Smoke in the tomb: ritual and medicine in the ancient world

Somewhere in the first millennium BCE, in places as far apart as the high valleys of the Pamirs and the dry hills of southern Judah, people placed cannabis on a source of heat and let it smoulder. We know this not because anyone wrote it down in a form we can trust, but because the chemicals in that smoke left traces that have survived for more than two and a half thousand years. This article follows those traces. It begins with the most famous ancient description of cannabis smoke, by the Greek historian Herodotus, and then moves to three archaeological sites where laboratory chemistry has turned a plausible story into hard evidence: a desert cemetery near Turpan in north-west China, a mountain cemetery on the eastern Pamir Plateau, and an Iron Age shrine at Tel Arad. It then turns to the regions where the evidence is mainly textual, including Egypt, Mesopotamia, South Asia, China and the Graeco-Roman Mediterranean, and asks how much weight those texts can bear.

The thread running through the story is a question of evidence. Some ancient cannabis claims rest on molecules extracted from a wooden bowl or a limestone altar; others rest on the translation of a single plant name in a long-dead language. These are not equally strong, though popular writing often treats them as if they were. The aim here is to show what each kind of evidence can and cannot tell us, without reading modern habits into ancient rites.

This article is the third in a series of six. The plant's botany and origins are covered in "origins", and its prehistoric spread as a fibre and seed crop in "prehistory". The medieval and early modern worlds come next in "medieval-trade".

Three kinds of evidence, and why chemistry matters

It helps first to be clear about what "use" means. The review on which this series is based distinguishes three broad categories of evidence (Long et al., 2017; McPartland, Hegman and Long, 2019):

• Pollen shows that the plant was growing somewhere nearby. It says nothing, by itself, about people.
• Macrofossils and fibres (seeds, stems, cordage, textiles) show that people handled the plant, but usually for food or fibre.
• Chemical residues can show that the plant's psychoactive compounds were present in a particular object, in a particular human context.

Only the third category speaks directly to whether cannabis was used for its effects on the mind. Until relatively recently, almost every such claim about the ancient world relied on interpreting texts. Now a handful can be checked in a laboratory.

What the plant leaves behind

Cannabis produces a family of compounds called cannabinoids, mainly in the sticky resin of the female flowering tops. In the living plant these compounds exist largely in an acidic form. The best known, THCA, is made inside the plant by an enzyme called THCA synthase. When THCA is heated, or slowly over time, it loses part of its structure and becomes THC (Δ9-tetrahydrocannabinol), the main psychoactive compound. Another cannabinoid, CBD (cannabidiol), is not intoxicating in the same way and is typically dominant in fibre-type hemp.

THC is not chemically stable over long periods. Exposed to air, light and time, it oxidises into a compound called cannabinol, or CBN. This matters a great deal to archaeologists. A sample that is thousands of years old will rarely contain much intact THC, but it may contain CBN in quantity. Finding CBN is therefore a little like finding ash where there was once a fire: it is the signature of something that has since changed. A high proportion of CBN relative to other cannabinoids suggests that the original material was rich in THC.

How residue chemistry works

The workhorse technique in most of the studies discussed below is gas chromatography–mass spectrometry, usually shortened to GC–MS. It sounds forbidding, but the principle is straightforward.

First, researchers scrape or extract tiny amounts of organic material from an object, such as the inside of a brazier or the surface of an altar, and dissolve it. In the gas chromatograph, this extract is vaporised and carried by a stream of gas through a long, thin column. Different molecules travel through the column at different speeds, depending on how readily they evaporate and how strongly they cling to the column's inner coating. The mixture is thus separated into its components, which emerge one after another.

Each component then passes into the mass spectrometer, which breaks the molecules into charged fragments and measures their masses. The resulting pattern of fragments is distinctive, rather like a fingerprint, and can be matched against known reference compounds. Put the two together and a laboratory can say, with considerable confidence, that a particular residue contains THC, CBN, CBD, or something else entirely, such as the resins of frankincense.

This is why chemical evidence is so much stronger than philological evidence. A word in an ancient text might mean cannabis, or it might mean a different plant, or a whole class of plants, or something the translator wished it to mean. A molecule with the fragmentation pattern of cannabinol is not open to that kind of reinterpretation. The chemistry does not tell us why people burned the plant, or what they felt, or what they believed about it, but it does tell us that they did.

What a gene can add

One study discussed below also examined the plant's DNA, looking for the gene that codes for THCA synthase. The presence of that gene means the plant had the inherited capacity to produce THCA, and therefore THC, rather than mainly CBD. It is a statement about the plant's potential chemistry, independent of how well the molecules themselves have survived. When chemical and genetic evidence point the same way, the case is much stronger than either alone.

The Scythians: a Greek historian and a frozen tomb

Herodotus in the felt tent

The oldest and most quoted written account of cannabis smoke comes from Herodotus, the Greek historian of the fifth century BCE whose Histories attempted to describe the peoples known to the Greek world. Book 4 is largely devoted to the Scythians, the horse-riding pastoralist peoples of the steppe north of the Black Sea, whom Greeks encountered through trade and colonies on that coast and through war.

Describing Scythian purification after a funeral, Herodotus says that they enclosed themselves in tents made of felt and threw hemp seeds onto red-hot stones. The seeds smouldered and gave off a vapour, and the Scythians, he wrote, would "howl with pleasure" (Herodotus, Histories 4.73–75, trans. Godley, 1921). He also notes that the Thracians made garments from hemp cloth, so he knew the plant's practical uses too.

Read literally, the passage presents a puzzle. Cannabis seeds, the hard little fruits that birds eat, contain very little in the way of psychoactive compounds. Scholars therefore generally read Herodotus' "seeds" as including the resinous fruiting tops of the plant, where seeds and resin-bearing flowers are bound up together. On that reading, the passage describes psychoactive fumigation: deliberately filling an enclosed space with cannabis smoke as part of a ritual (Ren et al., 2019).

Herodotus was not an anthropologist with a notebook in the tent. He was a Greek writing for Greeks, often relying on informants, and his accounts of foreign peoples mix observation with hearsay. His Scythian passage could easily have been dismissed as a traveller's tale. What changed that was archaeology.

Rudenko and the Pazyryk kurgans

In the Altai mountains of southern Siberia, the Soviet archaeologist Sergei Rudenko excavated a group of large burial mounds, or kurgans, at Pazyryk. These belonged to Iron Age horse-riding people who shared many features with the wider Scythian world, and they date to around the fourth to third centuries BCE (Rudenko, 1970).

What makes Pazyryk extraordinary is ice. Water seeping into the burial chambers froze and, in the cold climate of the high Altai, stayed frozen. The result was a natural deep-freeze that preserved materials which almost never survive in ordinary graves: wood, leather, felt, textiles and even human skin. Rudenko's account of these "frozen tombs" became a classic of archaeology.

Among the finds were braziers, stones and hemp seeds (Rudenko, 1970). The combination is strikingly close to Herodotus' description of seeds thrown on hot stones in an enclosed space, and these finds are widely taken as material confirmation of the practice he described (Rudenko, 1970; Ren et al., 2019).

It is worth being precise about what Pazyryk shows. It provides an archaeological context that matches a written description: the right objects, among the right kind of people, at broadly the right period. But the excavations took place long before modern residue chemistry, and the finds are best understood as strong circumstantial support. The decisive chemical evidence came later, and from further east.

(The broader role of steppe peoples such as the Scythians in spreading hemp as a fibre crop belongs to the prehistory of the plant, and is discussed in the "prehistory" article.)

Yanghai: a basket of cannabis in the Turpan desert

A landscape that preserves

The Turpan basin in Xinjiang, north-western China, is one of the driest and hottest places in Asia, and parts of it lie below sea level. Its extreme aridity has a remarkable effect on archaeology. Organic materials that would rot away within years in a damp climate can survive for thousands of years here, desiccated rather than decayed. The region is famous for naturally preserved human remains and textiles, and, as it turns out, for plant material too.

The Yanghai Tombs, a large cemetery near Turpan, have yielded hundreds of burials. Only two of the roughly 500 excavated tombs contained cannabis (Russo et al., 2008). One of them, tomb M90, became the subject of one of the most important studies in the archaeology of the plant.

Tomb M90

Tomb M90 held the burial of a man aged about 45. Radiocarbon dating places it at around 2,700 years before present. Alongside him was a leather basket and a wooden bowl containing 789 g of dried cannabis, preserved in unusually good condition by the dry conditions (Russo et al., 2008).

This was not a few stray seeds dropped by accident but a substantial deposit, placed deliberately in containers as part of a burial.

A multidisciplinary team led by Russo combined three lines of analysis: botanical examination of the plant material, phytochemical analysis of its chemistry, and genetic analysis (Russo et al., 2008). Together they found:

• THC, the principal psychoactive compound, still detectable;
• cannabinol (CBN), the oxidative degradation product of THC, which is exactly what one would expect after millennia of slow chemical change;
• the gene for THCA synthase, showing that the plants had the genetic capacity to produce the psychoactive precursor.

The authors concluded that the plant was "presumably employed by this culture as a medicinal or psychoactive agent, or an aid to divination" (Russo et al., 2008). They also observed that no hemp artefacts, such as textiles or cordage, were found in the tombs, which weakens any interpretation of the deposit as a stock of fibre (Russo et al., 2008).

What Yanghai does and doesn't prove

The Yanghai find is a landmark because it brings together chemistry and genetics on the same ancient material. Its central conclusion, that this was drug-type cannabis deliberately placed in a grave, is well supported.

Two points of caution are important, however.

The first concerns the person in the tomb. The publication and much of the coverage that followed refer to the man as a "shaman". That label is an interpretation, not a finding. The chemistry cannot tell us what social role he held, and "shaman" is a word heavy with modern baggage; it should be treated as a hypothesis.

The second concerns the date. Associated reports give dates of either about 2,500 or about 2,700 years BP, so the chronology should be cited with care. The difference is not enormous in the long view, but it matters when Yanghai is set alongside the other first-millennium BCE sites and someone wants to say which came "first".

There is also the question of purpose. The authors offer three possibilities, medicinal, psychoactive or divinatory, and sensibly do not choose between them; a plant placed in a grave may equally have been an offering or a marker of the dead person's role.

Jirzankal: braziers on the roof of the world

The Pamirs

The Pamir Plateau is a high, mountainous region where the great ranges of Central Asia meet, often described in English as "the roof of the world". Its eastern edge lies in what is now western Xinjiang. It is a harsh environment of high valleys and thin air, but it was not empty: mobile and semi-mobile communities lived and buried their dead there, and the region lay across routes connecting different parts of Asia.

Burning cannabis in the cemetery

At the Jirzankal Cemetery (also known as Quman) on the eastern Pamir Plateau, archaeologists recovered wooden braziers containing burnt stones from several tombs. Ren et al. (2019) analysed the organic residues from ten of these braziers, recovered from eight tombs. In their words, "cannabis plants were burned in wooden braziers during mortuary ceremonies at the Jirzankal Cemetery (ca. 500 BCE)" (Ren et al., 2019).

This is the most securely documented early evidence of burning cannabis for its psychoactive compounds. The distinction from Yanghai matters. At Yanghai the cannabis was placed in the grave as plant material; at Jirzankal we have the remains of the act of burning itself, chemically confirmed. The arrangement, plant material on heated stones in a container, is also strikingly reminiscent of Herodotus' description of seeds thrown on red-hot stones, even though Jirzankal lies far to the east of the Scythians he described.

A chemically unusual plant

The chemical profiles from the braziers held a surprise. They showed a higher proportion of THC breakdown products than is typical of wild cannabis plants (Ren et al., 2019). In other words, the plants burned at Jirzankal appear to have been more potent than one would expect from ordinary wild growth.

The authors offered possible explanations. The plants may have been physiologically altered through hybridisation or domestication, meaning that people had, deliberately or not, favoured more chemically potent forms. Alternatively, the result may reflect a poorly understood plasticity in the plant itself; cannabis chemistry can vary with environment, and the high-altitude conditions of the Pamirs might conceivably play a part (Ren et al., 2019).

Whether the people buried at Jirzankal cultivated their cannabis or gathered naturally high-THC wild plants remains unresolved. This is not a minor detail. If they were cultivating it, Jirzankal would be evidence of deliberate selection for psychoactive potency by the middle of the first millennium BCE. If they were gathering it, it would show that people knew where to find especially potent wild stands. Either way, the find suggests that, by this period, at least some communities were selecting or seeking out chemically potent plants.

Ritual, and a speculative step further

Ren et al. (2019) interpreted the burning as part of mortuary ritual. That is the reading most directly supported by the evidence: the braziers were in tombs, associated with burials.

They also proposed, more speculatively, that ritual "smoking" may have spread from elite to wider social groups in the eastern Pamirs. This is an interesting hypothesis, but it is a hypothesis. It goes beyond what ten braziers from eight tombs can establish on their own, and it should be presented as such.

The word "smoking" also deserves thought. To a modern reader it suggests a pipe or a rolled cigarette, and at Jirzankal there is no evidence of either. What we have is plant material burned on hot stones in a brazier, its smoke breathed from the air: closer to fumigation or incense than to anything modern.

Tel Arad: cannabis at a Judahite shrine

An Iron Age fortress and its shrine

Tel Arad lies in the northern Negev, on the southern margins of the Iron Age kingdom of Judah. The site was excavated in the 1960s. Within an Iron Age fortress, excavators uncovered a small shrine. Its innermost part was a small niche often described as the "holy of holies", and at the entrance to it stood two limestone altars of different sizes.

The shrine offers a rare physical glimpse of cultic practice in Judah outside Jerusalem, but what was burned on its two altars went unanswered for decades.

What the altars held

Arie, Rosen and Namdar (2020) applied gas chromatography–mass spectrometry to residues on both altars, which they dated to the eighth century BCE. The results were different for each altar.

Frankincense, a fragrant resin from trees that grow in southern Arabia and the Horn of Africa, was a prized incense across the ancient Near East and is well known from biblical descriptions of sacred incense. Finding it on an altar in a Judahite shrine is not in itself a surprise, though it does say something about long-distance trade.

The smaller altar was the surprise. Its residue contained THC, CBD and CBN, mixed with animal dung. The authors interpret the dung as a means of burning the cannabis at a low temperature (Arie, Rosen and Namdar, 2020). The logic is that the psychoactive compounds are released by moderate heat but destroyed by fierce combustion; smouldering dung would provide a slow, gentle heat that allowed the cannabinoids to vaporise rather than simply burn away. They argue that the cannabis was burned deliberately to induce altered states of consciousness during cultic ceremonies (Arie, Rosen and Namdar, 2020).

Why the Arad find matters

The Arad result extends securely identified psychoactive use of cannabis into the Levant, far from the Central Asian sites, and into the setting of an organised cult rather than a burial. It also shows two different aromatic substances burned side by side, each on its own altar, which suggests a deliberate and structured practice.

The limits of the Arad evidence

The chemistry is robust, but its interpretation has limits, several of which the authors themselves acknowledge:

• It concerns a single shrine. One site cannot establish a general practice across Judah, still less across the Levant.
• The excavation is old. The excavations took place decades before the chemical analysis, and no final excavation report has been published. Questions about exactly where and how the altars were found, and how they were handled and stored in the intervening years, are therefore harder to answer than they would be for a modern dig.
• The source of the cannabis is unknown. It may have been grown locally, or it may have been imported along the same trade networks that brought frankincense from the south. The residue cannot tell us which.

None of these points undermines the identification of cannabinoids. They do, however, limit how far one can generalise. The Arad find shows that cannabis was burned on an altar at one Judahite shrine in the eighth century BCE. It does not, by itself, show that cannabis was a standard part of Israelite or Judahite religion.

Egypt and Mesopotamia: claims requiring caution

If chemistry has strengthened the case for Central Asia and the Levant, the situation in Egypt and Mesopotamia is different. These two civilisations left enormous written archives, including medical texts. Egypt is famous for medical papyri such as the Ebers Papyrus; Mesopotamia left medical and pharmacological texts written in cuneiform on clay tablets. Popular histories frequently state that cannabis appears in both.

The trouble is that these claims rest mainly on contested identifications of plant names in the texts (Russo, 2007). Ancient plant vocabularies do not map neatly onto modern botanical species. A single ancient word might refer to one plant, to several related plants, to a plant product such as a fibre or oil, or to something we cannot now identify at all. Translators working in the nineteenth and twentieth centuries often proposed identifications based on context, resemblance to words in other languages, or simply plausibility, and those proposals have sometimes been repeated as fact long after the original scholars expressed doubt.

Russo (2007) uses Egypt as an archetype in his survey of cannabis history, but acknowledges the philological uncertainty. That is the crucial point. Unlike the Xinjiang and Arad cases, no published residue study has yet confirmed cannabinoids in a secure Egyptian or Mesopotamian ritual or medical context.

This does not mean the claims are false; it would not be surprising if cannabis was known in such well-connected civilisations. But plausibility is not evidence, and for now these claims should be treated as plausible hypotheses, not established fact. Residue chemistry offers a way of testing them, rather than simply arguing about translations.

South Asia: a sacred and medicinal plant

The Vedic word bhaṅgā

South Asia has perhaps the longest continuous religious and medical tradition of cannabis use in the world. Its starting point is usually traced to the Atharvaveda, one of the four Vedas, the oldest body of sacred literature in the Sanskrit tradition. Unlike the more liturgical Rigveda, the Atharvaveda contains a great deal of material concerned with everyday life, including charms and spells for healing and protection.

The Atharvaveda lists bhaṅgā among five plants that "free us from distress" (Russo, 2005). The resemblance to the later word bhang, which certainly does mean a cannabis preparation, is obvious, and many writers take the Vedic passage as the earliest Indian reference to the plant.

But caution applies here, too. Whether bhaṅgā meant cannabis in the Vedic period is debated (Russo, 2005). The same problem that afflicts Egyptian and Mesopotamian plant names applies: a word can change its meaning over the centuries, and the fact that a later word sounds similar does not guarantee that it referred to the same plant a thousand years earlier. The Vedic reference is suggestive, and the phrase about freedom from distress is evocative, but it is not proof.

Ayurveda and Tantra

Ayurveda, the classical medical tradition of South Asia, developed a sophisticated system of diagnosis and treatment with an extensive materia medica. Later Ayurvedic texts, together with texts from the Tantric traditions, describe medicinal preparations of cannabis (Russo, 2005; Russo, 2007). The word "later" matters: the firmest textual evidence for cannabis as a recognised medicine in South Asia comes from these works, not from the earliest Vedic layer.

Bhang, ganja and charas

Over time, three main kinds of cannabis preparation became established in South Asia, and their names have since travelled around the world (Russo, 2005; Russo, 2007):

• Bhang, made from the leaves and usually taken as a drink or in food;
• Ganja, the dried flowering tops of the plant;
• Charas, the resin, collected from the plant.

These became established consumables associated especially with Shaivite devotion, that is, with the worship of the god Shiva, and with religious festivals (Russo, 2005; Russo, 2007). In this context cannabis was not simply a drug in the modern sense, nor simply a medicine. It was woven into devotional life, hospitality and seasonal celebration.

A colonial witness: the Durga Puja

One striking glimpse of how ordinary these customs were comes from a much later source. In the 1890s the British colonial government in India appointed the Indian Hemp Drugs Commission to investigate the use of cannabis preparations. Its report documented the depth of these customs. It recorded that in Bengal an infusion of the leaves was "commonly offered to every guest and member of the family on the last day of the Durga Puja", the great autumn festival of the goddess Durga, without being considered reprehensible (Indian Hemp Drugs Commission, 1894).

This observation is many centuries removed from antiquity and should not be projected backwards uncritically. But it shows what texts about gods and medicines rarely reveal: a domestic, religiously sanctioned practice in which cannabis was offered as hospitality. (The Commission itself, and its importance for the later history of medicine and law, is discussed in the "medicine-prohibition" article.)

China: ma, materia medica and a disputed anaesthetic

Hemp as a staple crop

In China, hemp, called ma, was one of the classical staple crops. It was important for fibre, which was used for clothing and cordage; for food, in the form of its oily seeds; and later for paper, of which hemp was one of the early raw materials (Clarke and Merlin, 2013). For most of Chinese history, then, cannabis was primarily an agricultural plant of great practical value, more like flax in medieval Europe than like a drug.

It is easy, reading about Yanghai and Jirzankal, both within the borders of modern China, to imagine a continuous Chinese tradition of psychoactive use. The evidence does not support that: those sites belong to the cultures of Central Asia and its mountain borderlands.

The Shennong Bencao Jing

Chinese medicine did, however, take note of the plant. The Shennong Bencao Jing, often translated as "The Divine Farmer's Materia Medica", lists cannabis among its medicinal substances (Zuardi, 2006; Russo, 2007). The text is attributed to Shennong, a legendary emperor and culture hero credited in Chinese tradition with teaching agriculture and testing the properties of plants. In reality, the work was compiled in its received form in the early centuries CE. The attribution to Shennong is a way of claiming ancient authority, not a historical date.

As one of the foundational texts of Chinese pharmacology, its inclusion of cannabis shows that the plant was recognised as medicinal within a learned tradition. As with the Graeco-Roman writers below, that is not the same as describing widespread intoxicant use.

Hua Tuo and mafeisan

One of the most frequently repeated claims in the popular history of cannabis concerns Hua Tuo, a celebrated physician of around the second century CE, remembered in Chinese tradition as a pioneering surgeon. He is said to have used an anaesthetic called mafeisan to render patients insensible during operations. Many books and websites state confidently that this was a cannabis-based anaesthetic.

The claim rests on a translation of ma that sinologists dispute. The character can indeed refer to hemp, but it also carries other meanings, including a sense of numbness, which would make obvious sense in the name of an anaesthetic. Since no reliable recipe for mafeisan survives, there is no way to check the ingredients directly. The claim should therefore be regarded as unproven. It is a good example of how an attractive story, built on a single ambiguous word, can harden into an apparent fact through repetition.

The Graeco-Roman world: a fibre plant with medicinal uses

Hemp in the Mediterranean

Greek and Roman authors knew hemp mainly as a fibre plant, one of several that could supply the rope and coarse cloth on which Mediterranean economies, not least shipping, depended. Herodotus, as we saw, knew of hemp cloth among the Thracians, and described the Scythian smoke-bath precisely because it was unfamiliar to his Greek readers.

What the medical writers say

Medical writers also recorded uses of hemp seeds and juice (Russo, 2007; Zuardi, 2006). Three figures stand out, each a towering authority whose works shaped European and Islamic medicine for well over a thousand years:

• Dioscorides, a Greek physician of the first century CE, whose De Materia Medica became the standard Western reference on medicinal plants;
• Pliny the Elder, the Roman writer and official whose encyclopaedic Natural History gathered together much of what educated Romans knew about the natural world, before his death in the eruption of Vesuvius in 79 CE;
• Galen, the Greek physician of the second century CE, born in Pergamon and active in Rome, whose voluminous writings dominated medical theory for centuries.

These writers describe preparations for earache, flatulence and other complaints (Russo, 2007; Zuardi, 2006). Galen also noted that hemp seed cakes taken in quantity could affect the head (Russo, 2007; Zuardi, 2006).

That last observation is intriguing, because it shows that a careful clinical observer recognised that the plant could have effects on the mind. But it should be kept in proportion. These sources show that the plant's pharmacological activity was recognised. They give no evidence of widespread intoxicant use in the Mediterranean. Mediterranean societies had their own long-established intoxicant in wine, and their own elaborate cultures of drinking; cannabis was, for them, primarily a source of fibre and a minor medicine.

Ritual, medicine or "recreation"? Avoiding presentism

Having surveyed the evidence, it is worth stepping back to consider how it should be interpreted. One of the most common errors in cannabis history is what historians call presentism: reading the past through the categories of the present.

In modern debates, cannabis use tends to be sorted into two boxes, "medical" and "recreational", with "religious" occasionally added as a third. Popular writing often projects these boxes backwards, so that the Scythians become ancient party-goers, the Yanghai burial becomes evidence of an ancient stash, and the Arad altar becomes proof that "the ancient Israelites got high". Ancient ritual fumigation is read as "recreational use", or ancient medical uses as validation of modern products.

The evidence surveyed here does not fit those boxes well.

• Context points to ritual. At Pazyryk, Yanghai and Jirzankal, cannabis appears in funerary settings. At Arad it appears on an altar in a shrine, next to an altar used for frankincense. Herodotus describes Scythian smoke-baths as part of purification after a funeral. In every case where we have archaeological context, that context is ritual or religious.
• Ritual and medicine overlapped. In many ancient societies, healing, divination, purification and worship were not separate activities. Russo et al. (2008) were right to offer "medicinal or psychoactive agent, or an aid to divination" as alternatives rather than choosing one; the people concerned may not have seen these as different things.
• Pleasure is not the same as recreation. Herodotus says the Scythians "howl with pleasure", and it would be pedantic to deny that the experience may have been enjoyable. But enjoyment within a funerary rite is not the same as leisure use in the modern sense. Ecstatic or altered states have been central to religious experience in many cultures, and their value lay in what they were believed to reveal or achieve, not only in how they felt.
• Medical texts are not modern endorsements. When Dioscorides or a Chinese materia medica lists cannabis for an ailment, that tells us about ancient medical thinking. It does not validate any modern product or treatment, any more than ancient prescriptions of other plants validate them today.

Holding these distinctions in mind makes the ancient evidence more interesting, not less: people in the past related to this plant on their own terms, within systems of meaning quite different from ours.

What we know, and what we don't

The evidence for cannabis in the ancient world is uneven. In a few places it is excellent; in others it is little more than a contested word. The table below summarises the main regions discussed in this article and gives a rough rating of the strength of the evidence for psychoactive or ritual use.

What we can say with confidence

• By the first millennium BCE, people in at least two widely separated regions, Central Asia and the southern Levant, were deliberately burning drug-type cannabis in ritual settings (Russo et al., 2008; Ren et al., 2019; Arie, Rosen and Namdar, 2020).
• At Jirzankal, some of the plants used were chemically more potent than typical wild cannabis, suggesting that people were selecting or seeking out potent plants (Ren et al., 2019).
• Herodotus' account of Scythian smoke-baths, once easy to dismiss as a traveller's tale, is broadly consistent with archaeological finds (Rudenko, 1970; Ren et al., 2019).
• In South Asia, China and the Graeco-Roman world, learned medical traditions recognised that the plant had pharmacological effects (Russo, 2005; Russo, 2007; Zuardi, 2006).

What remains uncertain

• Who and why. We do not know the social role of the man in Yanghai tomb M90; "shaman" is an interpretation. We do not know whether the Jirzankal ritual spread from elites to wider society.
• Wild or cultivated. Whether the Jirzankal cannabis was grown or gathered remains unresolved, as does the source of the cannabis burned at Arad.
• The precise chronology. The Yanghai dates differ between reports, and the gap between the plant's early domestication and the first secure chemical evidence of psychoactive use spans millennia with little direct evidence (Ren et al., 2021).
• Egypt and Mesopotamia. Until a residue study confirms cannabinoids in a secure context, claims of use there remain hypotheses.
• The meaning of words. Whether Vedic bhaṅgā meant cannabis, and whether Hua Tuo's ma did, are questions philology alone may never settle.

The most promising way forward is more of what has already transformed the field: residue studies on securely dated vessels, and genetic work on ancient plant remains. Each could confirm a plausible story or retire a popular myth.

The next article in the series, "medieval-trade", follows cannabis into the medieval Islamic world, Africa and across the Atlantic.

Sources

Arie, E., Rosen, B. and Namdar, D. (2020) 'Cannabis and frankincense at the Judahite shrine of Arad', Tel Aviv, 47(1), pp. 5–28. doi:10.1080/03344355.2020.1732046.

Clarke, R.C. and Merlin, M.D. (2013) Cannabis: Evolution and Ethnobotany. Berkeley: University of California Press. doi:10.1525/9780520954571.

Herodotus (1921) Histories, Books III–IV. Translated by A.D. Godley. Loeb Classical Library 118. Cambridge, MA: Harvard University Press.

Indian Hemp Drugs Commission (1894) Report of the Indian Hemp Drugs Commission, 1893–94. Simla: Government Central Printing Office.

Long, T., Wagner, M., Demske, D., Leipe, C. and Tarasov, P.E. (2017) 'Cannabis in Eurasia: origin of human use and Bronze Age trans-continental connections', Vegetation History and Archaeobotany, 26, pp. 245–258. doi:10.1007/s00334-016-0579-6.

McPartland, J.M., Hegman, W. and Long, T. (2019) 'Cannabis in Asia: its center of origin and early cultivation, based on a synthesis of subfossil pollen and archaeobotanical studies', Vegetation History and Archaeobotany, 28(6), pp. 691–702. doi:10.1007/s00334-019-00731-8.

Ren, G., Zhang, X., Li, Y., Ridout, K., Serrano-Serrano, M.L., Yang, Y., Liu, A., Ravikanth, G., Nawaz, M.A., Mumtaz, A.S., Salamin, N. and Fumagalli, L. (2021) 'Large-scale whole-genome resequencing unravels the domestication history of Cannabis sativa', Science Advances, 7(29), eabg2286. doi:10.1126/sciadv.abg2286.

Ren, M., Tang, Z., Wu, X., Spengler, R., Jiang, H., Yang, Y. and Boivin, N. (2019) 'The origins of cannabis smoking: chemical residue evidence from the first millennium BCE in the Pamirs', Science Advances, 5(6), eaaw1391. doi:10.1126/sciadv.aaw1391.

Rudenko, S.I. (1970) Frozen Tombs of Siberia: The Pazyryk Burials of Iron Age Horsemen. Translated by M.W. Thompson. Berkeley: University of California Press.

Russo, E.B. (2005) 'Cannabis in India: ancient lore and modern medicine', in Mechoulam, R. (ed.) Cannabinoids as Therapeutics. Basel: Birkhäuser. doi:10.1007/3-7643-7358-X_1.

Russo, E.B. (2007) 'History of cannabis and its preparations in saga, science, and sobriquet', Chemistry & Biodiversity, 4(8), pp. 1614–1648. doi:10.1002/cbdv.200790144.

Russo, E.B., Jiang, H.-E., Li, X., Sutton, A., Carboni, A., del Bianco, F., Mandolino, G., Potter, D.J., Zhao, Y.-X., Bera, S., Zhang, Y.-B., Lü, E.-G., Ferguson, D.K., Hueber, F., Zhao, L.-C., Liu, C.-J., Wang, Y.-F. and Li, C.-S. (2008) 'Phytochemical and genetic analyses of ancient cannabis from Central Asia', Journal of Experimental Botany, 59(15), pp. 4171–4182. doi:10.1093/jxb/ern260.

Zuardi, A.W. (2006) 'History of cannabis as a medicine: a review', Revista Brasileira de Psiquiatria, 28(2), pp. 153–157. doi:10.1590/S1516-44462006000200015.