The Voynich Manuscript contains one of the most useful kinds of disorder.
A disorder with a possible before and after.
The pages usually called pharmaceutical show rows of plant fragments, short labels and elaborate container-like forms. At first glance they seem like one coherent visual programme.
Physically, they are not together.
One bifolium lives in Quire 15.
Two live later in Quire 19.
Herbal material intervenes.
Then another clue appears.
The containers in Quire 19 begin relatively simple and cylindrical. Across those pages, some forms gradually become more elaborate. The elaborate style then continues naturally into the pharmaceutical bifolium now sitting earlier in Quire 15.
That has led researchers to propose that the visual sequence once ran differently from the present binding.
The “pharmaceutical section” may preserve one of the clearest places where drawing style remembers an order the binding no longer does.
Quick Read
- The pharmaceutical-looking material occupies three bifolia rather than one continuous present-day gathering.
- One bifolium is in Quire 15 around folios 88–89.
- Two bifolia are in Quire 19 around folios 99–102.
- The Quire 19 material tends to begin with simpler, largely cylindrical containers.
- Container forms become more elaborate toward the end of the Quire 19 sequence.
- The elaborate vessel style continues naturally on the Quire 15 pharmaceutical bifolium.
- René Zandbergen and other codicological researchers therefore treat the likely visual sequence as Q19 material before the Q15 bifolium.
- A commonly proposed order is: Q19 inner bifolium (f100/f101) → Q19 outer bifolium (f99/f102) → Q15 pharmaceutical bifolium (f88/f89).
- Under this model, the two Q19 bifolia would not remain nested in exactly their current relation.
- This is a reconstruction from visual and physical evidence, not an established semantic reading order.
- Container-style progression can suggest sequence without telling us what the containers mean.
- The problem demonstrates why present quire number, current binding order and original conceptual order must be kept separate.
What Counts as “Pharmaceutical” Here?
The name is conventional.
These pages contain:
- small detached plant-like fragments;
- rows of those fragments;
- short Voynich labels;
- larger paragraph text;
- many ornate vessel- or container-like drawings.
Those features make comparison with historical pharmacy reasonable.
But the order problem does not require the word “pharmaceutical” to be literally correct.
We could call them:
the pages with fragment rows and large containers.
The same physical sequencing problem would remain.
Three Bifolia That Look Like One Programme
The surviving visual family is distributed across three physical bifolia.
That matters because a bifolium is not an arbitrary pair of pages.
It is one folded sheet.
If two pharmaceutical-looking surfaces belong to the same bifolium, their physical relationship is stronger than a similarity between unrelated folios.
The three relevant units are usually described as:
- the pharmaceutical bifolium in Quire 15 — f88/f89;
- the outer pharmaceutical bifolium of Quire 19 — f99/f102;
- the inner pharmaceutical bifolium of Quire 19 — f100/f101.
In present binding order, they are separated.
Visually, they look related enough to invite reconstruction.
Why the Containers Matter More Than the Plants for Ordering
Plant fragments are inherently repetitive.
Roots can resemble roots.
Leaves can resemble leaves.
A container style can preserve a different kind of continuity.
Shape complexity.
Neck design.
Body structure.
Tiering.
Decorative elaboration.
Across the Quire 19 pharmaceutical pages, researchers have noted a broad progression from simpler cylindrical vessels toward increasingly elaborate forms.
Then the elaborate visual vocabulary continues on the Quire 15 bifolium.
That makes container style a candidate sequencing variable.
The Proposed Sequence
René Zandbergen’s current illustration synthesis gives the proposed order as:
- inner bifolium of Quire 19 — f100/f101;
- outer bifolium of Quire 19 — f99/f102;
- pharmaceutical bifolium of Quire 15 — f88/f89.
Under this reconstruction, the two Quire 19 bifolia would not remain nested in the way their current binding presents them.
The proposal is driven by the style sequence of the containers rather than by deciphered text.
That distinction is central.
We are reconstructing a plausible production/viewing order of drawings, not translating a procedural sequence.
Simple Containers First
The earlier part of the proposed Q19 sequence contains many relatively simple, largely cylindrical forms.
Some look like open vessels.
Some may have lids.
Some resemble wooden, ceramic or composite containers only in broad visual terms.
The material identity is not established.
The important ordering feature is the drawing vocabulary.
These forms are generally less architecturally elaborate than those encountered later in the proposed sequence.
Then the Containers Become More Elaborate
Moving through the Quire 19 material, the container drawings increasingly acquire more complex profiles.
More tiers.
More ornate necks.
More decorative segmentation.
More structure that looks almost architectural.
By the time the visual sequence reaches the final part of Q19, the style begins to look much closer to the elaborate vessels of f88/f89.
That continuity is the central clue.
Then Quire 15 Looks Less Like an Earlier Section
Under current quire numbering, the f88/f89 pharmaceutical bifolium appears earlier than the Q19 material.
Under the vessel-style sequence, it looks later.
This is the mismatch.
Present physical address says:
Q15 before Q19.
Container progression suggests:
Q19 pharmaceutical material before Q15 pharmaceutical material.
The same manuscript therefore carries two possible order signals.
One is later codex organisation.
One may preserve earlier design continuity.
Why This Does Not Mean the Containers Tell a Story
A visual progression can result from many processes.
- The illustrator becomes more elaborate over time.
- A source exemplar changes style across a sequence.
- Different container classes are intentionally ordered from simple to complex.
- Different production stages use different drawing conventions.
- Several hands or copying sessions are involved.
- The apparent progression is partly subjective.
Therefore “simple → elaborate” does not itself tell us:
- wood → ceramic;
- raw ingredient → finished medicine;
- small dose → large dose;
- cheap vessel → expensive vessel;
- early process → late process.
Those are semantic stories.
The ordering evidence is weaker and more defensible:
the drawing style appears to have a continuity that current quire placement interrupts.
Why Current Quire Order Cannot Automatically Win
The Two Numbering Systems article already established that quire marks record an organised manuscript state, not necessarily original intellectual order.
Other evidence shows that at least some bifolia had likely already been disturbed before the quire-numbering sequence stabilised.
Therefore “Q15 is numbered before Q19” cannot end the discussion.
It tells us how the gatherings were ordered when the numbering system was applied.
If the pharmaceutical bifolia had already been separated, the quire numbers would simply preserve the separated state.
The Herbal Interruption Matters
In the present codex, herbal-looking folios occur between the pharmaceutical bifolia.
That is one reason the visual family feels interrupted.
But thematic interruption alone is not sufficient to prove misbinding.
A deliberately compiled manuscript can switch subjects.
The stronger point is cumulative:
- three bifolia share a distinctive fragment+container page architecture;
- they are physically split;
- container style appears to evolve across them in an order different from quire numbering;
- the manuscript elsewhere contains independent evidence of disturbed order.
Together, those facts make reordering a serious hypothesis.
Why the Two Q19 Bifolia May Not Belong Nested as They Are
The proposed sequence does something more radical than merely moving Q15 after Q19.
It also changes the relation between the two Q19 bifolia.
Instead of treating them simply as nested sheets in current page order, the visual sequence favours placing one after the other.
This matters because bifolium nesting determines which pages become neighbours.
A visual progression that crosses current nesting awkwardly may indicate:
- misnesting;
- a different original gathering architecture;
- copying order independent of final binding order;
- or a false impression of progression.
All four should remain visible until independent evidence discriminates among them.
Production Order and Reading Order May Still Differ
Suppose the vessel drawings were actually made in the sequence Q19-inner → Q19-outer → Q15.
Does that prove readers were meant to read the pages in that order?
No.
A workshop can produce components in one order and assemble them in another.
An exemplar can be copied sheet by sheet, then gathered according to a different functional plan.
Therefore the order vocabulary must remain precise:
- drawing order — sequence in which visual forms were produced;
- bifolium production order — sequence in which sheets were completed;
- binding order — sequence in which sheets were gathered;
- reading order — sequence intended for consultation;
- current folio order — sequence the modern codex presents.
One hypothesis can fit one of these without proving the others.
Could Container Complexity Be a Scribal or Illustrator Drift?
Yes.
An illustrator can change style over time.
They become more confident.
Or more hurried.
They move from one exemplar to another.
They change tools.
If the container progression is production drift, it could become a chronological marker independent of semantics.
A serious test would compare other drawing features across the same bifolia:
- line quality;
- perspective attempts;
- symmetry;
- neck construction;
- decoration density;
- plant-fragment style;
- label placement;
- paint usage.
If multiple independent visual traits drift together, the order hypothesis strengthens.
Could Several Illustrators Explain the Sequence Instead?
Also possible.
Simple vessels could belong to one illustrator.
Elaborate vessels to another.
That would make stylistic clustering a hand/class problem rather than a temporal progression.
To distinguish them, compare style boundaries with:
- scribal hand assignments;
- Currier regime;
- ink behaviour;
- plant-fragment drawing style;
- physical bifolia;
- text layout.
If container style changes while all other production variables remain stable, a functional sequence becomes more plausible.
If many production features change together, different hands or stages become more plausible.
The Plant Fragments Can Provide an Independent Test
The vessel sequence should not be allowed to test itself.
Plant fragments offer another data layer.
Ask whether fragment organisation changes in the same proposed order.
- Do fragment counts per row change?
- Do root-heavy versus leaf-heavy rows shift?
- Do label patterns change?
- Does drawing scale change?
- Do whole-plant correspondences cluster by proposed phase?
If an independent plant-fragment signal agrees with the vessel progression, confidence rises.
If it conflicts, the simple→elaborate vessel narrative may be only one stylistic dimension rather than a true sequence.
The Text Can Test the Visual Order Too
The same logic applies to Voynichese.
Take the proposed order:
Q19 inner → Q19 outer → Q15.
Then ask whether textual features drift in the same direction.
- token-family frequencies;
- label vocabulary;
- line-position behaviour;
- rare glyphs;
- paragraph architecture;
- Currier-like feature scores.
If the text also forms a smoother sequence under the visual order than under current binding order, the reconstruction becomes far more interesting.
If not, that difference may tell us production of image and text followed different histories.
This Is a Perfect Place for Held-Out Testing
Container complexity can be judged subjectively if the researcher already knows the desired order.
A stronger method would define features first.
- number of body tiers;
- number of neck constrictions;
- symmetry;
- decorative bands;
- lid-like features;
- body width/height ratios;
- presence of feet or base elements.
Then score containers without using folio order.
Only after scoring should the inferred progression be compared with candidate bifolium orders.
This turns “it looks smoother to me” into a falsifiable ordering problem.
The Section Appears Unfinished
René Zandbergen describes the pharmaceutical material as appearing unfinished.
That observation fits several possible histories.
- The project stopped before a planned gathering was completed.
- Missing bifolia once completed the sequence.
- The pages were working material never fully assembled.
- The present split reflects incomplete or disrupted production.
Again, “unfinished” should not become permission to imagine missing content.
It should become a question about observable completion states.
Why This Problem Is Bigger Than Pharmacy
The pharmaceutical order problem is a model for reconstructing the whole manuscript.
When present binding and internal sequence disagree, use several independent signals:
- physical sheet relationships;
- numbering layers;
- drawing style;
- text statistics;
- scribal hands;
- paint transfer;
- page architecture;
- continuity across folds.
The goal is not to reorder pages until the story feels nice.
It is to find the order under which independent evidence disagrees least.
What the Pharmaceutical Order Problem Does Not Prove
- It does not prove the pages are pharmaceutical in function.
- It does not prove the containers are literal pharmacy vessels.
- It does not prove simple containers are chronologically earlier in absolute time.
- It does not prove the proposed order was the intended reading order.
- It does not prove one illustrator produced all vessel drawings.
- It does not prove Q15/Q19 were originally one single quire.
- It does not translate the plant labels or prose.
- It does not establish why the bifolia became separated.
What It Does Give Us
- A coherent visual family split across two present quires.
- A candidate container-style progression that conflicts with quire order.
- A concrete test of whether drawing style can recover older bifolium sequence.
- A way to combine codicology, iconography and text statistics without decoding first.
- Another independent reason not to equate current binding order with original design order.
Primary School: Mix Three Pages of a Comic
Draw three machines.
Make each one slightly more complicated than the previous machine.
Shuffle the pages.
Ask a child to reconstruct the likely drawing sequence from complexity alone.
Then ask:
Does the smoothest visual sequence prove that was the story order?
No.
It gives a hypothesis to test with other evidence.
Secondary School: Blind the Folio Numbers
Present the container drawings without page numbers.
Ask several students independently to rank them from visually simple to elaborate.
Compare agreement.
Only then reveal current folio order and proposed codicological order.
The exercise demonstrates why blind scoring is stronger than retrospective storytelling.
JC and Adult Readers: Build Competing Sequence Models
Compare at least three sequences:
- current binding order;
- Q19-inner → Q19-outer → Q15;
- all other physically plausible orders.
Score each against independent features:
- container morphology;
- fragment layout;
- textual continuity;
- scribal features;
- paint/drawing sequence;
- bifolium constraints.
The best reconstruction should win across several feature families, not only the one that inspired it.
Reader Checklist: Before You Reorder the Pharmaceutical Pages
- Which physical bifolia are involved?
- Which surfaces are genuinely part of the same visual family?
- What feature defines “simple” versus “elaborate” containers?
- Was that feature defined before looking at the desired order?
- Does plant-fragment layout support the same sequence?
- Does text behaviour support it?
- Does hand or Currier classification conflict with it?
- Are you reconstructing production order, binding order or reading order?
- Does the proposed sequence violate any physical bifolium constraint?
- Could several illustrators explain the style change?
- What order would falsify the progression hypothesis?
Frequently Asked Questions
Where are the pharmaceutical-looking pages?
The core fragment-and-container material is spread across one bifolium in Quire 15 (f88/f89) and two bifolia in Quire 19 (f99–f102).
Why do researchers think they may be out of order?
The container drawings in Q19 appear to progress from simpler forms toward the elaborate style continued in the Q15 bifolium, creating a visual sequence different from current quire numbering.
What sequence has been proposed?
A commonly discussed reconstruction is Q19 inner bifolium (f100/f101), then Q19 outer bifolium (f99/f102), then the Q15 pharmaceutical bifolium (f88/f89).
Does that prove the pages describe a process?
No. It is first a visual/codicological order hypothesis. Semantic process order requires additional evidence.
Why is this important?
Because it provides a concrete place where internal design continuity may preserve an older arrangement that present binding and quire numbering obscure.
Research Foundations
- René Zandbergen — The Illustrations, for the pharmaceutical folios, container progression and proposed sequence.
- René Zandbergen — Special Topics: Origin, for the page-order implications.
- Nick Pelling — Voynich Codicology, for broader misordering arguments.
- Voynich Ninja — The Containers in the Pharma Section, long-running specialist discussion.
- The Vessels and Plant Fragments, the canonical semantic/visual owner.
The Final Idea
Some sequences are written as numbers.
Some are preserved in physical folds.
And some may survive only because an illustrator changed style gradually enough for us to notice.
The pharmaceutical pages may be one of those cases.
The binding tells us where the pages ended up. The vessels may still be whispering where the sequence once ran.