The term scientific ethics may refer to the ethics of doing science (Is one free to inject unwilling subjects with a pathogen so as to gain valuable scientific insights? or What role should animal experimentation play in biology?). In that sense, scientific ethics is a branch of applied ethics. The term may also refer to whether or not the methods and assumptions of science can be applied to the subject matter of ethics. The present entry is concerned with scientific ethics in the second sense—Can there be a science of norms?
Scientific ethics in this sense is often argued to be an oxymoronic term. Science deals in empirical facts, discovering what is the case, while ethics deals in normative matters, uncovering what ought to be the case. A scientific ethics would thus commit the naturalistic fallacy of confusing what is with what ought to be. Historically speaking, however, this distinction is as much the exception as the rule. Premodern ethical systems, such as the virtue theories of Plato and Aristotle, did not couch the debate about what ought to be done in a way that made facts and norms non-overlapping magisteria (Gould 2002). To understand the relationships between science and ethics, it is useful to begin with some working definitions.
Defining Ethics and Science
Ethics is divided into descriptive, normative, and metaethics. Descriptive ethics is the study of empirical facts related to morality, such as how people think about norms, use norms in judgment, or how the norms themselves evolve. There is a rich tradition of organizing knowledge about these things scientifically, ranging from the field of moral psychology (focusing on how people reason about norms) to some forms of sociobiology (studying how norms arose on evolutionary timescales).
Normative ethics is an attempt to organize knowledge about what human beings ought to do or intend, or what kind of people they ought to be—it provides guidance and advice. The three major versions of normative ethics are virtue theory, utilitarianism, and deontology. A virtue theoretic approach, such as found in Aristotle, focuses on the nature of persons or agents. Are they flourishing—functioning effectively as human beings—or failing to flourish? Virtue theorists focus on states of character (virtuous or vicious) and how they affect the ability to live the best human life. Utilitarians, such as Jeremy Bentham (1748–1832) or John Stuart Mill (1806–1873), focus instead on the consequences of an action, rather than the character of the person committing it. Specifically they look at the amount of happiness caused (or unhappiness prevented), with the happiness of all counting equally. Deontologists, such as Immanuel Kant (1724–1804), focus on the nature of the action itself rather than its consequences. Certain actions express appreciation for, and are done in accordance with, the demands of duty, respecting that which is the foundation of morality: rationality and autonomy.
Metaethical questions consider the scope and nature of moral terms. Do ethical terms such as good and bad refer to facts about the world, or merely to states of emotion in people making judgments? Does ethics constitute knowledge or not; is ethical knowledge illusory? What is the structure of ethical arguments? It is less controversial that science may influence metaethical positions (although that position is also debated) than that there can be a science of normative ethics.
Science likewise comes in three forms. In the weakest sense, a science is an organized body of knowledge. If this is what is meant by science in relation to ethics, then a science of ethics certainly exists. The major moral theories just mentioned are attempts to bring some organization to what is known about morality.
Normally, though, science means something stronger and refers to a set of epistemological canons that guide inquiry. In one form, these canons are called methodological naturalism: the methods of inquiry used by an empirical science such as physics or biology. These include observation of the world, hypothesis formation, intervention and experiment, iterative formation and improvement of a theory, and more. Such activities are constitutive of the scientific method. If such methods can produce knowledge about norms, then a science of ethics is possible.
An even stronger form of science is ontological naturalism: Only those entities, events, and processes countenanced by the existing sciences may be used in theory construction. Methodological naturalism is a weaker form of science than an ontological naturalism. Consequently the possibility of an ethics grounded in ontological naturalism is more controversial.
In the weakest sense, ethics is a science if it can be organized into a coherent body of knowledge; in the moderate sense, ethics is a science if it can use the traditional epistemological canons of science to gain moral knowledge; and in the strongest sense ethics is a science if in addition to using the methods of science it also makes reference only to the entities and processes accepted by the extant, successful natural sciences. Only nihilists or radical moral particularists (those who contend that moral theory is so situation driven that general principles are impossible) would deny that there could be a science of norms in a weak sense. The moderate position is more controversial. Some would contend that moral knowledge is not gained using the empiricist methodology of the scientific method. For example, Kant's deontological theory does not require that humans reason empirically about morality; rather he maintains that they can know what they must do a priori independent of any particular experience. The strong position is the most controversial: Whether a normative theory can exist that differs neither in scope or content from the empirical sciences is debatable.
The argument offered most often against the possibility of scientific ethics in the moderate or strong senses is the naturalistic fallacy. First articulated by David Hume in A Treatise of Human Nature (1739), the naturalistic fallacy occurs when one moves from a list of empirical premises to a conclusion that contains a normative component. Hume is "surprised" when authors writing about ethics who were previously reasoning in the "usual way" suddenly begin to substitute "oughts" in places where before only the copula "is" had been present (Hume, Book III, Part I, Section I, Paragraph 24). Hume appears to point out a flaw in attempts to reason from the empirical to the normative—one will make reference to an unexplained term in a conclusion that was nowhere present in the empirical premises of the argument. Such an argumentative structure is invalid; the truth of the premises does not guarantee the truth of the conclusion. G. E. Moore advanced a similar argument early in the twentieth century when he argued that naturalized ethical systems fall prey to the open question argument. After one has identified normativity with a natural property such as avoidance of pain, for example, one can still meaningfully ask whether it is good to avoid pain. This means that utilitarians have not successfully reduced goodness to the natural property avoiding pain.
Whether or not the naturalistic fallacy and the open question argument provide in principle rationales against a moderate or strong scientific ethic is itself an open question. There are several possible responses. For example, both arguments rely on an analytic/synthetic distinction (a distinction between sentences true by definition and sentences true because of the way the world is), and many philosophers think no such distinction exists (see Casebeer 2003a). In addition, Hume's argument applies only to traditional deductive and inductive arguments. It may well be, though, that the relationship between natural ethical facts and the norms they deliver is abductive; one may best explain—abduction is often called inference to the best explanation—patterns of certain facts by assuming that they are also natural norms. Finally the open question argument probably does not generalize; it really amounts to saying that the two ethical systems Moore examines (Spencerian evolutionary ethics and hedonism) are not good natural ethical theories, and all but partisans would agree.
Why Scientific Ethics
Given disagreements about whether a scientific ethics in the moderate or strong sense is possible, why might people want such a thing? There are four possibly interrelated reasons. First science seems to some to have undermined traditional ethics, and hence human beings should use science to re-create ethics on firmer foundations. Second scientific ethics might be driven by concerns about the coherence of worldviews. Third scientific knowledge is the only real kind of knowledge. Fourth the sciences provide a prestige model, and in a highly scientific society people always try to imitate that which is of greatest prestige.
The first rationale may reflect a praiseworthy desire to reconsider long-standing issues in ethics from the perspective of contemporary science; for instance, what does contemporary cognitive science say about the existence of a free will, and what impact might this have on the conception of ethics? As another example, sociobiologists sometimes veer towards eliminativist extremes about the subject matter of ethics (morality is an illusion fobbed off on people by their genes). Strong scientific ethics thus might be a path to reconstruct what is purportedly illusory, whether it be a notion of agency compatible with the sciences or a scientific defense of the genuine objectivity of ethics.
The second rationale is closely related: Researchers may hold out hope that human knowledge can be unified. At the very least, they may ask that it be consistent across spheres of inquiry. Concerns about consilience can thus drive scientific ethics (Wilson 1975). The third and fourth rationales are strongly linked: If scientific knowledge is on a firmer footing than folk knowledge or nonempirical inquiry, then it is no wonder that funding and prestige would attach to scientific pursuits rather than not. Researchers in ethics may thus be attracted to the epistemic roots of science and the research support flowing from them. Sometimes this attraction leads to pseudoscientific ethics (just as it leads to pseudoscience), as in, for example, the work of Madam Vlabatsky's theosophical scientific ethics or in the eugenics movement. A thoughtful scientific ethics rejects pseudoscience and the pseudoethics that might follow.
Of course science advances, changing as time passes. Will attempts to connect science and ethics undermine the certainty some strive for in morality? They may, but this is no objection to the enterprise; it might be that the best one can hope for even in ethics is something like the best guess hypothesis offered by the practicing scientist.
Examples of Scientific Ethics
What might a moderate or strong scientific ethics look like? Herbert Spencer (1820–1903) claimed to offer such a theory in his work; he derived an evolutionary account of morality that is basically utilitarian in nature: If humans but allow the mechanisms of nature to do their work, there will be natural social evolution toward greater freedom. This will in turn lead to the greatest possible amount of happiness. While widely acclaimed during its time, Spencer's theory was ultimately rejected owing in part to its scientific inaccuracies, and to attacks upon it by Henry Sidgwick, Thomas Huxley, and G. E. Moore. At its worst, Spencer read repugnant norms into evolution; for example, here is what he said about Great Britain's Poor Laws, which mandated food and housing for the impoverished: "... there is an habitual neglect of the fact that the quality of a society is lowered morally and intellectually, by the artificial preservation of those who are least able to take care of themselves ... the effect is to produce, generation after generation, a greater unworthiness" (Spencer 1873 , p. 313).
What might a more plausible scientific ethic look like? Such a theory might resemble that offered by the Greek philosopher Aristotle or the pragmatic philosopher John Dewey (1859–1952).
Aristotelian ethics is prescientific in the sense that the scientific revolution had not yet occurred; nonetheless, his method is empirical. For Aristotle, human flourishing is the summum bonum of existence; to say that an action is ethical or that a person is good is just to say that the action or the person contributes to or constitutes proper functioning. Contemporary ethicists have pursued this line of reasoning; for example, Larry Arnhart (1998) argues for a naturalized, Aristotelian ethical framework, and William Casebeer (2003a, b) argues that moral facts can be reduced to functional facts, with functions treated as an evolutionary biologist would (that is, as being fixed by evolutionary history). Leon Kass (1988) raises questions for such approaches; there are things that human passions and gut reactions say about the morality of certain actions that can never be captured with reason or the scientific method alone.
A related merging of science and ethics occurs in the work of the classic American pragmatists, such as Charles Pierce (1839–1914) and Dewey. Pierce argues that science itself is a form of ethics—it expresses respect for the values that underpin effective inquiry, and is subordinate to ethics insofar as it is human concerns about the efficacy of ideas that cause people to pursue science to begin with. Relatedly Dewey argues in his Ethics (1932) that the process of regulating ideas effectively—which is what science does in essence—enables human beings to become better able to express values and act upon them. This approach of replacing preexisting value with the creation of value and understanding what genuinely follows from that positing of value is called axiology (Casebeer 2003a).
Even if moderate and/or strong versions of scientific ethics seem implausible, almost everyone admits that scientific results may limit the possible space of normative moral theories. Only the most trenchant antinaturalist would think that facts about human beings and how they reason have absolutely no bearing on moral concerns. These facts should, at the very least, constrain moral theorizing. For instance, Owen Flanagan advocates the principle of minimal psychological realism, which states that the moral psychologies required by moral theories must be possible for humans: "Make sure when constructing a moral theory or projecting a moral ideal that the character, decision processing, and behavior prescribed are possible ... for creatures like us" (Flanagan 1991, p. 32). So the scientific study of the genesis, neurocognitive basis, and evolution of ethical behavior is relevant to normative moral theory even if the moderate and strong versions of scientific ethics are misguided or fail.
Contemporary Developments and Future Possibilities
There are five general areas in which scientific research has the potential to constrain moral theory: moral psychology, decision theory, social psychology, sociobiology, and artificial modeling of moral reasoning. Moral psychologists focus on the psychological processes involved in moral thought and action. They study such phenomena as akrasia (weakness of the will), moral development, the structure of moral reasoning, and the moral emotions. Some of the best known work in this area revolves around moral cognitive development; Lawrence Kohlberg, for example, has formulated an empirically robust theory of moral development whereby people progress through three stages of moral reasoning, each broken into two levels. In the first stage, one reasons by asking, What's in it for me? In the second, one asks, What does culture or society say? In the third, one asks, To what contract would I be a party? What do universal moral principles demand? Progress through these stages or schema is universal and (with some exceptions) invariant. If Kohlberg is right, then perhaps a normative moral theory that takes issues of justice seriously is more viable than one that does not (although his research has been criticized for this very reason; see Lapsley 1996 for a summary).
Other moral psychologists have been exploring the relationship between reason and moral emotions such as guilt or shame. One longstanding debate in moral theory has involved the relationship between having a moral reason to do something and whether that reason necessarily motivates an individual to take action. Internalists (such as Plato or Kant) argue that moral reasons necessarily motivate: If, morally speaking, one ought not to do something then one will, ceteris paribus, be motivated not to do that thing. Externalists (such as Aristotle) argue that a moral reason must be accompanied by an appropriate motivational state (such as an emotion) in order to spark action. If certain normative moral theories require either an internalist or externalist psychology in order to be plausible, then results from empirical research may constrain moral theory. For example, Adina Roskies (2003) argues persuasively that neurobiological data about the relationship between emotion and reason rules out internalism and makes a Kantian psychology implausible. Other issues in moral psychology will stand or fall with progress in the cognitive sciences; for instance, moral cognitive development and moral concept development may both be subsumed by research into cognitive and concept development in general.
Decision theorists study the determinants of human choice behavior. Traditional rational actor assumptions (such as possessing unlimited time and computational power, a well-ordered preference set, and indifference to logically equivalent descriptions of alternatives and choice sets) usually inform decision theory. Whether or not these assumptions apply to human reasoning when it is done well may affect whether normative moral theories must be essentially rational and hence whether they must respond to the same norms as those of reason traditionally construed. Much work in decision theory has revolved around either extending the predictive power of traditional rational actor assumptions, or in articulating alternative sets of rational norms to which human cognition should be responsive. For instance, Amos Kahneman and Daniel Tversky's (1982) heuristics and biases research program explores the shortcuts human beings take to achieve a reasonable result when under time pressure or when working with incomplete information. It may very well be that normative moral theories constitute sets of heuristics and biases.
Gerd Gigerenzer and the Adaptive Behavior and Cognition Research Group (2000) focus on ecological rationality, demonstrating that traditional rational canons can actually lead people astray in certain environments. While there is a rearguard action to shore up traditional rational actor driven decision theory, in all likelihood, progress on this front will require articulating a new conception of rationality that is ecologically valid and cognitively realistic. The results of this program may, in turn, affect the structure of normative moral theory in much the same way that the structure of normative rational actor theory has been and will be affected.
Social psychologists study human cognition and emotion in the social domain. Given that moral judgments are paradigmatically about how people ought to treat others, work in this area usefully constrains normative theorizing. One controversy regards whether or not the fundamental attribution error (the human tendency to undervalue the situational influences on behavior and overvalue the internal character-driven causes) undermines traditional approaches to virtue theory. If, as some social psychologists argue, there is no such thing as bravery as a general trait, but rather only such fragmented virtue-theoretic traits as brave while standing in the checkout line at the grocery store, then it may very well be that virtue theory will have to become much more sophisticated if it is to be plausible (see Doris 2002 for a comprehensive discussion, as well as Harman 2000; Doris and Stich 2003 also offer a useful survey). The social nature of moral reasoning means that the latest studies of social psychological behavior can, on the weakest view, usefully constrain normative theorizing, and on a stronger view can usefully coevolve with it.
Sociobiologists such as E. O. Wilson study the origin and evolution of (among other things) moral norms. They argue that genes keep moral culture on some sort of leash: At the very least, the capacities human beings use to reason about morality are evolved capacities and need clear connections to the environments in which these capacities evolved; maximally moral norms may be nothing more than norms that have enabled organisms and groups of organisms to increase their genetic fitness. Sociobiological approaches to human social behavior have been controversial, but have nonetheless shed much light on how both the capacity to reason morally and the structure of some moral norms came to be (Boehm 1999, for example, discusses the evolution of egalitarian norms). Game-theoretic work on the evolution of the social contract and other moral norms has illuminated aspects of ethical behavior ranging from the propensity to be altruistic to the temptation to defect on agreements in certain instances. Sociobiological study reinforces the notion that any accepted normative theory should have a describable evolution and a discernable way of maintaining its existence (see Binmore 1994).
Computer models at both the micro and macro level have usefully informed all these fields of research. Changes in technology have influenced what philosophers make of the possibility of scientific impact on ethics. For example, Rene Decartes's inability to reconcile how mental states could be identical to brain states drove, at least in part, his dualism. The advent of in vitro methods for identifying the neural machinery of cognitive activity, such as Positron Emission Tomography (PET) and functional Magnetic Resonance Imaging (MRI), may have headed off dualism at the philosophic pass if such technologies were available during his time. The spread of inexpensive and powerful computing technology has made possible everything from the simulation of artificial societies (and hence has influenced sociobiological approaches) to the simulation of moral reasoning in an individual (and hence has influenced moral psychology). On the social simulation front, promising work by Jason Alexander and Bryan Skyrms (1996) on the evolution of contracts has usefully informed moral theorizing. On the individual level, work by cognitive modelers such as Paul Thagard (2000) and Paul Churchland (2001) has highlighted areas where normative moral theory can intersect with cognitive modeling.
Is scientific ethics possible? Appropriately enough, this is an empirical matter. Should the promise held out by the rapidly progressing cognitive, biological, and evolutionary sciences be realized, there is reason to be sanguine about the moderate and strong programs for a scientific ethic. Science could reaffirm some of the prescientific insights into the nature of morality. But even if this very possibility is a misguided hope, scientific insights into human nature and cognition can usefully constrain the possible space of normative moral theory, and in this sense the existence of scientific ethics is a foregone conclusion. Science and ethics are indeed both magisterial, but they are, ultimately, overlapping.
WILLIAM D. CASEBEER
Aristotle. (1985). Nichomachean Ethics. Indianapolis, IN: Hackett. Aristotle's best-known work in normative ethics. Aristotle was a student of Plato, another foundational figure in virtue ethics.
Arnhart, Larry. (1998). Darwinian Natural Right: The Biological Ethics of Human Nature. Albany: State University of New York Press. Defends a contemporary version of Aristotelian ethics using evolutionary biology.
Bacon, Francis. (1605 ). The Advancement of Learning, ed. Stephen J. Gould. New York: Modern Library. Arguably the humanities/natural science split first occurred with Francis Bacon; a useful introduction to his thought.
Bacon, Francis. (1620 ). Novum Organum. Chicago: Open Court Publishing Company. A good statement of the canons of empirical science early in their developmental history.
Binmore, Ken. (1994). Game Theory and the Social Contract, Vol. 1: Playing Fair. Cambridge, MA: MIT Press. Using assumptions from game theory, argues against the possibility of a Kantian reconstruction of John Rawls's theory of justice.
Boehm, Christopher. (1999). Hierarchy in the Forest: The Evolution of Egalitarian Behavior. Cambridge, MA: Harvard University Press. Discusses the evolution of egalitarian tendencies in apes and hominids.
Caplan, Arthur L., ed. (1978). The Sociobiology Debate: Readings on the Ethical and Scientific Issues Concerning Sociobiology. New York: Harper and Row. While dated, nonetheless serves as an excellent introduction to the sociobiology debate.
Casebeer, William D. (2003a). Natural Ethical Facts: Evolution, Connectionism, and Moral Cognition. Cambridge, MA: MIT Press. Argues for a strong form of scientific ethics, recapitulating a neo-Aristotelian virtue theory using resources from evolutionary biology and cognitive neuroscience.
Casebeer, William D. (2003b). "Moral Cognition and Its Neural Constituents." Nature Reviews: Neurosciences 4: 841–847. Discusses the interplay between neuroscience and normative moral theory while reviewing the neural mechanisms of moral cognition.
Churchland, Paul M. (2001). "Toward a Cognitive Neurobiology of the Moral Virtues." In The Foundations of Cognitive Science, ed. Joa˜o Branquinho. Oxford: Oxford University Press. Reprint of Churchland's seminal 1998 article; discusses the affinity between virtue theory and neural network conceptions of cognition.
Dewey, John. (1932 ). Ethics. Carbondale: Southern Illinois University Press. Dewey's best known text in moral theory.
Doris, John M. (2002). Lack of Character: Personality and Moral Behavior. Cambridge, UK: Cambridge University Press. Surveys results from contemporary social psychology, arguing that they make implausible the kinds of traits required by virtue theory.
Doris, John M., and Stich, Stephen. (2003). "As a Matter of Fact: Empirical Perspectives on Ethics." In The Oxford Handbook of Contemporary Analytic Philosophy, eds. Frank Jackson, and Michael Smith. Oxford: Oxford University Press. Offers an excellent introduction to empirical issues in ethics.
Elman, Jeff L., et al. (1996). Rethinking Innateness: A Connectionist Perspective on Development. Cambridge, MA: MIT Press. Argues that neural network accounts of cognition can account for multiple facts about the development of language in humans.
Flanagan, Owen. (1991). Varieties of Moral Personality: Ethics and Psychological Realism. Cambridge, MA: Harvard University Press. Discusses the relationship between psychology and normative moral theory.
Gigerenzer, Gerd; Peter M. Todd; and the ABC Research Group. (2000). Simple Heuristics That Make Us Smart. Oxford: Oxford University Press. Discusses the heuristics and biases at work in human thought, including what impact this has for the human conception of rationality.
Gould, Stephen J. (2002). Rocks of Ages: Science and Religion in the Fullness of Life. New York: Ballantine Books. Discusses the relationship between ethics and science, arguing that the two are nonoverlapping magisteria.
Harman, Gilbert. (2000). "Moral Philosophy Meets Social Psychology: Virtue Ethics and the Fundamental Attribution Error." In Explaining Value and Other Essays in Moral Philosophy. Oxford: Oxford University Press. Argues that the fundamental attribution error demonstrates that there are no such things as character traits.
Hume, David. (1739 ). A Treatise of Human Nature, ed. Lewis A. Selby-Bigge. Oxford: Clarendon Press. Clearly articulates the naturalistic fallacy in this defense of noncognitivism about ethics.
Kant, Immanuel. (1785 ). Grounding for the Metaphysics of Morals, trans. James Ellington. Indianapolis, IN: Hackett. A concise summary of Kant's views on ethical issues and one of the best-known works in deontology.
Kahneman, Daniel; Paul Slovic; and Amos Tversky, eds. (1982). Judgment Under Uncertainty: Heuristics and Biases. Cambridge, UK: Cambridge University Press. The seminal work in the heuristics and biases research program.
Kass, Leon. (1988). Toward a More Natural Science: Biology and Human Affairs. New York: Free Press. Argues that science can go too far if it is not appropriately regulated by the wisdom contained in emotional reactions to certain technological advances.
Lapsley, Daniel K. (1996). Moral Psychology. Boulder, CO: Westview Press. An excellent survey of issues in moral psychology.
Moore, G. E. (1902 ). Principia Ethica. Amherst, NY: Prometheus Books. Articulates the open question argument against strong scientific ethics.
Roskies, Adina. (2003). "Are Ethical Judgments Intrinsically Motivational?: Lessons from Acquired Sociopathy." Philosophical Psychology 16(1): 51–66. Seminal article tackles the internalism/externalism debate (do moral reasons necessarily move people to act?) using neurobiological evidence.
Skyrms, Brian. (1996). Evolution of the Social Contract. Cambridge, UK: Cambridge University Press. Brings iterated game theory to bear on the evolution of the norms of justice and exchange; useful work in this area has also been accomplished by Jason Alexander, who studied under Skyrms.
Thagard, Paul. (2000). Coherence in Thought and Action. Cambridge, MA: MIT Press. Uses computer models to explain how concerns about coherence affect moral reasoning.
Wilson, Edward O. 1975 (2000). Sociobiology: The New Synthesis, 25th Anniversary edition. Cambridge, MA: Harvard University Press. Single-handedly established the field of sociobiology (or, in contemporary terms, evolutionary psychology).