Moral Reasoning and Value Statements Assessment – Professional Code of Ethics Paper
Timmy050607
Kallenberg.ch.3.pdf
!"
3
A G D
If ethics resembles design, and good design requires talking with each other, it should not surprise us that ethics too involves talking. And while our society o/en uses “moral” and “ethical” interchangeably, ethicists will sometimes distinguish between them this way: morality has more to do with what is right and wrong; ethics has more to do with those conversa- tions in which we defend what is right and wrong.
Before we begin thinking about ethics as argument, there are a few cautions to attend. First, I will generally follow the above distinction be- tween morality and ethics. But this distinction is not hard and fast. If a person is a whiz at constructing ethical arguments but lives like a schmuck, such a person would be a double failure—both a failure at morality and a failure at ethical argumentation. 0is is because, as it is commonly as- sumed, solid ethical reasoning ought to be self-persuasive. If someone is unconvinced by their own ethical argument and so lives like the devil, we rightly assume that something is wrong—wrong with the individual and quite possibly wrong with their arguments. Granted, our intuitions about human nature obviously prevent us from taking this too far, that is, from thinking that solid ethical argument will compel people to be good. A/er all, we say, “We’re only human!” (We’ll see in a later chapter that our intu- itions are correct on this point.) Yet we also intuitively reject the idea that ethical argument has no connection whatsoever with how we live. So, as important as it is to be clear about what makes for good ethical argument, ethics-as-argument is only a small part of a much bigger story about how to live well.
Second caution. Arguments—the kind that can be spoken or writ- ten—are generally evaluated along three lines: completeness, soundness,
B! D"#$%&
!#
and validity. 0e test of completeness asks, “Have we forgotten any rele- vant premises?” 0e criterion of soundness asks, “are all the premises true or factual?” And the metric of validity inquires, “does the logical linkage between premises work correctly?” Beginning students o/en think that the most di1cult task is ferreting out the logical form of an argument and discerning whether this linkage is valid or invalid. But in fact, a more common way to go astray is failing to ask, “Have we forgotten any relevant premise?” 0is is di1cult because getting it right depends on a combina- tion of the sharpness of one’s moral eyesight and how well formed one’s moral imagination is. Here, then, is the second caution: Don’t become so enamored with constructing ethical argument that you neglect the more important task (to which the middle and later chapters of this book are devoted), namely, embarking on the journey of clarifying your own moral vision and educating your own moral imagination. One can begin to mas- ter ethical argumentation in a matter of weeks or months, but moral vision and imagination is a much longer journey.
A P P A G B D
0e most concise form of an ethical argument is called a Value Statement.2 Here is an example:
Value Statement: “Professional Engineering (P.E.) licensure3 will land you a government job.”
0ere are two parts to this statement. First, there is a value claim. A value claim is an implicit or explicit claim that something is good/bad, obliga- tory/permitted/forbidden, right/wrong, etc. 0e italicized terms, as well as others such as should/shouldn’t and ought/oughtn’t, are generally thought to constitute moral vocabulary. In the above statement, we can easily de- tect that the speaker of this Value Statement is “thumbs-up” about getting one’s P.E. license. Although the moral term is implicit, we might rewrite the Value Statement in these ways:
Value Claim: “You ought to pursue P.E. licensure because . . .” Value Claim: “Achieving one’s P.E. license is good because . . .” Etc.
Arguing about Good Design
!!
Here the moral terms have been made explicit. We can tell the speaker approves of getting P.E. licensure because the Value Statement uses words like ought and good.
0e second part of the Value Statement is the ground(s) or reason(s) supplied in defense of the claim. In the initial statement, the reason has to do with securing stable employment with the government.
Value Statement: “Achieving one’s P.E. license is good because it will land you a government job.”
In this sentence, we are tipped o2 to there being some sort of special rel- evance or logical link between “Achieving one’s P.E.” and “landing a gov- ernment job” by the word because.
F!"#$% 3.1 Parts of a Value Statement
We might reverse the order of the terms and achieve the same e2ect:
Value Statement: “Earning one’s P.E. license will land you a govern- ment job, therefore, achieving one’s P.E. is good.”
But notice that the tip-o2 term is no longer because, but rather therefore.
F!"#$% 3.1a Parts of a Value Statement
Whichever order we write them, we see clearly that the two parts— the claim, the reason(s)—are linked together by words like therefore or then or because. 0ese linkage words show the logical movement of the Value Statement. Since the reason(s) is o/en said to “support” the claim, we say that the logical 3ow moves from reason(s) to claim. 0is is usually represented by an arrow.
B! D"#$%&
!$
F!"#$% 3.1b Diagramming the Flow of Logic
For the sake of consistency, we will diagram ethical arguments as moving from le/ to right.
E E A
We said at the outset that there are three kinds of questions for evaluating arguments in general. 0ese three also work for evaluating ethical argu- ments in particular. 1. Completeness: Are all the relevant reasons present? Are all the reasons
related to moral good rather than merely instrumental good? 2. Soundness/Veracity: Are all the reasons factual, true, reliable? 3. Validity: Does the logic 3ow correctly? Does the argument employ
proper logical form?
Let’s take these in order and ask them of ethical arguments.
Completeness
First, we can ask whether all the relevant premises are listed. In the case of achieving one’s P.E., perhaps landing a government job (R1) isn’t the only good reason for undertaking this rigorous four-year process. If it is statisti- cally true that P.E.s earn more money on average than their noncerti4ed counterparts (I do not know if this is true), then we could add a second reason (R2):
“. . . and because you’ll earn more money.”
I’ve also heard it argued that achieving one’s P.E. will improve the moral 4ber of the engineer’s character. Again, I don’t know whether this is true or false. But supposing it is true, it can be added as a third reason (R3). Perhaps it can also be shown that P.E.’s have better-looking spouses.
Arguing about Good Design
!%
0is sounds suspicious, but if the data bears this out, then our provisional Value Statement can be stated more completely:
Value Statement: “Achieving one’s P.E. is good because it will land you a government job (R1), you’ll earn more money (R2), it will improve the moral 4ber of your character (R3), and you will be more likely to have an attractive spouse (R4).”
Notice how the strength of the argument currently stands. If some- one wanted to attack this Value Statement, how would you expect them to proceed? 0e strength of the conclusion (“Achieving one’s P.E. is good”) is linked to the reasons. If the reasons are strong, the conclusion will be strong, too. If one or more reasons are untrue, then the conclusion may be likewise weakened. Sometimes a lone reason can carry all the weight. More frequently ethicists look for a cumulative case: all else being equal, the argument with the greatest number of reasons o/en (though not al- ways) is taken to be stronger.
For all we know, there may be a dozen additional reasons that sup- port the wisdom of pursuing P.E. certi4cation. But for the sake of simplic- ity, let’s say these four are all the relevant possibilities. We can diagram the logical 3ow this way:
F!"#$% 3.1c Diagramming the Flow of Logic
B! D"#$%&
!&
Soundness
0is brings us to the second test: Is each reason “sound” or “factual” or “true”? All of a sudden the task of evaluation becomes very tricky. R1 (gov- ernment job) and R2 (higher pay) are worded in such a way that we might possibly estimate their truthfulness by means of simple numerical data. But R3, moral 4ber, is very di1cult to assess. One possible correlative mea- sure might be data about white-collar crime: do P.E.s less frequently than non-P.E.s embezzle, steal intellectual property, and so on? But, of course, not every bad character commits a crime or gets caught when they do. So the data might turn out to be only crudely representative.
But now notice something else. As it stands, R3 (moral 4ber) is it- self a claim. In other words, R3 needs its own set of reasons before it can play a trustworthy role in our argument. 0is problem is not uncommon. Remember, like design work, we expect ethics to be an ongoing conversa- tion. So if the arguer is pressed by an opponent, he or she must supply an additional level of reasons:
F!"#$% 3.1d Second-Order Reasoning
If there are no second-order reasons for R3, then R3 must drop out of the argument. If secondary reasons (R3ʹ etc.) can be found then R3 stays—at least it stays until someone can mount a serious rebuttal. Imagine that we know of a study that shows there really is a signi4cantly lower incidence of white-collar crime among P.E.s. 0is R3ʹ means that R3 (moral 4ber) can stay. On the other hand, if the study that we hope will back up R3 turns out to have used a 3awed method of data sampling, then R3ʹ is out the window, and R3 may need to be abandoned for sheer lack of support.
Arguing about Good Design
!'
0e fourth reason, R4 (attractive spouse), presents a di2erent kind of problem. Suppose a group of listeners is willing to concede that P.E.’s have more beautiful spouses. “But what of it?” they ask. Does spousal beauty really make for a better spouse? Is having a beautiful spouse really a signi!- cant reason compared to qualities that outweigh beauty on the goodness scale?
In contemporary Western society, marriage for beauty alone is looked down upon, because beautiful people are not necessarily nicer, harder working, more faithful, and so on. So the beautiful person who is married for beauty alone is called a “trophy”—something easy to look at, but possi- bly having very little functionality. Since a trophy spouse is not necessarily a “better” spouse, R4 might indeed fail the test of soundness at the outset, not because we know R4 to be false, but because we are correct to doubt that R4 is a morally signi!cant reason. To say the same thing di2erently, R4 may be true, but somewhat beside the point. Of course, someone may one day concoct a good case for why trophy spouses are morally signi4cant. But this would have to be argued out. And until the day a strong argument is presented, R4 must be discarded as trivial in comparison to other goods.
One of the di1culties with analyzing for soundness is the fact that “moral” tends to be a relational term rather than the name of a property. For example, we say that the ability of grass to re3ect green light (say, 5500å) makes for its green color. 0is would be true no matter what coun- try the grass is growing in. Green is thus a property of the grass. But moral terms are more like spatial relations than inherent properties. A grassy 4eld is green in and of itself. But the grassy 4eld is “to the south of ” only in virtue of whatever is to the north. If a lake is to the north and a parking lot to the south, then “to the south of ” can be said of the 4eld with respect to the lake but cannot be said of it with respect to the lot. In other words, “to the south of ” is not an inherent property of the 4eld, but a property dependent upon context.
Frequently, though perhaps not always, a moral term functions more like “to the south of ” than “green.” Consequently, understanding a moral quality requires describing the wider context, usually by telling a story, just as the surrounding context of the grassy 4eld must be described in order to understand the truth of the property “to the south of.”
Let’s return to our argument about P.E.s. A common way to ask about wider context of “good” is: “Good with respect to what?” One audience mistakenly thinks that R1 (landing a government job) is an inherently good thing. But we would be right to inquire why is a government job a good
B! D"#$%&
$(
thing? Good with respect to what? A reasonable response may be “a gov- ernment job is good with respect to stability of income.” 0at may settle the question and the argument then can proceed. But a skeptic can push to the next level by asking whether stable income is necessarily a good thing. Again we ask, “Good with respect to what?” If the protagonist is a parent of very young children, the supervision of whom requires one parent to stay at home, then getting a stable job is a moral good that aims at providing for children during the years when only one parent can work. On the other hand, perhaps the job seeker wants a stable income to support an illegal drug habit rather than to support young children! Or perhaps the person wants the government job because the government gives out the most styl- ish ID badge. Okay, this seems silly. But if either of these were reasons, nether would be a good reason—an illegal drug habit is an evil rather than a good, and the stylish ID badge is only good in the most trivial sense. 0e landing of a government job would be instrumentally good, in the sense that landing a government job helps one achieve what one craves—illegal drugs or a cool ID badge. But what one craves does not appear to be tied to a good that has moral signi4cance. In telling the rest of the story, R1 (land a government job) is exposed as a distraction to the logical 3ow of the real ethical argument.
At this stage, you may be frustrated both at the tediousness of ethical argumentation as well as its highly contingent nature; an ethical argument can rarely, if ever, deliver a knockout punch. When, say, doing a d-e proof in calculus, the answer is categorically irrefutable. But ethics, like almost all of life, is not like mathematics. One must employ skilled judgment within a whirl of contingencies at every step, knowing one might be challenged and shown to be wrong. 0is is not to say that every ethical argument is end- less (though some appear to be). No. It really is possible to make a strong enough case that the audience—the boss, clients, coworkers, etc.—move toward consensus. But to understanding ethics-as-argument is to realize that every consensus is provisional and liable to challenge and revision on another day or by another audience.
Validity
0e third and 4nal class of measures for assessing arguments has to do with logical structure. 0e simplest way to ferret out the invalidity of an argument is to demonstrate that one or more reasons are simply irrelevant. But there are more complex ways that an argument can fail the test of
Arguing about Good Design
$)
validity. In logical terms, the validity of an argument has to do with its form. Most of us have met the logical form called modus ponens, or “the mode of putting.” 0e idea is that if we “put” the antecedent “P” into a conditional “If P then Q” we are guaranteed “Q” as the consequent.
F!"#$% 3.2 Modus Ponens
Logicians spend their careers identifying and proving valid forms of argument just as geometers identify and prove various corollaries in Euclidian geometry. But unlike geometers, logicians also expose certain attempts at argument as 3awed, fallacious, or invalid.4 Consider the fallacy called “a1rming the consequent.” A child knows that if it is pouring rain, then it must be cloudy outside. But even a child would not claim from the fact that it is cloudy outside that it must therefore be pouring rain at this present moment.
F!"#$% 3.3 &e Fallacy of A'rming the Consequent
B! D"#$%&
$*
0e fallacy of “a1rming the consequent” tries to squeeze a guarantee for the existence of P out of “Q” and “P implies Q.” But it cannot be done. My dog, Angie, is decidedly not a cat.5
0e logical 3ow of an argument is sometimes compared to the way movement is thought to be constrained by the material conditions of a mechanism. 0e direction and 3ow of an argument’s logical “mechanism” is symbolized by the arrow. In addition to “a1rming the consequent,” four of the more common ways the “logic arrow” can go awry include ad hominem (attacks on the speaker rather than on what is said); deny- ing the antecedent (If P then Q; not-P, therefore, not-Q);6 generalizing an anecdote (thinking that one example proves a general rule); and post hoc ergo propter hoc (confusing temporal succession with causality). Many of these faulty ways of arguing are well known to us from political campaigns and advertisements. As important as it is to be able to spot fallacious ar- guments, our focus will be more positive, namely, learning how ethical arguments succeed rather than fail. So, we will set aside issues of improper form. But before we can do some constructive work, we must deal with one enormously important objection philosophers once leveled against the logic of ethical arguments. As we shall see, this philosophical worry, though once widely held, can be set to rest by good engineering thinking.
M I O M F D
At stake is whether arguments that move from “fact” to “value” can work. Enlightenment period thinkers (circa eighteenth-century Europe) began to say there is no legitimate way to infer statements of value from state- ments of fact, or statements of “ought” from statements of “is.” 0e work of British empiricist David Hume (d. 1776) led philosophers to claim in all seriousness that the move from “is to ought” constitutes a logical fallacy. 0ey called it “the naturalist fallacy.” It is still believed in some quarters. But since the late 1950s, beginning with the work of Cambridge philoso- pher Elizabeth Anscombe and expanded upon by Notre Dame philoso- pher Alasdair MacIntyre, there has been growing acknowledgment that movement from “is to ought” under some conditions may be not only per- fectly legitimate, but also extremely important for ethical argumentation.7
0e legitimacy of moving from “is to ought” should not surprise those who live in the engineering world. We know that from an engineer- ing perspective, there are very good reasons for linking “ought” to “is.” 0e
Arguing about Good Design
$"
simplest way to see this is by considering an artifact like, say, the wrist- watch. If I say, “0is is a good wristwatch,” people will understand me to mean that it keeps time accurately. Why do people assume I mean this? Because we all know what wristwatches are for. To ask what something is for, is to ask about its intended purpose or intended function.8 When we say, “Wristwatches are for keeping accurate time,” we are stating the func- tional de!nition of the artifact. A functional de4nition is simply an answer to the question, what is this thing for? A functional de4nition enables us to easily move from statement of fact—“0is watch keeps accurate time”—to a statement of value: “0is is a good watch.” Notice, also, that from the functional de4nition we can easily infer how a watch ought to behave: “A wristwatch ought to keep accurate time.” Both descriptive and prescriptive moral vocabulary—both “is good” and “ought”—follow directly from the functional de4nition.
0e reason Hume denied a passage from “is to ought” in ethics was that he denied the possibility of there being a functional de4nition for human beings. But thoughtful people from time immemorial have been asking the question, what is human life for?—and have not produced a uni- vocal answer. But wait a minute. Isn’t this like design? Just because there is no perfect answer does not mean there is no answer at all! It doesn’t matter that our answers are always provisional or hotly contested. 0e point that confounds Hume is that very few people say that no answer is possible. So long as there is the possibility of local, provisional agreement about what life is for, there is a way for us to say how we ought to live. Had Hume said human beings do not have an unambiguous functional de4ni- tion he would have been right. It is true that there is no single, universally accepted explanation of what human life is for. Nevertheless, as in good design, there are a number of proposals on the table. Many of these pro- posals are religious proposals; but atheists, too, have proposals regarding what life is for. So long as we keep arguing about this question, we must concede that the concept of a functional de4nition is at least intelligible. And if a proposed functional de4nition is intelligible, the question about its counterpart, the ought, will also be intelligible.
0e detour into Hume’s so-called naturalistic fallacy has taken us far a4eld. All that engineering ethics needs is for the concept of functional de4nition to be intelligible. 0is gives us the leverage we need to move from facts to value, or in our lingo, from reason(s) to a value claim.
While the philosophers and theologians debate what human life is for, our task will be much easier. We are about the business of constructing
B! D"#$%&
$#
arguments (admittedly provisional ones) in defense of some value claim. So, we argue about whether “this bridge is good” or “that video game is bad,” by saying so and supplying supporting reasons.
W
0e technical term for moving from a factual reason to a value claim is “warrant.” You’ll notice that value claims employ moral vocabulary: good/ bad, obligatory/permitted/forbidden, should/shouldn’t, moral/immoral, and so on. 0is vocabulary—whether implicitly or explicitly expressed—is what distinguishes the sentence as containing a value claim. You’ll also notice that reasons given are shorn of any such special vocabulary. (0is absence is what led Hume wrongly to think that there is no way to draw conclusions about value from statements of fact.) Reasons use everyday terms—for example, “Government jobs pay more than nongovernment jobs.” 0e role of what we are calling here the warrant is to make explicit how the link is made between factual reasons and the moral vocabulary that appears in value claims. We have already met one of these logical links: the functional de4nition. A functional de4nition belongs to the class of logical links called “warrants.”
0e following example is so very obvious that no one would bother diagramming it. But because it is obvious, it can help us see the warrant in action. We are all familiar with the hydraulic cutter/spreader called “0e Jaws of Life” that rescue workers use to extract trapped persons from wrecked automobiles. As an artifact, we can state the value claim without hesitation: “0e Jaws of Life are morally good.” Here the word morally indicates that “good” means more than simply “it works e1ciently” (al- though e2ective performance is included in “good” design). 0is device is good both with respect to its means (it works e2ectively) and with respect to its aim or end (it aims to extract trapped human beings from danger).
If someone asked, “Is the thing called Jaws of Life good?” you could reply, “Yes! Because it saves lives.” So the obvious Value Statement is:
Value Statement: “0e Jaws of Life are good because they save lives.”
We’d be shocked if someone pressed us further: “Sure they save lives, but are they good?” We’d think that they were being sarcastic or perhaps that they do not speak English 3uently. Nevertheless, the clearest rebuttal to their wrongheaded question is simply: “Anything that saves human lives is morally good.”
Arguing about Good Design
$!
"is is the warrant. 0e warrant links the fact (“Jaws of Life save lives”) with the moral vocabulary of the value claim (“is good”). A more complete diagram, therefore, of ethical argumentation is as follows:
F!"#$% 3.4 Warrants Support the Logical Linkage
In the example just given, the warrant is so obvious no one would ever ask for it to be spelled out. Nevertheless, a warrant must be implicit in the argument in order for the argument to succeed. 0is warrant works, as it were, behind the scenes and out of view. But it is still in play.
We now see that an ethical argument can be analyzed into three pri- mary parts: the value claim (the conclusion), the reasons, and the warrant that links the reasons to the conclusion. 0e strength of an argument is a function of these three parts. In an earlier illustration we said that the strength of the conclusion (“Achieving one’s P.E. is good”) is dependent upon the strength of the reasons. We can now say that the strength of the conclusion is also related to the relevance of the warrant. If someone wanted to challenge a Value Statement, he or she might question one or more of the reasons. But they might also challenge the warrant. On such occasions, warrants need to be defended. (See section below called “Back- ing.”) Let’s see how warrants function in everyday discussions concerning design and technology.
Types of Warrants
As an example of a morally good artifact, the Jaws of Life is a nice and tidy example that few would oppose. More typically, ethical arguments about the moral worth of an engineering project will be messier and more complex. For that reason, it is important that we understand the
B! D"#$%&
$$
full complement of acceptable ways of moving from factual reason(s) to a value claim. For our purposes, there are four classes of warrants.
A. Moral De!nition
0e above defense of the Jaws of Life uses a warrant that we will call a mor- al de!nition: “Anything that saves lives is morally good.” It is a de4nition because it is universal in scope (“anything”). 0is de4nition is constructed in unambiguous, black-and-white terms that virtually anyone can under- stand. Not all de4nitions are like this. Some may be more analogical in nature, requiring skill to make the application. 0e more skill required, the more likely we are dealing with a heuristic than with a de4nition that anyone can understand. In many cases, a functional de4nition works like a moral de4nition. In other cases, the functional de4nition works like a heuristic. More on this ambiguity later.
B. Stipulations
0ere is much, much more to living well than following rules. Neverthe- less, from time to time, clear-cut rules that we will call stipulations do play a role in ethical argument. While de4nitions are warrants that use descrip- tive moral vocabulary such as good, stipulations o/en use modal moral vocabulary, words like ought and should. Unsurprisingly, people argue about rules: Which rules are binding? To what extent? Under what condi- tions? Are there any exceptions? And so on. But here are some common stipulations whose authority is widely acknowledged:
0e Ten Commandments, the Code of Hammurabi, and the Golden Rule U.S. federal and state and city laws—because these are far from per- fect, the laws of the land are not the place to end ethical argument, but the place to begin. Laws are best thought of as something like the ground 3oor, a level beneath which we’d best not sink.9
Discipline-speci4c federal regulations (e.g., building codes, material codes, licensure policies, etc.) Professional codes of ethics such as the NSPE Code, the ASME Code, etc.10
Arguing about Good Design
$%
In 2008 an engineer working for San Jose–based so/ware 4rm Quan- tum3D was found guilty of selling 4ghter pilot training so/ware to the Chinese government.11 Here is one possible Value Statement: “Mr. M’s act was wrong because selling his 4rm’s so/ware is intellectual property the/.” 0e value claim is thumbs-down; what Mr. M did was wrong. 0e reason supporting this claim is that his act was a form of stealing. In this instance the logical movement from “the/” to “morally wrong” is war- ranted by a stipulation—in particular, the eighth commandment (or its equivalent from other cultures), “0ou shalt not steal.” Of course, prohibi- tions against stealing are found in many more places than the Hebrew Bible—that is why the prohibition is widely accepted. We only need to quote one form of this warrant to make explicit how the argument works:
F!"#$% 3.5 Stipulation Warrant
C. Principles
A third class of warrants is called the “principle.” To the untrained eye, principles look like de4nitions. For example, one common principle is the utilitarian formula explained in Equation 1.1, namely that the Net Good- ness calculates the product of (goodness) × (likelihood) × (signi!cance). But principles, more than de4nitions, require sophisticated judgment to use appropriately. For this reason, principles may be thought of as a kind of heuristic. But principles di2er from heuristics in scope—principles o/en appear to have universal application and seem to claim universal assent. Whether moral principles have universal applicability and assent is a mat- ter of scholarly debate. For our purposes, a given principle’s applicability, like the usefulness of a given design, has to be decided on a case-by-case basis.
One must be careful when warranting an ethical argument by means of the principle of utilitarianism. 0is principle requires not only skill in
B! D"#$%&
$&
its application, but it also requires data; the utilitarian conclusion holds only when actual numbers are supplied for calculating outcomes. As we saw in chapter 1, this may be extremely di1cult to do. For example, sixty- 4ve years ago it was argued that 3uoridation of the public water supply would result in a marked decrease in tooth decay. 0is was a very heated political debate, because no one knew the long-term ill e2ects of regularly ingesting very small amounts of 3uoride. But today, sixty-4ve years later, the Center for Disease Control can say with con4dence: “For 65 years, community water 3uoridation has been a safe and healthy way to ef- fectively prevent tooth decay. CDC has recognized water 3uoridation as one of the 10 great public health achievements in the 20th century.”12 0e CDC Web site houses all the statistical data one could wish for. So from today’s vantage point, we can compose the following Value Statement: “Fluoridation of the public water supply is good, because it improves human health.” What warrants the move from the reason (“Fluoridation improves human health”) to the value claim (“Fluoridation of the public water supply is good”) is the net gain in human health. We know this net gain—a marked decrease in tooth decay, and undetectable negative side e2ects—to be real rather than imagined because someone along the way did the math and calculated Net Goodness =# (goodness) $ (likelihood) $ (signi!cance). 0e calculation—easy enough today, but di1cult in 1950— is a warrant, because the calculation translates fact (“improved health”) into quanti4ed units of value. (As we shall see below, the function of Rʹ is called “backing.”)
F!"#$% 3.6 Principle Warrant
Arguing about Good Design
$'
D. Heuristics
0e term heuristic was introduced in chapter 2. As you remember, Koen de4nes a heuristic as “anything that provides a plausible aid or direction in the solution of a problem but is in the 4nal analysis unjusti4ed, inca- pable of justi4cation, and fallible.”13 0ere is no straightforward way to apply heuristics to an ethical argument because proper understanding of heuristics depends upon proper development of the skills of the persons employing them. In this regard we are each and always in progress. Yet we can’t wait around until our training is 4nished before undertaking ethical argumentation. In point of fact, doing ethical argumentation, even im- perfectly, is itself part of the training regimen for improving our skills in using heuristics well. As they say, “Anything worth doing is worth doing . . . poorly, until one can learn how to do it better!” Let’s see how a heuristic might warrant an ethical argument about an engineering artifact, in this case, a bridge.14
In 1936, Robert Moses (1888–1981) was awarded the Cornelius Amory Pugsley Gold Medal Award for his work in developing the parks and parkways in Greater New York City. Moses held as many as twelve di2erent New York City and state jobs simultaneously and earned a repu- tation for being “the man who gets things done.” During his reign, he more than tripled the number of playgrounds in NYC—a good thing. But by all accounts he was a power-hungry, egocentric, violent, and racist man capable of great harm.15 Under Moses’ watch, clumsy intersections were replaced by scenic overpasses and elegant throughways that allowed traf- 4c to 3ow smoothly. Two hundred of these overpasses were constructed over the beautiful parkways that led to the parks and beaches. Now the nefarious part: these two hundred overpasses were built with a maximum clearance of only nine feet. Because city busses were twelve feet tall, low bridges e2ectively excluded those who rode busses—namely, poor people, especially blacks and Latinos—from having access to the posh public parks and beaches. Let’s be clear: racial discrimination is evil. Moses in- tended these overpasses for evil and the use to which they were put was evil. Moses is long dead, but the overpasses remain standing, perpetuating moral evil without any living human agent willing it so. One can almost imagine these overpasses as perpetual minions forever carrying forward the evil designs of their maker.16
We understand that these two hundred overpasses are a mixture of good and evil. In terms of safety and aesthetics, they are good: they work well and look beautiful. Yet upon re3ection, we instinctively recoil against
B! D"#$%&
%(
them as evil. In addition to 4nding Moses guilty of both evil intention and evil action, the artifacts themselves can be called morally evil, regardless of who designed and built them.
Value Statement: 0ese two hundred overpasses are evil (in part), because they e2ectively prevent the poor access to public parks and beaches.
0e structure of this value statement is clear: the value claim is thumbs-down—these overpasses are “evil.” 0e reason provided has to do with their discriminatory work: people are prevented access to public parks and beaches simply because they were poor.
What warrant links “preventing beach access” with the moral vo- cabulary “evil”? We can’t say that Moses’ designs were illegal, because they were built before the Civil Rights Act of 1964. We might hazard a moral de4nition: “denying beach access to the poor is evil.” But this de4nition doesn’t seem quite right. In the 4rst place it is too speci4c, too ad hoc.17 If we call this a moral de4nition, we will need to concoct an ad hoc de4nition for absolutely everything we want to call evil. 0e resulting moral system will be crystal clear, but in4nitely large—too large to learn, much less write down! In the second place, this moral de4nition might be opposed by Mo- ses himself. Moses might employ the utilitarian principle against our ques- tionable de4nition by insisting, “0e greatest good for the greatest number of the public is to keep the ri2-ra2 o2 the beaches.” Here we smell a rat; we should immediately and forcefully counter this utilitarian claim by argu- ing that “the poor” cannot be blithely equated with “the ri2-ra2.” But, alas! Do we have the sociological data at hand to show the utilitarian argument is 3awed? Remember, if we’re doing utilitarian consequentialism, we must have the numbers. Since sociology is not our expertise (assuming we are engineers rather than sociologists), our argument might not turn out to be very powerful. So we have two choices. We can either enlist a sociolo- gist for help in gathering data to prove that Moses’ use of the utilitarian principle is 3awed, or we can look for another way of warranting our claim that these two hundred overpasses are evil. 0ere is an alternative; it uses heuristics. To explain how a heuristic can warrant our Value Statement, we’ll need a quick detour to sharpen our skilled judgment.
As I wrote this paragraph, there was a blip in the news that I’m sure will be quickly forgotten. House Speaker John Boehner (R-OH) gave the 2011 commencement address to Catholic University of America in Washington, DC. Eighty-one Catholic academics signed a letter of pro- test. Why did they object? 0ey were protesting Boehner’s voting record;
Arguing about Good Design
%)
Boehner repeatedly and consistently opposed the interests of the poor, disenfranchised, and marginalized persons of society—including chil- dren, minorities, and women. 0us, these eighty-one professors claimed that “Boehner ought not address the graduating class.” A short time a/er the brouhaha died down, a political pundit invited two of the signatories, Dr. Vince Miller (University of Dayton) and Dr. Steve Schneck (Catholic University), to “explain” themselves on national television.18 Of course, talk shows don’t really want rational explanation; they want a spectacle. So at one point the pundit interrupted whoever was talking and demanded to know, “Are you socialists?” Without blinking, Dr. Miller replied, “No—we are Catholics.” If we pause here a moment, we can consider whether “We are Catholics” might function as a heuristic warrant in the implied ethi- cal argument. I think so. For sake of illustration, consider the way social identity functions in Amish moral life.
We can imagine an Amish child asking his mother, “Why can’t we wear zippers?” And we can imagine the parent’s reply: “Because we are Amish.” 0is may sound like a dodge to me. But remember, I am not Amish. And because I am not Amish, I don’t have a very deep under- standing of who the Amish are or how they think. I mean, what if there is a powerful and convincing explanation of why they don’t wear zippers, and what if the explanation cannot be spelled out in a sentence or two? Hmm. Perhaps if I study Amish life and culture carefully, and come to know in a deep way who they are as a people, I may eventually begin to get a handle on why the use of zippers is a big deal.19 Since the explanation of the no-zipper policy is complex and involved, the answer given to the child—“Because we are Amish”—is true and not trivial. When the Amish astounded the world by forgiving the murderer of 4ve Amish children in Nickel Mines, Pennsylvania, they could explain their marvelous, but strikingly odd, response by appeal to this very same heuristic: “Because we are Amish.”20 0e sentence “because we are Amish” is a heuristic, a sentence that requires skill and knowledge and hands-on experience to unpack fully.
As a warrant in ethical arguments, Amish identity is a heuristic because it requires skill—namely, a certain kind of 3uency in Amish practices, history and theology—to understand it. Similarly for Catholic Christians, their religious identity functions as a heuristic warrant. Grant- ed, being a Catholic Christian may not seem as odd to most Americans as being Amish. But there still is some oddness. At any rate, the conservative pundit interviewing Drs. Miller and Schneck clearly did not share this
B! D"#$%&
%*
understanding (or else appeared willfully to misunderstand them), since he mistook the theologians for socialists (socialist is a dirty word to most Americans). So Miller clari4ed his warrant by appealing to something that Catholics like John Boehner and the political pundit should know about Catholic Christianity: Catholics have a deep and abiding commitment to “the preferential option for the poor.” 0is phrase encapsulates Jesus’ teaching and is the cornerstone of Catholic Social Teaching. 0e basic idea is that the extreme poor, the marginalized, the disenfranchised are so disadvantaged that they can never catch up. It is too late to go back in time and grow up again with better nutrition to help brains and bodies develop; it is too late to go back and attend better schools; it is too late to go back and move into an area with better houses and better neighbors and a lower crime rate. 0erefore, enabling the impoverished to survive requires more than merely a level playing 4eld. A level playing 4eld would still give the rest of us a huge edge. Rather, Catholic Christian compassion requires the 4eld to be tipped in favor of the poor.21
0is warrant, the preferential option for the poor, is certainly con- testable—the interviewer thought it blatantly ridiculous, and probably many viewers did too. Nevertheless, it is undeniably the Catholic position, and in fact, the Christian position in general.
We can, at last, complete our simple argument against Moses’ two hundred overpasses. 0e link from “denies beach access” to the moral conclusion “are evil” is warranted by the Catholic identity heuristic: “We are the ones who give preferential option to the poor.”
F!"#$% 3.7 Heuristic Warrant
0is chapter aims to introduce the practice of ethical argumentation in order for us to argue fairly about engineering design. 0e examples have been intentionally simpli4ed so that we can more easily see the template
Arguing about Good Design
%"
for justifying value statements. We expect that real-time justi4cation will be more complex. 0e next level of complexity of ethical justi4cation is called “backing.”
B
Every ethical argument, once made explicit, has the three working parts with which we are now familiar: One or more reasons supply “support” for a claim about something’s value or goodness (thereby obliging humans to behave in certain ways) or its badness (thereby prohibiting certain human behaviors); the character of the “support” is spelled out in the warrant. As we have seen, the paradigm case is spelled out by Figure 3.4. But this diagram looks suspiciously simplistic. Surely there is more to ethics than three boxes and two arrows! And there is: the reason and warrant are o/en themselves the conclusions of other supporting arguments (Figure 3.1d).
F!"#$% 3.8 A Nearly Endless Regress of Justification
0eoretically, nothing prevents this process of justi4cation from ex- tending to the le/ forever. Practically speaking, however, real-time conver- sations between two people rarely extend backwards more than a step or two. Nor can printed “conversations”—that is, book-length treatments of a moral issue—go on forever. While an ethics book can extend justi4cation many more steps to the le/ than can ordinary conversations over co2ee, even a book truncates justi4cation eventually. But where to stop? If a Value
B! D"#$%&
%#
Statement is 4rst-order justi4cation, we will call everything else second- order justi4cation.
F!"#$% 3.9 First- and Second-Order Justification
We know that questions can always be raised against one or more parts of an argument; there simply is no such thing as knockdown proof in discussions about moral goods. 0e key heuristic for knowing when to stop adding more support is this: “Supply su1cient justi4cation to adequately address legitimate questions raised against a given Value Statement.” Rath- er than diagram second-order justi4cation in all its detail, we will adopt a conventional shortcut regarding all these steps to the le/, whether one or one hundred, and simply fold them under the term backing. As a result, the heuristic in play can be clari4ed: “Supply su%cient backing to answer legitimate questions raised against a given Value Statement.”
F!"#$% 3.10 Backing
Arguing about Good Design
%!
As illustrated in Figure 3.10, backing is shorthand for the whole pro- cess of settling questions raised against the original reasons and warrants. In sum, ethical arguments have not just three but four primary parts: claim, reason(s), warrant(s), and backing. Before concluding, let us exam- ine how backing works.
Recall the above illustration, “0e Jaws of Life are morally good.” 0e reason supplied was that “the Jaws of Life save lives.” Can this statement be questioned? Of course. But the question can also be answered. A convinc- ing response to someone who doubts the e1cacy of the Jaws of Life will require some research. Yet before doing research we can imagine what sorts of answers will work. For example, in this instance empirical data might provide backing for the stated reason, “0e Jaws of Life save lives.” A simple Google search takes us to the company Web site where we read that the Jaws of Life have saved “thousands” of lives since 4rst introduced by Hurst Performance, Inc., in 1972.22 Do these data (“thousands”) constitute adequate backing? Well, is this a reliable source? Maybe. 0en again, may- be not. A/er all, we realize that Hurst wants to sell its product, so it is in their best 4nancial interest not to mention the failure rate of the device in their marketing. But let’s suppose for the sake of illustration that the data (“thousands”) is reliable. We can still ask, is “thousands” speci4c enough to quell all objections about whether the device saves enough lives to back the stated reason? Admittedly, “thousands” is pretty vague. But perhaps a more important number is the statistic that thirty-4ve thousand rescue teams carry the portable Jaws of Life.23 Here the backing is not data about the number of lives saved, but an inference drawn from expert opinion: “0irty-4ve thousand teams of rescue experts carry Jaws of Life because they (the experts) think the device is likely to save lives.” Both the number of rescue teams (thirty-4ve thousand) and their status as experts supply backing to the reason. A third possible backing might be found in the fact that the sheer value of a single human life is so great that we do not need a large number but only a single instance of a life saved to back our reason. 0e same Google search reveals testimonial of a woman whose life was apparently saved by a rescue team employing the Jaws. Generally, argu- ment by anecdote is in poor form.24 But one testimonial is better than no testimonial. In 1972, Hurst faced the same di1culty as those dentists who wanted to convince cities to put 3uoride in the water in the 1940s. Months before the 4rst life was actually saved, Hurst likely claimed that the Jaws will in the near future save lives. While the prophecy turned out to be true, it was not a claim that was backed by data. Notice how much stronger the
B! D"#$%&
%$
argument becomes once that 4rst life has been saved. From that moment on, a testimonial in the form of “so-and-so’s life was saved by the Jaws of Life on (day/month/year)” may function as convincing backing.
0e previous paragraph supplied three possible strategies for backing the reason. 0e reason is not the only part of the argument that can be challenged by a doubter. 0e warrant also can be criticized. 0e stated warrant for moving from “saves lives” to “is morally good” is the moral de!nition, “Anything that saves lives is morally good.” Surely this is a warrant to which everyone subscribes. Or is it? Notice that the de4ni- tion leaves some stones unturned. Suppose that the manufacturing of this device requires rare minerals the mining of which statistically shortens the lives of the miners. Doesn’t their human su2ering and shortened life spans undermine our con4dence in the “anything” of “Anything that saves lives is morally good”? In other words, the objector can legitimately ask, “What does it actually cost human beings (in terms of human lives short- ened or physical su2ering incurred) to produce this device that perhaps only occasionally saves lives? Might there be collateral damage that comes close to outweighing the value of saving a life?” At stake is whether the manufacturing process resulting in thirty-4ve thousand Jaws of Life now in the 4eld may cost more lives than are saved by use of the device. 0e form of the challenge is logical, a denial of the universal (“Anything that save lives . . .”) on grounds that we live in a messy world where mechanical gains come at a price: the production of this device has real-world costs in terms of energy consumption, greenhouse gas emissions, health risks to those who mine the metals and breath the fumes, etc.
An adequate backing for this attack on a widely accepted warrant could conceivably take an empirical form. If we could provide real-world data that demonstrate that the costs of production are “acceptable” relative to the number of crash victims saved by the Jaws of Life, then the original universal moral de4nition still stands as warrant for the argument. Al- ternatively, the backing could conceivably take a logical form, observing that costs may detract from the relative weight of thing’s goodness, but costs are unable to change the nature of the thing’s goodness. Since the aimed-for end of the Jaws of Life is indeed the saving of human lives, it is morally good. High costs of production may make its context of use “morally tragic” (i.e., a legitimate good that entails unavoidable negative aspects), but may not make the device itself morally bad. If this kind of in- strumentalist argument works (why it does not need not distract us here),
Arguing about Good Design
%%
then su1cient backing derives from the logical point that costs do not alter the basically good nature of the device.
C
Ethical argument is an important part of ethics. (Remember, it is only one part.) Ethical argumentation would be easier to do if clear-cut de4nitions, stipulations, and principles were ready at hand. Perhaps sometimes they are. But more o/en than not we 4nd ourselves enmeshed in conversations that are murky, and the only warrants at hand are those slippery critters we’ve dubbed “heuristics.” We already know heuristics to be tricky things. But discarding heuristics in favor of something more cut-and-dried would be a grave mistake. Increasingly, scholars are realizing Koen’s right: de4ni- tions, stipulations, and principles are only clear-cut when all the mess is trimmed o2 and a problem is reduced to the ideal case. In the real world, where the mess cannot be trimmed away, pretending things are crystal clear won’t help a bit if things are irreducibly murky all the way down! When we take the messiness of the world seriously, we see that de!nitions, stipulations, and principles function “heuristically.” Ergo, we need to attune ourselves to the sort of di2erence it makes when ethical argumentation is approached through the lens of heuristics. In the next chapter we will look speci4cally at the di2erence it makes to read professional codes of ethics “heuristically.” 0en, in chapter 5, we’ll tackle the really big topic in ethics, which is not ethical argumentation but something called “practical reasoning.”
D Q
1. What do you think makes a reason a moral reason? Why?
2. What is a warrant? How do warrants function in ethical arguments?
3. What kind of utilitarianism is the calculation represented by Figure 3.7? (Recall that in chapter 1 we read about two forms of utilitarian consequen- tialism, namely, cost-bene4t and risk-bene4t analyses.)
4. Give examples of possible backing for the reason listed in Figure 3.7.
B! D"#$%&
%&
5. What kind of approach would you recommend for supplying backing for the warrant in Figure 3.8?
N
1. For a crystal clear introduction to ethical reasoning, see Murphy, Rea- soning and Rhetoric in Religion, chs. 1 and 8.
2. I am indebted to the discussion of these concepts in the opening chapter of Whitbeck, Ethics in Engineering Practice and Research.
3. Professional Engineering (P.E.) licensure requires four years of practi- cal engineering experience a/er the university degree plus at least two exams: the Fundamentals of Engineering (the “F.E.,” two four-hour portions covering 180 discipline-speci4c questions) and the Prin- ciples and Practices of Engineering (the “P.E.”). Achievement of licen- sure subsequently quali4es one to belong to the National Society of Professional Engineers and be governed by the NSPE Code of Ethics.
4. For a simple place to begin sorting out valid and invalid logical forms, see Weston, A Rulebook for Arguments.
5. If you thought hard enough, you might be able to come up with an example in which the a1rmation of a consequent seems to work. But as logicians are keen to point out, those cases where it seems to work are either (a) accidental coincidence of e2ects (rather than Q causing P) or (b) instances of P being a “necessary” condition for Q. But in the latter case, we are no longer dealing with a simple conditional.
6. For a very concise and helpful book on the logic of argumentation, see Weston, A Rulebook for Arguments.
7. Anscombe, “Modern Moral Philosophy.” See also the works of Ma- cIntyre, especially A&er Virtue.
8. 0e Greeks called a thing’s intended purpose its telos; hence this kind of moral reasoning is known as teleological reasoning. 0is can be confusing; one kind of teleological reasoning looks for expected outcomes (e.g., utilitarianism). But Aristotle’s much earlier version of teleological reasoning derives “ought” from the thing’s intended func- tion or purpose.
9. We also know some laws are downright immoral—there were once laws protecting slaves as the property of their holders. 0e di1culty
Arguing about Good Design
%'
with overturning immoral laws is that to do so will require “revolu- tion” rather than reform. For insightful discussion, see McCabe, Law, Love and Language.
10. See the Online Ethics Center for Engineering and Research, http:// www.onlineethics.org/Resources/ethcodes.aspx.
11. Associated Press, “Engineer Becomes First Sentenced under Eco- nomic Espionage Act.”
12. http://www.cdc.gov/3uoridation/ 13. Koen, Discussion of the Method, 28. 14. 0e following paragraph is borrowed from ch. 1 of Kallenberg, God
and Gadgets. 15. For some of the more shocking details, see Caro, Power Broker. 16. 0e suggestion I just made—that technology can take on a life of its
own—was very much at home in the NT. I will take up this discussion in more detail in chapter 5.
17. To solve a problem ad hoc means to do something that only works in this case. Good design aims to avoid the need for ad hoc 4xes once the artifact is in circulation.
18. 0e interview can be found on YouTube: http://www.youtube.com/ watch?v=x02b9dRWlfQ.
19. See, for example, Kraybill, Riddle of Amish Culture. 20. See Kraybill, Nolt, and Weaver-Zercher, Amish Grace. 21. McCarthy, Heart of Catholic Social Teaching. 22. See company Web site: www.jawso3ife.com/About. 23. Ibid. 24. For examples of bogus and poorly formed argumentation, see Weston,
A Rulebook for Arguments.
