19 May 2010

Gödel in Hong Kong

Two books on “x-phi”

  • Joshua Knobe and Shaun Nichols (Eds.). Experimental Philosophy. Oxford University Press. July 2008.
  • Kwame Anthony Appiah. Experiments in Ethics. Harvard University Press. January 2008.

Appeal to intuition has long been one of the core tools, if it can be called that, of the philosophical method. It is intuition, understood as that immediate operation of the mind by which knowledge is obtained without either observation of the world or inference from premises, that both distinguishes the work of the philosopher from that of the scientist, and motivates the familiar accusation that philosophy is a mere “armchair” discipline. Even those philosophers thick-skinned enough to ignore this accusation tend to recognize one deep problem with excessive reliance on the evidence of intuitions: any given intuition, considered in isolation, is only as reliable as the person who has it. But how can we determine that? One obvious way would be to check the intuition against several other intuitions. But then, inevitably, philosophy finds itself drifting into the territory of the social sciences, something the majority of philosophers steadfastly refuse to let happen.

This still very comfortable majority has, in the past few years, come under attack by a small cadre of professional philosophers who have dared to engage openly in the heretical practice of empirical inquiry. Their movement, which has come to be called “x-phi” by some of its adherents, proposes to create an experimental branch of the discipline that will challenge the armchair intuitions with which most philosophers have been content to work, by presenting empirical data showing the extent to which laypeople disagree with these intuitions.

As a historian of 17th-century philosophy, news of what’s hot among my ahistorical colleagues sometimes reaches me only with significant delay (sometimes I even start to imagine that I am a 17th-century philosopher, and that problems such as the nature of God or the sympathy between weapons and wounds are real ones, worthy of real solutions). Until recently I was so out of the loop that when I first heard mention of “experimental philosophy,” in late 2008, I thought to myself, well good, scholars are finally taking an interest in the work of Robert Boyle, Margaret Cavendish, and other great 17th-century philosophers who described their work by this same name. Cavendish for example is the author of Observations upon Experimental Philosophy, published in 1666, a work that touches upon a wide variety of questions, including the problem of the reproduction of flies, the possibility of the transformation of wheat into barley, and so on. For her, as for Boyle, there was simply no worry about the relationship between philosophy and science, since philosophy included science. Until the rise of university-based philosophy in the late 18th century, particularly in Germany, philosophy just was whatever it was that curious people took an interest in, and that helped them, even if only just a bit, to piece together the massive puzzle of the world.

Now, this might seem an exceedingly loose definition of “philosophy.” But the historical record could not be clearer: from at least the Renaissance until sometime in the 18th century, philosophy really was everything, or at least everything that involved “inquiring into things below the earth and in the sky,” to cite the famous accusation against Socrates in the Apology. This is the meaning implied in the title of the journal of the Royal Society of London, the Philosophical Transactions, also founded in 1666, and featuring until this very day, under its now archaic title, articles on everything from the larynx of chimpanzees to the existence of God.  So experimental philosophy is not new, even if “x-phi” is. (In his fine book, Experiments in Ethics, rightly considered to be one of the better measurings-up of the new movement, Kwame Anthony Appiah graciously acknowledges this earlier chapter in the history of experimental philosophy: this debt, one might say. More on this later.)

Cavendish and Boyle had their own battles to fight, to be sure, yet no one doubted that what they were doing counted as philosophy. How things have changed! Far from being the new normal science, x-phi in fact has only captivated a relatively small minority of reputable academic philosophers, who still have to go to great lengths to justify their very existence. A rather nervous self-justificatory pleading is on full display in Knobe and Nichols’ edited volume, the chapters of which, as promised by one of the book’s several blurbs, are sure to “inspire some philosophers and enrage many more.” Yet one might reasonably wonder whether we are not dealing with a somewhat manufactured controversy. X-phi is packaged as revolutionary, while its defenders seem eager to offer constant reassurances of its “continuity with traditional philosophy.” One cannot help but think of the teenager who cuts loose a bit, perhaps sporting the insignias of a radical party, or the symbols of “Eastern” spirituality, but never going so far as to call his right to the family inheritance into doubt. X-phi wants to stay within the comfortable bounds of mainstream Anglo-American philosophy, while also claiming to subvert many of its basic principles. It offers assurances of its continuity with traditional (i.e., 20th-century Anglophone) philosophy, while at the same time insisting upon its newness. Branding the movement as “x-phi” certainly makes it look new, anyway, in much the same way that the X-Games are new, while plain old games have been around forever.1

Is the philosophical program that emerges out of this double movement, between tradition and innovation, even coherent? It seems to me that x-phi simply cannot decide what it wants to do. On the one hand, it is attracted by the “striking range” of “doxastic diversity” out there in the world, and its practitioners seems to think that this diversity is in itself a good thing and worthy of study. But in the end it returns to the old philosophical yearning, vastly preceding the Anglo-American tradition or the “Cartesian” armchair tradition, and indeed going back at least to Plato, to arrive at the singular truth. Thus the limited interest of x-phi in “Eastern philosophical notions” is motivated by the same basic drive, as Knobe and Nichols explain, that leads “some Christian children to think there’s no rational basis for preferring Christian to Hindu beliefs” (again, x-phi as moderately rebellious teenager); so too, the editors continue, we might discover “that there’s no rational basis for preferring Western philosophical notions to Eastern ones.” In other words, if we take an interest in “Eastern” philosophical notions, this is because we are considering adopting them, or at least making enough of a show of considering adopting them to worry our parents a bit.

Some human-scientists have come up with rather more interesting questions about the diversity of beliefs out there in the world. Even Aristotle, come to think of it, understood that “it is the mark of an educated mind to be able to entertain a thought without accepting it.” After Aristotle, some have gone even further and held that it is the mark of an educated mind to be able to entertain a range of thoughts without looking to accept any of them. But the experimental philosophers are in the end only willing to venture into the range and diversity of beliefs as a step along the path towards warranted belief. “First,” Knobe and Nichols explain the procedure, “we use the experimental results to develop a theory about the underlying psychological processes that generate people’s intuitions; then we use our theory about the psychological processes to determine whether or not those intuitions are warranted.” Yet, again, there seems to be a degree of indecision, and on the following page the editors acknowledge that “even if we discover important differences between the philosophers and the folk, it would hardly follow that data from the folk are irrelevant. Rather, the whole pattern of the data might tell us something important about the ultimate source of the philosophical problems.”

What do Knobe and Nichols mean by this? Do they mean that a global survey of folk intuitions about x might yield something more interesting than an answer to the question, ought I believe that x? But if that is what they mean, why on the previous page do they claim that in the end x-phi is all about determining which intuitions, whether from folk or philosophers, are warranted? I think x-phi has not yet decided, and I think it is unable to decide because it wants to affirm its allegiance to its immediate forebears, even as it vaguely glimpses the tremendous interest of the delirious range of things that the folk out there in the East and elsewhere believe.  


The new experimentalists are in a bind: they want to keep philosophy’s borders as sharply defined as do their immediate armchair-bound forebears, yet unlike their forebears they recognize that philosophy needs real data in order to make real progress. Their solution is to create an experimental wing of philosophy from scratch, rather than simply to acknowledge their deep interdependence with other human, social, and natural sciences. Knobe and Nichols do note at one point that “at the turn of the century, anthropologists provided a catalog of the striking cultural diversity in moral views.” But they do not explain why they are not willing to just start valuing the kind of research that those anthropologists were doing at the turn of the century (nor do they make clear which century they have in mind), by taking up a place in a fruitful, interdisciplinary, human-scientific endeavor. 

For the new experimentalists, philosophy cannot stop at mere cataloguing, which is where the turn-of-the-century anthropologists are presumed to have stopped. “If philosophers gave up all other forms of thought and just spent all of their time running experiments,” Knobe and Nichols explain, “it really is true that they would never get anywhere.” But if we understand “experiment“in the broad sense to include “observation,” then indeed this is exactly what early modern experimental philosophers did, including Boyle and Cavendish, as well as Francis Bacon, who implored his contemporaries to “lay your notions by” and to just start cataloguing the particulars.

It should be clear by now that to the extent that I wish to attack x-phi, I wish to do so at its unguarded flank: the bulk of the criticism is coming from those who think philosophy must not open up to experimental methods at all, and so it is against this criticism that the new experimentalists have built up their defenses. I do hope the experimentalists win on that front. There is much to admire in the basic principles of the movement. As concerns the movement’s actual accomplishments, that is another question altogether.

At the end of their “manifesto” (i.e., the book’s introduction), Knobe and Nichols rightly express their doubt that “general considerations can provide any ultimate justification to sustain experimental philosophy. The real measure of a research program depends on whether the program generates exciting new discoveries.” The papers they present are meant to showcase x-phi’s ascendancy, but the actual results described are meagre indeed. By my count five out of twelve of the essays (including the opening “manifesto”) amount to viable experimental philosophy. The others consist either in interesting reflections on “the future of experimental philosophy,” or in bad science.


The most problematic contributions in the volume are the two papers in the section on “Cross-Cultural Differences in Intuition.” In the second of these, “Semantics, Cross-Cultural Style” (reprinted from a 2004 article in the journal Cognition) Edouard Machery, Ron Mallon, Shaun Nichols, and Stephen P. Stich present their experimental finding that the intuitions of East Asians (dubbed “EAs”) on the one hand and Westerners (“Ws”) on the other support different theories of reference. Because EAs are believed to think more “holistically,” and Ws more analytically, Machery et al. set out to determine whether these differences might underlie differences in semantic intuitions, and find to their own satisfaction that they do.

EAs, it turns out, tend to prefer a “descriptivist” account of reference that fits better with their holistic Weltbild, according to which the referent need not satisfy the associated description. To summarize the example preferred of philosophers ever since Saul Kripke came up with it in 1972, and indeed the example administered to the comparison classes of undergraduates in Hong Kong and in New Jersey by Machery et al.: suppose that, contrary to universal belief, it is not Kurt Gödel, but rather his associate, “Schmidt,” who discovered the incompleteness of arithmetic. If this is so, then every time we speak of “Gödel’s incompleteness theorem,” we are in fact speaking of a theorem discovered not by Gödel at all, but by Schmidt. To whom, then, does the name “Gödel” refer? A defender of the “causal-historical” view (that is, an opponent of the descriptivist view) thinks it must refer to Schmidt, whether we know it or not, while a descriptivist theorist accepts that it can refer to Gödel, since all that matters is that the referent figure in the causal history of the speaker’s current use of the word, and not that it satisfy the description.2

We now know, as a result of Machery et al.’s study, that English-speaking undergraduates in Hong Kong prefer the descriptivist theory, while undergraduates in New Jersey tend to prefer the causal-historical view. Have we, then, learned something new about the cultural specificity of semantic intuitions? If so, which cultures are we talking about, exactly? The authors claim to be sensitive to the concern that “East Asian” is not really, to put it bluntly, a denoting term. “There is a common concern,” they note understatedly, in a footnote, “that the labels ‘East Asian’ and ‘Western’ are too rough to do justice to the enormous diversity of cultural groups such labels encompass. We are sympathetic to this concern. However, the crudeness of these groupings does nothing to undermine the experiment we present.” They proceed to draw out what must be among the most glaringly invalid inferences of recent academic philosophy: “On the contrary, if we find significant results using crude cultural groupings, there is reason to believe more nuanced classifications should yield even stronger results.”

This is simply not true. If a comparative study of the responses of English-speaking undergraduates at a university in Hong Kong reveals that responses there are different from the responses obtained from undergraduates at Rutgers, this in no way increases the probability that Hong Kong-style responses will also be given in any future study of Hmong peasants in Cambodia, Mongol-Turkic nomads in Eastern Turkestan, indigenous Japanese Ainus, Filipino Negritos, North Korean labor-camp prisoners, Javanese Hindus, and so on. There is absolutely no reason to think that a single study at an English-speaking university in Hong Kong reveals anything at all about “Eastern ways of thinking.” Even if the authors wish to restrict the extension of “EA” to inhabitants of China, Japan, and Korea, as is suggested at one point, it is not at all clear why they see these three very distinct nations as having something in common that the other inhabitants of that half of the continent lack. It is even more unclear why one should suppose that within either of these three populations there should not be radical differences between the intuitions of people coming from different socioeconomic strata, generations, regions (particularly with respect to the famous urban/rural divide in modern China), ethnicities, and so on. As to ethnicity, it is never made clear whether a Uighur citizen of China, to choose just one of many examples, would count as “Chinese,” and thus as “EA,” or not. One comes away from this study with the sense that there has been a whole century of increasing sensitivity and subtlety in the other human sciences that has simply passed the new empirically inclined philosophers by, and that they’ve jumped back into the study of other ways of knowing as if Victorian-era caricatures were still the industry standard. The authors are interested in another culture, but not too interested.

The problem is not just that what English-speaking undergraduates at a university in Hong Kong have to say about the Gödel case says nothing about “Eastern thought.” The problem is not just the authors’ vast culturological conclusion drawn from such minimal data. The problem is that the study of cultures, which are inter alia what cross-cultural experimental philosophy is studying, requires not just the solicitation of answers to the scientist’s questions, but also a scientific interest in the questions asked within the culture. While it is at least possible that Anglophone undergrads in Hong Kong had already been thinking about the Gödel case at the time of the study, it is very unlikely that this problem is on the minds of North Korean labor-camp prisoners or Hmong peasants. A member of a community other than the one from which the experimental philosopher sets out might have different intuitions about the Gödel case, but he might also really not care, or he might be intimidated by the umlauts, or he might mistakenly presume that a competent answer to the question requires a mastery of the incompleteness theorem that he knows himself to lack. In part for these reasons, the kind of human science that has traditionally taken an interest in cross-cultural empirical study has recognized that beyond swooping in and soliciting answers to our questions, there is something to be learned from sticking around long enough to find out what their questions are. In the end, it is only by learning what their questions are that we can really make any sense out of why they give the answers they do, rather than just deferring to armchair speculations about the characters of civilizations.

A properly scientific study of the unity and diversity of the human mind stretches back to the “turn-of-the-century anthropologists,” and it is still being continued by philosophically minded cognitive anthropologists such as Scott Atran (the latter’s recent work on suicide bombing is a good example of a scholar taking an interest in questions that philosophers seldom ask, e.g., “how might I best give up my life for the glorification of God?”), and has been elegantly summarized by transdisciplinary scholars such as G. E. R. Lloyd. But in order to fully recognize the breadth of the project before us, and to consent to take it on, philosophers must stop policing their disciplinary boundaries and start recognizing a common project with scholars in other fields (beyond psychology). And this is something the new experimental philosophers do not seem ready to do.

Machery et al.’s surveys of students in Hong Kong rely heavily on prior work done by the prominent psychologist Richard E. Nisbett, who in his very well-known book The Geography of Thought: How Asians and Westerners Think Differently… And Why (2003), argued that “[t]he collective or interdependent nature of Asian society is consistent with Asians’ broad, contextual view of the world and their belief that events are highly complex and determined by many factors.” The individualistic nature of Western society, on the other hand, “seems consistent with the Western focus on particular objects in isolation from their context and with Westerners’ belief that they can know the rules governing objects and therefore can control the objects’ behavior.” Nisbett is a serious social scientist, whose research has yielded noteworthy results. But this does not mean that his interpretation of the results ought to be accepted wholesale. In particular, his presumption that “Asian” and “Westerner” name two real kinds is hasty, for reasons I’ve begun to sketch out.  

If these are not real kinds, though, then what kind of kinds are they? For one thing, they are newsworthy kinds: Nisbett purports to account for basic worldview differences between the young people who make up China’s rising work force, and their counterpart cohort in the United States, and American bookbuyers eat it up. It is no secret that there is now a massive industry of books attempting to account for the divergent fates of China and America in terms of civilizational character, and Nisbett’s work stands as a fine representative of this genre. If we are to agree with Knobe and Nichols that philosophers must not stop at experiment, but should do something more than experiment that is distinctively philosophical, perhaps that something might include critical investigation of the extrascientific forces that often help to shape scientific research programs. As it is, Machery et al. have added one minor footnote about semantic intuition to the vast genre of East-vs.-West literature, whose basic premises they accept wholesale, and whose complex history and underlying causes they seem wholly to ignore.


Once the two papers featuring Hong Kong undergraduates are out of the way, that’s all for the East, and for the remaining five chapters it’s back to America, where the authors investigate, among other things, intuitions about free will and determinism.

For the past few centuries, philosophers have worried about how it can be that human beings, who supposedly possess the remarkable ability to arbitrarily do one thing or its opposite in accordance with the dictates of their free will, can nonetheless exist in a world that is governed by some variety of determinism. As Kant put it, we are caught in an antinomy between the view that “there are in the world causes through freedom,” and its opposite, “there is no freedom, but all is nature.” The assumption has been that there is a problem here (though Kant himself believed he could solve the problem a priori), yet the three papers on lay views about moral responsibility and determinism show that in fact non-philosophers are perfectly comfortable holding simultaneously that one can be both constrained to act in a certain way and be morally responsible for that action. In other words, laypeople tend to be “compatibilists” about free will and determinism.

It doesn’t hurt to see empirical confirmation, but it also seems to be something philosophers could easily have anticipated from the armchair. They could also have learned of the relative rarity of incompatibilist worries by going back to the Greek philosophers, to whom they surely had at least an introduction as undergraduates. Or they could simply have read about Greek ideas concerning responsibility and fate, for example, in Bernard Williams’ masterful Shame and Necessity of 1994. As Williams shows, setting out from some intriguing insights of Nietzsche, the Greeks conceptualized justice more as a sort of payment that lifts a taint and restores harmony than as disapproval of a person’s having done one thing when he or she should have and could have done another. So much for the supposedly intractable antinomy between determinism and free will. Whoever came up with the vain illusion of the latter anyway? This conception, as Nietzsche rightly noted, emerged as a response to distinctly Christian preoccupations, and evolved.

Here we have a case where lay intuitions do help confirm that recent philosophers could very well be “thinking weird” about a problem, but where the same thing could have been learned by paying attention to areas of inquiry such as history, philology, and theology. Turning to empirical methods is valuable, but not at the cost of neglecting what we already know. And it is no argument to say that this knowledge is not scientific, for there is no reason in principle why we could not go about studying scientifically, even quantitatively, all the occurrences of references to responsibility and determinism in the entire corpus of Greek literature (the Germans see this sort of study as a part of Altertumswissenschaft or the “science of antiquity,” and it is technical, rigorous, and productive indeed). In other words, we could “survey” the Greeks to find out what they think about this problem.

But the past is a foreign country, and one suspects that the new experimentalists only want to linger there about as long as Machery et al. lingered among the EAs. Yet if we understand “experiment” in its rich, double sense—still retained in French but lost in English by the 18th century, of both “experiment” and “experience”—and if we accept that observation is a variety of experience, then there is potentially no limit to what may count as a source of experimental results. Yet x-phi, at least for now, appears content to limit itself to the results of questionnaires distributed to undergraduates. 


A similar concern about this rather narrow conception of “experiment,” and about the ability of philosophers to carry out empirical research that goes much beyond the questionnaire, informs Jesse Prinz’s elegant and thoughtful paper in the final section of the book, “Empirical Philosophy and Experimental Philosophy” (in their presentation of the paper, the editors mistakenly refer to this pair as “experimental philosophy” and “experimental psychology,” unless I’m missing some obvious, and perhaps revealing, synonymy). 

Prinz argues compellingly that observation is what underlies and unites different approaches to philosophy. He understands observation in a greatly expanded sense, so as to include even “armchair elicitation of intuitions,” which “qualifies as a form of observation in a broad sense of the term” insofar as it is a process of “introspective memory retrieval.” If observation so understood is philosophy’s true Leitmotiv, then there is no reason to exclude as non-philosophical either the most armchair or the most experimental approaches.  

Prinz highlights some significant limitations on the kind of research being done by experimental philosophers, limitations arising from the nature of their training. Experimental philosophers do not, for example, investigate connections between concepts by studying reaction times. They are interested in psychological methods, but real psychologists are leagues ahead of them in setting up complex tests that elicit what is going on in a subject’s reasoning. Why not, then, just defer to the psychologists when they are the ones with the skill and resources to conduct better tests? And why not, for that matter, defer to the anthropologists when we need help interpreting cultures (as Machery et al. clearly do)? And why not defer to the historian of philosophy for insight into the origins of the so-called problem of compatibilism? There is an anxiety running through this book that might provide a sort of answer to these questions. Walter Sinnott-Armstrong inadvertently discloses it with a question of his own, in his insightful armchair contribution entitled “Abstract + Concrete = Paradox.” He asks: “What’s left for philosophers?”

I do not have an answer to this question; but not having an answer worries me less than many of my disciplinary peers. Consider, again, the Philosophical Transactions. The conception of philosophy that emerged out of this is a discipline that does not need to police its borders in order to thrive, and that thrives best when it takes advantage of “experiment” in the very widest sense of this term, which includes everything from introspection to the “thick description” of a Geertzian ethnographer and the manipulation of eggs in petri dishes. That’s what philosophers once thought experimental philosophy was. It is heartening to see a sort of descendant of it taking shape, but I believe that it is not yet shaped enough, and that the papers in Knobe and Nichols’ volume for the most part show that it suffers from ignorance of its place in history, both the history of philosophy proper, to the extent that that can be isolated, as well as the history that philosophy shares with science natural and human.


Appiah, whose book represents a more promising path for the future development of experimental philosophy, certainly recognizes the value of history to our current philosophical projects, whether experimental or a priori. He begins his work with an elegant epigram from David Hume, who understood that “the study of history confirms the reasonings of true philosophy.” One part of history, of course, is the history of philosophy’s effort to carve out and defend a space for itself among the various disciplines; this struggle, what Kant called “the conflict of the faculties,” is in the end what gives philosophy the character it has. There is no thread of core concerns that weaves across the centuries, tying Plato and Paracelsus and Prinz together. It is true that they are all interested in some of the same things, but each is interested in many things that do not interest the others at all, and each holds these other things to be just as philosophical as what is shared by all.

Yet these days, as Appiah recognizes, philosophers tend to construe their identity in terms of what they are not: not empirical scientists, not theologians, not artists. In the end, this sort of self-definition is as baseless as it is unsatisfying. Part of Appiah’s project is to show “how newfangled that disciplinary self-conception is.” As he sharply notes, what is in fact novel “isn’t the experimental turn; what’s novel was the turn away from it.” Of particular importance for Appiah’s conception of experimental philosophy is that all of the canonical premodern and early modern philosophers in fact belong as much to the history of psychology as of philosophy. To take an interest as a philosopher in experimental psychology, in this respect, is simply to close a recent and artificial rift.

A sophisticated, new experimental philosophy will be one that understands the historical contingency of disciplinary self-conceptions, rather than supposing that there is a real and natural chasm between philosophy and psychology that x-phi is now boldly traversing. It will be open, as is Appiah, to history as well as to what we now call “psychology.” And it will cease with the pious deference to its immediate academic predecessors, whose conception of disciplinary boundaries is no more written in stone than the medieval quadrivium turned out to be.

1 I should point out that Knobe and Nichols are not themselves responsible for this flashy moniker, and to their credit avoid using it in their volume. Yet a quick Google search will show the very central role of branding in this movement’s current prominence. This might be the fault of overzealous graduate students, who understand best how to harness the power of the new social media. But it says at least something that no one is composing anthems and making music videos about, say, Thomas Aquinas on individuation.

2 An earlier version of this article mistakenly swapped the beliefs of the “EAs” and “Ws” in the study performed by Machery et al.

Image: Dana Schutz, Thinker, 2008. Oil and acrylic on canvas, 95.25" x 76", Courtesy Zach Feuer Gallery, New York.

Subscribe!

It’s the right thing to do.

Sign up now to start receiving the magazine that believes history isn't over just yet.

Subscribe now »

  • Justin E.H. Smith
    • Maneaters

      In July 2008, while travelling on a Greyhound bus between Edmonton and Winnipeg, Vincent Li beheaded his sleeping seatmate, a man he had never met, with a butcher knife. More…