Naturalistic Fallacy: 1, Sam Harris: 0

For those who don’t know Sam Harris, he is a rather famous critic of theism who often invokes science and broad rationalism in his arguments for the abandonment of organized religion. Along with Christopher Hitchens, Daniel Dennett, and Richard Dawkins he is sometimes known as one of the “four horsemen of the neo-atheiest movement” (Wilson 2011).

In the TED Talk that I post above, Harris tries to dissemble the well-known Is-Ought Problem (also known as Naturalistic Fallacy) of David Hume. For those not familiar with this critical idea, it attempts to separate fact from values. The universe works by a multitude of mechanisms producing a greater multitude of phenomena, and some of these phenomena can be observed and their mechanisms understood using the methods of scientific inquiry. Such scientific discovery tells us what ‘is’, but can we determine what ‘ought’ to be based on scientific understanding? Hume famously says ‘no’, and to make the jump from ‘is’ to ‘ought’ is to commit the naturalistic fallacy. Understanding how the universe produces outcomes does not tell us what outcomes we ought to value.

I tend to agree with Hume. Many exceedingly common biological phenomena are also morally repugnant to all but a few human beings. My favorite depiction of the real brutality and amorality of nature was provided by George C. Williams in his last book, Plan and Purpose in Nature (Williams 1996). In Chapter 9 (“Philosophical Implications”), Williams describes the kind of deception, murder, and coercion that occur in nature and then asks “Do you still think God is good?”. Nature is morally neutral, and so humans must construct values without guidance from nature; scientific discoveries will therefore never be moral discoveries.

The construction of values is one of humanity’s greatest cultural adaptations, an adaptation that allows us to possess and achieve intentions and goals that are somewhat independent of the natural world in which we live. While we cannot completely escape the clear limitations of what ‘is’, our constructed values (paired with the cooperative effort that these values inspire) allow us to navigate a much larger space of possible outcomes than other organisms. Humans have been fiddling around with values to see what kind of social outcomes they produce for millenia, and the societies of today are those whose values produced sustainable outcomes. Some scientists have argued that humans are not the only animals that have values (see Bekoff and Pierce 2009), but what is undeniable is that humans use constructed values to achieve mutual goals more than any other species.

In his talk, which is boldly entitled “Science can answer moral questions“, Harris tries to build the case that scientific understanding of what ‘is’ can help prescribe what ‘ought’ to be. He claims that as conscious organisms, we ought to be able to rationally deduce a set of universal morals based on scientific understanding of how our world works. In starting his talk this way, Harris joins a long line of philosophers and scientists who view human consciousness as our key innovation, the adaptation that has allowed us to move above and beyond other organisms. In my eyes this is a flimsy edifice upon which to build a debatable argument. We really do not know how ‘conscious’ other organisms are, although there are many signs of consciousness in our close relatives. Can we be so sure that it is our consciousness that makes us unique? Could it be that other traits — such as our extreme sociality and penchant for cultural exchange — actually underlie our success? To what degree is our consciousness aimed at deducing truths versus making optimal behavioral decisions based on incomplete and local information? Harris addresses none of these issues, which undermine his initial assumption.

The talk does not go too much farther before its logic completely unravels. Without even acknowledging that he is doing so, Harris constructs his own arbitrary value: human well-being. He begins talking about “how human communities flourish”, and suggests that there is a “moral landscape” across which human well-being is not uniform: some moral values produce greater overall human well-being than others. He uses several examples (which not at all coincidentally lionize secularism and demonize religion) to demonstrate that human well-being is not the outcome of all moral systems.

Jumping off of his moral landscape concept, Harris takes on moral relativism, suggesting that if we value human well-being we cannot possibly say that all cultures are morally neutral. Comparing the way women are treated in extremist Islamic cultures with the lot of women in western democracies, he challenges us to say that one culture produces the same outcomes for women as others. He suggests that if we cannot trust the Taliban on physics, then we cannot trust them on morals. The audience cheers as he calls for a “universal morality” aimed at improving the lives of all humans. In case we missed the other blunt-trauma examples, Harris asks us whether the values of human rights activist The Dalai-Lama and mass murderer Ted Bundy produce the same outcomes in terms of human well-being.

At this point, you may be wondering what my problem with Harris is. Don’t I support the improvement of human well-being through well-chosen values? Don’t I prefer the way that women are treated in the United States over the way that they are treated in Saudi Arabia? Don’t I think that the Dalai Lama is a better guy than Ted Bundy? Don’t I share Harris’ values?

Based on what he lays out in this talk, I do share Harris’ values when it comes to human well-being. But his intellectual and scientific values?

That is where we depart.

We are not completely at odds about science: from what I can tell, Harris believes that scientific understanding should be used to help achieve the goals implicit in our moral values. I can get behind that idea. But Harris is so confused about what is science and what is morality that he effectively blunts the power of science, joining the rest of the ascientific culture warriors in the semi-random and frequently-destructive process of design-free cultural evolution. He is a meme warrior and not a scientist, and this is polemic not scholarship.

Have you noticed how Harris leads with a particular value, human well-being? Where did that value come from? Is it, as the title of his talk seems to promise, derived from scientific understanding? Or is this Harris’ own (albeit widely-shared) value? I know of no scientific discoveries that suggest that the most successful culture will be the one that values the overall well-being of humans. In fact, an easy way to understand some of the cultures that Harris detests is to understand that their success comes from very different values than “overall human well-being”. For instance, the overall outcome that seems most valued by fundamentalist religion is not the overall welfare of its adherents, but the continued existence of the religious culture itself. Strangely Harris chooses not to pick on western capitalism, another cultural complex that produces outcomes that are far from Harris’ value of well-being for all. The outcomes produced by both western capitalistic and fundamentalist religious cultures clearly represent different values than Harris’, and there is no way for science to tell us which value we ought to maintain.

Essentially, Harris’ talk is a classic case of the naturalistic fallacy’s evil twin, what I would call the ‘ought-is problem’. He starts with what he values (what ‘ought’ to be) and then constructs his version of what ‘is’ to support this value. In doing so Harris joins a massive group of historical and contemporary ideologues who distort, bend, and misrepresent scientific fact to prop up their arbitrary ideas of what ought to be.

So what does science have to say about achieving valued outcomes? Can it tell us what cultural values are more likely to catch on with people? Can it predict cultural winners and losers? Can it guide us in achieving our own valued outcomes? I believe that the answer to all three of these questions is “yes”. Once you have arrived at what you want to achieve, science has the potential to at least inform your attempt to transform your desires into reality. We see this clearly in design fields. Working with a client, an architect creates a design around certain aesthetic and functional values. It is up to the engineer — trained in science — to make this design reality. Engineers turn imagined ‘oughts’ into real objects through their understanding of what ‘is’. The ‘is’ of the world we live in constrains design: tell your engineer that you want to build a three-story building out of corrugated cardboard and she will tell you that such a design will not work. But because of robust discoveries in physics and materials science, a wide array of architectural outcomes can be achieved. The science devoted to understanding human cultural evolution is not as advanced as physics or materials science, but it still has a lot to tell us about how to achieve our cultural goals.

Harris ignores almost all of this science. When I first saw the title of his talk, I assumed that he would talk about our evolved sense of justice. Frans de Waal’s Age of Empathy (de Waal 2009) provides an excellent overview of how human and non-human animals have genetic predispositions towards maintaining fairness. As mentioned above, Marc Bekoff and Jessica Pierce argue in their book Wild Justice (Bekoff and Pierce 2009) that moral values are as old as animal sociality itself. And a whole host of behavioral economics experiments suggest that humans are endowed cross-culturally with a unique set of pro-social instincts. Strangely, Harris does not seem to be thinking about the constraints and predispositions imposed by our evolved psychology at all. In not aligning his valued goal of equitable well-being with our evolved instincts, he misses the one chance to argue that there is actually some ‘is’ in the things we feel ‘ought’ to be, and places himself alongside all the other folks who foolishly believe that humans are blank slates ripe for infection by any cultural value.

Harris also seems to be oblivious to the fact that culture evolves. As such, culture itself is as blind to morality as any other evolved trait. Those cultures which are most successful at passing on their cultural values will be the ones that survive. That is the only rule of cultural evolution, and it can be used to explain why some cultures flourish while others do not (Harris’ initial stated goal). An understanding of cultural evolution could help establish Harris’ “moral landscape” indicating where overall human well-being is greatest, but this landscape would be useless unless placed alongside a landscape showing the cultural fitness of different values. Only where a peak indicating high human welfare on Harris’ “moral landscape” matches up with a peak on my “cultural fitness landscape” will he achieve his valued outcome.

Harris’ third scientific oversight revolves around group selection. Once relegated to the trash heap of evolutionary theory, group selection is now recognized by a significant number of scientists as a potential mechanism by which human cooperation can evolve (Bowles and Gintis 2011). Successful cooperation is what makes particular cultures more successful than others, and arguably successful cultures have been those that could increase their scale of cooperation above other cultural groups. A hallmark of successful group selection for cooperation is that it suppresses within-group selection so that between-group fitness is maximized. This suppression of within-group selection does not have to be totalitarian or coercive, but it could be. If Harris wants to maximize overall human well-being, he needs to come up with values that maximize the fitness of his cultural group. Cultural fitness? Competing cultural groups? Harris’ talk scarcely hints at these topics.

To demonstrate how Harris’ poor use of science limits his potential for achieving his goals, let me re-examine one of the cultures he objects to in his talk: fundamentalist Islam. Harris goes to great lengths point out how terribly women suffer in fundamentalist Islamic cultures; this suffering is at odds with his values, and it happens to be at odds with my values. But if our mutual goal is to end the sort of misogynist oppression perpetrated by these cultures, we had better figure out why they are so prevalent. In the case of fundamentalist Islam, we can look at this cultural complex in terms of how it interacts with evolved social instincts, how well it spreads culturally, and what group competitive advantages it confers. Harris decries the oppressiveness of the burqa, but is its prevalence facilitated by the uniformity with which it is applied to women in these cultures? Are the beliefs at the root of radical Islam particularly catchy, and thus more likely to spread? And do the values of fundamentalist Islam — even as they oppress large numbers of individuals — make it a cultural group that is particularly good at competing with other groups? These are the questions we should be asking if we want to undermine a culture and its values because we find the outcomes it produces abhorrent.

Notice how coming up with testable scientific hypotheses about why a particular culture is successful is completely different from deciding whether we value the outcomes produced by that culture. But in gaining some scientific understanding of cultures we wish to compete with, we gain an advantage in achieving our own valued outcomes. In order to achieve our ‘ought’, the culture we advocate must be informed by what ‘is’.

In the end, Harris becomes the butt-end of his own joke. He and I and many others may value human well-being, but how to get there? In Harris’ world, the way to get there is to convince other people that science is on your side, that scientific understanding favors the outcome that we commonly desire. In doing so, Harris actually ignores some basic well-understood concepts of evolutionary biology. This means that Harris, while advocating what ‘ought’ to be, is ignoring what actually ‘is’. We can hope that the values he constructs along the way are effective at achieving the outcomes he values, but without science to guide him his attempts at using cultural values to achieve valued outcomes amount to trial-and-error (just like other evolutionary processes that suffer from no design or poorly-informed design). What I am suggesting is that the real role for science to play comes after values have been constructed. As I indicated above, I do not think that our values are completely arbitrary cultural constructions, but inasmuch as they are, we can achieve them effectively only if we employ the best science available. Science tells us what ‘is’ — it defines obstacles, limitations, and challenges to achieving the outcomes we value — and thus provides us with the ability to predict what cultural values are most likely to survive. The question is not whether science favors a particular cultural complex over another, but which cultural complexes will more effectively harness scientific understanding to achieve their valued goals. Sadly the science devoted to understanding both cultural evolution and group selection — the key tools needed to predict how successful competing cultures will be — is presently under-developed. Using the basics of evolutionary theory we can see that some cultures will be more stable than others, but we need a lot better science in order to truly aid the quest for “better overall human well-being”. But if we do not pursue and use this scientific understanding, we have to pin our hope on cultural trial-and-error. Hearing Harris’ rhetoric, this leaves me worried.

Rather than trying to paint itself as a wholly different culture based solely on scientific rationalism, it seems to me that the modern atheist movement would present itself in a much more honest and accurate way if it admitted that it was just another culture. Being “just another culture” does not preclude being a more successful culture, and the success of the modern atheist movement will rely on having better values and better outcomes. What is “better”? In an evolving system of culture, better values and better outcomes are intertwined: your values need to be appealing to more people than competing value systems, and these values have to lead to better social and economic outcomes than competing value systems. It is really strange how self-avowed evolutionists spontaneously forget all the evolutionary principles they know when it comes to questions of human values, joining the fray with other cultural complexes that are not informed by scientific understanding. Would Richard Dawkins use an argument of moral superiority to explain why a cheetah runs fast? If not, why is Sam Harris trying to argue that the universe favors his morality irrespective of its fitness or stability?

I had heard a lot about Harris indirectly, but this talk was my first direct exposure to his work. As you might have picked up, I am not impressed. For someone aligned with science and sympatico atheist evolutionary biologists, he seems to be missing some major evolutionary concepts. Given how scientifically weak his arguments are, why has he ascended to the podium of the neo-atheist movement? Well, based on his performance in this video and the reaction of the audience, Harris’ success is based on his humor and wit. He is clever in the way he weaves together his story, and he knows how to drop a funny line (check out the part of this video where he manages to call himself the “Ted Bundy of string theory”). Funny guy, but pretty light on deep analysis. That he has pretty high cultural fitness (in other words, he is famous) is kind of ironic, because like the religions he criticizes he is much more catchy than he is insightful. At best, he is an entertaining hack. At worst, he is a sideshow barker who distracts his audience from a more nuanced and informed assessment of the real problems we face in constructing a moral system that maximizes human well-being.