Ravens: behavior & cognition
The common raven (Corvus corax) is one of the most studied birds in comparative cognition, and much of what is known about it comes from a small number of long-running research programmes, notably aviary work by Bernd Heinrich and Thomas Bugnyar and field observation across the Northern Hemisphere. This profile summarises three well-documented behaviour areas: problem-solving in food-access tasks, the social cognition that surrounds food caching and pilfering, and play.
These behaviours are described as ethologists report them, with the type of evidence and its limits noted for each. Many of the most striking findings come from captive or semi-captive birds, where conditions can be controlled but may not match wild life. Where interpretations are contested in the literature, that is stated rather than smoothed over.
Pulling food up on a string
In a much-cited series of experiments, ravens were presented with a piece of meat hung from a perch on a long string, a situation they would not normally meet in the wild. To get the food, a bird has to reach down, pull up a length of string with its bill, hold each loop underfoot, and repeat the sequence. Heinrich and Bugnyar reported that several naive ravens performed a coherent pull-and-step sequence soon after first encountering the setup, rather than only converging on it by slow trial and error.
The researchers framed this as possible evidence that ravens grasp something about the relationship between the string, the food, and their own actions, rather than acting on a fixed instinct. The behaviour is repeatable across individuals and is a standard reference point in corvid cognition, but it is best read as flexible problem-solving in a controlled task, not as proof of human-style reasoning.
Caveat: Whether string-pulling reflects genuine insight or rapid associative learning is actively debated, and not every bird solves it; results come from a small number of captive individuals and should not be generalised into claims about ravens being uniquely or measurably 'smart.'
Caching, pilfering, and watching who watches
Ravens store surplus food in scattered hidden caches and also raid caches made by others, which sets up a competition between hiding food and stealing it. Aviary studies by Bugnyar, Heinrich, and colleagues found that ravens that had watched another bird cache could later relocate those caches using observational spatial memory, and that cachers behaved differently depending on whether a competitor had been able to see the hiding event. Birds tended to cache out of view, delay, or move food when a potential pilferer was present and had the chance to observe.
Follow-up work reported that ravens distinguished between a competitor that had witnessed caching and one that had been prevented from seeing it, adjusting their own caution accordingly. Researchers discuss this in terms of tracking what another individual may or may not have seen. This is a careful, evidence-bounded claim about visual access, not a demonstration that ravens hold human-like beliefs about others.
Caveat: Most of these findings come from captive or semi-captive birds in designed tests; whether they amount to 'tactical deception' or theory of mind is contested, and simpler explanations based on attending to a competitor's behaviour are not fully ruled out.
Sliding, dropping, and tug-of-war with objects
Ravens are among the more frequently reported playful birds. Observers in Alaska and northern Canada have described ravens repeatedly sliding down snowbanks and snow-covered roofs, then walking or flying back up to slide again, and birds have been seen rolling down snow on their backs while holding a stick in the feet. In the air, ravens carry, drop, and re-catch sticks and other objects, and two birds may tug at the same stick in a contest sometimes likened to tug-of-war.
A study of free-flying ravens documented patterns in object play, including a decrease in object play with age and longer play bouts with novel objects, and play has also been recorded in nestlings. Ethologists treat these as play because the actions are repeated, varied, and not tied to an immediate survival payoff, while noting that play may help young birds practise motor skills or assess social partners.
Caveat: The function of raven play is not settled and is easy to over-read as 'fun' or emotion; much locomotor-play evidence is anecdotal field observation, and the more systematic object-play data come from particular study groups rather than the whole species.
How this profile is sourced
Behavior claims here are drawn cautiously from institution-backed references and described with their evidence context and limits. See animal research sources for the methodology, the behavior cluster hub for the wider topic, and animal senses & adaptations for the underlying biology.
Frequently asked questions
- Can ravens really solve puzzles, or are they just trained?
- In string-pulling experiments, naive ravens that had not been trained sometimes produced the full pull-and-step sequence soon after first seeing the apparatus, which researchers describe as flexible problem-solving. Whether this reflects true insight or fast associative learning is debated, and not every bird succeeds. These are controlled captive studies, so they show what ravens can do in a test setup, not a measurable intelligence score.
- How do ravens know where other ravens have hidden food?
- Aviary studies show ravens can remember the locations of caches they watched another bird make, an ability called observational spatial memory, and use it to pilfer those caches later. Cachers in turn adjust their hiding when a competitor could see them. This is documented as tracking visual access and a competitor's behaviour; researchers are cautious about whether it amounts to understanding another animal's mind.
- Do ravens play for fun?
- Ravens show behaviours ethologists classify as play, such as sliding down snow, dropping and catching objects in flight, and tugging sticks with another bird. Juveniles are especially playful and play less with age. Because the actions are repeated and not tied to an immediate payoff, they fit definitions of play, but the function is uncertain and labelling it 'fun' goes beyond what the evidence shows.
Last updated: