Editorial policy

TeaHQ describes tea. It does not tell you what tea will do to you. Everything published here is either transcribed from a record that can be checked, or clearly marked as a synthesis of common practice — and where we do not know something, the page says so rather than filling the gap.

The most useful thing about this site is not what it contains. It is that the absences are deliberate and labelled, so you can tell the difference between something we have checked and something nobody has written down.

Every factual field on a TeaHQ page comes from one of two places, and the distinction runs through the whole catalogue.

External evidenceis a document somebody outside TeaHQ published — a standards body, a research paper, a producer’s own documentation. Where a claim rests on one, we record which document, how thoroughly it was checked, and whether anyone actually read the passage being relied on rather than just the abstract.

TeaHQ synthesis is our own account of widely-attested practice, written by us and attributed to us. Most brewing parameters are synthesis: nobody has published a controlled study of how long to steep a Dan Cong, and the honest description of what we offer is informed common practice rather than a measurement. Where a parameter set is synthesis rather than recorded practice for that specific tea, the brew assistant says so on the page.

Synthesis can describe what people do. It cannot certify what a document says. A TeaHQ synthesis is never used as evidence that some other source made a claim.

The sources and evidence page explains the classes and tiers in more detail.

Parts of the catalogue exist in the model and are not published. That is a decision rather than an unfinished job, and there are two main cases.

Health and functional claims.TeaHQ holds a set of records about tea and health — what the evidence does and does not support — behind a human review checkpoint that has not been signed. Until a named person takes accountability for them, they stay unpublished. There are no “tea for sleep” pages, no functional benefit pages, and no health filters in any tool, because there is no field in the model that records what a tea is for.

Herbal infusions. Chamomile, rooibos and peppermint are not tea — they are not Camellia sinensis and share none of its chemistry. They carry their own safety model, including interactions and contraindications, and that surface is held to the same review requirement. Two herbal records exist in the catalogue and neither is published.

Individual records that fail a check. A handful of tea traditions are written and unpublished because their historical claims rest on sources nobody has verified to the required depth. Publishing them with a hedge would be worse than holding them.

TeaHQ publishes a caffeine estimate for 11 of 261 teas. That is not an oversight, and it is the single most common question we cannot answer.

A caffeine figure without its basis is decoration. The same leaf brewed two ways differs several-fold, so any number we publish has to state the leaf weight, the water volume, the temperature, the steep time and which infusion it refers to. Meeting that standard requires either a measurement or a citable source per tea, and we have neither for most of the catalogue.

We will not estimate caffeine from a tea’s category. Category predicts caffeine poorly — matcha and gyokuro measure above many black teas, and white tea is not low-caffeine despite near-universal belief that it is. A figure derived that way would look authoritative and be undetectably wrong.

The caffeine rule is one instance of a general one. TeaHQ does not derive a fact about a specific tea from a generalisation about its category, its origin or its price.

The same discipline shows up in the brewing tool. Parameters for one brewing style are never calculated from another: choosing gongfu for a tea that only has western parameters returns nothing at all, with an explanation, rather than a scaled number. Five grams for twenty seconds and three grams for three minutes are different extractions, not different sizes of the same one, and a plausible-looking wrong answer is worse than no answer.

A great deal of tea writing repeats founding stories as fact. Da Hong Pao’s robed monkeys, Tie Guan Yin’s dream of the goddess Guanyin, and the emperor Shennong discovering tea by accident are all told throughout the trade as though they were recorded events.

Where a legend matters to how a tea is understood, we include it and label it as tradition rather than record. We do not suppress it and we do not repeat it as history. A test in our build enforces this: a history note mentioning a legend must also mark it as one.

Cultural practices are described as practice — what people do, where, and now — rather than as history, unless the historical claim can be sourced. Where a tradition has regional variants, the page says which one it is describing instead of presenting one region’s form as definitive.

Uncertainty is a property of a field rather than of a page, so it is recorded field by field rather than with a blanket disclaimer.

• Brewing parameters are marked as recorded practice or as synthesis, and the brew assistant shows which.
• Oxidation is recorded as a descriptor, never a percentage. Oxidation figures are quoted everywhere and measured almost nowhere.
• Where a name is widely misapplied — Lao Ban Zhang, zheng yan, lao cong, Jin Jun Mei — the tea’s page says so, because the substitution is more useful to a buyer than the description.
• Where a claim is contested, both positions are stated. We do not resolve a live disagreement by picking one and staying quiet about the other.
• Where a field is empty, the page says it is not recorded rather than omitting the row.

If something here is wrong, we want to know, and we would rather hear about a small error than only about a large one.

Report it through the contact page. Include the page and, if you can, what the correct information is and where it comes from — a source we can check moves much faster than an assertion.

Factual corrections are made to the record itself, so every page reading that record is corrected at once. Where a correction changes the substance of what a page said rather than fixing a typo, the change is noted on the page rather than made silently.

TeaHQ describes 150 plants people drink and says almost nothing about what any of them does to a person. That is a deliberate structural decision rather than an omission waiting to be filled, and it is the single largest difference between this site and most writing about herbal drinks.

The reasoning is about evidence rather than caution. Describing what a plant is, where it grows, how it is prepared and what it tastes like rests on trade knowledge, cultural record and our own tasting. Describing what it does to a body rests on clinical and regulatory literature, and this project holds none of it. The source register contains 142 documents, of which 127 are taxonomic — authoritative about names and distribution and explicitly bounded from saying anything about physiology. Counting them as evidence for a health claim would be the exact error the register exists to prevent.

So the separation is enforced in the structure, not just observed in the writing. Health material lives in its own record type behind an unsigned human review checkpoint. The botanical schema has no field a safety claim could occupy, and a test asserts that no such field appears. Botanical prose may not assert that anything is safe — because “no safety information” and “safe” are different statements and only the first is one we can make. Where a plant’s search demand is overwhelmingly functional, the search register records a refusal rather than leaving a silent gap that looks like an oversight.

The cost is real and is not hidden: functional queries are the largest and most commercially valuable demand in this subject, and TeaHQ forgoes all of it. What you get instead is a catalogue where every page is the kind of claim we can actually support.

TeaHQ sells nothing, recommends no vendor, and carries no advertising or affiliate links. No producer, retailer or brand has paid for inclusion, placement or wording, and none has been offered the opportunity.

If that ever changes — an affiliate arrangement, sponsorship, or anything else with a commercial interest attached — it will be disclosed on this page and on every page it affects, before it takes effect rather than after.

TeaHQ is operated by HEKNO Ltd. Editorial decisions and the governance rules described here are set by TeaHQ rather than by any outside party.

Health-bearing content requires a named human to take accountability before it can be published, which is why none of it is published yet. More about the project.