Well if you are so squeamish, I will do it my- self." What, if anything, do I say then? What if he then proceeds to dump the waste? Each of the two avenues that I have outlined has its own ad-

vantages and disadvantages. It may not be possible for me to find out ahead of time which one is more

likely to work well with my particular supervisor.

Notice that many of my subsidiary judgments are about how far to go in convincing my supervisor of the error of his order. After I think of some

ways to do that, I evaluate those par- ticular actions. I must first think of

what further actions I might take be- fore considering whether to take those actions. This point is worth em- phasizing because so often ethical problems have been represented as whether to do x or y, as decision prob- lems between prescribed alternatives and thus implicitly as multiple-choice problems.

Notice, too, that nothing in the ar- gument for a multiplicity of accept- able ethical responses requires ethical

relativism. That is, I do not suggest

that the variety of acceptable solu- tions to complex moral problems stems from agents holding a variety of moral beliefs. Even if all agents had exactly the same moral beliefs, there is frequently a variety of responsible actions one can take. The various ad-

vantages and disadvantages of two ac- ceptable solutions may each better fit the situations of two individuals with

different life circumstances (but the same moral beliefs).

A third and final point about solu- tions to design problems that holds for responses to ethical problems as well is that they must do all of the following:

* Achieve the desired performance or end-for example, create a child seat that fits on a wheel-on-board-

suitcase or fulfill one's responsibility for environmental safety.

* Conform to given specifications or explicit criteria for this act-for ex- ample, the seat must fit inside the overhead rack and be a comfort-

able booster seat that straps into an airline seat; straightening out the toxic waste issue should not take so

much time that I fail in other major responsibilities.

* Be reasonably secure against acci- dents and other miscarriages that might have severe untoward conse- quences.

* Be consistent with existing back- ground constraints-for example, for the child seat do not require very expensive, scarce, or hazard- ous materials for its manufacture;

for any ethical problem do not vio- late anyone's human rights (so it goes without saying that even if fea- sible, killing off the supervisor is not an option).

So far the value of the analogy with design problems has been to draw at- tention to features of ethical prob- lems that often go unnoticed. These features are especially likely to be overlooked when a case is con-

structed to illustrate philosophical points but the discussion of that case is mistaken for one of how an agent should go about resolving a moral problem. Judith Thomson has con- structed a famous example to test the common assumption that the moral

permissibility of abortion turns on

whether the fetus is a person.3 The example does not, and was not meant to, simulate the problem faced by a woman who is considering abortion and enlightens such a woman only to the extent that her deliberations

are influenced by the belief that if the fetus is a person, then abortion is wrong.

Moral Lessons from

Design Problems

The value of the analogy with de- sign problems is that it gives us guid- ance on how to go about responding to moral problems; strategies for ad- dressing design problems have ana- logs for resolving ethical problems.

First, consider the examination of the situation and definition of the

problem. Some assessment is needed just to name the problem. In the case of design problems, the ambiguity is typically limited to lack of knowledge of what potential users might require in a device (and hence the constella- tion of features in it) and of what so- lutions are available already. Often it is not clear how far you can go in meeting some requirements and still satisfy others. For example, in the case of a child seat it would be desir-

able to accommodate large toddlers and three-year-olds, as well as average size two-year-olds, but the suitcase will only support a limited load.

The need at early stages of design to take account of ambiguity or uncer- tainty is illustrated even more clearly in the design of a complex device. As with the suitcase child seat, the exam- ple I choose of a complex device is one so novel that when it was de-

signed, there were no industry stand- ards for the characteristics of such a device. The device is one that auto-

mates testing for a variety of immune factors. At the initial stage designers had to decide such questions as how constant the temperature at which the device maintains chemical reac-

tions has to be: Should the specifica- tions be for a temperature of 37?C ? 1? or 37?C ? 0.1?? Once such specifi- cations were decided upon, the de- signers built a feasibility model, that is, a model that meets the specifica- tions and embodies the core features

of the technology. Such a feasibility model demonstrates that it is possible

12

This content downloaded from 128.193.42.7 on Thu, 13 Oct 2016 19:45:42 UTC All use subject to http://about.jstor.org/terms

Hastings Center Report, May-June 1996

to create the device in question but typically leaves open many questions about the device that will actually be manufactured and sold.

The initial phase of the design of the immunoassay device illustrates the task of problem definition. Engi- neers recognize the importance of allowing for as much flexibility as possible in the definition of the prob- lem, that is, to avoid foreclosing op- tions to change features or to add new ones in successive models to im-

prove safety, performance, reliability, or manufacturability. Comprehensive foresight prevents difficult or costly changes when far along in the proc- ess. For example, retooling for manu- facture (changing the manufacturing process) is very expensive.

The first lesson from design problems for moral problems, then, is to begin by considering the uncertainties in the situ- ation. In the case of ethical problems, the situation may be even more am- biguous, creating even more of a challenge for foresight. At least with a design problem it never happens that what seemed to be a problem of designing a bridge turns out to be a problem of designing a tower. In contrast, if one hears from one per- son that another is doing something wrong, it may be that the second is doing wrong or that the first is slan- dering the second. All that is certain at the beginning is that something is not as it should be, since the first is telling you that the second is guilty of something.

Appreciating ambiguities and un- certainties is important. These are often underemphasized in profes- sional ethics. For example, the origi- nal (1989) edition of On Being a Sci- entist (the National Academy of Sci- ences's handbook on research and

research ethics for young scientists) recommends that when one believes one has witnessed research miscon-

duct, one should talk it over with a

trusted experienced colleague and "[o]nce sure of the facts, the person suspected of misconduct should be contacted privately and given a chance to explain or rectify the situation" (p. 19). Two things are wrong with this piece of advice. One is that, as is now widely recognized, confronting a per- son who has committed misconduct

runs the risk of having him or her

destroy the data record or the like. However, the more general point is that it is often not possible to wait for certainty before acting. The advice to act only when one is certain is a li- cense to avoid action.

What is needed are ways of acting that will prove prudent and fair how- ever uncertainties are resolved. In

asked to dump toxic waste illegally, there would be a particular person who would be my supervisor whose character I might learn more about. There would be an actual organization (a company, a university) with particu- lar policies that I could investigate.

One of the important characteris- tics of a responsible or wise response

The advice to act only when one is certain is a license to avoid action.

that case, agents typically need to fig- ure out whether to gather more evi- dence, how to raise the issue (or gather more evidence) without being unfair to others, and how best to elicit support for their moral concern.

The second lesson from engineering de- sign for moral problems is that the devel- opment of possible solutions is separate from definition of the problem and may require more information. This is one of many features that distinguish moral problems from formal "deci- sion problems." Decision problems or problems in decision analysis include specifying the alternatives among which one is to decide; that is, a fully defined decision problem is a type of multiple-choice problem. This need to develop possible solutions implies, first of all, that statements of moral problems as open-ended situations do more justice to those problems than do statements of them as multiple- choices. (The writers of the latest [1995] edition of On Being a Scientist, took this point to heart and put most of their cases in an open-ended form.)

Furthermore, prior to proposing solutions, agents must frequently clar- ify the problem. Although open- ended statements do more justice to moral problems than do multiple- choice statements, even open-ended statements are only outlines of ethi- cal problems. If one had an actual ethical problem, there would be real details to be examined. For example, if I really had the problem of being

to a practical problem is appropriate investigation of the problem before attempting to solve it. Part of this in- vestigation for design engineers was already mentioned: investigation of the requirements of potential users. This is especially important for an en- tirely new device. If engineers are seeking to design a better mousetrap, say, then they engage in "benchmark- ing," that is, they gather information about the mousetraps already avail- able. Equally important, they investi- gate the demand for mousetraps with features not currently available and try to read the relative importance of features in the mind of the user.

Too often when statements of ethi-

cal problems are presented to stu- dents, their attempts to probe the complexity of the case are cut off. An- swering problems without seeking to investigate them is poor training for understanding and addressing actual moral problems.

From the place of brainstorming in the practice of engineering design we learn more about how an agent goes about developing responses. Brain- storming requires an uncritical at- mosphere in which people can pres- ent half-baked ideas that may later be refined or combined. Articulation of

any half-baked ideas is discouraged in the many ethics classes where adver- sarial debate is the primary method used. Although an adversarial format may provide some useful pre-law training, it does not help develop the

13

This content downloaded from 128.193.42.7 on Thu, 13 Oct 2016 19:45:42 UTC All use subject to http://about.jstor.org/terms

Hastings Center Report, May-June 1996

ability to think constructively about resolving moral problems.

A rather heroic capacity to brain- storm in the face of criticism is dem-

onstrated in the responses of Carol Gilligan's subject Amy, who was asked to respond to the "Heinz dilemma." When Amy is asked if Heinz should steal a drug he cannot afford in order to save the life of his wife, she pro- poses new alternatives to either steal- ing or letting Heinz's wife die:

Well, I don't think so. I think there might be other ways be- sides stealing it, like if he could borrow the money or make a loan or something, but he really shouldn't steal the drug-but his wife shouldn't die either.4

Asked why he should not steal the drug she replies:

If he stole the drug, he might save his wife then, but if he did, he might have to go to jail, and then his wife might get sicker again, and he couldn't get more of the drug, and it might not be good. So, they should really just talk it out and find some other

way to make the money. (p. 28)

The brainstorming (in this case "borrowing the money" or, as Amy elsewhere suggests, persuading the druggist to lower the price) and in- terrogation of the problem are not entirely separable activities. In ad- dressing design problems suggestions from potential users about their needs frequently stimulate new ideas, and ideas for approaches to the de- sign may stimulate new questions for potential users. For ethical problems, additional information gained through interrogating the problem frequently changes the desirability of possible responses.

This point is illustrated by consid- eration of the following situation: A highway safety engineer is allocating resources for safety improvements and considers two intersections. Both have the same number of fatal acci-

dents per year. However, one is in a rural setting and the other is an ur- ban setting. The urban intersection handles on average four times the number of cars as the rural intersec- tion and also has a higher rate of mi-

nor injuries and property damage

than does the rural intersection.

There is just enough money in the budget to improve one intersection. Which one should it be?

The choice of improving the urban intersection is often justified on the ground that there improvements will have the greatest overall reduction of injury and this choice is cited as illus- trating a utilitarian choice of "the greatest good for the greatest num- ber." The choice of the rural intersec-

tion is made on the ground that it is a more dangerous intersection in the sense that the likelihood of a fatal ac-

cident for a given use of the intersec- tion is four times higher. This consid- eration is taken to represent concern for fairness (presumably equal distri- bution of the risk of fatal injury asso- ciated with going through any given intersection) or even respect for indi- vidual rights.

What is relevant here is not how

well this story illustrates the philo- sophical distinctions between utilitar- ian and competing foundationalist schools of thought in ethics, but the danger that this example will be mis- understood as an example of prob- lem-solving. Notice first of all that the problem is presented as a forced choice between spending all of the re- maining resources on one intersec- tion and spending it all on the other. In fact, there would likely be many other choices. For example, putting up traffic signs at both intersections may be an alternative to installing traffic lights at either one. However, even ac- cepting the multiple-choice character of the problem as stated, there is a great deal of potentially relevant in- formation that the example does not tell us about the accidents. For exam-

ple, suppose that at one intersection, but not the other, in all serious acci- dents at least one of the drivers in-

volved was drunk (or fell asleep, or had a heart attack, etc.). Such infor- mation might show that the most cru- cial variable for reducing serious ac- cidents at one site is reducing driver impairment, while at the other it is the physical characteristics of the in- tersection and is best remedied by changing the intersection itself.

A third lesson from design problems concerns acting under time pressure. It is

often important to begin by pursuing sev- eral possible solutions simultaneously, so

that one will not be at a loss f one meets insuperable obstacles, but still avoid spreading one's engies too broadly This admonition applies both to the design of individual features of the device

from the feasibility stage onward, and to approaches to making changes in the design when obstacles are en- countered at later stages.

The need to act under pressure of time is also a common feature of ethi-

cal problems. In the face of time pres- sure it is reasonable to pursue several possibilities simultaneously in case one fails to prove practicable. Con- sider the ideas proposed by Amy, the child who rejects the forced choice of the Heinz dilemma and "brainstorms"

a variety of possible courses of action. The simultaneous pursuit of several options is a mark of good design strat- egy when there is any danger that one line of development may prove unfea- sible. Pursuing several options con- trasts with representing the moral problem as a static situation with static solutions so that the problem is sim- ply one of selecting the right alterna- tive and doggedly pursuing it.

Fourth and finally, the dynamic diar- acter of problem situations has further im- plications. The problem not only re- quires clarification, but even changes and develops over the course of time or is replaced by others. For example, in attempting to avert the Challenger accident, engineer Roger Boisjoly's problem situation began with evi- dence, in the form of blackened grease, that hot gas was escaping through the joints. The problem then became one of conducting experi- ments to test the effect of tempera- ture on the seals, and then one of getting a seal team formed to redes- ign the seals and of getting resources to do so. The final problem became one of stopping the flight in view of predicted record cold temperature.

When the dynamic character of the ethical situation is neglected, "making the best of a bad business" is often

confused with taking an action that is justified in the general case. For ex- ample, a colleague with whom I was working to formulate criteria for re- search ethics raised the question of whether gift authorship is ever ethi- callyjustified. (Gift authorship in a re- search context is the listing of a per- son as an author although the person

14

This content downloaded from 128.193.42.7 on Thu, 13 Oct 2016 19:45:42 UTC All use subject to http://about.jstor.org/terms

Hastings Center Report, May-June 1996

has not contributed substantially to the research reported in the paper.)

My colleague was thinking of a case in which he had an idea for a collabo-

rative project and proposed it to a European researcher who had done some work on a subject that estab- lished some of the ground work for the new effort. The European re- searcher at first expressed interest, but then failed to respond when my colleague actually proposed to start the work. After several communica-

tions brought no response, my col- league undertook the work with members of his own lab only. There was some delay because my col- league's group had to recreate some research materials that would have

been on hand at the European re- searcher's lab. In due course my col- league and one of his post-doctoral fellows completed the research and wrote a manuscript reporting the work. As a courtesy because the work was built in part upon the earlier work of the European researcher, my colleague sent a "pre-print," that is, a copy of the unpublished manuscript, to the European researcher. That re- searcher replied that my colleague could not publish the paper because they had used a virus obtained from the European researcher's lab for a different purpose, and they had not obtained permission to use it for this project. (The sharing of research ma- terials, or the means for making them, is encouraged in science, although no lab is expected to take on great bur- dens to supply others with materials. Some journals require that those who publish articles in their journal fur- nish to others the reagents and simi- lar materials necessary to replicate the work. In this case, however, my colleague had agreed to use the virus only for a single purpose. That pur- pose did not include making materi- als for the project described in the manuscript.)

Hoping to shame the European re- searcher into desisting from his com- plaints, my colleague wrote back ask- ing if there was someone that the re- searcher thought should be added as an author on the manuscript. To his dismay, his European colleague sent a letter back nominating both him-

self and a younger researcher in his lab as coauthors. My colleague's post-

doc was about to take a job in prox- imity to the European colleague. Be- cause this post-doc, who had done nothing wrong, would be vulnerable to retaliation, my colleague decided to go ahead with gift authorship, and added the names of the two Euro-

of smaller problems about what to do next in the face of multiple ambi- guities and uncertainties. To be re- sponsible requires consideration of how to treat others, and what becomes of others and oneself in addressing intermediate problems, as well as in

The analogy with design problems implies that we should expect that even excellent responses to a problem may be improved upon in many cases.

pean researchers to the list of authors on the manuscript. He did this de- spite his firm conviction that gift authorship is a corrupt practice.

I think that many would agree that he "made the best of a bad situation." This does not show that under some

circumstances gift authorship is justi- fied, however, because the situation it- self was one that, as an ethical matter, one should avoid. Going along with gift authorship was just the best, or least bad, thing my colleague could think of to do in this situation. In the

future he would take care not to get into this situation. He would be more careful to check the conditions under

which he received research materials

and would not again make the mis- take of offering gift authorship as an intentionally sarcastic communication.

Improving on Excellence

Part of the explanation for the mis- understanding and misrepresenta- tion of moral problems is that most of recent ethics and applied ethics have neglected the perspective of the moral agent. Instead, ethics has ex- clusively emphasized the perspective of the judge or that of a disengaged critic who views the problem from "nowhere" and treats it as a "math

problem with human beings." For the agent facing a moral problem, not only are possible responses unde- fined, but the nature of the problem situation itself is often ambiguous. As a result, the agent faces a whole series

the final outcome of the larger story in which the smaller problems are located.

Understanding design activity in engineering, especially in the ways in which it differs from merely analyzing existing designs, highlights the as- pects of the agent's response to moral problems that philosophy and ap- plied ethics have had difficulty illumi- nating. The multiply constrained nature of many problems in engineering de- sign provides an excellent model of challenging moral problems that in- volve many types of moral considera- tions, all of which must be taken into account. Many moral problems that are represented as conflicts are better understood as problems in which there are multiple constraints that may or may not turn out to be simul- taneously satisfiable.

The analogy with design problems implies that we should expect that even excellent responses to a prob- lem may be improved upon in many cases. I embrace this implication. To frame moral problems primarily from the vantage point of the judge or the moral critic, rather than from that of the person facing the problem, asso- ciates ethics with judgment and criti- cism and creates incentives for people to insulate themselves from criticism, either by narrowing the scope of the problems they address or by develop- ing ready rationalizations for their behavior. However, pressing prob- lems, both individual problems of how to be a good engineer, teacher,

15

This content downloaded from 128.193.42.7 on Thu, 13 Oct 2016 19:45:42 UTC All use subject to http://about.jstor.org/terms

Hastings Center Report, May-June 1996

parent, or friend, and social prob- lems, such as providing good health care or protecting the environment, are multiply constrained problems that require continuing input and oversight by many individuals and or- ganizations. Recognizing that good resolutions of moral problems can be improved upon should have the salu- tary effect of promoting open, con- structive, and nondefensive discus- sion of moral problems.

Acknowledgments

This article is based on Chapter 1 in the author's Understanding Ethical Problems in Engineering Practice and Research, Cam- bridge University Press, winter 1996.

3. Judith Jarvis Thomson, "Rights and Death," Philosophy and Public Affairs 2 (1973): 146-59.

4. Carol Gilligan, In a Different Voice (Cambridge: Harvard University Press, 1982), p. 27.

References

1. Stuart Hampshire, "Fallacies in Mor- al Philosophy," Mind 58 (1949): 466-82.

2. Caroline Whitbeck, "The Trouble

with Dilemmas,"Journal of Professional Eth- ics 1, nos. 1 & 2, (1992): 119-42.

The Henry Knowles Beecher Award 1996

K. Danner Clouser

In 1968, a year before The Hastings Center was founded, K. Danner Clouser, a Founding Fellow, became the first philosopher in the first depart- ment of humanities in a medical school. While

others were writing about getting a dialogue started between doctors and philosophers, Dan was actually talking with them. And he has been talking with them ever since. He has given hun- dreds of seminars, lectures, or workshops in medi- cal ethics. Unique among contemporary philoso- phers of medicine, his influence, like that of Socra- tes, is primarily oral. However, he has written some very influential articles on bioethics, both on what it is and what it is not, and he has helped shape the writing of others, serving on the editorial board of the first Encyclopedia of Bioethics, the Encyclopedia of Philosophy (Supplement), and as a charter member of the editorial board of The

Journal of Medicine and Philosophy. But mostly he has realized the importance of talking with peo- ple, of showing them, usually gendy, that they are overlooking some important distinction or failing to take into account some significant facts.

He has not been quite so gentle in his introduc- tions when he was master of ceremonies at work-

shops (often for The Hastings Center) or semi- nars or even the 25th Anniversary of The Hast- ings Center. The people he introduced could be

sure that everyone would arrive on time for no one wanted to miss Dan's introduction. Often these introductions would be remembered far

longer than the talk they introduced. Dan showed the complexity of human nature, combining an often irritating humility and self-effacement with an absolutely wicked sense of humor.

His influence in the field of medical ethics is in-

calculable. He has helped scores of colleges and medical schools initiate programs in the humani- ties and medical ethics. He was the Johnny Apple- seed of medical ethics, sprinkling the seeds of new programs throughout the country. He broad- ened his scope in 1991 when, in Cairo, Egypt, he gave an invited keynote address at the First Inter- national Conference on Bioethics in Human Re-

productive Research in the Muslim world.

He has served on the NIH Fetal Tissue Trans-

plantation Research Panel and the NIH National Advisory Council for Human Genome Research. He has received honorary doctorates from Gettys- burg College and Jefferson University. But the greatest tribute to him is the love and affection that he has inspired in his students and colleagues.

Robert Longly Chairman of the Board

The Hastings Center

16

This content downloaded from 128.193.42.7 on Thu, 13 Oct 2016 19:45:42 UTC All use subject to http://about.jstor.org/terms