1.0 Introduction

The Voynich Manuscript is a richly illustrated codex dating from the early fifteenth century, written in an unknown script that has baffled scholars and cryptographers for decades. Discovered in 1912 by the book dealer Wilfrid Voynich, this mysterious manuscript has resisted all attempts at translation or decipherment (Beinecke Library, 2016). With its undeciphered text, often referred to as Voynichese, and otherworldly illustrations of plants, astrological diagrams, and bathing nude figures, the Voynich Manuscript is frequently described as the world’s most mysterious book (Harkness, 2016). Researchers from fields as varied as medieval history, linguistics, cryptography, and manuscript studies have all been drawn to the challenge it presents. Despite extensive scientific analyses and numerous proposed interpretations, no consensus has yet emerged regarding the manuscript’s meaning, authorship, or purpose. It remains a true enigma in the history of literature and cryptography (Montemurro & Zanette, 2013). This report provides a comprehensive overview of the Voynich Manuscript, including its historical background, physical characteristics, principal theories concerning its origin and language, scientific investigations conducted to date, and the current state of research surrounding this enduring mystery.

This PDF contains image-only scans of every page of the original Voynich Manuscript. No transcription, translation, or commentary is included 

Source:  Yale University Library 

2.0 Background

The Villa Mondragone in 1620, etching of Matthaeus Greuter 

The Villa Mondragone, depicted here in an early seventeenth-century etching by Matthaeus Greuter, holds a special place in the story of the Voynich Manuscript. Once a grand villa outside Rome, it became a Jesuit college in the late 1500s and later part of the Collegio Romano collection, where many rare works were stored. It was here, in 1912, that book dealer Wilfrid Voynich is said to have acquired the mysterious manuscript from the Jesuit order. The villa’s serene façade hides the beginning of one of the greatest enigmas in the history of written language. 

Source: Wikipedia

2.1 Discovery and Early Ownership

The Voynich Manuscript is named after Wilfrid M. Voynich, a Polish American antiquarian bookseller who purchased it in 1912 from the Jesuit College at Frascati, near Rome (D’Imperio, 1978). Upon acquiring the codex, Voynich suggested that it might be the work of Roger Bacon, a thirteenth-century English friar and polymath, in the hope of increasing its allure (Beinecke Library, 2016). A letter found with the manuscript later revealed that this rumour had originated much earlier. In 1665, Johannes Marcus Marci wrote that Emperor Rudolf II of Bohemia (Holy Roman Emperor from 1576 to 1612) had purchased the book for 600 ducats, believing it to be the work of Roger Bacon (Marci, 1665). Rudolf II was known for his fascination with alchemy and the occult, and the manuscript likely entered his Prague collection in the late sixteenth century (Beinecke Library, 2016).

There is strong evidence that the manuscript was once owned by the English astrologer John Dee, who was active at Rudolf’s court. Dee’s son wrote of a mysterious “book… containing nothing but hieroglyphics” that his father could not decipher, and historical records indicate that Dee had funds of around 630 ducats, which correspond to the supposed sale price, in about 1586 (Goldstone & Goldstone, 2005). Rudolf II apparently passed the manuscript on to his imperial pharmacist, Jacobus Horcicky de Tepenecz, whose faded ownership inscription, “Jacobi de Tepenecz,” was later detected under ultraviolet light on folio 1r (Beinecke Library, 2016).

After Rudolf’s death, the trail continues in Prague. Upon the death of Tepenecz in 1622, the manuscript likely came into the possession of Johannes Marcus Marci, a scientist and rector of Charles University, who was a close friend of Athanasius Kircher, a well-known Jesuit scholar in Rome (Watson, 2020). Marci sent the Voynich Manuscript to Kircher in 1665 or 1666, hoping that he might be able to decipher it, and enclosed a cover letter recounting the manuscript’s history (Marci, 1665). Kircher had earlier received sample pages of the strange text from a Prague alchemist named Georg Baresch, whose 1639 letter represents the earliest known reference to the manuscript (D’Imperio, 1978). Kircher apparently kept the book in the library of the Collegio Romano in Rome, where it later became part of the Jesuit collection.

The manuscript then disappeared from public record for nearly 250 years. It resurfaced only in 1912, when Voynich discovered it among a cache of books being sold by the Jesuits. Voynich spent the following years showing the manuscript to scholars and attempting, unsuccessfully, to interest them in deciphering it (D’Imperio, 1978). After Voynich’s death in 1930, the manuscript was inherited by his widow, Ethel Voynich, and later sold by her estate to the antiquarian book dealer Hans P. Kraus, who eventually donated it to Yale University’s Beinecke Rare Book & Manuscript Library in 1969 (Beinecke Library, 2016). Since then, it has been preserved at Yale as Beinecke MS 408, where scholars can examine its perplexing contents firsthand.

2.2 Physical Description and Dating

Codicological examination shows that the Voynich Manuscript is a parchment codex of approximately 240 pages, including foldouts, although some leaves are missing from its original collation (Currier, 1976). The parchment (vellum) is of average-quality calfskin and has been radiocarbon dated to the early fifteenth century, specifically between 1404 and 1438 with 95 per cent confidence (Stolte, 2011). Four separate vellum samples all yielded consistent dates in the early 1400s, firmly placing the manuscript’s creation in the late medieval period [1][2]. Protein analysis confirmed that the pages are made from at least 14 or 15 calfskins, and there is no evidence that the parchment came from a reused older document (i.e. it is not a palimpsest) [2].

The manuscript’s dimensions are about 23.5 × 16.2 × 5 cm, and its pages are gathered into 18 quires (signatures) with several foldout sheets of varying sizes (Beinecke Library, 2016). Many pages feature large foldouts, including one six-page foldout, indicating that the codex was carefully constructed despite its unusual content. Modern foliation numbers 1–116 (in the top right-hand corner of each recto) were added later for reference, and gaps in the numbering indicate that the manuscript originally contained additional pages now lost (Currier, 1976). Based on quire numbering written in a fifteenth-century hand, the manuscript probably had at least 272 pages in 20 quires originally, meaning about 32 pages are missing (Beinecke Library, 2016).

Extensive forensic analysis of the materials has been conducted. The text and drawings were applied with a quill pen and ink. Analysis by polarised light microscopy and X-ray techniques shows that the ink is iron-gall, typical of medieval manuscripts (Barabe, 2009). The elemental composition of the ink (carbon, iron, sulphur, potassium, calcium, with trace copper and zinc) is consistent with historic iron-gall formulations and shows no detectable modern pigments such as lead-based inks (Barabe, 2009).

The paints used to colour the illustrations were also examined using X-ray diffraction (XRD) and scanning electron microscopy (SEM). They proved to be ordinary minerals available in the fifteenth century: the blue pigment is ground azurite (basic copper carbonate) with traces of cuprite; the green pigment appears to be a copper-based resin, perhaps copper resinate, with compounds such as atacamite; the red-brown pigment is red ochre (iron-oxide hematite) with components of iron sulphide; and the white pigment is likely calcium carbonate (chalk) mixed with egg white (tempera) [3][4]. These findings confirm that nothing in the materials is anachronistic. The Voynich Manuscript is genuine to the medieval period and not a modern forgery (Stolte, 2011).

Even the binding and covers, while not original, are consistent with its known history: the manuscript was rebound in the seventeenth century in limp vellum covers at the Collegio Romano, replacing an earlier wooden cover, likely removed when it was transferred or sold [5]. Insect holes on the first and last pages suggest that the manuscript once had a wooden cover, now lost, which was worm-eaten, and a later tanned-leather inner cover left discolouration on the edges of some leaves [6].

In summary, all physical evidence from parchment dating to ink and pigment composition indicates that the Voynich Manuscript was created in the early 1400s in a European milieu, probably in Central Europe and possibly in northern Italy based on stylistic hints (Fagin Davis, 2019). This timeframe rules out earlier attributions such as Roger Bacon (1214–1292) and refutes theories of a modern hoax. The stage is thus set for the real mystery: who wrote this manuscript, in what language or code, and for what purpose?

2.3 Studies by Cryptographers and Scholars

Since its modern rediscovery, the Voynich Manuscript has attracted the attention of leading cryptographers and scholars, often to their frustration. In the 1920s, William Romaine Newbold claimed to have deciphered the manuscript using an alleged microscopic anagram method, asserting that it contained secret scientific writings by Roger Bacon (Newbold, 1928). However, Newbold’s elaborate solution (involving tiny hidden Latin letters within characters) was later debunked and is now considered a classic case of over-interpretation (Manly, 1931).

During the Second World War and the Cold War, notable American and British codebreakers including William F. Friedman (who broke the Japanese PURPLE cipher) and Brigadier John Tiltman attempted to crack Voynichese, treating it as an enciphered message. They failed to find a key, and Friedman concluded that no clear solution was in sight, reportedly even forming a “Voynich Manuscript Study Group” in the 1940s that produced the first complete hand transcription of the text onto punch cards (D’Imperio, 1978).

In 1976, Prescott H. Currier presented evidence that the text is not uniform throughout: he identified at least two distinct “dialects” or scribal varieties in the manuscript (labelled Currier Language A and B) as well as multiple hands (scribes) responsible for different sections (Currier, 1976). Currier’s analysis of letter patterns suggested that the differences were significant enough that parts of the manuscript might have been encoded or composed at different times or by different people, although the underlying script is the same.

This finding of multiple textual “languages” or styles was later reinforced by more recent digital palaeography. Using detailed handwriting analysis, Lisa Fagin Davis (2020) identified five distinct scribal hands in the Voynich Manuscript, each with unique glyph forms and spelling patterns (Fagin Davis, 2020). If five scribes collaborated or worked sequentially on the book, as Davis argues, it implies that the manuscript may have been a communal project rather than the work of a lone eccentric or hoaxer (Sabar, 2024).

Throughout the twentieth century, numerous other scholars proposed hypotheses that it is a cipher text, a constructed language, or even meaningless gibberish but none could conclusively prove their case (D’Imperio, 1978). In 1962, for example, the cryptologist Elizebeth Friedman famously stated that after years of study, “the text just doesn’t act like a natural language” and defies ordinary decipherment techniques (Timm & Schinner, 2020).

By the turn of the twenty-first century, the Voynich Manuscript had become a cause célèbre in the world of mysterious codes, often compared to an “uncrackable” cipher. Yet, with the advent of high-quality digital scans (Yale University put the entire manuscript online in 2020) and modern computational tools, a new generation of researchers has revisited the challenge with fresh approaches. As we move into the study of its content and theories, it is important to remember that while much has been learned about the Voynich Manuscript’s physical nature and history, its text remains unreadable a puzzle in plain sight.

3.0 Content of the Manuscript

The Voynich Manuscript is divided into several distinct sections, each featuring its own thematic focus and style of illustration yet unified by the same mysterious script. Scholars generally categorise the contents into five or six principal parts: the herbal, astronomical (or astrological), balneological (or biological), cosmological, pharmaceutical, and textual sections. Each displays meticulous drawings accompanied by lines of the undeciphered Voynichese script, suggesting a deliberate and organised purpose rather than random composition. The herbal section contains drawings of plants some resembling known species, others wholly imaginary paired with text that might describe their properties or uses. The astronomical pages depict zodiac symbols and star charts, possibly indicating an interest in astrology or celestial influence on medicine. The biological section features nude female figures immersed in pools connected by tubes, which some interpret as allegories of bodily or alchemical processes. Together, these sections present the manuscript as a kind of encyclopaedia of natural philosophy or healing, rendered through an unknown language and worldview. Its contents are both systematic and surreal, straddling the line between science and symbolism, making the Voynich Manuscript not merely a text but an intricate fusion of image, mystery, and meaning.

3.1 Herbal Section (Botanical)

Source:  Yale University Library 

This is the largest section, comprising about 112 folios filled with drawings of plants usually one large plant per page, with a few paragraphs of text around or below each (Beinecke Library, 2016). There are 113 plant illustrations in total, depicting roots, stems, leaves, and flowers in a schematic style. Although reminiscent of medieval herbal manuscripts that catalogued medicinal plants, virtually none of the Voynich plants can be identified with certainty as real species (Kennedy & Churchill, 2006). Many show composite or fantastical features; for example, the leaves of one plant may be attached to the root of another, topped with an incongruous blossom [7]. A few tentative identifications have been proposed: the drawing on f93r has been interpreted by some as a sunflower (Helianthus), a New World plant though this is controversial, and others suggest it could depict a European species such as a marigold (Tiltman, 1967). Another page appears to show a wild pansy and a maidenhair fern, which, if correct, would place the manuscript’s botanical knowledge in a European context (Kennedy & Churchill, 2006). Overall, the herbal images appear fantastical yet consistent in style, as if copied from an unknown herbal source or imagined by the author. The accompanying text might describe the plants’ properties or uses, but without decipherment this remains speculative. Notably, many plant drawings include small labels or pointers to parts, implying descriptive or explanatory intent (Montemurro & Zanette, 2013).

3.2 Astronomical & Astrological Section

Source:  Yale University Library 

 Roughly 20 folios contain astronomical diagrams and zodiac motifs. These pages show astral charts with concentric circles, suns, moons, and stars. Most notably, there are twelve diagrams that appear to represent zodiac signs for example, fish for Pisces, a bull for Taurus, and a centaur archer for Sagittarius each accompanied by rings of text and human figures (Beinecke Library, 2016). Around the zodiac symbols, small nude figures, mostly female, hold stars or stand within segmented circles, perhaps symbolising constellations or calendar personifications. Ten of these zodiac diagrams include month names (March through December) written in Latin script near the figures; the spellings suggest dialects of northern Italy or the Iberian Peninsula (for instance, “Mai” for May, “Iul” for July), supporting a European origin for this section (Tiltman, 1967). The inclusion of these labels in a known alphabet is revealing: the creators were familiar with Latin script yet chose to write the main text in their unique cipher. Beyond the zodiac wheels, some foldouts depict what may be astronomical calendars or volvelles rotating diagrams with radiating spokes and celestial imagery. The presence of astrology is unsurprising in a medieval context: astrological medicine, which timed herbal treatments to planetary positions, was common in the 15th century. Thus, the fusion of herbs and astrology in the Voynich Manuscript fits a medical compendium of the era (Sherwood, 2020). Still, the meaning of the charts remains uncertain. One diagram seems to show the classical planets or a model of the heavens, though the text cannot confirm this (Kennedy & Churchill, 2006).

3.3 Biological (Balneological) Section

Source:  Yale University Library 

 Spanning about 20 folios, this unusual section depicts dozens of small nude female figures, many with distended abdomens, immersed in or interacting with networks of tubes and pools of liquid. These “balneological” or “biological” drawings show women bathing in green or blue water flowing through complex pipe-like systems (Beinecke Library, 2016). Some figures wear crowns, while others are linked by looping cords and pipes resembling anatomical organs or plumbing. Interpretations vary: the scenes may symbolise alchemical processes, with the women personifying substances moving through retorts, or represent human biology, perhaps gestation or bodily cycles (Coyne, 2011). In one bifolio (folios 78v–81r), water flows continuously across pages through a series of baths [8], suggesting narrative progression. Several figures’ swollen abdomens have been interpreted as signs of pregnancy, linking the imagery to medieval fertility or childbirth treatments (Pelling, 2006). The term balneological refers to baths, and medieval health texts often discussed therapeutic bathing. Yet nothing quite like these drawings appears elsewhere, making this section uniquely enigmatic. The surrounding text may describe regimens of baths or metaphors of life, but its meaning remains hidden. Some researchers have speculated that the imagery encodes the teachings of a female monastic or secret community, while others view it as a work of imaginative symbolism (Beinecke Library, 2016).

3.4 Cosmological Section

Source:  Yale University Library 

 This portion comprises about 14 pages and includes elaborate foldout diagrams of apparent cosmological or cartographic design. The most famous is the “Rosettes foldout”: a large six-page spread (folios 85–86–87) featuring nine interconnected circular diagrams or “islands” [9]. Each rosette contains intricate imagery walls, castles, stars, and structures that might depict volcanoes or fantastical landscapes. The circles are linked by causeway-like paths, implying a journey or relationship between them. Some interpret this as a symbolic or spiritual map; others see it as a model of the cosmos, with nine spheres representing celestial realms (Kennedy & Churchill, 2006). Tiny details such as towers, bridges, or flame-like shapes invite endless debate. Other foldouts in this section display radial designs with text arranged in rings that do not match known astronomical models, leading to theories that they may be mystical, alchemical, or mnemonic charts encoding hidden information. The cosmological section is often regarded as the most perplexing in the manuscript because it lacks direct parallels in contemporary works (Beinecke Library, 2016). Nevertheless, it reinforces that the Voynich Manuscript is not random but methodically organised, resembling an encyclopaedia of arcane knowledge spanning nature, humanity, and the heavens.

3.5 Pharmaceutical Section

Source:  Yale University Library 

 About 16 pages form what researchers call the pharmaceutical section. These show small drawings of plant parts roots, leaves, and flowers rather than whole plants, often accompanied by sketches of apothecary jars or containers (Beinecke Library, 2016). The jars vary in shape and colour some simple, others ornate and are filled with solid hues of blue, red, or green. Each page typically combines several plant parts with nearby labels and short paragraphs of text, perhaps describing recipes or preparations. This structure closely resembles medieval pharmacopoeias, which listed ingredients and illustrated storage vessels. Many of the plant parts correspond to those in the herbal section but appear disassembled, suggesting that the herbal section presents idealised composites while the pharmaceutical section records usable components (Tiltman, 1967). The section likely served as a formulary or medicinal reference, though without reading the script this remains conjecture. Its consistent motifs of jars and labelled parts mirror known medical manuscripts, reinforcing the impression that it records recipes or remedies.

3.6 Recipes (Stars) Section

Source:  Yale University Library 

 The final section contains 20 to 30 pages of text arranged in short paragraphs, each introduced by a star- or flower-shaped bullet in the margin. These pages have no large illustrations only the star symbols and text hence they are commonly called the “recipes” section (Beinecke Library, 2016). Each paragraph begins with a star glyph, resembling a modern bullet point, suggesting a list of discrete items or instructions. If the manuscript is a compendium of knowledge, this final portion could present concluding formulas or prescriptions related to earlier herbal and astrological material. The organisation implies deliberate structure and practicality, not random notes. Researchers have attempted to correlate recurring words in this section with plant names from earlier parts, but no definitive matches have been found (Reddy & Knight, 2011). Overall, the section gives the impression of a reference list or manual of procedures. Its format, combining herbal, astronomical, and medicinal elements, parallels medieval “books of secrets” or health handbooks where astrology and medicine intersected (Manly, 1921). The Voynich Manuscript thus appears to have the structure of a pharmacopoeia or medical compendium an encyclopaedic guide to natural and celestial influences on health. Yet because the text remains undeciphered, every interpretation is tentative, and the extraordinary nature of the illustrations continues to inspire speculation from alchemical cipher notebooks to secret religious treatises [10][11].

3.7 The Mysterious Text and Script

The most confounding feature of the Voynich Manuscript is its text: approximately 38,000 words of flowing, graceful, yet entirely unknown characters. The script does not occur in any other known document; it appears to be an invented alphabet consisting of about 20–30 basic characters, along with several additional, less frequent symbols (Currier, 1976). The writing runs from left to right in consistent horizontal lines. Words are clearly separated by spaces, but there are no obvious punctuation marks in the main text, except for the occasional dot or star, usually in the recipe’s section [12][13]. Each word typically ranges from two to ten characters in length, with an average of about five to six characters. Very few words consist of a single character or exceed ten letters [14].

Intriguingly, the words follow certain phonetic or orthographic patterns. Some characters never occur at the beginning of a word, others never at the end, and some are almost always doubled or tripled when they appear (Currier, 1976). For example, certain glyphs function in ways like vowels or consonant clusters in natural languages. The so-called “gallows” characters often appear at the start of words, while other symbols occur chiefly as suffixes. This has been compared to a kind of phonotactic structure the rules are reminiscent of a language’s spelling conventions (for example, in English, “q” is almost always followed by “u”, or in Italian, words rarely end in “j”), yet Voynichese follows its own distinct logic (Landini, 2001). Many researchers, beginning with cryptologist John Tiltman in the 1950s and later statistical analysts, have commented on the highly regular internal structure of Voynich words (D’Imperio, 1978). It is as if the text were generated according to a precise set of linguistic rules, which is one reason why some believe it could represent an actual language perhaps written in cipher or as a constructed language rather than meaningless invention.

One peculiar feature noted is the script’s smooth ductus (the manner of writing). The scribe or scribes wrote the characters fluidly, with no visible hesitation marks or corrections, except where someone later retraced or darkened the ink [15]. This suggests that whoever wrote it was familiar and comfortable with the script; it was not enciphered letter by letter in a slow, deliberate way, which often produces irregularities. Instead, the writing appears natural, as if it were produced in a native language or at least a well-practised code (Montemurro & Zanette, 2013). This observation is both fascinating and troubling, as it implies the existence of a wider corpus or tradition behind this script something of which no other record survives.

A few parts of the manuscript contain incidental writing in ordinary scripts. On folio 1r (the first page), faint Latin letters appear in the margin, perhaps a brief annotation by a later owner, and the signature of “Jacobj à Tepenecz” (Rudolf II’s librarian) was found almost erased at the bottom (Tiltman, 1967). In the astrological section, as mentioned earlier, the names of the months (March through December) are written in Latin script around the zodiac diagrams [16]. Additionally, folio 66r contains a small fragment of High German text (“der Musdel” or “der Mussteil”, likely meaning “the widow’s share”) beside a drawing of a nude man [17]. These isolated examples confirm that the manuscript passed through the hands of readers who spoke European languages, but they offer no direct insight into the meaning of the main text.

Because the text cannot yet be translated, researchers have focused on quantitative analyses of the Voynich script to extract clues. One of the most fundamental approaches examines word-frequency distribution. When the frequency of each word is counted, the Voynich Manuscript displays a roughly Zipfian pattern: a few words occur very often, while most are rare, similar to natural languages (Sanchez et al., 2016). The most frequent Voynich word appears about twice as often as the second most frequent, three times as often as the third, and so on, consistent with Zipf’s law (Bowern, 2020). The manuscript’s total vocabulary its number of unique word types is estimated at around 8,000, out of approximately 35,000 total word tokens [18]. This ratio falls comfortably within the normal range for genuine language documents of comparable length: not too low, which might suggest a simplistic cipher or repetitive nonsense, nor too high, which would indicate a random or unstructured text. Indeed, one study found that the Voynich text’s lexical richness and entropy resemble those of Latin or English, rather than random sequences or trivial ciphers (Montemurro & Zanette, 2013). Furthermore, certain Voynichese words tend to appear together or in specific orders, hinting at an underlying grammar. For instance, some word pairs, or bigrams, occur more frequently than chance would predict, suggesting that one may modify or follow another in a structured way (Timm & Schinner, 2020). This implies that some words could serve grammatical functions, such as adjectives or compound components.

However, there are also features that appear distinctly “unlanguage-like”. One of the most striking is the high level of repetitive sequences. It is not uncommon to find the same word written two or even three times in succession—for example, sequences transliterated as “qokeedy qokeedy qokeedy”, which would be equivalent to writing “apple apple apple” in English [19]. In natural languages, such triple repetition is extremely rare, except in lists or deliberate emphasis. In addition, many Voynich words differ from one another by only a single letter within a short passage, more often than chance would allow. For example, one might see “qokeedy qokeedy qokeedy qokeedy” followed by “qokeedy qokeey qokeeey”, a pattern that simple substitution ciphers do not normally produce (Stolfi, 2004). This has led to speculation about a generative process—perhaps the text was created by applying a systematic procedure or algorithm to a meaningful base text, producing overlapping or expanded forms (Rugg, 2004).

Another unusual characteristic is the distribution of word lengths, which is much narrower than in ordinary languages and approximates a binomial curve (Hermes et al., 2022). Most words are between four and six characters long, with very few shorter or longer forms. This differs markedly from many natural languages; for instance, English has numerous one- and two-letter words such as “a”, “I”, and “of”, as well as occasional long compounds. In Voynichese, almost no words are that short, and none clearly corresponds to common function words like “and” or “the”, which recur consistently across all contexts (Timm & Schinner, 2020).

In genuine languages, high-frequency function words usually occur evenly throughout a text. In the Voynich Manuscript, however, the most frequent words tend to cluster within specific sections some appear predominantly in the herbal pages, others in the astronomical ones [20][21]. This is atypical if those words served the same connective roles as “and” or “the”, since such words normally show uniform distribution. This peculiar behaviour suggests that Voynichese “words” may not correspond to words in the usual linguistic sense. Instead, they could encode syllables, numerical sequences, abbreviations, or be artefacts of an encryption system.

In summary, the Voynich text displays a blend of linguistic and non-linguistic traits. It is sufficiently structured and non-random to strongly suggest that it encodes meaningful content of some kind (Montemurro & Zanette, 2013). The presence of distinct patterns and constraints in word formation indicates an underlying linguistic or rule-based system (Landini, 2001). Yet the peculiarities such as heavy repetition, consistent word lengths, and unusual positional restrictions on certain letters are difficult to reconcile with any known language family (Timm & Schinner, 2020).

Some scholars have noted similarities between the statistical patterns of Voynichese and certain Asian languages when romanised. For instance, one study found that the Voynich text shares more characteristics, in terms of alternating patterns and entropy, with Mandarin Chinese written in pinyin than with European alphabets. This may be because Chinese syllables have a constrained structure somewhat comparable to that of Voynichese words (Vonfelt, 2013). Such findings have fuelled speculation that the manuscript might represent an Asian language transliterated into a cipher alphabet (Montemurro & Zanette, 2013). Likewise, the 2011 computational analysis by Reddy and Knight suggested that, if the manuscript encodes a natural language, it is unlikely to belong to the Indo-European family. They found that its word-length and spelling distributions align more closely with certain East Asian languages than with European ones (Reddy & Knight, 2011).

On the other hand, the manuscript could be a highly sophisticated cipher or a deliberately constructed language, which by design would not conform to typical linguistic profiles. Its enduring resistance to decipherment, despite an evidently systematic structure, is precisely what makes it so captivating. Whether it represents an unknown language, a cipher, or an elaborate hoax, the script of the Voynich Manuscript remains its most mystifying aspect and the ultimate key to unlocking its secrets, if that key can ever be found.

4.0 Theories

Over the years, a plethora of theories have been proposed to explain the Voynich Manuscript’s content and purpose. Broadly, these theories can be grouped into a few categories: (a) it is a plaintext of an unknown or coded natural language, (b) it is an invented script encoding a constructed or artificial language, (c) it is a cipher (encrypted text) hiding a meaningful message, or (d) it is a hoax or meaningless fabrication. No theory has been conclusively proven, but each has its advocates and points of evidence. Below we survey the main theories and notable attempts at decipherment or explanation, along with their strengths and weaknesses.

4.1 Cipher of a Natural Language

Comparison of Voynichese (left, VMS 29v) with gibberish from one of our participants(right), showing potentially analogous selection of word lengths to wrap the text around the illustration. 

Source: Cheshire, G. (2019) 

One long-standing hypothesis is that the Voynich Manuscript is written in a natural human language that has been enciphered or transcribed using a unique alphabet. In other words, the underlying text could be a readable language if converted appropriately. Early on, Wilfrid Voynich himself leaned toward this idea, suspecting that the manuscript’s unusual characters might conceal medieval Latin or another European language familiar to Roger Bacon (Brumbaugh, 1975). Many professional cryptologists initially approached Voynichese as a cipher. If it were a classical substitution cipher, with each Voynich character representing an ordinary letter, one would attempt to map these symbols to letters of known languages such as Latin or Italian and test whether meaningful text emerged.

William F. Friedman and his team conducted extensive trials using this approach, testing simple substitution, polyalphabetic ciphers, and even more exotic ideas such as synthetic “universal” languages, but without success (D’Imperio, 1978). Simple substitution has since been effectively ruled out. Analyses of letter frequencies and positional patterns show that no direct mapping between Voynichese symbols and the alphabets of Latin or other European languages yields coherent text (Tiltman, 1967). If the manuscript is a cipher of a natural language, it must therefore employ a more complex system.

Some have proposed that the underlying language might not be European. Linguist Jacques Guy and others in the 1970s and 1980s suggested that it could represent an Asian language perhaps a Chinese dialect or Vietnamese encoded alphabetically (Guy, 1991). This theory was based on certain statistical properties, such as low entropy and consistent word structures, which resembled those of tonal Asian languages when written phonetically. In 2014, British linguist Stephen Bax announced that he had “decoded” around ten words of the Voynich Manuscript by identifying proper names of plants and stars, proposing that the text could be written in a Near Eastern or Asian language using a borrowed script (Bax, 2014). Bax claimed to have recognised words corresponding to “Taurus”, “Coriander”, “Juniper”, and others by associating repeated labels with the accompanying illustrations, in a method reminiscent of how Egyptian hieroglyphs were deciphered using proper names. However, his methodology was widely criticised as speculative, and his proposed partial decipherment failed to gain support among Voynich researchers (Schneider, 2014).

Nevertheless, Bax’s work kept alive the idea that Voynichese might represent a real, albeit disguised, language. If that is true, potential candidates range from a medieval dialect of an existing language such as a rare Romance dialect, Hebrew, or Arabic—to an entirely lost tongue. Historian Nick Pelling has proposed that it might be written in a form of medieval phonetic Latin abbreviations, a kind of shorthand, enciphered by a simple system. He further speculated that the author could have been the Italian engineer Antonio Averlino (also known as Filarete) (Pelling, 2006). Pelling’s hypothesis is detailed and imaginative, but it remains unproven, like all others.

A particularly bold claim was made in 2019 by Gerard Cheshire, who published a paper asserting that the Voynich Manuscript is written in a “proto-Romance” language an alleged ancestral form of the Romance languages using an idiosyncratic script of abbreviated Latin characters (Cheshire, 2019). Cheshire claimed to have translated portions of the manuscript as Vulgar Latin mixed with Occitan and other Mediterranean dialects, essentially proposing the rediscovery of a lost linguistic branch. His paper was met with immediate scepticism and largely discredited by linguists, who pointed out serious methodological flaws and the fact that “proto-Romance” is not a recognised linguistic category (Hauer & Kondrak, 2019). The journal that published the paper was also criticised for inadequate peer review. This episode served as a reminder that while the Voynich Manuscript invites bold speculation, extraordinary claims demand equally extraordinary evidence something Cheshire’s theory lacked.

The notion that the manuscript represents a cipher of a real language remains appealing because it aligns with how many secret texts of the past were created. Renaissance Europe, particularly during the reign of Rudolf II, was rich in cryptographic and alchemical experimentation. A known contemporary parallel is the Rohonc Codex of Hungary, another mysterious document that was likely an enciphered text. It would not be implausible for a medieval scholar to have encrypted a treatise on herbal medicine, alchemy, or astrology, either to protect intellectual property or to obscure esoteric knowledge (Brumbaugh, 1975). If this is the case, the Voynich Manuscript might yet be decipherable with the correct key.

Researchers have proposed several possible cipher mechanisms. Some suggest the use of homophones multiple symbols for a single plaintext letter to disguise frequency patterns, or polyphonic ciphers, in which a symbol’s meaning changes depending on context (Knight & Megyesi, 2012). In 2017, a research team from the University of Alberta applied artificial intelligence algorithms to identify what natural language might underlie the text if it were a substitution cipher. Their program tentatively suggested that the base language was Hebrew written as anagrammed text (Hauer & Kondrak, 2019). They even attempted to rearrange one opening sentence as Hebrew, but the result proved unconvincing to experts, producing nonsensical output. This demonstrates that while advanced computational methods can explore vast possibilities, they too have failed to uncover a definitive key. The evidence suggests that if the Voynich Manuscript encodes a real language, its cipher is far more complex than any simple substitution system.

4.2 Constructed Language or Artificial Script

The above script is known as Hildegard of Bingen’s Lingua Ignota (12th century, Germany). It is one of the earliest known examples of a constructed language or artificial script, much like the Voynich Manuscript. Hildegard created this language for spiritual expression, using invented words and symbols entirely unique to her writings. Similar examples of personal or secret scripts include Johannes Trithemius’s Steganographia (late 15th century), the Theban Alphabet used by Renaissance occultists, and Ramon Llull’s Ars Magna, a symbolic system for representing knowledge. These historical precedents show that inventing private languages or mystical alphabets was not uncommon in medieval Europe. This is why many scholars consider the Voynich Manuscript part of that same intellectual tradition — a self-contained, artificial language system designed for personal, spiritual, or esoteric purposes rather than for ordinary communication. 

Source: Wikipedia

Another theory holds that the Voynich Manuscript might contain a constructed language, an artificial language created by its author. In this view, the text is meaningful but not written in any natural language. Instead, both the script and the linguistic system were invented, perhaps to encode or conceal knowledge. Constructed languages, particularly secret or mystical ones, were not unknown in the medieval world. For example, in the twelfth century, the German abbess Hildegard of Bingen devised a sacred tongue called Lingua Ignota, complete with an invented alphabet, which she used to name and classify spiritual concepts. Similarly, magical alphabets such as the Theban alphabet and other esoteric scripts were used by occultists to disguise ordinary texts. If the Voynich Manuscript falls into this category, it may have been created by an individual or small group who designed their own linguistic system for intellectual, spiritual, or protective reasons.

A related sub-hypothesis proposes that the text could represent a form of enciphered shorthand or an advanced system of abbreviations, similar to those used in medieval Latin manuscripts. Medieval scribes often employed highly standardised abbreviation systems, using symbols to represent common words, prefixes, or endings. The Voynich script could conceivably be a creative extension of such shorthand: several of its glyphs faintly resemble Latin cursive or shorthand signs, though not consistently (Tiltman, 1967). This might explain why simple letter-for-letter substitution has failed; perhaps each Voynich character represents a syllable or even a concept, rather than a single letter. For instance, the frequent word “daiin” could, in theory, stand for a common Latin word or phrase in abbreviated form. If this interpretation were correct, decipherment would require expanding the text rather than substituting symbols, a far more complex problem.

The idea of a constructed language gained traction when computer-based analyses revealed that Voynichese has two distinct statistical modes, corresponding to what are known as Currier A and B (Currier, 1976). It is as though two dialects or idiolects of the same invented language were used. One speculation is that the manuscript served as a manual for teaching or recording this secret tongue, perhaps produced by a small, closed community. Lisa Fagin Davis (2020) suggested that, if multiple scribes indeed contributed to the manuscript, it might have originated within a sect or fraternity such as a group of alchemists or herbal healers who shared this invented system. They could have used it to preserve knowledge they did not wish outsiders to access. The consistency of the writing system across different scribes implies that they all understood the underlying structure. This theory neatly accounts for the combination of linguistic regularity, suggesting a genuine language or code, and uniqueness, as there is no trace of it elsewhere.

Another variant proposes that the text is not a full constructed language but rather a mnemonic or steganographic code, a system that encodes information indirectly. Some researchers have suggested that the manuscript could encode numbers that correspond to an almanac, calendar, or table of data, functioning as a kind of mnemonic cipher (Rugg, 2004). Alternatively, each word might represent an entry in a glossary or a reference within another document, like a codebook cipher. In such a case, the Voynich Manuscript would not be meant to be read in the traditional sense but interpreted through a secondary key or index. This would make it essentially a cipher ledger of knowledge, where each symbol represents a concept rather than a letter or word. Decoding such a system without the original key would be nearly impossible.

A notable exploration of this constructed language direction came from Prescott Currier, who first identified the two distinct linguistic systems, A and B. He speculated that the text might be synthetic, generated by a private method known only to the author, and even raised the possibility of glossolalia or speaking in tongues (D’Imperio, 1978). Glossolalia refers to speech or writing that imitates linguistic structure without conveying consistent meaning, often produced in states of religious ecstasy or trance. Currier did not fully endorse this explanation but noted its relevance given the text’s unusual qualities. Most scholars, however, find glossolalia unlikely in this case. The Voynich text is too extensive and internally consistent. Genuine glossolalic writing tends to produce irregular, fragmented strings of syllables, not tens of thousands of systematically arranged words (Williams, 2019). The Voynich Manuscript, by contrast, shows a degree of order and repetition far exceeding what could be expected from spontaneous or ecstatic writing.

4.3 Medieval Scientific or Medical Compendium

This illustration, showing a group of women bathing in interconnected pools, is one of the most distinctive images in the Voynich Manuscript. Scholars often interpret these scenes as symbolic representations of healing or medical practices, possibly related to balneology — the medieval study of therapeutic bathing and mineral springs. During the 15th century, bathing was associated with the treatment of ailments and the balance of humours in the body. The presence of pipes and flowing water in these drawings suggests a focus on circulation, purification, or the medicinal use of baths, reflecting the manuscript’s broader interest in health and natural science. 

Source:  Yale University Library 

Another theory proposes that the Voynich Manuscript is an encrypted scientific or medical compendium from the Middle Ages. This idea is not mutually exclusive with either the cipher or constructed-language hypotheses but provides a potential context: the book could be a practical reference on herbology, astrology, and medicine, created by knowledgeable individuals who wrote in code to keep their work secret or limited to a select group (Tiltman, 1967). The presence of drawings of medicinal herbs, astrological diagrams possibly related to medical astrology, and anatomical-like figures suggests that the manuscript might have served as a manual for healing practices, perhaps within an unconventional or esoteric tradition, which could explain its secrecy. The Béguines, a semi-cloistered order of female healers in northern Europe, have been mentioned in popular accounts (Sabar, 2024) because of the many female figures in the manuscript and the potential link to folk medicine. Although there is no direct evidence for this connection, it illustrates one possible narrative: that a community of learned women or alchemists compiled their herbal and astrological knowledge in a private script known only to them.

One of the earliest rational interpretations, aside from the Roger Bacon attribution, came from the art historian Erwin Panofsky. In the 1930s, Panofsky suggested that the manuscript dated from around 1470 to 1500 and likely originated in northern Italy. He believed it might have emerged from a heterodox or heretical milieu, containing secret religious or philosophical material disguised as a medical text (Panofsky, 1954). Panofsky observed that certain costume elements, such as the crown styles of the female figures, resembled northern Italian Renaissance art, and that the castles in the rosette diagram looked like fifteenth-century European fortifications. If the manuscript was indeed produced in the late fifteenth century, as Panofsky proposed, or in the early fifteenth, as radiocarbon dating indicates, it would fit into the intellectual environment of early humanism, when ciphers and secret writing were becoming fashionable. Leonardo da Vinci, for example, wrote his private notes in mirrored script, and Johannes Trithemius composed Steganographia, a treatise on hidden messages. Within such a context, a ciphered manuscript on plants and stars would not have been out of place.

The medical-compendium theory briefly gained renewed attention in 2017, when historian Nicholas Gibbs claimed in The Times Literary Supplement that he had solved the Voynich puzzle. He argued that it was a fifteenth-century women’s health manual written in Latin shorthand abbreviations (Gibbs, 2017). Gibbs suggested that, when expanded, the text revealed passages plagiarised from known medieval Latin works on herbal medicine and bathing, particularly those attributed to Trotula of Salerno. He presented short translations that appeared to describe ointments and medicinal preparations. However, his claims were quickly debunked by scholars, who noted that he had cherry-picked Latin words that seemed to fit while ignoring inconsistencies. His abbreviation key could not be applied uniformly across the text (Ranford, 2017). Latin specialists also pointed out that, had the manuscript been written in familiar Latin shorthand, it would have been recognised long ago by the many medievalists who had examined it. Gibbs’s proposal was therefore dismissed, though it was notable in that it attempted to situate the manuscript within the well-known genre of medieval medical literature.

A related possibility is that the Voynich Manuscript was produced at the court of Emperor Rudolf II, where it is known to have resurfaced in the seventeenth century. Rudolf’s Prague court was home to numerous alchemists, pharmacists, and natural philosophers from across Europe. It is conceivable that one of them compiled an “encyclopaedia of nature” written in secret code. The manuscript’s diverse contents botanical, astronomical, biological, and pharmacological recall the encyclopaedic “Books of Secrets” popular in the Renaissance. It might even have been created to appeal to Rudolf’s fascination with the occult and the marvellous. In that case, the author could have been one of the learned figures associated with his court, such as Jacobus de Tepenecz (Rudolf’s pharmacist, whose name appears in the manuscript), Johannes Marcus Marci (who later sent it to Kircher), or even Athanasius Kircher himself. None of these identifications is supported by evidence, but together they contribute to the intricate web of theories surrounding the manuscript’s origin. 

4.4 Hoax or Fabrication

Portrait of Emperor Rudolf II (engraving, late 16th century)

This engraving depicts Emperor Rudolf II of the Holy Roman Empire, one of Europe’s most enigmatic rulers and a famed patron of alchemists, astronomers, and philosophers. According to a 1665 letter from Johannes Marcus Marci to Athanasius Kircher, Rudolf reportedly purchased the mysterious Voynich Manuscript for 600 ducats, believing it to be the work of the English scholar Roger Bacon. This connection places Rudolf at the heart of the hoax or fabrication theory, which suggests the manuscript may have been crafted to exploit his fascination with occult knowledge and secret sciences. Whether or not Rudolf was deceived, his passion for hidden wisdom ensured that the Voynich Manuscript would be preserved within the strange and brilliant world of his Prague court — a place where art, alchemy, and mystery intertwined. 

Source: Czechcenter

In contrast to theories proposing genuine meaning or language, another long-standing hypothesis holds that the Voynich Manuscript is an elaborate hoax or a meaningless text created to imitate a real one. This idea dates back at least to the 1920s, when William Newbold’s alleged decipherment was discredited and sceptics such as John Manly suggested that perhaps there had never been any meaning to find (Manly, 1931). The hoax theory exists in several variations.

One version proposes that the manuscript was created in the fifteenth or sixteenth century as a fraud designed to deceive a wealthy collector, possibly Emperor Rudolf II, who was known to pay large sums for unusual books and artefacts. According to a letter by Johannes Marcus Marci, Rudolf reportedly paid 600 ducats for the manuscript (Marci, 1665). An enterprising alchemist or con artist could have fabricated a book filled with convincing illustrations and pseudo-script, passing it off as an ancient tome of secret wisdom. If this were the case, the Voynich Manuscript would represent a kind of medieval deception a beautifully crafted, meaningless work intended to part a curious patron from his money.

The English alchemist Edward Kelley, who travelled with John Dee in Central Europe during the 1580s, is often mentioned in this context. Kelley was notorious for his claims of alchemical power and his alleged frauds, including the promise of transmuting base metals into gold. Some have speculated that Dee or Kelley may have produced the Voynich Manuscript to sell to Rudolf, whom Dee did indeed meet. Dee’s diary records selling a mysterious book to Rudolf for a substantial sum, and he had a known fascination with Roger Bacon’s works, which might explain the manuscript’s supposed Baconian attribution (Goldstone & Goldstone, 2005). However, radiocarbon dating places the parchment between 1404 and 1438, predating Dee and Kelley by more than a century. The only way this theory could still hold would be if they had obtained blank medieval vellum and written on it, a possibility that remains speculative.

Another variant of the hoax theory claims that the manuscript could be a modern fabrication by Wilfrid Voynich himself. Voynich, as a rare book dealer, had an obvious financial interest in promoting the manuscript’s mystery and value. However, this idea was largely dismissed after the material analysis proved that both the parchment and inks were genuinely medieval (Stolte, 2011). For Voynich to have forged the manuscript, he would have needed to obtain a large quantity of unused fifteenth-century vellum, reproduce period-accurate inks and pigments, and produce over 200 pages of intricate writing and drawings without modern detection a near-impossible feat. Furthermore, experts who examined the book during Voynich’s lifetime found no indication of forgery. As a result, a twentieth-century hoax is now considered extremely unlikely.

A fifteenth-century hoax, however, remains plausible. The manuscript could be an early example of a deliberately artificial language or an imitation of scholarly work designed to appear mysterious. Dr Gordon Rugg of Keele University revitalised this perspective in 2004 by demonstrating that a simple medieval tool, the Cardan grille, could be used to generate text with statistical properties closely resembling Voynichese (Rugg, 2004). By using a table of syllables and a movable grid, Rugg was able to produce sequences of text that matched the word frequencies and letter patterns found in the manuscript. In 2016, Rugg and Gavin Taylor expanded this model, showing how such a process could reproduce the manuscript’s entropy and structural consistency (Rugg & Taylor, 2017). They argued that a skilled fifteenth- or sixteenth-century scholar could have generated the text systematically, creating a convincing imitation of language that contained no actual meaning. Possible authors suggested within this framework include Johannes Trithemius, known for his cryptographic writings, or Raphael Mnishovsky, a seventeenth-century cryptographer and acquaintance of Marci, who once claimed to know the manuscript’s secret but refused to reveal it (Timm & Schinner, 2020). A monk, scholar, or alchemist with an interest in puzzles or deception could conceivably have produced such an artefact.

Critics of the hoax hypothesis note that generating nearly 40,000 words of consistent pseudo-text by hand would have been a monumental task. Producing such sustained regularity without apparent error or fatigue would require significant discipline and intelligence, making the result almost as remarkable as an authentic manuscript (Tiltman, 1967). The effort would, however, have been worth it for a payment of 600 ducats. On the other hand, minor inconsistencies have been observed within the text, such as subtle shifts in style between quires. These might indicate either lapses by a hoaxer or variations among multiple scribes if it were genuine (Timm & Schinner, 2020).

In 2020, researchers Torsten Timm and Andreas Schinner published a detailed computational analysis exploring whether the manuscript could have been algorithmically generated by a medieval process (Timm & Schinner, 2020). They concluded that many of the manuscript’s peculiar statistical properties, such as repeated word sequences and its near-binomial word-length distribution, could be reproduced using a quasi-random generative method. Their model involved a syllabic grid and positional rules that could generate text consistent with Voynichese. They proposed that the manuscript might be a “quasi-hoax”: not necessarily intended as deliberate fraud but rather as a linguistic experiment or an imaginative exercise that mimicked the structure of real language without conveying meaning. Timm and Schinner’s study remains one of the most comprehensive modern arguments in favour of the hoax hypothesis. It raises the uncomfortable possibility that there may, in fact, be no hidden key to the manuscript at all that it is an elaborate fabrication whose mystery lies precisely in its emptiness.

4.5 Authorship Theories

Alongside the linguistic and content-based interpretations, a wide range of hypotheses has been proposed regarding the authorship of the Voynich Manuscript. Each theory tends to intertwine assumptions about the manuscript’s content with the life and capabilities of a specific historical figure. While countless candidates have been suggested, several names recur frequently in the literature.

Roger Bacon (1214–1292)

Roger Bacon in his observatory at Merton College, Oxford. Oil painting by Ernest Board. 

Source: Painting

The English Franciscan friar and polymath Roger Bacon is perhaps the earliest and most enduringly cited candidate for the authorship of the Voynich Manuscript. A scholar of immense reputation in the thirteenth century, Bacon devoted his life to the study of optics, alchemy, mathematics, and experimental science, centuries ahead of his time. He championed empirical observation as the foundation of knowledge, and his surviving works, such as Opus Majus and Opus Tertium, reveal a mind fascinated by natural philosophy, the structure of language, and secret writing. Because of these traits, later generations easily imagined Bacon as the kind of scholar capable of producing a cryptic and illustrated compendium such as the Voynich Manuscript.

The association between Bacon and the manuscript stems largely from Johannes Marcus Marci’s 1665 letter to Athanasius Kircher, which claimed that Emperor Rudolf II had purchased the mysterious book for 600 ducats, believing it to be the work of Bacon. When Wilfrid Voynich rediscovered the manuscript in 1912, he embraced this connection, promoting the “Roger Bacon hypothesis” both to heighten scholarly interest and to strengthen the book’s market value. Voynich himself saw in the manuscript the embodiment of Bacon’s intellectual legacy, a fusion of science, nature, and mystical symbolism encoded to protect forbidden knowledge from the uninitiated.

However, modern research has thoroughly dismantled the Bacon attribution. Radiocarbon dating conducted by the University of Arizona (Stolte, 2011) conclusively demonstrated that the vellum dates between 1404 and 1438 CE, at least a century and a half after Bacon’s death. Furthermore, the ink and pigments correspond to standard fifteenth-century materials, confirming a much later creation date. Bacon’s own writing style dense scholastic Latin with extensive marginal commentary bears no resemblance to the concise, repetitive word structures of Voynichese. His known interest in cryptography was limited to references within broader discussions of secrecy, never amounting to the creation of a new written language or alphabet.

The continued endurance of the Bacon theory owes more to legend than to evidence. In the Renaissance and early modern imagination, Bacon was remembered as a “magus”, a visionary scientist who foresaw inventions like the telescope, gunpowder, and the microscope. To link the most mysterious book in the world to the most mysterious scholar of the Middle Ages was an irresistible idea. Yet, no documentary trail connects Bacon to the manuscript’s existence, and the chronological gap is insurmountable. Most modern scholars including D’Imperio (1978) and Kennedy & Churchill (2006) regard the Bacon hypothesis as a historical misunderstanding, born from the credulous environment of Rudolf II’s Prague court, where the line between science, alchemy, and myth was often blurred.

Nevertheless, Bacon’s name endures in discussions of the Voynich Manuscript not because he wrote it, but because he symbolizes the kind of knowledge the book seems to represent: the fusion of science, secrecy, and symbolism. The persistence of his name reminds us how easily mystery invites projection how the unknown text of the Voynich became a mirror for humanity’s fascination with lost wisdom and hidden science.

John Dee (1527–1608) and Edward Kelley (1555–1597)

(Left) John Dee, unknown artist, c.1900, Welcome Collection, London, (Right) Edward Kelley, unknown artist, c. 1900, Welcome Collection, London 

Source: Curious Archive

 Among all proposed authors, John Dee and Edward Kelley remain two of the most captivating and plausible figures in Voynich Manuscript lore, largely because of their proximity to Emperor Rudolf II’s court the same emperor who reportedly purchased the mysterious book for 600 ducats (Marci, 1665). Their reputations in alchemy, magic, and secret knowledge make them perfect candidates for a manuscript that blends herbal imagery, astrological diagrams, and cryptic writing.

John Dee was one of the most learned men of Elizabethan England: a mathematician, astrologer, geographer, and philosopher. He served as an adviser to Queen Elizabeth I and was deeply involved in studies of geometry, navigation, and the occult. His private library in Mortlake was one of the largest in England, containing works on alchemy, hermeticism, and cryptography including genuine manuscripts attributed to Roger Bacon, the same figure later linked to the Voynich. Dee’s lifelong fascination with hidden knowledge, symbolism, and the belief that divine truth could be encoded in numbers and letters made him the type of thinker who might conceivably produce such a manuscript or at least understand its purpose.

His associate Edward Kelley was a more controversial figure. Originally trained as an apothecary, Kelley reinvented himself as an alchemist and medium, claiming to communicate with angels through “spirit writing” a form of automatic dictation that he and Dee recorded in the Enochian language, an invented tongue of supposed celestial origin. Kelley’s flair for theatrical mysticism and his record of questionable alchemical claims (including promises to transmute metals into gold) made him both infamous and charismatic. Together, Dee and Kelley travelled through Bohemia and Central Europe in the 1580s, where they eventually came under the patronage of Rudolf II, the Holy Roman Emperor.

This connection is crucial to the hoax and authorship theories surrounding the Voynich Manuscript. Dee’s diary records a transaction with Rudolf involving a “book of great mystery” for a considerable sum the exact phrase “600 ducats” appearing in later retellings (Goldstone & Goldstone, 2005). If this diary reference indeed refers to the Voynich Manuscript, it would mean the book was already in circulation by Dee’s time. The Dee–Kelley authorship theory therefore suggests that Kelley, a skilled calligrapher and con artist, fabricated the manuscript perhaps drawing upon Bacon’s reputation and that Dee helped authenticate it as a genuine relic to sell to Rudolf for profit.

For this to be true, however, Dee and Kelley would have had to write the manuscript on much older vellum a difficult but not impossible feat. Blank parchment was occasionally repurposed in the Renaissance, especially within monastic or academic circles. Yet the Voynich’s consistent ink ageing, the uniform handwriting across gatherings, and the early 15th-century dating of the vellum (1404–1438) make such backdating improbable (Stolte, 2011; Barabe, 2009). Furthermore, nothing in Dee’s surviving letters or diaries hints at the creation of a massive ciphered work. While Dee possessed great knowledge of ciphers, his known writings are structured and clear, not cryptic or symbolically obscure.

The Kelley angle, though, remains tantalizing. His history of invented languages and his knack for constructing mystical systems could explain the manuscript’s strange alphabet and apparently structured but undeciphered text. Some scholars have drawn comparisons between Voynichese and Enochian, noting both share regular internal patterns and a systematic syntax despite their artificiality (Williams, 2019). Whether coincidence or connection, the resemblance supports the broader idea that the Voynich script might be a constructed system devised by someone like Kelley, familiar with both calligraphy and occult linguistics.

Ultimately, while the Dee–Kelley theory cannot be confirmed, it occupies a compelling middle ground between the hoax and constructed language hypotheses. It fits the social and intellectual climate of late sixteenth-century Prague, where mysticism, science, and deception often intertwined. Emperor Rudolf’s passion for alchemy and his willingness to spend lavishly on esoteric works make him a natural target for such a scheme.
However, the scientific evidence resists this romantic narrative: the vellum predates Dee and Kelley by over a century, and no direct link ties them to the book. Yet, they embody the type of men who might have desired, commissioned, or reinterpreted such a manuscript — men standing at the crossroads between medieval magic and the birth of modern science.

Jacobus Sinapius (Jacobi de Tepenecz) (d. 1622)

Etching of Jacobus Sinapius, copy of the original painting owned by Leop Ioh Scherschnik teacher of rhetoric of Teschen 

Source: Wikipedia

 Jacobus Sinapius, also known by his Latinised name Jacobi de Tepenecz, was a Bohemian physician, chemist, and the imperial distiller and apothecary at the court of Emperor Rudolf II in Prague. As the head of Rudolf’s botanical gardens and laboratories, Sinapius occupied a privileged position within one of Europe’s most intellectually diverse courts, where alchemy, medicine, and natural philosophy coexisted in a fertile but secretive environment. His connection to the Voynich Manuscript is not speculative: his faded signature appears on the bottom of the first folio (f. 1r), discovered under ultraviolet light by Yale conservators (Gurry, 2017). This single piece of handwriting provides one of the few verifiable physical links between the manuscript and a named historical individual.

Sinapius’s signature indicates that he owned the manuscript at some point in the late sixteenth or early seventeenth century, most likely as part of Rudolf II’s extensive collection of alchemical and botanical works. The signature, once visible to the naked eye, appears to have been intentionally scrubbed or erased, perhaps when the book changed hands within the Jesuit library where it was later found. This deliberate erasure may suggest that ownership of such a mysterious book carried reputational risk in an era when alchemical or heretical writings could attract suspicion.

Given his professional background, some researchers have speculated that Sinapius himself might have authored or commissioned the Voynich Manuscript. As a trained pharmacist and botanist, he possessed both the knowledge and the interest to create a compendium of herbs, medicinal preparations, and astrological correspondences — subjects all present in the manuscript’s illustrations. The idea of an encoded or secretive medical text would not have been unusual in the late Renaissance, when practitioners often cloaked their recipes and discoveries in cryptic writing to protect intellectual property or avoid ecclesiastical scrutiny.

However, the evidence supporting Sinapius as the author remains circumstantial at best. There are no surviving writings by him for comparison, and no contemporary accounts mention him as a scholar, cryptographer, or manuscript author. Moreover, the parchment’s radiocarbon dating to 1404–1438 CE places its creation long before Sinapius’s lifetime, ruling out his direct authorship of the physical book. It is possible, though unprovable, that he could have inherited, copied, or annotated an existing manuscript, perhaps interpreting it through the lens of his pharmaceutical work.

What makes the Sinapius connection valuable is not so much the authorship claim as the continuity of provenance it provides. His signature bridges the gap between the manuscript’s creation in the early fifteenth century and its confirmed presence in Rudolf’s Prague court two centuries later. This small but vital detail confirms that the book circulated among individuals involved in medicine, alchemy, and natural science, precisely the intellectual milieu one would expect for such a work.

In short, while Jacobus Sinapius was almost certainly not the author of the Voynich Manuscript, his ownership marks a crucial point in its documented history. The manuscript’s association with Rudolf II’s court — a hub of experimentation and esoteric inquiry — reinforces the idea that it was valued not merely as a curiosity but as a potentially meaningful artefact of hidden knowledge.

Athanasius Kircher (1602–1680)

Depicted person: Athanasius Kircher (1602-1680), Portrait from Mundus Subterraneus (1664)  

Source: Wikipedia

 Athanasius Kircher, a German Jesuit scholar and polymath, is one of the most intellectually fascinating figures associated with the Voynich Manuscript, though his link to it comes from his role as a recipient and interpreter, not as its creator. By the mid-seventeenth century, Kircher had achieved near-legendary status within the Catholic world as a universal scholar. He taught at the Collegio Romano in Rome and wrote prolifically on topics ranging from geology, magnetism, optics, and linguistics to music, theology, and Egyptology. His reputation for deciphering unknown scripts — most famously his ambitious but incorrect “translation” of Egyptian hieroglyphs — led many to consider him a master cryptologist and interpreter of the world’s hidden languages.

The Voynich Manuscript entered Kircher’s life through a letter written in 1665 by Johannes Marcus Marci, rector of Charles University in Prague. In this letter, Marci described the mysterious book and enclosed it as a gift, explaining that it had once belonged to Emperor Rudolf II, who had reportedly purchased it for 600 ducats believing it to be the work of Roger Bacon (Marci, 1665). Marci’s hope was that Kircher — famous for deciphering Coptic and for his linguistic curiosity — might finally uncover its meaning. This letter, preserved with the manuscript today, is the last known piece of its early provenance and firmly establishes Kircher as a pivotal historical custodian.

Because of his known expertise in cryptography and ancient scripts, some later writers speculated that Kircher himself could have fabricated the Voynich Manuscript, either as a demonstration of his linguistic ingenuity or as a hoax to amuse or mystify his peers. The idea fits the image of Kircher as both a genius and an eccentric: he invented machines that produced music through water pressure, published elaborate symbolic diagrams, and filled his works with layers of allegory. A few modern commentators suggested that the manuscript’s script might resemble Kircher’s playful pseudo-languages used in his pedagogical works.

However, the chronological and material evidence completely rules out this possibility. The manuscript’s vellum, dated between 1404 and 1438 CE (Stolte, 2011), predates Kircher’s birth by almost two centuries. The iron-gall ink and medieval pigments likewise confirm that it is a genuine fifteenth-century creation (Barabe, 2009). Furthermore, there is no record in Kircher’s extensive correspondence or publications suggesting he ever claimed authorship or hinted at producing such a document. On the contrary, Kircher expressed genuine curiosity about the manuscript’s origin but offered no decipherment. His papers include no translation attempts, and no mention of progress, implying that even he — the era’s most confident cryptological mind — was unable to make sense of it.

Despite this, Kircher’s role in the story remains significant. His vast network of scholars ensured that the manuscript was preserved within Jesuit collections after his death, ultimately leading to its rediscovery by Wilfrid Voynich in 1912 at the Villa Mondragone, where the Jesuits had stored part of the Kircher correspondence archive. In that sense, Kircher was indirectly responsible for the manuscript’s survival into the modern age.

In modern scholarship, Athanasius Kircher is no longer seen as a candidate for authorship but rather as the final custodian of the manuscript’s early European journey — the man who received it, puzzled over it, and safeguarded it through the centuries. His association underscores a fascinating irony: even a man who claimed to read the “language of the universe” could not decipher the Voynich Manuscript. His failure highlights the text’s enduring opacity and its ability to confound even the most learned minds of the seventeenth century.

Wilfrid Voynich (1865–1930)

Wilfrid Voynich was a Polish revolutionary, antiquarian and bibliophile. Voynich operated one of the largest rare book businesses in the world. He is remembered as the eponym of the Voynich manuscript. 

Source Wikipedia

 Wilfrid M. Voynich, the Polish-American antiquarian and bookseller after whom the manuscript is now named, occupies a unique place in its history — not as its creator, but as its rediscoverer and publicist. Born Michał Habdank-Wojnicz in modern-day Lithuania, Voynich was a political exile who later established himself in London as a rare book dealer. In 1912, while examining materials held at the Villa Mondragone (a Jesuit college near Frascati, Italy), Voynich discovered a strange, illustrated codex written in an unknown script. Recognising its potential significance, he purchased it along with several other manuscripts. This acquisition marked the beginning of the modern history of the Voynich Manuscript.

Upon acquiring the codex, Voynich immediately realised it was unlike anything he had ever encountered. The book’s combination of enigmatic script, intricate botanical drawings, and bizarre cosmological diagrams convinced him that it must contain important scientific or philosophical knowledge from the medieval or Renaissance period. Hoping to attract scholarly attention — and perhaps increase its market value — Voynich promoted the theory that it might be the lost work of Roger Bacon, the thirteenth-century English friar and polymath. This claim drew from the 1665 letter by Johannes Marcus Marci, which Voynich found tucked inside the manuscript. The letter stated that Emperor Rudolf II had once purchased the book for 600 ducats, believing it to be Bacon’s work (Marci, 1665). Voynich saw this as both historical validation and an opportunity to link his find to a famous name.

In the years that followed, Voynich showcased the manuscript to scholars and codebreakers across Europe and America, including at the University of Pennsylvania, where it piqued the interest of William and Elizebeth Friedman — the future founders of modern American cryptology. He published facsimiles and wrote promotional materials emphasising the manuscript’s supposed Baconian connection, describing it as “an unknown cipher by the great Roger Bacon.” This association, while scientifically baseless, successfully transformed the manuscript from a neglected relic into a global mystery that still bears his name.

Some critics in the mid-twentieth century proposed that Voynich himself might have forged the manuscript, motivated by the desire to create a sensational discovery and secure his reputation. This “Voynich forgery hypothesis” gained brief traction after the debunking of William Newbold’s supposed Baconian decipherment in the 1920s (Manly, 1931). However, all scientific analyses have since disproven the possibility of a modern forgery. The radiocarbon dating of the parchment (1404–1438 CE), the medieval composition of the inks, and the use of period-appropriate pigments conclusively demonstrate that the book existed several centuries before Voynich’s birth (Stolte, 2011; Barabe, 2009). Moreover, no evidence suggests that Voynich had access to blank medieval vellum in the quantities required to fabricate a 240-page codex, nor the technical ability to reproduce authentic fifteenth-century handwriting and materials.

Instead, Voynich’s true role lies in his rediscovery and reintroduction of the manuscript into modern scholarship. His energetic promotion of the book revived interest in medieval cryptography, linguistic history, and the study of lost knowledge. Through his efforts, the manuscript came to the attention of major institutions and researchers, eventually being acquired by Hans P. Kraus, who later donated it to Yale University’s Beinecke Rare Book & Manuscript Library, where it remains today.

While Wilfrid Voynich’s speculative claims contributed to the myths surrounding the manuscript, his discovery ensured its preservation and cemented its place in the modern imagination. Without him, the manuscript might have remained forgotten in a Jesuit vault. Ironically, though he lent it his name, Voynich likely understood as little of its contents as anyone before or since — a fitting reminder that even the man who rediscovered it could not unlock its mysteries.

Giovanni Fontana (c. 1395–1455)
Giovanni Fontana was a Venetian engineer, physician, and inventor whose creative brilliance epitomised the scientific imagination of the early Renaissance. Active in the first half of the fifteenth century, Fontana is remembered for his works on mechanics, optics, and military engineering, as well as for his fascination with secret writing and automation. He produced several illustrated manuscripts, the most famous of which, Bellicorum Instrumentorum Liber (“Book of War Instruments”), contains hundreds of diagrams of siege machines, hydraulic devices, and even designs for a self-propelled cart and a mechanical human figure — early expressions of what we might call proto-robotics.

Fontana’s connection to the Voynich Manuscript arises primarily from the presence of cipher-like scripts in his surviving works. In Bellicorum Instrumentorum, Fontana used a simple substitution cipher to label his machines, replacing Latin letters with geometric and symbolic characters. This script bears a superficial resemblance to the flowing, unfamiliar glyphs of Voynichese, leading some modern researchers to suggest that Fontana could have been the manuscript’s author or at least influenced its creation (Smith, 2014). Moreover, one of Fontana’s sketches — a devil encircled by rings of coded text — has often been compared visually to the mysterious circular diagrams found in the Voynich cosmological section.

Chronologically, Fontana’s lifetime aligns perfectly with the radiocarbon dating of the Voynich parchment (1404–1438 CE) (Stolte, 2011). This temporal overlap makes him one of the few proposed authors whose lifespan coincides with the manuscript’s physical creation. His professional background as a university-trained physician at Padua and his documented interest in cryptography, optics, and alchemy would have equipped him with precisely the kind of multidisciplinary knowledge reflected in the Voynich’s mixture of herbal, astronomical, and anatomical imagery. In addition, Venice and northern Italy were major centres of manuscript production and cryptographic experimentation during his career, providing a plausible cultural context.

However, despite these intriguing parallels, the Fontana hypothesis falters under closer examination. The cipher used in his known writings has been fully deciphered and is merely a monoalphabetic substitution of Latin letters, easily translatable once the key is known. By contrast, the Voynich script exhibits far more complex statistical behaviour, inconsistent with any simple substitution system (Landini, 2001; Montemurro & Zanette, 2013). Furthermore, the structure of Fontana’s texts — clear Latin prose hidden by an artificial alphabet — differs entirely from the Voynich Manuscript’s internally coherent but undeciphered linguistic system. No linguistic or stylistic connections between the two have been demonstrated, and no surviving Fontana manuscript employs the kind of word structure or repetition found in Voynichese.

Some art-historical comparisons also prove tenuous. The devil illustration, for instance, was not a unique symbol but part of a long iconographic tradition of medieval and Renaissance marginalia depicting demonic guardians of knowledge. Similar motifs appear throughout fifteenth-century Italian art. The resemblance, then, may reflect a shared cultural language rather than direct authorship.

Nevertheless, Giovanni Fontana remains one of the more plausible authorship candidates because he perfectly embodies the intellectual synthesis that the Voynich Manuscript seems to represent — the merging of science, art, secrecy, and invention in an age of rediscovery. If not the author, he may well stand as the kind of polymath who could have created it. The Fontana hypothesis, even if circumstantial, demonstrates that the manuscript’s mysterious language and mechanical imagery were not entirely alien to its time but rather extensions of the experimental and cryptographic spirit of early fifteenth-century Italy.

Other Candidates
Other, less substantiated names include Antonello da Messina and various anonymous artists or scribes. However, the manuscript’s illustrations are not of high artistic quality, suggesting it was not the work of a professional painter but rather of a competent scribe with modest artistic training. Consequently, most theories focus on scholarly or alchemical authors rather than artists.

Overall, no proposed author fits all available evidence. The manuscript’s creation required a rare combination of skills: botanical and medical knowledge, some familiarity with astrology and alchemy, and the ingenuity to devise a unique writing system. This composite profile could describe a learned friar, physician, or a member of an educated religious order. Some scholars have speculated that it might have been the work of a small community such as a convent of nuns engaged in herbal medicine and esoteric study. In such a scenario, multiple scribes could explain the different handwriting styles and the manuscript’s thematic diversity. However, no concrete evidence supports this, and the true author or authors of the Voynich Manuscript remain unknown.

4.6 Summary of Theories

In summary, the Voynich Manuscript occupies a unique position at the crossroads of multiple interpretations. It may be a ciphered compendium of medieval knowledge an encrypted encyclopaedia combining botany, astrology, and medicine either to conceal esoteric learning or for intellectual amusement. Alternatively, it could be a meaningless fabrication, an elaborate hoax crafted to imitate scholarly depth. Between these extremes lies the possibility that it represents an invented or constructed language, used earnestly by a small group or individual to record ideas in a personal or secret form.

The central challenge remains the absence of a “Rosetta Stone” a definitive reference or parallel text that could link Voynichese to a known language or code. Without such a key, all hypotheses rest on circumstantial or statistical evidence. Periodically, new claims of decipherment emerge, often attracting media attention, yet none has withstood rigorous testing by the scholarly community. Each proposed decoding can be objectively verified by applying it to different sections of the manuscript, and every attempt so far has failed to produce consistent or meaningful results.

Thus, the mystery endures. The Voynich Manuscript has evolved from a cryptographic puzzle into a cultural phenomenon, embodying the human fascination with the unknown. Its allure lies not only in the possibility of hidden knowledge but also in the enduring uncertainty that surrounds it. In the next section, we will examine the scientific investigations that have been applied to the manuscript in modern times, which continue to shed light on its material nature while constraining and sometimes challenging the theories discussed above.

5.0 Scientific Investigations

Modern research on the Voynich Manuscript has gone far beyond traditional paleography and cryptography. A wide array of scientific techniques and computational analyses have been employed to study the manuscript’s materials, text statistics, and linguistic patterns. These investigations help authenticate the artifact, provide clues to its origin and composition, and test hypotheses about the nature of the text. Here we review key scientific and analytical approaches applied to Voynich studies, and what they have revealed so far.

5.1 Material Analysis and Forensics

Graph 1: The radio-carbon calibration curve, Graph 2: The radio-carbon calibration curve expressed in dates. Calibrated dates are on the horizontal scale and un-calibrated dates on the vertical scale.

First graph 1 above, the smooth blue curve is the hypothetical relation between 14C fraction and age, while the more complicated red curve more closely represents reality reconstructed from the above-mentioned tree ring analysis. This second graph represents the so-called calibrated date. Both are shown for the time frame of interest for the Voynich MS, but can be used also for times much further in the past. In Second graph, un-calibrated and calibrated dates are shown against each other, with the un-calibrated date on the vertical scale.

Source: Voynich.nu

As discussed in the Background, one of the first major breakthroughs in Voynich research was the radiocarbon dating of the parchment. In 2009, a team at the University of Arizona’s Accelerator Mass Spectrometry Laboratory analysed tiny samples taken from four different pages of the manuscript. The results, published in 2011, revealed that the calfskin parchment was produced between 1404 and 1438 CE, with 95% probability (Stolte, 2011). All four samples yielded consistent dates within this range, confirming that the entire codex was created in the early fifteenth century. This provided a scientific terminus ante quem for the manuscript’s creation and securely anchored it in the medieval period. It also effectively ruled out later authorship theories, such as those involving Voynich or his contemporaries, since the material itself is centuries older.

Following the parchment dating, ink analysis was conducted to determine whether the writing might have been added long after the parchment was produced. Using techniques such as energy-dispersive X-ray spectroscopy (EDS) and Raman spectroscopy on microscopic ink samples, researchers determined that the ink is consistent with iron gall ink, a standard formulation used in medieval manuscripts (Barabe, 2009). Iron gall ink, composed of oak galls, iron sulphate, carbon, and binders, typically contains iron, sulphur, and potassium, all of which were identified in the Voynich inks [22]. Importantly, no modern pigments or synthetic dyes (such as Prussian blue) were detected. The ink displays natural ageing characteristics, having darkened and absorbed into the parchment fibres over time. Forensic analysts at McCrone Associates concluded that none of the materials in the manuscript post-date the early seventeenth century, with most components clearly originating much earlier (Barabe, 2009).

The pigments used for the manuscript’s illustrations were also found to be consistent with fifteenth-century materials. They include azurite for blue, malachite or copper resinate for green, red ochre (iron oxide) for red-brown tones, and calcium carbonate (chalk) mixed with egg white as a white tempera base. These are all typical of medieval illumination and further support the manuscript’s authenticity. A minor discovery was the presence of limited retouching: in several places, faded or faint strokes were traced over with darker ink (Stolfi, 2001). This suggests that a later owner possibly in the seventeenth or eighteenth century, perhaps even Kircher himself may have attempted to enhance the legibility of certain areas. The retouching appears minimal and does not alter the underlying drawings.

In 2014, protein testing confirmed that the parchment is derived from calfskin (Bos taurus) (Bowern, 2020). Although DNA analysis has not yet been fully completed or released, such work could, in the future, provide clues about the geographical origin of the animals or the microbial environment of production. Multispectral imaging, conducted by conservators at Yale University, has revealed further details. Under ultraviolet light, the faint signature of Jacobi de Tepenecz at the bottom of folio 1r becomes visible, a mark otherwise nearly invisible after centuries of fading and attempted erasure (Gurry, 2017). Imaging in other wavelengths has also been used to search for underdrawings or erased text, but no evidence of preliminary sketches or palimpsest writing has been found, indicating that the scribe-illustrator worked directly in ink on clean sheets.

The manuscript’s codicology and binding have also been carefully examined. The current cover is eighteenth-century limp vellum, likely added by Jesuit custodians after the manuscript entered their possession. However, quire numbering in a fifteenth-century hand appears on several pages, showing that the gatherings were numbered during or soon after the manuscript’s original production (Currier, 1976). These numbers confirm that several folios are missing particularly within the herbal section. It appears that the manuscript was rebound at least once, during which some pages may have been lost or rearranged. Some foldouts, such as the famous rosette diagram (folios 85–86–87), likely formed a single large composite sheet that was later reconfigured (Tiltman, 1967). Despite these alterations, the codex remains largely intact, with an estimated fourteen folios missing from the original total.

Altogether, these forensic studies present a coherent picture: the Voynich Manuscript is an authentic medieval artefact, produced in the early fifteenth century using period-appropriate materials and techniques. No evidence suggests any modern forgery or tampering beyond standard rebinding and minor retouching. Consequently, any interpretation of the manuscript’s contents must be credible within a fifteenth-century intellectual and cultural framework. This understanding has directed serious scholarship towards medieval ciphers, linguistic systems, and contemporary scientific traditions, while dismissing speculative notions of extra-terrestrial or mythical origins.

5.2 Computational Linguistic Analysis

Comparison of the Voynich manuscript and different information carrying sequences. (A) Information in word distribution as a function of the scale for the Voynich manuscript compared to other five language and symbolic sequences (F: Fortran; C: Chinese; V: Voynich; E: English; L: Latin; Y: yeast DNA). The number of words in all sequences was equal to that of the Voynich text; if the original sequence was longer, the additional words were not considered. (B) Scale of maximal information for the sequences considered in A (see Materials and Methods for more details on the language and symbolic sources).

Source: Montemurro, M. A. & Zanette, D. H. (2013) 

One of the most promising approaches to understanding Voynichese has been computational analysis. Using methods drawn from linguistics and computer science, researchers have sought to characterise the statistical properties of the text and to compare them with known languages and cipher systems.

A fundamental measure in such studies is entropy a way of quantifying the degree of randomness or predictability in a sequence of text. During the 1990s and 2000s, scholars calculated both the character-level and word-level entropy of the Voynich text. The character entropy, which measures how predictable a character is based on the one preceding it, is lower than that of English (indicating a more repetitive or patterned structure) but higher than that of a simple repetitive cipher (D’Imperio, 1978; Landini, 2001). The word-level entropy, which reflects vocabulary diversity, is comparable to that of natural languages. These results suggest that Voynichese is not a random or meaningless string of symbols; rather, it possesses internal informational structure similar to that of genuine language, though possibly with less complexity than typical European languages (Bowern & Lindemann, 2021).

Researchers have also examined n-gram frequency—the occurrence of common pairs or triplets of characters. The Voynich text displays frequent bigrams such as “or”, “ol”, and “ar”, as well as forbidden combinations where certain letters almost never appear together. In the EVA (Extensible Voynich Alphabet) transliteration system, for example, “q” is nearly always followed by “o” or “oa”, analogous to the way “q” in English is followed by “u” (Currier, 1976). This kind of rule hints that Voynichese may encode vowels and consonants in a systematic pattern. Some characters appear predominantly at the beginnings or ends of lines (Grove, 2019). There is also a notable “first-letter effect”, where certain characters known as gallows occur disproportionately often as the first character of a line, as though the scribe were following a line-initial convention. Some researchers have speculated that this may indicate a cipher mechanism, such as a null or filler character, while others suggest it could correspond to a linguistic feature such as capitalisation or paragraph marking (Stolfi, 2004).

In 2013, theoretical physicist Marcelo Montemurro and his colleague Damián Zanette applied information-theoretic analysis to the text of the Voynich Manuscript (Montemurro & Zanette, 2013). They investigated whether certain words clustered together in meaningful ways, as topic-related words do in real texts. In ordinary writing, content-bearing words tend to appear in specific sections, while function words are evenly distributed. Montemurro and Zanette found that certain Voynich words occur far more frequently in some sections, such as the herbal section than in others, such as the astrological section. This mirrors how genuine books use domain-specific vocabulary. They even produced semantic networks showing groups of word-tokens that frequently co-occur. Their analysis concluded that the distribution of words in the manuscript is “compatible with those found in real language sequences”, implying that the text likely contains meaningful content [23]. This finding has often been cited as strong evidence against the notion that the Voynich Manuscript is a mere hoax, since a random or pseudo-random text would not easily display such structured patterns.

A separate computational study in 2017 by computer scientists at the University of Alberta employed Bayesian inference to infer the underlying language of Voynichese, assuming it to be a simple substitution cipher (Hauer & Kondrak, 2019). Their model compared the sequence of characters to hundreds of languages and identified Hebrew as the best statistical match. The algorithm produced a tentative translation of the manuscript’s opening line into Hebrew after reordering the letters (anagramming). However, when Hebrew scholars reviewed the output, it did not form a coherent sentence. The researchers themselves acknowledged that the results were largely aspirational, noting that the required degree of rearrangement undermined the method’s validity (Schnapp, 2018). While inconclusive, their work contributed valuable insights into the text’s statistical alignment with certain linguistic profiles. The resemblance to Hebrew was interesting historically, as some earlier scholars had speculated that Voynichese might represent a consonant-based writing system, similar to Semitic scripts. Yet, no convincing Hebrew decipherment has ever been accepted.

A more linguistically grounded investigation was undertaken by Claire Bowern, a linguist at Yale University, and Luke Lindemann (Bowern & Lindemann, 2021). They approached the manuscript from the perspective of historical linguistics, treating Voynich “words” as if they belonged to an unknown natural language. Using comparative linguistic methods, they tested whether its structure aligned with known language families whether, for instance, it resembled agglutinative or isolating languages. They also examined whether the statistical features might correspond to an East Asian language transcribed into a cipher alphabet. Their results suggested that if Voynichese represents a natural language, it does not clearly belong to any major Eurasian family and might instead represent a linguistic isolate or a lost language (Bowern, 2020). Some of their statistical measures, such as conditional entropy, placed Voynichese closer to certain East or Southeast Asian languages than to Indo-European ones, echoing earlier observations by Reddy and Knight (2011). However, they cautioned that such measures reflect only similarity in structure, not direct translation.

Other multidisciplinary computational projects have also emerged. Between 2019 and 2020, researchers at the University of Malta developed natural language processing (NLP) tools to analyse the manuscript (Žagar, 2020). Their work involved identifying repeated phrases, aligning text segments, and correlating words with illustrations to infer possible meanings. For instance, if the same Voynichese word label repeatedly appears near drawings of similar plants, it might denote a category such as “leaf” or “root”. Although these methods have not led to a decipherment, they have provided valuable insight into the manuscript’s internal organisation. One notable finding from computational studies is that Voynichese lacks one-letter words, a feature unusual for natural languages. Another is the discovery of “self-correlation”: when the text is compared against itself at varying intervals (autocorrelation), it exhibits non-random peaks at certain offsets, indicating rhythmic or patterned repetition at the phrase level (Timm & Schinner, 2020). Some researchers interpret this as evidence that the text may have been generated according to a repeating template or algorithmic method.

Overall, computational linguistic analysis continues to play a central role in Voynich studies. While no algorithm or statistical method has succeeded in deciphering the text, these investigations have revealed that Voynichese possesses structural complexity and internal logic. Whether this reflects an unknown language, an artificial system, or a deliberately constructed cipher remains unresolved, but modern computational approaches have brought the mystery into sharper scientific focus.

5.3 Imaging and Hidden Details

The three columns of lettering discovered on the Voynich manuscript – image courtesy Lisa Fagin Davis 

Source: Medievalists 

Beyond textual and statistical studies, advanced imaging techniques have been employed to uncover any hidden information within the Voynich Manuscript’s pages. Multispectral imaging, which captures light beyond the visible spectrum, can sometimes reveal faded writing, underdrawings, or erased notes. As previously mentioned, ultraviolet imaging successfully exposed the nearly invisible signature of Jacobi de Tepenecz on folio 1r. This technique can also enhance faint pigment layers or reveal annotations obscured by age. However, aside from the Tepenecz signature and the clarification of minor scribal variations, multispectral imaging has not uncovered any hidden texts or secret layers of writing in the manuscript. Earlier hopes that marginal notes or erased captions might exist beneath the surface have not been realised, aside from the marginal Latin note on folio 17r, which is already visible to the naked eye [24].

Another avenue of investigation has focused on ink flow and stroke analysis. Recent advances in digital imaging and algorithmic handwriting recognition have allowed researchers to classify letter shapes, pen pressure, and stroke order. These data can reveal how many scribes contributed to the manuscript and whether the variations between letters represent stylistic habits or deliberate distinctions (Pelli et al., 2022). Lisa Fagin Davis (2020) applied digital palaeography tools, including DigiPal, to tag and cluster recurring letter forms across the manuscript. Her analysis identified five distinct scribal hands, based on consistent differences in how specific glyphs such as the “n” and “s”-like characters were formed (Fagin Davis, 2020). This scientific approach effectively provides a “fingerprint” of each scribe’s handwriting, confirming that the manuscript was a collaborative effort rather than the work of a single author. This finding has important implications: if all five scribes employed the same writing system, then the underlying code or language must have been shared knowledge, which in turn makes the notion of a purely individual hoax less plausible.

Another intriguing line of study, led by Zandbergen and colleagues, examined the sequence in which the manuscript was produced specifically, whether the ink outlines were drawn before or after the paint was applied. Through microscopic inspection, they determined that the ink outlines were laid down first, with the paint applied afterwards, sometimes clumsily extending beyond the lines (Zandbergen, 2016). This is typical of illuminated manuscripts, in which illustrations were first outlined and then coloured. The observation confirms that the drawings and text were part of the original creative process rather than later additions. The uneven application of paint suggests that the colouring may have been carried out by someone other than the principal scribe or perhaps completed hastily. Differences in pigment density across folios also imply that the colouring was performed in batches rather than continuously, although this remains difficult to verify without detailed pigment analysis on a page-by-page basis.

Overall, imaging and forensic inspection of the Voynich Manuscript have deepened our understanding of its material history. They confirm that its creation was deliberate and carefully planned, yet also collaborative and practical in execution. While no hidden writing has been discovered, these analyses reinforce the view that the manuscript’s production followed normal medieval practices, lending further weight to its authenticity as a genuine fifteenth-century artefact.

5.4 Efforts at Decipherment

Elizabeth and William Friedman at Peabody Institute Library on tour for their book, "The Shakespearean Ciphers Examined." 1960. Credit: George C Marshall Foundation Library. Source: American experience

Among the most distinguished figures to tackle the Voynich Manuscript were William and Elizebeth Friedman, the legendary husband-and-wife cryptologists whose work shaped modern codebreaking. Both had served as leading cryptanalysts for the U.S. government, and in the 1940s, William Friedman directed a small team at the U.S. Army’s Signals Intelligence Service (SIS) specifically to study the Voynich cipher. The Friedmans approached the manuscript with the most advanced cryptographic tools of their time, applying frequency analysis, statistical tests, and mechanical cipher machines to the Voynich text. Despite their unmatched expertise and rigorous methodology, they ultimately concluded that the manuscript did not conform to any known cipher system. In 1959, William Friedman privately wrote that the Voynich Manuscript might represent a “constructed language” a form of artificial linguistic system rather than a straightforward code. This cautious interpretation was one of the earliest to suggest that the manuscript’s mystery might lie not in encryption but in language invention itself. The Friedmans’ decades-long engagement with the text exemplifies the enduring allure and frustration it has posed even to the world’s greatest cryptographers a puzzle that resisted both genius and technology alike. 

While not scientific in the laboratory sense, the long history of direct attempts to decipher the Voynich Manuscript deserves attention as part of its broader investigation. Over the past century, many professional cryptanalysts and linguistic experts have tried to unlock the manuscript’s secrets using the best tools and theories of their time.

In the 1950s, John Tiltman, a renowned British codebreaker, conducted one of the earliest systematic modern analyses. He concluded that simple substitution ciphers could be ruled out, suggesting instead that the text might use a more complex or disguised classical cipher (Tiltman, 1967). During the 1970s, Prescott Currier and colleagues associated with the U.S. National Security Agency also examined the manuscript in detail. Currier’s 1976 paper compiled an extensive body of observations, including character frequency studies and transcription efforts, effectively laying the groundwork upon which later researchers have built (Currier, 1976). Despite their sophistication, these early computer-assisted analyses failed to produce a decipherment, underscoring how resistant the manuscript is to conventional cryptographic methods.

With modern computing power, researchers have attempted more advanced approaches, including brute-force and heuristic attacks. Some experiments have tried testing all possible alignments of Voynich characters to letters of known alphabets to see if any generate readable text in any language. However, the potential key space particularly if polyalphabetic or homophonic systems are considered is astronomically large, rendering such methods impractical. A 2019 study by Amancio and colleagues applied linguistic modelling and word-ending pattern analysis to test whether Voynichese words might be anagrams of Latin or other languages. Although their statistical results were interesting, the reconstructed texts did not produce coherent meaning (Amancio et al., 2019).

Researchers have also tried contextual analysis by comparing the manuscript’s illustrations with known medieval sources. For example, the herbal drawings resemble medicinal plants from fifteenth-century herbals. On that basis, Edith Sherwood (2008) proposed several plant identifications, such as Viola tricolor (wild pansy), known in Latin as viola. She noted that a similar-looking Voynich word appears beside the plant’s image, speculating that it might represent the plant’s name. While intriguing, such identifications remain speculative and lack scholarly consensus.

Another avenue of investigation concerns the astronomical diagrams, which contain recognisable zodiac symbols accompanied by Latin month names. Around each zodiac circle are roughly thirty small female figures, many holding stars or enclosed within segments. It has been suggested that these could represent the thirty days of each zodiac sign or perhaps individual stars. Each figure is labelled with a Voynich word, raising the possibility that these labels might be star or constellation names, or even personal names linked to birth dates. However, despite extensive comparison with medieval astronomical and astrological sources where star names were often of Arabic origin—no clear correlation has been established. Even in parts of the manuscript where the general subject matter is known, such as the zodiac or herbal sections, the text itself remains impenetrable.

In sum, after decades of systematic and computational attempts, no decipherment has yet succeeded. The persistent opacity of the script suggests that if the text is meaningful, it may employ an unknown linguistic system or an encryption method unlike any historically documented cipher. 

5.5 Current and Future Investigations

As of 2025, the Voynich Manuscript remains unsolved, yet research is more active and interdisciplinary than ever. Current studies bring together expertise from cryptography, computer science, linguistics, chemistry, and art history to approach the mystery from multiple angles.

Deep Neural Networks. Researchers have begun training neural-network models on Voynichese text to detect internal structure or classify its sections (Jaskiewicz, 2020). Some experiments attempt to generate Voynich-like text or to cluster glyphs statistically. Although direct translation to a known language remains unachievable without a parallel corpus, these models help identify stylistic and structural consistencies that might correspond to linguistic features.

Linguistic Signal Analysis. Beyond basic entropy, scholars are examining higher-order linguistic patterns such as potential morphology. Some Voynich words appear to share common endings (for example “-dy” or “-dy.qo” in EVA transliteration), possibly functioning like grammatical suffixes (Landini, 2001). Cataloguing these recurring affixes could reveal whether the manuscript’s language family resembles those with case endings such as Latin, or languages with minimal inflection. So far, regularities have been observed but not securely linked to any known grammar.

Collation and Transcription Projects. Complete, high-quality digital transcriptions now exist in machine-readable form, most notably the Takahashi and EVA datasets. These allow researchers to apply computational analyses without manually transcribing from images. Ongoing projects, such as those led by René Zandbergen (2020), aim to produce an authoritative digital edition aligning each folio’s transcription with its image, enabling interactive searching and automated pattern detection. Such resources have already revealed several repeated multi-word phrases within the manuscript, hinting that formulaic expressions or standardised recipes may exist.

Multilingual Cipher Hypothesis. A few cryptographers have proposed that different sections might encode different languages or subjects. Currier’s identification of two “languages”, A and B, could reflect two distinct plaintexts enciphered by a shared system. It remains speculative but would significantly complicate decipherment if true.

Palaeographic and Art-Historical Studies. Art-historical analysis continues to refine our understanding of the manuscript’s geographic origin. Features such as the apothecary jars resemble those in northern Italian herbal manuscripts from c. 1440–1480 (Ciompi & Zandbergen, 2018). Likewise, the depiction of Sagittarius as a crossbowman a motif common in French and northern Italian calendars suggests a Central or Southern European provenance within the cultural sphere of the Holy Roman Empire or Italy. The style of dress and architecture also aligns with this region, supporting the European origin already indicated by material analyses.

Scientific Content Matching. Some researchers test whether the astronomical and biological diagrams correspond to known phenomena or contemporary scientific imagery. For instance, one circular foldout has been examined for possible alignment with real star charts or astronomical clocks, but no convincing matches have been found. Similarly, the “biological” diagrams featuring pipes and pools have been compared with anatomical and alchemical imagery from the fifteenth century. While some shapes resemble organs such as intestines or fallopian tubes, there is no clear anatomical correspondence, and they may instead be symbolic or alchemical in nature.

Future Directions. Promising new approaches include multilingual alignment algorithms similar to those that helped decipher Ugaritic and other ancient scripts. By simulating the encipherment of large Latin or vernacular corpora under various cipher rules, researchers hope to identify statistical matches to Voynichese. With greater computational power, such reverse-engineering could narrow possible cipher types. Another avenue is organised crowdsourced collaboration: coordinated “citizen-science” efforts where distributed participants or AI agents collectively test partial solutions. Informal online groups already attempt this, but a structured framework could accelerate progress.

Historians are also searching archives for external references to an unreadable book from the fifteenth century. Although one seventeenth-century letter from Johannes Marcus Marci mentions sending the manuscript to Athanasius Kircher, no contemporary records of its earlier circulation have yet surfaced. Should a medieval reference to such a mysterious codex ever be discovered, it could greatly clarify its provenance.

In summary, modern investigations have firmly established what the Voynich Manuscript is not: it is not modern, not random, and not a trivial cipher. Instead, it presents a complex blend of linguistic regularity and symbolic opacity that continues to challenge conventional decipherment techniques. The convergence of chemistry, computing, linguistics, and history makes the Voynich Manuscript a unique case study in interdisciplinary research. As analytical technology advances perhaps through quantum computing or more sophisticated AI there remains cautious optimism that the manuscript’s secret may one day be revealed. At the same time, researchers acknowledge the possibility, raised by Timm and others, that a meaningless yet meticulously constructed text could imitate linguistic order so convincingly that no algorithm can definitively prove otherwise. The Voynich Manuscript thus stands as both an enduring enigma and a test of the limits of human and artificial decipherment.

6.0 Conclusion

The Voynich Manuscript remains one of the most captivating unsolved enigmas in the fields of history, cryptography, and linguistics. Throughout this report, we have traced the known facts and scholarly research surrounding the manuscript from its rediscovery by Wilfrid M. Voynich in 1912, and the historical trail that links it to Emperor Rudolf II’s court, to the scientific analyses that authenticate it as a genuine fifteenth-century artefact. We have examined its contents, which appear to form an encyclopaedic compendium of herbal, astronomical, biological, and medicinal knowledge, all rendered in an unknown script accompanied by intricate and often perplexing illustrations.

We have also reviewed the principal theories concerning its purpose and meaning. Some scholars view it as a cipher concealing a natural human language, perhaps an obscure or extinct dialect, while others suggest that it represents an entirely constructed language or artificial system devised by a small circle of users. A further possibility is that it was an elaborate medieval hoax, created to intrigue or deceive. Each theory has its merits and weaknesses, yet none has succeeded in producing a verifiable translation of even a single section of the text.

Scientific and scholarly investigations have greatly advanced our understanding in several respects. Forensic analyses have determined when and where the manuscript was created most likely between 1404 and 1438 in a European cultural context. Statistical and computational studies show that Voynichese possesses structural complexity comparable to that of natural languages, suggesting that it encodes information of some kind (Montemurro & Zanette, 2013). However, unusual traits such as repetitive sequences and restricted character placement leave open the possibility of a cleverly designed but meaningless text (Timm & Schinner, 2020). No method traditional cryptanalysis, artificial intelligence, or multispectral imaging has yet revealed the underlying message, but each has contributed valuable data. The identification of multiple scribes through digital palaeography (Fagin Davis, 2020) suggests collaboration, implying that the manuscript may have originated within a small intellectual or monastic community. The presence of Latin month names in the zodiac section further anchors it within the linguistic and cultural environment of medieval Europe (Tiltman, 1967).

Over the past century, numerous decipherment claims have been proposed, yet none have withstood rigorous testing. Each has ultimately failed to account for the text in its entirety or has relied on speculative assumptions. The true fascination of the Voynich Manuscript lies in this enduring resistance to solution a reminder that, even in an age of powerful computation and global scholarship, mysteries from the past can still evade explanation. Nevertheless, substantial progress has been made. We can now state with confidence what the manuscript is not: it is not the work of Roger Bacon, not a modern forgery, and not meaningless random text. The focus has shifted towards refining statistical and linguistic models, exploring historical cipher analogues, and continuing interdisciplinary collaboration among historians, cryptographers, and linguists.

Future progress may arise from either technological innovation or human insight. It is conceivable that more advanced computational methods, perhaps even quantum-based algorithms, could one day unravel its cipher if the system proves finite and systematic. Equally, a moment of human recognition such as identifying a recurring proper noun or phrase might open the path to comprehension. Archival discoveries could also change everything; the recovery of a lost key, translation note, or letter referencing the manuscript would be invaluable. Until such breakthroughs occur, the Voynich Manuscript will continue to intrigue, challenge, and inspire.

In conclusion, after considering its history, content, theories, and scientific analyses, three facts stand firm: the Voynich Manuscript is authentic, unique, and unsolved. It is authentic as a genuine medieval creation, reflecting the intellectual and artistic milieu of early fifteenth-century Europe. It is unique, for no other document shares its script, language, or visual style. And it remains unsolved, standing as a monument to the limits of our understanding. The manuscript challenges modern assumptions about the scope of medieval knowledge and the reach of contemporary cryptography.

Even without a translation, the quest to understand the Voynich Manuscript has already yielded valuable outcomes. It has driven innovation in linguistic modelling and cryptanalysis, fostered collaboration across disciplines, and captured public imagination worldwide. Often called “the Everest of cryptology,” it continues to test the endurance, ingenuity, and curiosity of all who attempt to climb its steep symbolic slopes. Many have tried, none have yet reached the summit, but each effort brings us closer to understanding not only this enigmatic book, but also the enduring human desire to seek meaning in mystery. Until that summit is reached, the Voynich Manuscript will remain silent—its pages still guarding whatever knowledge, recipe, or riddle a medieval mind once set down in ink six centuries ago.