Why Chinese Physicians Kept the Seed
One Plant, Two Medicines
Modern accounts of cannabis in Chinese medicine often rest on one early source and a few approving lines taken from it. The source is the Shennong Bencao Jing, the Divine Farmer's Classic of Materia Medica, and the lines most often repeated are the ones that make the plant sound mild and restorative. Read in full, the text is less obliging. The same entry that credits part of the plant with benefiting the five viscera also records that eating too much of it makes people see ghosts and run about wildly, and that taking it over a long period lightens the body and opens a way to the spirits. Praise and warning sit a few words apart. Modern retellings often preserve the approving language more readily than the warning.
The book itself calls for caution before any single line does. Shennong, the Divine Farmer said to have tasted hundreds of plants to learn their effects, belongs to legend. The text bearing his name was compiled anonymously, most likely during the Eastern Han dynasty in the first or second century CE, from material that was probably older. No original copy survives. The version read today is a reconstruction drawn from later works that quoted it, especially the edition annotated by the Daoist physician Tao Hongjing in the early sixth century, who marked his own comments in a different ink from the text he had received. Every phrase credited to the Divine Farmer has passed through centuries of copying before reaching a modern translation.
Elsewhere, the history of cannabis medicine is usually told as a story about the flowering plant and its resin. South Asian physicians worked with those materials for centuries, and when William O'Shaughnessy began his trials in Calcutta in the 1830s, the preparations he adopted came from that local practice. Nineteenth-century Western pharmacopoeias listed preparations made from the flowering tops. In that telling, the plant's medicine lies in its most potent parts.
Chinese medicine tells it differently. Hemp had been grown in China since Neolithic times, and the earliest materia medica already recorded a part of the plant that could unsettle the mind. Yet the part that held its place in Chinese prescriptions for the next two thousand years was the seed, a food grain valued for moistening and nourishing the body. A tradition that knew the plant's power over the mind came to build its medicine around something else, and its texts record how that happened.
Two Cannabis Entries
The Bencao Jing does not treat cannabis as a single medicine. It describes one part of the plant under the name mafen, giving mabo as another name, and it describes the seed separately as mazi. The two descriptions read almost as though they concern different plants.
What mafen denoted is not settled. The classic gives no physical description, and commentators disagreed about it from at least the sixth century onward. Tao Hongjing described it as cannabis without fruit, while the older Erya dictionary had treated the term as referring to the fruit. Modern scholars have proposed the immature female inflorescence, the mature seed-bearing flower head, and the bracts surrounding the fruit. Most of these candidates are resin-bearing reproductive structures, but the name alone cannot establish which part the compilers meant, or whether every later author meant the same thing by it.
The entry itself is brief. Mafen is acrid in flavor and balanced in nature. It is said to govern the "five taxations and seven damages," traditional categories of depletion and injury with no clean modern equivalent, to benefit the five viscera, and to bring down blood and cold qi. Then come the lines about excess and long use.
The seed entry speaks a different vocabulary. It gives the seed a sweet flavor and the same balanced nature, credits it with supplementing and strengthening the body, and promises that long use keeps a person robust. Nothing in it concerns spirits, ghosts, or altered perception. The notes that physicians later attached to the classic sharpened the division. They labeled mafen toxic and the seed nontoxic, and they assigned the two different harvest times, mafen in the seventh month and the seed in the ninth.
One word shared by both entries has caused persistent misreading. Balanced translates ping, a term for a substance's thermal nature, meaning it neither warms nor cools. Modern writers have read it as harmony or equilibrium in the body and on that basis described cannabis as a balancing medicine. In the bencao vocabulary the word says something much narrower, and it applies to the troublesome mafen as fully as to the seed. Any claim about what the classic thought of cannabis has to name which entry it relies on and read its terms in the sense the compilers gave them.
Reading the Warning
A modern reader tends to see the mafen entry as a dose threshold, with a mild benefit turning into harm at higher amounts. The entry is arranged differently. It separates quantity from duration. Eating a great deal produces the ghost-seeing frenzy, while steady use over a long time lightens the body and opens a way to the spirits. Promises of a light body recur throughout the classic's entries for substances taken over years in pursuit of longevity. Read on its own terms, the mafen entry describes a substance with two faces, one dangerous and one prized by a particular kind of practitioner.
Later commentators show how physicians responded. Tao Hongjing noted in the early sixth century that mafen was rarely used in formulas. He also recorded a claim from adepts, probably Daoist practitioners, that taking the flowers with ginseng granted knowledge of things not yet come. A thousand years later Li Shizhen, compiling his Compendium of Materia Medica in the late sixteenth century, repeated the claim, called it an overstatement, and recast the combination as a remedy for forgetfulness. His Compendium also preserved an older recipe in which cannabis was pounded with two other plants into pills, one to be taken each morning facing the sun for a hundred days by anyone who wished to see ghosts.
The history of the name says something about how the entry was transmitted. Commentators went on arguing over mafen for centuries, and Li Shizhen's attempt to settle the matter by splitting mafen and mabo into separate entries introduced new contradictions that modern reference works still carry. A dispute that persisted for more than a thousand years may point to declining familiarity with the material itself, and to later authors preserving older descriptions of something they seldom handled directly.
The sources never explain why physicians kept their distance. One plausible reading, offered as interpretation, is that bencao medicine valued substances whose effects stayed within predictable limits over long use, and a material that could deliver either a light body or a frenzy was a poor fit for that ideal.
How the Seed Endured
The seed followed a steadier course. Around the end of the second century, the physician Zhang Zhongjing recorded a hemp seed pill for dry, stubborn constipation, and the formula remains in common use today. Under the name huomaren, the ripe fruit, commonly referred to as hemp seed, is an official drug in the Pharmacopoeia of the People's Republic of China, where it serves as a moistening laxative. It is the part of the cannabis plant that has retained an established place in modern Chinese clinical practice.
The written record shows the shift in proportion. Early herbals often headed their cannabis entries with mafen and treated the seed as an addendum. By Li Shizhen's time the balance had reversed, and his Compendium gives roughly three times as much space to the seed as to mafen. Many Qing dynasty herbals dropped mafen altogether. A modern encyclopedia of nearly 97,000 historical Chinese formulas lists 69 built around the seed and no named formula centered on the flowering part.
The flowering part did not vanish entirely. From the Song period onward, a few texts paired cannabis flower with datura as an anesthetic for painful procedures. These were narrow, specialized uses, and they never gave the flower anything like the seed's place in routine prescribing.
Part of the explanation may lie in what the seed was outside medicine. Herbals frequently classed hemp among the grains, and its oily seed was eaten long before it entered any formula. Its enduring medical role centered on nourishment and complaints of the bowels. Those uses gave results that could be repeated from one patient to the next and carried little risk over long use, which matched the classic's ideal of a medicine safe to take for years. The flowering part offered the opposite, a strong effect on the mind that went two ways depending on quantity and duration.
The outcome shaped what Chinese medicine actually handed down. Its working inheritance from cannabis is an inheritance of the seed. The familiar claim that Chinese physicians used cannabis for thousands of years is literally true and misleading whenever cannabis is taken to mean the flowering plant.
The South Asian Path
South Asian practice went the opposite way with the same plant. There, the flowering tops and their resin became the basis of medicine and were used for pain, spasm, and other difficult conditions. By the time O'Shaughnessy arrived in Bengal, those materials already held an established place in local practice, and he built his tinctures from them.
Both traditions knew that part of the plant acted powerfully on the mind. What separated them was whether that part of the plant became a routine medicine.
Present-day genetic and archaeological research suggests one possible explanation, and it belongs to modern scholarship rather than to the historical texts. Hemp in central China was cultivated for millennia as a fiber and seed crop, and some scholars have proposed that long selection for those uses diluted resin-rich drug types in the Chinese heartland, where archaeology yields abundant hemp fiber and little evidence of drug use. In modern China, drug-type plants are reported mainly in the far northwest and southwest. If the proposal is correct, many Chinese physicians may have had less potent flowering material to work with than their South Asian counterparts. The question remains open. The herbals themselves never distinguished fiber plants from drug plants, so this account rests on genetics and archaeology rather than on anything the physicians wrote. The bencao preference for gentle, long-term medicines and the flower's association with spirit-seeking adepts may also have pulled in the same direction.
None of these explanations accounts for the divergence alone. What the sources do establish is that medical reliance on the flowering plant was regional. Two literate medical cultures, both well acquainted with the plant, reached different conclusions about which part of it belonged in routine care.
What the Texts Support
That divergence limits what the Divine Farmer's Classic can be asked to support. The enduring Chinese medicine of cannabis is a seed medicine, and hemp seed contains at most trace amounts of cannabinoids. Its long record as food and laxative says nothing about THCA or any other acidic cannabinoid, and it cannot serve as historical precedent for acidic cannabinoid preparations.
The mafen entry is the classic's only candidate for such a precedent, and it cannot bear the weight. The classic does not say which part of the plant mafen was, how it was prepared, or how much counted as excess. It does not say whether the material was taken fresh or dried, cooked or raw. Modern chemistry treats those details as decisive, since heat converts THCA into THC and the two compounds act differently in the body. That distinction belongs to the twentieth century. The compilers never drew it, and their words cannot be mapped onto it without supplying information the text does not contain. Reading the ghost-seeing warning as a description of THC, or the word balanced as a reference to acidic cannabinoids, imposes modern chemistry on a source that knew nothing of it.
What the record does offer is historical, and it should stay historical. It shows compilers who separated the parts of a plant carefully, who distinguished heavy use from prolonged use, and whose entries record a potent part of a familiar crop drifting out of routine medicine over more than a thousand years. Those are observations about how one tradition handled cannabis. They are not evidence about how modern THCA preparations behave, and they should not be presented as such.
The texts also preserve a fact that sits uneasily with modern efforts to enlist them. A medical tradition that knew cannabis well kept recording the part of the plant that affected the mind, and over many centuries its medicine settled on the seed. The texts are best read for what they record about that tradition, and not as support for modern claims their compilers had no way to make.
References & Citations
Brand and Zhao, Frontiers in Pharmacology, 2017 — historical review of cannabis in Chinese medicine, including plant-part terminology, changing medical use, and unresolved questions in the classical record. (Frontiers)
Informs: Distinguishes the historical use of cannabis seed from mafen and related reproductive material while clarifying the uncertainty surrounding traditional plant-part terminology and the later decline of the inflorescence in routine Chinese medicine.
Hui-Lin Li, Economic Botany, 1974 — archaeological and historical account of cannabis cultivation and use in China. (JSTOR)
Informs: Provides historical context for hemp as an ancient Chinese fiber and grain crop and for the long-standing importance of the seed outside medicine.
Hui-Lin Li, Botanical Museum Leaflets, Harvard University, 1977 — study of hallucinogenic plants described in historical Chinese herbals. (Biodiversity Heritage Library)
Informs: Traces later herbal descriptions of cannabis, including psychoactive effects, spirit-related uses, and the transmission of earlier claims through successive materia medica.
O’Shaughnessy, Provincial Medical Journal and Retrospect of the Medical Sciences, 1843 — revised account of Indian hemp preparations and their medical use in Bengal. (BMJ)
Informs: Establishes the South Asian comparison by documenting the flowering and resinous cannabis preparations O’Shaughnessy encountered and adapted for medical use in nineteenth-century Bengal.