Listen on Apple | Listen on Spotify | Listen on Amazon | RSS Feed
An interstellar exploration crew mistakes Earth for a world of ferocious, weapon-wielding giants after misinterpreting humanity's broadcasted radio signals and television shows.
Become a paid subscriber to listen to all of our episodes completely ad-free! You’ll also get access to a deeper dive on the stories we share and the authors!
There’s a particular comic structure that science fiction does better than almost any other genre: the alien civilization, vastly more advanced than humanity in every measurable way, undone entirely by a failure of interpretation. “Rabbits Have Long Ears” is one of the cleaner executions of that structure, and it’s aged into something almost prophetic — a story about an advanced intelligence drowning in data it cannot actually understand, written eleven years before the moon landing and about seventy years before anyone coined the phrase “garbage in, garbage out” as a critique of machine learning.
The Computer That Knows Everything and Understands Nothing
The Kaar have a genuinely formidable piece of technology: a computer that ingests the entire radio spectrum of a planet, decodes its languages, and cross-references the results into something resembling a comprehensive understanding. It’s the kind of capability that would make any modern data scientist weep with envy.
It is also catastrophically, hilariously wrong, and the story is very precise about why.
The computer pulls its information from human broadcasts without any sense of genre, register, or context. A western pulp novel passage about Blinky drawing his gun gets classified with the same epistemic confidence as a dictionary entry. The excerpt about Lon and his rabbit stew gets treated as documentary evidence about a “rabbit” rather than as one guy’s hunting anecdote. The computer doesn’t know it’s reading fiction. It doesn’t know the difference between a how-to manual and a cowboy story. It has ingested an enormous quantity of human transmission and concatenated all of it into a single undifferentiated account of what humans are like — guns are “five and a quarter man-lengths long” because that’s the only gun it happened to capture, and the dredge operator’s one specimen becomes the computer’s entire dataset for an entire weapon class.
This is, almost literally, the modern problem of training a model on the internet and getting back something that’s confidently, fluently, and completely wrong because it cannot distinguish a folk tale from a fact. Willard wrote this joke in 1958 with a literal alien spaceship, and it still lands as a critique of how we build knowledge from aggregated text today.
The Anthropologists Keep Dying
There’s a small running joke buried in the story’s setup that’s worth pulling out: the Kaar are down to their last anthropologist, having lost the others in earlier disastrous first contacts, and this scarcity is treated as a serious operational problem — more serious, apparently, than losing a radio operator, of which they have plenty on ice.
This detail does double duty. It’s a joke about academic specialists being undervalued and overexposed to risk, which is funny on its own. But it also quietly establishes that the Kaar’s approach to contact has a track record, and the track record is bad. They lost three men to “high I.Q. slugs” in their last encounter, after initially misjudging those slugs as something harmless. They have a documented habit of getting the basic nature of alien species wrong and discovering the error only after people have died.
By the time we get to the rabbit, the story has already trained us not to trust the Kaar’s judgment. We just don’t expect the error to compound itself so completely, or so amusingly.
The Stasis Beam and the Wrong Kind of Caution
Commander Losure is, within the logic of his own civilization, behaving responsibly. He’s cautious. He doesn’t want to lose more crew. He refuses to use the dredge on anything that registers as a high order of thought, which is a genuinely admirable ethical constraint for a species capable of grabbing specimens from a planet’s surface with a force beam. He second-guesses the computer’s conclusions more than once, suspecting — correctly — that something is being lost in translation.
And he’s still wrong, comprehensively, because the one piece of physical evidence he gets — an actual living Earth rabbit, yanked off the lunar surface — happens to be wearing horseshoes. Or what looks like horseshoes. Iron crescents nailed to its feet, which Montresig notices on the way to the airlock and which neither of them bothers to investigate, because by that point they’ve already decided humanity is a race of gun-toting giants and they’re heading home.
The detail is the story’s best joke and its sharpest one. The rabbit is, of course, an Earth rabbit on the lunar surface — meaning the “shoes” can only be something humans put there, evidence of contact and tool use and presence on a celestial body the Kaar assumed was untouched. It’s the single most important piece of information in the entire story, a literal smoking gun about human capability, and Montresig deletes the question from the computer’s memory because “a rabbit’s a rabbit, shoes or no shoes.”
The Kaar had the truth in their hands and threw it out the airlock along with the rabbit.
What the Story Knows About Confirmation Bias
Once the Kaar decide humans are dangerous giants, every subsequent piece of evidence gets folded into that conclusion rather than examined on its own terms. The gun is evidence of giants. The rabbit with long ears, mislabeled by a computer parsing a hunting anecdote, becomes further evidence of a savage, warlike species — even though, if anyone had paused to think about it, an actual rabbit hopping around a hold and getting spooked by approaching crew members should have been a pretty strong clue that something had gone wrong with the data pipeline.
Nobody pauses. Montresig, in particular, has already made up his mind, and every new data point gets read through that lens. Did you hear that? Is there any question in your mind now about the size and ferocity of these people? He’s not reasoning from evidence to conclusion. He’s reasoning from conclusion backward into evidence, the same cognitive shortcut that makes conspiracy theories so durable and first impressions so hard to dislodge.
The Commander is the story’s voice of mild skepticism, and even he gives up in the end — not because he’s convinced, but because the dredge breaks down and he’s out of options and patience. I don’t believe in giants, much less in scientific ones, he says, and then leaves anyway, taking the wrong conclusion home because correcting it would require more effort than the mission has left in it.
The Punchline Underneath the Punchline
The story’s surface joke is the misidentified rabbit. Its real joke, the one that lingers, is that an extremely advanced civilization — four thousand years of spaceflight, hundreds of colonized planets — leaves Earth no wiser than when they arrived, while Earth gains a genuinely valuable strategic outcome without ever knowing contact happened at all.
The Kaar will go home and advise caution rather than evacuation, based on a conclusion that happens, by sheer accident, to keep them away from a planet that — per their own ethical code — they wouldn’t have harmed anyway. Humanity is saved not by any human action, but by a fictional cowboy story and a poorly contextualized hunting anecdote, processed by a computer with no sense of genre.
It’s a deeply silly story, and the silliness is the whole appeal. But there’s something underneath it that holds up: that vast intelligence and enormous data do not, by themselves, produce understanding. You still need someone to ask whether the rabbit is wearing shoes for a reason.
