Crows Have Been Observed Holding Funerals for Their Dead—And Then Investigating the Cause

For years, birdwatchers have reported strange behavior: when a crow dies, other crows gather around the body, calling loudly, sometimes touching the corpse with their beaks, then flying off in apparent agitation. Scientists called these gatherings "cacophonous aggregations" and assumed they were simply alarm responses. But new research from the University of Washington's Avian Behavior Lab suggests something far more interesting: crows may be holding a kind of funeral, and then conducting an investigation to learn about potential threats.

Dr. Kaeli Swift, who has studied crow behavior for over a decade, designed a clever experiment. Her team placed dead crows (ethically sourced from wildlife rehabilitation centers) in urban and suburban locations frequented by wild crows. They then observed the reactions of live crows from a hidden blind, recording every behavior with high-speed video and directional microphones.

The results, published in Animal Behaviour, show a consistent three-phase sequence. First, the discovery phase: one or two crows spot the body and emit a loud, repetitive "scolding" call that draws others from as far as a kilometer away. Within 10 minutes, the aggregation can swell to over 50 individuals, all calling in a chorus that Swift describes as "strikingly different from their normal mobbing calls."

Second, the investigation phase: one or more crows (usually the ones that discovered the body) approach the corpse cautiously, often from downwind. They may peck gently at the body, pull at feathers, or roll the corpse over with their beaks. In 14% of trials, a crow picked up a small object—a twig, a piece of litter, a pebble—and dropped it onto the corpse. Swift calls this "offering behavior," though she cautions against assuming human-like mourning. "We cannot know what they feel," she said. "But the behavior is ritualized and consistent. They are not just reacting to a novel stimulus. They are following a script."

Third, the avoidance phase: after the investigation, the crows fly to nearby perches and fall silent. For the next 24 to 48 hours, they avoid the area where the corpse was found, even if that area contains abundant food. Swift tested this by placing hot dogs (a crow favorite) near the corpse site. Crows that had participated in a funeral avoided the hot dogs for two days, while control crows that had not seen a corpse ate immediately.

"That avoidance is the key," Swift explained. "If the funeral was just a general alarm, you would expect them to avoid the entire area forever, or at least until the perceived threat leaves. But they only avoid for two days. That suggests they are doing something more specific: they are learning about the cause of death, then determining how long that cause remains dangerous."

To confirm the learning hypothesis, Swift ran a second experiment where the dead crows were placed alongside a stuffed hawk (a predator) or a stuffed songbird (non-threatening). Crows that saw a corpse with a predator present showed prolonged avoidance—five days instead of two—and their alarm calls were louder and more frequent. Crows that saw a corpse with a harmless songbird present showed normal two-day avoidance. The crows were clearly associating the corpse with the type of nearby animal.

Most startlingly, crows that found a corpse in an area where a human had previously trapped and banded them (a negative experience) avoided that specific human for over a month. If the same human appeared in a different location without a corpse, the crows ignored them. The crows remembered the face, the location, and the context, and they updated their risk assessment accordingly.

"That is not just learning," Swift said. "That is reasoning. They are asking: what killed this crow? Is that thing still here? If I see it again, should I be scared? Those are exactly the questions a detective asks at a crime scene."

The research has implications beyond academic curiosity. Crows are expanding their range into cities worldwide, and human-crow conflicts are increasing. If crows can identify individual humans who threaten them and communicate that information to other crows—which Swift's previous work has shown they can—then a single negative encounter with a person could poison that person's relationship with an entire neighborhood crow flock for years.

"Do not throw rocks at crows," Swift advised. "They will remember your face. They will tell their friends. And their friends will tell their friends. Ten years from now, crows you have never met will dive-bomb you because your reputation precedes you. That is not a superstition. That is science."

As for the question of whether crows grieve, Swift remains cautious. "Grief implies an internal emotional state we cannot measure," she said. "But the behaviors—the gathering, the silence, the avoidance—those are the same behaviors we see in humans who have lost someone close. Whether we call it grief or not, it is a response to death that is far more sophisticated than anyone expected from a bird."