Chinese materia medica

Herbs are almost never given singly — the unit is the formula, built like a court with an emperor, a minister and an envoy.

What it is

The Chinese herbal tradition — bencao, literally "roots and herbs" — comprising several thousand substances, of which a few hundred are in routine use.

Its defining feature, and the thing most often missed in Western accounts, is that the unit of treatment is the formula, not the herb. A practitioner does not prescribe astragalus; they prescribe a named formula containing astragalus among several others, adjusted for the individual. Single-herb prescription exists and is unusual.

The classical formula structure is described by analogy with a court:

Emperor (jūn) — the principal herb, addressing the main pattern
Minister (chén) — supports the principal or treats a secondary pattern
Assistant (zuǒ) — moderates the principal's harshness, or treats an accompanying symptom
Envoy (shǐ) — directs the formula to a particular part of the body, or harmonises the rest

Substances are classified by four natures (hot, warm, cool, cold), five flavours (pungent, sweet, sour, bitter, salty), the channels they are said to enter, and their direction of action — rising, falling, floating, sinking.

Where it came from

The Shennong Bencao Jing, compiled in the early centuries CE and attributed to the mythical emperor Shennong, lists 365 substances graded into three classes by toxicity and intended duration of use. The Shanghan Lun of Zhang Zhongjing, around 200 CE, is the foundational formula text and its prescriptions are still in use.

Tao Hongjing expanded the materia medica around 500 CE. The Tang government issued what is generally regarded as the world's first state pharmacopoeia in 659. Li Shizhen's Bencao Gangmu (1596) is the great synthesis and has its own article.

The twentieth century reshaped it substantially. What is now taught as Traditional Chinese Medicine was systematised under the People's Republic from the 1950s — a standardisation of a diverse and regionally varied body of practice into a teachable curriculum, with some material selected and some dropped. Scholars including Paul Unschuld and Volker Scheid have documented this in detail. It is a real tradition with a real history and it is not the unbroken five-thousand-year transmission the marketing describes.

How it is practised

Diagnosis proceeds by pattern differentiation — biàn zhèng — using observation, pulse and tongue examination and questioning, to identify a pattern such as spleen qi deficiency or liver qi stagnation. The formula follows the pattern, not the Western disease name, and two people with the same diagnosis may receive different formulas. Preparations are traditionally decoctions; granules and pills are now common.

This describes the method. It is not instruction and contains no formulas or doses.

What we can and cannot say

We can say the formula structure is a genuine pharmacological idea, not just poetry. The assistant's role of moderating the principal herb's harshness is exactly what a modern formulator would call reducing adverse effects, and the classical pairing of aconite with liquorice and ginger — which is done specifically to reduce aconite's cardiotoxicity — has a plausible basis. Whether the reduction is sufficient is a separate question, and aconite deaths still occur.

We can say this makes the tradition hard to test, and that both sides use that. A trial of a single herb is not a test of the practice, because the practice does not use single herbs. A trial of one fixed formula against one Western diagnosis is not a test either, because the practice individualises. Trials that do individualise lose the standardisation that makes results interpretable. Practitioners are right that the standard design misses what they do; critics are right that a claim which cannot be tested in any available design is in trouble. Nobody has resolved this.

We can point to the one unambiguous success, which came out of this corpus: artemisinin, found by Tu Youyou reading a fourth-century text and noticing that the preparation method mattered. That is a real vindication of taking the literature seriously, and the drug that resulted is an isolated compound given at a measured dose.

We can say the safety problems are specific and serious. Aconite (fù zǐ) is used in classical formulas and is acutely cardiotoxic; deaths from inadequately processed preparations are reported regularly. Aristolochia species entered the materia medica under names that overlap confusingly with safer plants, which is precisely how the Belgian kidney cases happened. Imported preparations have repeatedly been found to contain undeclared pharmaceuticals and heavy metals.

And we can say the conservation problem is severe and is not incidental. Demand for animal products in this materia medica — tiger bone, rhinoceros horn, pangolin scale, bear bile — drives poaching that has pushed several species toward extinction. Chinese authorities have banned the trade in several of these, enforcement is uneven, and many practitioners and professional bodies inside the tradition campaign against them and prescribe plant substitutes. Some wild plant species are under the same pressure.

We cannot say whether pattern differentiation identifies anything real. What can be said is that this is the most theoretically elaborated herbal system in existence, that its formula logic contains ideas modern pharmacology would recognise, and that its safety and conservation problems are as real as its sophistication.