The two stories, told as stories
Repeated very widely and not evidenced. Recorded because the story is itself part of the culture, not because it happened.
The first has a picker or a hunter distracted — by a snake, by a deer, by an errand — so that the leaf is left in the basket and bruises on the way home, and the resulting partly oxidised tea turns out to be delicious; in the Anxi version the man is called Wulong and the tea is named after him. The second has a passing army occupying a tea factory in the Wuyi hills at the height of the season, the leaf sitting far too long, and the maker drying it hurriedly over pine fires to salvage it, producing the first smoked black tea. Both are repeated in vendor copy, in books and on packets, and neither is evidenced: they surface in the written record much later than the periods they describe, they have no named source, and they share the structure of a folk tale rather than of a record. TeaHQ tells them here as legends because they are unavoidable, and labels them in the sentence rather than in a caveat.
Why the accident story is structurally suspect
TeaHQ's own reading of well-attested general knowledge, with no single source behind it.
Consider what would actually have to happen. A batch spoils in a novel way; someone tastes it rather than discarding it; someone identifies which of several uncontrolled variables produced the result; someone reproduces it deliberately, repeatedly, and at a quality worth selling; and a district then adopts the method. That is not an accident, it is a development programme with an anecdote at the front of it. The same objection applies to every accidental-invention story in food and drink, and the reason they persist is that they are memorable and they flatter the product with luck as well as skill. What the accident stories are actually recording, if anything, is that the boundary between categories is a matter of degree — leaf left longer oxidises further, and the difference between a green, an oolong and a black tea is in large part a difference in how long and how far a maker lets that go.
What is reasonably well established instead
TeaHQ's own reading of well-attested general knowledge, with no single source behind it.
That partial oxidation as a controlled technique, with withering, repeated bruising of the leaf edge, and firing, developed in the Fujian hills and in the Wuyi range and spread from there to Anxi, to Guangdong and to Taiwan. That the Wuyi area is the source both of the rock oolongs and of the earliest black teas. That the technique reached Taiwan with Fujianese migration in the nineteenth century and developed there into a distinct tradition. And that these developments belong to the Ming and Qing periods, in a general sense that TeaHQ is confident about and cannot date precisely. That is a much less quotable account than the snake and the soldiers, and it has the advantage of being about the thing that actually determines what is in the cup: a set of decisions about withering time, bruising intensity, firing temperature and roast. It is also well established that the two categories are commercially entangled from the start. The same hills, often the same makers and sometimes the same gardens produced both, and the export demand that built the black tea trade drew on labour and material that would otherwise have gone into oolong. Treating them as separate traditions with separate origins is a consequence of how they are sold abroad rather than of how they were made.
Smoke was a drying method before it was a flavour
TeaHQ's own reading of well-attested general knowledge, with no single source behind it.
The pine-smoke character of the Tongmu-area souchongs is now the defining feature of a style, and it began as a solution to a problem: in a damp mountain climate at the end of a season, a maker who cannot get leaf dry will lose it, and a pine fire under a drying loft is what is available. Smoke follows. What is genuinely interesting is the second half of that history — that a technical necessity became a flavour the export market wanted, so it was intensified and standardised, and then largely rejected by the domestic Chinese market and by the modern speciality trade, so that unsmoked or lightly smoked versions of the same tea are now the prestige product in its home district. A style can therefore travel from accident to signature to embarrassment inside one tea, and the marketing story stays fixed at the first stage. The same trajectory is visible in other teas and is worth naming as a general pattern. A necessity becomes a signature, the signature is intensified for an export market, the export market moves on, and the intensified version survives abroad as the definition of a style that its home district has already abandoned. Anyone buying a smoked souchong should know they may be buying a nineteenth-century export specification rather than a local one.
What the legends cost a buyer
TeaHQ's own reading of well-attested general knowledge, with no single source behind it.
Two things, both practical. First, an origin legend crowds out the information that would help: which cultivar, which altitude band, how heavy the roast, how recent the firing. A vendor who spends the description on the snake has not spent it on the roast. Second, the legends encourage a false picture of category boundaries as accidents of history rather than as continuous variables a maker controls. A reader who understands that an oolong is defined by choices about oxidation extent and roast, and that those choices are made deliberately for a market, can taste comparatively and learn something. A reader who understands that an oolong happened because somebody saw a deer cannot. `guide-tea-origin-myths` makes the same argument across the whole family of tea legends. There is a third cost, and it is to the makers. A story that attributes a category to an accident implicitly denies the skill of the people who developed it, and the districts concerned have generations of accumulated technical knowledge that the accident story writes out entirely. Correcting the legend is not only a matter of accuracy; it puts the work back where it was done.