Kaeru-chan — cooperation that has to be carried

A frog that cannot move on its own is a strange thing to build a study around. That is the whole point. One hundred small figurines, each named Kaeru-chan, are trying to get home to a community makerspace in the mountains of Komoro, Nagano. Every one of them is stranded until a person decides to help — and kaeru is the Japanese word for both “frog” and “to return home,” so the pun is doing some of the work too.

A frog that cannot move on its own

The rule is deliberately austere. A Kaeru-chan moves only when a person who finds it makes a choice: carry it to the makerspace directly, or entrust it to someone else who will carry it a little further. Nobody is staffing the journey. There is no schedule, no route, and no central authority deciding where any frog goes next. The objects themselves do the asking, and every leg of the trip is a small act of cooperation between strangers who may never meet again.

The frogs are seeded at 25–35 partner sites and released into the country. Some will travel town to town, coast to coast; the study aims to reach as many as 3,000 people before it is done. And the design has teeth: if a chain breaks — if a frog is set down and forgotten, or entrusted to someone who never carries it on — that Kaeru-chan re-spawns at its starting point and the chain begins again. Cooperation here is not assumed. It has to be re-earned every time it lapses, which is exactly what makes the successful chains worth measuring.

Why a cognitive science lab is co-PI

On its face this reads like an art project, and in one register it is. But it is also one of the cleanest field instruments we have seen for a question the lab studies constantly in the abstract: how does cooperation propagate through a network that has no coordinator? No dispatcher routes the frogs. No reputation system scores the carriers. What holds a chain together is trust handed from one person to the next, mediated entirely by an object that carries its own story and its own provenance rather than by any authority above the participants.

That is a decentralized, staffless network in the wild — and the frog is the mechanism that recruits people into it. Each Kaeru-chan is a way of eliciting spontaneous cooperation among strangers, at a scale and in a naturalness that is almost impossible to manufacture inside a lab. We can build agents that pass favors down a graph and read off when cooperation survives and when it collapses; what we rarely get is thousands of real people doing the passing, in real towns, with real stakes about whether a small ceramic frog ever makes it home. Kaeru-chan gives us the second thing, and lets us ask whether the first thing predicts it.

The object as instrument

Every frog carries a small NFC chip, and the chip does double duty. For the human helper, it is a guide: it tells the frog’s story, shows where the frog has already been, and lets each new caretaker pick up the thread from whoever held it last. For the research team, that same tap is a measurement. It records the collective journeys the frogs trace across the map and captures the social interactions each handoff sets in motion — who passed to whom, how far, how fast, and where the chains thrive or stall.

So the object is at once the narrator and the sensor. The provenance a carrier reads to decide whether to help is the same provenance that becomes our data. There is no separate instrumentation layer sitting outside the cooperation; the thing that makes the cooperation legible to the next stranger is the thing that makes it legible to us. For a lab interested in how information and trust travel through social structure, that collapse of medium and measurement is the design’s best idea.

The collaboration

Kaeru-chan is a joint project with Michelle Huang, an artist-researcher and founder of Akiya Collective — a nonprofit social R&D lab that turns Japan’s abandoned rural houses (akiya) into installations, residencies, and research sites. As a Visiting Researcher at Chiba Institute of Technology, she is co-PI with the lab on this nationwide participatory study of cooperation through social objects. Her broader body of work is at michellekhuang.com. The makerspace the frogs are trying to reach is not a metaphor; it is a real place in the mountains, and the journey home is the experiment.

Where this goes

The frogs let us watch cooperation assemble itself, one handoff at a time, across a whole country — the kind of large-N, naturalistic social behavior our other work usually has to approximate with models and small studies. As the chains accumulate, the questions become tractable ones: what makes a stranger decide to carry a frog, what kind of network topology emerges when no one is drawing it, and which structures let cooperation persist rather than re-spawn.

Related: our work on teaching by changing the world asks a neighboring question about how coordination arises between agents when no one can simply issue instructions — Kaeru-chan is that question let loose on three thousand people and a hundred frogs.