The Voynich Manuscript

The first axis considers the manuscript as a sequence of symbols that can be converted into numbers and then into bits. Several conventions were tested because there is no independent reason to choose one unique character-to-number mapping.

Structures appear: bands, repetitions and density regions. However, these patterns are compatible with page layout, glyph repetition and the chosen representation. No obvious hidden message was recovered.

The binary sequence was then examined as a discrete series to look for repetitions or preferred frequencies, with a broader analysis across the manuscript's 209 pages. The result is strongly negative in the narrow sense: no single periodicity emerges as an indisputable signature of an external encoding. The experiment nevertheless supports examining the manuscript as a data sequence rather than only as a text.

Words and characters were converted into positions, lengths and values to test whether the spatial organization could generate a figure when the presumed meaning of the glyphs is removed.

About fifty pages were subjected to a search for regularity and symmetry. The highest scores remain mainly explainable by lines, margins and text blocks. No sufficiently stable hidden drawing was identified.

Binary groups (4 to 16 bits) were converted into X/Y coordinate pairs to look for a map, constellation or spiral that might emerge from the noise.

The resulting point clouds do not produce a clearly recognizable map, constellation or spiral. The main value of this branch is methodological: it provides useful negative controls for evaluating the other avenues.

A central idea in the research was to treat a binary sequence as temporal information. Two binary states can be represented by two distinct frequencies: this is the principle of FSK (Frequency Shift Keying).

An initial experimental construction used frequencies around 900 Hz and 1900 Hz. Several WAV/PCM and MP3 files were generated and examined using spectrograms.

After generating the signal, the bits were grouped into 8-bit bytes. Several offsets were tested, together with bit inversion, bit order reversal and several textual representations.

One experimental sequence obtained was:
Y..9n.92.A=.P.0..Qe0S...D.L..9..2..

This string contains printable ASCII characters, but it does not constitute a readable message. Broader searches were also performed using UTF-8, hexadecimal and Base64. Conclusion: no stable linguistic string was obtained — the result is exploratory rather than a decoding.

The next idea was to reduce the information to four states, allowing a natural correspondence with the four DNA bases: A, C, G and T. Mathematically, four states correspond to two bits. With four symbols there are 4! = 24 possible permutations; the different correspondences were compared, and one particular mapping, noted GACT, stood out on the studied data (max ORF = 583 codons).

An obtained excerpt does indeed resemble a biological sequence, for example GTCTCTTCAAGTGATATCCACATCCTCGCGGTCCAATGGATGGGCCAAT… However, such a resemblance is expected whenever an arbitrary sequence is converted into four letters.

The sequence was then examined for motifs such as ATG, reading frames and potentially long sequences, compared with shuffled sequences. The critical point: an ORF can occur by chance. For a biological hypothesis to become convincing, signatures would need to clearly exceed random distributions and reproduce across independent sections.

The DNA branch therefore remains an interesting experimental hypothesis, but it does not currently support a conclusion that a genome or real biological information is present.

Another line of inquiry looked for a proportion close to the golden ratio φ ≈ 1.618 in the manuscript's circular diagrams (the "Rosette").

Three prominent radii measured were approximately 29 px, 62 px and 78 px, giving ratios of about 2.138, 1.258 and 2.690 — they do not constitute a golden-ratio signature. The logarithmic fit performed on a blue motif was approximately r = 227.8 exp(−0.0122θ), with a factor per turn close to 0.926: clearly different from a classical golden-growth spiral.

Conclusion: the circular geometry is real and interesting, but the golden ratio was not demonstrated by these measurements.

The astronomical branch is qualitatively different from the digital transformations: it relies on elements explicitly drawn in the manuscript. The celestial folios contain stars, circles and organizations that can be compared with configurations of the sky.

The seven-star group on f68r3 is a particularly interesting starting point for astronomical analysis. Precise identification of every star and dating of the drawing remain separate questions.

Exploratory astronomical calculations showed that there are historical dates on which the Moon can lie very close to the Pleiades region. Such a coincidence is interesting, but it is not sufficient to date the manuscript: several celestial objects must be compared simultaneously.

The comparison arose from a visual resemblance between some Voynich glyphs and modern alphabets presented under the name "Light Language": loops, hooks, spirals and repetitive forms.

The resemblance remains graphical. No Voynich → Light Language correspondence table has been found that decodes the text using a reproducible grammar and vocabulary. This branch must therefore remain separate from any historical conclusion.

The idea was to test whether information might be hidden by incorrect synchronization or an incorrect line width, like a scrambled image or signal. No stable content was recovered.

The experiment nevertheless highlighted an essential principle: strongly changing the representation of data can produce highly convincing patterns even when no message is actually present.

The comparison with the Wow! Signal is conceptual. The goal is not to claim a historical relationship, but to test a general idea: information can be encoded in a representation that resembles noise or a signal until the appropriate decoding method is applied.

No data from this research demonstrates that the manuscript contains the Wow! Signal or an extraterrestrial transmission. The comparison serves as a framework for thinking about coding, carriers, synchronization and weak-signal detection.

The extraterrestrial hypothesis formed part of the initial questioning and partly explains the search for unconventional codes, signals, astronomical structures and multi-modal representations.

It cannot, however, be inferred from the current results. Even a future discovery of an unusual code would not, by itself, identify its origin. Independent evidence concerning the support, dating, technology, historical context or provenance would be required.