Humans are ambivalent about sex because they fear death

Terror-management theory (TMT) postulates that denying personal mortality is a primary human motive (Greenberg, Solomon, & Arndt, 2008). Humans, uniquely among all animals, are intelligent enough to have a sense of self and also to understand that they as agents are certain to eventually stop existing. Thus they create and endorse ideals, philosophies, and social structures that imbue life with symbolic meaning, providing a kind of symbolic immortality. This broad idea is originally due to Ernest Becker, an anthropologist who was inspired by Otto Rank and hence by Freud, but it was first examined experimentally by Jeff Greenberg, Tom Pyszczynski, and Sheldon Solomon.

Basic TMT experiments take the form of making mortality salient to subjects assigned to the experimental group, while making some other topic salient to subjects in the control group. Possible mortality cues include having subjects describe in writing "the emotions that the thought of your own death arouses in you" or subliminally1 presenting the word "dead". Control cues may prime a totally innocuous concept, such as television, or an aversive but nonlethal concept, such as dental pain or flunking a test. Then the researchers measure subjects' adherence to or defensiveness about supposedly death-denying constructs, such as social groups or religions. Sure enough, mortality salience makes people more clannish: death-primed Christians view Christians more favorably and Jews more negatively (Greenberg et al., 1990), death-primed whites express more positive attitudes towards white racists (Greenberg, Schimel, Martens, Solomon, & Pyszcznyski, 2001), and death-primed Scots judge the English more negatively (Castano, 2004). An example of a subtler mortality-salience effect is provided by Taylor (2012), who found that mortality salience increased preference for TV shows with themes of law and justice, like Law & Order. Furthermore, among Taylor's death-primed subjects, watching an episode of Law & Order ameliorated a self-enhancing bias observed in subjects who saw no episode or an edited version of the episode in which justice was thwarted. Thus, seeing justice being done seems to comfort death anxiety, and people choose what media to consume accordingly.

So where do sex and the body come in? You would expect that mortality salience could increase defensiveness about sexual norms just as for any other social norm, as in Rosenblatt, Greenberg, Solomon, Pyszczynski, and Lyon (1989), in which death-primed municipal court judges recommended higher bonds for indicted prostitutes. The thing about the body, though, is that it is itself a reminder of mortality. The fact that our minds are so tightly bound to bodies undermines the symbolic immortality we strive for. We may then be threatened not just by illness and injury but also by pleasures of the flesh like food and sex. And the picture is further complicated by how bodies and sex can themselves take on symbolic and therefore terror-management value: think of athletes who strive for an ideal physique, or men who boast of their sexual conquests.

The complex relationship between death and sex in the human mind has been examined primarily by Jamie Goldenberg. To begin with, Study 1 of Goldenberg et al. (1999) found that mortality salience made the "physical aspects" of sex less appealing, in that death-primed subjects gave lower appeal ratings to experiences during sex such as "Having an orgasm" and "Feeling my partner's sweat on my body". Ratings for the "romantic aspects" of sex, like "Blending of selves" and "Expressing love for my partner", were unchanged. Conversely, in Study 2, priming subjects with physical but not romantic aspects increased the accessibility of death (i.e., how readily death-related knowledge was brought to mind), as measured by how subjects filled in word fragments such as "C O F F _ _" that had death-related and non–death-related solutions ("coffin" versus "coffee"). So it appears that we associate sex with death, and this association is what makes sex potentially unsettling. The significance of the romanticization of sex was best demonstrated by Study 3, which showed that additionally priming subjects with the concept of romantic love removes the increase in death accessibility brought about by a sex prime. So romance can act as a means of whitewashing sexuality: as Goldenberg et al. put it (p. 1,176), "Romantic love transforms sex from an animal act to a symbolic human experience, thereby making it a highly meaningful part of one's cultural worldview and obscuring its threatening link to mortality."

A catch of the effects just described is that they were found only in people relatively high in neuroticism, one of the Big Five personality traits. There are plausible theoretical reasons for this, such as neuroticism either causing or being caused by difficulty with terror management. But the plausibility of TMT as an explanation for large-scale features of human society suffers, since we'd expect that low-neuroticism people have a lot of influence on social norms, too. Fortunately, Goldenberg, Cox, Pyszczynski, Greenberg, and Solomon (2002) were able to find a reduction of the appeal of the physical aspects of sex that wasn't influenced by neuroticism, with the help of an additional manipulation: having subjects read an essay that emphasized the similarity of humans to other animals (by saying things like "the boundary between humans and animals is not as great as most people think"). The idea is that maintaining an ideological distinction between humans and animals is another way to defend against the mortality salience of corporeality. When this defense is undermined, even low-neuroticism people can be intimidated by the connection between lust and death.

To extend this thinking to bodily things other than sex, consider Goldenberg et al. (2006). Again, the effects of interest were found only in high-neuroticism people. In Study 2, death-primed subjects spent less time using a foot massager. In Study 1, death-primed subjects lasted for a shorter time in the cold-pressor task, that is, the keeping-your-hand-in-ice-water task, which has a long history of use in experimental psychology as a pain stimulus; in this study, however, subjects "were told that different individuals find the experience 'exhilarating, uncomfortable, pleasurable, or unpleasant.'". The value of this paper is that the dependent measures concern tactile sensations, one pleasurable and one painful, that aren't related to sex and to which few social norms apply.

As an example of how even the physical aspects of sex can manage terror, consider Study 2 of Goldenberg, McCoy, Pyszczynski, Greenberg, and Solomon (2000). Subjects high on a measure of "body esteem" (how positively they viewed their appearance, strength, sensations, etc.) found the physical aspects of sex more appealing after a death prime. (Neuroticism wasn't measured.)

Finally, McCallum and McGlone (2011) examined a perhaps more down-to-earth dependent variable: euphemism. Subjects had to write descriptions of nine photographs, supposedly for another subject who would try to answer questions about each image using only the description. Between subjects, the seventh image was randomly assigned to be a picture of dogs urinating, defecating, or copulating. Death-primed subjects were more likely to describe these actions euphemistically.

That was a lot to digest, so, to review and synthesize: sex, insofar as it is fleshly and creaturely, can be upsetting by its association with mortality. To defend against these mortality concerns, we can distance ourselves from them by euphemizing bodily functions or thinking of ourselves as distinct from animals, or we can whitewash sexuality with socially valued notions like romantic love, or we can try to squeeze some symbolic immortality out of the body itself. While these phenomena are thoroughly entangled with social norms in practice, the core death–body link seems not to require social norms for its existence.

Altogether, this is a gorgeous theoretical package. The universality of mortality explains the universality of erotophobia, and the diversity of ways in which terror can be managed explains the diversity of sexual rules. The biggest gap is a problem inherent to any attempt to use experimental psychology to explain social trends: external validity. Is the congruence between the laboratory findings of TMT and real-world erotophobia enlightening or misleading? To give a concrete example: we know that sex can arouse mortality concerns, but are mortality concerns really part of why we wear clothes in practice? Another: we know that romance can ease mortality concerns aroused by sexuality, but is this effect really part of why romance is valued to begin with? Such hypotheses about cause and effect in societies as wholes are difficult to test observationally and practically impossible to test experimentally. (Thinking about these things makes one a bit more sympathetic to the methods of sociology, doesn't it? Sociologists seek answers to some very hard questions.) So until some very ambitious studies are run, I have no better answer to such questions than the following: affirmative answers are most consistent with TMT, which is supported by experiment.

It's stupid that we go to such lengths to deny the inevitable. Becker himself argued that the denial of death is responsible for much of the evil in the world: war between ideological factions, for example, may arise because one ideology challenges the soundness of another (e.g., Christian and Islamic beliefs contradict each other) and thus undermines it as a source of symbolic immortality. Frankly, TMT is itself terrifying. However, we should not assume that all behavior that happens to be motivated by mortality concerns is bad. With respect to sex, we'll see in the chapters to come why wariness of human sexuality may be justifiable, and even why it might be wise to dress modestly.

Still, it is reasonable to ask what we can do about whichever aspects of erotophobia we've decided are harmful. (Indeed, as I see it, the primary value of knowing the causes of erotophobia is to help control erotophobia.) I've argued that TMT implies erotophobia in general cannot be vanquished without changing human nature itself. But the diversity we've seen in how people manage terror suggests that, by means of clever social engineering, we may be able to replace people's terror-management strategies with more useful ones, which focus on the actual dangers of sexuality without spilling over into excessive sex-negativity. (For example, imagine if people felt condoms and consent were what made sex palatable.) Obviously, this is easier said than done.

More broadly, I hope you agree that a certain healthy fear of death is a good thing. If we value anything we can do while we're alive, we should value life. Not to mention that terror management may be a key ingredient of human ambition and achievement. Goodness knows that I'm trying to get some symbolic immortality out of doing science—that awareness of my own finitude is what motivates me to value my own experiences less than what happens to humanity as a whole. So it's possible to fly from death more prosocially than Voldemort, the Harry Potter villain who cast evil spells in the pursuit of immortality.

It has sometimes been suggested that mortality salience is not merely one basic human motive among many, but the most important one, or even the most fundamental one, from which all behavior originates. For example, Greenberg et al. (2008) entertain the notion (although they cite it to Yalom, 1980) that "all psychopathology is (at least in part) the result of ineffective terror management" (p. 128).

I don't like this line of thinking one bit. It smacks of Freud; it's one way that TMT's origin in Becker, a non-scientist, may keep it from developing properly. TMT is a big, powerful theory, but there's hardly evidence that it's equipped to be psychology's theory of everything. Nor should any sane person expect such evidence to arise. One simple reason why: to return to the theme of Goldenberg et al. (2002), human behavior is in many substantial ways similar to the behavior of other animals, but TMT doesn't apply to animals.

One could try to explain erotophobia in terms of disgust. Sex is disgusting because it's unclean. Besides sex being truly unhygenic because of the transmission of bodily fluids like sweat, saliva, and semen, we may mentally associate sex with excretion because the penis excretes urine, the anus excretes feces, and the vagina excretes menses. But we need to overcome this disgust to have sex, so (it appears) evolution has arranged for sexual arousal to dampen disgust reactions, or at least disgust reactions towards sexual stimuli (Stevenson, Case, & Oaten, 2011, observed this in men and Borg & de Jong, 2012, observed it in women). So the overall prediction is that we'll find sexual things disgusting whenever we don't find them arousing. If we combine this line of thinking with the observation I belabored in the chapter on preferences, namely, that sexual preferences are idiosyncratic, it follows that what we find sexually disguting—our "sexual antipreferences", as it were—should be similarly idiosyncratic. Hypothesize some way by which particular sexual antipreferences could (a) dominate a culture (the same way that some sexual preferences, like a preference for large breasts, dominate cultures) and (b) generalize to negative reactions other than disgust (such as fear), and you have a process for generating kinds of erotophobia as weird as the ones we observe.

I think that this idea is original. There are no direct tests of it. But my reading of the literature (see appendix: http://arfer.net/w/esa/disgust-theory) is that support for it is mixed. It doesn't seem equipped to replace TMT any time soon.

I've seen reasonable theories for many classes of the social norms that I've characterized as erotophobic. To wit:

• When resources are scarce, as they have been in most societies, reproduction is risky. If there's too many mouths to feed, someone will starve. Without good contraceptives, regulating coitus is the only reliable way to regulate reproduction. (This is the line of thinking favored by Stearns, 2009.)
• Similarly, one way to avoid STDs is to avoid sex.
• In a society in which paternity is important (e.g., because possessions are inherited patrilineally), paternity has to be known. And the only way to keep track of paternity without paternity testing is to regulate who copulates with whom.
• More generally, one way people might reinforce their social dominance over other people (e.g., men over women, or clergy over laity) is to make the latter follow a lot of unpleasant or degrading rules.
• People may be partly aware of the ways in which sexual stimuli and feelings can have undesirable influences on human behavior (again, see the chapters to come). Rules like nudity bans could be an attempt to minimize such harm. (See DeForest, 2011, for a humorous exposition of this idea.)

While any number of these theories may be right, they're not enough, even as a package. I mean, I can imagine how functional norms, as people lost sight of their purpose over time, could grow into monsters. For example, rules that were intended or that memetically evolved to keep sexual feelings from interfering with cognition may have given birth to a less pragmatic anxiety about sexuality that in turn created Victorianism, even though extreme anxiety about sexuality is a solution worse than the problem. What these theories can't explain, and where TMT shines, is that regulations about sexuality having no obvious function are universal, but obvious fully general anxiety about sexuality is not. (If the horny nude Marquesan girls who swam out to greet sailors (Bolin, 2004) were anxious about sexuality, they sure as heck didn't know it.) The regulations imply some kind of crazy unconscious inconsistent-yet-fully-general anxiety about sexuality, which only TMT (and perhaps the disgust theory) can explain. Also, these theories don't enjoy the experimental support of TMT.