Unlike anthropologists, philosophers do not tend to go on field trips. But we do go on trips. I therefore allow myself to start with an anecdote.
A few years ago, I had the privilege of making a short visit to some of the predominantly Muslim neighborhoods in one of the two Spanish enclaves on the North African mainland. During this brief visit, we made a stop to eat what the Spanish call pinchos morunos in a local restaurant by the sea. As we left the establishment and headed for our car, my host, who was a member of the armed forces, had a brief and congenial interchange with one of the clients who was seated at a table outside the restaurant entrance. The client in question was a physically imposing person, even by the standards of this heavily militarized enclave, and it seemed to me as if the two men might be connected by some tie of collegiality, so I asked who the client in question was. “The Legion” was the answer.1 Drawing the obvious demographic implications, I asked how they, as fellow members of saliently different demographic groups, find it working together. “Perfect. No problem,” was the answer. “And what if war were to break out between Spain and Morocco? What would you do?,” I asked. Without a moment's hesitation, my host responded: “We would kill them before they kill us.” Knowing that my host was speaking as an experienced UN peacekeeper, who had seen the deadly results of violent social collapse, there was no way I could dismiss this as mere bluster. I shall return to the question of how to understand this reply in the final section of the article.
The preceding vignette is an accurate depiction of a historically real event (or so I insist). Yet even if I had just made the whole thing up, it illustrates a basic fact about the necessary conditions of our ethical understanding of the social world. It matters not only what we think about what people do. It also matters what we think about what they would do if, that is, in counterfactual circumstances, some of which may be very unlikely to occur, at least as articulated in a schematic thought experiment. Without having at least some handle on what to say about a significant range of possible but historically unactualized and sometimes very unlikely circumstances, we would be unable to maintain our ethical bearings in our social world as it actually is. But not just that. In order to retain our ethical bearings in our social world as it actually is, we not only have to consider our response to those counterfactual circumstances themselves, but we also have to consider how to respond to the question of which counterfactuals we ought to consider; of when to do so; and of how to do so. Once again, I shall return to the significance of this fact in the final section of the article.
Three Levels of Counterfactual Engagement
There is more than one problematic aspect of ethical thought experiments. Here, I shall mention three. In each case, my point of departure will be the now (in)famous “trolley problem,” in which an individual agent is normally presented with two scenarios in which they can cause the death of one person in order to save five, but where the manner in which this outcome is achieved is widely expected to cause significantly more reluctance in some cases (for example, where the agent has to directly push an innocent victim) than in others (for example, where what the agent would have to do is flip a switch that causes an innocent victim to be pushed.) (See, e.g., Kamm 2016; Lillehammer 2023.)
The first problematic aspect of ethical thought experiments is whether there is an interestingly narrow range of responses that will tell us what people do think or, if not what they do think, then what they ought to think, about the counterfactuals in question. There has been much criticism of the so-called “trolleyology” literature on this score over the half-century or so since Philippa Foot incidentally introduced her trolley example in the context of an argument that was primarily about the morality of abortion (Foot 1967; cf. Fried 2012). It would be natural to see this kind of criticism articulated by anthropologists, which—indeed—it has been (see, e.g., Keane 2015, 2021). I shall only make two brief remarks about this aspect of “trolleyology” here. First, there is by now a significant literature that surveys people's responses to the standard trolley cases, some of which is cross-cultural and some of which surveys responses from people from over 40 different countries.2 What this literature suggests is that many human beings in many different places are less inclined to approve of someone causing death by directly pushing someone than by means of some form of indirect action.3 With respect to this particular question at least, any allegedly “urgent need for further research” is at this point in the proceedings very much open to question. Either way, this challenge to “trolleyology” is rarely raised by critics in isolation; it is normally brought as part of a wider barrage of challenges, such as that of asking what we are supposed to infer from people's responses to imaginary and sometimes highly artificial thought experiments even if we have been able to reliably collect and record them. The anthropologically interesting challenges to “trolleyology” arguably lie elsewhere.
The second problematic aspect of thought experiments in ethics is whether, given the responses to counterfactuals described in a certain way, there is an interestingly narrow range of norms and principles that either explain or justify those responses. Perhaps it would be less natural to expect to hear this criticism articulated by anthropologists, although this does not mean that anthropology has nothing useful to contribute to the discussion of it. One example will have to suffice. The philosopher Frances Kamm—who is perhaps the most prominent living exponent of “trolleyology” in contemporary moral philosophy—has devoted several decades to the articulation of a unifying principle that will explain and justify the ethical difference between a range of trolley cases, some of them relatively mundane, others quite extraordinary.4 According to one answer, which Kamm calls “The Doctrine of Productive Purity,” there is an ethically significant difference between causing a harm to produce a new component of a greater good on the one hand and causing a harm to sustain an existing component of a greater good on the other, where even if the former is not permissible, the latter might be.5 Consider by comparison the practice of Jain renouncers described by James Laidlaw, according to whom the Jain proscription against killing living things—however small—has led Jains to develop a variety of sophisticated strategies to distinguish between different ways of causing death, some of which are thought to be ethically less problematic than others (Laidlaw 1995, 154–159, 165–166, 216–217, 306–309; 2005; 2007). Laidlaw's ethnography presents us with examples of Jain renouncers who will only eat food if it is cooked by someone else; or only drink water after it has been previously boiled. One obvious question is how it can make sense to claim that it is better that another living creature be caused to die by someone else rather than oneself or by being previously boiled rather than by being imbibed. Laidlaw correctly notes that the norms enacted by Jain renouncers have a strongly “non-consequentialist” flavor and goes on to suggest that living by these norms might be aimed to better promote the good of the renouncers themselves even if it does not aim to best promote the good impartially, universally, or overall (Laidlaw 2007, 166–169; cf. Kamm 2023). To conjecture that there is a “doctrine of productive purity” at work here is hardly far-fetched. The question is what this doctrine might be and whether it is a sensible aspiration to articulate this “doctrine” at the level of theoretical sophistication and consistency that is standardly found in the work of Kamm and other moral philosophers who share her methodological outlook.6
The third problematic aspect of ethical thought experiments is to what extent it is justified to either entertain or elicit responses to certain kinds of counterfactuals at all. This problematic aspect can be usefully divided into two separate questions, namely: (a) to what extent the consideration of such counterfactuals ought to be considered an advisable aspiration; and (b) for whom it should be thought of as an ethically advisable aspiration if it is one. It cannot be assumed that the answer to (b) is “just about anyone,” or “just about anyone at any time.” After all, at various times and in various places it has been seriously entertained by intelligent human beings that comprehensive ethical insight is importantly esoteric.7 It would be natural to expect this aspect of ethical life to also have been explored by anthropologists, which—indeed—it has been (see, e.g., High et al. 2012).
That leaves question (a), that is, whether anyone should ever entertain certain kinds of counterfactuals at all. This is also an aspect of ethical life that it is natural to think would be of some anthropological interest. Yet the response of at least some contemporary anthropologists would seem to have been that the appropriate perspective from which to address this question when it comes to the kinds of counterfactuals discussed in the “trolleyology” literature is as one group of academic practitioners sitting in judgment over another group of academic practitioners, to whom the following challenge is then raised, namely: “Why should serious intellectuals busy themselves with such things?”
No doubt, this is a perfectly legitimate question. It is not, however, a question that anthropologists are better placed to answer than anybody else. An alternative approach is to consider ethical thought experiments as another field of anthropological inquiry, about which one might want to ask questions like the following: (1) What kind of ethical life (academic or otherwise) is it that comes to understand itself as being encumbered with an aspiration to develop maximally consistent and explanatorily coherent responses to an arbitrary range of ethical thought experiments? (2) What are the social conditions at work where this kind of aspiration has been observed to arise? After all, and as many philosophers would agree, there is no historical inevitability that ethically serious subjects who are “trying to engage cognitively . . . with the real world” will aspire to this kind of ethical life (Geuss 2010: 59). If so, what is it that makes some moral philosophers, at certain times and in certain places, either pursue or reject this aspiration? This is a question that someone could be seriously interested in even if they ultimately conclude that this tendency toward the pursuit of maximal generality and consistency in ethical thought is based on “a profound mistake” in spite of the fact that, as the philosopher Raymond Geuss has argued, it “holds an almost invincible dominion over the modern imagination” (2010: 59). Incidentally, taking this view does not stand in the way of endorsing the appeal to ethical thought experiments even of a “trolleyological” variety for a variety of didactic purposes, as when Geuss himself describes the example of someone having it explained to them that a bomb has been planted on an airplane ‘which will explode unless I perform some action, such as flipping a switch’ (Geuss 2017: 289–290).8
Connecting the Threads
I shall conclude this brief discussion of the problematic aspects of ethical thought experiments where I began, namely with my host's response to the What if? question. One obvious way of interpreting his response is as a prediction of how he would behave in a set of extremely unfortunate and (one would hope) extremely unlikely circumstances. Another obvious way of interpreting his response is an ethical judgment about how one ought to behave, or how it would be better to behave, in such a set of extremely unfortunate and (one would hope) extremely unlikely circumstances. Yet these two interpretations do not exhaust the available alternatives. A third alternative is that what my host was doing was putting a stop to the conversation before it had really got started, in a similar way to someone who explicitly says: “You might think that's an interesting question; but I'm not going there; and certainly not with you.”
I do not know whether this is what my host intended by the response I quoted at the outset of this article. Either way, there is a vast range of potential reasons why someone might refuse to consider an invitation to entertain an ethically disagreeable counterfactual possibility of the kind that is commonly articulated in the ethical thought experiments familiar from contemporary “trolleyology” and the like; from its complete lack of ethical interest or practical relevance at one extreme to its overwhelming ethical interest or practical relevance at the other. What I have tried to suggest is that what those reasons are is a potentially respectable topic of anthropological investigation.
Suppose, finally, that I return to my host and ask for his response to my interpretation of our earlier conversation. Would he agree to its accuracy? Or would he reject it as reading way too much into what was, after all, a very brief exchange; or would he even deny that it ever took place as described in my account? There is at least some reason to think that I would be met with a negative response. As people are said to have joked in Stalinist Russia, the past is unpredictable. I can imagine that some anthropologist colleagues may occasionally have entertained similar retrospective concerns about their own ethnographic fieldwork, whether involving the discussion of imaginary thought experiments, or otherwise. In my own case, the concern in question introduces yet another set of counterfactuals, the epistemic and ethical credentials of which undoubtedly merit critical scrutiny.
Acknowledgments
I am grateful to Paolo Heywood, Adam Reed, and an anonymous referee for comments and suggestions that led to the improvement of this article.
Notes
The Spanish Legion, or Legión Española was created in the 1920 under King Alfonso XIII as a Spanish variation on the French Foreign Legion. The Legion played an important part in the Rif War from 1921 to 1926 and—on the Nationalist side—in the Spanish Civil War from 1936 to 1939 (Franco was a Legion officer). In the constitutional monarchy that is contemporary Spain, the Legion has been deployed in Lebanon, the former Yugoslavia, Afghanistan, and Iraq. For many Spaniards today, the Legion is also associated with its famous parades during national festivals (e.g., Holy Week) and with the nonhuman mascots kept by its units (e.g., goats or monkeys). In its current iteration, the Legion is open to men and women, primarily—although not exclusively—of Spanish origin.
See, e.g., Gold (2023); Greene (2023). The survey in Gold (2023) would appear to suggest that there has been a lack of effort to seek responses from people based on the mainland of Sub-Saharan Africa. (There is some data available from Madagascar.)
The evidence for uniformity in disapproval of causing death by directly pushing someone is more mixed. See, for example, Gold (2023).
For what might be the largest selection of thought experiments of this general kind, see Kamm (2007).
In Kamm's words: “Sustaining a component [of the greater good] seems not to require causal purity in the same way that production does” (2007: 164; see also Thomson 1985, who invokes the similar, but not equivalent, idea of a “redistributive exemption”).
The case of Jain renouncers is by no means unique in this respect. A range of rabbinic prohibitions would appear to have a similar structure. No doubt the list could be further expanded, but there is no space to do so here.
See, e.g., Williams (1995: 153–171) for discussion of an eminent example taken from nineteenth-century Anglophone philosophy.
Geuss makes use of this example to illuminate the ethics of being “negatively” or “positively” causal of someone's death. The wider context is his attempt to explain the philosopher Theodor Adorno's account of how oppressive social structures have a tendency to present ethical or political choices as less complex than they really are (see Geuss 2017: 274–293).
References
Foot, Philippa. 1967. “The Problem of Abortion and the Doctrine of Double Effect.” Oxford Review 5: 5–15. https://doi.org/10.1093/0199252866.003.0002.
Fried, Barbara H. 2012. “What Does Matter? The Case for Killing the Trolley Problem (or Letting It Die).” The Philosophical Quarterly 62 (248): 1–25. https://doi.org/10.1111/j.1467-9213.2012.00061.x
Geuss, Raymond. 2010. Politics and the Imagination. Princeton, NJ: Princeton University Press.
Geuss, Raymond. 2017. Changing the Subject. Cambridge, MA: Harvard University Press.
Gold, Natalie. 2023. “Cross-Cultural Responses to Trolley Problems and Their Implications for Moral Philosophy or: How I Learned to Stop Worrying and Love (Constructivist) Relativism.” In The Trolley Problem, ed. Hallvard Lillehammer, 182–210. Cambridge: Cambridge University Press.
Greene, Joshua. 2023. “Trolleyology: What It Is; Why It Matters; What It's Taught Us; and How It's Been Misunderstood.” In The Trolley Problem, ed. Hallvard Lillehammer, 158–181. Cambridge: Cambridge University Press.
High, Casey, Ann H. Kelly, and Jonathan Mair, eds. 2012. The Anthropology of Ignorance. Basingstoke, UK: Palgrave Macmillan.
Kamm, Frances M. 2007. Intricate Ethics. Oxford: Oxford University Press.
Kamm, Frances M. 2016. The Trolley Problem Mysteries, ed. Eric Rakowski. Oxford: Oxford University Press.
Kamm, Frances M. 2023. “Non-Consequentialism in Light of the Trolley Problem.” In The Trolley Problem, ed. Hallvard Lillehammer, 159–178. Cambridge: Cambridge University Press.
Keane, Webb. 2015. Ethical Life. Princeton, NJ: Princeton University Press.
Keane, Webb. 2021. “To Kill or Let Die: How Americans Argue about Life, Economy and Social Agency.” In Pandemic Exposures: Economy and Society in the Time of Coronavirus, ed. Didier Fassin and Marion Fourcade, 177–192. Chicago: HAU Books.
Laidlaw, James. 1995. Riches and Renunciation. Oxford: Oxford University Press.
Laidlaw James. 2005. “A Life Worth Leaving: Fasting to Death as Telos of a Jain Religious Life.” Economy and Society 34 (2): 178–199. https://doi.org/10.1080/03085140500054545
Laidlaw, James. 2007. “The Intension and Extension of Wellbeing: Transformation in Diaspora Jain Understandings of Non-Violence.” In Culture and Well-Being: Anthropological Approaches to Freedom and Political Ethics, ed. Alberto Corsín Jiménez, 156–179. London: Pluto Press.
Lillehammer, Hallvard, ed., 2023. The Trolley Problem. Cambridge: Cambridge University Press.
Thomson, Judith Jarvis. 1985. “The Trolley Problem.” Yale Law Journal 94: 1395–1415. https://openyls.law.yale.edu/bitstream/handle/20.500.13051/16338/56_94YaleLJ1395_1984_1985_.pdf?sequence=2&isAllowed=y.
Williams, Bernard. 1995. Making Sense of Humanity. Cambridge: Cambridge University Press.