Viewpoint

Apr 2015

Making Ethical Progress without Ethical Theories

Dien Ho, PhD
AMA J Ethics. 2015;17(4):289-296. doi: 10.1001/journalofethics.2015.17.4.spec1-1504.

The Couple

A young couple plans to start a family. A month into her pregnancy, the mother-to-be realizes that balancing work, education, and starting a family will be more than she is ready to handle. She talks to her partner, and he expresses his desire for her to carry the child to term and reminds her that she did commit to the pregnancy. Should she be allowed to terminate the pregnancy against her partner’s wishes [1]?

The Problem

The typical way ethicists tackle an applied question like the one above is by deploying a favored ethical theory and finding out what answer falls out after plugging in the “values for the variables.” Most health care ethics textbooks and courses subscribe to this approach and design the pedagogy by first introducing a plethora of ethical theories before tackling applied ethical issues such as patient confidentiality, euthanasia, abortion, and so on. This approach relies on an implicit belief that learning ethical theories is a necessary if not sufficient part of solving everyday ethical problems. The popular text book (that I use in my courses) Ethical Issues in Modern Medicine by Bonnie Steinbock, Alex London, and John Arras, for instance, outlines fifteen ethical and meta-ethical theories in its introduction [2]. These theories range from moral nihilism, to rule-utilitarianism, to feminist ethics, to casuistry. But this crash course on ethics and meta-ethics does not tell us which of the competing ethical and meta-ethical theories we should use to figure out what one ought to do. The gap between theories and application exists because ethical theories often make competing recommendations and we do not know which theory we ought to use. In the example above, a utilitarian might conclude that, in the long term, terminating the pregnancy would maximize utility for the most people, and therefore the mother-to-be ought to do it. A deontologist, on the other hand, might disagree and conclude that the mother-to-be has a duty to keep her promise. It seems that unless we first settle on the “right” moral framework, we can never solve any ethical problems.

Steinbock et al. clearly recognize the pedagogical and conceptual gulf between theories and practice, and they suggest that having so many ethical theories contributes to

an extraordinarily complex moral reality. Still, in deciding what to do, one must decide when utilitarian considerations should prevail, or when one ought to adhere to absolutist principles; when to appeal to principles and when to seek guidance in virtues; when to abide by universal, impartial considerations and when to concentrate on personal relationships and feelings [3].

Steinbock et al. echo what many philosophers have suggested: we need to consider all the ethical frameworks and attempt to achieve coherence among our moral beliefs.

Simply maintaining coherence is not difficult, but it does not resolve moral questions. When presented with merely two conflicting moral beliefs, coherence demands only that we reject at least one; it does not dictate which moral belief we ought to reject so long as we reject at least one. Imagine the many possible coherent outcomes when an ethical problem presents multiple mutually incompatible propositions. Coherence will not suffice in helping us determine what we ought to do.

Moreover, the demand for coherence is itself a moral value system, in which consistency is prioritized above other possible concepts of the right or good. It might prove psychologically comforting to strive for greater coherence in one’s reasons and actions, but a recommendation based on coherentism is not necessarily right without a prior assumption that coherentism is the correct moral theory.

Ethical Theories Are Secondary in Resolving Everyday Moral Disputes

The idea that we need ethical theories to tell us what we ought to do might strike most laypersons as awkward and artificial; e.g., consider how odd it sounds to decide whether one ought to continue a pregnancy by seeing if it maximizes utility [4]. The fact is we frequently resolve ethical disputes in our daily lives, often paying little attention to ethical theories. To be sure, when challenged, many of us justify our moral choices on the basis of some ethical theories, but the appeal to theories is often a form of post hoc rationalization. Psychological studies done by Jonathan Haidt support this view [5]. Of course, the fact that we use moral theories to rationalize our choices post hoc does not entail that moral theories are merely window dressing. However, if one goal of the study of ethics is to provide a relatively accurate description of our moral reality (e.g., how we reason morally), then surely it should account for the glaring absence of appeal to ethical theories in everyday moral reasoning. To suggest—as introductory ethics courses do—that moral reasoning must flow down from broad theories to specific moral recommendations is to fly in the face of how we actually decide what we ought to do. It makes one wonder whether such a model in fact concerns our ethics [6].

A comparison with empirical sciences can perhaps illuminate the relationship between deep ethical theories and ordinary moral discourse. In science, researchers rarely engage questions of why certain fundamental laws are true. Instead, they deploy heuristics and “midlevel” theories or generalizations to tackle their problems. For instance, a biologist might rely on the rules of Watson-Crick base pairing that the purine adenine always pairs with the pyrimidine thymine and the pyrimidine cytosine always pairs with the purine guanine. The regularity of the pairing can be explained by understanding hydrogen bonding, but this deeper explanatory step is rarely, if ever, needed in conducting everyday research in genetics. Typical moral decision making relies on heuristics much like the rules of Watson-Crick pairing: these heuristics allow us to identify which moral beliefs we should hold but they float above the explanatory project (ethical theorizing) of trying to figure out why these beliefs are true.

So far I have only sketched a negative thesis; that is, when we make moral decisions in our daily lives, we rarely appeal to broad ethical theories. Nevertheless, when we attempt to convince ourselves and others that a particular moral decision is appropriate, we need to find some way to justify it. Appealing to a specific ethical theory for justification only means that we then need to provide further justifications for choosing that particular ethical theory. What is required to justify our moral choices must lie beyond specific ethical theories. In order to identify the appropriate justificatory source, let us reexamine why we need ethics in the first place.

Why Ethics and the Default Principle?

We often appeal to ethics to adjudicate disagreements such as those between two parties whose interests cannot both be satisfied, for instance, an asymmetric moral disagreement in which A wants to do X and B wants her not to [7]. Of course, we can adjudicate disagreements in a number of ways—we could resort to violence, drawing straws, bribery, guilting or shaming one party or another, and so on—but when we choose to resolve disagreements by appealing to ethical considerations, we agree to allow reasons to be the ultimate arbiter of what we are permitted to do. This broad characterization of ethics as rational conflict resolution generates some interesting results. For starters, it tells us that one party can only impose its will on another by appealing to reasons. Moreover, in the absence of a compelling reason not to, individuals are permitted to do what they wish [8]. Call this the default principle (DP).

DP follows a recognition that we offer reasons to ethically justify our actions only when these actions affect another person. In an asymmetric disagreement, there are logically two possible ways the use of reasons might resolve the conflict: either A offers a reason for why she is permitted to X or B offers a reason for why his prohibition of A doing X is warranted. Of course, to say that A must offer a reason for why she is permitted to X is to say that unless a reason is offered, she cannot X. But surely this is absurd. Imagine if A is all by herself on a desert island. She does not need to offer a reason to morally justify her actions before she does them. Her actions only require justifications when they run into the interests of another party. Thus, in an asymmetric disagreement, B has to offer a reason for why his prohibition of A doing X is warranted; which is to say, unless there are reasons to think otherwise, A gets to do what she wants (i.e., the DP). The insight that ethics arise only when interests collide is behind much of the Enlightenment philosophers’ appeal to the state of nature [9].

The DP has what might be called a “permissive bias”: the burden of proof always rests on the side that wishes to restrict another’s autonomy. Thus, regardless of whether A can articulate why she should be permitted to do what she wants, as long as B fails to supply a good reason why she should not, A is permitted to do what she wants. Moreover, suppose for a moment that a genuine impasse exists between the two parties; e.g., the disagreement rests on a difference of values—something that is typically based on subjective preferences. By definition, no reason exists to limit A’s autonomy; hence, A must be permitted to do as she pleases. Both of these results—B fails to supply a reason for A’s not doing as she wishes, so she does as she wishes, and A and B differ in their values, so rational argument is a hopeless task, and A gets to do as she wishes—reflect a permissive bias. Indeed, they echo a typical liberal slant that conservative ethicists have complained about. If the permissive bias is in fact a product of liberal values, then their criticism is correct. One would need to provide an argument to support the allegation that the permissive bias derives from liberal values, and, given the fact that they are values, arguing to establish their exclusive source in liberal thinking appears to be an impossible task.

Our permissive bias stems not from the acceptance of liberal values or, for that matter, any specific ethical framework. It comes from the very ground rules of appealing to reasons to resolve conflicts; that is, the need to provide a reason to morally justify one’s actions only arises when it compromises the interests of another party. In this respect, the permissive bias transcends ethical frameworks, and it helps us avoid the earlier problem of how we go about justifying a moral decision without locating the “right” ethical theories, at least in the cases of absence of arguments.

The scenario outlined at the start of this essay provides an example of how the default principle can guide us through an ethical conundrum. Given the default principle, the side that wishes to override the woman’s desire to terminate her pregnancy must supply reasons to justify that position. If, after they are evaluated, the reasons offered remain inconclusive, she must be permitted to terminate. Her failure to offer any reasons to defend her preferences does not undermine her right to proceed. In this sense, the default principle provides a permissive bias without assuming liberal values. Of course, we still need some way to evaluate reasons offered that does not presuppose a specific ethical framework. In the next section, I will derive a justification for appealing to arguments by parity from the default principle.

Generating Arguments by Parity from the Default Principle

The default principle justifies a certain kind of moral coherentism by way of its use of arguments by parity. In determining what moral attitude we should take towards a particular practice, it is common for us to identify the moral belief we have towards similar practices. For instance, the argument in favor of legalization of same-sex marriages centers primarily on the idea that same-sex and heterosexual marriages are very similar and therefore equally acceptable. The proponents of legalizing same-sex marriages have argued that there are fundamentally no reason-based differences between the two types of marriages that are relevant to making decisions about whether it is ethical. That is, the standard reason given for allowing marriage only between members of different sexes is that the purpose of marriage is procreation, but infertile couples are allowed to get married; thus, same-sex couples’ inability to procreate without external assistance cannot justify prohibiting same-sex marriages. The force of this argument is that logical consistency demands that either we reject a deeply entrenched belief or change our minds and assign the same moral attitude towards the practice in question.

An argument by parity does not need to rely on some broad ethical framework. A commitment to logical consistency is all that is necessary to appreciate the argumentative force, and this commitment comes not from a blind worship of logical consistency but from the default principle. Suppose there are no practical differences between X and Y that would cause us to judge them in ethically different ways. To prohibit X but not Y would be to prohibit X without a reason (since if there were a reason, it would suffice to prohibit Y as well). A rejection of a sound argument by parity is thus a rejection of the default principle.

Default Principle Limitations

The default principle has some limitations.

The requirement of some shared beliefs. Arguments by parity require there to be some shared moral beliefs to provide sufficient purchase to make the analogy work. If two individuals share no moral beliefs, arguments by parity would be useless. Exactly how far we can go with only a few shared beliefs is an important question that demands empirical answers. My suspicion is that we can make a great deal of progress with a handful of shared basic moral beliefs (e.g., all else being equal, one ought not choose a course of action that causes more suffering).

The possibility of expanding, rather than closing, gaps in belief. Another serious problem with arguments by parity is that in the face of logically incompatible moral beliefs one always has at least two choices: accept the new moral belief and abandon the firm belief that one once held, or retain the firm, long-held belief and reject the new belief. In other words, logical consistency merely insists that one cannot stay at a particular doxastic spot; it does not tell one whether to move closer to or further from the beliefs of the other disputant. If the anchoring belief is firm enough, the cost of revising one’s web of beliefs would be so high that the agent would abandon the new belief as the path of lesser resistance. But of course psychological ease is no guide of morality or rationality [10].

The inability to compel one’s opponent to move morally closer (as opposed to further) is an obvious possible outcome of utilizing arguments by parity. Nevertheless, I suspect that this feature is less a dialectical defect than a variance of Hume’s problem of justifying rationality, to wit, there is no absolutely secured foundation upon which to attempt to justify rationality without using reason—the function we are attempting to justify. All reasoning must involve constant adjustments among competing beliefs in a sort of reflective equilibrium [11].

The permissive bias does not apply when autonomy is not restricted. There will be disputes in which no one side is clearly restricting the autonomy of another. For instance, if two individuals argue over which of them should receive a liver transplant, their statuses vis-à-vis the default principle would be symmetric. This is a symmetric disagreement. In this respect, the permissive bias is irrelevant, and we must evaluate the relative merits of both of their positions. This shortcoming is not unexpected given the fact that the permissive bias and subsequent default principle come from accepting the ground rule that the imposition of will can only be permitted when there is reason to justify it. When no party is attempting to impose his or her will on another, the asymmetry necessary to generate the permissive bias is absent.

Implications of the Default Principle

The diversity of ethical frameworks that participants in a moral dispute hold tells us that, if solving a moral dispute or question means deriving a proper course of action on the basis of the “right” moral framework, no moral problem can be solved. Yet we are often capable of solving moral problems. A great deal more research is needed to get a better understanding of how we make moral judgments, but a few things should be apparent. Firstly, an understanding of the nature of ethics as consisting of appealing to reasons to solve conflicts generates some important justificatory and methodological consequences that have been for the most part unnoticed. Agreeing to settle disputes by recourse to reason provides us with a rough roadmap of how to proceed without presupposing a particular moral framework. This allows us to avoid the justification problem that plagues those who subscribe to the orthodox “trickle-down” view.

The efficacy of the default principle to resolve ethical disputes without necessitating recourse to a particular moral framework suggests that we should reevaluate the way clinical ethics is taught and conducted. The standard pedagogical approach of introducing a plethora of ethical theories as the beginning of training in ethics is unnecessary and most likely unhelpful in providing clinicians with the tools to tackle real-life ethical problems. Teaching students how to identify asymmetric arguments, understand their structures, and use practical reasoning skills, and familiarizing them with mid-level moral principles, are far more useful. Likewise, the hesitation (e.g., the view of the American Society of Bioethics and Humanities) about ethics consultation services that make clear recommendations is unwarranted if the hesitation stems from worries that various ethics consultants might impose their own ethical assumptions and values on those who do not share them. The default principle tells us that, as long as we agree to resolve our moral conflicts by appealing to reason, there are methodological principles that transcend particular ethical frameworks.

Read More

References

  1. This is the core conflict in Planned Parenthood of Central Missouri v Danforth 428 US 52 (1976).

  2. Steinbock B, London AJ, Arras JD. Ethical Issues in Modern Medicine: Contemporary Readings in Bioethics. 8th ed. New York, NY: McGraw-Hill; 2013.

  3. Steinbock, London, Arras, 35.

  4. We often use midlevel moral generalizations (e.g., do what maximizes the benefits of most people) to help us reason morally. These midlevel moral generalizations are not ethical theories in the traditional sense of providing a foundation for morality. Rather, they are rough rules of thumb that we do not hesitate to abandon when we intuit that using them in a particular situation is inappropriate. In other words, these midlevel rules do not serve the deep explanatory or justificatory role of any grand moral views.

  5. Haidt J. The emotional dog and its rational tail: a social intuitionist approach to moral judgment. Psychol Rev. 2001;108(4):814-834.
  6. For any philosophical investigation, we often have to strike a fine balance between the descriptive component (the way X is) and the prescriptive component (the way X ought to be). If a study merely provides a descriptive component, we would not be doing philosophy but, say, sociology. Likewise, if it is purely prescriptive, then one wonders what the project has to do with the initial subject matter. Ethics is unique in that all of us are intimately familiar with it. We are, to borrow a phrase from linguistics, “competent speakers” of the language of ethics. An analysis that tells us we are radically wrong in how we reason morally would be akin to a linguist who studies a native French speaker and concludes that she actually doesn’t speak French.

  7. It is an asymmetric disagreement because the object of B’s want is A’s refraining from X. If A wants X and B wants Y and X and Y cannot be jointly satisfied, then we have a symmetric disagreement.

  8. We must accept this statement for now without listing explicit criteria for what makes an argument “compelling” in order to further develop the thesis that the default principle (DP) can function as a nontheory-based approach to ethical problem solving. It is as though all we are saying here is that the person or team with the higher score wins; we have yet to say what constitutes “scoring.”

  9. The formal argument would look like this:

    1. When there is an asymmetric moral conflict either A has to justify why she is allowed to X or B has to justify why A is not permitted to X.
    2. “A has to justify why she is allowed to X” entails that, unless there are reasons to justify why she is allowed to X, she is not allowed to X.
    3. “B has to justify why A is not permitted to X” entails that, unless there are reasons to justify why B is allowed to prohibit A from doing X, B is not allowed to prohibit A from doing X.
    4. B’s not being allowed to prohibit A from doing X entails that A is permitted to X.
    5. “Unless there are reasons to justify why A is allowed to X, she is not allowed to X” entails that, if A is all by herself, she is not allowed to X unless she can offer a reason to X.
    6. It is not true that, if A is all by herself, she is not allowed to X unless she can offer a reason to X.
    7. Thus, it is not true that, unless there are reasons to justify why she is allowed to X, she is not permitted to X (5 and 6).
    8. Therefore, when there is an asymmetric moral conflict, (3) is true (1 and 7).
    9. Therefore, when there is an asymmetric moral conflict, A is permitted to do X unless B justifies why A is not permitted to X (8 and 4). Conclusion (9) is just the DP.
  10. The agent may choose in the other direction—“I see that my belief that straight marriage should be permissible but gay marriages should not is logically inconsistent, so I am going to abandon my long-held belief that straight marriages are permissible.” In this case, the agent retreats to his or her long-held belief. The fact that some counties in the US have stopped issuing marriage licenses altogether because they don’t want to issue marriage licenses to gay couples shows that the retreat option is not that rare or weird.

  11. Rawls J. A Theory of Justice. Cambridge, MA: Harvard University Press; 1971:46-53.

Citation

AMA J Ethics. 2015;17(4):289-296.

DOI

10.1001/journalofethics.2015.17.4.spec1-1504.

Acknowledgements

I wish to thank Joe Demasi, Susan Gorman, Jane Leo, and Ken Richman. I also wish to extend my deepest gratitude to the editors at AMA Journal of Ethics; particularly, Faith Lagay, Hannah Kushnick, and Shirley Martin.

The viewpoints expressed in this article are those of the author(s) and do not necessarily reflect the views and policies of the AMA.