cháyè xiāngxíng

香型 (xiāngxíng) literally means “aroma type.” It is a working axis of the Chinese tea trade and tasting description: a way to say in one word which direction a tea’s smell and taste is oriented, without listing the entire set of notes. When a seller says “this is 清香型,” they are communicating not a cultivar, not a region, and not a price, but an approximate aroma zone, behind which stands a specific technology: degree of roasting, depth of oxidation, presence of aging, method of drying.

It is important at the outset to distinguish two different uses of the word 香型, because in Russian-language texts these are constantly confused.

The first is technological types common to many classes of tea: 清香 (qīngxiāng), 浓香 (nóngxiāng), 蜜香 (mìxiāng), 陈香 (chénxiāng), 樟香 (zhāngxiāng), 枣香 (zǎoxiāng), 药香 (yàoxiāng), 松烟香 (sōngyānxiāng). They describe the result of processing and storage, and in principle apply to different raw materials.

The second is varietal aromatic types within a single group of teas, where 香型 denotes the strain of the bush and its inherited aroma. The classic case is Fenghuang dancong (凤凰单丛, fènghuáng dāncōng): the monograph 《凤凰单丛》 (chief editors 黄瑞光 and 桂埔芳, China Agriculture Press, 2020) lays out the strains into ten canonical types — 黄枝香 (huángzhīxiāng, also spelled 黄栀香 in another orthography), 芝兰香 (zhīlánxiāng), 蜜兰香 (mìlánxiāng), 桂花香 (guìhuāxiāng), 玉兰香 (yùlánxiāng), 姜花香 (jiānghuāxiāng), 夜来香 (yèláixiāng), 茉莉香 (mòlìxiāng), 杏仁香 (xìngrénxiāng), 肉桂香 (ròuguìxiāng) — plus a section on newly bred cultivars. Another reference on the same material, the “Atlas” by Chen Shaoping, gives seventeen aromatic sections — the same ten plus seven secondary groups (黄茶香, 柚花香, 橙花香, 杨梅香, 附子香, 苦种, “other 清香”) — plus a supplement (增录) as an eighteenth section; the supplement collects entries not assigned to aromatic groups and is not itself an aromatic type. That is, even within one narrow tradition, the number of types and their boundaries depend on the source.

Further in the article, the discussion is mainly about the first, technological axis — the one by which tea is described in the broad trade.

The Nature of the Scale: Descriptive, Not Measurable

An aromatic type is not the result of an instrument. It is a sensory category assigned by a person, and assigned by a combination: the smell of the dry leaf, the smell of the warmed vessel, the aroma of the infusion, the aftertaste, sometimes the color of the dry leaf. From this follow consequences worth keeping in mind.

The boundaries between types are blurred and mobile. One and the same Wuyishan oolong may be called 中火 (zhōnghuǒ, medium fire) by a seller oriented toward light roasts, and 轻火 (qīnghuǒ, light fire) by a seller of the traditional school, and the verbal description will shift from “dense” to “clean.”

Some types are fixed in industry normative documents as official subcategories, and there are more such cases than is commonly thought. For Anxi tieguanyin (安溪铁观音, ānxī tiěguānyīn), products are divided by the method of finishing into three types — 清香型, 浓香型, and 陈香型; the third was added by the latest edition of the relevant standard, and it has its own prescribed technological process and its own sensory indicators. For Taiwanese-style (台式) oolong, the normative division is also threefold — 清香型, 熟香型, and 蜜香型, with 蜜香型 having both a definition in the text of the product standard and a separate technological line in the production standard. Finally, 槟榔香 (bīnglángxiāng, “betel nut aroma”) is defined in the standard of terms for sensory evaluation of tea — as an aroma that arises in liu bao after aging. The specific standard numbers are not given in the text: they should be checked against the current editions, not from memory. For 樟香, 枣香, and 药香, however, we found no normative definitions — they are absent from the standard of terms for sensory evaluation, and this is purely market language.

One tea may carry several types at once, and in descriptions they are combined: “陈香带药香,” “枣香浓郁, 略带陈味.” A type is a dominant, not the only note.

清香 — Clean, Fresh

What it means. Light, high, “green” or floral aromatics: fresh greens, flowers, milky and creamy shades, sometimes a note of freshly cut grass. The infusion is light, yellowish-greenish, the taste is light, with pronounced freshness.

What causes it. A low degree of oxidation (for oolongs) and light roasting or its absence. In Taiwanese high-mountain oolongs (高山茶, gāoshānchá) and in modern Anxi tieguanyin, the 清香 line results from a combination of short 做青 (zuòqīng, shaking and withering), light drying without prolonged 焙火 (bèihuǒ, heating over fire), and sometimes cold storage of the finished tea. In green teas, aroma of this kind is the norm, and the term is applied to them less often because it is not distinguishing.

Where it occurs. Oolongs of Taiwan and Anxi; some lightly roasted Wuyishan oolongs; some dancongs that did not make it into the ten in the reference books and are described simply as “other 清香”; conditionally — green teas.

What it does not mean. It does not mean “higher quality” or “more natural tea.” It does not mean a young harvest: 清香 is a consequence of processing, not of leaf age. It does not mean suitability for long storage: light roasts usually lose their freshness faster than dense ones.

浓香 — Dense, Strong Roast

What it means. Dark, dense, roasted aromatics: caramel, dried fruit, nuts, baked goods, toasted rice, sometimes cocoa. The infusion is darker, amber or reddish-brown, the body is denser, bitterness and astringency are smoothed.

What causes it. Roasting — prolonged or repeated, with charcoal (炭焙, tànbèi) or in an electric oven. During heating, caramelization and Maillard reactions take place; part of the original floral volatiles leaves, new, “baked” ones appear. In Wuyishan oolongs, the stages are designated as 轻火 / 中火 / 足火 (zúhuǒ, full fire). In Anxi tieguanyin, 浓香型 is one of the three normative subcategories; historically it is closer to how tea was made before the fashion for light lines.

Where it occurs. Wuyishan rock oolongs (武夷岩茶, wǔyí yánchá), tieguanyin of the 浓香 line, traditionally roasted dancongs, some Taiwanese oolongs (凍頂 and variants reworked toward 熟香), some black teas.

What it does not mean. Roasting does not “fix” poor raw material, although it masks some defects — which is also why it is harder to judge raw material from 浓香 than from 清香. It does not mean burnt: scorching (焦味, jiāowèi) is a defect, not a degree of the type. It does not mean the tea is automatically old: a heavily roasted tea from last month is still a young tea; it is usually allowed to rest so that the “fire leaves” (退火, tuìhuǒ).

蜜香 — Honey

What it means. Honeyed, berry-honey, sometimes fruity sweetness in the aroma and aftertaste, often with a sensation of thick, “sticky” sweetness in the infusion.

Normative status. Contrary to a common notion, 蜜香 is not only a market word. For Taiwanese-style (台式) oolong, 蜜香型 is an official subcategory on a par with 清香型 and 熟香型: the product standard defines 蜜香型 as a type whose aroma carries a honey-fruity note (茶叶产品香气带有蜜果香), and divides the entire product range into exactly these three types. The technology standard provides a separate processing line for 蜜香型 — with deeper withering (weight loss of 20–30 %) and a high degree of fermentation (60–65 %). The closest parallel in meaning on the mainland is the description of 东方美人 as a Taiwanese honey-type oolong. For the other technological types in this article, apart from 清香 and 浓香, we found no normative definitions.

What causes it. By two different routes, which are often mixed up.

The first is leaf damage by the small green leafhopper (小绿叶蝉, xiǎo lǜ yè chán). The leaf reacts to the bites, the composition of its aroma precursors changes, and upon subsequent strong withering and oxidation the characteristic honey note appears. Taiwan’s 东方美人 (dōngfāng měirén) and Taiwanese honey black teas 蜜香红茶 (mìxiāng hóngchá) are built on this principle; part of such notes in some mainland teas is explained by the same logic. The raw material for such teas is not treated against the insect — otherwise the mechanism does not start.

The second is simply the degree of oxidation and the withering regime. Honey notes regularly arise in black teas and heavily oxidized oolongs without any participation of the leafhopper. It is precisely this route that is described in the technological standard for Taiwanese oolong: deep withering plus high fermentation.

Where it occurs. Taiwanese oolongs and black teas, some mainland black teas, individual oolongs with a high degree of oxidation. In dancongs there is a varietal type with a similar name but a different nature, 蜜兰香 — there it is the inherited aroma of the strain, not a consequence of bites.

What it does not mean. A honey aroma does not prove that the tea was bitten by a leafhopper — this is the most frequent semantic distortion around 蜜香. It also does not mean honey was added. It does not mean high mountain: the link between altitude and the honey note is not direct. The existence of the normative subcategory 蜜香型 for Taiwanese oolong also says nothing about the mechanism: the standard sets the technology and the sensory profile, not the participation of an insect.

陈香 — Aged

What it means. The aroma of old tea: dry wood, the dust of a dry warehouse, a bookcase, dry leaves, sometimes a light mushroomy or earthy note. The taste is softer and smoother than in a young tea of the same type, astringency is dulled.

What causes it. Time and storage conditions. Slow oxidative and microbiological processes rework the original substances; in pu-erh this is described by the principle 越陈越香 (yuè chén yuè xiāng, “the older, the more aromatic”), on which the entire collecting logic of sheng pu-erh rests. A distinction is made between 干仓 (gāncāng, dry warehouse) — the slow, clean, safe path — and 湿仓 (shīcāng, wet warehouse) — rapid “aging” with the risk of mold and spoilage; historically, wet storage is linked with Hong Kong practice (港仓). For shu pu-erh, 越陈越香 also works, more gently, and over the years yields a dense 醇厚陈香.

Where it occurs. Sheng and shu pu-erh, liu bao (六堡茶, liùbǎochá) and other heicha, aged white tea (老白茶, lǎobáichá), aged oolongs, and Anxi tieguanyin of the 陈香 line. The last case deserves emphasis: for Anxi tieguanyin, 陈香型 is not only a market label but also a normative product subcategory, introduced in the latest edition of the relevant standard, with its own finishing process and its own sensory indicators.

What it does not mean. It does not mean a specific age: “陈香” is a quality of aroma, not a number of years, and it appears quickly under wet storage, while under dry storage it may not appear over the same time. It does not authenticate the label: the eras 号级茶 (private trading houses before the 1950s), 印级茶 (zhongcha of the 1950s–70s, with the 中 stamp in various colors), 七子级 (from the 1970s, the era of the article codes 唛号, màihào) are reference points for attribution, not a consequence of smell. It also does not mean the absence of defects: mustiness, moldy and “wet” smell (霉味, méiwèi) do not belong to 陈香, although on the market they are sometimes presented that way.

樟香 — Camphor

What it means. A cool, menthol-woody note compared to the camphor tree; in the infusion it gives a sensation of freshness “in the throat.”

What causes it. Reliably — not established. The most widespread explanation in the trade: old tea trees grow in mixed forest next to camphor trees (樟树, zhāngshù), and the leaf supposedly takes on their aroma. This explanation is popular but not causally proven; at least some of the cases are more simply linked to prolonged aging of sheng pu-erh under dry storage, where cool woody-menthol notes appear on their own.

Where it occurs. First of all aged sheng pu-erh; sometimes mentioned for individual heicha.

What it does not mean. It does not confirm the age of the trees, nor their “wildness,” nor a specific locality. It does not mean camphor or any flavorings were added. As an independent marker of the authenticity of old pu-erh, this note does not work. There is no normative definition for 樟香 — it is not described in the standard of terms for sensory evaluation.

枣香 — Jujube

What it means. The sweet aroma of dried Chinese jujube (red date), sometimes with a shade of dried fruits and a light “cooked” sweetness.

What causes it. Technologically — dense aging of heavily compressed or coarse raw material. It is regularly described in aged shu pu-erh and especially in pressed forms of liu bao and adjacent heicha, as well as in aged white tea from coarse leaf — above all in pressed 寿眉 (shòuméi), where 枣香 has become almost a mandatory item in commercial descriptions. The exact chemical nature of the note is not covered in this article.

Where it occurs. Shu pu-erh, heicha, aged white tea (mainly pressed and from coarse leaf).

What it does not mean. It does not mean anything was added to the tea. It does not mean a specific aging period: in trade, “枣香” is often claimed for teas of different ages. It does not mean a high grade of raw material — rather the opposite: this note is more often linked with coarse leaf than with bud-picked material. There is no normative definition for the term.

药香 — Medicinal, Herbal

What it means. The aromatics of a pharmacy and Chinese herbs: roots, dried herbs, licorice, sometimes balsamic and slightly bitterish shades.

What causes it. Prolonged aging, more often in teas with post-fermentation or a high degree of oxidation. It appears in old sheng and shu pu-erh, in aged liu bao, in old white tea, and in heavily aged oolongs. In liu bao, a close semantic neighbor is 槟榔香 (bīnglángxiāng, “betel nut aroma”), and it is the only one of the neighboring terms that has a normative definition: the standard of terms for sensory evaluation of tea describes 槟榔香 as an aroma similar to betel nut that arises in liu bao after aging. 药香 itself, however, has no normative definition.

Where it occurs. Aged pu-erh, liu bao and other heicha, 老白茶, aged oolongs.

What it does not mean. It does not mean the tea has medicinal properties — this is a description of smell, not pharmacology; no medical claims follow from 药香. It does not mean herbs or ingredients were added. It does not always mean a virtue: in some schools of description, 药香 is considered a desirable sign of maturity; in others, a sign that the tea has over-sat or was stored too wet.

松烟香 — Pine Smoke

What it means. A distinct smoky note: pine resin, a campfire, smoked goods, sometimes “lacquered” wood.

What causes it. Direct contact of the leaf with smoke during drying or re-drying over pine (less often other) firewood. This is the only type among those listed whose cause is fully technological and unambiguous. The classic case is the smoky version of zhengshan xiaozhong (正山小种, zhèngshān xiǎozhǒng) and its smoked derivatives; smoky notes also traditionally arise in a number of Hunan heicha, where drying was carried out over an open fire with pine firewood. In modern versions of xiaozhong, the smokiness is often reduced or removed entirely, so a “lapsang” without smoke is not a counterfeit but a different production line.

Where it occurs. Black teas of the Wuyishan xiaozhong line, some heicha, individual local green and yellow teas with fire drying.

What it does not mean. It does not mean flavoring, nor a defect. It does not mean high or low quality by itself: smoke can both emphasize good raw material and mask weak material. The smoky note does not “air out” completely, but over the years it usually becomes softer and recedes into the background.

Summary Table

Type	Main cause	Where typical	Frequent false interpretation
清香	weak oxidation, minimal roasting	Taiwanese and Anxi oolongs, some Wuyishan	“fresher = better,” “young leaf”
浓香	prolonged or strong roasting	rock oolongs, tieguanyin 浓香型, dancongs	“old tea,” “scorching”
蜜香	leafhopper bites or deep withering with high fermentation	东方美人 and other Taiwanese 蜜香型 oolongs, Taiwanese honey black teas, black teas	“necessarily leafhopper,” “honey added,” “purely market word”
陈香	aging, warehouse regime	pu-erh, heicha, 老白茶, old oolongs, tieguanyin 陈香型	“proof of age and authenticity”
樟香	not established; linked to aging	aged sheng pu-erh	“grew next to a camphor tree” as fact
枣香	aging of coarse and pressed raw material	shu pu-erh, heicha, pressed 寿眉	“jujubes added,” “sign of elite raw material”
药香	long aging, post-fermentation	old pu-erh, liu bao, 老白茶	“medicinal tea,” “herbs added”
松烟香	drying over pine firewood	xiaozhong, some heicha	“flavoring” or “defect”

Natural Aroma, Scented, and Added

These three things differ not in the strength of the smell but in provenance, and they should not be mixed.

Natural aroma is the result of the bush cultivar, growing conditions, and processing. All eight types above belong here, including the smoky one: smoke is a production method, not an addition to the finished tea. The varietal types of dancong also belong here, where “桂花香” or “茉莉香” means that the leaf’s aroma resembles osmanthus or jasmine, although neither flowers nor extracts are in the tea. This is a constant source of misunderstanding: 蜜兰香 dancong contains neither honey nor orchids.

Scented tea — 窨制 (yìnzhì): finished tea is repeatedly layered with fresh flowers, the flowers give their aroma, then they are sifted out, the tea is re-dried, and the cycle is repeated as necessary. This is how jasmine tea (茉莉花茶, mòlì huācha), osmanthus tea (桂花, guìhuā), and a number of others are made; for osmanthus tea there exists an industry standard, as there is for jasmine tea. This is a traditional and fully open technology: the tea is honestly called flower tea, and the number of scenting cycles is an ordinary quality parameter. The number of layers, flower consumption, and process durations differ by producer, and figures should not be given without reference to a specific standard.

Added aroma is the flavoring of finished tea with essences, essential oils, extracts. This is a separate product category; it does not belong to 香型 in the original sense, and designating it with terms like “蜜香” is incorrect. There are no reliable everyday ways to distinguish flavoring from natural aroma “by eye”; indirect signs are an unnatural sameness of smell from infusion to infusion, a mismatch between the aroma of the dry leaf and the infusion, and a sharp smell that does not change as it cools. Indirect signs remain indirect: they are a reason for a question, not a conclusion.

Why the Boundaries Wander Among Different Sellers

There are several reasons, and all of them are systemic, not a consequence of anyone’s dishonesty.

Some types are normatively fixed, but unevenly: for Anxi tieguanyin and Taiwanese oolong the division into types is written into standards; for liu bao the term 槟榔香 is fixed; but for 樟香, 枣香, and 药香 there are no normative definitions. The rest is the language of custom, and there the boundaries are held only by established practice.

The terminology is regionally heterogeneous. The Chaozhou, Wuyishan, Anxi, and Yunnan schools of description developed separately; the same word in them may rest on different reference points. Even the orthography wavers: 黄枝香 and 黄栀香 are the same dancong type in two reference works.

The reference points of tasters are personal. A type is assigned by comparison with experience, and experience differs; two people of equal qualification need not agree on the boundary between 中火 and 足火.

The market influences the language. When 蜜香 or 枣香 become selling words, the frequency of their use grows faster than the frequency of the teas that actually correspond to them. The existence of a normative subcategory does not prevent this: a standard describes a specific product of a specific category, while the word circulates across the whole market.

The practical conclusion is simple: an aromatic type is useful as a first filter — it says in which direction the tea is oriented and allows one to cut off what is clearly not that. It is neither a guarantee of quality, nor proof of origin, nor an identifier of a specific tea.

What This Article Does Not Claim

On chemistry. It does not give the specific volatile compounds responsible for each type, nor their concentrations. For some types such data exist, for others they are fragmentary; they should be checked against specialized literature, not trade descriptions.

On standards. The numbers and editions of national and industry standards are deliberately not given in the text, but the normative facts themselves are named: for Anxi tieguanyin, the division into 清香型, 浓香型, and 陈香型 is fixed, with 陈香型 added by the latest edition of the relevant standard; for Taiwanese-style (台式) oolong, the division into 清香型, 熟香型, and 蜜香型 is fixed, with 蜜香型 having both a definition in the product standard and a separate process in the production standard; 槟榔香 is defined in the standard of terms for sensory evaluation of tea as applied to liu bao. For 樟香, 枣香, and 药香 we found no normative definitions — they are absent from the standard of terms for sensory evaluation. Exact references and current editions in any case require verification against the documents in force.

On the leafhopper. The link between leaf damage by the small green leafhopper and the appearance of the honey note is recognized and lies at the basis of individual production lines, but the quantitative side (what degree of damage yields what result) is not described here, and the reverse assertion — “every 蜜香 comes from a leafhopper” — is false. The normative definition of 蜜香型 describes the aroma and the technology, not the insect.

On the camphor note. The explanation through proximity to camphor trees is presented as widespread in the trade, not as an established fact. The mechanism of 樟香 is not treated in the article as clarified.

On age and storage. Year of production, region, and authenticity cannot be derived from aroma. The correspondence of eras (号级, 印级, 七子级, modern) belongs to attribution by packaging, article code, and the history of the batch, not to sensory analysis. Auction price levels for rarities are not discussed in the article because they do not belong to 香型.

On health. 药香 is a description of smell. No medical, preventive, or therapeutic properties of tea follow from the presence of this note, and none are claimed in the article.

On the number of types. The eight listed are the most commonly used part of the scale, not a complete list. Individual traditions have their own sets: the ten canonical varietal types of dancong according to the monograph 《凤凰单丛》 (2020), and seventeen aromatic sections in Chen Shaoping’s “Atlas” plus the supplement 增录 as an eighteenth section, 槟榔香 in liu bao, local terms in descriptions of individual teas. There is no universally accepted closed nomenclature of 香型.

On persons and productions. The information about the authors, given as the provenance of the sources, is taken from the imprint data of the monograph: its chief editors are 黄瑞光 and 桂埔芳; among the deputies is 黄柏梓, who, according to the publication’s reference, surveyed 5120 bushes and identified 123 名丛 (míngcóng, “named bushes”). There are no other names, estates, quotations, or figures in this article, and they should not be attributed to it.

the teas described here are available from teamotea