[This essay is a different take on thoughts about the concept of genocide by NYU philosopher Paul Boghossian. We have previously published Massimo Pigliucci’s analysis of the same topic in Scientia Salon on 15 October 2014, under the title The Philosophy of Genocide.]
Most of us think we understand what genocide means. Perhaps this is the case for the lay concept of genocide, but it might not be for the legal definition of it adopted by the United Nations. And I think this latter possibility is disconcerting.
The UN defines genocide as any of the following acts committed with the intent to destroy, in whole or in part, a national, ethnical, racial or religious group, as such:
(a) Killing members of the group;
(b) Causing serious bodily harm;
(c) Deliberately inflicting on the group conditions of life calculated to bring about its physical destruction in whole or in part;
(d) Imposing measures intended to prevent births within the group;
(e) Forcibly transferring children of the group to another group.
At first glance, the UN definition appears to be very clear. We might then be tempted to believe that when determining which actions or events count as genocide, deploying this concept accurately would be pretty straightforward. However, closer examination runs us into several practical and philosophical problems as it may admit of too many vague terms, leading to ambiguities and moral puzzles such that we might think it best to just set the term aside and not employ it in legal or day-to-day judgments.
In fact, it has recently been argued that these problems run deeply through the UN definition of genocide. In his paper, “The concept of genocide”, Paul Boghossian of New York University has made the point very persuasively.
In this essay I will take a look at some of the problems that Boghossian points out with the UN definition of genocide. I will not discuss everything Boghossian addresses, but rather select a few issues that I find to be the most philosophically and practically worrisome. The goal is simply to try to offer a plausible framework for solving three major problems, sort of a prolegomenon to a full solution. As we will see, my view here is not entirely new, but I think it could use more attention, and hope to steer people in a direction that is worth pursuing further, if the concept of genocide is to be salvaged at all.
Boghossian’s three problems
Boghossian is quick to point out that there are significant issues with the condition that an agent must “intend to destroy in whole or in part” in order for genocide to have occurred. First, it seems that there have been almost no cases in which an agent or party has intended to destroy a group in whole . Even Nazi Germany, one of the paradigmatic examples of a party that committed genocide, would permit the existence and survival of some Jews when these Jews brought Nazi Germany pragmatic benefits. Interestingly, Hitler personally spared a Jewish-Austrian named Eduard Bloch, even going so far as to issue him protection during World War II, since Bloch had formerly been Adolf Hitler’s family physician .
Since the Jewish Holocaust is considered one of the most obvious examples of genocide and yet it doesn’t satisfy the “in whole” condition, it seems that such condition excludes too many, even possibly all, cases that are commonly considered instances of genocide. To correct this problem, perhaps we should say that Nazi Germany satisfied the “in part” condition (they intended to destroy the Jewish group in part), and had thereby committed genocide.
But then Boghossian asks us to consider what would count as satisfying the “in part” criterion of the UN definition. While “in whole” excludes too many cases, “in part” seems to conversely include too many. Consider the instance of the Jewish revenge gangs that targeted Germans (presumably former Nazi officers) after World War II, resulting in the death of some 1,500 Germans . The action of these gangs seem to satisfy the UN’s definition of genocide. They intended to destroy in whole or in part the national group of Germans simply because they were Germans. However, we probably still wouldn’t say that the Jewish revenge gangs committed genocide, most likely because too few members of the German population were killed. So, the “in part” condition needs to be either revised or done away with.
Boghossian proposes that we could try amending the condition to say “in whole or in substantial part.” Unfortunately this will not do: he points out that the idea of a “substantial part” is hopelessly vague. What counts as a substantial part of a group? A certain percentage of the members? Some number threshold, of, say, 3,000 individuals? Boghossian notes that the attacks on the World Trade Center might count as genocide since over 3,000 people were killed, but it is not clear if this was as a “substantial part” of Americans or not.
Here I think it is worth pausing to examine what is going on in Boghossian’s arguments. He is claiming that it is difficult to determine, i.e., it is vague, when a party or person intends to destroy a substantial part of a group of people. However, my suspicion is that this is a problem when it is the case that a substantial part of a group has actually been destroyed. The UN definition of genocide includes both a consequence and an intention condition. It is not enough for an agent to just intend to commit some crime; the agent must also actually commit the crime in order for the crime to have occurred.
The difference between the consequence and intention conditions can come apart philosophically. It is certainly not obvious that what it means to intend to destroy a group in substantial part is the same as what it means to actually destroy a group in substantial part. Let’s say I believe that a group is composed of 300 members. It turns out, unbeknownst to me, that the group is actually composed of 1,000,000 members. If I proceed to kill 200 members of this group, we can all agree that I have satisfied the conditions for intending to destroy a group in substantial part while failing to actually destroy a group in substantial part. This suggests that what it means for an agent to intend to destroy a group in substantial part has something to do with an agent’s belief system, while what it means to actually destroy a group in substantial part has to do with the way the world actually is, independent of what we believe.
I think it is important to keep in mind how the two issues interplay in the UN definition. Clearly we think that for an action to count as genocide it must meet at least two different necessary conditions: an intention condition and a consequence condition.
From now on, let’s assume that the vagueness problem that Boghossian raises for the concept of genocide is a problem with the consequence condition of the definition, and we shall see that the distinction between intention and consequence turns out to be very important.
The problem of figuring out when the consequence condition has been met, i.e., when a high enough percentage, or a high enough number, of people have been killed is what I call the vagueness and convention problem. This is the first practical and philosophical problem that Boghossian raises. We intuitively think that a certain threshold of people killed must met in order for genocide to have occurred. However, we don’t want it to be the case that the consequence condition’s threshold for satisfaction rests solely upon a convention, some arbitrary percentage that we agree on without any good reason other the pragmatic one of allowing us to use the concept of genocide in judgments and legal proceedings.
A similar kind of problem arises when we consider which groups are potential victims of genocide. The UN definition lists these groups explicitly: national, ethnical, racial, or religious. But why are these groups the only ones that can be potential victims of genocide? Boghossian suggests these groups may be chosen as potential victims of genocide because they all admit of indelible traits, traits that one cannot change from birth, and not traits that one has blamelessly acquired (a trait like, say, being wealthy). However, he then points out that it is not clear why, if indelibility is what matters, gender is not included while religion is. In other words, if indelibility is what makes groups potential targets of genocide, then the UN appears to have excluded some groups in the set of potential victims of genocide and included some groups that don’t meet the criteria of indelibility.
Additionally, why should we think that indelible traits determine which groups can be potential victims of genocide in the first place? Why should we think that you could only be a potential victim of genocide based on what traits you were born with, but not based on what traits you have blamelessly acquired over the course of your life? For example, why couldn’t the group constituted by people who have acquired the status of being “wealthy” be one capable of being subjected to genocide? It seems that holding the position that the indelibility criterion is what matters for determining which groups can be victims of genocide would require further argumentation, one that looks like a difficult road to travel.
So maybe we should sidestep the indelibility criterion’s problems and try something else. Perhaps the groups whose members identify themselves as part of that group are potential victims of genocide. This would allow for groups constituted by members who have shared indelible traits as well as groups constituted by members who have traits that people have blamelessly acquired to be potential victims of genocide.
But there is a problem with this too, one of inclusion. For there may be members of some groups, such as a group of people who attend a regular book club, who identify themselves as part of that book club group and would now fit the bill for being potential victims of genocide. Intuitively, the book club group doesn’t seem to be a potential victim of genocide.
This problem can be called the grouping and convention problem. We are inclined to believe that the set of groups that are potential victims of genocide shouldn’t be too inclusive or too exclusive, and we also don’t want it to be the case that the groups included in this set are decided merely by convention.
The last problem will not require an elaborate exposition. The badness problem is the problem of what makes genocide distinctively heinous. We tend to think that some actions have a greater moral magnitude than others. Lying is morally bad, but murder is far worse. What is it about an act of genocide, then, that makes it morally worse than mass murder? Why is it that targeting a people for the very reason of targeting a specific group has such great negative moral magnitude?
A sketch of a solution
I think that the badness problem and the grouping and convention problem are intimately related, so I’m going to tackle these first. It may be useful to think carefully about what it is that makes genocide a distinctive crime, a crime that is different from mass murder. Boghossian himself thinks that the concept of genocide was introduced precisely to point out such distinctiveness .
A good starting point might be to look at clear cases where we consider two crimes distinct and ask what makes them different. It is a crime to inhumanely kill domestic animals; it is also a crime to torture and kill humans. What makes these two crimes different? Maybe part of what makes them different is that the crime necessarily takes away properties from a human that we value that it is not necessarily taking away from a domestic animal. A human has emergent properties that arise from his functional composition, such as mental properties like beliefs, desires, and intentions . When you torture and kill a human, you necessarily take away these properties, whereas when you kill a domestic animal, you do not. Similarly, I think that there are emergent properties of certain groups that you necessarily take away when you target those groups that you don’t necessarily take away when you randomly kill a large number of people .
The classic example of an emergent property in physics is liquidity. Individual H20 molecules do not have the property of liquidity, but hydrogen-bonded groups of H20 molecules do. What are, by analogy, the emergent properties of certain groups of individuals? The are the properties that constitute a culture. A people is different from a group insofar as a people is a group that has the emergent property of a given culture. They have a shared set of mental states that lead to shared habitual behaviors. Devout Christians attend church weekly to worship their conception of God, southern European ethnicities tend to value family and consequently routinely make time to eat together, etc.
What makes genocide a distinctive crime, then, is that you necessarily take away the emergent property of a culture, whereas you do not necessarily do this when you commit mass murder. It is possible for a particularly extreme case of mass murder to result in cultural destruction, but this is not necessary.
Even so, what makes genocide distinctively bad?
In order to tackle this, we need to talk a bit about paralysis and murder, as tangential as this may at first seem. I have frequently heard people say that they would prefer death to paralysis or loss of autonomy. This seems to be rooted in the fact that paralysis prevents an individual from performing his/her normal daily functions and satisfying certain goals and desires.
Now, we tend to think somebody cannot both paralyze and murder another person at the same time. We typically think of a person as paralyzed only when that person is alive and has a loss of motor function(s). Conversely, we don’t think of a person as paralyzed, in the ordinary sense, when that person is dead and consequently has a loss of motor functions.
What I’d like to suggest is that while an individual cannot be paralyzed and murdered at the same time, a group can, due to its emergent properties. When genocide is committed against a people, the members of this group are unable to do certain things that we value: engage in daily customs and practices, perform behaviors that reflect their shared value systems, uphold certain goals of their group, etc. In other words, they are culturally paralyzed. In the case of the Jews during World War II, the killing of a large number of that group prevented its members from being able to go to synagogue, uphold the Sabbath, or celebrate Jewish holidays. In this sense, the Jewish people were in a state of what we can call “cultural paralysis.” The word “paralysis” is particularly appealing to use in this context, as it denotes a condition that isn’t necessarily permanent. Paralyzed individuals can recover, and so — a fortiori — can cultures. The Jewish culture has at least partially recovered from its paralysis, for instance.
What makes genocide distinctively bad, then, is that it consists in the conjunction of two morally repugnant actions at once: mass murder is bad, but mass murder that results in cultural paralysis is worse.
Since mass murder can sometimes result in cultural paralysis even though it wasn’t intended, more must be said here. I do think that some cases of mass murder can be just as morally bad as genocide in the sense that they can both yield the same consequence of cultural paralysis. However, there is still a different sense in which genocide is “distinctively heinous.”
Judith Jarvis Thompson draws a distinction between assessing an action and assessing an agent. She argues that intentions are used for assessing the moral character of an agent, while consequences are used for assessing the moral status of an action . While genocide and certain extreme cases of mass murder (those that result in cultural paralysis) carry the same action assessment, they carry a different agent assessment. In genocide, the consequences of the actions that lead to cultural paralysis are on a moral par with mass murder that leads to cultural paralysis, but the intentions of the agent is what makes genocide particularly heinous. The agent intended to commit the conjunction of both murder and cultural paralysis, while the mass murderer only intended the former. This means that the consequences of a genocidal agent’s actions are necessarily either morally worse than or equal to those of a mass murderer, but the moral character of a genocidal agent is always worse. In this sense, then, genocide can be said to be distinctively heinous and deserving of a worse punishment.
It is important to note that Boghossian discusses how some have tried to explain the badness of genocide in terms of a violation of group rights, but argues that this seems like a path best avoided. Accordingly, I have made no reference to a violation of group rights in my explanation. I have attempted to explain the badness of genocide solely in terms of the destruction of something we value, namely, a particular culture, linking it to the agent’s assessment of the action. Hopefully this will allow me to circumvent the complexities of tackling group rights issues, though of course it is possible that my idea of a culture being an emergent property opens up its own set of difficult problems.
This position about what makes genocide distinctive and distinctively heinous is certainly not radically novel. It has long been recognized that cultural destruction has an important role in what we think genocide is. Raphael Lemkin, the man who coined the term “genocide,” thought that what he called “cultural genocide” should be incorporated as a constituent part of the definition of genocide . While what he meant by cultural genocide may be somewhat different from what I term cultural paralysis, the underlying theme is the same, and similar ideas have been expressed by David Nersessian . If our intuitions are ever going to correspond with a definition of genocide, we have to include a type of cultural paralysis/destruction component into our definition.
Incorporating the idea of a culture in our concept of genocide helps to offer a solution to the grouping and convention problem too, since the groups of people that have the emergent property of a culture turn out to be the groups that are potential victims of genocide. On this account, we can say that religions, ethnicities, genders, sexual orientations, class status, intellectuals, etc. can all be potential victims of genocide, while groups such as faculty members at NYU and the book club cannot be because they don’t have sufficiently robust cultural emergent properties.
Granted, I have grossly oversimplified what a culture is, and consequently have likely oversimplified just how easy it may be to decide which groups can be potential victims of genocide. There may be another vagueness problem lurking here, concerned with when exactly a group has the property of a culture, but it still seems to me that this account doesn’t works better than the others that we discussed above. At the very least it looks like a more promising solution to the grouping problem, and one that should be explored further.
So, my provisional definition of genocide looks something like this:
Genocide has occurred when a group’s culture has been paralyzed or destroyed via the murder of members of that group with the very intention to cause such paralyze or destruction by violent means.
This definition focuses not on the number of members of a group killed, but on whether or not cultural paralysis/destruction has occurred. In this way we sidestep the practical issue posed by the vagueness and convention problem. However, as I hinted at above, we do run into another sort of vagueness problem, and not a trivial one at that: when does a group count as having a culture? What is a culture? Did the terrorist attacks of 9/11/01, for instance, result in temporary cultural paralysis? Still, I think that if the concept of genocide is to be salvaged, this is the vagueness problem we will need to focus on.
Since our conditions for when genocide has occurred don’t revolve around quantity of people killed, but instead revolve around cultural paralysis, the number of people killed becomes a lens we can look through to see if a culture has been paralyzed. No doubt there are some clear cases where a certain number of people killed necessarily means that cultural paralysis or destruction has occurred, but when number-related evidence underdetermines whether or not genocide has occurred, we must look to other facts about cultures to come up with an answer, which seems promising.
I said at the beginning of this essay that I was only going to offer a prolegomenon to a solution. I have sketched an account of what we need to do in order to salvage the concept of genocide. I don’t think this account is detailed enough, nor is it likely to be right as presented, but I hope it is a step in the right direction. The first thing we need to do is include cultural paralysis/destruction in the definition of genocide, which would help explain our intuitions about the grouping and convention problem and the badness problem. The second thing to do is to solve the culture-related vagueness problem: when does a group count as having a culture, and what is a culture? This is a pretty deep problem, but I think it is time we revive discussion of it.
Daniel Tippens received his Bachelors of Arts in Philosophy at New York University. He is now a research technician at New York University School of Medicine in the S. Arthur Localio Laboratory.
 Boghossian, P. (2010) The concept of genocide. Journal of Genocide Research.
 Ibid. 10.
 The Nizkor Project, interview with Dr. Eduard Bloch.
 Boghossian, op. cit., 7.
 I recognize that an account of the mind that claims the it is an emergent property of the brain is controversial, as is the account of the mental life of domestic animals that holds that they don’t have the same morally relevant emergent properties that humans do. I am not trying to take a stand on either one of these issues; I am merely employing them for illustrative purposes.
 I thank Massimo Pigliucci for helping me articulate this point during one of his dinner and philosophy meetups.
 Thomson, J.J. (1999) Physician-Assisted Suicide: Two Moral Arguments. Ethics 109: 497-518.
 Cultural genocide, Wiki entry.
 Nersessian, D. (2005) Rethinking Cultural Genocide Under International Law. Carnegie Council, 22 April.