Theme 6. Ethics in Science

Resnik, David B. (2005). The Ethics of Science. An Introduction. Taylor & Francis e-Library.

CHAPTER 2. Ethical Theory and Applications

“If we want to understand ethics in science, the three most fundamental questions to ask are “what is ethics?,” “what is science?,” and “how are science and ethics related?” People have proposed many different answers to these questions and I do not pretend to have the only defensible ones.

Ethics, Law, Politics, and Religion

In order to answer the first question, we need to distinguish between ethics as a subject matter and ethics as a field of study (or

moral philosophy). Ethics are standards of conduct (or social norms) that prescribe behavior.

551

Standards of conduct do not describe our actual behavior, since people often violate widely accepted standards. For example, most people in the United States accept the idea that one ought to tell the truth, yet many people lie all of the time. Even though people lie all the time, we indicate our endorsement of honesty as a standard of conduct by defending honesty in public, by teaching our children to be honest, and by expressing our disapproval of lying (Gibbard 1986).

Ethicists (or moral philosophers) study standards of conduct. Ethics as a field of study is a normative discipline whose main goals are prescriptive and evaluative rather than descriptive and explanatory (Pojman 1995). Social scientists offer descriptive and explanatory accounts of standards of conduct; ethicists criticize and evaluate those standards (Rest 1986). While the sociologist attempts to understand how often people commit suicide in the United States, the ethicist tries to determine whether suicide can be rationally justified.

In thinking about standards of conduct, it will be useful to distinguish between ethics and morality. Morality consists of a society's most general standards. These standards apply to all people in society regardless of their professional or institutional roles (Pojman 1995). Moral standards distinguish between right and wrong, good and bad, virtue and vice, justice and injustice. Many writers maintain that moral duties and obligations override other ones: if I have a moral duty not to lie, then I should not lie even if my employment requires me to lie. Moral standards include those rules that most people learn in childhood, e.g. “don't lie, cheat, steal, harm other people, etc.” Ethics are not general standards of conduct but the standards of a particular profession, occupation, institution, or group with-in society. The word “ethics,” when used in this way, usually serves as a modifier for another word, e.g. business ethics, medical ethics, sports ethics, military ethics, Muslim ethics, etc.

Professional ethics are standards of conduct that apply to people who occupy a professional occupation or role (Bayles 1988). A person who enters a profession acquires ethical obligations because society trusts them to provide valuable goods and services that cannot be provided unless their conduct conforms to certain standards. Professionals who fail to live up to their ethical obligations betray this trust. For instance, physicians have a special duty to maintain confidentiality that goes way beyond their moral duties to respect privacy. A physician who breaks confidentiality compromises her ability to provide a valuable service and she betrays society's (and the patient's) trust.

Not all standards of conduct are what we would call “ethics,” however. Hence, it is important to distinguish between ethics and other social norms, such as law, politics, and religion. There are several reasons why ethics is not law. First, some actions that are illegal may not be unethical. Speeding is illegal, but one might have an ethical obligation to break the speed limit in order to transport someone to a hospital in an emergency. Second, some actions that are unethical may not be illegal. Most people would agree that lying is unethical but lying is only illegal under certain conditions, e.g. lying on an income tax return, lying when giving sworn testimony, etc. Third, laws can be unethical or immoral (Hart 1961). The United States had laws permitting slavery in the 1800s but most people today would say that those laws were unethical or immoral. Although we have moral and ethical obligations to obey the law, civil disobedience can be justified when immoral or unethical laws exist. Since we can appeal to morality and ethics to justify or criticize laws, many writers maintain that the main function of a legal system is to enforce a society's moral and ethical consensus (Hart 1961).

Fourth, we use different kinds of mechanisms to express, teach, inculcate, and enforce laws and ethics (Hart 1961). Laws are expressed publicly in statutes, penal codes, court rulings, government regulations, and so forth. Although ethics and morals are sometimes made explicit in religious texts, professional codes of conduct, or philosophical writings, many ethical and moral standards are implicit. Laws are also often expressed in highly technical and complex jargon, and we 552

often need specially trained people—lawyers and judges—to interpret them. Ethics and morals, on the other hand, tend to be less technical and complex.

Some standards of conduct are best viewed as political rather than ethical or moral in nature. For example, the principle of “one person, one vote” is an important rule for organizing voting precincts and apportioning elected representatives. The difference between this standard and an ethical or moral norm is that this political maxim focuses on the conduct of groups or social institutions, whereas ethical and moral standards focus on the conduct of individuals. Political standards take a macro-perspective on human affairs; ethical and moral standards adopt a micro­perspective. Politics as a subject of study, on this view, includes those disciplines that take a macro­perspective on human conduct, e.g. political science and political philosophy. However, the distinction between ethics and politics is not absolute since many actions, institutions, and situations can be evaluated from an ethical or political point of view (Rawls 1971). For example, abortion raises moral issues concerning the status of the fetus and a mother's right to self-determination, and it also raises political issues concerning the state's authority to intrude in personal decisions. Since the distinction between ethics and politics is not absolute, this book will explore political issues in science even though it will focus primarily on ethical questions.

Finally, it is important to distinguish between ethics and religion.

Moral Theory

Moral philosophy includes the sub-disciplines normative ethics, applied ethics, and meta­ethics (Frankena 1973). Normative ethics is the study of moral standards, principles, concepts, values, and theories. Applied ethics is the study of ethical dilemmas, choices, and standards in various occupations, professions, and concrete situations and the application of moral theories and concepts in particular contexts (Fox and DeMarco 1990). We have already mentioned several areas of applied ethics in the previous section, namely medical ethics, business ethics, etc. Meta-ethics studies the nature and justification of moral standards, values, principles, and theories and the meaning of moral concepts and terms.

In order to get a better understanding of ethical theory and its applications, it will be useful to introduce several key concepts of moral philosophy. Each person in society gets exposed to a commonsense morality. This morality consists of a wide variety of standards of conduct, duties, obligations, values and principles that come from disparate sources, such as parents, teachers, peers, religious leaders, professionals, literature, music, the media, and so forth (Pojman 1995). Ethicists 553

call these standards a “commonsense” morality because they are the norms that most people learn and practice without any explicit theorizing or deeper analysis. Some of these commonsense morals include principles like “do unto others as you would have them do unto you,” “keep your promises,” “be fair,” “always do your best,” and so on. Some of these commonsense values include happiness, honesty, justice, charity, courage, integrity, community, love, knowledge, and freedom.

A moral theory, on the other hand, attempts to provide a justification or foundation for commonsense morality. Moral theories describe, unify, explicate, and criticize commonly accepted morals (Fox and DeMarco 1990). Moral theories start with a database of commonsense moral convictions, but they can go beyond commonsense by drawing on findings from psychology, sociology, biology, economics, and other sciences. Once we have developed a moral theory, we can use that theory to challenge some of our commonsense moral convictions; these convictions can be revised in light of deeper reflections and analysis. Thus, commonsense morality can change in response to critical reflection and need not be taken at face value. For example, the interpretation of the commonsense maxim “spare the rod and spoil the child” has changed as a result ofcomplement ethics, even though ethics need not be tied to any particular religion or its teachings. If psychological research and deeper reflection on the ethics of disciplining children. In the 1800s this maxim implied that parents and teachers were justified in whipping, beating, and hitting their children for insubordination. Today, research in child development and moral discussions have led us to take a very different view of corporal punishment. Conduct that was perfectly acceptable in the 1800s would now be viewed as child abuse. After using moral theories to change commonsense morality, we can then revise those theories so that they cohere with this new database. This process of revising commonsense morality in light of theory and vice versa can continue indefinitely, and is known as the method of wide reflective equilibrium (Rawls 1971). Most ethicists believe that this method provides the best way of justifying moral theories.

Philosophers and theologians have defended a wide variety of moral theories, each with its own particular slant on morality. Some theories emphasize individual rights and dignity; others emphasize the common good. Some theories are secular; others are religious. Some theories focus on obligations and duties; others focus on virtues and character. Some theories establish moral ideals; others settle for practical principles. Some theories assess consequences in judging actions; others assess motives. Some theories are human-centered; others place human beings in a larger ecological framework. Rather than take the reader through the labyrinth of moral theories that scholars have constructed and refined over the last few centuries, I will provide brief summaries of some of the most influential ones. I will state the basic assumptions these theories make but I will offer no deeper analysis or criticism.

The divine command theory holds that the rightness or wrongness of an act depends on God's commands: an act is right if it conforms to God's will; an act is wrong if it goes against God's will. Many of the world's most popular religions hold some version of this theory in that they base their morals on God's commands. As I mentioned earlier, I believe that it is possible to develop ethical standards that do not depend on particular religions or theologies, but the divine command theory denies this claim. While I think the theory can provide some insight into moral choices and standards, I will not rely on this theory in my analysis of the ethics of science, since I am attempting to develop a secular account of ethics. The remaining theories I consider provide ethics with secular foundations.

Utilitarianism holds that we should act in such a way that we produce the greatest balance of good/bad consequences (or utility) for all people in the long run (Mill 1979). There are two main types of utilitarianism, act-utilitarianism, which holds that individual actions should maximize utility, and ruleutilitarianism, which holds that actions should be based on a system of rules that maximize utility. According to popular caricatures of utilitarianism, this view implies that the ends justify the means and that the good of the many outweighs the good of the few. However, sophisticated versions of this theory can eschew or buffer these implications (Pojman 1995).

According to a view developed by the German Enlightenment philosopher Immanuel Kant (1981), one should always treat rational beings as having intrinsic value or worth, not as mere instruments or objects having only extrinsic value. Kantianism also holds that moral standards should be universalizeable: moral principles are rules that would be followed by all rational beings with a good will. (A person with a good will is motivated by the desire to do her duty for duty's sake.) For 554

Kant, actions must be done for the right reasons in order to be worthy of moral praise. Kantianism implies that individuals should not be sacrificed for the common good, that we have moral duties that do not depend on the consequences of our actions, and that motives matter in assessing the morality of human conduct.

The natural rights theory, like Kantianism, emphasizes the importance of individual rights and freedoms. According to this view, all people have natural rights to life, liberty, and property, and everyone is morally permitted to do anything they wish provided that their actions do not violate the rights of other people (Nozick 1974). Moral rights, on this view, can be thought of as like trump cards in that legitimate rights claims can be used to criticize any actions that would violate those rights. Rights are usually understood in a negative way: people have rights not to have things done to them but they do not have rights to have things done for them. Thus, the right to life implies a right to be killed but not a right to be saved from dying. This view is sometimes characterized as a “minimal morality” because it holds that we have no moral duties to help other people; we acquire obligations to help others through agreements or by initiating special relationships, such as husband-wife or parent-child.

Natural law theories hold that morality is founded on human nature: if an action has a basis in our natural instincts, emotions, or social relations, then it is right; if an action goes against our natural instincts, emotions, or social relations, then it is wrong. Natural law theories also maintain that we should strive to produce or achieve natural goods and eliminate or avoid natural evils (Pojman 1995). Natural goods include life, health, and happiness.

Social contract theorists propose that morality consists of a set of rules that we agree are important to regulate society; it is a kind of social contract we make in order to live together in a society. In justifying moral rules, social contract theorists imagine people as existing in a state of nature prior to the formation of society. In order to live well, people must cooperate; and in order to cooperate, they need some rules for conduct. These rules are the rules of morality, politics, and the law (Pojman 1995).

Virtue approaches to ethics have a long history dating back to Aristotle (1984), and they have been revived in recent years after a prolonged hiatus from the philosophical scene. According to the virtue approach, the central question in morality is not “what ought I to do?” but “what kind of a person should I be?” Our main task in life is to develop certain traits of character known as moral virtues. Some of these virtues might include honesty, integrity, courage, humility, sincerity, kindness, wisdom, temperance, and so on. One develops these virtues in the same way that one develops other character traits, i.e. through repetition and practice. Becoming a virtuous person is thus like becoming a good basketball player or musician. Most virtue theories imply that the virtuous person will follow moral rules and principles as they develop their characters, but character, not duty or obligation, defines ethical conduct.

The ethics of care, a theory inspired by feminist approaches to morality, rejects traditional approaches to ethics on the grounds that they place too much emphasis on duties, rights, and justice. Such theories are too abstract, legalistic, and uncaring, according to this view. As an alternative, the ethics of care holds that our main task in life is to love and care for ourselves and other people. We should cultivate loving and caring relationships in our conduct instead of relying on abstract concepts and principles (Gilligan 1982). In some ways, the ethics of care provides a modern rendition of Jesus' instruction to love your neighbor as you love yourself, and his critique of the Pharisees' legalistic morality.

One final type of theory worth mentioning has been inspired by the environmentalist movement. The deep ecology approach to morality is unlike all of the other approaches to ethics in that it is not human-centered. Humancentered moral theories frame important questions about nature in terms of human interests, rights, obligations, and so on. Thus, many writers maintain that pollution is undesirable only because polluted environments can harm people or prevent them from realizing various goods. Deep ecologists hold that humancentered ethics cannot adequately deal with moral issues involving other species, the land, eco-systems, the atmosphere, and oceans, since there are values in nature that are independent of human interests or rights (Naess 1989). Thus, an eco-system is worth preserving because it has intrinsic, moral value, not because we happen to value it for its economic or social uses. Animals have rights, according to some writers, because they also have 555

intrinsic moral worth and are not mere instruments for the promotion of human interests (Regan 1983).

As you can see, scholars have developed many different moral theories. These different approaches reflect some of the different insights and tensions that we find in our thinking about human beings, society, and nature and they are based on different insights into the nature of morality. Is one of these theories the correct approach to morality? Since we can use the method of wide reflective equilibrium to test moral theories, any theory worth its salt needs to adapt to the evidence provided by science and reflections on common sense. Thus, although these theories look very different at first glance, they often end up supporting similar standards and values after we use wide reflective equilibrium to revise them. The upshot of this is that most theories have similar practical implications.

Since there are a wide variety of moral theories that yield similar results, the most reasonable approach to moral theorizing, I believe, is to accept some kind of pluralism. According to pluralism, there are a number of basic moral standards (or first principles) that can conflict with each other (Ross 1930). Each of these standards has some prima facie justification from commonsense morality and most of these standards are supported by different moral theories. Thus, utilitarians, Kantians, and social contract theorists can all agree that we should not harm other people, that we should not lie, and so forth. Since these principles have a broad base of support, they are usually less controversial than the particular moral theories that support them (Beauchamp and Childress 1994).

Many philosophers who study applied ethics prefer to work with general, ethical principles rather than moral theories because one can use principles to support an ethical decision or a social policy without defending an entire (possibly controversial) moral theory (Fox and DeMarco 1990). Another reason for employing general principles is that they are easier to understand, to teach, and learn than moral theories. Finally, since principles are expressed in very general terms, they can be applied to a variety of cases and interpreted in different ways. This kind of flexibility allows one to apply principles to diverse cases without ignoring important details. Some of these basic moral principles are as follows (Fox and DeMarco 1990):

Nonmalificence: Do not harm yourself or other people. Beneficence: Help yourself and other people.

Autonomy: Allow rational individuals to make free, informed choices. Justice: Treat people fairly; treat equals equally, unequals, unequally. Utility: Maximize the ratio of benefits to harms for all people.

Fidelity: Keep your promises and agreements. Honesty: Do not lie, defraud, deceive, or mislead. Privacy: Respect personal privacy and confidentiality.

Moral philosophers interpret and elaborate these principles by spelling out concepts like “harm,” “benefit,” “fairness,” “rational,” and “deception,” but I will not undertake that task here. For our purposes, we need to recognize that some set of moral principles can be applied to a wide variety of moral choices.

These principles should be viewed as guidelines for conduct rather than hard and fast rules. We should follow these principles in our conduct but exceptions can be made when they conflict with each other or with other standards. When two principles conflict we may decide to follow one principle instead of another. For example, when asked to give our opinion of someone's cooking we may decide to be less than completely honest in order to avoid harming that person. Since conflicts among various principles and standards can arise, we must frequently exercise our moral judgment in deciding how we should act.3 In order to exercise our judgment, we need to understand the particular features of a given situation. Thus, there is an important sense in which ethics are situational: although some general, ethical principles should guide our conduct, we need to base our decisions and actions on the facts and values inherent in particular situations. The next section outlines an approach to moral judgment and reasoning.

Moral Choices

We make choices every waking moment of our lives. Some of these choices are trivial; others are profound. Some choices are informed by personal preferences, tastes, or mere whimsy. Others are based on standards of conduct. Standards of conduct can regulate our actions by providing guidance 556

for many of the choices we face in living. For example, a principle of honesty would obligate a person to tell the truth when faced with the choice of lying about their golf score or telling the truth. It is not easy to follow standards of conduct all of the time since they often conflict with each other or with our personal interests. For example, a person who could win a great deal of money by lying about their golf score would face a conflict between their interest in money and the obligation to tell the truth. People often violate accepted ethical or moral standards for personal gain, but we usually label such actions as immoral and selfish and we disapprove of such conduct.

But a different kind of situation arises when standards of conduct appear to conflict with each other. People often must choose not between ethics (or morality) and self-interest but between different moral, ethical, legal, political, religious, or institutional obligations. In these circumstances, the key question is not “should I do the right thing?,” but “what is the right thing to do?” These problematic choices are known as ethical (or moral) dilemmas. Thus, an ethical dilemma is a situation in which a person can choose between at least two different actions, each of which seem to be well supported by some standard of conduct (Fox and DeMarco 1990). These choices may be between the lesser of two evils or the greater of two goods. Sometimes these choices involve two different ethical standards. For example, suppose a person has made a promise to a drug company to keep its trade secrets, but she discovers that the company has falsified and fabricated some of the data it has provided to the Food and Drug Administration (FDA) on a new drug it has been testing. She faces a conflict between a duty to prevent harms to the public and duties of fidelity and loyalty to the company. Sometimes these choices involve conflict between ethics and the law. For example, a driver who is trying to get someone to the hospital in an emergency situation faces the choice between breaking the speed limit and preventing harm to someone. I believe that most people, including scientists, face ethical dilemmas on a daily basis as they attempt to balance their various obligations and commitments.

So how does one “solve” an ethical dilemma? A solution to an ethical dilemma would be a decision or choice about what one ought to do. A fairly simple method for making any kind of choice also applies to ethical choices. This method is a kind of practical reasoning that proceeds as follows (Fox and DeMarco 1990):

Step 1: Frame a set of questions. Step 2: Gather information.

Step 3: Explore different options.

Step 4: Evaluate options. Step 5: Make a decision.

Step 6: Take action.

Posing a set of questions is usually the first step in any decision. A question can be as simple as “should I do X or not do X?” Once questions have been raised, one can gather information about the relevant facts and circumstances. Obviously, we can paralyze our decision-making abilities by expending too much time and effort gathering information, so we need to also decide which information is relevant. Since not having enough information can also result in poor decisions, it is often better to err in favor of too much information than in favor of too little. After one has gathered information, one also needs to explore different options. What are the different courses of action one can take? This step often requires some imagination and open-mindedness because often we fail to explore all of our options, and this causes us to overlook attractive alternatives. Some-times we can even avoid dilemmas altogether by fulfilling conflicting obligations at different times.

The fourth step is often the most difficult of all the steps, since it requires us to assess different choices in light of our various obligations and the information we have at our disposal. In evaluating these options it is important to ask these questions in the following order: are any of these different actions required or forbidden by the law? Are any of these actions required or forbidden by special (institutional or professional) ethical standards? Do moral principles have some bearing on these different actions? Although I accept a pluralistic approach to morality, and I believe that people who face difficult moral choices need to balance competing standards, these different norms can be ranked (or prioritized) according to the following rule of thumb: other things being equal, moral obligations take precedence over ethical obligations, which take precedence over legal ones. Since all industrial societies depend on the rule of law, we have a moral obligation to obey the law even though breaking 557

the law can be justified in some cases. Laws should be broken only when they require us to do something patently unethical or immoral or when civil disobedience is justified. Since socially valued professions and institutions can function only when people who occupy professional or institutional roles adhere to professional or institutional standards of conduct, we have a moral obligation to follow special ethical standards. However, professional or institutional standards can be violated if they require us to do something illegal or patently immoral. The key point here is that the burden of proof lies on someone who would violate the law or special ethical standards, since these standards have a sound foundation. In an ideal case, a “best” decision will emerge as a result of considering how various standards—legal, ethical, and moral—apply to a given situation. After reasoning through an assessment of the various alternatives, the next step is to make a decision and then take action.

Sometimes it will not be difficult to make a decision as a preponderance of evidence will favor one option over another. However, sometimes two or more options will still appear to be equally good (or bad) after we have conducted a thorough evaluation. If we encounter this difficulty, we can go back to some of the earlier steps in this process. After evaluating our options, we may discover than we need some more information or that there is another option than we did not consider in our initial assessment.

After going back to earlier steps, we can proceed to later ones, and then go back again. Thus, although I have portrayed this method as proceeding in a linear fashion, it often proceeds cyclically. An example of a difficult decision will help illustrate this method. A professor is about to turn in her final grades for the semester when a student comes to her asking about his final grade in her course. She tells him that he will receive a C, and he begs her to change the grade to a B. He explains that he needs to receive a B in the course in order to continue his scholarship at the university. He tells the professor that he will do extra work in order to get a higher grade. The question the professor asks might be “should I allow the student to do extra work for a higher grade?” The information-gathering part of the professor’s decision might involve reviewing the student's grades for the semester and determining whether she made any mistakes in grading. It might also involve looking at the overall grade distribution to see how close the student’s grade of C is to other students’ B grades. Her choices might be: (a) change the grade, provided the student does extra work; (b) change the grade with no conditions; (c) don’t change the grade. In evaluating her decision, she would consider her different obligations. As a professor, she has an obligation to grade fairly, and a grade change might be unfair to the other students; the student does not deserve a better grade. She also has an obligation, as a professor, to try to help her students, and one might argue that the best way to help this student is to change his grade, since he will have to drop out of school if he loses his scholarship.

On the other hand, one might also argue that the student needs to learn some lessons about academic responsibility, hard work, studying, and so forth, and that he can only learn these lessons by getting the C grade. Upon further reflection, the professor decides to gather more information and she asks more questions. She discovers that the student has approached his other professors with the same story and that none of them have agreed to change his grades. The professor decides not to change the student’s grades because (1) she has an obligation to be fair; and (2) although she has an obligation to help the student, it is not at all clear that changing his grade (under any conditions) is the best way to help him. This example seems to be an “easy” case because one solution emerges as the “best” way of solving the dilemma. But sometimes several different solutions can still seem to be just as good, even after we have conducted a thorough evaluation; and there may not be a single, optimal solution (Whitbeck 1995a). Moral decisions thus resemble engineering design problems: just as there may be more than one right way to build a bridge, there may also be more than one right way to solve a moral dilemma. None of this implies, however, that “anything goes” when it comes to solving ethical dilemmas, since some decisions are clearly wrong or unacceptable and some may be better than others with respect to specific standards. A bridge design that does not result in a stable structure would be unacceptable, and we can evaluate different designs with respect to costefficiency, reliability, durability, and other criteria. Likewise, a moral choice that violated most commonly held moral principles would not be acceptable, and we can evaluate different choices in terms of their adherence to beneficence, justice, autonomy, and other moral and ethical standards.

The possibility of multiple solutions also suggests that there can be some justification for allowing non-rational factors to influence moral decisionmaking. When reasoning yields several 558

equally good solutions, and we need to act on one solution, we can appeal to other methods, such as intuition, emotion, or chance. Sometimes a gut feeling may be the final arbiter in a difficult dilemma. This does not mean, of course, that reasoning should not play an important role in decision-making, since the earlier stages of difficult decisions should involve reasoning. It would be foolish to rely on a gut feeling before gathering information or evaluating options. On the other hand, it would be imprudent to fail to make a decision due to one's inability to find a single, optimal solution to a dilemma. Reasoning can be counterproductive when it prevents us from taking effective action.

Before reaching a decision, the following questions can be useful in helping us evaluate our decision process: “Could I justify this decision to a public audience?,” “Can I live with this decision?,” “Can I rely on anyone else's experience or expertise to help me make this decision?” The first question addresses a concern about public accountability. Sometimes people act unethically or simply make poor decisions because they do not believe that their choices or actions will become public. But we often must justify or explain our choices and take responsibility for our actions. This is especially true in a professional setting, where clients and the public demand accountability (Bayles 1988). Focusing on accountability does not guarantee a result that will please everyone, since some people may still disagree with a decision. However, a person who can defend her decision to other people will stand a better chance of making a good choice than a person who cannot.

The second question addresses a concern about personal integrity. Part of living a good life is being proud of one's conduct and character. We often feel shame when we realize that we have made poor choices, that we have acted unwisely, or that we are not acting like the kind person we would like to be. Good choices, actions, and character traits shine like jewels in our lives and give us a sense of pride in our conduct. Good character results from and manifests itself in our deliberations about difficult moral choices. The final question is simply a reminder that we can and should seek advice when faced with difficult choices. It is likely that we know someone who has faced a very similar dilemma and that we can benefit from their wisdom and experience. Of course, we must take responsibility for any decision that we make—we should not let other people make choices that really belong to us—but it is often helpful to know that we are not alone. Scientists can rely on peers or mentors when making difficult ethical choices.

Relativism

Before wrapping up this thumbnail sketch of moral philosophy, I need to address a concern that plagues all discussions of ethics and morality. The concern is that the standards I discuss in this book may be nothing more than the social customs or conventions of a particular society or culture. What is right or wrong, ethical or unethical is relative to a particular society or culture. Hence, any claims I make about the ethics of science will apply only to those cultures or societies that happen to accept the values and standards I discuss here. Worries about relativism arose in my earlier discussions of the application of moral principles and ethical dilemmas. Although I do not have the space to explore this topic in detail, it is important to say a few words about relativism to help clarify discussions of science and ethics.

To focus the issues, I will distinguish between three different types of relativism:

(1) Legal relativism (LR): legal standards are relative to a given nation or society.

(2) Special ethical relativism (SER): special ethical standards are relative to a particular social institution or profession.

(3) General ethical relativism (GER) (or moral relativism): all standards of conduct are relative to a particular society or culture.

LR would seem to be the least controversial of the three views. A nation's sovereignty depends on its ability to make and enforce its own laws. When US citizens make laws, these laws do not apply to other countries and vice versa. Various nations can agree on international laws and treaties, but even these laws would only apply to nations that enter these agreements. If laws are relative to particular nations and their agreements, our obligation to obey the law is therefore also relative to a particular nation (or its agreements with other nations): when in Rome, obey Roman law. Since moral or ethical considerations may justify breaking the law in some cases, we can appeal to these standards in criticizing a nation's laws, however.

SER also seems like a reasonable position, so long as we understand that special ethical standards must still answer to the law and morality. For example, it seems clear that physicians and lawyers can and should obey different ethical standards, since their professions have different goals and provide different goods and services. Although some standards may hold in both professions, e.g. legal and medical ethics both include a duty to maintain confidentiality, lawyers and physicians should obey the standards of their own professions (Bayles, 1988). When practicing medicine, follow medical ethics; when practicing law, follow legal ethics. However, standards of medical or legal ethics should not require physicians or lawyers to violate the law on a routine basis nor should they require physicians or lawyers to do things that are patently immoral. To see this point more clearly, consider some hypothetical standards of conduct in a street gang—“Gang Ethics.” “Ethical” conduct in a gang might include assault, theft, arson, vandalism, murder, racketeering, and other illegal and immoral activities. Could we treat “Gang Ethics” as a professional standard akin to medical ethics? The answer is clearly “no,” since these norms would violate legal and moral standards. We could call gang standards of conduct “social norms” but not “ethical norms.”

This brings us to GER, the most controversial of the three different versions of relativism. Hundreds of years ago, most people in the Western world would have rejected GER as heresy, since morality was viewed as based on God's commands, which do not vary across cultures. Just as there is one, true God, there is also one, true morality, according to this view. People who do not recognize these moral standards are uncivilized, immoral, or irreligious. But things have changed in the last few hundred years, and probably more people in the Western world believe in moral relativism today than at any previous time in history. Several factors can help explain this rise in the belief in moral relativism (Pojman 1995):

(1) A decline in religion: People are less religious today and therefore less apt to believe in an absolute morality based on God's will.

(2) A reaction to the abuses of colonialism: During the colonial era (and before) Western settlers and explorers conquered and converted the “uncivilized” people they found in the Americas, Africa, and the Pacific Islands. These abuses destroyed cultures and exploited indigenous peoples in the name of Western religion and morality.

(3) Multiculturalism: People have become more aware of cultural diversity as a result of immigration, global communication, and empirical data from anthropology.

(4) Science: Many people today view science as the sole arbiter of truth and distrust disciplines, such as moral philosophy, that do not have the same level of objectivity. Further, many widely accepted scientific ideas, such as Darwin's theory of evolution, undermine the belief in objective, moral standards (Dennett 1995).

(5) Philosophy: Many philosophers during the last century have challenged the objectivity of ethics and have defended versions of moral relativism. Some of them, such as Nietzsche and Sartre, have had an impact on Western literature and culture.

It is not my aim in this chapter to give a thorough critique of moral relativism - that task would require at least a whole book in itself. However, I will provide the reader with a brief assessment of this view and its implications for science and ethics.

One of the main arguments for moral relativism is the fact of cultural diversity (Benedict 1946). This fact is a descriptive thesis: different societies have different customs, mores and conventions. Anthropologists note that some cultures worship cows while others eat them; some cultures accept polygamy while others require monogamy; some cultures approve of female circumcision while others abhor this custom; some cultures practice cannibalism while others condemn the eating of human flesh, and so on. The second main argument for moral relativism is the fact of moral disagreement. This fact is also a descriptive claim: many people within a country and across national boundaries disagree about moral issues, such as abortion, euthanasia, capital punishment, human rights, and so forth. From these descriptive claims, those who defend GER draw a prescriptive conclusion: moral standards are relative to a particular society or culture (Pojman 1990). Relativists find additional support for their view in its implications for moral tolerance: if we understand that moral standards are relative to a particular society, then we will tolerate each society's values and standards. We will not pass judgment on other cultures or try to convert them to the “correct” way of living or thinking.

The two facts cited in the last paragraph are virtually beyond dispute, but do they support moral relativism? Critics of moral relativism attempt to debunk these facts in the following ways. First, cultural diversity may not be as great as appears at first glance. Anthropologists have also provided us with some evidence that many cultures have some standards in common. For instance, nearly all cultures have prohibitions against murder, assault, theft, rape, dishonesty, and incest (Pojman 1990). Cultures disagree about who belongs to the moral community, however. A culture might condone killing someone from a different moral community, e.g. someone from a different tribe or ethnic group, yet it might condemn killing someone from within the moral community. Although some moral standards exhibit a great deal of variation, such as standards pertaining to sexual practices, marriage, and personal liberties, one might argue that there are some basic standards held by all societies. We can refer to these standards as a core morality. There may even be a plausible scientific explanation for a core morality: the core morality consists of those standards which are necessary for the survival of any society (Pojman 1990). Without some common standards, societies would dissolve. We may even find that some of these core moral standards have a strong basis in common instincts and emotions, and that these traits have an evolutionary basis (Alexander 1987).

Furthermore, in many cases moral diversity can be explained away by appealing to economic conditions or background beliefs. For instance, people in the US abhor infanticide but people from other countries do not. Many of the cultures that do not abhor infanticide are far worse off eco­nomically than the US and often simply cannot afford to feed defective or unwanted infants. In the US, however, people have more than enough resources to care for defective or unwanted infants. If these countries that accept infanticide had more resources, it is likely that they would not condone this practice. In the US people eat cows; in India people worship them. These different attitudes result, in part, from different metaphysical beliefs about cows: most people in the US regard cows as unintelligent animals; most people in India regard cows as reincarnated people.

Second, people may share some common moral standards even if they disagree about controversial cases and issues. People who disagree about the morality of abortion may nevertheless agree that killing is wrong and that women should be granted the freedom to control their own bodies. At the center of this dispute is the question of the fetus's membership in the moral community—do fetuses have the rights we typically assign to newborn infants? Critics of moral relativism usually attempt to explain how moral disputes can arise even if most people accept a core morality (Pojman 1990).

Critics also point out that the relativist's inferences from descriptive claims about social customs to prescriptive ones about standards of conduct commit the naturalistic fallacy. According to most logicians, it is a fallacy to infer an “ought” from an “is” or a “value” from a “fact.” From the fact that many people do smoke cigarettes, we cannot infer that many people should smoke cigarettes. So, it is a fallacy to infer moral relativism from cultural diversity. From the fact that different cultures have different social customs, we cannot infer that they should have those customs (Frankena 1973).

Finally, critics also attack the relativist's appeal to tolerance by arguing that relativism does not imply tolerance (Pojman 1990). Suppose tolerance is a social custom in society S1 but not in society S2. It follows, if we accept relativism, that people in S1 should be tolerant but people in S2 have no obligation to be tolerant. Therefore, if people in S2 attempt to destroy or change the customs of S1, the people of S1 have no grounds for appeal, since the obligation to be tolerant applies only if you live in a tolerant society. This seems like a very unsatisfactory defense of moral tolerance, but it follows from moral relativism's assumptions. To provide tolerance with a more substantial foundation, one needs to hold that at least one moral standard, tolerance for other cultures, applies to all cultures. The irony here is that a thorough and complete relativism undercuts one of the values relativists usually defend, tolerance.

The alternative to moral relativism is some form of moral objectivism. For the purposes of this discussion, we can distinguish between two versions of objectivism, strong and weak (Pojman 1990). Strong objectivism, also know as absolutism, holds that (1) there are some universal moral standards, and (2) these standards have no exceptions; they are hard and fast rules. Weak objectivism holds (1) but denies (2): moral standards, though universal, are guidelines to conduct, not absolute rules. The weaker version of objectivism would appear to be more plausible than the stronger version, since weaker versions are better at accommodating the facts of cultural diversity and moral 561

disagreement. Moral differences and disputes can arise because different societies interpret and apply universal standards in different ways; general principles have various exceptions. Of all the moral theories that I discussed in the section headed Moral Theory, the pluralistic approach fits best with weak objectivism (Pojman 1990).

Although I think moral objectivism is highly plausible, it has its own difficulties. The main problem for the objectivist is to provide an objective basis for morality. How can some ethical standards apply across cultures? What is the foundation for these universal moral values or principles? There are three traditional answers to this request for an objective foundation for morality:

(1) Naturalism. Morality is based on human biology, psychology, sociology, and so on.

(2) Rationalism. Morality is based on reason itself; moral standards are the rules that any rational, moral agent would accept.

(3) Supernaturalism. Morality is based on God's will.

Each of these approaches generates its own problems. Naturalists must face the naturalistic fallacy and they need to explain how their doctrine does not simply make morality relative to the human species and its needs and goals; rationalists owe us a non-circular, informative, and practical account of rationality; and supernaturalists need to justify belief in the existence of God and explain the relation between morality and God's commands. I will not probe these metaethical problems and issues in more depth here.

As promised, I will bring this discussion back to our main topic, the ethics in science. What implications, if any, do various versions of relativism have for ethical conduct in science? First, let's consider legal relativism. If we apply this view to science, it holds that scientists should follow the laws of the nation in which their research takes place (or the laws that pertain to their research). This is a reasonable view that most scientists probably accept. (Some interesting legal questions arise when research spans over different nations, of course.) Aside from the moral legitimacy of laws, scientists also have some good practical reasons for obeying the law: scientists who violate the law can get into trouble with local authorities and the public at large. Although scientists may still violate laws under extraordinary conditions, they still have a general duty to conform to local laws.

Special ethical relativism also makes sense in science, provided that science's ethics conform to laws and commonly accepted moral standards. As professionals, scientists should follow the standards of their profession, not the standards of other, non-scientific professions. Although accountants and lawyers practice secrecy as a part of their profession, secrecy is generally frowned upon in science, except in some circumstances. When it comes to secrecy and openness, scientists should follow scientific ethics, not legal ethics. Thus, unless they have moral reasons for violating their professional standards, scientists should adhere to those standards. When practicing science, do as the scientists do.

This brings us to the last version of relativism, moral relativism. If scientists give priority to legal and special ethical standards, then concerns about moral relativism will usually play only a small role in scientific conduct. It should matter little to scientists whether honesty is valued in different cultures across the world; their main concern should be that honesty is valued in science. As long as scientists stay within their professional roles in science and do not interact with society at large, concerns about moral relativism should not be a major issue. However, since scientists must often interact with society at large, vexing moral questions arise in science, and it is these questions that summon the specter of moral relativism. Thus, moral relativism can be a real issue for scientists when they decide how to treat human subjects in a different country, how to share intellectual property rights across cultures, whether to conduct research on cloning human embryos, how to discuss research on homosexuality with the public, and so on.

CHAPTER 4. Standards of Ethical Conduct in Science

In the last chapter I argued that science's ethical standards are based on the goals of the scientific profession, which include the quest for knowledge, the elimination of ignorance, and the solution of practical problems. Many of the standards of conduct in science also have a moral foundation. For example, fabricating data is unethical in science because it is a form of lying, which is morally wrong, and because data fabrication promulgates errors and destroys the atmosphere of trust that plays a key role in science. Scientists should practice social responsibility in order to satisfy 562

moral obligations and to secure the public's support for science. Thus, ethical standards in science have two conceptual foundations, morality and science. Ethical conduct in science should not violate commonly accepted moral standards and it should promote the advancement of scientific goals. In this chapter I will defend twelve principles of ethics in science, which apply to different aspects of the research process. After discussing these principles, I will make some additional remarks to clarify my approach to ethics in science. The principles are as follows.

Honesty

Scientists should not fabricate, falsify, or misrepresent data or results. They should be objective, unbiased, and truthful in all aspects of the research process. This principle is science's most important rule because if this principle is not followed, it will be impossible to achieve science's goals. Neither the search for knowledge nor the solution of practical problems can go forward when dishonesty reigns. Honesty also promotes the cooperation and trust necessary for scientific research. Scientists need to be able to trust each other, but this trust breaks down when scientists are not honest (Committee on the Conduct of Science 1994, Whitbeck 1995b). Finally, honesty is justified on moral grounds: all people, including scientists, should be truthful.

In order to understand dishonesty in science, we need to distinguish between dishonesty and error (PSRCR 1992). Dishonesty and error produce similar consequences but they spring from different motives: a dishonest act always involves the intent to deceive an audience that expects to be told the truth. Deception can occur when someone lies, withholds information, or misrepresents information. Dishonesty does not occur when the audience does not expect to be told the truth: novelists do not lie when they tell tall tales. It is important to define dishonesty in terms of motives because motives play a key role in our assessment of human conduct. If scientists were instruments or mechanical devices, then we would only expect them to be reliable: a thermostat can give an accurate or inaccurate reading but it cannot tell the truth or lie. Indeed, since scientists are human beings, we forgive honest errors and reserve our harshest judgments for lies and intentional deceptions.

Many kinds of dishonesty in science involve the production and analysis of data. Fabrication occurs when scientists make up data; falsification occurs when scientists alter data or results (PSRCR 1992). In the Baltimore affair, ImanishiKari was accused of fabricating or falsifying data in her research team's experiments on mice.2 Misrepresentation occurs when scientists do not truthfully or objectively report data or results. The most common forms of misrepresentation are known as trimming, cooking, and fudging (Babbage 1970). Trimming occurs when scientists fail to report results that do not support their hypotheses. Fudging occurs when scientists try to make results appear to be better than they really are. Scientists “cook” the data when they design tests or experiments in order to obtain results they already have good reasons to suspect will be positive or when they avoid conducting tests that are likely to yield negative results.

Most scientists view fabrication and falsification as serious violations of scientific ethics, but there is some disagreement about the seriousness of misrepresentation because the line between misrepresentation of data and good methodology is sometimes ambiguous (Sergestrale 1990). Scientists sometimes have good reasons for eliminating or ignoring recalcitrant data; a certain amount of trimming can be a part of good scientific practice. For instance, some scholars have argued that Millikan trimmed data when he classified results as “good” or “bad” and only reported the “good” results in a paper discussing his oil drop experiments. (Millikan attempted to determine the charge of the electron by measuring the charge on drops of oil and using this measurement to calculate the minimum charge, the charge on an electron.) Others claim that Millikan had good reasons for distinguishing between “good” and “bad” results and for omitting the “bad” ones. Millikan understood his experiments and equipment and exercised his scientific judgment in evaluating his data (Committee on the Conduct of Science 1994).

The same point also applies to fudging and cooking data. Today's scientists often need to use statistical methods to convert masses of disorganized, meaningless data into meaningful, organized numbers or figures. If scientists are justified in using various statistical techniques in analyzing, organizing, and presenting data, then scientists need to exercise judgment and discretion in choosing those techniques. Scientists who misuse statistics can be accused of fudging; those who do not are 563

simply practicing good science. It is even acceptable to design tests in order to get positive results, so long as one does not avoid tests that might yield negative ones. Since there are no explicit rules for trimming data sets, choosing statistical methods, or for designing tests or experiments, scientists must exercise their judgment in deciding how to collect and analyze data. The ability to make judgments about the proper treatment of data can be acquired through experience in the laboratory and by following examples of good science.

Since the line between accurate representation and misrepresenting is often not clear cut, how can we tell when someone is unethically representing data or results? Correct representation of data involves the exercise of scientific judgment. We could rely on the judgment of experienced scientists to determine whether an action counts as misrepresentation. However, since even experts may disagree, we will also need to appeal to the motives or intentions of scientists in order to determine whether they are behaving improperly (PSRCR 1992). If a scientist trims data with the intent of deceiving other people; then he is being dishonest; if he trims data with the intent of reporting results in a clear fashion, then he is not. If a scientist uses statistical techniques in order to give a clear and objective picture of the data, then she is acting ethically; if she uses statistics simply as a rhetorical device to deceive her audience, then she is acting unethically. Of course, it is not always easy to determine a person's intentions.

Although honesty is most important in the production, analysis, and reporting of data and results, honesty also applies to many other aspects of the research process. For instance, in writing research proposals, scientists sometimes stretch the truth in order to have a better chance of obtaining funding (Grinnell 1992). Scientists, engineers, and public relations officials greatly exaggerated the scientific and economic importance of the Super-Conducting Super-Collider in defending this expensive project before Congress (Slakey 1993).

Scientists who act dishonestly may sometimes have reasons for behaving this way. According to my definition, parody is a form of dishonesty, though it might not be unethical. Consider physicist Alan Sokal's (1996a, 1996b, 1996c) parody of cultural studies of science. In order to defend science from social constructivist critics, Sokal put together a paper that parodied their jargon, rhetoric, and reasoning. The paper contained numerous “red flags,” such as errors in reasoning and unintelligible sentences, but the editors of Social Text published it. Sokal later revealed his “experiment” in Lingua Franca. His spoof challenged the judgment of the editors of Social Text as well as the intellectual standards of the entire field of cultural studies of science. Although many people laughed at the editors of Social Text, Sokal's quarrel was not with those editors or their journal. Sokal wrote his parody as a plea for reason, evidence, and logic. (Many people who work within the field known as cultural studies of science argue that reason, evidence, and logic play only a minimal role in scientific discovery, and they argue for the subjectivity of knowledge, truth and reality.) Were Sokal's actions unethical? Although dishonesty is usually unethical, one might argue that parody is not unethical when it is used to expose corruption and scandal in politics and the academy; satire is often the best way of revealing the truth (Rosen 1996). However, since even any lie can damage the integrity of the research process, honesty is the best policy in science and deviations from this standard require a special justification.

Carefulness

Scientists should avoid errors in research, especially in presenting results. They should minimize experimental, methodological, and human errors and avoid selfdeception, bias, and conflicts of interest. Carefulness, like honesty, promotes the goals of science in that errors can hinder the advancement of knowledge as much as outright lies. As we noted earlier, a lack of carefulness is not the same thing as dishonesty, since carelessness need not involve the intent to deceive. Carefulness also is important in promoting cooperation and trust among scientists and the efficient use of scientific resources (Whitbeck 1995b). When relying on someone else's work, scientists normally assume that the research is valid. This is an important assumption to make because it would be an incredible waste of time to check every piece of research one uses for errors. When errors plague the research process, scientists cannot make this important assumption, they cannot trust each other, and they must waste time and energy checking for errors.

Many scientists do not view error as a serious crime against science, even though errors are more prevalent than fraud. A scientist who publishes a paper containing numerous errors might be viewed as incompetent but not as unethical. However, even though carelessness is not as serious an offense as dishonesty, it is still very important to avoid carelessness, since errors can waste resources, erode trust, and result in disastrous social consequences. Errors in applied research, medicine, and engineering, can cause a great deal of harm. A miscalculation in a drug dosage can kill dozens of people, and a defect in a bridge design can kill hundreds. Thus, although some errors can be treated as honest mistakes or incompetence, serious and repeated errors can be viewed as a form of negligence (Resnik 1996b). The proper response to discovering an error in one's published or submitted work is to admit the mistake and publish a correction, erratum, or retraction (Committee on the Conduct of Science 1994).

In discussing carefulness, it is important to distinguish between different types of errors in the research process. Experimental errors are those errors relating to the use of scientific instruments in collecting data. Every instrument produces some noise, distortions, and false readings, though some instruments are more precise and reliable than others (Kyburg 1984). It is standard practice in all scientific disciplines to take these errors into account when reporting data and results. Methodological errors include errors relating to the interpretation and analysis of data by statistical methods or the use of theoretical assumptions and biases in inference. Most scientists learn that statistical methods can yield very deceptive results and it is always important to use the statistical techniques appropriate to an area of research. The use (or misuse) of theoretical assumptions and biases can also lead to errors. For instance, astronomers who accepted Copernicus' heliocentric system struggled for years to make it fit planetary observations because they assumed that all heavenly bodies must move in perfect circles (Meadows 1992)...

Openness

Scientists should share data, results, methods, ideas, techniques, and tools. They should allow other scientists to review their work and be open to criticism and new ideas.

The principle of openness promotes the advancement of knowledge by allowing scientists to review and criticize each other's work; science's peer review system depends on openness (Munthe and Welin 1996). Openness prevents science from becoming dogmatic, uncritical, and biased. Openness also contributes to the advancement of science by helping to build an atmosphere of cooperation and trust in science and by enabling scientists to use resources effectively (Bird and Houseman 1995). Knowledge can be obtained more effectively when scientists work together instead of in isolation, when they share data, research sites, and resources, when they build on previous research, and so forth. Another reason in favor of openness in science is that secrecy undermines the public's trust in science (Bok 1982). When scientific activities are not open and accessible, people can begin to suspect that scientists are dishonest or untrustworthy, and the scientific profession can suffer many adverse consequences when public support for science erodes. Finally, insofar as all people have a moral duty to help other people, and the sharing of data and resources constitutes a form of help, scientists have a general, moral obligation to avoid secrecy in addition to their scientific duties to be open.

Although openness is a very important principle of scientific conduct, exceptions to this rule can be justified in some situations. For instance, many scientists avoid openness in order to protect ongoing research (Grinnell 1992). In order to protect her reputation, a scientist may not wish to share her data or results before her experiments are complete or before she has had time to think about her work in some detail. A scientist may also not wish to share her data, ideas, or results in order to guarantee that she receives proper credit, recognition, and compensation for her work (Marshall 1997). Once a study is complete, however, the need to protect ongoing research no longer exists, and the results should become a matter of public record, especially if the study is supported by public funds.

All of these arguments in favor of limited secrecy assume that scientists are justified in wanting to receive credit, recognition, or compensation. One might even argue that this kind of self­interest in science plays a key role in the advancement of knowledge (Merton 1973, Hull 1988). Science thus involves a trade-off between self-interest and egoism and cooperation and trust. Indeed, 565

since scientists are rewarded for making original contributions, and these contributions advance the goals of science, it is likely that science's reward system works like an “invisible hand” for the benefit of science: individual scientists may unwittingly contribute to the overall good of science by simply seeking to accomplish their own, personal goals, e.g. prestige and respect (Merton 1973). Most scientists would agree, I think, that secrecy should be the exception rather than the rule in research. However, one might argue that scientists may sometimes have other obligations that override their obligations to science. For instance, scientists who work for private companies may be obligated to keep trade secrets (Bok 1982, Nelkin 1984), and scientists who do military research are obligated to protect classified information (Nelkin 1972, Bok 1982). Thus, openness, a scientific value, may conflict with business and military values. These issues raise many difficult questions, which I will return to later.

Freedom

Scientists should be free to conduct research on any problem or hypothesis. They should be allowed to pursue new ideas and criticize old ones.

Great battles in the history of science have been fought over this principle. The struggles of Galileo, Bruno, Vesalius and the Soviet geneticists all attest to the importance of freedom in science. The principle of freedom promotes the attainment of scientific goals in several ways. First, freedom plays an important role in the expansion of knowledge by allowing scientists to pursue new ideas and work on new problems. Second, intellectual freedom plays an important role in nurturing scientific creativity (Kantorovich 1993, Shadish and Fuller 1993). Scientific creativity stagnates in oppressive, authoritarian, overly structured environments. When societies attempt to limit scientific research or direct it in certain areas, they risk undermining science itself (Merton 1973). Third, freedom plays an important role in the validation of scientific knowledge by allowing scientists to criticize and challenge old ideas and assumptions. Freedom, like openness, helps science from becoming stagnant, dogmatic, and biased (Feyerabend 1975). For instance, during this century genetics in the Soviet Union stagnated because Soviet geneticists were not allowed to challenge Lysenko's ideas about heredity (Joravsky 1970). Finally, we should also note that morality provides a rationale for freedom in research: freedom of thought, expression, and action imply freedom in inquiry.

Although the principle of freedom is crucial to science, one might argue that minor restrictions on freedom can be justified under certain conditions. In order to understand limitations on scientific freedom, we should distinguish between restrictions on actions, restrictions on funding, restrictions on publication, and restrictions on thought and discussion. It is important to understand these distinctions because they have different moral and ethical ramifications. First, most types of research involve actions by scientists and these actions can be restricted in order to prevent scientists from harming people or violating their rights. Even the strongest defenders of autonomy recognize that my right to do as I please stops at your nose. Thus, there are sound moral reasons for not allowing scientists to perform research that harms human subjects or violates their rights to autonomy. Most scientists would not consider protocols for the use of human subjects in research to be a significant or worrisome restriction on scientific freedom. Second, most scientific research requires a great deal of money that scientists obtain from government agencies, businesses, private foundations, universities, or the military (Dickson 1984). Agencies allocate their funds according to the demands of their constituencies: businesses fund research in order to generate profits, government agencies must answer to Congress and the public, and so on. Given these political and economic realities, it is frequently the case that funding decisions restrict research: unfunded research does not get done. For example, Congress's decision to terminate the Super Conducting Super Collider stopped many proposed experiments in high energy physics (Horgan 1994). Although these experiments may be conducted at some later time, Congress effectively put a great deal of research on “hold.” Should we consider the failure to obtain funding for research to be a significant restriction on scientific freedom? Probably not. Although these funding decisions hamper research, scientists cannot legitimately claim to have a “blank check” to fund their pet projects. Funding is a privilege, not a right. Scientists who fail to obtain funding are still free to discuss their ideas or pursue funding at some later time. Although it is important for societies to fund scientific research in order to create an environment that nurtures

scientific creativity, the failure to fund a particular scientific project does not cause significant damage to this environment.

On the other hand, some restrictions on research should be taken very seriously and can cause significant harm to science. During the height of Lysenkoism in the Soviet Union, scientists were not allowed to do research that would challenge Lysenko's views, they were not allowed to publish papers that challenged Lysenko, and they were not allowed to teach or even discuss views that contradicted Lysenkoism, such as Mendelian genetics. Since censorship, moratoriums, and other more severe limitations on the discussion of scientific ideas can have a detrimental effect on science and violate basic rights and liberties, we have good reasons for avoiding these kinds of restrictions on research. However, even these more serious limitations on science can be said to be justifiable under dire conditions. For example, one might argue that research can be censored to protect national security or that some kinds of research, such as studies on cloning human embryos, should be banned in order to prevent adverse social consequences. Thus, the issue of freedom of inquiry often requires scientists and society to balance the advancement of knowledge against other social goals (Cohen 1979).

Credit

Credit should be given where credit is due but not where it is not due. I have already alluded to the principle of credit in discussing secrecy and openness in science.

Although this principle does not directly promote the advancement of knowledge or science's practical aims, it is justified insofar as it motivates scientists to conduct research, it promotes cooperation and trust, and it insures that the competition for science's rewards will be fair (Hull 1988). Rewards in science include recognition, respect, prestige, money, and prizes. When a principle of credit does not operate in science, scientists may be less motivated to do research, and they may be reluctant to share information since they may be afraid that their ideas will be stolen. Credit also plays an important role in punishing scientists or allocating blame. If a piece of research is flawed, one needs to know who is responsible for it, so that the errors can be corrected or the person(s) can be punished. Thus, responsibility and credit should be viewed as two sides of the same coin: a person should be given credit for a piece of research only if they can take responsibility for it (Kennedy 1985). Finally, credit can be justified on moral grounds as well: standards of fairness imply that all people, including scientists, should be given just rewards for their contributions and effort.

Plagiarism and honorary authorship represent two opposing types of unethical conduct in credit allocation. Plagiarism occurs when someone falsely represents another person's ideas as his own through irresponsible citation, attribution, or paraphrasing. Plagiarism embodies a failure to give credit where it is due. Plagiarism can also be viewed as a form of dishonesty, since plagiarizers make false or misleading statements pertaining to authorship (PSRCR 1992). On the other extreme, sometimes scientists grant honorary authorship to a person who has not made a significant contribution to a paper (LaFollette 1992). Honorary authorship may be granted in order to compensate a lab director or senior researcher, help a friend or colleague, or add some prestige to a paper. Honorary authorship is unethical because it grants credit where it is not due. Although most scientists agree that plagiarism and honorary authorship are unethical, there is less agreement once we move away from these two extremes. How much must a person contribute to a piece of research in order to receive credit? Should authors take credit for different sections of papers or different parts of a research process?..

Education

Scientists should educate prospective scientists and insure that they learn how to conduct good science. Scientists should educate and inform the public about science.

Education includes recruitment, formal instruction, training, and mentoring. A principle of education is important in science since the profession will grind to a halt if it cannot recruit, train, and educate new members. Recruitment is important in attracting new people to the scientific profession. Although formal science instruction also occurs in high school and lower grades, scientists do not usually become actively involved in this part of science education. However, scientists do have a duty to offer their suggestions and input into science education at these lower levels, and they have an obligation to educate people who intend to teach science at the K-12 level. Training is a type of 567

informal instruction that involves imitation, practice, and apprenticeship. It involves the acquisition of various skills and an intuitive understanding of scientific practice. A well-trained scientist has tacit knowledge of her subject that extends far beyond what she can learn in textbooks or lectures (Kuhn 1970, 1977, Kitcher 1993). Scientists also have an obligation to support the effort to educate the general public through popular books, magazine articles, television appearances, and so on. This is an important part of science education as well, since the general public needs to gain an understanding of science. Since science depends on public support, science benefits when the public has a sound understanding of science and suffers when ignorance about science abounds.

Although education is important in science, different scientists may decide to participate in science education in different ways. Some scientists may focus on graduate education, others may focus on undergraduate education. Some scientists may mentor many students, others may not. Some scientists may become actively involved in recruitment, others may not. Some scientists may write popular works or make media appearances, others may not. Some scientists may decide completely to opt out of science education in order to pursue academic, military, or business research. As long as enough scientists are involved in education, the scientific community can afford the luxury of having some pure researchers.

Social Responsibility

Scientists should avoid causing harms to society and they should attempt to produce social benefits. Scientists should be responsible for the consequences of their research and they should inform the public about those consequences.

The general idea behind this principle is that scientists have a responsibility to society (Lakoff 1980, Shrader-Frechette 1994). Scientists should not conduct inquiry with the attitude that someone else can worry about the consequences of research or science’s impact on society. Social responsibility implies that scientists have an obligation to conduct socially valuable research, to participate in public debates, to give expert testimony (if asked), to help make science policy, and to debunk junk science. Some scientists may reject the notion of social responsibility on the grounds that scientists should pursue knowledge for its own sake and should let politicians and the public deal with the social consequences of research. Responsibility for the social impacts of science falls on the media, politicians, and the public, not on scientists. Although this attitude has become less common in recent years, it still has a significant enough influence that it is worth refuting.

There are several reasons why scientists should take responsibility for the social impacts of research. First, although scientists cannot be held responsible for the unanticipated consequences of research, scientists can be held responsible for the consequences that can be anticipated. Second, scientists are also members of society who have moral duties toward other people, such as beneficence, nonmalificence, and utility. Third, scientists have professional duties to promote benefits and avoid harms. As professionals, scientists are expected to produce socially valuable goods and services and they are accorded a great deal of authority, responsibility, and trust. Social responsibility recognizes and honors this public trust (Shrader-Frechette 1994). Finally, social responsibility benefits science by increasing the public’s support for science: socially responsible science promotes public support for science; socially irresponsible science undermines public support for science (Slakey 1993). By serving society, scientists can fight negative images of socially irresponsible scientists, e.g. Mengele, Frankenstein, and replace them with positive ones (Nelkin 1995).

However, although scientists have a duty to be socially responsible, this duty should be carried out with discretion. As we noted earlier, scientists should not disclose information prematurely; information should be validated by other scientists through the peer review process before it is disclosed publicly. When research is disclosed prematurely, two types of bad consequences may occur. First, people may be harmed. For instance, a person might try a new kind of treatment if scientists say that it works, even if it has not been thoroughly tested, and she might suffer harmful side effects from the treatment. Second, science’s image can be harmed. When the public learns of an important discovery or cure that upon further inspection turns out to be a sham, they will be disposed to view scientists as incompetent or irresponsible. (The cold fusion debate serves as an unfortunate example of this effect.) Third, premature disclosure of scientific information can disrupt 568

the process of credit allocation in science, but the general public is usually not qualified to assess priority disputes (Longino 1990). (A concern for priority probably helped to motivate the premature disclosure of cold fusion research.) Researchers who present their results to the public may receive undeserved recognition and credit if it turns out that some more conscientious researchers have reported the same results but have submitted their work to a scientific journal.

Finally, we should note that the principle of social responsibility is like the principle of education in that some scientists may (at times) decide to set aside social responsibilities in order to pursue other goals. Some scientists may wish to be less outspoken than others; some may choose careers that generate few results with social implications, and so forth. Social responsibility is a shared obligation that can be met by different scientists at different times.

Legality

In the process of research, scientists should obey the laws pertaining to their work.

....all people, including scientists, have moral obligations to obey the law. Moreover, science may suffer great damage when scientists disobey the law: scientists may be arrested, equipment may be confiscated, funding may be denied, the public support for science may erode, and so forth. Laws pertain to many different aspects of research, including the use of hazardous and controlled substances, the use of human and animal subjects, the disposal of wastes, hiring practices, the appropriation of funds, and copyrights and patents (PSRCR 1992). Although scientists have strong moral and ethical duties to obey the law, this standard of conduct, like the other ones, can have exceptions. One might argue that sometimes scientists can break the law in order to gain important knowledge or benefit society. Throughout the history of science legal restrictions have impaired the advancement of knowledge. For instance, in Medieval Europe there were many legal restrictions on human dissections, and those who wanted to learn more about the human body had to conduct underground research. In Galileo's day, the Catholic Church imposed sanctions on teaching Copernicus' heliocentric astronomy. Although scientific civil disobedience can be justified in some instances, I should reiterate the point that burden of proof lies with those who would break the law (Fox and DeMarco 1990).

Opportunity

Scientists should not be unfairly denied the opportunity to use scientific resources or advance in the scientific profession.

A principle of opportunity can be justified on moral or political grounds: if all people in a society should not be unfairly denied opportunities, then scientists (as members of society) should also have these opportunities (Rawls 1971).

This principle can also be justified on the grounds that it promotes scientific objectives. Opportunity is like a principle of openness writ large because it opens the scientific community to new people and ideas. In order to overcome biases and dogmas and achieve objective knowledge, science needs to examine and consider a diversity of hypotheses, ideas, approaches, and methods (Kuhn 1977, Longino 1990, Solomon 1994). Although people from similar backgrounds might generate this epistemological diversity, it is more likely that people from different back-grounds generate the diversity of opinions required for the advancement of knowledge. Objectivity is more likely to emerge from the clash of different cultures, personalities, and styles of thought than from the consensus of like minds.6 This principle supports several important science policies. A great deal of government money these days goes to large scientific projects and to prestigious labs (Martino 1992). Scientists who work on smaller projects or at less prestigious places can thus be denied research opportunities. While there are some legitimate reasons for putting lots of money into large projects at the expense of smaller ones and for funding prestigious labs, science funding policy should not be so large-scale or elitist that many deserving scientists are denied research opportunities. A principle of opportunity implies that in funding science it pays to spread the wealth around.

Second, although women and minorities have made significant advances into science's professional ranks, there still are very few women and minorities among the scientific elite, i.e. Nobel prize winners, members of the NAS, full professors, etc. Indeed, there is some evidence that a type of “old-boy-network” or “glass ceiling” exists in science when it comes to career advancement and 569

prizes (Holloway 1993, Etzkowitz et al. 1994). Although there is nothing inherently wrong with relying on personal relationships to achieve career advancement, these relationships become a problem when they tend to exclude deserving people from a profession’s upper echelons. A principle of opportunity implies that scientists should recruit, employ, and reward underrepresented groups, such as women and minorities.

Third, the principle of opportunity also implies a general prohibition against discrimination in science, since discrimination can unfairly violate a person’s opportunities. Thus, scientists should not discriminate against colleagues or prospective colleagues on the basis of race, sex, national origin, nationality, age, or other characteristics not directly related to scientific competence (Merton 1973). This prohibition applies to a wide range of decisions scientists make in a professional context including hiring and promotion, admission, recruitment, resource allocation, and education. In addition to being unethical, many forms of discrimination are illegal. Although discrimination should be avoided in science, one might argue that some types of discrimination, e.g. preferential hiring, are justified in order to promote diversity in science or to correct for past injustices...

Mutual Respect

Scientists should treat colleagues with respect.

Although this principle can be justified on moral grounds, it can also be justified in that it is important for achieving scientific objectives: the scientific community is built on cooperation and trust, which will break down when scientists do not respect one another (Whitbeck 1995b). Without mutual respect, the social fabric of science unravels, and the pursuit of scientific aims slows down. The principle implies that scientists should not harm one another, either physically or psychologically, that they should respect personal privacy, that they should not tamper with each other’s experiments or results, and so forth. Although one might argue that some of the best scientists have not respected their colleagues (Hull 1988), the stereotype of the aggressive, mean-spirited, though highly successful scientist does not hold for science as a whole. While science can operate effectively when some people behave this way, I doubt whether it could operate effectively if all scientists behaved this way.

Efficiency

Scientists should use resources efficiently.

Since scientists have limited economic, human, and technological resources, they must use them wisely in order to achieve their aims. Although this principle seems somewhat trivial and obvious, it is still important in that many practices can be viewed as ethically questionable because they waste resources. Several practices related to publication can be viewed as unethical because they are inefficient. The “least publishable unit” is a phrase coined by William Broad (1981) that refers to the smallest piece of writing that can be published. Research that could probably be reported in one paper is sometimes divided up into three, four, or five papers. Additionally, scientists also sometimes use the same results for several different papers simply by making some minor changes in writing or presentation. Both of these practices can be regarded as unethical because they waste the scientific community’s resources (Huth 1986). It is not difficult to understand why scientists engage in these wasteful activities, since tenure and promotion committees tend to stress quantity over quality in assessing publication records.

Respect for Subjects

Scientists should not violate rights or dignity when using human subjects in experiments. Scientists should treat non-human, animal subjects with appropriate respect and care when using them in experiments.

This principle can be justified on moral grounds. If we hold that human beings have inherent moral dignity and some basic rights, then scientists should not violate these rights and dignities when using human beings in experiments (Jonas 1969). If we hold that non-human animals also have some moral standing, then scientists should treat animal subjects with appropriate respect and care (LaFollette and Shanks 1996). Furthermore, since both of these principles reflect the public’s concern for the ethical treatment of research subjects, they can be justified in that they help secure public 570

support of science: scientists who fail to demonstrate proper respect for human and animal subjects may incur the public's wrath. Since many societies have laws protecting human and animal subjects, scientists also have legal obligations pertaining to research on humans and animals. This principle needs to be spelled out in more detail, of course, since it can be interpreted in different ways, depending on how we understand notions like “treat with respect and care,” “human rights and dignity,” and so forth...”

David Resnik is a bioethicist at National Institute of Environmental Health Sciences, USA.

Control questions

1. How are science and ethics related?

2. What is scientific ethics?

3. What is scientific honesty?

4. What problems arise from a mismatch between resources and opportunities?

5. Impact of science and technology on human lives.

6. What challenges of modern science do you know?

7. Do scientists need a special ethicl approach?

8. What are normative, applied and meta-ethics?

9. How do science and technology impact on humans' lives?