On Oct. 7, 2023, as Hamas terrorists attacked the Nova music festival, a young man named Itamar ran from the festival grounds, avoiding gunmen until he found relative safety in a ditch. “I lay in the ditch with my face looking up to the sky,” Itamar wrote. “I said Shema yisrael adonai elohaynu adonai echad” (“Hear, oh Israel, the Lord our God, the Lord is One”). This prayer, known as the Shema, figures in many stories from Oct. 7, as it’s a prayer that many Jews say in times of mortal danger, and also on one’s death bed.
Saying such a prayer before death is common in religious traditions around the world. As I discovered while researching language at the end of life, biographical lore and cinematic depictions have primed us to expect a dying person to have a spontaneous, idiosyncratic thing to say. But the witty bon mot isn’t likely to happen, nor is it something that most people expect. In fact, untold numbers of people are raised in religious traditions that encourage the dying person to say a specific prayer, the name of a god, or some sacred sound that will guarantee the transition of their soul or spirit to a providential next place—which believers may call salvation, heaven, the land of ancestors, freedom from infinite cycles of reincarnation, or something else altogether. Whether or not they actually say such things is beyond the point. The expectation is that when they die, they don’t say their own thing; they say what everyone else before them has said.
For example, in some denominations of Islam, dying people endeavor to say the Shahadah, an Arabic affirmation of faith whose rough approximation in English is, “There is no god but God and Muhammad is his prophet.” These are the words one utters to convert to Islam, and these are the first words whispered in a newborn’s ear. A devout person speaks them throughout her life.
For those of us unfamiliar with such a ritual, we might think it counts only if it’s rigidly observed. Indeed, the Shahadah must be the absolute last thing someone utters—if the dying person talks about another topic, according to some views of Islamic death practices, the affirmation must be repeated. But I find it interesting that the practice of the ritual can be flexible, such that if a person is unable to speak, they can hold up their index finger, and if they’re unable to hold up a finger, someone else can raise it for them. Muslims whom I’ve spoken with say these all “count.” In a way, these variations are beautiful. But to me, as a linguist, they are not surprising, for this all reflects how ordinary language works: you can say “yes,” you can say “yes” and nod, or you can simply nod. That is, you can shift among these linguistic modes depending on the circumstances, something that people do constantly.
If a person dies agitated or delirious, they’re thought to be morally corrupted, their soul doomed. But if they’re calm—and maybe even have the presence of mind to say something intelligible—they’re judged to be bound for heaven.
At the same time, ritual last-word traditions are embedded in religious belief systems so vast and complex that it would never be reasonable to look for a uniform approach among their contemporary followers or even across the centuries. Some of the current beliefs are relatively recent. According to historian Elliott Horowitz, observant Jews weren’t expected to say the Shema on their deathbeds until the late sixteenth century or so. The entire prayer consists of three separate passages (only the first passage, from Deuteronomy 6:4, is the deathbed one), all of which is traditionally spoken twice a day in prayer, and at bedtime. The Deuteronomy passage only became associated with dying—particularly under conditions of martyrdom—because a first century rabbi, Akiva, was reciting it when he was struck down by Roman centurions in 135 C.E.
Other ritual last words also amount to a community’s dynamic responses to trends and events. Christians of medieval plague times adopted the script of the ars moriendi, or “the art of dying,” instruction manuals for dying. Some of them prescribed that the dying person say “yes” (in response to questions about their faith) and repeat the last words of Jesus Christ (“into Your hands I commend my spirit”). The explicit instructions were intended for ordinary people who had to fill in for priests whose ranks had been devastated by plagues.
Communities even inspired each other—in the seventeenth century, Jews in Italy began producing books for the sick and dying, possibly because of a resurgence of interest in ars moriendi by Italian Catholics. (I know about this resurgence because it was the context for the death in 1680 of the famous Baroque sculptor, Gian Lorenzo Bernini, who invented a system of hand signals to communicate the ars moriendi with his spiritual confessor and death consultant. A fascinating story for another time.)
But how capable were dying people of uttering the prescribed words? Many of the traditions adapted themselves to the physiological realities of dying. There’s the raised finger of Islam, for instance, that counts as the Shahadah. And ars moriendi gave instructions about what to do if someone couldn’t speak or reason. Sometimes it would be enough for a dying person to “answer in their hearts.” Medieval Japanese Buddhists were told to chant a divine “seed” syllable, “ah,” which evolved (scholars say) because repeating the “ah” increased the likelihood that the actual moment of death would coincide with an “ah.” Also, a single syllable is easier to pronounce than an entire sutra. Meanwhile, in Hinduism, it is more critical that people hear sacred prayers or music than that they speak. Among the priestly Brahmin caste, the karna mantra, or ear mantra, is to be chanted into the dying person’s ear. Rather than expecting spoken language under threatening conditions, this tradition lowers the demands for the dying person.
Despite this flexibility, there’s something of a shared thread of moral judgment running through these traditions, in the sense that how a person comports themself on the deathbed is taken to be a sign of their moral virtue. Scholar Jacqueline Stone, who studied Japanese Buddhist death traditions, called this “moral physiology.” In other words, if a person dies agitated or delirious, they’re thought to be morally corrupted, their soul doomed. But if they’re calm—and maybe even have the presence of mind to say something intelligible—they’re judged to be bound for heaven. Say what you will about the medical turn in approaches to death and dying, but one clear benefit has been the ability to sidestep moral explanations such as these that may inflame emotions and hinder the grieving process.
The resemblances among these traditions matter, I think, for a couple of reasons. I’m not a theologian, so I can’t speak to how they might be connected to foundational views of divinity or creation. But as a linguist, I find that this attention to language underscores how our ability to communicate with intention is taken to mark the limit of the person, at the point that they become unable to participate in social life. There’s a risk, though: if we elevate speech to be the prime vehicle of communication and consciousness at the expense of the nonverbal—or even judicious silence—we might be writing off a person. It also reflects how language, at its core, is a social activity. For our whole lives, we coordinate our activities through language, and it stays that way right to the very end. In this way, no one’s last words are really his or her own—we make them together.
It also means that last words, whether they’re spontaneous and unscripted or passed down as a thing you must do, don’t mark the end of language as much as they do its continuation. And in that, life continues its quickened pace.