Points of Discovery

Two Sheng of Water, Wrung Out

The story says Tu Youyou found the cure for malaria in a 1,600-year-old book. She did read the book — but the sentence that mattered was an instruction about water temperature, and it only mattered because 190 previous samples had failed.

The myth

The version that circulated after October 2015, when the Nobel Assembly announced that the Prize in Physiology or Medicine would be shared with an 84-year-old Chinese pharmacologist almost nobody outside China had heard of, went something like this. Tu Youyou, tasked by her government with finding a cure for malaria, turned away from modern laboratory screening and went instead to the classical texts of Chinese medicine. In a manuscript more than 1,600 years old she found a remedy for fevers — sweet wormwood, qinghao — that the modern world had forgotten. She tested it. It worked. Ancient wisdom, patiently consulted, had solved a problem that Western pharmaceutical science could not, and a drug derived from that plant has since saved millions of lives.

It is an appealing story and it is not a fabrication. Every element in it is drawn from something that happened. Tu did consult the classical literature; the text is real and can be dated; the plant is the right plant; the drug is real and the lives saved are real. A reader who has met this story in a headline or a documentary or a school assembly will recognise the shape of it.

What the shape leaves out is the part where the plant did not work.

The problem, and why the state was solving it

The programme that employed Tu was not a research grant. It was a military project, and it had a date for a name. On 23 May 1967, in response to a request from North Vietnamese leadership for help with malaria — which was disabling soldiers on both sides of a war being fought in mosquito country, and against which the standard drug chloroquine was failing as resistance spread — the Chinese government established a secret national programme. It was called Project 523, after the date of its founding meeting. Over its life, from 1967 to 1980, it involved more than 500 scientists, military personnel and medical practitioners, according to the account of the project given in the Lasker Foundation's 2011 profile of Tu published in the Journal of Clinical Investigation.

This is the first thing the myth omits, and it changes the register of the story. The work was done during the Cultural Revolution, in a country where universities had been disrupted and intellectuals were a suspect class, under a programme whose existence was classified. Tu was named head of a research team at her institute — the Institute of Materia Medica in Beijing, where she would work for more than fifty-five years — in early 1969. She had graduated from the School of Pharmacy at Beijing Medical College between 1951 and 1955 and had since been trained in traditional Chinese medicine, which is why the assignment fell to her: her brief was to screen the traditional pharmacopoeia for antimalarial candidates.

She was not a lone reader stumbling on a forgotten remedy. She was a professional pharmacologist running the traditional-medicine arm of a classified state programme that had already been searching for four years.

Two thousand recipes, and then a failure

The screening effort was industrial in scale. Tu's team assembled roughly 2,000 recipes drawn from some 640 herbs, and narrowed these to 380 extracts prepared from 200 herbal preparations, each of which had to be tested against malaria parasites in animal models. These figures are given in the Lasker profile and are the least romantic and most important numbers in the story.

Qinghao — Artemisia annua, sweet wormwood — was among the candidates. It had been tested. It had shown some activity, and then it had failed to reproduce. On the ordinary logic of a screening programme, that is the end of a candidate: an extract that does not reliably kill the parasite is discarded and the team moves to the next of 380.

The reason this matters is that it inverts the myth's causality. In the popular version, the ancient text is where the plant comes from — the source of an idea nobody had had. In the documented version, the plant was already on the list and had already disappointed. What Tu went back to the literature for was not a new candidate. It was an explanation for a failure.

The sentence

The text she returned to was the Zhouhou Beiji Fang — A Handbook of Prescriptions for Emergencies — compiled by the physician and alchemist Ge Hong, who lived from 284 to 363 CE. Ge Hong's handbook was a practical manual, a book of remedies for the sort of crises a household or a travelling physician might meet, and it contains a passage on the use of qinghao against intermittent fevers.

The Lasker profile reproduces the passage in translation:

> "Another recipe: qinghao, one bunch, take two sheng [2.2 liters] of water for soaking it, wring it out, take the juice, ingest it in its entirety"

Read as a recipe, this says nothing a screening chemist needs. Read as a procedure, it says one thing very loudly: there is no boiling in it. Chinese herbal preparations are conventionally decocted — simmered in water, often for a long time — and Tu's laboratory extractions had used heat as a matter of course. Ge Hong's instruction is to soak the plant in cold water and squeeze the juice out of it, and to drink that.

Tu's own account of what she concluded is preserved in the Lasker profile: that "the heating involved in the conventional extraction step we had used might have destroyed the active components, and that extraction at a lower temperature might be necessary to preserve antimalarial activity."

That is the whole insight, and it is worth being precise about what kind of insight it is. It is not a claim about a plant. It is a claim about an experimental artefact — that the negative result her team had recorded was a property of their method rather than of the material. The fourth-century text did not tell her that qinghao cured malaria; her own screening had already suggested it might, and then unsuggested it. The text told her why her instrument might be lying to her.

Sample 191

Acting on it, the team moved to solvents with low boiling points — ethyl ether boils at about 35°C — so that the plant could be extracted at reduced temperature, and separated the acidic and neutral fractions. In September 1971 they began this work. On 4 October 1971 they obtained a sample that inhibited parasitaemia completely in rodent and monkey models.

It was numbered 191.

The number is the part of the story that gets lost, and it is arguably the most instructive detail in it. One hundred and ninety previous preparations had not done this. The insight from Ge Hong did not deliver a cure; it delivered a corrected method, which then had to be run through the same grinding sequence as everything before it until one of the attempts worked.

What followed was not a publication. With clinical testing needed and the safety of the extract unestablished, Tu and two other members of her team took it themselves. The Lasker profile records the self-experimentation plainly and without embroidery, and it is worth resisting the urge to dramatise it: no contemporary account available here describes the scene, and the honest statement is simply that three researchers tested an unproven plant extract on their own bodies before giving it to patients. Trials followed in Hainan Province in 1972, in 21 patients infected with Plasmodium vivax, P. falciparum, or both, where fevers cleared faster than in the chloroquine controls.

In 1973 Tu obtained the pure compound and produced a derivative, dihydroartemisinin — C₁₅H₂₂O₅, molecular weight 282 — whose structure was confirmed in 1975 as a sesquiterpene lactone. The peroxide bridge in that structure was unlike anything in the existing antimalarial pharmacopoeia, which is precisely why nothing in the modern drug literature could have predicted it.

The credentials she did not have

When the Nobel announcement came in 2015 — Tu was the first mainland Chinese scientist to receive a Nobel in a scientific category — Chinese internet users gave her a nickname that stuck: the "three-nos" laureate. No doctorate. No period of study at a foreign university. No election as an academician of any of China's national academies.

This was not a curiosity. It set off a public argument in China about what its academic evaluation system actually measured, serious enough to be discussed in The Lancet in 2016 under the title "Tu Youyou's Nobel Prize and the academic evaluation system in China."

It is also worth saying what the "three nos" framing risks flattening. Project 523 involved over 500 people, and artemisinin's development from a promising extract into a deployable therapy was the work of many of them across more than a decade. The Nobel was awarded to Tu for the discovery; the drug that reached the world was not made by one person, and Chinese accounts of the programme have not always agreed about the allocation of credit. That disagreement is part of the record and is not resolved here.

Why the truth is the better story

The myth is a story about the past being wiser than we assume. Its moral is that we should look backwards more often.

The documented version is about something harder: the difficulty of telling a failed hypothesis from a failed measurement. Tu's team had the right plant in their hands and had recorded it as a disappointment, because the standard preparation method destroyed the thing they were looking for. Nothing in the chemistry as understood in 1971 would have flagged that. The correction came from an unexpected direction — a fourth-century instruction to use cold water — but it was only legible to someone who already had 190 negative results and was looking for a reason not to trust them.

That is the discipline the story illustrates, and it is not "consult the ancients." It is closer to the opposite: hold your own negative results loosely enough to ask whether the apparatus produced them. The old book was useful because Tu could read a household remedy as a statement about extraction temperature, and she could do that because of the modern pharmacological training the myth quietly writes out of her.

There is a version of this story in which ancient wisdom triumphs over modern science. The record supports a better one, in which a state programme, a classical text, an animal model, a low-boiling-point solvent and 190 failures were all necessary — and the person who assembled them held none of the credentials her own country used to identify people capable of it.

On sources. The quotation from Ge Hong's Zhouhou Beiji Fang (A Handbook of Prescriptions for Emergencies) — "Another recipe: qinghao, one bunch, take two sheng [2.2 liters] of water for soaking it, wring it out, take the juice, ingest it in its entirety" — and Tu Youyou's own statement that "the heating involved in the conventional extraction step we had used might have destroyed the active components, and that extraction at a lower temperature might be necessary to preserve antimalarial activity" are both quoted from the Lasker Foundation profile published as "From branch to bedside: Youyou Tu is awarded the 2011 Lasker~DeBakey Clinical Medical Research Award for discovering artemisinin as a treatment for malaria," Journal of Clinical Investigation, 2011, available open-access via PubMed Central (PMC3195493). That same profile is the source for Ge Hong's dates (284–363), the founding of Project 523 on 23 May 1967 and its Vietnam War origin, Tu's appointment to head her institute's team in early 1969, the screening figures (2,000 recipes from 640 herbs, narrowed to 380 extracts from 200 preparations), the ethyl-ether and acidic/neutral separation method, sample 191 and its date of 4 October 1971, the self-experimentation by Tu and two colleagues, the 1972 Hainan trials in 21 patients with P. vivax and/or P. falciparum, the 1973 isolation and dihydroartemisinin with its formula and molecular weight, the 1975 structural confirmation as a sesquiterpene lactone, and Tu's education at the School of Pharmacy of Beijing Medical College (1951–1955) and her long tenure at the Institute of Materia Medica. The scale and duration of Project 523 — more than 500 participants, 1967 to 1980 — is as reported in that profile and in contemporaneous reporting of the 2015 Nobel announcement. The "three-nos" nickname and its three components are as reported in Chinese and English-language coverage following the 2015 prize, and the ensuing debate about China's academic evaluation system is the subject of a 2016 correspondence piece in The Lancet, "Tu Youyou's Nobel Prize and the academic evaluation system in China."

Two limitations should be stated rather than smoothed. First, nobelprize.org's biographical pages and the Britannica entry both returned access errors during the preparation of this piece and could not be read directly; where this article relies on prize-body or encyclopedic material it does so through the secondary sources named above rather than through unverified paraphrase, and readers wanting the official account should go to those pages themselves. Second, the account of qinghao having been tested and having failed to reproduce before Tu returned to the classical literature is the standard reading of why the Ge Hong passage mattered — the insight is explicitly about correcting a prior extraction method, which presupposes a prior extraction — but the specific sequence and number of earlier qinghao attempts is not established by the sources consulted here, and no figure for them is asserted. Nothing in this piece reconstructs a scene, a conversation or an interior state that a named source does not record; where the record is thin, as with the self-experimentation, it is left thin.