I TRIED TO LISTEN TO THE VOYNICH MANUSCRIPT
What if we can’t read the world’s most mysterious manuscript because part of its language was never meant to be written?
For more than twenty years, I have sometimes slept with pages of the Voynich Manuscript under my pillow. It has produced exactly zero breakthroughs. But it has left me with a question I cannot shake: what if we have spent a century trying to read something that was never meant to be read the way we read language today? What if the strange writing, plants, stars, women, pools and tubes are parts of a system whose missing piece was sound, rhythm, memory, or even a way of sensing the world that modern humans no longer share?
I became fascinated with the manuscript around 2003. By then it had already defeated generations of historians, code breakers, language experts, botanists and mathematicians. That was part of the attraction. The vellum dates to the early 1400s, yet nobody has convincingly shown what the writing says. Strange plants fill the pages. Women stand in green pools. Stars turn inside wheels. Tubes connect things that seem to have no reason to be connected. The writing flows around all of it almost like a song whose tune has disappeared.
I worked for the railroad, so copies of those pages began traveling with me. During layovers I would spread them across motel beds and stare at them before going to sleep. In Portola, California, I once put a page beneath my pillow and wondered if my sleeping mind might notice something my waking mind had missed. Later, high in the Sierra Nevada, I would look at those same pages while we waited for our turn to move through the long Summit Tunnel. I even tried sleeping near them during breaks, hoping that seven thousand feet of mountain air or gravitational change might somehow shake something loose. Nothing happened. The manuscript stayed silent. My coworkers mostly gave me strange looks.
At home I had another idea. I was married then to a concert pianist with an almost frightening understanding of music. She could play works that seemed to require four hands. I spent years trying to find something that would truly challenge her and rarely succeeded. Ordinary life was another story. She could forget that bath water was running until it overflowed. A freezer door might remain open until everything inside began melting onto the floor. Cooking could turn into an experiment nobody had planned. Her explanation was simple: wasting creative energy on small daily things interfered with the important work. I learned to respect the argument. Our home was filled with classical music. The children occasionally conducted secret missions involving pop songs their mother considered evidence of cultural decline.
Somewhere in that house I began wondering whether the Voynich Manuscript might be music. Not music in the normal sense, but a system in which the symbols carried pitch, rhythm, change, repetition or relationships. I tried turning the writing into notes. My wife looked at some of what I was doing and pointed out a problem: parts of the system I was imagining would fall outside what a human ear could hear. In other words, I had created music for a listener who could not listen.
That should have ended the idea. Instead, it opened a bigger one. Why was I assuming the original listener sensed the world exactly as I do? Humans see only part of the light around us. Other animals hear frequencies we cannot hear, see colors we cannot see, or smell a world we barely know exists. Reality contains far more information than our bodies allow into our minds. We do not experience everything that exists. We experience what our senses are built to receive. So what if an old communication system made distinctions that were once obvious to its users but meaningless to us?
Even ordinary speech already works this way. Take the words “I love you.” “I love you?” is not the same message as “I love you!” Stress the word “I” and the meaning shifts again. Stress “you” and it changes once more. The words remain the same while pitch, timing, stress and context carry extra information. Music does this constantly. A few notes can create a feeling that might take a paragraph to explain.
Another clue came from a courtroom. I was serving on a jury and became fascinated by the court reporter. People spoke quickly, interrupted each other and changed direction halfway through sentences, yet her hands kept moving. She was not writing English in the ordinary way. She was turning speech into another system built for speed, one that could later be turned back into language. That stayed with me. A written sign does not have to mean one letter. It might represent a sound, a phrase, an action, a change, a relationship, or several things at once. Suddenly my question changed. Instead of asking, “What letter does this Voynich symbol mean?” I began asking, “What is this symbol telling the reader to do?”
Years later I discovered that other people had also tried connecting Voynich to music. My first reaction was childish: Damn, somebody got into my rabbit hole. My second thought was better. If several people independently looked at this strange writing and wondered whether its repetition might carry something besides normal prose, perhaps the important question was not which musical theory was right. Perhaps the important question was why we insist that the answer must sound like a modern language at all.
Imagine that the Voynich Manuscript is simply a copy. Nothing supernatural is required. No aliens, no angels, no Atlantis. Just a copy of another document. Then imagine that earlier document was also copied from something older. Do that for enough generations and the symbols might survive after the knowledge needed to understand them disappears. Think of finding sheet music after everyone has forgotten what musical instruments were. Future scholars could count the notes, measure every pattern and publish brilliant studies about their order while missing the simplest fact: somebody was supposed to play it.
Then there are the pictures. I had long suspected that parts of the manuscript involve birth, reproduction, health or some kind of change in the body. There are simply too many women, plants, roots, liquids, containers and cycles for those ideas to be ignored. But perhaps I was still making the same mistake. Maybe these were not scientific drawings for experts. Maybe they were meant to be simple because simple images are easier to remember.
That brought me to nursery rhymes. A cow jumps over the moon. A girl loses her sheep. An egg falls off a wall. The pictures are absurd, yet children remember them for generations because rhythm and images stick in the mind. Now imagine finding one of those rhymes five thousand years from now after the culture that created it has vanished. Scholars might spend careers debating the astronomy of the cow that jumped over the moon. They would not be stupid. They would simply be missing the tune.
So I tried another idea. What if parts of Voynich were songs, memory aids, children’s teaching verses, instructions for midwives, plant knowledge, seasonal sayings, women’s traditions, stories about the stars, or several of these at once? What if the writing did not begin in the fifteenth century at all? The parchment tells us when this surviving copy was made. It does not tell us when every idea inside it was first created. That does not prove an ancient original, Atlantis, or anything supernatural. Those belong to imagination, not evidence. But a copy can preserve ideas much older than the copy itself.
Recently I returned to the problem while working on something completely unrelated. Now I had artificial intelligence available, and that made an old experiment easier. The idea had taken me twenty years to state clearly: do not translate Voynich into English first. Ask what kind of communication system could have produced it.
Maybe the path should be SYMBOL → PATTERN → SOUND OR RELATIONSHIP → IDEA → HISTORICAL LANGUAGE → MODERN ENGLISH. English should come last, not first. Languages do more than give different names to the same things. They divide the world differently. Some ideas that require one word in one language may need a sentence in another. A perfect English translation can therefore still fail to express what the original speaker meant.
I then created another experiment: imaginary missing pages. They are not real Voynich pages and they are not reconstructions. They are visual questions. I asked what a page might look like if the women, water, plants, stars and repeated symbols belonged to a memory system or story rather than a scientific textbook. The results looked strangely believable.
A Visual Experiment with the Missing Leaves of the Voynich Manuscript
The image above is NOT a page from the Voynich Manuscript.
It is something more deliberately speculative: an attempt to imagine what a missing leaf might have looked like if we stopped trying to “solve” the Voynich Manuscript for a moment and instead asked what kind of book its original readers may have believed they were looking at.
That led to an even stranger idea because I happened to be researching Renaissance Rome for another project. What would Caravaggio have thought if somebody had shown him one of these pages? Not a modern scholar who already knows that the Voynich Manuscript is famous. Not a code breaker determined to solve it. Just Caravaggio, busy painting, irritated that somebody has interrupted his light.
So I imagined thirty seconds that never happened. A merchant enters the studio hoping Caravaggio can introduce him to a wealthy collector. He brings a strange vellum page as a sample from a larger book. Caravaggio keeps painting. The merchant talks. Caravaggio keeps painting. The merchant finally moves close enough to block the light falling across the model. That gets Caravaggio’s attention. He glances at the page: strange writing, impossible plants, little women in green water, stars and pipes. The merchant asks what he thinks it is.
“Nursery rhymes.”
“From where?”
“Atlantis.”
“Why would you say that?”
“Because you wanted it to be old.”
Then Caravaggio goes back to painting.
That scene is entirely fiction. There is no evidence that Caravaggio ever saw the Voynich Manuscript or one of its missing leaves. But my imaginary Caravaggio accidentally taught me something. He looked at the page without knowing he was supposed to be impressed.
After twenty-three years, I still cannot read the Voynich Manuscript. The page beneath my pillow said nothing. The mountain air said nothing. The pianist found no melody I could use. The court reporter gave me no secret code. Artificial intelligence has not solved it for me. Atlantis has not returned my calls.
Speculative Historical Reconstruction, Rome, c. 1599
Digital photographic reconstruction, 2026
An imagined interruption in the working world of Michelangelo Merisi da Caravaggio. While the painter attempts to recover an expression from his model, a merchant seeks his attention with two unidentified vellum leaves and, more importantly, an introduction to a wealthy Roman collector.
But perhaps all those failures produced a better question. When we meet an unknown system of communication, how much of what we think we see actually comes from the system, and how much comes from our own minds? We assume writing carries speech because ours usually does. We assume pictures explain words because ours often do. We look for nouns, verbs, alphabets and grammar because those are tools we already know. We keep translating everything toward ourselves.
What if the map was never drawn for us?
Maybe Voynich is a cipher. Maybe it is medicine. Maybe it is a memory system whose missing part was never written down because everybody who used it already knew the song. Maybe it is nonsense. I do not know.
After twenty-three years, I am comfortable saying that.
The little women still stand in their impossible green water beneath impossible stars, waiting for something we may no longer know how to give them.
A voice, perhaps.
The right one.
Maybe somebody was supposed to sing.
Get Phil Tufi’s stories in your inbox
Join Medium for free to get updates from this writer.
VOYNICH SONIFICATION EXPERIMENT
Voynich f75r, heard rather than read. Using its standard EVA transcription, I assigned each unique glyph a fixed frequency in alphabetical order, beginning at 20 Hz; each following glyph rises by ײ⁰·⁷⁵ (about 1.6818). Each glyph plays for 0.10 seconds, with 0.015-second gaps between glyphs, 0.16 seconds between words and 0.45 seconds between lines. Nothing was rearranged.
The full range reaches 81,920 Hz, well beyond human hearing. For the audible version, frequencies below 110 Hz were doubled and those above 1,760 Hz halved by octaves until they entered that range, preserving their basic pitch relationships.
This is not a translation. It is a reproducible sonification of Voynich f75r.
“I didn’t translate the Voynich Manuscript into music. I gave every repeated symbol a repeated sound and listened to the structure.”
