The plant sets odds, not outcomes
It is worth being precise about what a cultivar does and does not decide, because both overstatements are common. A large-leaf assamica shoot carries more of the material that oxidised tea is built from and makes a better black tea than a delicate spring sinensis bud would; a shaded sinensis leaf grown for savoury depth makes a poor black tea and an excellent green one. Producers plant what suits what they intend to make, which is why the association between plant and category is strong in practice. But the association is commercial and agronomic, not chemical destiny. The same bush, picked on the same morning, can be sent down three different routes and will produce three teas that a drinker would not guess were related. That is the fact that separates tea from wine and from coffee as a subject, and it is why a tea catalogue organised around varieties would be organised around the wrong thing.
The clearest case: a Fujian oolong cultivar made three ways
Several of the aromatic Fujian and Taiwanese oolong cultivars are made as oolong, as a black tea and occasionally as a green tea, and the three are sold side by side. Made as a rolled oolong the leaf is sun-withered, shaken repeatedly with long rests, fixed at a moment judged by smell, compacted into pellets and lightly baked, and the cup is floral and thick with a long aftertaste. Made as a black tea the same leaf is withered for hours, rolled to rupture it comprehensively, oxidised to the end and dried, and the aromatic character that the cultivar is prized for reappears as a sweet, honeyed, fruit-driven note over a red liquor. Made as a green tea it is fixed within hours and shaped, and what survives is a clean, faintly floral cup with none of the depth either of the others has. Three teas, one plant, and the aromatic signature of the cultivar is legible in all three — which is exactly the point. The plant contributes a recognisable accent and the route contributes the language.
The Japanese case: one intake, four teas, and no route change at all
A second and quite different version of the same lesson. A Japanese grower's intake goes down one route and produces aracha — crude tea, correctly steamed, rolled and dried, and not yet a product. A refiner then cuts it to length, screens it, winnows it and pulls the stem out with an electrostatic separator, and what comes off that line is four teas: sencha from the leaf fraction, kukicha from the stem, mecha from the small rolled buds, konacha from the fines. Nobody changed the route. The separation happens at the end, and the four products differ because they are different parts of the same plant rather than because different decisions were taken. It is worth holding this alongside the oolong case, because the two are opposite demonstrations: there, one plant made three teas by three routes; here, one plant and one route make four teas by a sorting operation.
Where a cultivar genuinely does constrain the route
The honest qualification. Some routes place real demands on the material and a plant that cannot meet them is not a candidate. Shaded production wants a cultivar that responds to low light without stalling; ball-rolled oolong wants leaf with enough substance to survive hours of compaction; a route that ends in months of charcoal roasting wants leaf that will not collapse under it; and pu-erh's whole proposition rests on large-leaf material with enough in it to still be changing in twenty years. So the relationship is asymmetric and worth stating that way round. A route can be applied to almost any leaf and the result will be recognisably that route; whether it is any good depends on whether the leaf could take it. The plant does not choose the category, and it does set a ceiling on how well a given category can be executed.
Why this matters when you are buying
Two practical consequences. First, a cultivar name on a packet tells you about an accent, not about a style, and two teas from the same named cultivar can be entirely different products. Reading the route — fixation method, shaping, roast, drying — tells you far more about what will be in the cup. Second, when a producer offers the same cultivar as three teas, that is a genuinely useful thing to buy: it is the cleanest available experiment, because everything except the route is held constant. Very few subjects offer an experiment that clean at a domestic scale, and drinking three such teas side by side teaches more about processing than any amount of reading, including this.