Abstract: It is an old philosophical idea that if the future self is literally different from the current self, one should be less concerned with the death of the future self (Parfit, 1984). This paper examines the relation between attitudes about death and the self among Hindus, Westerners, and three Buddhist populations (Lay Tibetan, Lay Bhutanese, and monastic Tibetans). Compared with other groups, monastic Tibetans gave particularly strong denials of the continuity of self, across several measures. We predicted that the denial of self would be associated with a lower fear of death and greater generosity toward others. To our surprise, we found the opposite. Monastic Tibetan Buddhists showed significantly greater fear of death than any other group. The monastics were also less generous than any other group about the prospect of giving up a slightly longer life in order to extend the life of another.
---
3.2. Tradeoff
Many in the Buddhist tradition, following classical Indian Buddhist philosophers such as Kamalasıla and Santideva, argue that one of the important consequences of realizing the absence of self and impermanence is the reduction in egocentricity, reflected in the increase in generosity, care, love for others, and the self disappearing as an object of special concern (Garfield, 2015). So, one would expect, if this view is right, that those who see the self as insubstantial and impermanent would be less likely to value their own lives over those of others. Indeed, canonical descriptions of non-egocentricity, generosity, and care are replete with examples of individuals who willingly sacrifice their own good for that of others. The next study explores this with our populations.
Many in the Buddhist tradition, following classical Indian Buddhist philosophers such as Kamalasıla and Santideva, argue that one of the important consequences of realizing the absence of self and impermanence is the reduction in egocentricity, reflected in the increase in generosity, care, love for others, and the self disappearing as an object of special concern (Garfield, 2015). So, one would expect, if this view is right, that those who see the self as insubstantial and impermanent would be less likely to value their own lives over those of others. Indeed, canonical descriptions of non-egocentricity, generosity, and care are replete with examples of individuals who willingly sacrifice their own good for that of others. The next study explores this with our populations.
3.2.1. Method
Participants were given a task developed from those used to assess the influence of intertemporal self-connectedness on future decision-making (e.g., Bartels & Urminsky, 2011). In this task, participants have to decide the extent to which they would trade off months of their own life for an increasing amount of time for another person. Participants were shown the following instructions:
Participants were given a task developed from those used to assess the influence of intertemporal self-connectedness on future decision-making (e.g., Bartels & Urminsky, 2011). In this task, participants have to decide the extent to which they would trade off months of their own life for an increasing amount of time for another person. Participants were shown the following instructions:
Imagine that you have a terminal disease that will kill you in 6 months unless you take a medication. There is only one dose of the medication available. If you take the medication, it will prolong your life by 6 months. So if you take the medicine, you will live for 12 months instead of 6. If you don’t take the medication, it will go to someone else who has the same condition and will die in 6 months. This person is very much like you but a stranger whom you will never meet or be in contact with.
Participants then indicated whether they would choose to give the medicine to themselves or the stranger if the medicine had varying degrees of efficacy for the stranger. The shortest possible extension of the stranger’s life by the medicine was “1 month or less”; the longest was “more than 5 years,” with eight additional intervals (2 months, 3 months, 6 months, 1 year, etc.) in between. The medicine was always said to prolong the participant’s life by 6 months. Each participant therefore rendered 10 responses (one for each self-other paired decision). Responses were scored by counting the number of times a person opted to give the medicine to themselves; scores ranged from 0 to 10. A score of zero would mean a subject would give up their medicine to the stranger regardless of how long it would prolong their life; a score of 10 would mean a subject would keep the medicine for themselves even if the medicine would prolong the stranger’s life by 5 or more years.
Participants then indicated whether they would choose to give the medicine to themselves or the stranger if the medicine had varying degrees of efficacy for the stranger. The shortest possible extension of the stranger’s life by the medicine was “1 month or less”; the longest was “more than 5 years,” with eight additional intervals (2 months, 3 months, 6 months, 1 year, etc.) in between. The medicine was always said to prolong the participant’s life by 6 months. Each participant therefore rendered 10 responses (one for each self-other paired decision). Responses were scored by counting the number of times a person opted to give the medicine to themselves; scores ranged from 0 to 10. A score of zero would mean a subject would give up their medicine to the stranger regardless of how long it would prolong their life; a score of 10 would mean a subject would keep the medicine for themselves even if the medicine would prolong the stranger’s life by 5 or more years.
One thousand one hundred and eighty-one participants (701 Americans, 240 Indians, 240 Tibetan monastics; 42.2% female) took the tradeoff task. Of the American participants, 311 identified as Christian and 276 identified as nonreligious. The 41 American participants who indicated other religions were excluded from this analysis, as were 73 Americans who gave incoherent answers (i.e., choosing to keep the medicine that would keep the stranger alive for more time when they had already chosen to give away the medicine when the stranger lived for less time). One Indian subject was excluded for having missing data. We did not collect data from Tibetan or Bhutanese lay Buddhists.
[...]
3.2.3. Discussion
There are several ways to interpret this study, some more optimistic than others. One is that Buddhists are more honest about what they would do with medicine if they were sick. While this possibility cannot be ruled out, it would mean that all three comparison groups (Indians, Christian Americans, and nonreligious Americans) are more dishonest than the Tibetans, and we have no good reason to suspect this. Furthermore, the social desirability of answers on the Tradeoff task is far from obvious. Another possibility is that the Tibetan monastics consider their lives more valuable than that of a stranger, because they are more likely to contribute to the world than a person selected at random.
To rule out these possibilities, future studies should measure actual egocentric behavior, see if performance changes when members of the in-group are used (e.g., trading months off your life vs. a Buddhist you have never met), and collect self-report data. Perhaps such work will reveal a principled reason for why the monastic Tibetans gave such egocentric responses.
On the other hand, future work should also explore what the upper bound is for the egocentric responses among the monastics. Because the majority of Tibetan responses were at ceiling for egocentricity, this study may fail to capture just how dramatic these differences are.
There are several ways to interpret this study, some more optimistic than others. One is that Buddhists are more honest about what they would do with medicine if they were sick. While this possibility cannot be ruled out, it would mean that all three comparison groups (Indians, Christian Americans, and nonreligious Americans) are more dishonest than the Tibetans, and we have no good reason to suspect this. Furthermore, the social desirability of answers on the Tradeoff task is far from obvious. Another possibility is that the Tibetan monastics consider their lives more valuable than that of a stranger, because they are more likely to contribute to the world than a person selected at random.
To rule out these possibilities, future studies should measure actual egocentric behavior, see if performance changes when members of the in-group are used (e.g., trading months off your life vs. a Buddhist you have never met), and collect self-report data. Perhaps such work will reveal a principled reason for why the monastic Tibetans gave such egocentric responses.
On the other hand, future work should also explore what the upper bound is for the egocentric responses among the monastics. Because the majority of Tibetan responses were at ceiling for egocentricity, this study may fail to capture just how dramatic these differences are.
3.3. Discussion
Contrary to our central prediction, we did not find that Buddhist monastics showed less fear of self-annihilation than Abrahamics and Hindus. Indeed, to our astonishment, we found that the monastics showed greater fear of self-annihilation than any other group, including lay Tibetans and Western nonbelievers. Furthermore, the Buddhist monastics were less generous with months of their own lives than the Hindus, the Abrahamics, and the Western nonbelievers.
In light of how surprising our results are, it is worth reviewing some of the limitations and alternative explanations of the research. One possibility is that the results on the increased fear of self-annihilation among monastics is a fluke of the sample. We think this is unlikely partly since we sampled from two geographically distinct monasteries, but more importantly because the results on the self-annihilation measure cohered with the results on tradeoff task. In both studies, the monastic Tibetans exhibited significantly higher concern for self-preservation.
Another explanation for our findings might be that the monastic Tibetans did not really believe the no-self view. We think there might be something right about this, but caution is required. Almost all Tibetan monastics in our sample report that they rely on the no-self doctrine as reassurance that death should not be feared. Furthermore, across several measures, we found the monastics to maintain that there is no self. And this is an absolutely central part of their religious tradition. It is as central to Buddhism as the belief that Jesus is the son of God is to Christianity. So, if the monastics were told, “You don’t believe in the no self view,” this would likely inspire strenuous rebuttals. Nonetheless, there might be another sense in which the monastics have not internalized the no-self view deeply enough to deflect fear of death. Indeed, we suspect that something like this is the case, as we will discuss in the final section.
Another limitation of our study is that we focused on a single Buddhist tradition, and it will be important to see whether the findings hold for other traditions as well. Buddhist traditions differ from one another in doctrinal detail, with respect to practice, with respect to relations between lay and monastic communities, and in degree of piety. We have examined only one of these traditions—the Tibetan tradition as it is represented in the Indian exile community and in Bhutan. It will be important in future studies to examine other Buddhist traditions to determine whether the phenomena we have found are general, or whether they are driven in part by properties of the Tibetan community or Buddhist doctrine, per se. If the latter, further study will be required to determine which features of Buddhist doctrine are responsible for these results.
Contrary to our central prediction, we did not find that Buddhist monastics showed less fear of self-annihilation than Abrahamics and Hindus. Indeed, to our astonishment, we found that the monastics showed greater fear of self-annihilation than any other group, including lay Tibetans and Western nonbelievers. Furthermore, the Buddhist monastics were less generous with months of their own lives than the Hindus, the Abrahamics, and the Western nonbelievers.
In light of how surprising our results are, it is worth reviewing some of the limitations and alternative explanations of the research. One possibility is that the results on the increased fear of self-annihilation among monastics is a fluke of the sample. We think this is unlikely partly since we sampled from two geographically distinct monasteries, but more importantly because the results on the self-annihilation measure cohered with the results on tradeoff task. In both studies, the monastic Tibetans exhibited significantly higher concern for self-preservation.
Another explanation for our findings might be that the monastic Tibetans did not really believe the no-self view. We think there might be something right about this, but caution is required. Almost all Tibetan monastics in our sample report that they rely on the no-self doctrine as reassurance that death should not be feared. Furthermore, across several measures, we found the monastics to maintain that there is no self. And this is an absolutely central part of their religious tradition. It is as central to Buddhism as the belief that Jesus is the son of God is to Christianity. So, if the monastics were told, “You don’t believe in the no self view,” this would likely inspire strenuous rebuttals. Nonetheless, there might be another sense in which the monastics have not internalized the no-self view deeply enough to deflect fear of death. Indeed, we suspect that something like this is the case, as we will discuss in the final section.
Another limitation of our study is that we focused on a single Buddhist tradition, and it will be important to see whether the findings hold for other traditions as well. Buddhist traditions differ from one another in doctrinal detail, with respect to practice, with respect to relations between lay and monastic communities, and in degree of piety. We have examined only one of these traditions—the Tibetan tradition as it is represented in the Indian exile community and in Bhutan. It will be important in future studies to examine other Buddhist traditions to determine whether the phenomena we have found are general, or whether they are driven in part by properties of the Tibetan community or Buddhist doctrine, per se. If the latter, further study will be required to determine which features of Buddhist doctrine are responsible for these results.
4. Conclusion
Our results suggest a paradoxical effect of Buddhist teaching. Buddhism encourages the belief that there is no persistent self, and this is taken to be a reason not to fear death. We find that monastic Buddhists explicitly deny the existence of a persistent self, in line with Buddhist thought. But contrary to the Buddhist promise of reduced fear of death, the monastic Buddhists showed dramatically increased fear of self-annihilation and valued their own lives over others to a much greater degree.
On every measure we used, the monastics deny the existence of the self. So why do they fail to show the expected reduction in fear of death? We think that it is because, despite their training and explicit claims, they retain a powerful sense of personal identity across the biological lifespan. In particular, like everyone else, Tibetan monastics engage in episodic retrospection and prospection, and this generates a robust sense of personal identity with the past experiencer (see, e.g., Nichols, 2014). Even if you have changed enormously since your first kiss, it will still seem like you had that experience. We propose that even for the monks and nuns, there remains a persistent and powerful sense of identity yielded by episodic memory and prospection within biological life.
The claim that episodic memory generates a sense of personal identity even among monastics is reinforced by looking at work within Tibetan Buddhism. Autobiographies are a primary genre in Tibetan literature. The autobiographies are by people who are held to be of high spiritual attainment (e.g., Gyatso, 1998, 103). It might seem incoherent for an enlightened Buddhist to write an autobiography—how can one affirm an autobiography while denying the self?5 It is certainly clear that these texts make liberal use of the first person singular. The official rejoinder to this alleged incoherence is that these works treat the author as merely a “conventional” person, not an enduring ultimate self. It is possible to speak of persons in this merely conventional fashion, but Tibetan autobiographies suggest that this is not always consistently upheld. Often in these works, the author is reporting a past experience, and the recollections certainly do not seem to present the distanced perspective afforded by thinking that there really is no persisting self.
Rather, they suggest a clear identification with the past experiencer. Consider, for instance, the most famous work in this tradition, The Life of Milarepa. We find the author describing a scene from years earlier in which he had returned to his ancestral home and found human bones among a heap of rags. He writes,
On every measure we used, the monastics deny the existence of the self. So why do they fail to show the expected reduction in fear of death? We think that it is because, despite their training and explicit claims, they retain a powerful sense of personal identity across the biological lifespan. In particular, like everyone else, Tibetan monastics engage in episodic retrospection and prospection, and this generates a robust sense of personal identity with the past experiencer (see, e.g., Nichols, 2014). Even if you have changed enormously since your first kiss, it will still seem like you had that experience. We propose that even for the monks and nuns, there remains a persistent and powerful sense of identity yielded by episodic memory and prospection within biological life.
The claim that episodic memory generates a sense of personal identity even among monastics is reinforced by looking at work within Tibetan Buddhism. Autobiographies are a primary genre in Tibetan literature. The autobiographies are by people who are held to be of high spiritual attainment (e.g., Gyatso, 1998, 103). It might seem incoherent for an enlightened Buddhist to write an autobiography—how can one affirm an autobiography while denying the self?5 It is certainly clear that these texts make liberal use of the first person singular. The official rejoinder to this alleged incoherence is that these works treat the author as merely a “conventional” person, not an enduring ultimate self. It is possible to speak of persons in this merely conventional fashion, but Tibetan autobiographies suggest that this is not always consistently upheld. Often in these works, the author is reporting a past experience, and the recollections certainly do not seem to present the distanced perspective afforded by thinking that there really is no persisting self.
Rather, they suggest a clear identification with the past experiencer. Consider, for instance, the most famous work in this tradition, The Life of Milarepa. We find the author describing a scene from years earlier in which he had returned to his ancestral home and found human bones among a heap of rags. He writes,
When I realized they were the bones of my mother, I was so overcome with grief that I could hardly stand it. I could not think, I could not speak, and an overwhelming sense of longing and sadness swept over me. (Quintman, 2010, p. 118; see also Shabkar, 1994, p. 32; Kongtrul, 2003; 172–3)
This passage is hardly a dispassionate report that a conventional person consisting of fleeting traits included a set of perceptions. Instead, it seems to be a recollection of a devastating personal experience. It is most plausible that Milarepa, in reflecting on this terrible event, could not suppress the sense that he had the experience of discovering his mother’s bones, even if, in a different register, he would deny that there is any self in which he consists, or that he is now the same person who endured that experience.
The sense of continuity over time within a biological life may be resistant to the ideological conviction that that sense is delusional. Nonetheless, the common view that there is self-persistence between lives may well succumb to the no-self doctrine. If this is the case, it may be that the prospect of death is the prospect of the end of the only kind of self in which one has any conviction. This would be a more dramatic denouement than that anticipated in death by Christians and Hindus, each of whom retains a conviction in the survival of an immortal soul. If monastics do indeed fail to extirpate the sense of continuity, this would help explain both their heightened fear of self-annihilation and their increased egocentricity in tradeoff tasks.
Indeed, this explanation actually fits with a traditional Buddhist distinction between innate self-grasping and philosophical self-grasping. The latter is the conviction in the reality of the self as a result of philosophical or religious doctrine, and it is regarded among Buddhist philosophers as eliminable simply through philosophical reflection. The former, however, is regarded as immune to mere philosophical reflection, and it is argued that only prolonged meditation can dislodge it. None of the participants we studied were long-term meditators (Tsongkhapa, 1991), and one important question for future research will be whether highly experienced practitioners of meditation would in fact show reduced fear of self-annihilation.
The Buddhist may be correct in thinking there is no persisting self and hence that it is irrational to fear death. Nonetheless, as Buddhists themselves recognize, our sense of identity across the biological lifespan is resilient, and perhaps the thought of self-annihilation triggers fears too primitive to be easily tamed by the philosophical belief that there is no persistent self.
The sense of continuity over time within a biological life may be resistant to the ideological conviction that that sense is delusional. Nonetheless, the common view that there is self-persistence between lives may well succumb to the no-self doctrine. If this is the case, it may be that the prospect of death is the prospect of the end of the only kind of self in which one has any conviction. This would be a more dramatic denouement than that anticipated in death by Christians and Hindus, each of whom retains a conviction in the survival of an immortal soul. If monastics do indeed fail to extirpate the sense of continuity, this would help explain both their heightened fear of self-annihilation and their increased egocentricity in tradeoff tasks.
Indeed, this explanation actually fits with a traditional Buddhist distinction between innate self-grasping and philosophical self-grasping. The latter is the conviction in the reality of the self as a result of philosophical or religious doctrine, and it is regarded among Buddhist philosophers as eliminable simply through philosophical reflection. The former, however, is regarded as immune to mere philosophical reflection, and it is argued that only prolonged meditation can dislodge it. None of the participants we studied were long-term meditators (Tsongkhapa, 1991), and one important question for future research will be whether highly experienced practitioners of meditation would in fact show reduced fear of self-annihilation.
The Buddhist may be correct in thinking there is no persisting self and hence that it is irrational to fear death. Nonetheless, as Buddhists themselves recognize, our sense of identity across the biological lifespan is resilient, and perhaps the thought of self-annihilation triggers fears too primitive to be easily tamed by the philosophical belief that there is no persistent self.
Posts
