Can Science Determine Moral Values?
A challenge from and dialogue with Marc Hauser about The Moral Arc
by Michael Shermer
Marc Hauser: The Moral Arc (TMA) is a tour de force, a celebration of our moral progress, and an inspiration for times when we see the world as dark and dangerous. But as in any tour de force, there are sections that are controversial and others that seem to provide less than the explanatory adequacy championed. I want to take up the latter, focusing in particular on the fundamental thesis that runs throughout the book, articulated succinctly on page 3 of the Prologue:
… we can trace the moral arc through science with data from many different lines of inquiry, all of which demonstrate that in general, as a species, we are becoming increasingly moral. As well, I argue that most of the moral development of the past several centuries has been the result of secular not religious forces, and that the most important of these that emerged from the Age of Reason and the Enlightenment are science and reason, terms that I use in the broadest sense to mean reasoning through a series of arguments and then confirming the conclusions are true through empirical verification.
In brief, though I fully agree that reasoning by means of careful argumentation has and will continue to serve us well on our path to moral progress, I disagree that science and scientific evidence will settle or even help settle many of the moral challenges we face as individuals and as a species. As should be clear from the empirical work that I and many of my students and collaborators have carried out on the nature of moral judgments (see, for example, Joshua Greene’s book Moral Tribes)
I fully support scientific inquiry into morality. But I don’t believe that this work will settle key moral problems; rather, it will illuminate the nature of our moral instincts, together with the role that cultures may play in bending both our judgments and our actions. In fact, as I see it, some of the primary challenges to your thesis come from the moral dilemmas and scenarios that have been used in the scientific work of late, and that you discuss in TMA. Let me begin, therefore, close to where you begin in the book, with the famous trolley problem.
The moral philosopher Philippa Foot first introduced the trolley problem in a paper focused on abortion and the doctrine of double effect. Fantasy dilemmas such as the trolley problem are introduced in philosophy to help us think through central moral issues in the absence of our real world connection to the problem. Thus, instead of wrestling with abortion per se, Foot wrestled with the reasons why we might or might not consider the means by which we carry out an action as more or less significant than the consequences of such actions; this distinction maps on, to a first approximation, to difference between deontological as opposed to utilitarian perspectives.
The original trolley problem asks us to consider whether the driver of a runaway (no brakes) trolley should allow it to continue down the track where it will run over and kill 5 workers or steer it onto a sidetrack where it will kill 1 worker. If one focuses on consequences, the answer is easy: turn the trolley, killing 1 worker but saving the lives of 5. But there are other considerations, including: the 1 worker on the side track is safe, so by turning the trolley, the driver is deciding that his life is less valuable than any single individual on the main track; the driver doesn’t intend to kill the 1, rather, he intends to save the 5, so the 1 worker’s death is a byproduct or side-effect; and so on.
In response to Foot’s case, the moral philosophers Judith Thomson and Frances Kamm spun off a railroad station’s worth of cases, all designed to disentangle the factors that might be in play when we are faced with competing moral outcomes. One of the most vivid spinoffs considered the case where a runaway trolley is empty, but a bystander is standing on the side of the track next to a very heavy man. If the bystander pushes this man onto the track, his weight will stop the trolley from advancing and killing the 5 workmen, but this man will of course die. Here, though the numbers are the same as in Foot’s case – 1 vs 5 – many of us feel a substantial difference. In particular, though the consequences seem to dominate our decision-making in the original case, the means seem to dominate our decision-making in this second case: though we may feel that turning the trolley makes good moral sense in the first case, pushing the man seems morally wrong in the second case.
Much has been discussed about these cases, and many others like them (see below). But the conclusions that have been drawn, including connections to applied issues such as abortion and euthanasia (see below) derive from careful deliberation, reflection and philosophical expertise, not scientific evidence. Where science has played a role-including some of my own work, as well as the contributions of John Mikhail, Joshua Greene, and some of the remarkable students I have had the privilege to work with (Fiery Cushman, Liane Young)-is in revealing how different factors influence people’s judgments. Consider one example, one that is highly relevant to a number of issues raised in TMA. Based on responses from thousands of subjects, judging hundreds of different dilemmas (Hauser et al., 2007; Banerjee et al, 2010), judgments about right or wrong were not influenced by gender, education, political affiliation or religious background, including a contrast between atheists and all those with some kind of religious background. This kind of evidence is of interest in terms of our understanding of the factors that may or may not guide moral judgments, but they don’t enable us to decide whether we should punish those who push the heavy man or reward them because they have contributed to greater human flourishing! The debate between deontological and utilitarian reasoning rages on, and science won’t decide which wins the day.
Consider next a different set or class of moral situations: active vs. passive euthanasia and incest. Each of these issues have fascinated philosophers and scientists, and much progress has been made in thinking about them. But none of the scientific evidence provides the means for deciding which position is the more morally progressive view, or is more likely to lead to human flourishing.
In many countries, including the United States, it is legally permissible to allow someone who is in pain and suffering from an incurable disease to die (passive euthanasia), but legally forbidden to cause this person to die through lethal injection (active euthanasia). Countries that allow both active and passive euthanasia, such as the Netherlands, have taken this route because of careful reasoning, including arguments from philosophers. Scientific evidence didn’t push the Netherlands in this direction, though there is scientific evidence from some of my own work (Hauser et al., 2009) comparing how Dutch and American subjects judge the moral permissibility of actions as opposed to omissions in unfamiliar cases. Interestingly, the results show that although the Dutch have explicitly decided to endorse both the action of ending someone’s life and the omission of allowing someone to die, like Americans, they see actions as morally worse than omissions when the scenarios are unfamiliar. This is interesting with respect to the nature of our moral judgments, and the relative immunity of this system to cultural influences, but it doesn’t tell us whether the Americans are backwards or progressive in terms of the legality of euthanasia.
Scientists such as Jonathan Haidt have carried out terrific work showing how the emotion of disgust plays a role in guiding our moral judgments, including cases of incest. In his most famous case, when you tell people about a brother and sister who decide to have protected intercourse and keep it a secret, many find this morally wrong and definitely yucky, despite the fact that it is protected sex and thus, without reproductive consequences! But the fact that you might find incest disgusting and morally wrong, whereas someone else may not, doesn’t tell us whose view is morally superior, preferred, or more likely to lead to human flourishing. In fact, given that incest among relatively distant relatives, such as second cousins, is unlikely to lead to negative consequences for the developing fetus, one might argue that we should support those who are more tolerant of incest. And tell others to just get over their disgust!
In sum, scientific evidence provides increasingly interesting information on the nature of our moral judgments and actions, including a wealth of evidence that extends from genes to behavior. But most of this work plays no role in shaping the moral conversation. Scientific evidence can illuminate how human nature and nurture shape our moral judgments to these cases, but will not provide the ballast for adjudicating between arguments.
Thank you for taking the time to read my book carefully Marc, and for articulating so clearly what most scientists and philosophers believe about the wall separating science and values.
First, I conjecture that it is “science and reason”-not just “science” in the narrow sense of running experiments in a lab and collecting data—that have been the major drivers of moral progress because they can and have determined moral values. (I reject the “philosophy is dead” notion recently proffered by a few popular scientists, because the philosophical tradition of reason and logic underlies all of science, and scientists use reason when deducing general principles from specific observations.) Ever since the Scientific Revolution, when scientists such as Bacon, Copernicus, Kepler, Galileo, Harvey, and Newton discovered that the world is governed by natural laws and principles that can be revealed, understood, and used to make predictions and test hypotheses, thinkers in other fields during the Age of Reason and the Enlightenment sought to understand the laws and principles that govern political, economic, legal, social, and moral systems, which they then used to make predictions and test hypotheses about how best we should live. Thomas Hobbes, Charles Montesquieu, Cesare Beccaria, John Locke, Jeremy Bentham, Thomas Jefferson, Benjamin Franklin, Thomas Paine, Immanuel Kant, François Quesnay, David Hume, Adam Smith, and others were all, in my reckoning, scientists who employed the best empirical and rational tools of their age. The term “scientist” didn’t exist then, so they are often referred to as philosophers or natural philosophers, but whatever terms we use my point is that they placed supreme value on reason and scientific inquiry, from which they discovered or derived such concepts as human natural rights, equal treatment under the law, individual autonomy, freedom of thought and expression, and other principles related to equality and liberty, on top of which they built a diverse, cosmopolitan worldview of Enlightenment Humanism.
Hobbes’s Leviathan, considered the most influential political treatise ever written, begins with atoms in motion and builds on observations and first principles to devise a rational- and empirical-based social system (he called himself the Galileo of civil society). In his book Esprit des Lois (The Spirit of the Laws), Montesquieu invoked Newton when he compared a well-functioning government to “the system of the universe” that includes “a power of gravitation” that “attracts” all bodies to “the center” (the monarch), and he employed the deductive method of Descartes: “I have laid down first principles and have found that the particular cases follow naturally from them.” By “spirit” Montesquieu meant “causes” from which one could derive “laws” that govern society. “Laws in their most general signification, are the necessary relations derived from the nature of things,” he wrote. Quesnay—physician to the King of France—and his followers (the French physiocrats) undertook a systematic study of the economy from which they gathered empirical evidence and derived rational principles that underlie how economies grow or shrink as a function of government policies (and from where the French term laissez faire—“leave alone”—comes). This led to the Scottish Enlightenment philosopher Adam Smith to compose the founding text of economic science, which everyone knows as The Wealth of Nations. Its full title, in fact, is An Inquiry into the Nature and Causes of the Wealth of Nations. It is a scientific inquiry to discover the true nature and causes of wealth, straight out of the tradition of the scientific revolution.
So historically, we have already been using science to determine such moral values as the best way to structure a polity, an economy, a legal system, and a civil society, in the same way that physicians have developed improved medical science and epidemiologists have worked to build better public health science in order attenuate plagues, disease, and other scourges of humanity. If you agree that it is better that millions of people no longer die of yellow fever and smallpox, cholera and bronchitis, dysentery and diarrhea, consumption and tuberculosis, measles and mumps, gangrene and gastritis, and many other assaults on the human body, then you have offered your assent that the way somethingis(diseases such as yellow fever and smallpox kill people) means weoughtto prevent it through vaccinations and other medical and public health technologies. Analogously, if you agree that millions of lives have been saved over the past couple of centuries by a reduction in violence (war, torture, homicides, etc.) due to improved understanding of causality in these areas and the application of appropriate policies based on those causes, then you might well concur that applying the methods of the social sciences to further attenuating war, crime, and violence is also something we ought to do.
Why are these science-based policies morally good? Because they lead to the survival and flourishing of sentient beings, which is a moral starting point grounded in evolutionary biology. By survivalI mean the instinct to live, and byflourishingI mean having adequate sustenance, safety, shelter, bonding, and social relations for physical and mental health. I claim that any organism subject to natural selection will by necessity have this drive to survive and flourish. If it didn’t, it would not live long enough to reproduce and would no longer be subject to natural selection. BysentientI meanemotive, perceptive, sensitive, responsive, conscious,and therefore able to feel and to suffer.
Finally, to your point about utilitarianism and the trolley dilemma, by the “moral arc” of progress I mean an improvement in the survival and flourishing ofindividualsentient beings. I emphasize the individual (the 1 worker on the track) over the collective (the 5 workers on the track) for four reasons: (1) Natural selection operates on individual organisms, not groups. (2) It is the individual who survives and flourishes or who suffers and dies, not the group, tribe, race, gender, state, nation, empire, or society. Individual sentient beings perceive, emote, respond, love, feel, and suffer, not populations, races, genders, groups, or nations. (3) Historically, immoral abuses have been most rampant, and body counts have run the highest, when the individual is sacrificed for the good of the group. The utilitarian calculus that it is permissible to kill 1 to save 5 is too easy to ratchet up to kill 1 million to save 5 million, and that is the basis of genocide and why utilitarianism fails in certain real-world situations (as opposed to hypothetical moral dilemmas), and therefore… (4) The rights revolutions of the past two centuries have focused almost entirely on the freedom and autonomy of individuals, not collectives—on the rights ofpersons, not groups. Individuals vote, not races or genders. Individuals want to be treated equally, not races. Rights protect individuals, not groups; in fact, most rights (such as those enumerated in the Bill of Rights) protect individuals from being discriminated against as members of a group, such as by race, creed, color, gender, or—soon—sexual orientation and gender preference.
Legal systems have evolved to follow this line of reasoning and historical development. Analogous to the trolley problem, if a surgeon has 1 healthy person in her waiting room and 5 patients in operating rooms each dying of an organ failure that the harvesting of the 1 will save the 5, if she were to carry out the surgeries resulting in the death of the 1 healthy person to save the 5, she would go to jail for murder. The moral arc has bent, in part, because our legal system has followed our intuition that the intentional harm or murder of an individual against their will feels wrong, and your own research confirms that most people would not push 1 man off a bridge onto the track to stop the trolley from killing 5 workers. Natural rights theory trumps utilitarianism based on my moral starting point of the survival and flourishing of individual sentient beings.
As for your real-world examples, euthanasia is resolvable by natural rights theory (which I consider to be science-based): as long as the individual consents to allowing herself to die or instructs someone to initiate an assisted suicide (here videotaped consent should be mandatory to prevent abuse of the law), it is morally acceptable regardless of whether or not it leads to the greatest good for the greatest number. Incest, in part, follows natural rights theory because the incest taboo, which anthropologists have shown is a human universal, was selected for because of the genetic harm from too much inbreeding, and from modern psychological research showing that incestuous relationships between, for example, fathers and daughters, can be severely damaging to the child. Of course, if you alter the conditions such that the incestuous relationship is consensual, between distant cousins, and does no one harm, then it may be considered morally acceptable. But here the problem is that the exceptions are mostly in the realm of philosophical thought experiments designed to nudge our intuitions to come into conflict with our reasoning. In conclusion, science and reason can and have helped us determine moral values.
I started off my comments by noting a distinction between reason and scientific evidence.I specifically said that reason, rational discourse, etc., has been essential to moving our discussions of morality. I also noted that science has informed important aspects of how humans judge moral situations and how we act, and what can lead to universality as opposed to cross-cultural variation; the latter can be important as knowing human biases can inform policies, as Kaplow and Shavell have argued in their legal treatise comparing fairness as opposed to individual welfare discussions.My worry is that I don’t feel as though you engaged with the core part of my comment which is that scientific evidence can’t adjudicate between different moral perspectives when different moral perspectives have validity on their own. In other words, if you are a utilitarian you put more weight on consequences than means, and if you take on a deontological perspective, you see the means as more important than the consequences; you see the world through a lens of well reasoned “reasons.” So reasoning yes.But the key part of the quote from your book is that we determine that our moral “conclusions are true through empirical verification.” But evidence wouldn’t convert a utilitarian over to the dark side of deontology, and vice versa.A good counter-example would convert some, especially if it led to a slippery slope dragging in other cases.So it is not that I reject reasoning, and it is not that I reject the role of science in some cases. For example, if you can show that a vaccine saves thousands of lives-the evidence-then it should be possible to argue based on this that people ought to take the vaccine. And yet, even here, if your culture promotes a perspective of using only traditional medicine, as opposed to artificial chemicals, but you do so knowing the risks, could we mandate this as policy?But toward the end of your commentaries, when you engage in my cases, your response is couched in terms of good reasons, but not in terms of evidence from scientific observations and experiments that could adjudicate between the options. A smart open minded utilitarian could be convinced to change as a function of a good logical argument, but if you showed him evidence that, say, significantly more people consider the means over the consequences, I doubt this would have any impact. I raise this latter point, because this is precisely the response I received from many distinguished philosophers in response to my empirical work. In fact, Frances Kamm, the distinguished Harvard ethicist told me in a seminar on my work that she didn’t really care if 5 million people voiced a different judgment from her own on a particular trolley problem, because her own reasons were principled and considered in the context of a broader view of right and wrong. In brief, the evidence was irrelevant.
The challenge, in brief, is for you to point to work in either your book, or elsewhere, or even in principle, that could flip things around.This is the challenge that I posed to Sam Harris as he was writing his book, and I don’t feel that it has been addressed.
I think a lot of moral thought experiments along the lines of the trolley problem, or the “lifeboat ethics” dilemmas given to students to suss out the various moral problems inherent in any ethical system, may not be ultimately resolvable through science, in the sense you are using the term to mean empirical evidence.To come at this in a slightly different manner, given the diversity of human interests and moral foundations it may be that there are, as Sam Harris articulated it in his book The Moral Landscape, “multiple peaks on the moral landscape.” For example, in my public debates with John Lott over gun control (he wants almost none and I want some), it became clear to me that there are a lot of Americans who simply don’t care how many people die from gun violence each year (tens of thousands), they cherish their freedom to own a gun over the carnage that piles up as a result. What a science-and-reason based society has done is allow us to establish a system that can be changed in response to these differing values so that there are multiple peaks from which to choose. Science may help you choose which one is best for you, and science may help society design its moral systems to be as optimal as possible for these differing peaks.
I think the core issue, or difference between us, boils down to this. I think your sense of science, and what it can contribute, boils down to reason, even though you explicitly stated that from reason we decide with “empirical verification.” For me the challenge has always been whether science, in terms of evidence, experiments, observations, and modeling can decide between competing moral views in the hard cases we have been discussing.I don’t think it can, and so far, you haven’t provided any cases where it has, or walked through some plausible scenarios for how it might. When philosophers get in your grill, or Sam’s, or mine and Josh Greene, it is about the strong version of the claim: science and its evidence, not the role of reason. Sam’s multiple peaks don’t help. Sure there are multiple peaks, and this is what philosophers and other great minds have considered for a long time.I don’t think that is new.
If, as you say, it is about how individuals decide what is best for them, then we are not in the game of morality but about individual choice.If individuals within a society recognize the significance of a finding, and decide to follow its implications, then that would be different. But if I decide for myself what pieces of evidence are worth picking and what pieces I can ignore, and I base my morality on such selective picking, we won’t have a system of morality that can operate with others.
Now, imagine that we find scientific evidence that determines, without doubt, that some animals not only feel pain, but can think about their future selves, understand what it is like for another to feel pain, and act on the basis of it. They have, in essence, some of the critical ingredients of sentience as well as moral agency (as opposed to moral patience).I understand this evidence and I decide that since these animals are like my children, I can neither support eating them nor carrying out experiments. You, on the other hand, also understand this evidence, but decide that it is irrelevant to your moral decisions because they are not humans.In essence, you decide that humans have the moral right to control other animals, including hunting them and using them, because the core issue is our well being.Someone else decides that the evidence is important, but insufficient. That for another species to count as a moral agent, they would need to understand, or at least come to understand (as in human babies) the moral issues in play.The evidence can’t decide who is morally right because there are different standards.Three peaks? If so, fine, but science isn’t going to adjudicate.
So, I guess it boils down to this for me.If the moral arc is guided by reason, as opposed to gut feelings and appeal to the supernatural, I am totally on board, and I can’t imagine a philosopher on the planet who wouldn’t be.On the other hand, if you want to take the stronger position that all aspects of the moral arc have been and will be guided by scientific evidence, of the kind that biologists, chemists, physicists, etc., collect in their daily lives, then I don’t think you have shown this, either in terms of prior work or in principle work.
Perhaps not all aspects of the moral arc have been bent by science. But what has happened is that in the same way that Galileo and Newton discovered physical laws and principles about the natural world that really are out there, so too have social scientists discovered moral laws and principles about human nature and society that really do exist. Just as it was inevitable that Kepler would discover that planets have elliptical orbits-given that he was making accurate astronomical measurements, and given that planets really do travel in elliptical orbits, he could not have discovered anything else-scientists studying political, economic, social, and moral subjects will discover certain things that are true in these realms. For example, that democracies are better than autocracies, that market economies are superior to command economies, that slaves don’t like being enslaved and oppressed, that torture and the death penalty do not curb crime, that burning women as witches is a fallacious idea, that Jews did not cause the Black Death, that blacks are not intellectually inferior to whites, or that women are not too weak and emotional to run companies or countries.
My view is that ever since the Age of Reason and the Enlightenment the idea that individual sentient beings have natural rights has outcompeted other ideas that place the group, tribe, nation, race, gender, or religion above the rights of the individual. These rights have expanded around the globe because individual sentient beings want them, and they want them because it is part of their nature to want them-it is instinctive-and a proper scientific understanding of human nature has revealed this fact. Knowing that, we then have a moral obligation to expand those rights where we can, and to help people whose rights are being violated.
Jonathan Haidt’s six moral foundations (described in his book The Righteous Mind) are an interesting test case because, as he argues, they are part of our nature, evolved features of our minds as social primates.
- Liberty/oppression, related to our desire for freedom and autonomy and our resentment of bullies and oppressors who try to restrict our liberty.
- Harm/care, related to our long evolution as mammals with attachment systems and an ability to feel (and dislike) the pain of others. This foundation underlies virtues of kindness, gentleness, and nurturance.
- Fairness/reciprocity, related to the evolutionary process of reciprocal altruism. This foundation generates ideas of justice, rights, and autonomy.
- In-group/loyalty, related to our long history as tribal creatures able to form shifting coalitions. This foundation underlies virtues of patriotism and self-sacrifice for the group.
- Authority/respect, shaped by our long primate history of hierarchical social interactions. This foundation underlies virtues of leadership and followership, including deference to legitimate authority and respect for traditions.
- Purity/sanctity, shaped by the psychology of disgust and contamination. This foundation underlies religious notion that the body is a temple that can be desecrated by immoral activities and contaminants.
As Haidt’s data shows, liberals tend to emphasize the first three, conservatives tend to value the second three, and libertarians focus more on the first foundation over all others. I am prepared to argue that one of the drivers of moral progress as I have defined it is that the second three foundations that have been the backbone of groups, tribes, nations, and religions are being outcompeted by the first three that are the central core of the rights revolutions of the past two centuries. The reason for their success is that the worldview of Enlightenment Humanism and Classical Liberalism that embraces liberty, care, and fairness over in-group loyalty, authority, and purity is more likely to lead to the survival and flourishing of individual sentient beings.
Exceptions such as ISIS, al Qaeda, and others who want to return to a 7th century Caliphate built on Sharia prove the generalization: they reject the Enlightenment values of science, reason, openness, tolerance, and individualism. I claim that these moral foundations-these truths about our moral nature-are discoverable by science, and once discovered can be used in the service of the betterment of humanity. It is in this sense that rights theory trumps utilitarian theory, at least in these cases and others I document in The Moral Arc.