If – like me – you missed this fascinating 2023 study on the development of prehistoric cannabis cultigens by Rita Dal Martello et al., you can download the pdf here: Cannabis differentiation and Haimenkou.
The authors examine archaeological cannabis seeds with the aim of illuminating the early development of the crop into distinct drug-type and fibre-type domesticates. Their idea is that fibre- and grain-type domesticates can be distinguished from drug-type based on seed morphology. By comparing known archaeological and modern samples, they analyse over 800 cannabis seeds dated to between 1650 and 400 BCE that were excavated from a site named Haimenkou to the north of Dali, Yunnan (Southwest China).
Haimenkou is located close to “the core area of the 1st millennium BCE polity of the Dian”, which is closely linked to Tai peoples such as the Shan and Lao, who may have originated in or near the Dian kingdom before migrating to regions such as Burma and Laos.
Because the sizes of the analyzed cannabis seeds “mostly plot in the range of overlapping psychoactive/fibre types”, the authors suggest that “the cannabis assemblage from Haimenkou is indicative of a crop beginning to undergo evolution from its early domesticated form towards a diversified crop specialized for alternative uses, including larger oilseed/fibre adapted varieties” – and of course, drug varieties.
For researchers keen to investigate the seed morphology of drug-type landraces in closer detail, there’s extensive data on seed weights and shapes here, including over 140 accessions mostly from Central, South, and Southeast Asia that were collected by The Real Seed Company. Among them are ganja landraces from Laos, Isan, Mae Hong Son, and Burma, as well as Manipur and Orissa, and a hemp landrace from Northeast Laos. The site is best viewed on a laptop.
Yunnan itself is a region where cannabis is currently cultivated as a fibre and grain crop (‘hemp’), and also occurs as a widespread weed in feral forms, possibly even as a truly wild plant. Immediately to Yunnan’s south and west are ethnically Tai regions such as Laos, Thailand, and Shan State where the most common form of the cannabis crop has historically been ganja, though hemp cultivation can be found among non-Tai highland ethnic groups such as the Lanten and Hmong.
Yunnan was favoured by the cannabis breeder David Watson (Skunkman) as the most likely origin of the cannabis plant itself, an opinion he arrived at after growing a wide range of landraces from China, Thailand, and Afghanistan and noticing how their aromas and other traits all pointed to Yunnan as their likely shared origin. An increasing number of studies using genomics and fossil pollen suggest that at least his hunch as to the origin of the species was broadly correct and do indeed point to the contiguous area of Tibet as being where Cannabis sativa L. probably speciated from the hop tens of millions of years ago.
But what’s most fascinating about the present study is how it sheds light on a likely course of development for the ancient ancestors of Tai landraces, particularly the drug-type forms now cultivated for ganja in Thailand, Laos, and areas of Burma like Shan State.
Cultivation of drug-type landraces is associated with the South-Western Tai area, particularly along the Mekong River in Isan and Laos, but also in Shan areas of Burma.
In addition, its insights raise the compelling possibility that a Tai kingdom that ruled much of what’s now Northeast India may have played a role in introducing drug-type cannabis to the Sanskritic cultures of medieval India through Assam and into regions that would later become historic ganja cultivation centres, most notably Bengal and Odisha. That’s the Tai kingdom of Ahom, which over the course of its history controlled territory along the Brahmaputra River from the frontiers of Tibet and Upper Burma around what’s now Arunachal Pradesh up to the borders of Bengal, ruling from 1228 to 1826 CE.
For a scholarly look at the question of when exactly drug-type cannabis arrived in India and first shows up in Sanskrit literature (two separate questions, note) check out the work of Gerrit Jan Meulenbeld. Along with most Indian scholars of Ayurveda, Meulenbeld argues that cannabis drugs only unambiguously appear in Indian medical and religious literature during India’s Persianate era, occurring no earlier than about 1000 CE and under the gramatically feminine name bhaṅgā (भङ्गा). In this view, the earliest definitive reference to cannabis as an ‘intoxicating’ substance is generally agreed to be by Vaṅgasena, an important author in the traditions of Ayurveda, who was located in Bengal and flourished around 1050 – 1100 CE. That medieval date is likely to strike some aficionados as implausibly late and inspire talk of the Vedas and so forth. But the contention of these academics is that although the masculine and neuter forms ‘bhaṅga’ (भङ्ग) such as are found in the Atharvaveda or Pali canon may perhaps refer to fibre-type cannabis (‘hemp’), they could just as likely be names for any number of other things or plants.
In the more expansive view of one of the authors of this 2023 paper on Haimenkou, the archaeobotanist Professor Dorian Fuller, it’s around the Late Harappan Horizon (3900 to 3500 BCE) that “cannabis comes to India […] as part of the “Chinese horizon”, which is really just piece meal adoption of various things coming in via Central Asia including crops and technologies (harvest knives) from China, peaches, apricots, millets, japonica rice.” It’s plausible, Fuller notes, that cannabis at this date was also “already utilized from wild populations in the Himalayas.”
Among Sanskritists, the prevailing orthodoxy remains that drug-type cannabis, like opium, was probably introduced into India with Islam, particularly with Persianate forms of Islam from Central Asia such as blossomed in the Samanid Empire. But absence of cannabis drugs from Sanskrit literature is not the same thing as their absence from India altogether, as noted by Professor James McHugh.
Over a century ago, Nikolai Vavliov proposed that cannabis was very likely independently domesticated at multiple times and places across Eurasia, an opinion still shared by most experts. Per Dal Martello’s study, the earliest cannabis domesticates were surely multipurpose, and it’s notable that multipurpose landraces are to this day the main type of cannabis cultivated in the Indian and Nepali Himalaya. Very tall and typically with good bast fibre and large oily seeds, multipurpose Himalayan landraces produce sufficient THC for use as charas or edibles. Incidentally, cannabis was described as a multipurpose crop by the agronomist Ibn Wahshiyya, who died c. 930 CE in eastern Iraq, and a multipurpose landrace named ‘Smyrna’ was common throughout the late Ottoman Empire. Suffice to say, where and when drug-type domesticates may have developed within South Asia and the Indian and Nepali Himalaya long prior to the Muslim era is looked at in this 2023 study and others by Fuller such as Between China and South Asia.
Compatible with this picture of multiple domestications and diffusions within and without the subcontinent, Dal Martello et al.’s interpretation of Haimenkou provides a plausible origin story for a “both… and…” scenario, one where Tai cultures also introduce drug-type cannabis into the Sanskritic culture zone of medieval India by a highly frictional overland route through Zomia and Assam. In this way, the crops and products that come to be known as bhaṅgā and gañjā (गञ्जा) emerge from across India’s Central Asian mountain frontiers to the north, west, and east: from the Himalaya, through the passes of the Hindu Kush with Islam, and along the course of the Brahmaputra from the Tai mueang that were then expanding out of the foothills of the Southeast Asian Massif. If so, then the literature of the Ahom Kingdom could have much to tell us about this eastern chapters of this history.
Broadly contemporary with the expansion of Ahom culture through Assam is a crucial text of Indian cannabis culture which was compiled sometime between the 12th and 15th centuries at Srisailam, a Shakta–Shaiva temple complex 100 miles inland from the Coromandel Coast in the jungle highlands of what’s now Andhra Pradesh. This is the Ānandakanda or ‘Root of Bliss’, an alchemical compendium that expounds in great detail on the characteristics, effects, and uses of cannabis, and sets out techniques for producing potent, resinous ganja, strongly implying the ‘sinsemilla technique’ of roguing out male plants. Its compilers were Nath Shaiva alchemists whose metier necessitated engaging with trade across the frontiers of India from Central Asia, the Himalaya, and Indian Ocean in order to acquire commodities such as mercury, coral, spices, and precious metals. Though most closely associated with Bengal, the siddha milieu of these Shakta yogi–pharmacists involved a world of power places and temples that extended from Sindh and Peshawar to Sumatra and Assam. This linked the Nath alchemists into trade networks such as those of Persian Muslims, Sogdian Zoroastrians, and Tibetans, and inevitably also with Ahom Tai culture at sites such as Kamakhya.
This interconnectedness across the subcontinent and its frontiers is why the cosmopolitan culture of the Nath siddhas at the particular node of India’s ‘economy of the sacred’ in 15th century Srisailam could produce a text such as the Ānandakanda – and just as ganja was on the brink of becoming a truly global crop and commodity. How maritime trade across the Bay of Bengal during the Age of Sail (mid-C16 to mid-C19) may have further shaped the cannabis of the East Indies is set out at the introduction for our accessions from Southeast Asia. Suffice to say for now that preliminary genome trees suggest close genetic connections between Chinese landraces, Himalayan multipurpose landraces, and the tropical and subtropical ganja landraces of South and Southeast Asia, so botanical reality may yet prove to align with the picture outlined here. For drug-type cannabis, multiple independent domestications in and around China and Central Asia probably created several distinct landrace groups, closely defined by product and region, with some interesting relationships and overlaps that could yet further illuminate their history.
.
