Translation is not understanding; it is an act of care.

We study whales—particularly sperm whales—because they are, in many ways, already alien. They left the land, became mammals, and then, some thirty million years ago, decided to return to the sea. They live in a world we can barely imagine: dense, dark, and almost entirely acoustic. What they know, they know through sound. Every gesture, every encounter, is heard before it is seen.

Whales allow us to test what it means to recognize intelligence in another form—one built on resonance rather than vision. They are enormous, social, migratory creatures whose minds share knowledge across oceans. Historical records suggest that they even learned to evade whalers: early hunters found them easy prey, but within a few generations, the whales adapted, warning each other of danger. Communication spread faster than the ships themselves. Knowledge moved through sound. That story already contains a theory of language.

The sperm whale’s clicks and codas are not songs but patterns exchanged through water. They carry tempo, rhythm, and rubato; they speed up, slow down, and modulate. Communication here is not in the symbols but in the music. Each coda is a phrase of social life. We have found identity codas that function as clan signatures and others that seem to mark proximity, recognition, or belonging. When whales from different clans meet in the same waters, their accents drift toward one another. They begin to speak more similarly—a kind of acoustic convergence that mirrors how human dialects blend at a border.

We don’t have a Rosetta Stone, so we listen. We can detect syntax but not semantics; we can describe the music but not the message. Yet even this limited form of translation reveals something profound about communication itself. The structure is there—the repetitions, the timing, the shared rhythm—but its meaning remains submerged. What we can do is attend to that structure, to learn what understanding might sound like before it arrives.

At Project CETI, we build models that listen the way ears hear melody. They find patterns that recur in specific social contexts, motifs that correspond to certain behaviors. We can now generate synthetic codas so convincing that even whale biologists cannot tell them from real ones—and still, we do not know what the sounds mean. That uncertainty is the humbling part. Science moves forward through error; philosophy moves through doubt. Between them lies the practice of listening.

The Whale Relic emerges from this work. It is not a translator that explains, but an instrument that listens. Each point of light within it is drawn from an actual recording—a coda that once rippled through the waters off Dominica or the Galápagos. When activated, it plays the original sound and produces an imagined human response: Recognition, Gathering, Unknown Signal. The brightness of each glyph shifts with the model’s confidence, echoing how our algorithms test their own understanding. It is a conversation that never resolves, a translation machine that knows its limits. Some signals are stable, others blur or dissolve entirely. The Relic shows not what we know, but what we are still learning to hear.

To translate whales is not to decode them; it is to meet them in uncertainty. The ethical act is to listen before speaking. Understanding here is a spectrum, not a destination. The Relic captures that gradient—the threshold between data and empathy, between sound and sense. Each pulse of light becomes a hypothesis about connection, each click a small act of care.

Perhaps, as when Roger and Katy Payne first recorded whale songs in the 1970s, a new wave of empathy begins here—not from what we understand, but from what we finally learn to hear. To translate another species is to learn humility at the scale of the planet. What we learn from whales is not their language, but our own capacity to listen.