Three kinds of evidence, three different questions
TeaHQ's own reading of well-attested general knowledge, with no single source behind it.
A written reference establishes that somebody, at some date, wrote a word. An excavated residue establishes that plant material was present in a particular place. A population-genetic analysis establishes how the cultivated plant relates to its wild relatives. Those are three different questions, and none of them is the question most readers want answered, which is when people began drinking tea as a habit. The gap between what each method can support and what it gets used to support is where nearly all confident early-tea chronology comes from. This guide is about the gap rather than about a date. TeaHQ takes the position that the honest summary of early tea is short: the plant is native to a broad zone across south-west China and the adjacent uplands of northern Myanmar, north-east India and northern Laos and Vietnam; people in that zone used its leaf for a long time before anybody wrote about it; and the transition from local use to a traded, taxed, written-about commodity happened somewhere in the first millennium of the common era and is much better evidenced at its end than at its beginning.
What a text can establish, and the word problem
TeaHQ's own reading of well-attested general knowledge, with no single source behind it.
The standard early Chinese references are cited constantly and are harder to use than their reputation suggests. The best known is a household contract, the Tong Yue, conventionally dated to the first century before the common era, which lists among a servant’s duties the buying and the boiling of something usually translated as tea. The difficulty is the character. Early Chinese writing used tu for a range of bitter plants, and the graph that unambiguously means tea is a later differentiation of it. So a pre-differentiation reference may be about tea, or about a bitter herb that is not tea, and the argument for reading it as tea is generally contextual — the place it is bought, the vessel it is boiled in, the company it keeps in a list. That argument is often good. It is not the same as the text saying tea. TeaHQ hedges every early textual date for this reason, and takes the eighth-century Cha Jing as the first point at which the written record is unambiguous, detailed and internally consistent enough to build on.
What excavated material can establish
TeaHQ's own reading of well-attested general knowledge, with no single source behind it.
Physical evidence is the strongest kind and the narrowest. Where plant material survives in a tomb or a storage vessel and can be identified, it establishes that the material was there — which is a real and valuable fact, and is not a fact about consumption, preparation or frequency. A funerary deposit tells you what was thought worth burying. It does not tell you what people drank on an ordinary day, and grave goods are systematically unrepresentative because they are chosen. Vessel typology is weaker still. A bowl of a shape later used for tea is not evidence of tea unless something was in it, and the history of tea equipment is full of forms borrowed from wine and from food service. Reports of early tea residues are periodically published and periodically over-read in general accounts, so TeaHQ records the shape of the finding — material identified, place, approximate period — and declines to convert it into a claim about when tea drinking began.
What the plant’s genetics can establish
Attested in the historical record, and sourced to it.
Genomic work on Camellia sinensis answers questions the other two methods cannot touch, and answers none of the questions about human practice. A haplotype-resolved assembly of the tea plant published in Nature Genetics reports population-genomic analysis across a large panel of Camellia accessions and finds independent evolutionary histories and parallel domestication in the two widely cultivated varieties, var. sinensis and var. assamica, together with extensive introgression contributing to the diversity of modern cultivars. That is a substantial result for tea history: it says the small-leaf and large-leaf cultivated plants were not one domestication that later diverged, and it says modern cultivars carry material from more than one lineage. What it does not say is when anybody drank anything. Domestication is a process visible in the plant; drinking is a process visible only in texts and artefacts. Reading a domestication result as a date for tea culture is the commonest error in popular accounts of this literature.
The legend, and why it is durable
Repeated very widely and not evidenced. Recorded because the story is itself part of the culture, not because it happened.
The story that the emperor Shennong discovered tea when leaves blew into his boiling water is repeated in nearly every introduction to the subject, and it is a legend rather than a record: Shennong is a mythological culture hero credited with agriculture and with the medical literature generally, the story is attached to him long after the period it describes, and there is no evidential path back to an event. TeaHQ says this in the sentence that tells the story, rather than in a footnote, because the story is almost always told in the same voice as the checkable parts of the account. Its durability is worth understanding. A single named discoverer, an accident, a moment — that shape is memorable, quotable and fits on a packet, and the real answer, which is that a plant in a large forested zone was used by a lot of people over a long time in ways nobody wrote down, does not. `guide-tea-origin-myths` collects this and five relatives and says what each is worth.
What TeaHQ will and will not say about early tea
TeaHQ's own reading of well-attested general knowledge, with no single source behind it.
It will say that the plant is native to a zone straddling several modern borders, that both cultivated varieties have long independent histories, and that use of the leaf in that zone certainly predates any surviving description of it. It will say that the written record becomes usable in the Tang period and that the earliest unambiguous technical account is the Cha Jing. It will not supply a date for the first cup, a country of invention, or a single centre of origin, because the first is unknowable, the second is a modern political question rather than a historical one, and the third is contested among people who work on the plant directly. Where a source elsewhere in this catalogue gives a firmer answer than this one, that is a difference in confidence rather than in knowledge, and a reader is better served by knowing which parts of the story rest on what. Every dated statement in the rest of this file is written on the same principle.