The Cookbook in Ancient and Medieval China

The Cookbook in Ancient and Medieval China

Donald Harper, University of Chicago

(paper prepared for the conference on “Discourses and Practices of Everyday Life in Imperial China,” Columbia University, October 2002)

This paper examines the evidence of cookbooks as a genre of technical literature during the period from the Western Han to the Tang dynasty. The cookbooks themselves are lost. However, the three chapters of Jia Sixie’s 賈思勰agricultural treatise, Qimin yaoshu齊民要術(composed ca. 540 A.D.), devoted to food preparation (chapters 7-9) include roughly 280 recipes ranging from fermentation techniques and condiments to stews, roasts, and noodles and breads. The recipes record with remarkable precision the ingredients, amounts, and preparation methods as well as the presentation of the finished dish and other aesthetic observations related to gastronomy. Moreover, it is probable that 160 of the 280 recipes are quoted from two older cookbooks: a Shijing食經(Culinary canon, 110 recipes; see Table 1); and a Shici食次(Culinary procedures, 50 recipes; see Table 2).[1] The term shijing “culinary canon” was a label applied to several categories of writing about food and diet in medieval times, and works including the term in their title are classified under medical literature in the Suishu隨書, Jiu Tangshu舊唐書, and Xin Tangshu新唐書bibliographic treatises, listed along with medical recipes, dietetics, and materia medica. The Qimin yaoshu recipes, in which there is no sign of medical or dietetic concerns, are clear evidence that one category of shijing was a guide to refined cuisine in the form of a precise written account of how to prepare it.[2]

Jia Sixie does not provide bibliographic details about the Shijing and Shici that he quotes. Most scholars date the works to slightly earlier, perhaps the fifth century, with the understanding that a number of recipes may go back to antiquity. In 1999, the excavation of the tomb of Wu Yang 吳陽(d. 162 B.C.) at Huxishan 虎溪扇, in Yuanling 沅陵, Hunan, revealed fragments of a bamboo-slip manuscript cookbook written in the same style as the Qimin yaoshu recipes. According to the preliminary report on the manuscript by Zhang Chunlong 張春龍and Guo Weimin 郭偉民, nearly 300 slip fragments have been recovered and there is not a single unbroken slip.[3] Each recipe in the manuscript is preceded by a number and a title, with 148 being the highest number among the extant fragments (title illegible). Zhang Chunlong and Guo Weimin estimate that the fragments account for only about one third of the original manuscript. Although there are no textual parallels between the Qimin yaoshu and the transcription of several fragments from Wu Yang’s cookbook that are provided by Zhang Chunlong and Guo Weimin in their report, the second century B.C. manuscript is clearly an example of the same genre of cookbook represented by the Shijing and Shici recipes quoted in the Qimin yaoshu; that is, we now have concrete evidence of continuity in the textual transmission of cookery recipes from antiquity into the medieval period.

The significance of these cookbooks should be considered in the light of other genres of technical literature that circulated in ancient and medieval times, were mostly lost after the Tang, and beginning in the twentieth century have become much better known through the recovery of medieval manuscripts at Dunhuang 敦煌and the archaeological excavation of ancient manuscripts, mostly from Warring States, Qin, and Han tombs. To cite several examples, hemerological manuscripts have been found in many ancient tombs whose contents have parallels in Dunhuang manuscripts.[4]One of the hemerological manuscripts from tomb 11at Shuihudi 睡虎地, in Yunmeng 雲夢, Hubei (burial dated ca. 217 B.C.), includes a section on demonology that is related to the medieval demonography Baize tu 白澤圖(Diagrams of White Marsh), known to us from two Dunhuang manuscript fragments, P2682 R° and S6261.[5] Medieval writings on sexual cultivation and love magic now have ancient parallels in medical manuscripts from tomb 3 at Mawangdui 馬王堆, in Changsha 長沙, Hunan (burial dated 168 B.C.).[6]

We have known from the bibliographic treatises of the Hanshu漢書, Suishu, and Jiu/Xin Tangshu that such technical literature existed in great variety and in manifold types of compilations -- some anonymous, some ascribed to legendary figures, and some authored by known individuals. The literature attests to the value placed on written records concerning various aspects of everyday elite life, and the recovered examples of the literature allow us to trace the influence of writing on the transmission of knowledge and on the forms of cultural experience in ancient and medieval elite society. Not only are we better able to gauge the quality of imagination that informed the activities in which the elite engaged, and thus to appreciate their mental and physical world, but we must also consider how technical literature contributed to that imagination.

Perhaps the most noteworthy observation based on comparison of related ancient and medieval technical manuscripts is that the text -- key words and sentences -- often remains essentially the same while the particular compilation in which it occurs has changed. In part this phenomenon is related to the nature of the information, which typically is encapsulated in textual units that can be combined and recombined without detriment to the information. The “recipe” -- an accurate rendering of fang方, especially with regard to the recipe as a unit of written information -- is the ideal form of technical literature. Authorship is not an issue, and yet there is an assumption of textual authority that we must recognize if we are to appreciate the place of technical literature in ancient and medieval elite life. I do not slight the role of oral transmission in the maintenance of belief and practice and in the introduction of novelty, and yet the impulse to commit information of all kinds to writing was a constant element of ancient and medieval elite life.

While it remains unknown when the first cookbook was written in China (probably not before the Warring States), I suspect that when Wu Yang’s cookbook was copied in the second century B.C., the decision to compile a cookbook was not an exceptional event but rather was something that many members of the elite might do. Coming back to the Qimin yaoshu and the two cookbooks whose recipes it quotes at length, their attribution remains uncertain. One speculation associates the Shijing with the Shijing by Cui Hao崔浩listed in the two Tangshu bibliographic treatises (perhaps to be identified with the Shijing of a Mr. Cui in the Suishu bibliographic treatise).[7] Cui Hao served as chief minister to the Northern Wei state in the first half of the fifth century. Only the preface to his Shijing survives, in which Cui states that he has recorded the recipes for the dishes prepared by the women of the family; his purpose was to preserve for later generations the family culinary traditions.[8] Even allowing for correspondence between Cui family traditions and the Shijing quoted in the Qimin yaoshu, the Qimin yaoshu source is not likely to have been the collection of recipes of a northern elite family.

On the contrary, Miao Qiyu繆啟愉, the modern editor and commentator of the Qimin yaoshu, argues convincingly in his commentary that the Shijing and Shici recipes quoted in the Qimin yaoshu are strongly influenced by the cuisine of the South during the period following the end of the Han, and that both cookbooks are likely to have been compiled by a southerner.[9] At the same time, Miao Qiyu notes evidence that the Shijing quoted in the Qimin yaoshu was not necessarily a single copy of a single book, but designated either several copies of a cookbook by this title or perhaps several recensions of recipes all equally referred to as Shijing.[10] Certain linguistic features, however, serve to indicate a shared language for all the recipes in the Qimin yaoshu that were derived from the Shijing and Shici, and to distinguish these recipes from Jia Sixie’s other recipes (which appear to reflect his own language). Miao Qiyu’s analysis of the relevant linguistic features makes it possible to identify the Shijing and Shici recipes even when the Qimin yaoshu text may not provide an explicit reference to the source of each recipe.[11]

One term in particular serves as a clear marker of Shijing and Shici recipes and suggests an element of formality that distinguished a body of “canonical” recipes. Only the Shijing and Shici recipes provide instructions concerning the presentation of finished dishes to the diners, using the verb dian奠. As Miao Qiyu notes, dian originally denoted the ritual presentation of sacrificial food. I concur with his judgment that “in ancient times there was no distinction between banqueting and sacrificing; while the term dian in the Shijing and Shici originated in the offering of sacrifices, to judge from Lu Chen’s account [a lost work on sacrifice quoted in the Taiping yulan太平御覽] it seems likely that such foods were first offered as sacrifices and afterwards served for dining.”[12] A simple recipe for carp stew (liyu hu鯉魚臛) from the Shijing provides a good example of the usage of dian:

Use a large fish. Scale, prepare, and cut into one inch by one inch squares, five fen thick. Boil and season like murrel stew. Add cooked cereal made from whole grains. When serving (dian), remove the cereal and serve (dian) [the dishes filled] halfway. If it is served (dian) over grain, it does not accord with the standard model.[13]

Jia Sixie offers a number of other recipes, some no doubt culled from his own family traditions, but the preponderance of Shijing and Shici recipes in the three Qimin yaoshu chapters is an indication of his concern for transmitting the standards of fine cuisine as set down in the tradition of canonical cookbooks. The recipes themselves must have circulated in various forms, both oral and written, and yet the conception of culinary culture was surely influenced by the existence of the written recipe literature. Who wrote the recipes, who read them, and how popular were the dishes whose exact preparation the recipes described? Since literacy is not a requirement for cooking, it seems likely that cookbooks were destined mainly for an elite readership for whom the recipes served several functions. The account of preparing and serving carp stew, which was probably a common dish, was, for those who could read, an effective way to capture the essence of carp stew without actually being the cook. The recipe gave the reader a sense of knowing the dish more fully than just through the experience of eating it, whether it was a familiar dish in everyday life or perhaps a dish that was unfamiliar and yet knowable in the written substitute.

Jia Sixie probably never ate many of the dishes quoted from the Shijing and Shici. Yet he would have been keenly aware of the culinary and cultural heritage embedded in the recipes. Recipes for dishes he had eaten were recorded alongside recipes for dishes of an earlier era or for dishes belonging to a different cultural region -- for example, the South in the period of the Northern and Southern Dynasties, whose dishes were not part of everyday experience for readers in the North.

Let me offer one example of the cookbook as a catalyst for the cultural imagination. Among the dishes described in the Qimin yaoshu recipes are wild rice (gu菰)[14] and water mallow soup (chun geng蓴羹)[15] with fish mincemeat (yu kuai魚膾). Both dishes evoked the region and culture of Wu in the Northern and Southern Dynasties. The several accounts of Zhang Han張翰, who in 301 left his home in Wu to serve in the administration of the ruler of Qi in Luoyang洛陽, are representative. As recorded in his Jinshu晉書biography, when the autumn wind blew in Luoyang, Zhang Han,

thought of Wu’s wild rice and water mallow soup with sculpin mincemeat and said, “What a man values in life is to obtain what suits his wishes. How can he be tied to an official post several thousand li away [from home] to pursue fame and rank?” -- whereupon he ordered his carriage and returned home.[16]

Here is the Shijing recipe for water mallow soup and fish:

The Shijing states: “Water mallow soup. The fish is in two-inch lengths but the water mallow is not cut. For murrel, put the water mallow into cold water. For culter, put the water mallow into cold water. When it boils, add the fish and salted black beans.” It also says: “The fish is three inches long and two and one half inches wide.” It also says: “Pick over the water mallow carefully and rinse it with hot water. Split the murrel down the center. Slice it thinly across the grain, making uniform slices two inches wide and the breadth of the halved fish. Cook, letting it come to the boil thrice, and drop in the water mallow all at once along with clear black bean liquid and dissolved salt.[17]

And here is the Shijing recipe for preparing wild rice:

Method for cooked wild rice. Put whole wild rice in a leather sack. Crush a porcelain dish with a pestle, being sure not to pound the bits too fine. Add the porcelain to the leather sack until full. Roll the sack back and forth on a board, and extract the grain. You can process one and a half sheng of grain at a time. Cook like rice.[18]

A far more detailed account of how to prepare water mallow soup with fish mincemeat precedes the Shijing recipe in the Qimin yaoshu, and is presumably by Jia Sixie; he also includes an account of water mallow cultivation in his section on raising fish (fish ponds are an ideal place for water mallow cultivation).[19] Whether water mallow was commonly grown in the North is uncertain; in his water mallow soup recipe, Jia Sixie includes advice about greens that may be substituted when there is no water mallow. Wild rice was surely a rare item in the North. Of course, the fact of a written recipe from the Shijing for water mallow soup with fish and for wild rice placed both dishes within a classical tradition of cuisine; that is, the dishes were neither exotic nor barbaric (and they were known to have been eaten by the Chinese elite prior to the Northern and Southern Dynasties).[20] Yet the perception of difference remained in effect in the Tang Dynasty, as for example when Yuan Zhen元稹wrote verses warning Bo Juyi白居易of the discomfiture that Bo Juyi must experience in the Southeast, including having to eat “mixed water mallow -- with quantities of split eel” and “a melange of millet -- half steamed wild rice.”[21]

In this sophisticated play on foodways and culture in ancient and medieval elite society, I would like to propose that cookbooks had a vital role. When the contemporary elite named particular dishes in prose and poetry, they may have known the dishes from personal experience, but cookbooks reinforced personal experience and extended their knowledge to unfamiliar dishes. At the very least, it seems likely that both Yuan Zhen and Bo Juyi read recipes for the dishes that Bo Juyi must now daily endure. The wealth of detail in cookbooks had multiple uses -- and like other genres of technical literature, cookbooks were the repository of a wisdom that informed daily life. No doubt the elite found cookbooks useful in developing culinary taste and ensuring the talent of kitchen staff. At the same time, the fact that the elite were reading cookbooks informed their aesthetic responses, and perhaps even abetted the extensive exploitation of the language of cuisine in literature.

As with other examples of ancient and medieval Chinese technical literature that are being restored in greater quantity in modern times through the discovery of manuscripts, I would argue that cookbooks were popular among the elite. Certainly the collection of recipes in the Qimin yaoshu suggests popularity. The few number of works listed in bibliographic treatises for the ancient and medieval periods that include the term shijing in their title -- not all of which are cookbooks of the type treated in this paper --[22] is not a useful indicator of popularity, but rather reflects the priorities of book collection and bibliographic classification. An anecdote in the tenth century Qingyi lu清異錄provides a vignette of an early ninth century gastronomer, Duan Wenchang段文昌, that suggests a possible pattern for the Tang elite:

Duan Wenchang was a great connoisseur of cuisine. A calligraphic plaque hanging in the kitchen of his mansion read “Hall for the Refinement of Delicacies.” And the street outside was known as the “Passage to the Bureau of Delicacies.” The household was run by an old family maid, who taught kitchen management to the female servants. They called her the “Matriarch of Cuisine.” Over a period of forty years she trained a hundred women, and only nine were talented enough to inherit her methods. Duan Wenchang compiled his own Shijing in fifty juan卷, which at the time was called “The Lord of Zouping’s Model Statutes of Cuisine” (Zouping gong shi xianzhang鄒平公食憲章).[23]