Theologising the Fear of Death, Pt 1
In this essay—divided into two parts (see Part 2 here)—Fr Jonathan reviews what psychologists have learnt about the fear of death, and what Augustine might teach us about how to live with the fact of our mortality.
It is not, these days, sexy to defend the fear of death, except perhaps as a transient stage in a process of grieving, à la Kubler-Ross, which is the last time her name will be mentioned in this paper. Perhaps it never has been sexy. All the same, I would like to try to provide a defence of sorts of the fear of death, not as good per se, but at least natural, rational, appropriate, and useful. To do so, I will draw mainly from two sources: the social psychological literature on our attitudes toward death and St Augustine. The first body of work will help me to answer an empirical question, about whether people do fear death. St Augustine will help me with the theologising—as promised in the title of this paper—with the goal of working out how people should respond to the fact of their mortality.
Now, living on this side of Hume’s is-ought distinction, we are rightly wary of moving from observations about what is normal to judgements about what is normative. Even among natural law theorists, there are disagreements—sometimes rather vicious ones—between those who insist that facts can tell us something about norms (e.g., Henry Veatch, Anthony Lisska) and those who deny that this inference was ever a part of natural law theory in the first place (e.g., John Finnis, Germain Grisez). So, it is not obvious that both my questions should be addressed in a single paper.
This is not the time and I am not the person to defend any particular view of the fact-value dichotomy as it pertains to natural law theory, but here are two brief observations. First, even natural law theorists who deny that the theory requires inferences from is to ought nevertheless assert that human beings are naturally inclined towards the good. In other words, there are some facts—facts about the things to which human beings are naturally inclined—that are relevant to doing moral philosophy. Thus, facts about our natural inclinations or aversions regarding death are, in principle, potentially relevant to moral philosophy. Second, while Hume is certainly and obviously right that “ought” cannot be deduced directly from “is”, in real life, facts do not exist in metaphysical and hermeneutical vacuums. One’s theology or philosophy of nature may well supply the premises required to generate values from facts. If so, then facts cannot just be ignored. Rather they have to be interpreted, theologically or otherwise.
On whether people fear death, Pt 1
There are several reasons to suppose that people do fear death. You might think, for example, that death is inherently scary, in much the same way that sugar is inherently sweet. Or, to put things in Darwinian terms: we evolved to fear death. As evidence for this claim, we might argue from first principles, to say that death is the paradigmatic impediment against every organism’s evolutionary prerogative: it is quite difficult to reproduce if you’re dead, unless you are a male praying mantis for whom being eaten by one’s mate mid-coitus contributes to her nourishment, which on average doubles the number of eggs she subsequently lays . Even given such reproductive benefits of sexual cannibalism, male mantises still show preferences, albeit mild ones, for less aggressive sexual partners .
We need not appeal to so exotic an example to observe that, generally speaking, animals avoid death. Furthermore, even if we are reticent to attribute complex emotions to non-human animals, it would—given similarities in our sympathetic nervous system responses to acute threat—be mean to deny fear or something very much like it to at least our fellow mammals. And on this note of inter-species similarity, we might reason from other animals’ struggles against death that we too share an instinct to remain alive, the converse of which is the fear of death or something very much like it, enough at least to deserve the name. We might and we do reason thus. This essay would be more tedious than it already is if I were to enumerate examples of this kind of reasoning, but let me indulge in just two.
The first is, appropriately enough, St Augustine, who in Sermon 172 (on 1 Thess. 4.13), constructs the following a fortiori argument: “[i]f animals which are so created as to die each in its own time, flee death, love life, how much more man”. My second example is much more recent: the cultural anthropologist Ernest Becker, whose 1973 book The Denial of Death spawned the psychological research programme that forms the background of this paper. Becker argues that the fear of death is an expression of the instinct of self-preservation common to very many animals. Rather than rely on a theological claim about human beings being made for eternity, Becker’s a fortiori argument is built on the fact that human beings know that we are going to die: this is what exacerbates our fear. Other animals are mortal, and struggle to live when they are threatened, whereas we have to live with the constant awareness—even if ignored, even if suppressed—that we will inevitably face death.
Venerable as they may be, these plausibility arguments—either from evolutionary first principles or by analogy with other animals—should not satisfy scientists, good empiricists that we are. We want data. There are, very roughly speaking, two ways to collect data about whether someone is afraid of death: we can ask them directly, or we can observe what happens when they are faced with death or, more likely to pass ethics committees, thoughts of death.
The first option is by far the most common, not least because it is the easiest. Even so, population-level data is hard to come by, which is how we would establish a sense of what the normal level of death anxiety is, or, put another way, whether the fear of death is normal. As you may know from your own experience with personality tests or intelligence tests, psychologists are greatly enamoured of psychometric instruments: and indeed, there are multiple tools available in English alone. Using such tools, studies have consistently found that, at least in the Western populations where such studies are often run, levels of death anxiety are low to moderate, generally below the midpoint of a scale, if you imagine that the low end of the scale indicates total indifference towards death and that the high end indicates abject terror: in other words, people do not strongly deny that they fear death, but nor do they tremble at the thought, at least not while sitting comfortably at a desk to fill out a questionnaire. My own attempts to cast the empirical net more widely have found similarly middling levels of death anxiety in Russia, South Korea, Brazil, and the Philippines.
Things get a little bit more interesting when we begin to ask whether proximity to death brings with it increased self-reported fear. There have not, as far as I am aware, been any studies in which researchers interview vast numbers of participants in their final hours of life, but there are studies specifically looking at the elderly, the terminally ill, and those sentences to death.
There are several American studies that suggest that the fear of death declines with age; my own data confirms this result too, though the finding did not replicate in South Korea, the Philippines, Brazil, and Russia, all selected for their significant Christian populations. I also have data from countries with much smaller Christian representations—like China and Japan—and they do not evince this pattern either. But perhaps age is a too equivocal a proxy for proximity to death, especially as the over-75s tend to be under-represented in these studies for practical reasons. To address this problem, researchers have turned to terminally ill patients, and those otherwise near to death.
The research on terminally ill patients—mostly those diagnosed with HIV/AIDS or cancer—have largely found that they report no higher levels of death anxiety compared to the general population, after controlling for elevated levels of generalised anxiety. Furthermore, a study from the 1970s dug a little deeper, asking patients why they said they were not afraid of death: by far the most common reason given was that death was inevitable, and therefore that it was not worth being anxious about .
More recently, Amelia Goranson and her colleagues’s reported two ingenious studies. In the first, they collected blog posts written by terminally ill patients, diagnosed with cancer or amyotrophic lateral sclerosis (Lou Gehrig’s disease). This allowed them to subject the texts to sentiment analysis, which they did using a tool call Linguistic Inquiry and Word Count. [This next part might horrify humanities scholars, and I won’t be offended if you plug your ears for a few minutes.] What the tool does is that it categorises and counts words based on predefined and preselected dictionaries. In this study, the researchers were interested in how positive or negative people felt: and so, the software searches for and counts words like “happy” and “sad”, and generates positivity and negativity scores based on the frequency with which such words are used .
Having collected these blog posts, the researchers also asked a group of nonpatients to imagine that they had been diagnosed with a terminal illness and that they had created a blog to record their experience with the illness. Then, they were asked to write a blog post for this imagined blog. So now, the researchers have two corpuses: blog posts written by people who are actually terminally ill, and blog posts written by people who are imagining being so. Armed with their little bit of artificial intelligence, they could compare the two groups, which they did, finding that the nonpatients were more negative than the patients. That is, we imagine that terminal illness is worse than it is. [Before you complain too much about the unreliability of the software, I should add that the researchers also recruited human coders for the job, and found more or less the same thing.] Longitudinal analyses on the patients’ blog posts, also showed that patients’ blog posts became more positive (and not more negative) over time—that is, as they approached death.
Their second study looked at the last words of inmates on death row. It turns out that the Texas Department of Justice makes available executed prisoners’ last words going back to 1982, from which the researchers extracted 396 items. The researchers also obtained collections of death row poetry, which is exactly what it sounds like, of which they found 188 poems. Finally, they also recruited noninmates to imagine being on death row and writing their own last words, analogous to the previous study. Using the same computer- and human-coding techniques, the researchers again found that actual last words were less negative than death row poems, which were in turn less negative than imagined last words.
Now, of course, none of these studies are perfect. They are not ethnographies of the death bed. They rely on self-report, which may be unreliable under these strange circumstances. The comparisons made between those who are dying and those who are merely imagining are blunt and problematic on various fronts. All the same, this is the evidence base that we do have, and it suggests that the fear of death is more imagined than real.