eduKateSG · VOYNICH RESEARCH LIBRARY · TANGENTIAL VOYNICH · RESET LENS
Tangential Voynich | Reset to Zero: The Alien Eye, the Period Eye and the Parallax Trap
Before we interpret an unread manuscript, we inherit the eyes that do the interpreting.
That sounds obvious until we notice how much arrives inside those eyes. A modern reader sees a book and expects text. We see spaces and expect words. We see plant-like forms and reach for botany. We see circles, stars and human figures and reach for astronomy or astrology. We see naked figures near pools or channels and reach for bathing, medicine, anatomy or reproduction. We see repeated marks and reach for alphabet, cipher, code, language or generated pseudo-text.
None of those moves is absurd. Most are reasonable. That is exactly why they are dangerous.
A reasonable first interpretation can become an invisible coordinate system. Once the coordinate system is installed, every later observation is measured from it. The manuscript may not move at all, but the apparent location of its meaning shifts because the observer has shifted.
This is the parallax problem.
Tangential Voynich has spent a long time deliberately placing Beinecke MS 408 inside systems that almost certainly have nothing to do with its historical function. The reason is not to make strange claims. The reason is to expose what our preferred model makes invisible. A railway controller asks different questions from a botanist. A compiler engineer notices different boundaries from a medievalist. A warehouse designer notices retrieval and grouping. A choreographer notices reachable transitions. A network engineer notices routing, framing and state.
Then every borrowed world is removed.
The surviving measurements return to the manuscript.
This article attempts a more severe reset. We will use two observers that pull in opposite directions.
- The Alien Eye knows nothing about terrestrial categories and therefore cannot call anything a word, plant, star, zodiac, bath, recipe, book, person, medicine or language until evidence earns the category.
- The Period Eye knows too much about one historical world: the visual habits, fashionable objects, manuscript forms, clothing, diagrams, medical and astrological conventions, devotional books, practical compilations and artistic tastes available around roughly 1350–1450.
One observer removes culture. The other restores historical culture.
If a feature survives both, it deserves our attention.
Voynich Research Library · Tangential Voynich: The Wrong-System Protocol · The Observer Is Part of the Instrument · The Null Model Is Not Neutral
Research Contract: This Is Not an Alien Theory
The first firewall must be explicit.
This article does not claim that the Voynich Manuscript was made by aliens, describes aliens, records alien knowledge, contains extraterrestrial language, or requires any extraterrestrial explanation.
The alien is a measuring instrument.
It is a thought experiment designed to destroy inherited categories. The alien observer has no idea what a European manuscript is supposed to be. It does not know that ink marks arranged in horizontal lines are probably writing. It does not know that a green form with roots and leaves is probably intended to evoke a plant. It does not know that twelve repeated radial divisions may invite a zodiacal interpretation. It does not know that a naked human-like figure is naked, because clothing itself is a cultural category it has not yet learned.
That ignorance is artificial. It is also useful.
The second firewall is temporal.
The often-cited radiocarbon work on four parchment samples places the parchment in the early fifteenth century; a commonly reported combined interval is 1404–1438 at about 95% probability. That dates the animal skin from which sampled parchment was made. It does not directly date the application of every ink stroke, establish authorship, establish place of manufacture, establish assembly sequence, or prove that all later additions are contemporary. The dating discussion itself explicitly warns that parchment creation and writing are different clocks. See the radiocarbon dating summary and documentation.
For this article, 1350–1450 is therefore a comparison window, not a provenance claim. It gives us enough temporal width to look backward and forward around the parchment horizon without pretending that a single year, city or court has been established.
The third firewall is historical.
There was no single medieval eye. A Parisian court painter, a physician in Germany, a university student in Padua, a monastic copyist, a merchant, an apothecary, a professional scribe, a noble patron and an ordinary urban reader did not inhabit identical visual worlds. “The period eye” in this article is therefore not a claim about one medieval mentality. It is a family of historically constrained observer models.
The purpose is comparison.
Not conclusion.
Why Reset to Zero?
Voynich research has an unusual difficulty. The manuscript is extremely easy to recognise and extremely hard to identify.
Its pages are visually memorable. Its script looks organised. Its illustrations look referential. Its sections look different from one another. Its repeated forms invite counting. Its diagrams invite geometry. Its human figures invite narrative. Its plants invite identification. Its codex form invites the assumption that the object is meant to be read as a book.
Every one of those invitations may contain signal.
Every one may also contain observer bias.
Our own Voynich work has repeatedly encountered the same warning: physical, scribal, textual, visual, geometrical and distributional divisions do not necessarily collapse into one clean semantic map. A page can belong to one visual family, another scribal family, another textual distribution and another codicological unit at the same time. The fact that two boundaries coincide does not prove that they represent the same hidden concept.
That is why “start again from zero” does not mean forgetting evidence.
It means separating evidence from the names we have attached to it.
Zero is not ignorance. Zero is the point at which observation is preserved but interpretation loses its automatic privilege.
The Parallax Trap
Parallax is normally a shift in the apparent position of an object caused by a change in the observer’s position.
For Voynich, the object is the same manuscript. The observer moves.
- A cryptographer sees symbol distributions, substitution possibilities, homophones, transpositions and constraints.
- A linguist sees positional regularities, token morphology, syntax-like recurrence and distributional structure.
- A botanist sees leaves, roots, flowers, graft-like junctions and impossible plant composites.
- A historian of medicine sees materia medica, regimens, bathing, pharmacology and practical compilations.
- An art historian sees line, pigment, costume, iconography, workshop convention and stylistic borrowing.
- A codicologist sees quires, bifolios, foldouts, ruling, sewing, rebinding, missing leaves and production sequence.
- A statistician sees sampling frames, null models, clustering, dependence, multiple comparisons and overfitting.
- A modern software engineer sees tokens, states, parsers, protocols and representations.
None of these observers is wrong merely for having expertise. Expertise is useful precisely because it notices structure.
The problem begins when the noticed structure quietly becomes the ontology of the manuscript.
A plant expert may correctly see that an illustration contains botanical-looking components and still be wrong that the page’s primary function is botanical identification. A cryptographer may correctly detect non-random positional constraints and still be wrong that the text is a cipher. A medievalist may correctly identify familiar visual conventions and still be wrong that the manuscript belongs to the expected genre.
Tangential Voynich therefore treats observer frames as controlled perturbations.
Move the observer.
Watch what seems to move.
Keep what does not.
Camera One: The Alien Eye
Imagine a competent observer arriving from somewhere that has never seen Earth.
It understands measurement, recurrence, topology, geometry, physical materials, information, uncertainty and causality. But it has no terrestrial cultural dictionary.
Place a high-resolution Voynich folio in front of it.
What can it say?
Not “plant.”
It can say that a central coloured form exhibits branching, repeated subforms, bilateral or approximate symmetry, terminal elements, a lower branching region and an upper branching region. It can record whether adjacent mark clusters occur near the form, whether those clusters occupy consistent positions and whether similar forms recur elsewhere.
Not “word.”
It can say that small mark sequences recur, that gaps partition sequences, that line beginnings and endings may differ in distribution, that some marks prefer particular positions and that sequence families vary across regions of the object.
Not “zodiac.”
It can say that a circular structure is divided radially, that repeated human-like forms occur around an annulus, that star-like or emblem-like marks recur, and that labels or short mark strings are spatially associated with repeated figures.
Not “bath.”
It can say that multiple human-like forms are spatially associated with bounded regions, channels, containers or flowing line structures, and that the page topology differs from the plant-dominant pages.
Not even “book,” initially.
It can say that flexible animal-skin sheets have been marked, folded, nested, bound and later handled as a sequential object. The codex category can be earned from construction evidence rather than assumed from appearance.
What the Alien Is Allowed to Measure
The alien observer is not stupid. It is simply denied semantic shortcuts.
- Recurrence: which forms repeat and at what scales?
- Position: which forms prefer beginnings, endings, margins, centres, rings, labels or particular page regions?
- Boundary behaviour: what changes at line, paragraph, page, diagram, section, bifolio or quire transitions?
- Co-occurrence: which forms tend to appear together?
- Exclusion: which combinations are unexpectedly absent?
- Directionality: do transitions behave differently in one direction than another?
- Geometry: what is radial, nested, connected, aligned, repeated or transformed?
- Topology: what is inside, outside, connected, isolated, bridged or enclosed?
- Scale: does a pattern exist at glyph, sequence, line, page, bifolio or codex level?
- Production trace: which features could arise from hand movement, layout planning, copying, ruling, binding, painting sequence or later intervention?
- Representation: which observed categories are robust to alternative segmentation of marks, spaces, images and page regions?
- Uncertainty: which distinctions are visible and which are imposed by the annotator?
This sounds austere.
It is.
But austerity creates a clean baseline.
If the alien independently recovers stable page families, repeated boundary states, radial organisation, position-sensitive mark classes and consistent associations between image regions and short mark strings, those are stronger observations than “this looks herbal” or “this looks astrological.” They are not more meaningful. They are less semantically loaded.
The Alien Eye Also Has a Failure Mode
Removing culture can remove information.
A wedding ring is not understandable from its geometry alone. A traffic sign cannot be fully inferred from pigment and shape. Money cannot be reconstructed from paper chemistry. A map’s function cannot be derived from line topology if the coordinate convention is unknown. A medical symbol can look like decoration to an observer who lacks its tradition.
So the alien eye is excellent at preventing premature semantic commitment and poor at recovering culturally encoded meaning.
That is why we need the opposite camera.
Camera Two: The Period Eye
Now reverse the experiment.
Instead of an observer with no terrestrial priors, imagine observers living inside the broad European manuscript world around 1350–1450.
The question changes from:
What can be seen without culture?
to:
What would have been visually familiar, fashionable, prestigious, ordinary, technical, old-fashioned, regional, devotional, medicinal or strange to people near the manuscript’s material horizon?
This is not about dressing the Voynich in medieval costume and declaring the mystery solved.
It is about rebuilding the comparison set.
What Was Fashionable Around 1400?
“Fashion” should be understood broadly.
Clothing changed. Artistic styles changed. Book ownership changed. Patronage changed. Manuscript illumination changed. Practical scientific and medical compilations continued to circulate. Astrology and medicine overlapped. Old visual traditions were copied while new naturalisms emerged. Courtly luxury existed beside utilitarian working books.
There is no single fashionable 1420.
There are multiple overlapping fashions.
1. International Gothic and the Courtly Book
The Getty’s survey of courtly manuscripts from 1380–1450 describes an international exchange among princely courts including Paris, Prague and Milan. Artists and books moved. Styles travelled. The resulting International Gothic aesthetic favoured graceful figures, fashionable court dress, patterned surfaces, bright colour and refined display.
The Metropolitan Museum’s Belles Heures, made around 1405–1408/9 for Jean de France, duc de Berry, shows what the upper end of manuscript splendour could look like: a richly illuminated object designed for a royal patron and private devotion.
That does not make luxury illumination the norm. It makes it a reference point.
If a theory places Voynich inside elite courtly production, then the absence or presence of features associated with elite display becomes evidence to examine rather than atmosphere to borrow.
2. Clothing Was Visually Loud
Around the turn of the fifteenth century, elite clothing in parts of western Europe could be spectacular. The Morgan Library’s Illuminating Fashion material describes the houppelande with large sleeves, costly textiles, elaborate headwear, belts, fur and ornament as conspicuous markers of late-fourteenth- and early-fifteenth-century fashion. By the 1420s and 1430s, styles shifted, simplified or changed silhouette while Burgundian influence remained powerful.
Why should Voynich research care about sleeves and hats?
Because depicted clothing is a time-sensitive interface between image and social world. If a manuscript repeatedly depicts contemporary people in social scenes, fashionable clothing can become useful comparative evidence. If it instead depicts nude, minimally dressed, schematic or timeless figures, that absence is also informative—but only after we compare like with like. Technical diagrams, zodiacal personifications, saints, allegories and medical figures may follow conventions different from court scenes.
The period-eye test therefore asks not “does Voynich clothing look like 1420?” but “what kind of scene would have been expected to display current dress, and what kind would not?”
3. Books of Hours Were Everywhere in Elite and Aspirational Manuscript Culture
The Met describes the Book of Hours as the most popular book form of its time. These private devotional books were customizable, locally inflected and often illuminated. A surviving Book of Hours made around 1410 for the English market shows how manuscripts could be expanded and supplemented later.
This matters even if Voynich is not devotional.
Popular book forms shape the visual literacy of their period. Readers become accustomed to calendars, repeated liturgical structures, image-text relationships, decorated initials, cycles, personal additions and portable private books. A period observer does not approach every codex as a modern printed monograph with one author, one genre, one stable edition and one intended reading path.
4. Medical, Pharmacological and Astrological Knowledge Could Share a Manuscript World
Modern academic departments separate medicine, botany, astronomy, astrology, pharmacy, religion and natural philosophy. Medieval manuscripts did not always respect our departmental boundaries.
Wellcome’s MS.105, copied in Germany before 1435, combines medical and pharmacological treatises, rules for health, food and drink, fevers, diseases, plant- and animal-derived drugs, Avicennian material and regimen traditions. Wellcome MS.8515, dated around 1425 and associated with southern England, is described as a practical manual containing calendrical computation, astronomy, astrology and later medical recipes.
British Library manuscript catalogues likewise preserve fifteenth-century compilations in which medicine and astrology coexist. Harley MS 2375 includes medical and astrological texts and an astrological diagram. Harley MS 1736 includes surgery, recipes and a medical-astrological text in Middle English.
The lesson is not “Voynich is a medical-astrological manual.”
The lesson is that our modern genre boundaries are themselves a possible parallax error.
5. Plant Images Did Not Need to Behave Like Modern Botanical Plates
Herbal illustration carried long copying traditions. Wellcome’s discussion of medieval plant portraits notes that patterns of illustration could be copied across manuscripts for centuries and that images might balance recognition, instruction and decoration rather than modern botanical naturalism.
The northern Italian Tractatus de herbis tradition, represented by a manuscript produced perhaps around 1440, shows a world in which medicinal plants, labels, inherited forms and manuscript illustration were already deeply developed.
This gives the period eye an important correction.
A plant-like image that is not photographically accurate is not automatically evidence of fantasy, incompetence or encryption. But neither is it automatically a herbal illustration. The right comparison is against historical representation practices, including copied, schematic, composite and decorative tendencies.
6. Astronomy and Astrology Were Not Separate Modern Departments
A useful overview from the Getty-linked Smarthistory essay Written in the Stars explains how astronomy and astrology were intertwined with education, medicine and broader medieval understandings of the cosmos. Zodiacal signs, planetary relations, calendars and medical practice could inhabit overlapping intellectual spaces.
Again, that does not identify Voynich’s circular diagrams.
It changes what counts as a plausible period comparison.
The Period Eye Is Not the Same as the Fashion Eye
There is another trap here.
When we search for “what 1400 looked like,” the surviving objects most likely to be photographed, digitised, exhibited and celebrated are often the spectacular ones.
Royal books. Gold. Famous painters. Cathedral commissions. Luxury costume. Masterpieces.
That can create a museum parallax.
Most working knowledge was not necessarily packaged like the Belles Heures. A practical medical compilation, university text, recipe collection, astrological manual, apothecary reference or personal working book could be visually plainer, composite, heavily used, amended or organised for function rather than spectacle.
So our comparator library must deliberately include at least four strata:
- Elite / courtly: expensive objects designed to display patronage, devotion, status and artistic refinement.
- Professional / technical: books for medicine, astronomy, law, university study, pharmacology, computation or skilled practice.
- Workaday / composite: compilations, recipe books, miscellanies, added notes and evolving practical objects.
- Transitional / hybrid: manuscripts that combine older inherited forms with emerging fifteenth-century naturalism, humanism, regional style or new modes of book production.
If Voynich is compared only with masterpieces, it may look crude.
If compared only with technical books, it may look decorative.
If compared only with herbals, it may look astronomical.
If compared only with astrological manuals, it may look botanical.
The comparison set creates the anomaly.
That is exactly what The Null Model Is Not Neutral warned about.
Three Eyes on the Same Folio
We can now construct a controlled comparison.
| Feature | Modern Eye | Alien Eye | Period Eye | Possible Neutral Residue |
|---|---|---|---|---|
| Plant-like page | Botanical illustration, herbal, encoded plant | Branched coloured structure associated with repeated mark sequences | Could be compared with herbals, materia medica, plant portraits, practical medical books | Stable image-text association; repeated morphology; page family |
| Radial circular page | Zodiac, astrology, astronomy | Nested/radial partition with repeated figures and labels | Could be compared with calendars, zodiac imagery, astronomical/astrological diagrams | Radial classification; repeated slots; figure-label mapping |
| Human figures in bounded shapes | Bathing, medicine, anatomy, reproduction | Human-like units interacting with containers, channels and boundaries | Could be compared with balneological, medical, allegorical, devotional or diagrammatic traditions | Human-unit topology; containment; flow; repeated scene grammar |
| Running script | Language, cipher, code, pseudo-text | Repeated mark sequences with gaps and positional constraints | Could be compared with scribal scripts, abbreviations, technical notation and manuscript-specific conventions without assuming equivalence | Sequence regularity; positional classes; boundary effects |
| Foldout | Large diagram or map | Expanded physical surface enabling a larger connected representation | Could be compared with tables, diagrams, genealogies, astronomical schemes, maps or complex visual summaries | Scale expansion; cross-panel continuity; special handling state |
The key is the final column.
The neutral residue is not “the truth.”
It is what remains after the observer-specific story has been subtracted.
What the Alien Eye Sees That the Period Eye May Miss
A historically informed observer can be too good at recognition.
Show a medievalist a circular arrangement with twelve-ish divisions and familiar iconographic cues and a historical category may appear immediately. That is valuable, but it can also suppress alternative structural descriptions.
The alien asks rude questions.
- Are the divisions actually equal?
- Are there always the same number of repeated units?
- Does label placement follow geometry or content?
- Are repeated figures truly repeated, or are we grouping visually different objects under one human category?
- Does the central object control the surrounding arrangement?
- Which features survive if the page is rotated?
- Which relations survive if colour is removed?
- Which relations survive if labels are masked?
- Is the diagram still recognisably the same family under those perturbations?
Those questions can break a familiar iconographic identification without proving an alternative.
That is useful.
What the Period Eye Sees That the Alien Eye Cannot
The period observer asks a different set of rude questions.
- Would this kind of diagram have been culturally legible?
- Would a contemporary reader expect image and text to interact in this way?
- Is the object mixing domains that modern scholars separate but period manuscripts often combined?
- Does the illustration inherit a conventional form that should not be judged by modern naturalism?
- Would contemporary costume normally appear in this scene type?
- Is the absence of gold, framing, decorated initials or polished figural modelling meaningful for the proposed social setting?
- Does the object look like a presentation manuscript, a working manual, a compilation, a private book, a workshop object or something outside those comparators?
- Which visual features were fashionable, which were ordinary, which were old-fashioned, and which were unusual in different regions?
The alien can tell us that two symbols differ.
The period eye may tell us that the difference mattered to a fifteenth-century maker.
Neither is sufficient alone.
The Reset-to-Zero Protocol
The mechanism can now be made operational.
Step 0 — Freeze What Is Actually Known
Resetting does not permit us to erase material facts merely because they are inconvenient.
- The object exists and is held as Beinecke MS 408.
- It is a codex made from parchment with ink and pigments.
- Its script remains unidentified and not convincingly read.
- Its illustrations include recurrent plant-like, circular, human-figure and other visual families.
- The manuscript has foldouts, complex codicology, missing leaves and later custody history.
- Scientific testing constrains aspects of the material history.
- Radiocarbon evidence dates sampled parchment, not every act of inscription or later intervention.
The Beinecke Library provides high-resolution access and collection context, including links to the manuscript images and material analysis.
Step 1 — Strip the Semantic Labels
Temporarily replace “word,” “plant,” “zodiac,” “bath,” “recipe,” “astronomy,” “medicine” and “language” with neutral descriptions.
This is not because the traditional labels are foolish. It is because we need to know which observations depend on them.
Step 2 — Atomise the Evidence
Break a striking impression into small observations.
“It looks like a medicinal plant” becomes:
- one central coloured form;
- lower branching structure;
- upper repeated lateral elements;
- terminal structures;
- several colour regions;
- adjacent short mark sequences;
- longer mark sequences in a neighbouring region;
- similar page architecture elsewhere.
Now several hypotheses can compete for the same evidence.
Step 3 — Build a Lookalike Set
Do not ask for the favourite explanation.
Ask which alternatives could generate the same surface.
A plant-like image might be literal botanical representation, copied herbal convention, mnemonic composite, symbolic diagram, decorative pseudo-botany, a transformed technical object or something else. The point is not to make the list infinite. The point is to prevent the first plausible category from becoming the only candidate.
Step 4 — Seek the Discriminator
The next useful question is often not “what evidence supports my theory?”
It is:
What observation would separate the two strongest lookalikes?
If two models both explain the page, evidence shared by both cannot choose between them.
Find the feature on which they disagree.
Step 5 — Run the Alien Pass
Remove historical names and test the geometry, recurrence, boundaries, exclusions, production traces and positional behaviour.
Any claim that disappears completely when its semantic vocabulary is removed is lens-sensitive.
Step 6 — Run the Period Pass
Now compare the same neutral features against a historically stratified corpus from roughly 1350–1450.
Not one famous manuscript.
A corpus.
Include elite, technical, medical, astrological, herbal, devotional, university, practical and composite objects where possible. Record geography, date confidence, social function, material quality and whether the manuscript is a presentation object or a working object.
Step 7 — Intersect the Residues
Features that survive both passes become stronger candidates for genuine structure.
For example:
- The alien pass may recover a stable radial partition.
- The period pass may show that radial zodiacal or calendrical organisation was culturally available.
- The intersection supports “deliberate radial classification in a historically plausible visual vocabulary.”
It still does not prove astrology.
That final distinction matters.
Step 8 — Use Negative Evidence
A good period model must predict absences as well as similarities.
If a proposed elite-courtly origin should normally produce expensive pigments, extensive gilding, fashionable dress, high-finish figural work or heraldic display, and those are weak or absent, the model must explain why.
If a proposed technical manual should normally include tables, numerals, recognisable labels, conventional diagrams or explicit organisational markers, and those are missing, that matters too.
Absence is not automatically decisive. Damage, copying, regional style, workshop stage, private notation and incomplete preservation can all complicate it.
But a model that can explain every presence and every absence after the fact is not being tested.
Step 9 — Allow HOLD
The process must be allowed to end without a winner.
“Undecidable with current evidence” is not a failed answer.
It is protection against forced certainty.
Step 10 — World Return
Delete the alien.
Delete the fashionable court.
Delete the modern specialist.
Return only the measurements and historically bounded constraints that survived.
Worked Example: The Plant Pages
The modern eye sees plants first.
That is reasonable. Many images have roots, stems, leaves or flowers.
The alien eye refuses the noun.
It records branching architecture, repeated leaf-like subforms, colour placement, root-like lower complexity, adjacent labels, page-level recurrence and variation between image families.
The period eye then asks whether comparable manuscripts represented medicinal or useful plants schematically, copied inherited forms, combined recognition with symbolic convention, or modified older prototypes.
Now the research question improves.
Instead of:
Which modern species is this?
we can ask:
- Which image properties were stable in contemporary herbal traditions?
- Which were free to vary under copying?
- Were roots or flowers exaggerated for identification, doctrine, memory or decoration?
- How often do labels occupy similar positions in comparator manuscripts?
- Do Voynich image families preserve a consistent grammar that is stronger than individual species resemblance?
- Are the most “impossible” forms actually more anomalous than comparable historical plant portraits?
A period-aware answer may reduce some of the apparent weirdness.
If it does, that weirdness was partly modern parallax.
If the forms remain highly anomalous even against broad historical comparators, the anomaly becomes more interesting.
Worked Example: The Circular Pages
The modern eye says zodiac almost immediately on some folios.
The period eye has strong reasons to consider that family of interpretations because zodiacal, calendrical and astrological imagery was culturally available and could connect to medicine and practical computation.
The alien eye does not care.
It asks how many sectors exist, how repeated figures are distributed, whether labels align to figures or sectors, whether radial spacing is regular, whether the central element changes the surrounding pattern, whether apparent sequence is directional and whether the same layout grammar recurs across folios.
The useful return may be weaker than “zodiac.”
But it may also be more defensible:
A family of radial classification diagrams associates repeated figures, short labels and central motifs in a way that is compatible with—but does not by itself prove—period astrological or calendrical visual practice.
That sentence leaves room for evidence to do more work later.
Worked Example: The Human Figures and Pools
Some Voynich illustrations are routinely described using balneological language because human figures appear in or around pool-like, vessel-like and channel-like structures.
The alien pass strips that away.
We now have repeated human-like units, bounded spaces, connected shapes, possible flow paths, colours, gestures and local labels.
The period pass then asks which visual traditions around 1350–1450 used human figures inside schematic systems: bathing, medicine, anatomy, cosmology, allegory, devotion, astrology, fertility, humoral theory, diagrams or other practices.
The crucial question becomes discriminative:
Which specific spatial relation would be expected under one period comparator and not another?
Without that discriminator, resemblance remains resemblance.
Worked Example: The Script
The script is where the reset becomes most uncomfortable.
Our modern reading machinery is powerful. We segment glyphs, tokens, lines and paragraphs almost automatically. But What If Voynich Is Not a Text? and The Inverse Representation Problem already showed why the representation itself must be tested.
The alien eye asks:
- What are the smallest recurring mark units?
- Are spaces reliable boundaries or merely visual gaps?
- Do mark sequences have stable internal positions?
- Does line position alter the distribution?
- Do visual page families alter mark distributions?
- Do certain marks occur only in limited structural states?
- Can alternative segmentation explain the same recurrence?
The period eye asks another set:
- What kinds of scripts, abbreviations, notational systems, multilingual practices and specialist marks existed in comparable manuscript environments?
- How much local convention could a professional or private manuscript tolerate?
- Which visual behaviours would be ordinary for writing, and which would be unusual?
- What external reading aids would a contemporary user normally need?
Neither pass is a decipherment.
Together they tighten the question before decipherment.
A Better Meaning of “Fashionable”
The most interesting use of fashion is not to hunt for a single matching sleeve.
It is to ask what a culture was selecting for.
Around 1400, different manuscript environments selected for different things.
- Courtly display selected for refinement, luxury, identifiable patronage and fashionable visual culture.
- Private devotion selected for portable cycles, repeated prayer structures, personal use and image-assisted contemplation.
- Medical practice selected for useful classification, recipes, regimen, substances, diagnosis and treatment traditions.
- Astrological practice selected for calendrical, planetary, zodiacal and prognostic organisation.
- University and professional books selected for dense transmission, commentary, reference and teaching.
- Working compilations selected for usefulness over purity of genre.
That is a more powerful period eye than “people wore houppelandes.”
Clothing is one visible marker.
The deeper fashion is what kinds of objects a society valued, copied, commissioned, carried, consulted and considered legible.
The Absence Test
Historical comparison becomes much stronger when we ask what is missing.
Suppose Voynich were proposed as a courtly luxury object.
Then we should compare it with courtly manuscripts not only for similarities but for expected prestige signals: materials, gilding, framing, costume, finish, heraldry, patron display, narrative polish and workshop execution. A mismatch does not kill the model automatically, but it creates explanatory debt.
Suppose it were proposed as a practical medical compilation.
Then we should look for the organisational behaviours of practical books: repeated entry structures, ingredients, measurements, cross-references, headings, indexes, diagrams, tables, section markers, recipe syntax or other recurrent interfaces. If these are absent or hidden, the theory must explain why.
Suppose it were proposed as an astrological manual.
Then the circular imagery is not enough. We would want structure that discriminates astrological organisation from generic radial design.
The absence test prevents historical familiarity from becoming historical proof.
The Parallax Matrix
A useful Voynich feature can be scored conceptually across four questions.
- Observer stability: does the feature survive when experts from different domains describe it?
- Representation stability: does it survive reasonable changes in segmentation, transcription, colour, rotation, scale or boundary definition?
- Historical compatibility: does it fit at least one well-supported period practice without requiring anachronism?
- Predictive survival: does the feature help distinguish competing models on new or held-out material?
A feature that survives all four is much more valuable than a feature that is merely evocative.
Call this the parallax rule:
The more an observation depends on where the observer stands, the less semantic weight it should carry until independent evidence stabilises it.
This does not eliminate interpretation.
It makes interpretation earn its weight.
The Most Important Possibility: Some Voynich Weirdness May Be Ours
There is a seductive assumption in mystery research:
If we find an object strange, the object must have been strange to its makers.
That does not follow.
Some things that look bizarre now may have been conventional then.
A copied plant form may look botanically incompetent to us and perfectly legible inside a manuscript tradition. An astrology-medicine linkage may look like category confusion to us and ordinary to a period practitioner. A composite manuscript may look like several unrelated genres to us because our libraries have separated those disciplines for centuries.
The reverse is also possible.
Some features that modern Voynich culture has normalised may have been deeply unusual in the fifteenth century.
We do not know which until comparison is systematic.
That is why the period eye matters.
And Some Historical Familiarity May Be Ours Too
There is an equal and opposite danger.
Once we learn that medieval manuscripts contained herbals, zodiac diagrams, medical recipes and strange marginal images, almost anything in Voynich can begin to feel “medieval.”
That is not enough.
A historically available motif is not evidence of a specific origin.
A historically available genre is not evidence of a specific function.
A fashionable style is not a signature.
A similar diagram is not identity.
Period plausibility is a filter.
Not a verdict.
The Alien and the Medieval Reader Meet
The strongest part of the experiment arrives when the two observers disagree.
The alien sees repeated radial slots.
The period reader says zodiac.
Good.
Now ask which measurable properties are specifically predicted by zodiacal organisation and which are merely radial.
The alien sees branching structures with labels.
The period reader says herbal.
Good.
Now ask which label-image relationships, ordering conventions or morphological priorities distinguish herbals from decorative plant imagery.
The alien sees human units in containers and channels.
The period reader offers bathing, medicine, astrology, allegory or another visual tradition.
Good.
Now the competing historical models must predict different structures.
The disagreement is not a problem.
The disagreement generates the next discriminating test.
A Future Comparator Test: 1350–1450 Without Voynich
The cleanest next experiment would begin without showing researchers the Voynich Manuscript at all.
Build a comparator corpus spanning roughly 1350–1450 with controlled representation of:
- herbals and materia medica;
- medical and surgical compilations;
- astrological and astronomical manuals;
- calendars and computistical books;
- Books of Hours and devotional manuscripts;
- university and professional texts;
- courtly presentation manuscripts;
- recipe and practical miscellanies;
- diagram-heavy scientific works;
- regional and lower-finish working manuscripts.
Then code the corpus before Voynich comparison.
- page architecture;
- image/text ratio;
- label placement;
- diagram topology;
- figure repetition;
- colour use;
- costume specificity;
- decoration level;
- entry structure;
- tables and numerals;
- plant representation style;
- marginal versus central imagery;
- foldout use;
- genre mixing;
- later additions and corrections.
Freeze the codebook.
Only then bring Voynich back into the room.
This reverses a common research habit. Instead of searching history for things that look like Voynich, we first map what the period actually contains and then ask where Voynich falls inside or outside that space.
That is a much harder test.
It is also much safer.
A Future Blind Observer Test
Another experiment follows directly from the alien lens.
Give masked page regions to several observer groups without telling them the preferred section names.
- manuscript and art specialists;
- natural-form specialists;
- engineers or systems observers;
- unprimed general observers.
Require neutral coding first.
Only after the neutral features are frozen should historical context be revealed.
The question is not which profession is “right.”
The question is which features survive observer rotation.
If engineers, artists, botanists and unprimed observers all recover the same boundary or repeated component, that feature earns more confidence as an observed structure. If only one group sees it, it may still be real—but it is lens-dependent and should be treated accordingly.
What Resetting to Zero Does Not Permit
- It does not permit us to ignore radiocarbon evidence.
- It does not permit us to pretend the manuscript has no historical context.
- It does not permit us to call every visual resemblance meaningful.
- It does not permit anachronistic comparators to become provenance claims.
- It does not permit period plausibility to become proof.
- It does not permit modern categories to be banned permanently; they may return if evidence earns them.
- It does not permit statistical structure to be called language without an additional bridge.
- It does not permit iconography to be called semantic translation.
- It does not permit a good story to outrun the manuscript.
Zero is a calibration state.
Not a belief system.
What This Changes
The usual Voynich question is:
What is this?
The reset changes it into a sequence.
- What is directly observable?
- Which parts of that observation depend on our categories?
- Which alternative categories fit the same surface?
- What would distinguish the strongest alternatives?
- Which structures survive an observer with no culture?
- Which structures become ordinary or meaningful inside historically plausible comparison worlds?
- Which expected period features are missing?
- Which observations survive both passes?
- What new prediction follows?
- What evidence would defeat it?
That sequence is slower than guessing.
It is faster than spending ten years inside the wrong coordinate system.
The Strange Gift of the Alien
The alien does not solve Voynich.
It gives us something more modest and possibly more valuable.
It reminds us how much of the manuscript we have already named.
Words.
Plants.
Zodiac.
Baths.
Recipes.
Sections.
Language.
Those names may ultimately be correct.
But the alien asks us to earn them again.
The Strange Gift of the Period Eye
The period observer gives the opposite gift.
It reminds us that we are not the original receivers.
We do not share the maker’s visual habits.
We do not share the original reader’s genre expectations.
We do not share their boundaries between medicine and astrology, practical knowledge and symbolic knowledge, copied tradition and observation, private notation and public script, luxury and utility.
A manuscript can be strange to us because its meaning is hidden.
It can also be strange because the receiver who once knew how to see it is gone.
That is not mysticism.
It is an interface problem across six centuries.
World Return
Now remove the alien.
Remove the houppelande.
Remove the Burgundian court.
Remove the herbal.
Remove the zodiac.
Remove the bath.
Remove the cipher.
Remove every borrowed answer.
What remains?
- a materially constrained historical object;
- multiple overlapping visual and textual provinces;
- repeated marks with strong positional and distributional behaviour;
- recurrent image families;
- radial, nested and connected geometries;
- image-label relationships;
- production, binding and sequence questions;
- a missing original receiver;
- a missing cultural key;
- and an observer—us—whose own categories are part of the measurement.
That is not zero.
That is a better starting point.
The safest Voynich theory may begin by asking which parts of the mystery disappear when we change the eyes looking at it.
And then, only after the parallax has been measured, we move forward again.
Sources and Comparison Rooms
- Yale Beinecke — The Beinecke Cipher (Voynich) Manuscript
- Voynich.nu — Radiocarbon dating documentation and explanation
- J. Paul Getty Museum — Fit for a King: Courtly Manuscripts, 1380–1450
- The Metropolitan Museum of Art — The Art of Illumination: The Belles Heures
- Morgan Library & Museum — Illuminating Fashion: Luxury in a Time of Madness
- Wellcome Collection — MS.105, medical and pharmacological treatises
- Wellcome Collection — c.1425 calendar, astronomy and astrology manual
- Wellcome Collection — Plant portraits and medieval herbal illustration
- British Library — Harley MS 2375, medical and astrological compilation
- Smarthistory / Getty context — Astronomy and Astrology in Medieval Manuscripts
Return to the Voynich Research Library →
Tangential Voynich: The Wrong-System Protocol →
Tangential Voynich: The Observer Is Part of the Instrument →