It is hard to grasp the Compendium of Materia Medica as an idea until you first meet it as an object. Fifty-two juan (chapter-fascicles) of text, plus two additional juan devoted to illustrations, is not a “book” in the casual sense. It is a designed environment for storing and retrieving knowledge, made to survive repeated handling, repeated copying, and repeated disagreement. Its scale announces its ambition: to gather an entire world of substances and their uses into a navigable order. [1]

The numbers that accompany the Compendium are not decorative statistics. They describe an editorial project that treated pharmacology and natural knowledge as something that could be stabilized through structure: about 1.9 million Chinese words, 1,892 medicinal items, and 11,096 prescriptions attached to them, reinforced by 1,109 pictures. Even before one asks whether any particular remedy “works,” one is confronted with the more historical question the Compendium forces on its readers: how does a medical culture decide what it knows, what it calls things, and how it prevents confusion from harming practice? [1]

Li Shizhen (1518–1593) sits at the center of that question not because he occupies a mythical origin point, but because he turned an inherited tradition of materia medica into a machine for correction, comparison, and classification. The awe his name still provokes is best earned the way his book earns it: by specificity, method, and the visible labor of organizing a difficult inheritance. [1] [2]

The Shock of Scale: A 52 Juan Encyclopedia Built to Be Used

First definitions: A bencao is a genre of “materia medica,” a tradition of writing that catalogs substances believed to have therapeutic value, describing their identification, properties, and uses. A juan is a unit of organization often translated as “fascicle” or “chapter,” a practical division for large works that had to be copied, printed, and consulted in parts. The Compendium is a bencao, but it is also something larger: it treats medicine, natural history, and textual scholarship as one continuous domain of work. [1]

Compendium at a glance: The work’s headline figures are unusually stable across serious descriptions because they are not interpretive claims; they are features of the artifact itself. Where sources differ is not usually in the counts, but in how they date “publication” (preface date, block-carving date, first complete printing, first wide circulation). The Compendium’s internal engineering, by contrast, is visible on the page and consistent: a fixed hierarchy of categories, standardized entry segments, and a systematic spread of illustrations that aims to reduce mistakes of identity. [1] [2]

Feature	What it indicates historically	Anchor figures
Text size	Encyclopedic ambition and long compilation time	About 1.9 million Chinese words [1]
Structure	Retrieval design for a working reference	52 juan of text + 2 juan of illustrations [1]
Visual program	Identification support and morphological comparison	1,109 pictures [1]
Inventory	A claim to completeness within a tradition	1,892 medicines [1] [2]
Prescriptions	Integration with practice and recipe literature	11,096 prescriptions [1]
New material	Not just copying, but extending the record	374 entries added by the compiler [2]

Knowledge callout 1: Scale is a historical argument. A compendium that can be lifted in one hand can be read cover to cover; a compendium that requires dozens of fascicles trains its reader to consult, compare, and return. The Compendium’s size therefore shapes a habit of use: it encourages medicine as reference work, not memory performance. [1]

The Compendium begins, importantly, not with substances, but with orientation. Its early volumes summarize earlier pharmacological works and lay out essential theory and constraints, including cautions and contraindications. Only after that scaffolding does the reader enter the main inventory of items. This sequencing matters: it frames the catalog as a disciplined domain, not a free-for-all of recipes and wonders. [1]

A modern reader should resist one temptation at the start: to treat the Compendium as a single, static text. In practice, it functioned as a platform. It could be reprinted, corrected, abridged, annotated, and criticized. That is part of why it lasted. An encyclopedia that can be argued with, because it is legible and structured, becomes a durable authority even when later scholars disagree with parts of it. [2] [9]

From Examination Culture to Physician-Scholar: A Life Shaped by Failure and Access

Formation: Li Shizhen was born in 1518 in Qizhou (in present-day Hubei), into a family tied to medical work. Like many ambitious sons in late imperial China, he was educated in the classics and pushed toward the civil service examinations, the prestige ladder that promised official position and social elevation. He passed an early county-level stage, but failed at higher levels multiple times, and redirected his life into medicine. The shift is important not because it is romantic, but because it placed him at the intersection of two skill sets: the disciplined reading practices of a would-be literatus, and the practical demands of diagnosis and therapy. [3]

Early practice and reputation: By his mid-twenties, Li was seeing patients independently. Accounts converge on a key turning point in the early 1550s: he entered the orbit of the Prince of Chu’s household in Hubei as a physician, and was credited with saving the prince’s son during a sudden collapse. The episode matters historically because it explains access. Elite households and the capital’s institutions could open libraries, networks, and rare texts that were otherwise difficult to consult. Compilers of big reference works often need exactly that kind of access. [3]

Capital experience: In the mid-1550s, Li was recommended to serve as an imperial physician in the capital. This appointment is often presented as a short detour, but its intellectual consequences were long. It placed him near rare medical books, competing textual lineages, and the everyday fact of contradiction: different classics said different things, used different names for the same item, or described different items under the same name. If the Compendium later reads like a war on confusion, the capital appointment offers one plausible origin for that war. [3] [4]

Return and resolve: After roughly a year, he resigned and returned home. Biographical accounts frame the decision as disinterest in wealth and fame, but even a more cautious phrasing captures the core: he chose the slower prestige of scholarship and clinical reputation over the faster prestige of office. The rest of his life can be read as an attempt to rebuild the reliability of the materia medica tradition from within, using the tools he had acquired as a reader, a practitioner, and now a compiler with unusually broad access. [3]

A compact way to picture the biography is not as a ladder climbed, but as a loop completed:

• Examination training built habits of close reading and textual comparison. [3]
• Clinical work exposed the consequences of classification errors and name confusion. [3]
• Elite access exposed the breadth of earlier literature and the depth of its contradictions. [4]
• Withdrawal from office created the time and autonomy needed for a decades-long compilation. [3]

Knowledge callout 2: The Compendium is often called “traditional,” but its author’s social trajectory is partly defined by a modern-sounding problem: information overload. When a tradition accumulates hundreds of source texts, the bottleneck becomes organization and verification. The Compendium is a solution to that bottleneck, built with the tools of a scholar who learned, through failure, how to read with discipline. [9]

Why Materia Medica Needed Repair in the Ming World

The problem is not ignorance: By the sixteenth century, Chinese materia medica writing was already ancient. The tradition had accumulated lists of substances, commentaries, regional notes, and recipe collections over many centuries. The challenge was not the absence of material. It was the presence of too much material, unevenly organized, transmitted through copying, and entangled with local names and shifting administrative geography. A later reader could inherit abundance and still be endangered by it. [9]

Concrete failure modes: Li Shizhen’s critique, as summarized in early scholarship on the project, emphasized several practical faults: muddled classification, inconsistency between illustrations and text, and contradictions across earlier authorities. The stakes were not abstract. If a physician misidentifies a substance because two regions use the same name for different plants, or because a diagram does not match the written description, the error can travel directly into treatment. In this sense, compilation becomes a form of patient safety. [3]

A compact text chart helps clarify the “repair” Li aimed at:

Starting problem	Leads to	Then to	Outcome
Name confusion	Misidentification	Wrong substance selected	Unpredictable outcomes [3]
Contradictory sources	Unstable guidance	Improvised synthesis by readers	Uneven practice [3]
Muddled categories	Poor retrieval	Missed substitutes or contraindications	Clinical risk [3]

Why classification is not neutral: In a materia medica, classification is not just filing. It can also be argument. To place “water,” “fire,” “earth,” and “metal” at the head of an inventory is to say something about the continuity between the inorganic and organic world, and about what counts as a “substance.” To group items by grasses, grains, vegetables, fruits, and wood is to treat the living world as a set of retrievable domains. The Compendium’s structure therefore makes claims about nature as much as about medicine. [1] [3]

A note on evidence: The Compendium’s method, as later summarized by scholars of the text, draws on both literary and non-literary sources. It is built from citations, but also from travel and local inquiry. The result is a reference work that simultaneously preserves earlier textual traditions and competes with them, correcting or re-ranking them where necessary. This is not a rejection of tradition; it is tradition turned into an arena for disciplined comparison. [4] [9]

The key historical insight is that the Compendium’s “cool facts” are often facts about information itself: how names attach to things, how errors replicate through copying, and how a culture attempts to stabilize knowledge without a single centralized laboratory system. Li’s answer was to make the page function like an instrument: standardized segments, repeated across entries, that guide a reader through identification, processing, and use claims in a consistent sequence. [1]

Li Shizhen and the Compendium as a Designed Knowledge System

Information architecture: The Compendium’s internal organization is so explicit that it can be summarized without interpretation. Early volumes introduce earlier pharmacological works and essential theory, then a small set of volumes treats drugs by therapeutic categories, and the bulk of the text classifies individual items by domains ranging from the elemental and mineral to the botanical and zoological, ending with the human body as a source of medicinal material. This is not only a table of contents; it is a worldview rendered as a retrieval system. [1]

The volume-to-category mapping is unusually clean:

Volume range	Primary content domain	Why it matters
1–2	Earlier pharmacological works and essential theory; cautions and contraindications	Trains the reader to read with constraints, not just curiosity [1]
3–4	Drugs organized by disease categories	Connects the inventory to practice and case categories [1]
5–11	Individual drugs: water, fire, earth, metal, and related inorganic domains	Begins with the non-living and “elemental” as foundational substance classes [1]
12–21	Individual drugs: grasses	Turns the vast botanical world into a searchable domain [1]
22–25	Individual drugs: grains	Links staple cultivation to medicinal classification [1]
26–28	Individual drugs: vegetables	Anchors medicine in everyday plant knowledge [1]
29–33	Individual drugs: fruits	Treats diet, seasonality, and therapy as overlapping concerns [1]
34–37	Individual drugs: wood	Extends “plant” into bark, resins, and timber-related substances [1]
38	Individual drugs: clothes and utensils	A reminder that “materia” includes crafted things, not only organisms [1]
39–42	Individual drugs: insects and worms	Makes small life part of pharmacology, not marginal folklore [1]
43–44	Individual drugs: scaled animals	Builds zoology into medical reference categories [1]
45–46	Individual drugs: shelled animals	Separates morphological domains for retrieval and substitution [1]
47–49	Individual drugs: poultry	Treats domesticated birds as a distinct pharmacological domain [1]
50–51	Individual drugs: beasts	Consolidates mammals and large animals as higher-order inventory [1]
52	Individual drugs: parts of the human body used as medicine	Ends with the most ethically and conceptually charged domain [1]

Entry anatomy: A single entry, no matter its substance domain, is designed to be readable in parts. The Compendium describes each drug under repeated subtitles: Name and Explanations; Processing Method; Flavor and Taste; Known Cure; Observation Notes; Relevant Prescriptions. This segmentation allows a reader to jump directly to identification, handling, or use claims, while still keeping those elements bound together as one unit of knowledge. [1]

A compact entry schematic looks like this:

• Name and Explanations -> what it is called, and why names differ. [1]
• Processing Method -> how the substance is handled before use (descriptive, not instructional here). [1]
• Flavor and Taste -> classificatory descriptors used to connect item to therapeutic reasoning. [1]
• Known Cure -> claims of indications attached to the item. [1]
• Observation Notes -> commentary, comparison, and critical remarks. [1]
• Relevant Prescriptions -> how the item appears in formula literature. [1]

What counts as evidence on the page: The Compendium’s authority is often misread as simple accumulation. Its deeper authority comes from its repeated demand that claims be anchored in something: a named earlier text, a comparison across regions, an explanation of why a name maps to an item, or an editorial note that flags uncertainty. Later scholarship has emphasized how broad its source base is, including hundreds of titles listed at the beginning and many more quoted or mentioned across the body. The implication is that Li treated the book not only as a container of facts, but as a map of the tradition’s own internal disagreements. [9]

Evidence as a layered stack

The Compendium’s method can be described as a layered stack rather than a single proof system.

Textual layer: Earlier bencao works and recipe collections are cited, compared, and sometimes corrected. [9]

Geographic layer: Place of origin and regional variation become part of the description, reflecting an assumption that location shapes a substance’s qualities. [9]

Morphological layer: Illustrations and descriptive cues are deployed to stabilize identification, especially where names float across regions. [1]

Editorial layer: The compiler’s notes function as a running commentary, recording uncertainty, resolving contradictions, or ranking alternatives. [3] [9]

This stack matters historically because it creates a format in which later scholars can intervene. A later critic can challenge a citation, propose a different identification, or add a new item, without destroying the structure. That is how encyclopedias become living institutions: they are built to be revised. [2] [9]

Knowledge callout 3: Volume 38 is a quiet shock. By giving “clothes and utensils” a formal place among medicinal domains, the Compendium treats crafted materials as pharmacological objects. It signals that the boundary between “natural” and “made” is not the organizing boundary of the genre; usability and tradition are. [1]

Fieldwork and Textwork: Verification on the Move

Travel as method: Li Shizhen’s compilation did not proceed only by sitting near a library. Accounts of his work emphasize decades of travel and inquiry, including interviews with working experts (farmers, fishermen, hunters, and local healers) and the collection of specimens and recipes. Historically, this is significant because it counters a stereotype: that premodern compilers only recycled books. Li’s project treated local knowledge as a legitimate input, then tried to stabilize it by inserting it into a repeatable entry format. [3] [4]

Workflow cycle (designed to absorb disagreement): [3] [4] [9]

• Read widely
• Extract notes
• Identify contradictions
• Travel or inquire
• Compare specimens and descriptions
• Revise entries
• Repeat

The point is not that travel “proved” everything. The point is that it gave the compiler a second channel of data when the literary channel disagreed with itself, and it gave him access to names and practices that had not yet been stabilized in writing. [9]

A method vignette: the problem of names that drift

Localism and recognition: One recurrent problem in materia medica is that a single substance can carry multiple local names, while a single name can attach to multiple substances. Later discussion of Li’s methods includes a vivid example: an anesthetic-associated flower whose naming varied by region, making it hard for readers to recognize what older texts were describing. Li treated the issue as a practical identification puzzle and conducted repeated investigations to clarify the relationship between name, description, and observed effect. In historical terms, the key achievement is not the drama of the anecdote but the epistemic move: he refused to let name alone serve as identity. [3]

Why this matters beyond one plant: In a world without standardized botanical nomenclature, the act of fixing a name-to-thing mapping is itself a medical intervention. It reduces the likelihood that a reader will take the wrong item into practice because a word drifted across provinces. It also allows the compendium to function as a shared reference across regions, not merely a local handbook. This is one reason the Compendium’s repeated emphasis on explanation, observation notes, and illustration is not decorative; it is a safety device built into the genre. [1] [3]

Knowledge callout 4: The Compendium’s “empiricism” is often less about laboratory-style proof and more about disciplined identification. If you cannot reliably say what an item is, you cannot responsibly debate what it does. Li’s repeated attention to names, morphology, and origin therefore sits upstream of every therapeutic claim in the book. [3] [9]

Printing, Editions, and Authority: How a Manuscript Became a Standard

Completion versus publication: A central anchor for the Compendium’s history is clear: Li completed a major draft in 1578 after roughly 27 years of work. The more delicate question is how to date “publication.” Some descriptions emphasize the Wanli-era Jinling block-printed edition associated with Hu Chenglong and date it to 1593, presenting it as the first printing and framing it as posthumous. Other catalog descriptions point to a preface dated 1596. Still other narratives describe a sequence in which blocks were carved around 1593, with complete publication or wider circulation associated with 1596. The safest historical handling is to treat these dates as different “publication events” within the same edition history rather than as mutually exclusive claims. [1] [2] [5] [9]

A compact edition landmark table makes the sequence legible without forcing one definition of “published”:

Date	What the date can mean	What is broadly agreed
1578	Completion of a major draft after decades of compilation	The core manuscript was finished by this point [2]
1593	A Wanli-era Jinling block-printed edition associated with Hu Chenglong; also treated as first publication in some records	Woodblock printing activity and an early edition are tied to this year [1] [2]
1596	Preface date in some catalog descriptions; also used by some scholars as “published” year	Some evidence anchors a 1596 milestone in the edition’s paratext [5] [9]
1603	Major reprint in another region; later editions often trace back to it	A documented reprint that shaped transmission [2]

Authority as a social process: A reference work becomes authoritative not only by being “correct,” but by being printable, retrievable, and stable enough that many readers can cite the same passage. The Compendium’s structure made it easy to consult. Its segmentation made it easy to quote. Its category system made it easy to locate an item even if one’s local term differed. Those are the mechanics of authority. The result is that later scholars and printers could extend the work’s reach through reprints while keeping its internal logic intact. [2]

Recognition as documentary heritage: In 2011, the Compendium was inscribed on UNESCO’s Memory of the World Register, a form of institutional recognition that treats the work as part of global documentary heritage. Historically, this matters less as a modern award than as a reminder of the Compendium’s dual nature: it is at once a medical text and a material artifact of printing, illustration, and information design. [2] [6]

Knowledge callout 5: The Compendium’s survival is not accidental. Once a work becomes a stable print template, it can migrate into many collections, reappear in multiple editions, and outlive shifts in medical fashion. Its authority is therefore partly an achievement of format: a book built to be copied tends to be copied. [2]

East Asian Afterlives: Japan and the Growth of Natural History Reading

Arrival and renaming: By the early seventeenth century, the Compendium had entered Japan and became known there under a Japanese reading of its title, often rendered as Honzo komoku. A Japanese exhibition account describes the text as present in Japan no later than 1604 and emphasizes its influence on local naturalists. Scholars of early modern Japanese knowledge have similarly described the work as a template that shaped later investigations and publications. The key point is that the Compendium traveled not simply as “Chinese medicine,” but as an encyclopedia of substances, descriptions, and classifications that could be used for natural history as well as therapeutics. [8] [7]

Reprinting and domestication: Once imported, the work did not remain a foreign monument on the shelf. It was reprinted, excerpted, and incorporated into local projects. Japanese authors produced works of materia medica that borrowed its categories and methods while adapting them to local flora, fauna, and scholarly priorities. The Compendium thus became a cross-border reference system: Chinese in origin, but East Asian in operation. [7] [8]

A practical way to see “influence” here is to track what gets copied:

Category logic: organizing substances by large natural domains (inorganic to botanical to zoological). [7]

Morphological attention: description and image as aids to identification. [8]

Citation habits: using earlier texts as a layer of evidence to be compared, not merely repeated. [7]

Named points of engagement: Later Japanese scholarship in materia medica and natural history includes figures who engaged directly with the imported tradition, producing works that either explained, adapted, or extended it for Japanese readers and environments. Examples frequently discussed in scholarship include Kaibara Ekiken and Ono Ranzan, among others, who helped turn materia medica study into a recognized scholarly domain in Japan. The Compendium’s role in this story is not that it supplied every fact, but that it supplied a system and a scale that reorganized what it meant to “know” natural substances. [7]

What changed in transit: When a text crosses borders, its categories can acquire new functions. In Japan, the Compendium’s classifications could support the study of local plants and animals even when therapeutic traditions differed. In other words, the Compendium’s afterlife was not only medical; it was epistemic. It helped stabilize a shared language of description and comparison that could be used in scholarship, collection, and print culture. [7] [8]

Modern Scholarly Afterlife: Translation, Dictionaries, and Why the Book Still Pays Rent

Why researchers still care: Modern scholarship often approaches the Compendium less as a handbook for treatment and more as a record of how a society described nature, organized substances, and encoded practical knowledge. One modern characterization emphasizes that even when some substances have “left the platform” of clinical use, the associated data remains valuable for studying past customs and culture, and that the Compendium functions as a broad encyclopedia of natural science for contemporary researchers. This is a historical justification for continued work on the text even in contexts that do not treat it as clinical guidance. [9]

Translation as infrastructure: The Compendium’s size makes casual translation insufficient. Modern translation efforts have therefore tended to be multi-volume, annotated, and tool-like, designed to support cross-referencing rather than to produce a single readable narrative. This matters because it mirrors the Compendium’s original function: it was built to be consulted, not merely read once. Translation that preserves the retrieval design of the original is, in effect, an attempt to rebuild the instrument in another language. [10]

Dictionaries and indexing as modern analogs of the Compendium’s method: One modern scholarly approach has been to build specialized reference tools around the text, such as dictionaries of historical illness terms and geographical designations. The rationale is practical: the Compendium assumes a reader who can decode dense terminology, shifting place names, and a long chain of earlier sources. Indexing and dictionaries therefore do today what the Compendium did in its own time: reduce confusion by stabilizing reference points. [9]

A compact table shows how modern research often “unwraps” the Compendium:

Modern research task	What it unlocks in the Compendium	Why it aligns with Li’s original problem
Terminology dictionaries	Makes historical disease terms and metaphors legible	Reduces misreading due to language drift [9]
Geographical mapping	Translates old place names into identifiable locations	Stabilizes origin claims and regional rankings [9]
Source tracking	Identifies earlier texts and quotation chains	Supports correction and critical comparison [9]
Annotated translation	Preserves structure and segmentation across languages	Keeps the book usable as a reference tool [10]

A final note on limits: It is possible to admire the Compendium without turning it into an infallible monument. Later scholars have pointed out errors and omissions, and the very scale that makes the work impressive also makes it vulnerable to miscopying and misattribution in a tradition of quotation and transmission. The more historically useful stance is the one the Compendium itself encourages: treat it as a structured record of claims, reasons, sources, and editorial judgment, then ask how that structure shaped what later readers could do with it. [9]

Timeline and Working Glossary

Orientation: The dates below are not only a biography, but a map of how a manuscript becomes an institution. The glossary is intentionally brief and functional, aimed at making the compendium’s internal logic easier to follow on first encounter.

Timeline:

• 1518: Birth in Qizhou (present-day Hubei). [3]
• 1540s: Repeated failures at higher-level civil service examinations; formal turn toward medicine. [3]
• Early 1550s: Service connected to the Prince of Chu’s household; rising reputation as a clinician. [3]
• Mid-1550s: Short period of service as an imperial physician in the capital; access to rare texts and contradictions. [3] [4]
• 1550s–1570s: Decades of compilation, note-taking, travel, interviews, and specimen inquiry. [3] [4]
• 1578: Completion of a major draft after about 27 years of work. [2]
• 1593: Wanli-era Jinling block-printed edition associated with Hu Chenglong; also treated as a key publication milestone in multiple records. [1] [2]
• 1596: Preface date in some catalog descriptions; often treated as another publication milestone in the edition history. [5]
• 1603: Major reprint shaping later transmission. [2]
• Early 1600s: Entry into Japan no later than 1604, where it became known as Honzo komoku and influenced natural history study. [8]
• 2011: Inscribed on UNESCO’s Memory of the World Register. [2] [6]

Mini-glossary

Term	Brief meaning in this context
Bencao	The genre of materia medica: catalogs of substances and their properties/uses. [9]
Materia medica	A systematic inventory of medicinal substances, often tied to recipes and practice. [9]
Juan	A fascicle or chapter-unit used to divide large works for consultation and printing. [1]
Woodblock printing	A method of printing by carving text/images into wood blocks, then printing on paper. [1]
Wanli	The reign era used to date late-sixteenth-century milestones in the Compendium’s printing history. [1]
Jinling	A historical name for the Nanjing area, associated with a key early printing of the Compendium. [1]
Paratext	Prefaces, guides, and other framing material that can carry dates and editorial claims. [2] [5]
Honzo komoku	A Japanese reading of the Compendium’s title, associated with its early modern Japanese afterlife. [8]
Contraindications	Warnings and constraints on use, highlighted in the Compendium’s early volumes. [1]

Key Takeaways

• The Compendium of Materia Medica is an engineered reference system: 52 juan of text plus 2 juan of illustrations, about 1.9 million words, 1,892 items, and 11,096 prescriptions. [1]
• Li’s lasting contribution is methodological: standardized entry structure, correction of contradictions, and verification practices designed to reduce misidentification and clinical risk. [1] [3]
• The work’s authority grew through edition history and cross-border use, shaping natural history and materia medica study in East Asia and earning modern documentary-heritage recognition. [2] [7] [8]

References

Primary documents and catalogs

[1] UNESCO Memory of the World Register nomination file. “Ben Cao Gang Mu (《本草纲目》Compendium of Materia Medica).” PDF. (accessed December 30, 2025).
[2] Library of Congress. “Ben cao gang mu : Wu shi er juan, fu tu er juan.” Item record and description. (accessed December 30, 2025).
[5] HathiTrust Digital Library. Catalog record: “Ben cao gang mu / [Li Shizhen]. 本草綱目.” Note includes preface dated Wanli 24 (1596). (accessed December 30, 2025).
[6] UNESCO. “Ben Cao Gang Mu (《本草纲目》 Compendium of Materia Medica).” Memory of the World Register page. (accessed December 30, 2025).

Scholarly articles and studies

[3] Li, Min, and Yongxuan Liang. “Li Shizhen and The Grand Compendium of Materia Medica.” Journal of Traditional Chinese Medical Sciences 2 (2015): 215–216. PDF. (accessed December 30, 2025).
[7] Marcon, Federico. “The ‘book’ as fieldwork: ‘textual institutions’ and nature knowledge in early modern Japan.” BJHS Themes 5 (2020). (accessed December 30, 2025).

Modern academic editions

[9] “Prolegomena” excerpt PDF for an annotated English-language scholarly project on the Ben cao gang mu, discussing source breadth, terminology tools, and travel-based data collection. PDF. (accessed December 30, 2025).
[10] Book information page for a modern annotated English-language volume in a multi-part translation project of the Ben cao gang mu. (accessed December 30, 2025).

Biographical and exhibit context

[4] Harvard Magazine. “Biographical sketch of Chinese pharmacologist Li Shizhen.” January 1, 2010. (accessed December 30, 2025).
[8] National Diet Library (Japan). NDL Gallery. “The beginning of progress in the 17th century.” Notes arrival in Japan no later than 1604 and influence as Honzo komoku. (accessed December 30, 2025).