The Whisk, the Foam, and the Emperor: Tea Battles of Song China

Tea History & Culture｜Author：祥興茶行｜Published：2026-07-15

Nine centuries before latte art, Song-dynasty tea lovers ground their tea to powder, whisked it into a snow-white foam, and competed over whose froth clung to the bowl longest. Here is how diancha worked — and what a modern lab learned by actually trying it.

Picture a contest. Two competitors each hold up a bowl of tea, and the first whose white foam collapses — whose bowl reveals a ring of waterline along the wall — loses. This is not a latte-art throwdown. It is doucha (鬥茶), "tea battling," the craze that swept Song China more than nine hundred years ago. Song people did not steep their tea. They whipped tea powder into a dense, milky foam in a full-blown performance of the hands — and the (reputed) emperor himself wrote an operating manual for it.

A Literary Gathering (Wenhui tu 文會圖), attributed to Emperor Huizong, shows scholars at an elegant tea party; along the bottom edge, attendants tend the brazier, mind the water, and whisk the tea. Ink and color on silk, National Palace Museum, Taipei.

From Boiling to Whisking: Six Steps Before the First Sip

In Lu Yu's Tang dynasty, tea went into a pot and was boiled (see [Lu Yu and the Classic of Tea](/articles?slug=lu-yu-cha-jing)). By the Song, the mainstream had shifted to diancha (點茶): put powdered tea in a bowl, "point" hot water onto it in controlled bursts, then beat it into foam with a bamboo whisk [1][2]. The Northern Song calligrapher Cai Xiang set down the whole procedure in his Chalu (茶錄, Record of Tea) — a book that predates the Daguan chalun by about half a century, written as a briefing for Emperor Renzong after Cai's posting in Fujian brought him face to face with the Jian'an tribute teas [2].

Following the Chalu, the tea-battle workflow breaks down into six steps [2][4]:

```mermaid flowchart LR A[Roast the cake dry the tea cake over heat] --> B[Grind mill it to fine powder] B --> C[Sift pass through a fine sieve] C --> D[Warm the bowl heat the cup first] D --> E[Whisk the tea paste, then pour and beat] E --> F[Judge inspect color and foam] E -.in detail.-> G[Make the paste: one qianbi of powder blended with a little hot water] G --> H[Add water in stages] H --> I[Whisk in circles raise the milky foam] ```

The details are exacting. The dose is "one qianbi of tea" — a qianbi being a small tea scoop, roughly 1.5 grams according to Chen Jinying's annotation [2][4], less than a modern teabag. The bowl must be preheated. Cai Xiang's own words: 「凡欲點茶，先須熁盞令熱，冷則茶不浮」 — whenever you would whisk tea, first toast the bowl until it is hot; if it is cold, the tea will not float [2]. Xie (熁) means to warm over fire. A bowl that is not hot enough will not hold a foam — the same logic as warming your cup before pouring latte art.

💡 A note on the word: the dian in diancha is not a fingertip dab. The scholar Li Kin Sum (Li 2022) argues it works like the dian in diantou, "to nod" — you tip the ewer in a quick bow to release one thin jet of water, then snap it upright to cut the flow, a rhythmic gesture repeated seven times per bowl [3]. Diancha, in other words, is a continuous choreography of pour, stop, whisk.

Seven Waters: One Bowl, Seven Pours, Seven Rounds of Whisking

The "Dian" chapter of the Daguan chalun, attributed to Emperor Huizong, divides the whisking into "seven waters" — seven pours and seven rounds of beating — and describes what the tea's surface should look like at each stage in poetic imagery [1][3][4]:

The seven waters of the Daguan chalun, round by round[1][3][4]

Pour ｜ The original image ｜ State of the surface

First ｜「疏星皎月」 sparse stars, a bright moon ｜ Lays the foundation of the foam

Second ｜「周回一線」 one thin ring, poured fast and cut fast ｜ Color begins to open — "pearls scattered in profusion"

Third ｜「粟文蟹眼」 millet grain and crab eyes ｜ The color six or seven parts achieved

Fourth ｜「輕雲漸生」 light clouds gathering ｜ The foam layer thickens

Fifth ｜「結凝雪」 congealing into snow ｜ The color fully revealed

Sixth ｜「乳點勃然」 milky curds surging up ｜ The froth mounds high (slow the whisk)

Seventh ｜「乳霧洶湧」 a billowing fog of milk ｜ Foam blankets the bowl and grips the rim — "biting the bowl"

Here is the catch: the Song originals give only these qualitative images. How fast should you whisk? How much water? The Song left no numbers. In 2022, a team at Hong Kong Baptist University led by Li Kin Sum did something rare in tea historiography — they ran hands-on reconstruction experiments, testing parameter sets one by one and showing you the results [3].

⚠️ Read this first: every number below comes from Li 2022 — a single modern reconstruction team's best results under its specific experimental conditions. These are not historical protocols recorded in Song sources. The Song originals offer only qualitative description and left no figures at all [3].

The Li 2022 team's most successful reconstruction parameters[3] ・Tea cake ground in a mortar about 600 strokes (the finer, the readier the foam), yielding 8–16 grams of fine powder ・Water starting at about 85°C and falling to roughly 50°C by the end (the "fish eyes and crab eyes" readiness stage measured at about 80–90°C — not boiling water) ・About 170 milliliters poured per bowl in total, averaging around 24 milliliters per pour ・Each pour whisked vertically about 70 strokes within 60 seconds (slowing for the sixth and seventh waters, following the Daguan chalun) ・The result: the thickest, most persistent milk-white foam — under those experimental conditions it covered the bowl for more than eight hours

The failures are just as interesting. The same team found that 120 milliliters and 270 milliliters both failed; only somewhere around 170 milliliters worked. Thirty whisk strokes were not enough, stretching a round to 90 seconds was too slow, and angled whisking lost to vertical [3]. Li draws a delicious conclusion: precisely because the variables are so many that nobody could reliably reproduce the best result, that unpredictability is the very essence of tea battling as a sport — doucha is fun, in other words, because it is hard [3]. The team even examined the spouts of Song ewers: a forward-curving, slender spout with a sharply tapered opening keeps the jet controllable and the cutoff clean — essential, because a single splashed droplet breaking the foam's surface makes a complete milky crown nearly impossible to rebuild [1][3].

The Tea Battle: First Water-Mark Loses

The primary source for the rules sits in the "Whisking Tea" section of Cai Xiang's Chalu. A properly whisked bowl must show 「面色鮮白、著盞無水痕」 — a surface fresh and white, leaving no water-mark on the bowl — to count as superb. Among the tea men of Jian'an, 「以水痕先者為負，耐久者為勝」: whoever's water-mark appears first loses, and whoever's foam endures wins, with the margin of victory counted as "one water apart, two waters apart" [2][4].

In plain terms, two criteria [3][4]. First, the foam should be white and even. Second, watch the bowl's wall. Foam that holds up and grips the rim without receding was called yaozhan (咬盞), "biting the bowl"; the moment it disperses, a ring of waterline shows along the inner rim — the shuijiao (水腳), or "water-foot." Su Shi has the very line in a poem: 「水腳一線誰爭先」, whose water-foot's thin line will show first? [4] First water-foot to appear, first to lose.

〔圖〕Diagram of biting the bowl versus the water-foot

A quick note on the word's paper trail. The earliest attestation of "doucha" appears in the Yunxian zaji (雲仙雜記), attributed to the Tang writer Feng Zhi: 「建人謂鬥茶曰茗戰」 — the people of Jian call tea battling "tea war." But that miscellany has been suspected of being a forgery ever since the Siku quanshu zongmu tiyao raised doubts (Yu Jiaxi, for his part, defended it). Bai Juyi's poems do confirm that competitive tea judging existed by the mid-Tang — just without the word "doucha." The careful formulation, then: tea battling "arguably traces to the Tang," but the etymological source itself is of disputed authenticity [4].

The standard for whiteness only tightened over time. Under Renzong, Cai Xiang ranked 「以青白勝黃白」 — bluish white beats yellowish white; by Huizong's reign, the Daguan chalun ruled 「以純白為上真，青白為次」 — pure white is the true summit, bluish white merely second [1][2][4]. This aesthetic even reached backward to dictate how tea was made. Because battling prized whiteness and a light, clean taste, the Beiyuan tribute workshops deliberately pressed the juice out of the leaf — the Beiyuan bielu says flatly, 「建茶惟恐其膏之不盡」: with Jian tea, the only fear is that the paste is not fully expressed, for paste left behind makes color and flavor heavy and murky [4]. White foam also needed a dark stage. The Daguan chalun declares 「盞色貴青黑，玉毫條達者為上」 — in bowl color, blue-black is prized, and bowls with jade-hair streaks running clear and unbroken are best [1]; Cai Xiang agrees: 「茶色白，宜黑盞」 — the tea being white, a black bowl suits it [2][6]. And so the black-glazed hare's-fur bowl became standard battle equipment. How that bowl came by its hare's fur and its yohen shimmer is another splendid story, saved for [Jian Ware and Tenmoku](/articles?slug=jian-ware-tenmoku).

A Tea Book Under the Emperor's Name

The Daguan chalun was completed in the Daguan era (around 1107): a preface plus twenty chapters, running from growing regions and harvest through utensils all the way to whisking, aroma and flavor, and storage-roasting. The preface has the author remark that "happening upon a day of leisure, I investigated the fine points," and the work was titled simply Chalun — the "Daguan" was pinned on by later generations [1][5].

But was the author really Emperor Huizong? Here the scholarship splits into two camps, and this article lays both out honestly. The Song shi's bibliographic treatise and the major catalogs never recorded the book, and doubts that Huizong wrote it himself go back a long way [5]. Li's 2022 study states plainly that he does not believe the emperor wrote the entire book alone, and leans toward reading it as a collectively compiled text — issued under Huizong's name with courtiers' hands at work [3]. Chen Jinying's 2018 study of tea battling, by contrast, follows the traditional account and simply calls it "imperially composed" [4]. The rigorous middle path is the one this article adopts: "attributed to Emperor Huizong" [1][3][4][5].

Li goes further and reads the book as a standardization project. The Northern Song ran a state tea monopoly, and by 1112 the monopoly revenue of Huizong's reign already exceeded four million strings of cash. By exalting Jian'an white tea, celebrating Jian ware, and laying down a single standard for whisked tea, the court dramatically raised tea's cultural and economic value [3]. And there is one detail to savor: this book, revered as the bible of tea battling, never once uses the word "doucha" [3]. It sets criteria and says nothing of winning or losing. Its posture is regulation, not sport.

Eastward Passage, and the Final Bow

Whisked-tea culture also crossed east along the Zen monastery routes. In the Southern Song, the Japanese monk Eisai made two journeys to China and carried tea seeds and the temple practice of drinking milled tea back to Japan, writing the Kissa Yojoki, Japan's oldest surviving tea book [8]. The tea rite of Jingshan Temple in Hangzhou — the "Jingshan tea ceremony" — is traditionally said to have reached Japan through pilgrim monks (one account has Enni Ben'en traveling to Jingshan in 1235; another has the Zen master Shomyo carrying bowls home east — the names in the records diverge, and no single figure can be settled on) [4][7]. But beware a common romantic shortcut: today's Japanese matcha ceremony is not a "living fossil" of Song diancha. It is an evolution and re-creation that took centuries, settling into its familiar form only with Sen no Rikyu [7][8] — a story we leave to [The Japanese Tea Ceremony and Rikyū](/articles?slug=japanese-tea-ceremony-rikyu).

In China itself, whisked tea headed for its exit. Under the Yuan, the standing of the literati fell and refined connoisseurship lost its soil; then in 1391, the twenty-fourth year of the Hongwu era, the Ming founder decreed 「罷造龍團，惟採茶芽以進」 — cease making dragon-cake tea; submit only picked buds. With caked tea cut off at the source, there was nothing left to grind, whisking withered, and steeped loose-leaf tea took the stage for good [4]. Chen Jinying offers a local footnote worth savoring: Taiwan's modern famous-tea competitions, run under the guidance of the Tea Research and Extension Station, can be read as a continuation and transformation of the old tea-battling spirit [4]. That is her interpretive claim, to be fair — but the next time you see the sealed label on a competition tea, spare a thought for that bowl in Jian'an, and the contest over whose white foam bit the bowl longest.

The Last Sip

🍵 Three sentences to take away

Song tea meant diancha: grind the tea to powder, warm the bowl, work a paste, then pour and whisk in stages until a milky foam rises. Battles were judged by whose foam "bit the bowl" longer and whose water-mark showed later — the firsthand rules come from Cai Xiang's Chalu [1][2][4].
The concrete seven-waters parameters (600 grinding strokes, 170 milliliters, 70 whisk strokes per pour, and so on) are the results of Li 2022's modern reconstruction experiment, not Song-dynasty fact; the originals offer only qualitative images like "sparse stars, a bright moon" [1][3].
Whether the Daguan chalun was written by Huizong himself remains contested — "imperially composed" and "collectively compiled under his name" both stand in the scholarship, and the official bibliographies never recorded the book. The rigorous phrasing: "attributed to Emperor Huizong, originally titled Chalun" [3][4][5].

References

（傳）趙佶（北宋，約 1107），《大觀茶論》（原名《茶論》，序＋二十篇），維基文庫校錄本，https://zh.wikisource.org/zh-hant/大觀茶論；ctext 本 https://ctext.org/wiki.pl?if=en&res=204389 （〈色〉篇「純白為上真」據 ctext 底本，一作「邑」為形近之訛）
蔡襄（北宋，皇祐年間撰、1064 刻石），《茶錄》，維基文庫校錄本，https://zh.wikisource.org/zh-hant/茶錄（〈點茶〉「一錢匕」從校注，通行本一作「七」為形近之訛）
Li, Kin Sum（李建深）(2022). "A Bowl of Good Tea in the Northern Song Dynasty: Using Modern Examples to Understand the Daguan chalun." _East Asian Science, Technology, and Medicine_ 54(1): 3–41. Brill. https://scholars.hkbu.edu.hk/en/publications/a-bowl-of-good-tea-in-the-northern-song-dynasty-using-modern-exam/
陳金英（2018），〈論宋代「鬥茶」茶俗－以宋代茶書為主的探討〉，《高雄師大國文學報》27 期，頁 67–109。https://www.airitilibrary.com/Article/Detail?DocID=18164692-201802-201805280011-201805280011-67-109
中國科學院自然科學史研究所（2003），〈《大觀茶論》作者問題的探討〉，agri-history 站點託管，http://agri-history.ihns.ac.cn/scholars/yxl/yxl40.htm
國立臺灣大學圖書館（策展），「宋代茶盞特展」展覽介紹，https://www.lib.ntu.edu.tw/events/song/intro.htm
中國非物質文化遺產網·數字博物館，「徑山茶宴」項目說明（2022 列入人類非遺），https://www.ihchina.cn/project_details/15309/
白井（武藏野大學博士論文），〈日本における「喫茶養生」の軌跡——なぜ栄西は「茶祖」と呼ばれるのか〉，https://www.musashino.ac.jp/mggs/wp/wp-content/uploads/2024/03/hakase_shirai_2shou.pdf