What the Lemmings Could Not Do: On Suicide, Cognition, and the Mortal Imagination

Of all the acts a human being can perform, suicide is the strangest. It requires the actor to picture a world without itself, judge that world preferable, and execute a plan whose author will not survive to see the result. No other behavior in the human repertoire so cleanly inverts the survival logic that built every body and every brain. The question of whether other animals do the same thing is a question about cognition. The behavior is downstream of cognition, and beneath cognition runs the question of meaning. To kill oneself one must first have the kind of self that can be killed by its owner.

The strict scientific answer is that no verified case of suicide has ever been recorded in a non-human species. A longer answer notes that the conventional framing of “no” collapses a real continuum of self-destructive behavior into a category fight. The fight has stayed politically charged for over a century, because the underlying question concerns human exceptionalism.

Émile Durkheim drew the modern boundary in Le Suicide (1897). He argued that self-destruction in animals could not count as suicide because animals lack the conscious foresight required, and he did so against a body of nineteenth-century writers who claimed the opposite. Bidie’s 1874 report of scorpions stinging themselves to death when surrounded by fire was the most cited case for the other side; later experimental work showed the behavior was reflexive heat escape, an early test of what became Morgan’s Canon, the principle that simpler psychological explanations should be preferred when they fit the evidence. Durkheim’s line held for almost a century. Then came the lemmings.

The lemming case deserves a careful retelling because it shows how completely a culture will manufacture a suicide story when it wants one. Walt Disney’s 1958 documentary White Wilderness, part of the True-Life Adventure series, won the Academy Award for Best Documentary Feature on the strength of a sequence in which lemmings appeared to plunge en masse off a cliff into the sea. The CBC’s The Fifth Estate sent Bob McKeown to investigate in 1982. His crew demonstrated that the footage had been filmed at the Bow River near Canmore, Alberta, a landlocked province where the lemming species shown does not even live, that Inuit children in Manitoba had been paid to wrangle the animals, and that the crew had placed them on a snow-covered turntable and pushed them off a cliff in batches. James R. Simon was the principal photographer responsible for the sequence. Lemmings disperse during population booms and occasionally drown crossing water; what White Wilderness presented as collective despair was accidental death during ecological dispersal.

The cultural afterlife of that fraud is its own data point. “Lemming” entered English as a synonym for self-destructive conformity. Political campaign ads borrowed the image for decades. A staged scene of rodents being shoved off a cliff became the parable for everything from groupthink to mass psychology, and the same parable then became the convenient cudgel used to dismiss every anecdotal claim of animal suicide that came afterward. The story did double work. It gave us a nature to project our pathology onto, and then it gave us a permission slip to deny that nature any inner life at all. Both moves were sloppy and both were revealing. We invented animal suicide because we wanted to find ourselves in nature, and we denied animal suicide because we wanted not to find nature in ourselves.

What survives serious scrutiny falls in two directions. In one direction sit the cases that look like suicide and are not. Pacific octopuses stop eating and waste away after reproducing, but every individual of the species does this, and the behavior is triggered by the optic gland; surgical removal of the gland prevents the post-reproductive decline. Pacific salmon die after spawning by programmed senescence, an evolutionary economy that converts the parent’s body into nutrients for the offspring’s stream. Honeybees die when they sting because the barbed stinger tears free of the abdomen, an outcome of anatomy alone. Pea aphids of the species Acyrthosiphon pisum sometimes burst themselves with sticky secretions to gum up predators threatening their colony, an act of kin-selected altruism described by Hamilton’s rule. Carpenter ants infected with the fungus Ophiocordyceps unilateralis climb to elevated leaves, clamp their mandibles in a death grip, and die as the fungus erupts from their heads, an outcome produced by parasitic manipulation of host nervous tissue. Each case is biology functioning as biology should, with no inner authorship of the death.

In the other direction sit the cases that resist easy dismissal. A male bottlenose dolphin named Peter, a subject in John C. Lilly’s interspecies-communication experiments at the Communication Research Institute on St. Thomas in the early 1960s, formed a deep bond with the volunteer researcher Margaret Howe Lovatt and stopped breathing after the program lost funding and he was moved to a smaller tank in Miami. Cetaceans are conscious breathers, which is to say each breath is voluntary, which makes voluntary cessation at least conceptually available to them. Ric O’Barry, the trainer who worked with the dolphins on the 1960s television show Flipper, has given a similar account of Kathy, one of the dolphins who played the title role, suffocating herself before him in 1970, an event he describes as the moment that turned him from trainer to abolitionist. Both accounts are anecdotal and unverified by independent observation, both involved experimental conditions that were themselves ethically indefensible, and both nonetheless describe a behavior, voluntary breath cessation in a self-aware mammal, that resists reduction to reflex.

The list extends. Captive primates show chronic self-injurious behavior at rates that climb with social isolation and confinement, and while self-injury is a separate clinical category from suicide, the despair underneath it is documented across decades of zoo and laboratory medicine. Domestic dogs in well-attested veterinary cases refuse food and water after the loss of an attachment figure and die. The Greyfriars Bobby tradition romanticizes this pattern, and the romanticization should not blind us to the underlying fact, which is that mammalian grief can be lethal. Elephants display extended mourning behavior at the bodies and bones of their dead. Corvids hold what observers have called funerals around dead conspecifics. The animal world is rich in recognition of death; what we cannot find in it, on the evidence available, is the deliberate authorship of one’s own.

That distinction is where cognition does its work. To commit suicide in the human sense requires four cognitive capacities, and they are not optional. The actor must possess an autobiographical self, a continuous narrator who carries the thread of identity across time. A mortality concept is also required, a working understanding that the self in question is a thing that can stop. Future-oriented planning enters next, the capacity to model a state of affairs that does not yet exist and to act in service of bringing it about. Add to all of this what Thomas Joiner has called the “acquired capability for lethal self-injury,” the ability to override the survival reflexes that every nervous system installs by default. The combination is rare. Some elements appear in elephants, in great apes, in cetaceans, in corvids. The full package, by the best evidence we have, appears only in Homo sapiens.

The current scholarly disagreement runs along that seam. David M. Peña-Guzmán argues, in his 2017 Animal Sentience paper and in subsequent work, that suicide should be reconceived as a continuum of self-destructive and conservation-withdrawal behaviors rather than an act gated by reflective intent, and that the intent requirement is conceptually shaky even for human suicide research. Carolyn Ristau, a cognitive ethologist working in the Donald Griffin tradition, replies that Peña-Guzmán reads death recognition into behaviors that admit simpler explanations. Reviews of the ethological literature by Denys de Catanzaro (1980) and Antonio Preti (2007) found no verified field case meeting the strict criterion. The dispute will not resolve cleanly, because the question is partly empirical and partly definitional. If you require reflective intent, the answer is no. If you require only self-destructive behavior with a recognizable affective component, the answer is yes for several species. The disagreement is about which definition deserves the word.

The deeper question, the one this entire literature circles without quite naming, is what the human capacity for suicide tells us about meaning. To take one’s own life requires the antecedent capacity to find life meaningful or meaningless. An animal cannot judge its existence and reject it because it cannot stand outside its existence to judge. Human beings can, and the cost of that capacity is the possibility of its extreme exercise. The mortal imagination is the same faculty that lets us write wills, plan retirements, build cathedrals across centuries, and pre-arrange our own funerals. It is also the faculty that lets us decide we are done. The capacity for transcendent meaning and the capacity for terminal despair are the same capacity. They are mortared together at the cognitive foundation, and you cannot have one without the architectural possibility of the other.

This is why the animal cases matter in a direction opposite to the one usually claimed. Recognizing that no other species commits suicide does its real work by showing what the achievement costs. The animals that grieve, that starve themselves after losing a bonded partner, that beat their heads against the walls of their cages, are showing us what suffering looks like before it has been organized into a self-killing project. They feel, and what they feel can kill them, but the killing is not their work. The killing is what despair does to a body that has not yet been given the cognitive equipment to do it on purpose.

Human beings have that equipment. We are the species that can review our lives and find the verdict wanting. We are the species that can imagine the world after our absence and weigh it against the world during our presence and choose. The lemmings could not do this. Peter the dolphin came closer than any other animal we have observed, and even that case sits inside enclosures we built and trauma we inflicted, which means we cannot be sure we are watching despair rather than cardiac collapse under stress. The line between human suicide and animal self-destruction runs along cognitive grounds, and treating it as a moral hierarchy has caused most of the historical confusion. Both kinds of dying are real, and only one is authored.

What follows from this for the meaning of life is unsentimental. The capacity for suicide is a load-bearing feature of the same cognitive structure that lets a human being recognize life can have meaning at all. Removing it would require dismantling the architecture that makes meaning possible in the first place. Meaning is the conscious assignment of value to a stretch of time. To assign value, one must be able to imagine the alternative. The animal who cannot conceive its own non-existence cannot assign meaning to its existence either; it can only live the existence it has. Meaning, in the strong human sense, is reserved for the species that can also negate it. The suicide line is the meaning line. They run along the same cognitive ridge, and they were always going to.

These observations describe the architecture of being human, the room we occupy, the cognitive shape that makes our particular kind of meaning possible. If you can ask whether your life is worth living, you are already inside the cognitive architecture that makes the question intelligible, and so is everyone you love. The animals around us suffer in their own register, and that suffering deserves the moral seriousness we usually reserve for our own. They cannot kill themselves. They can be killed, slowly, by what we do to them, by what is done to them, by losses they cannot articulate and cannot escape. Their catastrophe occupies its own register. The line that separates it from human suicide is drawn by authorship alone, with no hierarchy of worth implied.

The human alone holds the pen. The human alone can close the book. That this is true confers no special elevation on our species; the only thing it explains is why our literature has always been so preoccupied with whether to keep reading.