Animal Empathy
When Rats Stop to Help a Stranger
A rat, with no reward on offer and no social bond at stake, will reliably free a trapped stranger — and scientists are still arguing about what that means.
The Idea
Empathy has long been treated as one of humanity's defining achievements — the cognitive and emotional capacity to register another's suffering as if it were your own. But the more carefully researchers look at other animals, the more that definition starts to strain at its seams. The question isn't whether animals feel something when another of their kind is in distress. The evidence that they do is now substantial. The more interesting question is what kind of thing they feel — and whether the distinction between 'true' empathy and something functionally identical actually matters. Ethologists now describe empathy as a spectrum with distinct layers: emotional contagion at the base (catching another's distress, the way yawning spreads), then sympathetic concern, then targeted helping. Most mammals appear to operate somewhere in the middle of that spectrum. What's shifted in the last two decades is the rigour with which this has been tested. Rather than relying on anecdote, researchers have designed controlled paradigms — situations where an animal must actively choose to help, incur a cost to do so, and demonstrate that the helping is triggered by the other's distress state specifically, not by some unrelated cue. Under those conditions, rats, elephants, ravens, and several great apes have all cleared the bar. The old assumption that empathy required a theory of mind — a conscious model of another's inner life — is increasingly hard to defend as a prerequisite.
In the World
In 2011, neuroscientist Inbal Ben-Ami Bartal at the University of Chicago set up one of the cleanest tests of animal empathy ever run. She placed a rat inside a small transparent tube — restrained but unharmed — and let a free rat roam the same enclosure. The free rat would show visible signs of distress: increased activity, attempts to dig at the tube, apparent agitation. Over several sessions, most free rats figured out how to open the door and release the trapped one, then spent time investigating their freed companion. Crucially, this happened even with strangers — rats who had never met — though it occurred faster between cage-mates. When the free rat was given a choice between liberating the trapped rat or accessing a cache of chocolate chips (a powerful motivator for rats), it frequently chose to free the other rat first, sometimes sharing the chocolate afterwards. The team subsequently showed that the behaviour was suppressed when the free rat was given an anti-anxiety drug — suggesting the helper was acting to relieve its own empathic distress, not just responding to a social signal. Critics raised counterarguments about social facilitation and learned behaviour. Bartal's group addressed them, methodically, over the following years. What remains is a picture of a small mammal registering another's confinement, being motivated by that registration, and taking deliberate action to end it — with no obvious return benefit.
Why It Matters
The scientific story here does something quietly important to how we think about the boundaries of mind. If empathy is not a single capacity that evolved once in one lineage but rather a cluster of mechanisms that assembled, in different configurations, across many branches of the animal tree — then the question of who deserves moral consideration starts to look different. Not because animals are 'just like us', but because the argument that they lack the relevant inner life is harder to sustain with confidence. This isn't a call to sentimentality. It's a call to epistemic honesty. When we design spaces for animals — whether in research, farming, or urban environments — we are making implicit assumptions about what they experience. The rat studies, and dozens like them, suggest those assumptions have been doing a lot of quiet work that the evidence no longer supports. Sitting with that is uncomfortable in proportion to how many decisions it might implicate. But that discomfort is probably the right response to a genuine finding.
A Question to Ponder
If a creature acts to relieve another's distress at a cost to itself, without any obvious reward, what would it take for you to call that something other than empathy — and is that distinction doing philosophical work or just protecting a comfortable boundary?
Get a new one of these every morning.
Start learning with Thinkable