Intuition and Philosophy

If anything can be pursued in an armchair, philosophy can," the esteemed Oxford philosopher Timothy Williamson told the Aristotelian Society, of London, a few years ago. That may sound like an innocuous truism: No one pictures Bertrand Russell doing his philosophical cogitation anywhere but in a club chair, or perhaps in bed, postcoitally (given his adventurousness in that arena). But, in fact, Williamson's remarks are fighting words these days, thanks to the rise of a cohort of philosophers who believe that the armchair arguments of philosophers need to be probed and tested through surveys of ordinary people and laboratory experiments using human subjects. If philosophers want to demonstrate that their arguments comport with how the mind really works, say the proponents of experimental philosophy, they need to get off their duffs.

Does that sound like an incendiary charge? Indeed, an armchair in flames has become the informal symbol of the experimental-philosophy movement, also known as "x-phi." Online you can buy a burning-armchair T-shirt or watch a burning-armchair YouTube video, accompanied by the x-phi alt-rock anthem. The aggressive symbol is only partly tongue-in-cheek: Some experimental philosophers believe they are simply augmenting and supplementing traditional philosophical work, but others view themselves as overturning a significant number of philosophical projects. They think they are calling into question methodologies that philosophers have made use of "for 2,400 years," as the Rutgers philosopher Stephen Stich, a pioneer of x-phi, puts it.

Experimental philosophy has percolated on the edge of the discipline for several years, yet remains dogged by this question: Will it mature into a central, respected strand of philosophy or remain a semi-fringe endeavor, viewed from the mainstream as the kind of work that should be done (if at all) in psychology departments — with the results perhaps pondered later by "real" philosophers?

At the heart of experimental philosophy lies a suspicion of so-called "intuitions." An intuition in philosophy is something far more potent than it is in ordinary discourse. Intuitions rear their heads when philosophers write such things as, "In this case, we would surely say …," or, "It would be natural to say …" (for example, that killing a man to harvest his organs is wrong). It is a deeply rooted sense, tested from multiple angles and honed through thought experiments and dialogue. The trustworthiness of intuitions (whose roots can be traced back to Plato and Socrates, who thought they represented glimpses of the true, ideal world usually hidden from us) hardly goes undebated by traditional philosophers — quite the opposite — but the experimental philosophers apply a new kind of pressure. They think that by studying human minds, using empirical techniques, and drawing on the insights of modern psychological science, they can get a better sense of where intuitions come from, and whether or when they should be granted credence.

Experimental philosophy has suggested, for example, that people from East Asian cultures may have different intuitions on very basic philosophical questions — reference (what nouns refer to in certain situations), morality, epistemology (what it means "to know" something) — than members of Western societies do. Experimental philosophers also draw on work by contemporary psychologists demonstrating just how malleable human cognition is, how easily redirected and reshaped it is by external cues, even as the conscious mind remains blissfully unaware. Opinions on crime and punishment, for instance, can be altered by placing people in a dirty room designed to trigger feelings of disgust: Subjects in such experiments respond more punitively when asked what should be done to certain hypothetical criminals.

"If we keep getting the same kind of results with the right kinds of controls and right kind of experiments," says Stich, "then there is a problem with the central method that philosophers have used throughout the 20th century, and for a long time before that": the reliance on armchair intuitions.

Understandably, such claims have met with resistance. "A philosophical problem is not an empirical problem," writes Judith Jarvis Thomson, the noted MIT moral philosopher, in an e-mail message to The Chronicle, "so I don't see how their empirical investigations can be thought to have any bearing on any philosophical problem — much less help anyone to solve a philosophical problem."

When several philosophers, including Stich and Joshua Knobe, an assistant professor at the University of North Carolina at Chapel Hill, along with a few psychologists, including the University of Virginia's Jonathan Haidt, proposed to Oxford University Press a new journal focusing on empirical studies of moral philosophy, they got back one particularly scathing anonymous review: "This group," it said, "is overly impressed by dubious functional MRI studies purporting to demonstrate the neurophysiological underpinnings of moral thinking, and by small sample, 'rinky-dink experiments' conducted by philosophers who are not trained experimentalists."

While much of the proposal authors' work was "perfectly philosophically respectable," the reviewer said, "a great deal of their interest lies in what I can only describe as the desire to eliminate morality (or at least the study of morality) from the discipline of philosophy itself." Other reviews were more favorable, although Oxford has not made a decision about the journal.

There are signs, however, that experimental philosophy is coming into its own. One was the publication in June of Experimental Philosophy (Oxford), a greatest-hits anthology edited by Knobe and Shaun Nichols, of the University of Arizona. What's more, the National Endowment for the Humanities recently awarded Nichols and Ron Mallon, of the University of Utah, $180,000 to invite some top experimental philosophers to Utah next summer to present the basics to professors unacquainted with the subject. The University of St. Andrews, in Scotland, held a conference in September at which Stich and Williamson were to argue for and against the armchair, while the University of Cologne, in Germany, sponsored a gathering last month titled "The Armchair in Flames?" And the chair of the American Philosophical Association, Kwame Anthony Appiah, of Princeton, has written about x-phi as a positive development.

X-phi'ers were especially pleased that Caltech mentioned experimental philosophy as a subfield of interest — an apparent first — when it recently posted an ad for a tenure-track job in philosophy. "I just came out of grad school in 2005," says Thomas Nadelhoffer, who teaches at Dickinson College and runs the influential x-phi blog Experimental Philosophy, which has some 70 contributors. "We were universally told back then to do the best we could to hide that we do this stuff. Now people wear it on their sleeve."

In "An Experimental Philosophy Manifesto," the opening chapter of Experimental Philosophy, Knobe and Nichols say their critics would be right to reject the approach if the parody version of it reflected reality: But no one, they say, believes that philosophical questions — What acts count as "intentional"? Is it permissible to kill one innocent person to save five? — can be resolved by studying the answers ordinary people give. Rather, Knobe and Nichols argue, "The suggestion is just that, whatever else we do, we should also be looking at people's intuitions … as a way of coming to a deeper understanding of how the human mind works."

If Stich is the provocateur of the movement, upending 2,400 years of philosophizing with a single blow, Knobe is the diplomatic young star, emphasizing in a gentle voice that he merely wants philosophers to embrace "new tools" and that, for the most part, the reception of x-phi has been "a testament to the community of philosophy, and how open they are to different methodologies." A precocious 2006 Ph.D. from Princeton, he'd done some postgraduate work in a psychology lab at the University of Oregon under Bertram Malle, before enrolling in grad school. He finds himself drawn, he says, to the sorts of questions philosophers debated until the 19th century, as opposed to those of the 20th century, when the discipline got diverted into narrower, logic-chopping channels.

"From the beginning," he says, "Nietzsche was inspirational to me. I was interested in working on the kinds of questions he was interested in: how people make moral judgments, how moral judgments affect the way we understand our world." But while most Nietzsche scholars wrote essays on the philosopher's work, Knobe continues, "I was interested in doing experiments to answer the questions that he asked." Indeed, Knobe has argued, in a paper written with the University of Chicago's Brian Leiter, that Nietzsche deserves a spot alongside Aristotle and Kant in the pantheon of moral philosophers — in fact, above them. That's because Nietzsche argued that morality is the product of neither a virtuous upbringing, focused on creating firm character, nor the pure application of reason, but comes from individual, innate, sometimes unalterable attributes of humans — the sort of traits experimental philosophers like Knobe now pursue.

Knobe's best-known papers demonstrate how ordinary people's understanding of such things as intention, which philosophers have traditionally treated as value-neutral (one either tries to do something or doesn't, independent of the moral content of the act), is in fact shot through with moral considerations. Consider this scenario, used in numerous scholarly articles: A corporate chairman is presented by a vice president with a proposal for a new project. The VP explains that the project will increase profits but hurt the environment. The chairman replies, "I don't care at all about helping the environment. I just want to make as much profit as I can. Let's start the new program." They do, and predictably the environment is harmed.

Did the CEO intentionally damage the environment? In one of Knobe's most cited studies, some 82 percent of students said yes.

Then take this scenario: Same chairman, same VP, but this time the VP says the program will help the environment. The CEO, again, replies that he doesn't care; his only concern is money. He gives a thumbs-up and, again, as predicted, the environment is helped. This time only 23 percent of students say the CEO intentionally helped the earth, although the scenarios are logically identical.

That people's judgment about intentionality is shaped by the degree of harm done by the act being evaluated, also known as the Knobe effect, has been demonstrated in subjects from India, among 4-year-olds, even among people who suffer from deficits in their emotional processing. Some traditional philosophers had suggested in passing — using their intuition — that moral judgments might be shaping judgments about intention, but Knobe offered the striking data.

The study of intentionality is not purely ivory-tower stuff. Every time a U.S. bomber kills civilians in the course of combating the Taliban, the laws of war demand that the military investigate whether the pilots and their commanders intentionally caused the deaths and should be punished.

Knobe has found this intertwining of moral judgment with analytic assessment across a range of human thought: when we try to ascertain what people truly "value," for example, and when we judge people who have acted in situations in which they had little choice but to do what they did. (Presented with a hypothetical deterministic universe, in which every action can be predicted in advance, test subjects still say, oddly, that they would hold wrongdoers responsible.)

One could argue that the asymmetry in judgments about intentionality amounts to simple error, just as clearly as when people flip heads nine times in a row and then say they are "due" for a tails. But Knobe says it makes more sense to say that "intention" and "moral blame" cannot be neatly disaggregated without doing violence to the way the mind functions (as opposed to how philosophers think it should function). "From the beginning, your whole process of figuring out what is going on in the world is saturated with moral judgment," he says.

Other philosophers, even those of an experimental bent, are quicker to say that the test subjects are mistaken. Nadelhoffer, for example, in an essay in Experimental Philosophy, argues that courts may want to make jurors aware of this psychological puzzle, lest defendants who cause harm unintentionally (under the traditional understanding) receive punishments better suited to purposeful, coldblooded criminals.

One way that researchers have followed up Knobe's work is to parse even more finely the answers people provide in these surveys. Adam Feltz, a philosopher at Florida State University, and Edward T. Cokely, of the Max Planck Institute for Human Development, in Berlin, recently found that subjects' answers to the chairman question correlated with personality traits: Extroverts were more likely to say that the CEO who harmed the environment did so intentionally than introverts were. Extroverts, being more socially minded, may be more primed to punish those who violate societal norms.

In fact, while Knobe speaks of pinning down "folk philosophy," or how ordinary people make sense of the world, the latest developments seem to underscore just how many folk philosophies there are. Alfred Mele, of Florida State, has investigated the Knobe effect and identified a multitude of subgroups, each of which responds to the scenarios in a distinct way, worthy of analysis. Although he's a sometime participant in x-phi, and hardly an enemy, he's still not sure where this work is headed. "I think the project of finding and understanding folk concepts of this and that is interesting," he says. "But it won't settle a lot of deeper philosophical questions."

That was precisely the point that New York University's J. David Velleman made in one of the most-famous takedowns of experimental philosophy. After Slate published a brief article about the trend in 2006, Velleman said (on the now-defunct Web site Right2Left.com) that many of the so-called x-phi discoveries "are not exactly news to traditional philosophers."

"Of course," he went on to say, "it's useful to know what most people think about intentional action and moral responsibility. In philosophizing on these topics, we can't stray too far from what people think. … Even Aristotle relied on endoxa — received opinions — as a starting point of his inquiries.

"Maybe Aristotle was the first 'experimental philosopher,' then? No. Aristotle knew that the real philosophizing starts after the endoxa have been surveyed. His view remains true today."

Today Velleman says he does not want to be painted as an enemy of experimental philosophy, or to be interviewed on the subject. But via e-mail, he affirms that he is dubious about whether philosophy departments ought to be training students in experimental methodology (none currently does so systematically), given limited resources. Still, he says, he is willing to let the marketplace of ideas settle the question.

X-phi'ers now tend to pick up their experimental chops by wandering over to psychology or cognitive-science departments, collaborating with psychologists, or working with someone already part of the movement. Sean Kelly, the first philosopher with a laboratory at Harvard since William James (Kelly studies visual cognition), suggests that they're working in a Jamesian spirit. When James wanted to learn about mammalian neuroanatomy, he went out, bought a sheep's head, and started sawing and slicing. Similarly, Kelly says, young experimental philosophers are picking up whatever statistical and other tools they need, any way they can. "You shouldn't be afraid to do something because you haven't got a professional degree in it," he says.

Thanks in part to its experimental wing, philosophy is becoming a less-insular discipline. Granted, its insularity can be overstated; bioethicists regularly interact with physicians. Yet many scientists and social scientists find philosophy to be a hermetic field, and experimental philosophy opens a door. One meeting place is the Moral Psychology Research Group, founded in 2003, which includes academic psychologists like UVa's Jonathan Haidt, Richard Nisbett (of the University of Michigan), and Josh Greene (Harvard), as well as empirically minded philosophers.

"Something that has always frustrated me with reading philosophy is that it's a mixture of logical inference and just description of intuitions," says Haidt. "And often my intuition would differ from that of the philosophers. It would seem like a step from tight link to tight link. but one-quarter of the links were up for grabs." Now that the x-phi'ers have embraced experimental techniques, however, they "are producing work of interest to our side of the fence as well."

Haidt, who studies the balance between intuitive or emotional responses and reasoning in moral decisions, has come up with the following x-phi-like results: Low-income Americans are more likely to think there is something seriously wrong with cooking and eating the family dog, should it be hit by a car, than are high-income Americans (who may be overriding their revulsion with the reasoned argument that no one is hurt by the dinner). More recently he has explored the differences in moral reasoning between liberals and conservatives, finding (unshockingly) that conservatives are more interested in principles of loyalty and purity than liberals are. Significantly more conservatives than liberals, for example, say they would never accept a transfusion of blood from a child molester, no matter how much money they were offered.

And John Doris, a philosopher at Washington University in St. Louis and another member of the moral-philosophy working group, has drawn on findings from social psychology to question the very underpinnings of Aristotelian "virtue ethics." Perhaps no theory that relies on stable, stout character can survive once psychologists have shown that even seminary students who have just heard a sermon on the Good Samaritan will step over a man in need if they are in a hurry.

A key question for experimental psychology is how it proposes to get from descriptive findings to philosophically sound conclusions about what we ought to do, or what ought to count as sound knowledge.

So far, it appears, the discipline tends to be better at tearing down such conclusions, or trying to. Mallon, Nichols, Stich, and Edouard Machery, of the University of Pittsburgh, for example, have taken aim at one of the most-famous philosophical theories regarding "reference," which comes into play every time people open their mouths to speak. Only a few decades ago, most philosophers adhered to the theory that a term refers to a person or thing when it offers the best fit with the qualities or properties of that person or thing.

The philosopher Saul Kripke, of the Graduate Center of the City University of New York, upended that theory, arguing that for a reference to be successful, it merely needed to be part of a chain of comprehensible usages of the term, passed from person to person. No one in the chain need know any of the descriptive details about the figure in question — which is why people who could not relate a single detail about Cicero's life can still use his name and be referring to the same historical figure a classical scholar knows intimately.

In one thought experiment, Kripke asked readers to think of Kurt Gödel, the logician who devised the numerical "incompleteness theory." What if an unknown figure named Schmidt had, in fact, created that theory, and Gödel had stolen it? Once we knew that, if we said "Gödel," would we then be referring to Schmidt, who, after all, now fit the description of Gödel better than Gödel himself? No, Kripke argued, we would still be referring to Gödel. Case closed, most philosophers have agreed.

In surveys, however, Machery, Mallon, Nichols, and Stich found that while American college students tended to respond to the thought experiment as Kripke predicted, English-speaking students in Hong Kong answered quite differently. A majority said "Gödel" would actually refer to Schmidt. A substantial minority of American students also agreed with that intuition.

Why, Stich and his collaborators conclude, should we privilege the intuitions of Western philosophers, or even Westerners in general, over those of Hong Kong students — and perhaps over those of most Asians? (His thinking has been influenced by the work of the psychologist Richard E. Nisbett, who has argued that Asians and Westerners categorize their observations of the world quite differently.) "I'm not going to hold my breath," says Stich, "while you come up with a theory why a third of the world's population is wrong on this question."

Commentary on this paper and related ones by the same authors has been flourishing. In one draft paper, CUNY's Michael Devitt offers several objections. One is that the Gödel example is a minor one among many that Kripke offered to defend his theory. Devitt also argues that, in the end, the question remains: Which theory of reference best explains the way humans use language — a question that, he suggests, may yet require some philosophical training and argumentation to resolve. As Kirk Ludwig, a professor of philosophy at the University of Florida, put it last year, commenting on the same papers, "[We] must ultimately return to the first-person approach to achieve the goals we set ourselves in philosophy." In other words: Back to the armchair.

At a recent gathering of the Society for Philosophy and Psychology, Brian Scholl, a Yale psychologist who studies visual perception, raised an even more fundamental objection to the kind of work typified by many x-phi papers: They are overreliant on surveys, which are unreliable precisely for the reasons experimental philosophers notice in other contexts: The people who take them can be swayed by all sorts of un-pin-downable factors. Although he thinks much work in x-phi is "fantastic," he wrote in comments posted on the Experimental Philosophy blog this month, "I worry that these sorts of methods end up telling you less about the psychological processes that underlie the philosophically relevant intuitions — and more about the psychology of how people cope with being asked strange questions."

Expect the experimental philosophers to argue back, adjust some of their approaches in response, and soldier on. They say they are well past the point where an objection or two can cause the movement to implode. "This is one of the best parts of experimental philosophy," says John Mikhail, who teaches at the Georgetown University Law Center, and whose work translates people's complex responses to moral thought experiments into algorithms. "Younger people are not taking 'no' for an answer."