Embodied knowledge is a body of knowledge
What greek lament, embrace of the serpent and the dreaming ones still teach us
Inspired by discussions with my friend Anastasia Sylaidi.
There’s a scene in Embrace of the Serpent I can’t shake.
The shaman Karamakate—last of his people, last speaker of his language—places his hand on a rubber collector’s chest and listens. Not with a stethoscope. Not with data. He listens with a cosmology. And when he finds only hollowness, only forgetting, he recoils. A body that no longer holds memory is a body unmoored.
That moment lands hard because I’ve seen it before—in a different register—in the Mani peninsula of southern Greece. There, for generations, women have sung the moiroloi—funeral laments not as performance, but as practice. They sang to remember. Not just the dead, but the contours of the land, the lineages of kin, the echoes of war, exile, and drought. These were cartographies of grief, sung with accuracy and improvisation. The voice itself became archive—held not in paper or code, but in the tremor of breath, the timing of tears, the communal act of mourning.
No written record could substitute for this transmission. The knowledge lived in muscle memory, in cadence, in the practiced tremor of the voice—passed from mother to daughter not through schooling, but through presence. To witness a lament was to witness continuity.
Today, the moiroloi are fading, and you go cry to your psychologist…Not because they ceased to matter, but because the conditions that held them—village time, intergenerational proximity, women’s ritual authority—have been dissolved. What remains is documentary footage, academic papers and occasional reenactments, mostly for anthropologists and tourists.
Or, take the loss of indigenous languages. What we see in embodied knowledge systems is often high information density with low redundancy—unique associations between practices and language, held by specific people, in specific places, through specific rituals. A landmark study in PNAS (Cámara-Leret & Bascompte, 2021) found that over 75% of medicinal plant knowledge in regions like the Amazon and New Guinea is linguistically unique—known to only one language. When the language dies, the knowledge dies. There is no redundancy. No backup. Each language is a Rosetta stone to a particular way of healing, sensing, coexisting. The tragedy isn’t just that we are losing information—it’s that the only way to keep it alive was to live it.
We see this also echoed in e.g buddhist lineages, where teachings are passed orally, body to body, in lines of transmission. Not because they couldn’t be written down—but because they shouldn’t be severed from the energetic, ethical, and embodied readiness required to receive them. When a master dies without transmitting a teaching, it doesn’t go missing—it ceases to exist.
What connects the Amazonian shaman, the Mani lamenters, and the disappearing languages of the world is this: they each hold forms of knowledge that are not informational, but transformational; sensorial epistemologies—ways of knowing that emerge from immersion, from co-evolution with landscape, from the embodied intelligence of those whose lives have been shaped by the land and who shape it in return.
Real knowledge in these systems is metabolised. Knowledge that doesn’t simply tell you something—it does something to you. It calibrates the senses. It thickens your relation to place. It reorganizes what you pay attention to. If it doesn’t, it’s not knowledge—it’s artefact.
Here, I’m reminded of the Ongee people of the Andaman Islands, where dreams are not private reveries but acts of social weaving. The anthropologist Vishvajit Pandya describes how, each night before sleep, the community would recount their dreams from the night before—retracing the paths their inner selves had walked across the island, collecting smells and memories left behind during the day. These dream narratives were not merely reported; they were negotiated, collectively adjusted so that each person’s dream aligned with the others’, until the individual webs formed a shared architecture. “When the entire community talks through their dreams before sleeping then the individual webs are woven together into a single web over the whole community.” We often forget that dreams can be infrastructural. This is a distributed cognition that is spatial, mnemonic, and deeply ecological. It is through such architectures that memory is consolidated, intersubjectivity maintained, and cosmology made real.
And yet, our modern institutions—scientific, philanthropic, educational—are structured to valorise what can be extracted, replicated, scaled. We call this objectivity. Modernity is accelerating the collapse of such architectures—not just through cultural loss, but through a deeper severance of knowledge from life. We digitise chants, model traditions with AI, and file sacred gestures into institutional repositories. We store more than we can ever embody. Archive more than we can metabolise. As if preservation was the point. As if uploading a practice was the same as understanding it. It isn’t. You can’t preserve metabolised knowledge. You can only transmit it.
And transmission requires context: slowness, presence, repetition, ritual, relationship. It requires bodies and breath and contradiction. It refuses scale. So the real crisis isn’t that knowledge is disappearing. It’s that we’ve stopped investing in the conditions that allow it to stay alive.
If you are a steward of resources—time, money, space—the invitation is this: shift from collection to companionship. From recording knowledge to restoring the lifeworlds that sustain it. Instead of funding the archive, we must fund the people. Not only the database of chants, but the grandmother who still sings them. Not only the plant index, but the elder who knows when to harvest under which moon. Not only the metadata, but the dreamweaver who still speaks in relational tongues. There is, I think, a different kind of innovation hidden here. One that listens not for novelty, but for continuity. It begins not with the impulse to save, but with the humility to witness.
And perhaps this is where we begin, too. If you are simply a human trying to remember how to remember, begin here:
What knowledge lives in your body?
What practices changed how you grieve, speak, notice, move?
What have you inherited—not in books, but in gestures, rhythms, silences?
And where are you complicit in systems that flatten, extract, translate without listening?
Then begin small. Sit in the room where the last lament is sung—not to record it, but to bear witness. Support the teacher who refuses to publish. Learn a plant by tending it. Value the healer who teaches through apprenticeship, not PDFs. Let some things remain untranslatable.
Because in the end, the goal is not to save every song. It is to live in a world where songs still shape the way we live.