Printer Friendly

A confluence of humors: Ayurvedic conceptions of digestion and the history of Chinese "phlegm" (tan).

When one peruses medical case histories from the Qing dynasty, one sees constant references to phlegm, tan, as both cause and consequence of disease. That is to say phlegm figures as a central, indispensable concept in the late imperial Chinese imagination of the body and its pathologies. Curiously, however, when one goes back to the Huangdi neijing the earliest and foundational classic of Chinese medicine (ca. 1st c. B.C.E.), the term tan does not appear at all. Despite the central importance of phlegm in late imperial (and contemporary) Chinese medicine, only a few scholars have noticed this puzzle, and even fewer have engaged with it. Some have argued that the concept developed from the fluid yin. Others speculate that its development is owed to Ayurvedic conceptions received through Chinese translations of Indie Buddhist texts. Yet to date no one has put forward a sustained historical study of the origin and the development of the concept of phlegm in China.

This article is a first step in that direction. The first to consider both Chinese- and Sanskrit-language material, it investigates the earliest, formative period of the concept of phlegm in China. It argues that the initial emergence of the substance tan that later became "phlegm" was an indigenous development, while the formation of the concept of phlegm in Chinese medicine was influenced by Ayurveda. The influence hinges on the coincidence of Indie and Chinese intuitions about digestion.

The discussion proceeds along the following lines: (1) i first give a brief outline of indigenous Chinese theories of yin fluids and suggest how "phlegm" may have derived from them. (2) I then outline Indie physiological concepts relevant for discussions of phlegm in Buddhist texts. (3) Subsequently I consider the reception of Indie physiological concepts in Chinese Buddhism. (4) This leads to a short excursion into perceptions of phlegm (and bile) in contemporary Ayurvedic texts, and (5) a close reading of one sample passage in one of the Ayurvedic classics. (6) I argue that the divergent terminology for phlegm and bile in Chinese translations reflects a temporal shift in the perception of humors in India itself. (7) I then conclude by considering the influence of Ayurvedic conceptions of phlegm on Chinese medical traditions and suggest how the transfer of these conceptions from the corpus of Chinese Buddhist translations to the Chinese medical tradition could have taken place.


It is often stated that the character tan--later "phlegm"--first occurred in Zhang Zhongjing's (142-220? c.E.) Jingui yaolUe (ca. 200 C.E.). Evidence from treatises with older extant editions than the Jingui yaoliie suggests, however, that tan was originally written dan in the oldest versions of the treatise. (1) Dan moreover, first occurs in conjunction with yin, in the compound danyin which is part of a series of four yin diseases: "overflowing yin," yiyin; "hanging yin," xuanyin "blocked yin," zhiyin and danyin. (2) In other words, at first dan was a type of yin.

The primary meaning of yin is "drink." It is used both as a verb and a noun, meaning "drinking" and also "that which is drunk," "a drink." Instances of this meaning of yin go back as far as the Rites of Zhou (Zhou li). (3) A more specialized meaning of yin, however, evolved in medical usage; it began to refer to ingested fluids, specifically to harmful ingested fluids: when something went wrong with what one had drunk. The first occurrence of this kind of pathological yin fluid goes back to the Huangdi neijing. Excluding the doubtful Suwen Yunqi qipian chapters, (4) the treatise has three instances of "overflowing yin." "Overflowing yin" there refers to ingested surplus fluid, which the body is unable to process, so that it "overflows" (see the quote at the end of section 1 below). While "hanging" and "blocked" are also quite straightforward description of respective states of yin fluid, the meaning of dan is not so clear. Historical discussions of Chinese and Japanese scholars--principally Taki Motoyasu (1755-1810), Taki Motokata (1795-1857), Mo Meishi (1837-1907) and Zhang Shanlei (1873-1934)--revolve around the question how dan should be understood and how tan developed from dan.

As Mo and Zhang observe, dan means "clear." Hence, one would assume that danyin means "clear yin fluid" in the same manner in which yi, xuan, and zhi denote different states of yin fluids. However, in the Jingui yaoliie, danyin is also described as "sloshing around in the bowels." Thus Mo and Zhang ask: how could the author, Zhang Zhongjing, possibly have known or stated that yin fluid inside the bowels is "clear"? They suggest that dan is a scribal error for liu, "flowing." The shape of the character liu is close enough to dan to allow for this mistake, and it would make more sense to call a fluid sloshing around in the bowels "flowing" rather than "clear." Indeed, Chao Yuanfang (ca. 550-630) and Sun Simiao (581-682) list an independent category of "flowing yin" (liuyin) disease, whose definition is almost identical to the Jingui yaoliie's definition of danyin. This corroborates Mo and Zhang's argument. (5)

Like Mo and Zhang, Taki Motokata had also argued that yi, xuan, and zhi are descriptive of a state of yin fluid. But for him, dan is neither a miscopied liu nor does it mean "clear." Rather, it means "sloshing" (yaodong), an attested if obscure meaning of dan. In his definition, dan functions as a homonym and homophone of dan "sloshing fluid," and was used as a vulgar variant for the more complicated character. (6) In support of this claim, Taki Motokata cites evidence from a variety of sources: the Shuowen jiezi (121 c.e.), which defines dan as "water that is being moved around" (7) a passage from the Lingshu, "water that agitates beneath the heart in conjunction with fear that one will be apprehended by someone" (8) as well as a number of commentaries on poems that attest to the meaning of "agitated water" and to the interchangeability of dan with dan There is also evidence that tan and dan had the same meaning in the Xiaopinfang (ca. 445-73) and in Tao Hongjing's (456-536) Xiaoyan fang two lost works partially transmitted in the Japanese classic Ishimpo

   [TEXT NOT REPRODUCIBLE IN ASCII]: Dan in the chest, violent dry
   retching, and a bothered feeling. (9)

   [TEXT NOT REPRODUCIBLE IN ASCII]: Dan water in the chest. (10)

These two phrases, which use tan in place of dan, are extremely common in later phlegm-related literature. (11) Mo, Zhang, and the two Takis agree in arguing that tan developed from dan through a change from the water radical to the illness radical.

Following Taki Motoyasu, Taki Motokata observes that tan is not listed in either the Shuowen or Yupian (543) dictionaries, and medieval Buddhist dictionaries define dan and tan in almost identical terms as "ye body fluid in the chest" (12) and "water sickness above the chest". (13) The Song writer Huang Bosi (1079-1118), moreover, complains that his contemporaries frequently miscopy: "nowadays people write tan for dan; this is incorrect" [TEXT NOT REPRODUCIBLE IN ASCII]. (14) These instances show that dan and tan functioned as homonyms. (15)

Taki Motoyasu was also the first scholar to argue for Indie influences in the formation of the Chinese concept of phlegm. He noted that the initial location of dan in the Jingui yaoliie was the abdomen, whereas later texts primarily locate dan and tan in the chest. Taki therefore concluded that dan (and tan) of the later texts are not quite the same as the early instances of dan in the Jingui yaoliie, which essentially referred to an abdominal fluid. In his opinion, the later description of the location of tan in the chest is the sign of Indie influence. (16)

What is most interesting and pertinent for discussing the early understanding of dan tan in Chinese medicine is that Mo, Zhang, and the Takis all agree in understanding danyin as a subtype of yin, rather than ascribing any independent meaning to dan /tan as a substance in itself. Indeed, it is possible to construe a meaningful relationship between the definition of yiyin in the Huangdi neijing and the definition of danyin in the Jingui yaolue:

   Soft and dispersed liver pulse with lustrous complexion indicates
   overflowing yin. Overflowing yin is when, being thirsty, one drinks
   copiously and suddenly, so that [the ingested liquid] {overflows
   and) (17) enters the muscles and the skin outside the stomach and
   the intestines. [TEXT NOT REPRODUCIBLE IN ASCII] (Suwen 17, "Maiyao
   jingwei lun")

   If a person who was once thriving is now thin, and water goes
   around in his intestines, making a sloshing sound, it is danyin.
   [TEXT NOT REPRODUCIBLE IN ASCII] (Jingui yaolue, "Tanyin kesou bing
   maizheng bing zhi")

Yiyin and danyin are both harmful body fluids; they are both found in and around the stomach and intestines and seem both to be related to an intake of external fluids (this is spelled out clearly in the case of yiyin, and implied by the abdominal location in the case of danyin). The emergence of dan /tan can therefore cogently be described as a system-internal evolution from yin: yin and dan first denoted a gastro-intestinal pathology, and--as perceptively noted by the Takis--there is no indication that the earliest dan /tan had any relationship with the new substance of "phlegm."


There is a possible alternative way to explain the provenance of dan /tan in Chinese medicine: the simultaneity of the emergence of danyin in medical texts with the emergence of Buddhist translations in China could indicate that danyin shows an Indic influence in Chinese medicine. This explanation is plausible, because phlegm features prominently in Buddhist physiology and large-scale scriptural transmissions of Indic thought only commenced with the arrival of Buddhism to China. (18) Yet, unlike in Tibet, where this transmission also entailed direct translations of Indian medical works, China received her knowledge of Ayurvedic medicine almost exclusively through the intermediary of Buddhist texts. (19) This means that Chinese doctors and physicians knew of Indie medical concepts as they were used and portrayed in Buddhist scriptures, without being able to consult the Ayurvedic classics for additional detail or clarification. In order to gauge possible Indie influences in the concept of phlegm in Chinese medicine, we therefore need to understand how phlegm was conceived in Indie Buddhist contexts, and how it was transmitted to China from there.

In Buddhist literature phlegm principally appears in two contexts: as part of a three- or fourfold group of pathogenic agents (bile, phlegm, wind, and a combination thereof) closely related to the Ayurvedic theory of the tridosa and as member of a list of thirty-two body parts, enumerated for the purpose of meditation. I will give a brief overview of these two contexts first.

Wind, Bile, Phlegm and the Tridosa

Generally speaking, medical concepts in Buddhist contexts are very similar to medical concepts discussed in classical Indian medicine: classical Ayurveda itself developed in the context of early Buddhist asceticism from the first century B.C.E. to the second century C.E. (20) Of course Ayurveda is a full-fledged medical science, while Buddhist text only have occasional discussions of the body, sickness, and its cure. But apart from minor variations, both traditions share the same physiology, which stems from a common matrix of Indie traditions. Yet while the earliest redactions of the Ayurvedic classics only date from the fifth century C.E., the Pali canon began to be committed to writing in the first century B.C.E. Therefore one difference between medical concepts in early Buddhist texts and medical concepts in classical Ayurveda is that, due to their temporal priority, Buddhist medical passages lack the high level of explicitly theoretical elaboration with which key concepts are presented in the physiological and--as I shall argue in sections 4 and 5 below--younger sections of the Ayurvedic classics. But the key difference is that Buddhist texts bear witness to very early developmental stages of Ayurvedic concepts. They record traces of medical conceptions that predate the time of the recension of classical Ayurvedic texts and have been overwritten or erased in Ayurvedic texts as we know them today. What is usually known under the name of tridosa is a case in point.

In classical Ayurveda the tridosa doctrine functions as the main conceptual tool for explaining the etiology of illness and the preservation of health. Etymologically, dosa means 'fault' (< [??]dus 'become corrupted'), but classical and post-classical Ayurvedic treatises state that there are three dosas--wind, bile, and phlegm--which function equally as body constituents (dhatu) and as pathogenic agents (dosa). In their normal (that is, un-corrupted) state, these three dosas are understood as necessary, constitutive parts of the body, but when corrupted, vitiated, accumulated, riled, or expanded, they are held responsible for a panoply of diseases. Consequently, the vast majority of therapeutic measures in Ayurvedic medicine revolve around expelling, appeasing, curbing, yet curiously also strengthening the individual dosas in order to bring them back to balance and effect a cure. (21)

Scharfe has shown that, while "there are several references in the Buddhist canon to wind, bile, and phlegm as the cause of illness; the term dosa is also found in the canonical texts, but wind, bile, and phlegm are, as such, never so identified, either individually or jointly, and no reference to a tridosa theory occurs." (22) In other words, wind, bile, and phlegm started to be associated with the term dosa and referred to as a "triad of faults" (dosatraya) only relatively late, in the Bower manuscript (a Buddhist text dating from ca. 500-550 c.E.). Yet in this manuscript, and even in the older parts of the Caraka-samhita (one of the two foundational Ayurvedic classics redacted in the fifth c. c.E.), the term dosa has strictly negative connotations. (23) Only in the Sus'ruta-samhita (the other classic redacted in the fifth c. C.E.) and in the younger Astangasamgraha and Astangahrdaya (seventh c.) do we begin to see a positive re-valuation of the tridosa that is superimposed over their earlier understanding as pathogens. (24) Scharfe therefore convincingly argues that the positive valuation of the tridosa in the Sus'ruta-samhita was a younger development (the result of an attempt to achieve a neat correlation of the three dosas with the three gunas of Sarnkhya philosophy). In other words, the passages that describe the dosas as constitutive elements of the body, as well as passages that deal with strengthening the dosas in order to restore them to balance, are a secondary addition to the classical compendia. In the attempt to align Ayurvedic concepts with contemporary philosophical theory, the positive understanding of the dosas was superimposed on older textual layers, which focused exclusively on therapeutic measures for expelling and eliminating the dosas from the body. (25)

In the philosophical sections of Ayurvedic treatises and in many Buddhist treatises, the three dosas are correlated with four out of the five "great elements" (mahabhuta), which form the material building blocks of body and world in classical Indie theories of matter: earth (prthivi), water (apas), fire (tejas), wind (vayu), and space (okasa). (26) Subversion (Skt. parivarta) of the water element gives rise to phlegm, subversion of the fire element gives rise to bile, subversion of the wind element gives rise to wind. Sometimes a "combination" (samnipatika, samnipatavyadhi) of any of the previous three bhUtas is ascribed to subversion of the earth element. (27) Only the space element does not cause the rise of any particular dosa on its own.

Following Scharfe's above-outlined argument, it is anachronistic to discuss these pathogenic agents as "tridosa." They were not always known as dosa (the change in terminology postdates Pali and early Northern Buddhist texts), nor were they always a triad (as I shall also argue below). (28) However, this group of pathogens is well known and has been widely discussed under the label of "tridosa." For ease of comprehension, I have therefore opted to retain the term in lieu of a long-winded neologism.

Buddhist Lists of the Thirty-two Body Constituents

Altogether Buddhist physiologies enumerate forty-two body constituents (dhatu), which are correlated with the above-described four great elements of earth, water, fire, and wind. Here we are only concerned with the earth and water constituents, because it is these thirty-two constituents that are contemplated in Buddhist meditation exercises on the transitory and repulsive nature of the body. While the enumeration of thirty-two body constituents for the purpose of meditation is particular to Buddhism, lists of body constituents themselves are much older: they go back as far as the Vedas, where they were compiled in the context of sacrifices. (29) Sometimes the earth, fire, water, and wind elements themselves were added to the Buddhist list. This is why we get a discrepancy of thirty-two vs. thirty-six body constituents typically enumerated in these body constituents lists. The Mahasatipatthana-sutta (Digha Nikaya 22) contains just such a list of thirty-six body constituents:

   kesa, loma, nakha, danta, taca, mamsa, naharu, atthi, atthiminja,
   vakka, hadaya, yakana, kilomaka, pihaka, papphasa, anta, antaguna,
   udariya, karisa, pitta, semha, pubba, lohita, seda, meda, assu,
   vasa, kheja, singanika, la?ika, mutta, pathavidhatu, apodhatu,
   tejodhatu, vayudhatu (DN 22.4-5, 2293-94).

   Hair of the head, hair of the body, nails, teeth, skin, flesh,
   sinews, bones, bone marrow, kidneys, heart, liver, pleura, spleen,
   lungs, bowels, intestines, stomach, excrement, bile, phlegm, pus,
   blood, sweat, fat, tears, grease, saliva, mucus, serous fluid,
   urine. Earth dhatu, water dhatu, fire dhatu, wind dhatu. (30)

Both phlegm and bile are classified as water elements in this scheme.

Indie versions of these lists enumerate the body constituents in relatively stable order and employ a consistent and standardized anatomical nomenclature. (31) By contrast and perhaps not surprisingly, translations of the Indian body constituents lists in the Chinese Buddhist canon feature irregularities in numbers, order, and terminology.


There are three opposing scholarly positions regarding the reception of the tridosa in China; those of Unschuld, of Salguero, and of the Japanese research team of Jiro Endo, Teruko Nakamura, Hidehiko Yamaki and Hirokazu Miyamoto (hence forth abbreviated as "Endo et al."). According to Unschuld, the four great elements and the tridosa remained poorly understood in Chinese Buddhist translations themselves and were never integrated into Chinese medical theory because they were based on Indie conceptions of matter, foreign to the Chinese. (32) Salguero takes issue with Unschuld's argument. He argues that the variations in the Chinese terminology should not be seen as failed attempts at literal translation of Indie terminology, but rather as deliberate, strategic choices made by translators who were choosing terms that would resonate best with their intended Chinese audiences. For Salguero, the agenda of the translators was to enlist Buddhist medicine as an aid in the project of disseminating and proselytizing Buddhism in China, not to produce the most accurate rendering of Indie physiology. (33)

Salguero's reminder that we need to take into account the social and cultural contexts of the translation project and that the Chinese Buddhist translators had their own agendas, which differed from our own philological aspirations, is valid and well taken. Yet, it is questionable whether we can presume that "positioning] Buddhism in the Chinese religio-medical market place" (34) and "sell[ing] Indian medicine to the public" (35) were indeed the prime motivations of the translators in transmitting buddhavacana. For one, such a presumption amounts to imputing contemporary views of the world as a market place onto actors of 2000 years ago; and, as I shall demonstrate below, there is substantial evidence that the translators were extremely concerned with the literal accuracy of their terms.

The argument of Endo et al., which is based on a detailed diachronic study of the terminology for phlegm in Chinese Buddhist translations, both in the context of the tridosa and of the body constituents lists, is similar to Unschuld's position. Endo et al. note that phlegm was rendered with increasing accuracy as a substantive pathogen in Chinese translations. The very concept of a substantive pathogen, they submit, was introduced to China from India, through the transmission of the concept of phlegm. Like Unschuld, they blame the discrepancy between Chinese and Indie conceptions of matter for an alleged lack of understanding of the tridosa in early China, noting that Chinese conceptions of matter were attributive, while Ayurvedic conceptions were substantive. (36) In order to discuss Endo et al.'s argument, we need to review their findings.

Looking at table 1, we can, together with Endo et al., make a couple of straightforward observations. The changes in the terminology for phlegm in Buddhist translation literature can roughly be divided into three time periods. It was first translated as "cold" (han M) from the Later Han to Western Jin dynasties. In other words, it was translated by referring to its attribute. From the Eastern Jin to the Northern Wei dynasty, it was translated as "cold" (leng and "water(y)" (shui)--"water(y)" being already more substantial in character than "cold." From the Liang and Chen dynasties onwards, it was translated as tanyin, dan and tan. Based on this trajectory, Endo et al. argue for a gradual progression from an attributive to a substantive translation for phlegm in China.

Table 2 shows that in the early translation period before the Sui and Tang dynasties, translations of the Indian Buddhist body constituents lists mostly did not render phlegm. Endo et al. claim that this is because phlegm and saliva (tuo) were not differentiated at this stage, and so they were both translated as tuo. Endo et al. further observe that in later periods the character tan often does double duty to represent phlegm and bile, as in "yellow and white tanyin" (yellow tan = bile; white tan = phlegm).

Combining this observation with evidence from two other sUtras, which use dan and tan to translate "bile" rather than "phlegm," (37) Endo et al. advance their hypothesis regarding the formation and early meaning of the character dan. Though unacknowledged, their argument draws heavily on the above-outlined prior work of the Japanese scholars Taki Motoyasu and Taki Motokata. Just as in their case, Endo et al.'s argument hinges on the interchangeability of danand dan (both meaning "sloshing fluid"). Endo et al. extend their argument by explaining the provenance of the character dan. They note that in the Chinese medical texts and in Buddhist translations dan initially referred to "gallbladder" as well as "bile." Since this proved too ambiguous for the representation of bile (a fluid, after all), the flesh radical of the character dan became replaced with the water radical. The result was the character with the meaning of "sloshing bile." (38) Since dan was used interchangeably with dan (as the Takis had already established), the character dan (and later tan) came to be used for the representation of bile. In short


Endo et al.'s argument is exactly the same as that of Taki Motoyasu and Taki Motokata, except that they make dan logically prior to dan by positing its origin in dan ("gallbladder") and by arguing that dan first referred to "sloshing bile" rather than to "phlegm." Only a secondary, latent opposition between dan and dan, they argue, gradually led to a divergence between dan and dan and so these characters started to represent "bile" and "phlegm" respectively. (39)

Up to this point Endo et al.'s argument is elegant, even though their evidence is perhaps a little thin. The conclusions they draw from their findings, however, are problematic. Seeking to explain why phlegm was first translated with a character that meant bile, Endo et al. claim that Ayurveda, like Greek medicine, classified and observed phlegm and bile as expectorates and that this must have led the Chinese translators to believe that they belonged to the same category of fluids. Therefore, they argue, the translators rendered them with the same term (dan) until they achieved a better understanding of the opposed nature of the two fluids in the Tang dynasty. (40) Even a quick glance over Ayurvedic classics shows, however, that Indian doctors did not primarily understand phlegm and bile as expectorates, but as body constituents and/or waste products (see the detailed discussion of Ayurvedic digestion in section 4 below). Thus the alleged Ayurvedic understanding of phlegm and bile as expectorates cannot serve as an explanation why Chinese translators should have translated them with one and the same character.

Endo et al.'s conclusions, moreover, do not account for the discrepancies between tables 1 and 2. Although the Indie terms for phlegm and bile are the same (kapha/s'le?man and pitta), the development of the terminology for phlegm and bile in the Chinese translations follows completely different trajectories in the context of the trido?a doctrine and that of the Buddhist body constituent lists. In the case of the trido?a, we see the first translation of phlegm as tan in the Chen dynasty, and bile remains translated attributively as "hot," re or "yellow," huang throughout. By contrast, in the body constituents lists, phlegm is translated as dan almost three centuries earlier, during the Eastern Jin. In the case of the body constituents lists, by contrast, there is no single instance where bile or phlegm is translated attributively. They are rendered as dan, dan, dan or tan throughout. Endo et al.'s argument, that the translation of phlegm as dan and tan marks the point in time when the Chinese translators grasped phlegm as a substantive pathogen, would imply that these translators grasped the concept of a pathogen about three hundred years earlier in the context of the body constituents lists than in the context of the tridosa. Needless to say, this is highly unlikely; all the more so because many of the respective body constituents lists and tridosa occurrences are found in the same texts (see tables 1 and 2).

The discrepancy in the Chinese translations of phlegm (and bile) in the context of the body constituents lists and the tridosa rather suggests that contemporary Chinese translators perceived these substances differently in these two different settings. But what was this difference? I argue that the key to this question lies in Ayurvedic conceptions of digestion.

Using Chinese translations of Indie Buddhist texts as a basis to reconstruct the history of central Ayurvedic concepts would be highly problematic. Many of the Chinese translations were not directly transmitted from the Indian subcontinent; their routes of transmission involved Central Asian and South Asian cultures and translators. Often translators were not translating directly from classical Sanskrit or Pali but from regional dialects, and many of them were working from oral rather than textual transmissions. (41) Therefore I do not base my argument on evidence from Chinese Buddhist translations, but rather on evidence from the central Ayurvedic classics, Sus'ruta-samhita and Caraka-samhita themselves. I examine these classical compendia with the cues obtained from close reading of the Chinese translations, and show that the evidence in the oldest layers of these Ayurvedic classics agrees with and confirms the evidence from the Buddhist texts. I then go on to construct my argument from there.


In Ayurvedic pathology the dosas occur as single units. They cause disease as vata, pitta, kapha, or as a combination thereof. In Ayurvedic physiology, by contrast, the dosas have multiple functions, which are arranged typologically (see table 3 and 4 below). These functions are described unevenly within and across the early classics, Caraka and Sus'ruta. It is only in the treatises of Vagbhata that they start to be enumerated and named in a five-fold, standardized order. (42) This unevenness of the typologies shows that in the first centuries ce, when the Ayurvedic classics were composed, they had not yet matured into a standardized doctrine. (43) The development of the typologies thus falls roughly in the same time period as the early Buddhist translations. These nascent typologies should therefore give us a good indication of the range of perceptions of phlegm and bile that prevailed during that time. In other words, the body constituents lists can help clarify the differences in contemporary perceptions of phlegm and bile in the context of the tridosa and the Buddhist body constituents lists in China. This requires, however, to determine the type of bile and phlegm to which the bile and phlegm of these lists corresponded.

To do this, we need to take a closer look at the Ayurvedic conception of digestion. In Ayurvedic descriptions of the digestive process, ingested food is successively transformed into body constituents (dhatu) and waste products (mala, kitta). As in the case of the fivefold dosic typologies, Caraka's and Sus'ruta's descriptions of these digestive products are still uneven. (44) But, typically, digestive transformation involves seven body constituents, namely chyle, blood, flesh, fat, bones, marrow, and essence. Beginning with the transformation of food into chyle, each body constituent transforms successively into the next, while also producing a waste product in the process. In this scheme, phlegm is a waste product of the transformation of food into chyle, bile is a waste product of the transformation of chyle into blood, and so on (see tables 5 and 6).

While these lists of digestive products do not predate the Ayurvedic compendia, the Buddhist body constituents lists are much older. Yet both lists are a part of a shared Indie repertoire of body parts that goes back as far as the Vedas. Shared across different traditions and genres of the subcontinent, lists of body parts were assembled for different purposes and contexts and varied in the number and kind of parts they included. (45) Yet they were still put together from the same inventory. In other words, the products of the digestive process hail from the same storehouse as the constituents of the Buddhist lists. This becomes obvious when we draw up a list of the primary and secondary products of digestion and the body constituents lists and compare them with each other:

   Primary and secondary digestive products: chyle, blood, flesh, fat,
   bones, marrow, essence [and] phlegm (slesman), bile (pitta),
   secretions of the nine orifices (i.e., tears, earwax, snot, saliva,
   excrement, urine), sebum, sweat, nails, hair, oiliness of the skin,
   dirt of the eyes (Susruta Samhita su 46.527 and Caraka, ci

   Body constituents list: hair of the head and body, nails, teeth,
   skin, flesh, sinews, bones, bone marrow, kidneys, heart, liver,
   pleura, spleen, lungs, bowels, intestines, stomach, excrement,
   bile, phlegm, pus, blood, sweat, fat, tears, grease, saliva, mucus,
   serous fluid, urine (DN 22.4-5, 2293-94)

Clearly, apart from the solid organs and a few other items, the body parts of the two lists are identical.

Since digestive products and body constituents came from the same inventory, phlegm and bile in these two contexts must also be of the same kind. We already know that digestive phlegm and bile are produced during digestion; therefore in terms of the five-fold typologies they must be "moistening phlegm" (kledaka kapha) and "cooking bile" (pacaka pitta). Hence I deduce that bile and phlegm of the body constituents lists are also "moistening phlegm" and "cooking bile." In other words, I submit that the phlegm and bile of the body constituents lists are products of digestion. My argument is supported by a definition of the body constituents lists' bile and phlegm from within the tradition, Buddhaghosa's Visuddhimagga (5th c.):

The free bile spreads like a drop of oil on water all over the body ... The local bile is situated near the flesh of the liver between the heart and the lights. It is to be found in the bile container (pittakosa), which is like a large kosataki gourdpipe ... [Phlegm] is to be found on the stomach's surface (udarapatale). Just as duckweed and green scum on the surface of water divide when a stick or potsherd is dropped into the water and then spread together again, so too, at the time of eating or drinking, when the food, drink etc. fall into the stomach, the phlegm divides and then spreads together again. And if it gets weak, the stomach gets utterly disgusting with a smell of ordure, like a ripe boil or a rotten hen's egg, and then the eructations and the mouth reek with a stench like rotten ordure rising from the stomach so that the man has to be told, "Go away your breath smells." But when it grows plentiful, it holds the stench of ordure beneath the surface of the stomach, acting like the wooden lid of a privy. (Visuddhimagga VIII, 128) (46)

Buddhaghosa does not connect phlegm and bile to the transformation of body constituents--the materials he discusses predate emerging conceptions of digestion--but he does locate them firmly in the digestive tract.

I argue that the preceding evidence shows that, in contrast to the bile and phlegm of the body constituents lists, the bile and phlegm of the tridosa were perceived as constituent parts of an abstract triadic system that could manifest itself in the shape of a multitude of different functions, rather than as simple digestive fluids. (47) But this is only half of the answer.

The body constituents lists are not only considerably older than the five-fold typologies of dosas in the Ayurvedic classics; they are older than the concept of tridosa itself. This means that the understanding of bile and phlegm reflects an older layer of perception of these substances and that the understanding of phlegm and bile in the body constituents lists can tell us about the perception of these substances prior to and during the earliest stages of the tridosa theory. Hence, based on the understanding of phlegm and bile in the body constituents lists, I conclude that the earliest notion of phlegm and bile in the medical classics was that of digestive waste products.

We know well, of course, that bile and phlegm are associated with digestion. After all, this is clearly stated in the Ayurvedic classics. Dasgupta, in particular, has discussed how Caraka described the tridosa as "internal waste-products of unassimilated food-juice at the different stages of its assimilation" (he claims that Sus'ruta had a more "elemental" understanding of these substances). However, the perspective he adopted was essentially synchronic. Consequently, he was forced to make a rather odd argument, namely that waste-products "have important physiological functions ... for the preservation of the process of the growth of the body, when they are in due proportions." (48) In other words, Dasgupta saw the dosas functioning as simultaneously destructive and supportive of the body, instead of ascribing these two different and arguably quite contradictory functions to different developmental stages of their formation. This peculiar understanding of the dosas' dual function was to dominate most of the secondary scholarship following Dasgupta.

Muller, by contrast, was one of the few scholars who took seriously the layered nature of the classical compendia and sought to discern diachronic trajectories in the development of Ayurvedic concepts. He came very close to claiming that the earliest understanding of the dosas was that of digestive waste products. He argued, however, that the early negative valuation of the dosas as "faults" could not be reconciled with their early perception as digestive products. Hence he claimed that an even earlier understanding of the tridosas as "faults" must have been superimposed on later theories of digestion. (49)

Unlike Muller, I do not see any problem with a negative valuation of digestive products, and I believe that the understanding of phlegm and bile as digestive products shows clearly in the discussion of dosic pathologies across all of early Ayurvedic and other contemporary texts, not only in the explicit discussions of digestion in Caraka, as argued by Dasgupta. In order to demonstrate my contention, I will tease apart the different layers of meaning ascribed to the dosas in excerpts of one sample passage, the beginning of Susruta, sutrasthana chapter 21.


As mentioned above, classical Ayurvedic compendia, as they have come down to us, have been thoroughly redacted and supplemented by later commentators and do not predate the fifth century GE. (50) In the absence of text-critical studies, an accurate differentiation and dating of the textual layers in these works remains an impossible task. (51) Yet the composite nature of these compendia still manifests itself clearly through their heterogeneity, which allows isolating different strata of meaning in the texts. (52)

Ekavims'itamo 'dhyayah/ (53)


Athato vrana-prasnam adhyayam vyakhyasyamah //1// [section] 1 Now we will discuss the chapter that deals with the question of sores.

Yatha uvaca bhagavan dhanvantarih //2// [section] 2 As venerable Dhanvantari said.

Vata-pitta-slesmana eva deha-sambhava-hetavah / tair eva avyapannair adho-madhyaUrdhva-sannivistaih sarlram idam dharyate 'garam iva sthunabhis tisrbhir alas ca tristhunam ahur eke / ta eva ca vyapannah pralaya-hetavah tad ebhir eva sonita-caturthaih sambhava-sthiti-pralayesv apy avirahitam sarlram bhavati //3//

[section] 3 Vata, pitta, and slesman are the cause for the existence of the body. When uncorrupted, they are located in the lower, middle, and upper parts [of the body], and they support this body just like a house is supported by three pillars. This is why some call [the body] the "three pillared one." If these very same ones are corrupted, they are the cause of death. From birth, through life, and until death, the body is never separated from them, or from the fourth one, blood.

bhavati ca atra / narte dehah kaphad asti na pittan na ca morutat / sonitad api va nityam deha etais tu dharyate //4//

[section] 4 Regarding this [there is the verse] "there is no body without kapha, pitta, marut (wind), and blood; they constantly uphold the body." (54)

[[section] 5]

dosa-sthana-anyata urdhvam vaksyamah/ tatra samasena vatah sronl-guda-samsrayah / tadupary adho nabheh pakva-amasayor madhyam pittasya / amasayah slesmanah /1611 [section] 6 In what follows we shall discuss the locations of the dosas in turn: Among them, in brief, vata is situated between the pelvis (sroni) and the rectum (guda); pitta is situated above pelvis and rectum but below the navel (nabhi), between the amas'aya and the pakvas'aya, and slesman is situated in the amasaya.

atah param pancadha vibhajyante / tatra vatasya vata-vyadhau vaksyamah / pittasya yakrtplihanau hrdayam drstis tvak purvoktam ca slesmanas tu urahsirahkanthasandhaya iti purvoktam ca / etani khalu dosanam sthanany avyapannanam //7// [section] 7 Next [Joias] are divided into their five-fold divisions. We will discuss those of vata under 1wind diseases. Those of pitta are the liver (yakrt) and the spleen, the heart (hrdayam), the pupils of the eyes (drstis), the skin (tvak), and the aforesaid [that is, the region between the amasaya and the pakvasaya]. Those of slesman are the chest (urah), head (sirah), throat (kantha), the joints (sandhaya), and the aforesaid [that is the amasaya]. These are indeed the seats of the uncorrupted dosas.

visarga-adana-viksepaih soma-surya-anila yatha / dharayanti jagad deham kapha-pitta-anilas tatha //8//

[section] 8 Through emission, reception, and dispersal, kapha, pitta, and wind (anila) support the body, just as moon, sun, and wind (anila) support the earth.

[[section] 9]

tac ca adrsta-hetukena visesena pakva-amasaya-madhya-stham pittam caturvidham annapanam pacati vivecayati ca dosa-rasa-mutra-purisani tatra stham eva ca atma-saktya sesanam pitta-sthananam s'arirasya ca agni-karmana 'nugrahatn karoti tasmin pitte pacako 'gnir iti samjnah ... //10//

[section] 10 Because of particular invisible causal factors, pitta, which is located in the region between the repositories of the raw and the cooked (amasaya and pakvasaya), cooks the four kinds of food and drink and breaks them up into dosa, rasa, urine, and excrement. Situated there, [pitta] with its own power upholds the other pitta seats by virtue of the action of the body fire. This pitta is known as "cooking fire." ...

[[section] 11]

Ata urdhvam slesma-sthanany anuvyakhyasyamah / tatra amasayah pitta-as'ayasya uparisthat tatpratyanikatvad urdhva-gatitvat tejasas candra iva adityasya sa caturvidhasya aharasya adharah sa ca tatra udakair gunair aharah praklinno bhinna-samghatah sukhajaras ca bha vati //12//

[section] 12 Now we shall explain further the seats of slesman as mentioned above. The repository of the raw (amasaya) is above the repository of pitta (pittasaya). Because of their opposition to each other and because of the upward movement [of the heat of the pittasaya], it is like in the case of the moon, where the heating energy of the sun (tejasas adityasya) [is going upward toward the moon], [The repository of the raw] is the receptacle of the four kinds of food. There, through [the receptacle's] watery quality, food becomes moistened, broken up into its components, and easily digestible in this place.

madhuryat picchilatvac-ca prakleditvat tatha eva ca / amasaye sambhavati slesma madhura-sitalah //13//

[section] 13 From the sweetness, sliminess, and moisture [that exists there], sweet and cold phlegm are produced in the amasaya.

[[section] [section] 14-21]

tad usnair usna-kale ca meghante ca visesatah / madhyahne ca ardha-ratre ca jiryaty anne ca kupyati//22/

[section] 22 Through its heat, [bile] swells particularly during the hot season, at the end of the rainy season, at noon, at midnight, and while food is being digested.

[[section] 23]

sa sitaih sita-kale ca vasante ca visesatah / purvahne ca pradose ca bhukta-matre prakupyati //24//

[section] 24 Through coolness, [phlegm] becomes agitated particularly in the cold season, in spring, in the forenoon, and evening, and immediately after having eaten.

[[section] [section] 25-40]

Muller has already argued that the understanding of the dosas as three pillars that support the body ([section] 3), the description of bile, phlegm, wind, and blood ([section] 4), as well as the five-fold typology ([section] 10 (55)) represent late strata of the chapter. He based his argument on the positive valuation of the tridosa in these passages and on the contradiction of these introductory passages with the chapter heading, "Question about Sores." (56) Since this heading only refers to the second part of the chapter ([section][section] 25-40), which is on pathology, but not to the above translated discussion of physiology, Muller suggests that this discussion on physiology represents an interpolation into a chapter that was originally on pathology. That is to say, for him the pathological discussion represents the older layer of the text. (57)

I agree with Muller's argument, but what I am interested in is the curious redundancy with which the chapter describes the location of the dosas in the body and the recurrent emphasis it places on the digestive system as the site of bile and phlegm. The seats of the dosas are described in [section][section] 3, 6, 7, 10. While [section] 3 states that phlegm, bile, and wind are located in the "upper, middle, and lower part of the body," [section] 6 locates wind between the pelvis (ironi) and the rectum (guda), bile between the amas'aya and the pakvasaya, and phlegm in the amas'aya. Literally, amas'aya and pakvasaya are the repositories {asaya) of the raw (ama) and the cooked (pakva); they are located somewhere in the area of the middle and lower abdomen, (58) and in terms of function they roughly correspond to "stomach" and "intestines" of the Graeco-Latin tradition. [section] 7 gives the expanded array of the seats of phlegm and bile in their five-fold division ("liver, spleen, heart, pupils of the eyes, skin" for bile; "chest, head, throat, joints" for phlegm) and reiterates, once again, the seats of bile and phlegm in and between the repositories of the raw and the cooked. Reading these two paragraphs in conjunction gives the impression that [section] 6 and its description of the location of bile and phlegm in the repositories of the raw and the cooked was prior to [section] 7 with its enumeration of the five-fold division of seats (otherwise [section] 6 and [section] 7 could have been collapsed into one). This priority agrees with the above-argued temporal priority of the body constituents lists and Ayurvedic conceptions of digestion--both of which locate bile and phlegm in the digestive tract--vis-a-vis the five-fold typologies of the tridosa, which locate phlegm and bile in multiple locations throughout the body.

The following paragraph describes the five-fold typologies of bile and phlegm ([section] 7, see table 3 above), but not without discussing--for the fourth time!--the location of bile and phlegm in the repositories of the raw and the cooked, this time in close relation with the role of these substances in the process of digestion ([section][section] 10, 12-13). [section] 10 reiterates the location of digestive bile (pacakagni) between the repository of the raw and the cooked and its function as "cooking the four kinds of food and separating them into dosa, rasa, urine, and excrement." [section] 12 reiterates the location of phlegm in the repository of the raw, stating that the repository of the raw is located above the repository of bile (pittasaya) and that digestive phlegm, "because of its watery quality, moistens the four kinds of food and breaks apart its components so that they become easily dissolved." [section] 13 states that the repository of the raw, "being sweet, slimy, and moist," produces phlegm (slesman) that is of equally "sweet and slimy" nature.

Note that the text discusses digestive phlegm and bile both in terms of their function and in terms of their production: bile is "cooking food" but also turning food into a dosa [biles], rasa, urine, and excrement" ([section] 10). Phlegm is both "moistening and dissolving" food in the repository of the cooked ([section] 12), as well as being produced from the "sweetness, sliminess, and moisture" that exists there ([section] 13). We find, as we might expect, that conceptions of transformation of body constituents coincide with conceptions of digesting food.

Close examination of the location of bile and phlegm in the excerpted couplets also draws attention to a shift in the location of phlegm and bile in Indie physiology. In a paragraph that belongs to a comparatively late layer of the text, phlegm, bile, and wind are said to be located in the upper, middle, and lower part of the body ([section] 3), and this allocation of the dosa in terms of a tripartite division of the body is adopted by the vast majority of classical and post-classical Ayurvedic texts. Older layers of the text ([section][section] 6, 7, 10, and 12), however, locate phlegm and bile in and between the repositories of the raw and the cooked. [section][section] 7, 10, and 12 do so in the context of discussing particular instances of the dosas (the five-fold typologies and digestion), but [section] 6 makes a general statement about the location of dosas in the aggregate: it locates the primary seats of tridosic bile and phlegm in and between the repositories of the raw and the cooked and sees the digestive process as the source of their production. (59) This connection between the digestive process and the production of the dosas in the older layers of the text constitutes the reason for the emergence of pathologies of bile "after having digested" ([section] 22) and for illnesses of phlegm "right after a meal" ([section] 23).

These are by no means isolated observations about the particular timing for complaints of phlegm and bile. The close relationship of phlegm and bile with the digestive system reappears across a number of pathological sections in the classical Ayurvedic compendia as well as in Buddhist texts. The medical chapter of the Suvarnabhasottamasutra (ca. 300 C.E.), for example, states in the context of a discussion of the four elements, that

   kapha-adhikah kupyati bhukta-matre pitta-adhikah kupyati jiryamane/
   vata-adhikah kupyati jirna-matre ity esa dhatu-tritaya-prakopah

   Abundance of phlegm erupts right after a meal; excess of bile
   erupts during digestion, excess of wind erupts as soon as one has
   digested. These are the vitiations of the triad of constituents
   (dhatu). (60)

Caraka considers purgation and emesis the best cures for bile and phlegm because these treatments clear out the repository of the raw and thereby extract the "entire root" of bile and phlegm:

   virecanam tu sarvopakramebhyah pitte pradhanatamam manyante
   bhisajah; tad dhy adita eva amas'ayam anupravisya kevalam
   vaikarikam pitta-mulam apakars'ati, tatra avajite pitte 'pi
   sarira-antar-gatah pitta-vikarah pras'antim apadyanse, yatha 'gnau
   vyapodhe kevalam agnigrharn siti bhavasi tadvat //16// (Caraka
   Samhita su 20.16) (61)

   Doctors regard purging as the most excellent of all medicines for
   pitta. Beginning when [the purgative] has entered into the
   repository of the raw (amasaya), it indeed extracts the entire root
   of the corrupted pitta. When [the corrupted] pitta has been
   conquered there, pitta illnesses that have gone inside the body
   become extinguished, just like a fire-chamber, which becomes
   completely cold when the fire is removed. [section] 16

   vamanam tu sarvopakramebhyah slesman i pradhanatamam manyante
   bhisajah; tad dhy adita eva umasayam anupravisya uro-gata (62)
   kevalam vaikarikam s'lesma-mulam urdhvam utksipati, tatra avajite
   slesmany api sarira-antargatah s'lesma-vikarah prasantim apadyante,
   yatha bhinne kedarasetau sali-yava-sastika-adiny
   an-abhisyandyamanany ambhasa prasosam apadyante tadvad iti //
   (Caraka Samhita su 20.19)

   Doctors regard emesis as the most excellent of all medicines for
   slesman. Beginning when [the emetic] has entered into the
   repository of the raw (amasaya), the root of the corrupted slesman,
   which is located in the chest, is completely thrown up. When
   slesman has been conquered there, slesman illnesses that have gone
   inside the body dry up, just like crops such as paddy, barley, and
   sixty-day rice dry up when the ridge of a paddy field is broken and
   they are not being irrigated by water.

Again, the particular timing of the eruption and the specific site of extraction of phlegm and bile that are described in this passage only make sense if they are understood as substances that arise at the time and place of digestion, when and where foods are transformed into chyle and blood. (63)

In view of the evidence in the above-quoted examples, I submit that the close relationship of phlegm and bile with the process of digestion in the pathological sections of Sus'ruta and Caraka--that is, the older layers of these compendia, and in early Buddhist texts from ca. 300 c.e.--illustrates an early layer of understanding of tridosic bile and phlegm. According to this early understanding, phlegm and bile were conceived as digestive waste products that were produced and located in the repositories of the raw and the cooked. I further argue that the understanding of tridosic phlegm and bile as digestive waste products was prior to the five-fold typologies of the tridosa (with their extended locations in other parts of the body [[section] 7 above] and that it also predated the tripartite arrangement of the tridosa in the upper, middle, and lower part of the body.

Following the development of this tripartite arrangement of the tridosa in the body it is possible to discern an upward movement of the main seat of phlegm from its original location in the amas'aya to the chest (urah). Due to its layered nature, Caraka is already vacillating between an early pathological conception of amas'aya as the prime location for phlegm (Ca Sam sit 20, quoted above) and a later physiological conception of the primary seat of phlegm in the chest.

   urah siro griva parvany amasayo mesas ca slesma-sthanani tatrapy
   uro visesana slesmasthanam (Caraka Samhito su 20.8) (64)

   Chest, head, neck, joints, amasaya, and fat are the seats of
   slesman, thereof particularly the chest.

By the time of the Astangasamgraha (7th-8th c. C.E. (65)), the primary seat of phlegm had clearly shifted from the amasaya to the chest (urah) and kledaka kapha ("digestive phlegm")--the phlegm that is produced during the digestive process in the amas'aya--had been superseded by avalambaka kapha ("supportive phlegm")--the kind of phlegm that is seated in the heart--as the main type of phlegm that nourishes and replenishes phlegm in its other seats in the body. (66) I suggest that this replacement represents an effort to align the main seat of phlegm with the new tripartite location of dosas in the body, according to which phlegm is located in the chest, and to make intelligible the new location of phlegm, which simply did not make sense in terms of the original understanding of phlegm as product of digestion.

In the Astangahrdaya (7th-8th c. C.E.), Vagbhata finally integrated the connection of the tridosa with digestion into an elaborate architecture of tripartite divisions of body, life, day and night, and a meal:

   vayuh pittam kaphas ceti trayo dosah samasatah //6// vikrtavikrta
   deham ghnanti te vartayanti ca / te vyapino 'pi hmtsabhyor
   adhomadhyordhva samsrayah //7// vayo 'horatribhuktanam te
   'ntamadhyadigah kramat/ tair bhaved visamas tsksno mandas cagrsih
   sarnaih samah //8// (Vagbhata, Astangahrdaya, su 6-8)

   Wind, phlegm, and bile are the three dosas. In their abnormal
   state, they kill the body; in their normal state, they cause it to
   exist. /6/ Even though they are spreading everywhere, the dosas are
   seated below, between, and above the navel and the heart. /7/ They
   appear consecutively at the end, the middle, and the beginning of
   life, day, night, and a meal./ Through them the (digestive) fire
   may become irregular, violent, and sluggish, or balanced, if they
   are regular. /8/ (67)

Vagbhata's presentation masks the incongruity of the causal connection of the dosas with digestion and their newly established seats: it now appears fortuitous that phlegm and bile appear at the time of a meal. However, despite the revised locations for wind, bile, and phlegm, Vagbhata's dosic pathologies remain centered on the digestive system and the extraction of the dosas from the digestive tract still remains the best of all possible cures:

   sarirajanam dosanam kramena param ausadham / vast ir vireko
   vansanam tatlsa tailam ghrtam madhu //25// (68)

   For the dosas produced in the body in the (above) order, the best
   remedies are purgation, sweating, and emesis, and likewise sesame
   oil, ghee, and honey /25/

This latent therapeutic emphasis on the digestive tract as the site of extraction of phlegm despite its physiological upward move to the chest is yet another sign of the anteriority of the association of the dosas with the digestive system. (69)


As I have shown, the different terminology for bile and phlegm in the context of the body constituents lists versus the tridosa in the Chinese translations rendered a difference in conception well established in the Indic texts. This is why, in the context of the body constituents lists, the characters dan and tan could do double duty for the representation of bile and phlegm, while in the context of the tridosa they could not. In the body constituents lists, bile and phlegm were first and foremost imagined as waste products and not as opposing poles of a triad. Therefore it made sense to represent them with the same character. The contemplation of the body constituents was less about imagining phlegm and bile as opposed poles of cold and heat than about perceiving them as impure digestive fluids. The classification of both bile and phlegm as water elements in the body constituents lists (discussed in section 2 above) must also have helped to emphasize their shared character as fluids over and above their opposition in terms of temperature. By contrast, in the context of the tridosa, phlegm and bile were not and could not be represented with the same Chinese character, because their meanings largely derived from their opposition into qualitatively different functions.

The Chinese translators' choice of employing a different terminology for the translation of bile and phlegm in those two different contexts, despite the prevailing use of the identical terms of "pitta" and "kapha" for both contexts in the Indic source texts, has preserved for us an old distinction in the meaning and perception of bile and phlegm in Indic physiology. This distinction is not only qualitative but also temporal, and it points to the gradual development of divergent perceptions of phlegm and bile over time. The early Buddhist translations predated the first redaction of the Ayurvedic compendia and coincided with the development of divergent conceptions of the dosas, such as the five-fold typologies, the tripartite division of seats in the body, etc. Therefore the different character of phlegm and bile in these two different contexts was still obvious to those who translated the texts into Chinese at that time.


The preceding section has shown that--despite their complicated provenance and their translation into another language and cultural context--the Chinese Buddhist translations rendered Indic conceptions of phlegm with meticulous care and with astonishing faithfulness to their original meaning in the Ayurvedic classics. The question is, however, whether these Indic conceptions of phlegm in the Chinese Buddhist translations actually ended up influencing the formation of ideas about phlegm in medical texts or not.

Endo et al. have sought to answer this question by comparing the trajectory of phlegm in Chinese medical texts and Buddhist translations. Their analysis of the development of "phlegm" (location, symptoms, treatments, and range of attributes) in Chinese medical texts during the first four hundred years of its emergence shows the gradual "solidification" of a fluid. Initially, danyin did not yet have any viscosity; it was essentially a water pathology ("water goes around in the abdomen, making a sloshing sound") (70) In the third century it became a dense fluid that could form watery lumps (dan/tan pi in the hypochondrium. (71) From the fourth century onwards dan referred to a dense fluid that caused watery lumps and fullness (pi, piman, tanman), was prone to accumulations (yu, ji, shi, jie, yong), and could even become "like glue". (72) Endo et al. argue that this trajectory of gradual solidification in the medical texts mirrors the trajectory from attributive to substantive translations of phlegm in Chinese Buddhist translations, which shows that Indic conceptions of phlegm did influence the development of phlegm in Chinese medical texts. As noted above, Endo et al.'s argument is problematic. It does not account for the divergence of the trajectories of phlegm in the contexts of the tridosa and the body part lists. The trajectory of phlegm in the context of the body constituents lists shows absolutely no development from an attributive to a substantive translation that would match the development of the terminology for phlegm in the medical texts.

In contrast to Endo et al., and in agreement with the above-mentioned Chinese scholarphysicians, I therefore argue that the appearance of danyin in the Jingui yaoliie was an indigenous development. For one, the appearance of the character dan in the Jingui yaoliie actually predates the appearance of phlegm in the Buddhist translations by more than a hundred years. According to evidence from table 1, the earliest instances of dan and tan in Chinese Buddhist translations are found in the Madhyamagama (Zhong'ahanjing) and Ekottaragama (Zengyi'ahanjing), both translated in 397 C.E. The terminus ad quem for the composition of the Jingui yaoliie, however, is 219 C.E. This precludes the possibility of Indic influence on the initial occurrence of dan in the Jingui yaolue or makes it at least extremely unlikely (we cannot entirely exclude the existence of earlier translations containing the character or that phlegm in the Jingui yaoliie was actually an interpolation while the original version did not yet contain the character dan at all).

But the most important piece of the puzzle is the compound tanyin (danyin), which occurs prominently during the Eastern Jin among the earliest Buddhist translations of phlegm in the context of the body constituents lists (table 2) and later, during the Tang dynasty, also in the context of dosa translations (see table 1). Tanyin is an extremely rare compound, which appears almost exclusively in Buddhist texts; (73) it is also homophonous with tanyin (danyin) of the medical texts. The homophony of these two compounds and the fact that they appear in exactly the same contexts suggest that tanyin is a derivative of the tanyin (danyin) compound of the medical texts. (74) The use of a compound that derives from tanyin (danyin) to translate "phlegm" in Buddhist texts suggests that the Buddhist translators borrowed the notion of danyin fluid from the Chinese medical context to translate the notion of digestive phlegm of the body constituents lists, rather than the other way around (see table 7).

The fact that tanyin was used almost three centuries earlier in the context of the body constituents lists than in the context of the tridosa also shows that the translators must have perceived a specific similarity between danyin and the digestive kind of phlegm and bile of the body constituents lists (rather than between the bile and phlegm of the tridosa typologies).

What was this similarity that let them choose danyin over other possible candidates among body fluids? In view of the evidence, I suggest the following hypothesis regarding the origin and development of the concept of phlegm, tan, in Chinese medicine. The original location of digestive bile and phlegm in stomach and abdomen provided the basis for the initial identification of Indic digestive phlegm (kledaka kapha) in the body constituents lists with the fluid of danyin (tanyin), as discussed in the Jingui yaoliie and other early Chinese medical texts. Since yin fluids already functioned somewhat like a pathogen in these early medical texts, they naturally suggested themselves as a suitable candidate for the translation of the negatively valued digestive phlegm and bile of the Indic body constituents lists. Digestive phlegm and bile of the body constituents lists (as well as of early Ayurvedic pathology) were equally located in stomach and abdomen, in exactly the same locations where the danyin (tanyin) fluid of Chinese medical texts could be found. Hence out of all the different subtypes of yin fluids, the compound danyin suggested itself naturally as a suitable term for the earliest translations of Indic phlegm. Interestingly, the compound danyin (tanyin which designated a type of yin fluid--and not the isolated character dan /tan, which subsequently took on the meaning of phlegm--was chosen for the earliest translations of Indic phlegm. In other words, in the initial translations, Indic phlegm was not identified as dan/tan but as a type of the Chinese fluid yin.

Once the identification of Indic phlegm with the danyin fluid was achieved, Indic conceptions of phlegm had an easy inroad into Chinese medicine. The initial translation of Indic phlegm in Buddhist translations as tanyin, a compound that derived from the medical term danyin created the conditions under which influence from Buddhist translations to Chinese medical theory could and did indeed take place. Evidence of this influence includes the move of the primary site of phlegm from the digestive system to the chest in later Chinese physiology, which directly parallels the similar shift in Ayurveda described above; an increasing use of the single character dan/tan, independently of its use in a compound with yin; and an increasing viscosity of phlegm in the Chinese medical tradition (left unaccounted for when explaining the Chinese concept of phlegm as an entirely indigenous development). The initial associations of danyin (tanyin) with digestive Indic phlegm (kledaka kapha) continued, however, and account for the persisting ties of tanyin with digestion in later Chinese medical understanding.

This outline of Ayurvedic connections in the formation of phlegm--one of the major concepts in post-classical Chinese medicine--shows that the scope of Indic influences on Chinese medicine is far greater than Unschuld assumed. It also shows that, contrary to Endo et al.'s claims, Chinese medicine did have a clear notion of "substantive" matter and a "substantive" terminology for body fluids prior to Indic influences, and that Chinese Buddhist translators did not hesitate to use this substantive terminology when they saw fit. My findings also show that translators did not create and manipulate variations in the terminology for phlegm and bile primarily with a view to their intended audiences, as contended by Salguero. Quite the contrary: The terminology crafted by the Chinese translators was so accurate and faithful to the Indic source texts that it preserves for us an ancient and long-forgotten distinction in the perception of phlegm and bile in early India and alerts us to a temporal shift in the meaning of Ayurvedic terminology that was all but lost to us today.

Natalie Kohle

Australian National University

This article is a synopsis of two chapters of my dissertation "Phlegm (tan): Toward a History of Humors in Early Chinese Medicine" (Harvard 2015). First and foremost, I thank my advisors, Shigehisa Kuriyama. Leonard van der Kuijp, and Janet Gyatso for their inspiration and guidance. I also thank Li Jianmin of Academia Sinica for working with me, and for his advice. I am grateful to my friends Marcus Bingenheimer and Elisabeth Kaske for reading through and commenting on my manuscript, and to Stephanie Jamison, Anne Monius. Mari Jyvasjarvi-Stuart, and Dan Stuart for their time spent reviewing and editing my Sanskrit translations. I thank Pierce Salguero for inviting me to present my research for this article at his panel at the 24th International Congress of History of Science. Technology, and Medicine in Manchester in 2013, and I thank Dominik Wujastyk. Volker Scheid, and all the other panelists for their comments and encouragement. Research for this article was made possible by the financial support from the Taiwanese Ministry of Foreign Affairs ("Taiwan Fellowship"), the Chiang Ching-Kuo foundation ("Doctoral Fellowship"), and the Harvard Graduate Society ("Dissertation Completion Fellowship").

(1.) The oldest quotes of Zhang Zhongjing's Shanghan zabing lun (the predecessor of the Jingui yaoliie) are found in Tamba Yasuyori's (911-995) lshimpo (984) and in Wang Shuhe's (fl. 210-58) Maijing. The surviving lshimpo ms. copy dates back to Heian times (794-1185), the earliest surviving edition of the Maijing dates from 1330, and the earliest surviving edition of the Jingui yaoliie itself dates from 1340.

(2.) Jingui yaolue, "Tanyin kesou bing maizheng bing zhi" (ed. Li Keguang (Taipei: Zhiyin, 2002, reproduced in Scripla Sinica).

(3.) See Chen Zhiping, "Dui Jingui de 'tanyin' bing pian de chubu tantao", Jiangsu Zhongyi 8 (1958): 14.

(4.) These chapters are probably a later accretion to the Huangdi neijing. See Paul Unschuld, Huang di nei jing su wen: Nature, Knowledge, Imagery in an Ancient Chinese Medical Text (Berkeley and Los Angeles: Univ. of California Press, 2003), 393; Catherine Despeux. "The System of the Five Circulatory Phases and the Six Seasonal Influences," in Innovation in Chinese Medicine, ed. Elisabeth Hsii (Cambridge: Cambridge Univ. Press, 2001), 128-29.

(5.) Chao Yuanfang, Zl?ubing yuanhou lun (610), juan 20: Sun Simiao, Qianjin yaofang, juan 18: Zhang Shanlei. "Mo Meishi Yanjingyan 'Shi tan pian'", in Zhang Shanlei yiji Yanyi kaozhengji (Beijing: Renmin weisheng chubanshe, 1996), 403-4; Mo Meishi, "Shi Tan" in Yanjingyan (Beijing: Renmin weisheng chubanshe, 1990), 63-64. The identity of danyin and liuyin was already noted by Yu Yan (1879-1954). See his "Tanshuo" and "Yinshuo", in Yu Yan, Yu Yunxiu Zhongyi yanjiu yu pipan (Anhui: Anhui daxue chubanshe, 2006), 83, 88.

(6.) See Taki Motokata, "Tanin gaisobyo myakusho narabini chi dai juni", in Kinki gyokkan yoryaku jutsugi, Mankyudo Hanabusa Heikichiro, Sonsei Yakushitsu sosho (1854), fascicle 2. folia 6a-b.

(7.) Shuowen jiezi, juan 11, Shuihu

(8.) Lingshu 4, "Xieqi zangfu bingxing" Siku quanshu [SKQS], Wen yuan ge Si ku quan shu dianzi ban (Hong Kong: Zhongwen daxue chubanshe. 2002). This sentence is also discussed in Endo et al., "tan no kigen II", Nihon ishigaku zasslti 39 (1993): 547. Compare the very similar lines in Lingshu 19, "Sishiqi" (SKQS).

(9.) Xiaopin fang, quoted from Tamba Yasuyori, Ishimpo (Shenyang: Liaoning kexue jishu chubanshe, 1996, reproduced in Scripta Sinica), juan 20, section 29.

(10.) Xiaoyan fang quoted from Ishimpo, juan 9, section 7.

(11.) At the time, the Xiaopin fang was lost, so Taki Motokata quotes from an excerpt in the Ishimpo. Recently, one juan from an old manuscript of the Xiaopin fang has surfaced, but not the juan that contains the recipe in question. Chen Yanzhi. (4/5th c.), Shohinho Kotei naikyo meido (Tokyo: Kitasato Kenkyujo Fuzoku Toyo Igaku Sogo Kenkyujo, 1992), 70-71. The Xiaoyan fang is also lost.

(12.) Ye is a normal, non-pathological physiological body fluid. The Huangdi neijing describes it as the first transformation product from water and grains ([TEXT NOT REPRODUCIBLE IN ASCII]), which subsequently transforms into more specific body fluids. Lingshu 36, "Wu long jinye bie"; Suwen 23, "Xuanming wuqi" (SKQS). Ye is usually discussed in conjunction with the body fluid jin, and it appears most commonly as the compound jinye Jin and ye are conceived as two different kinds of body fluids, however. Jin denotes a thin, mobile body fluid that travels quickly and lightly throughout the body, whereas ye denotes a viscous, less agile fluid: "Jin and ye each have their own paths, therefore when the three burners issue qi, that which warms muscles and flesh and fills the skin is jin, that which flows but does not travel is ye." Lingshu 36, "Wu long jinyue bie" (SKQS). For a preliminary discussion of jin and ye body fluids in a Western language, see Steven Clavey, Fluid Physiology and Pathology in Traditional Chinese Medicine (Melbourne: Churchill Livingstone, 1995), 1-38.

(13.) Taki Motoyasu. Isho, in Huanghan yixue congshu, vol. 13 (Shanghai: Shanghai Zhong-yixueyuan chubanshe. 1993), yuan shang, 31-32: Taki Motokata; Kinki gyokkan, fascicle 2. folia 6a-b. comes from a quote of Ruan Xiaoxu's (479-536) Wenzi jilUe quoted under the entry for feng huang dan re in Huilin's (737-820) Yiqiejing yinyi--(808), T.54n2128_p0452b20, quoted from the Chinese Buddhist Electronic Text Association (CBETA), accessed 8/29/2015. [TEXT NOT REPRODUCIBLE IN ASCII] is an entry under dan in Chen Pengnian's (961-1017) Da Song chongxiu guangyun (1008), juan 2, section 23, folio 75a. Quoted from The Chinese Text Project, (accessed on 8/29/2014).

(14.) Huang Bosi, Fade kanwu juan xia, folio 5b [TEXT NOT REPRODUCIBLE IN ASCII]. The line in question is "[TEXT NOT REPRODUCIBLE IN ASCII], The quote is from one of Wang Xizhi's. (303-361) calligraphies, Zuxiage ruchang tie. See Wang Xizhi, Xingshu caoshu huihian (fu shi wen)(Beijing: Beijing shuji chubanshe, 1990), 349, 351.

(15.) It is, however, difficult to determine when the change from dan to tan took place, because to date the vast majority of extant medical text editions do not predate Ming. Fabien Simonis notes that "the dominance of tan over dan in Chinese translations of Indian Buddhist texts suggests that tan was already more common than dan in Tang times but by no means proves it, because the version of the Buddhist canon through which we are reading these texts was edited by people who took the form tan for granted and may therefore have 'corrected' dan for tan as a matter of course" (Fabien Simonis, "Ghosts or Mucus? Medicine for Madness: New Doctrines, Therapies, and Rivalries," in Modern Chinese Religion I, Song-Liao-Jin-Yuan [960-1368 AD], vol 1, ed. John Lagerwey and Pierre Marsone [Leiden: Brill, 2014], 612-13 n. 30). By contrast, that an extant late-eleventh/early-twelfth century edition of the Waitai miyao (752) still has dan where later editions have tan suggests that the change did not take place until the Song period or after. Simonis, "Mad Acts, Mad Speech, Mad People in Late Imperial Chinese Law and Medicine" (PhD diss., Princeton Univ., 2010). 55-56 n. 76.

(16.) Taki Motoyasu, Isho. See also Taki Motokata, Zatsubyo koyo vol a, p. 226, ed. Zhang Liping (Beijing: Xueyuan chubanshe, 2009).

(17.) Song commentators of the Suwen note that Huangfu Mi's (215-282) Jiayijing has yi instead of yi; hence yi should be understood as yi. See Unschuld, Huang di neijing suwen, 290. My translation is based on that reading.

(18.) Earlier, pre-Buddhist Sino-Indian cultural exchanges are attested by the presence of Indian loan words in Old Chinese. Yet, due to lack of historical records, very little is known about these contacts. On early Sino-Indian cultural contact see Victor Mair, "Foreword: The Beginnings of Sino-Indian Cultural Contact," Journal of Asian History 38 (2004): 81-96: "[The] File [on the Cosmic] Track [and Individual) Dough[tiness]: Introduction and Notes for a Translation of the Ma-wang-tui Manuscripts of the Lao Tzu [Old Master]," Sino-Platonic Papers 20 (1990): 1-68; "India and China: Observations on Cultural Borrowing," Journal of the Asiatic Society (Calcutta) (30.3-4) (1989): 61-94. On loanwords, see Hoong Teik Toh, "Notes on the Earliest Sanskrit Word Known in Chinese," Sino-Platonic Papers 201 (2010); and Zhu Qingzhi, "Some Linguistic Evidence for Early Cultural Exchange between China and India," Sino-Platonic Papers 66 (1995). The use of South Asian substances, such as chenxiang (Aquilaria agallocha), in pre-Buddhist Daoist and medical practice also suggests early cultural contacts of China with South Asia, but research on this fascinating topic has yet to be undertaken.

(19.) There are few exceptions: the Kumaratantra of Ravana and the KoSyapa-Samhita were directly translated into Chinese. However, these two translations are rather late, describe a kind of demonological pediatrics, and do not belong to the corpus of classical Ayurveda. See Jean Filliozat, etude de demonologie indienne: Le Kumaratantra de Ravana et les textes paralleles indiens, chinois, cambodgien el arabe (Paris: Imprimerie Nationale, 1937) and "Le Kumaratantra de Ravana," Journal asiatique 226 (1935): 1-66; Prabodh Chandra Bagchi, "A Fragment of the Ka.<yapa-Samhita in Chinese," Indian Culture 9 (1942): 53-64, and "New Materials for the Study of the Kumaratantra of Ravana," Indian Culture 7 (1941): 269-86. On Tibetan translations of Ayurvedic texts, see Claus Vogel, Vagbhata Astangahrdaya-samhita: The First Five Chapters of Its Tibetan Version (Wiesbaden: Stciner Verlag, 1965); Palmyr Cordier, "Introduction a l'etude des traits medicaux sanserifs inclus dans le Tanjur tibetain," in Gustave Lietard et Palmyr Cordier: Travaux sur l'histoire de la medicine indienne. Un demi-siecle de recherches ayurvediques, ed. Arion Rosu (Paris: College de France, Institut de Civilisation Indienne, 1989), 513-39.

(20.) This was first argued by Kenneth Zysk in his Asceticism and Healing in Ancient India: Medicine in the Buddhist Monastery (New Delhi: Motilal Banarsidass, 1998), passim.

(21.) On the tridosa, see Gerrit Meulenbeld "Some Neglected Aspects of Ayurveda or the Illusion of a Consistent Theory," in Mathematics and Medicine in Sanskrit, ed. Dominik Wujaslyk (Delhi: Motilal Banarsidass, 2009), 105-17, and "The Characteristics of a Dosa," Journal of the European Ayun'edic Society 2 (1992): 1-5; Reinhold F. G. Miiller, "Uber Pitta oder Gaile unter Bezug der Tridosa-Lehre der altindischen Medizin," Janus 38-40 (1934): 77-106. "Die Bewertung der Drei Fehler dosa durch die indischen Arzte," in Jahrbuch des Museums fur Volkerkunde zu Leipzig (Berlin: Akademie Verlag, 1960). 17: 76-93, and "Eigenwertungen in altindischer Medizin," Nova Acta Leopoldina 138.20 (1958): 1-132; Dhirendranath Ray, The Principle of Tridosa in Ayun'eda (Calcutta: Banerjee, 1937); Surendranath Dasgupta. A History of Indian Philosophy (London: Cambridge Univ. Press, 1932), 2: 325-39.

(22.) Hartmut Scharfe, "The Doctrine of the Three Humors in Traditional Indian Medicine and the Alleged Antiquity of Tamil Siddha Medicine," JAOS 119 (1999): 612.

(23.) This is also implied by the etymology of the term dosa. See Miiller, "Die Bewertung der Drei Fehler."

(24.) Scharfe. "The Doctrine of the Three Humors," 625-29.

(25.) Scharfe, "The Doctrine of the Three Humors," 625-29; Miiller, "Die Bewertung der Drei Fehler" and "Eigenwertungen," 72-96.

(26.) Gerald James Larson, "The Concept of Body in Ayurveda and the Hindu Philosophical Systems," in Self as Body in Asian Theory and Practice, ed. Thomas P. Kasulis, Roger T. Aimes, and Wimal Dissanayake (New York: Suny Press, 1993), 113-16, 118-19, and "Ayurveda and the Hindu Philosophical Systems," Philosophy East and West 37.3 (1987): 245-56; Scharfe, "The Doctrine of the Three Humors," 627-29; Victoria Lyssenko. "The Human Body Composition in Statics and Dynamics: Ayurveda and the Philosophical Schools of Vaisesika and Samkhya," Journal of Indian Philosophy 32 (2004): 35-36; Dasgupta, History of Indian Philosophy. 1: 213-17. 280-81. 2: 369-73; Paul Demieville. "Byo," in Hobogirin Dictionnaire encyclopedique du Bouddhisme. d'apris les sources chinoises el japonaises, vol. 3, ed. Sylvain L?vi et al. (Paris: Maison Franco-Japonaise, 1937), 250-55: Jyotir Mitra, A Critical Appraisal of Ayurvedic Material in Buddhist Literature, with Special Reference to the Tripitaka (Varanasi: Jyotiralok Prakashan, 1985). 40-57. For a recent, concise overview of the historical development of the theory of the mahabhUtas, see Karin Preisendanz, "Mahabhutas," in Brill's Encyclopedia of Hinduism, vol. 2. ed. Knut A. Jacobsen et al. (Leiden: Brill, 2010), 806-18. I thank Dominik Wujastyk for this reference. Hindu cosmologies tend to have five mahabhutas, whereas Buddhist cosmologies usually have four. The discrepancy between the five- and four-element theories in Hindu and Buddhist cosmologies is interesting and deserves further study, especially in view of possible Greek influences on the four bhiita theory. So far, see the preliminary contributions by Jiro Endo and Teruko Nakamura, "Ayurveda no tridosa setsu to bukkyo igaku no shidai fucho setsu hikaku kento,", [TEXT NOT REPRODUCIBLE IN ASCII] Kagakushi kenkyU 34.193 (1995): 1-9; and Kido Sugita, "Bukkyo to yiryo to no kakawari" Nihon ishigaku zasshi 40.1 (1994): 1-4.

(27.) Demieville, "Byo," 249-55. In some cases, a prevalence of the earth element leads to "heaviness" (Skt. guru) instead of a "combination" of wind, bile, and phlegm: see table 1 below.

(28.) Scharfe, "The Doctrine of the Three Humors."

(29.) Kenneth Zysk, "The Evolution of Anatomical Knowledge in Ancient India, with Special Reference to CrossCultural Influences," JAOS 106 (1986): 689; Reinhold F. G. MUller, "Zur anatomischen Systematik im Yajus," Sudhoffs Archiv fur Geschichte der Medizin und der Naturwissenschaften 27 (1934): 28. For further discussion of these lists, see Katsumi Fukunaga, Bukkyo igaku jiten (Tokyo: Yuzankakku Shuppan, 1980). 37-45; Mitra, A Critical Appraisal, 45; Michael Radich, "The Somatics of Liberation: Ideas about Embodiment in Buddhism from Its Origins to the Fifth Century CE" (PhD diss., Harvard Univ.. 2007), 1402. 1405-7; Philipp Maas. "The Concepts of the Human Body and Disease in Classical Yoga and Ayurveda," Wiener Zeitsclirift fur die Kunde Siidasiens 51 (2007-8): 144 n. 69; Demi?ville, "Byo," 250-51.

(30.) For translations of this passage of the Mahasatipatthana-sutta see Maurice Walshe, The Long Discourses of the Buddha: A Translation of the Digha Nikaya (Boston: Wisdom Publications. 1995), 337-38; Zysk, Asceticism and Healing in Ancient India, 34; Steven Collins, "The Body in Theravada Buddhist Monasticism," in Religion and the Body. ed. Sarah Coakley (Cambridge: Cambridge Univ. Press, 1997), 192: Scharfe, "The Doctrine of the Three Humors." 614. For other sutras of the Pali canon where versions of this list can be found, see Scharfe, "The Doctrine of the Three Humors," 614 n. 47.

(31.) For further discussion of these thirty-two (thirty-six) elements in the Pali canon, see Fukunaga, Bukkyo igaku jiten, 443-57; Mitra, A Critical Appraisal. 45; Radich, "The Somatics of Liberation," 1402. 1405-7. Eric Greene discusses the translation of these lists into Chinese. See his "Of Bones and Buddhas: Contemplation of the Corpse and Its Connection to Meditations on Purity as Evidenced by 5th Century Chinese Meditation Manuals" (MA thesis, Univ. of California. Berkeley. 2006), 31, 34.

(32.) Paul Unschuld, Medicine in China: A History of Ideas (Berkeley and Los Angeles: Univ. of California Press. 1985), 132-53. and "The Chinese Reception of Indian Medicine in the First Millennium A.D.." Bulletin of the History of Medicine 53 (1979): 339-41.

(33.) Pierce Salguero. Translating Buddhist Medicine in Medieval China (Philadelphia: Univ. of Pennsylvania Press. 2014). 55-66; "Mixing Metaphors: Translating the Indian Medical Doctrine Tridosa in Chinese Buddhist Sources." Asian Medicine: Tradition and Modernity 6 (2010/11): 67-70; and "Buddhist Medicine in Medieval China: Disease, Healing, and the Body in Crosscultural Translation (Second to Eighth Centuries c.E.)" (PhD diss., Johns Hopkins Univ., 2010), 123-33.

(34.) Salguero, Translating Buddhist Medicine, 60.

(35.) Salguero, Translating Buddhist Medicine, 141.

(36.) Endo et al., "Tan no kigen I (--)," 336. They argue that the idea of a substantial internal (or physiological) pathogen was foreign to the early Chinese translators, because in early Chinese medicine illnesses are caused by qualitative changes of yin Pe and yang, cold and heat, etc., whereas Ayurveda was based on a substantive conception of matter, in which world and body are composed of essentially different and mutually antagonistic substances.

(37.) Zhong'ahanjing (Madhyamagama) and Zengyi'ahanjing (Ekottaragama). For the Zhong'ahanjing the evidence comes from the comparison of two parallel passages in juan 7 (T.ln0026_ p0465a27-28) and 42 (T.ln0026_p0690c22-23). (N.B. the juan numbers given by Endo et al. for these two passages are both incorrect.) Endo et al. argue (correctly, I think) that these parallel passages reveal that dan is used in the meaning of bile, not phlegm, in the list of the thirty-two body constituents of juan 42. For the Zengyi'ahanjing, which uses tan in the meaning of bile (T.02n0125_p0604b01-03), see table 1 above. The Zhong'ahanjing and the Zengyi'ahanjing were both translated by Gautama Sarighadeva in 397 and from 379-98 c.e. respectively. See Lewis Lancaster, The Korean Buddhist Canon: A Descriptive Catalogue by Dr. Lewis R. Lancaster (1979, made digitally available by CBETA). Unfortunately, neither of these sutras is extant in an Indic language, so we cannot ascertain whether dan was indeed used in the sense of bile.

(38.) "Sloshing" because the prior meaning of dan was "sloshing water." Endo et al., "Tan no kigen II," 547.

(39.) Endo et al., "Tan no kigen II," 548.

(40.) Endo et al., "Tan no kigen II," 339-40.

(41.) On early Chinese Buddhist translations see Jan Nattier, A Guide to the Earliest Chinese Buddhist Translations: Texts from the Eastern Han and Three Kingdoms Periods (Tokyo: International Research Institute for Advanced Buddhology, Soka Univ., 2008); Daniel Boucher, Bodhisattvas of the Forest and the Formation of Mahayana: A Study and Translation of the Rastrapalapariprccha-sutra (Honolulu: Univ. of Hawai'i Press, 2008); "Dharmaraksa and the Transmission of Buddhism to China," Asia Major 19 (2006), 13-37; "Gandhari and the Early Chinese Buddhist Translations Reconsidered: The Case of the Sadharmapundarikasutra," J A OS 118 (1998): 471-506; "Buddhist Translation Procedures in Third-Century China: A Study of Dharmaraksa and His Translation Idiom" (PhD diss., Univ. of Pennsylvania, 1996); Antonio Forte, The Hostage An Shigao and His Offspring: An Iranian Family in China (Kyoto: Instituto Italiano di Cultura, Scuola di Studi sull'Asia Orientale. 1995); Salguero, Translating Buddhist Medicine, 44-66.

(42.) As shown in tables 3 and 4, Suiruta describes five functions of pitta and kapha, but names only those for pitta. Caraka enumerates a number of functions that clearly are variants of the functions in Susruta, but they remain unnamed. See also MUller, "Eigenwertungen," 89; "Uber Pitta Oder Galle," 88-90; Ray, The Principle of Tridosa, 65-84.

(43.) Muller, "Uber Pitta oder Galle." 88.

(44.) Maas, "The Concept of the Human Body," 136-44; Rahul Peter Das, The Origin of the Life of a Human Being: Conception and the Female according to Ancient Indian Medical and Sexological Literature (Delhi: Motilal Banarsidass, 2003), 273-84.

(45.) For example, Maas has shown that lists that were adapted for digestion tended to exclude the constituent of "skin" and "sinews" in favor of "chyle," a substance that is intrinsically tied to ideas about transformation of food. Maas, "The Concept of the Human Body," 136-44.

(46.) Bhikkhu Nyanamoli, tr., The Path of Purification: Visuddhimagga (Berkeley: Shambala Publications, 1976), 280.

(47.) In an aside, Muller noted the qualitative difference of digestive bile in contrast to tridosic bile a long time ago, but he did not proceed to explain this difference. Muller. "Zur Feuerwertung in altindischer Medizin," Sudhoffs Archiv fur Geschichte der Medizin und Naturwissenschaften 36 (1953): 317.

(48.) Dasgupta, A History of Indian Philosophy. 2: 331.

(49.) Reinhold F. G. Miiller, "Uber indische Vorstellungen von der Verdauung," Mitteilungen des Instituts fiir Orientforschung 7 (1959-60): 213.

(50.) Meulenbeld, History of Indian Medical Literature, 1A: 114.

(51.) The project group Philosophy and Medicine in Early Classical India is working on a critical edition of the Caraka Samhita Vimanasthana and Sarirasthana. Needless to say, these two chapters represent only a fraction of the classical Ayurvedic texts.

(52.) The heterogeneity of the classical Ayurvedic compendia is generally acknowledged, but not much researched. See Dominik Wujastyk, The Roots of Ayurveda (Oxford: Penguin, 2003), xxxii-xxxiv.

(53.) Y. T. Acarya, ed. Susrutasamhita, Susrulena viracita, VaidyavaraJridalhanacaryaviracitaya Nibandhasamgrahakhyavyakhyaya samullasita, Acoryopahvena Trivikramalmajena Yadavasarmano samiodhita (Mumbayyam: Nirnayasagaramudrayantralaye, 1938), quoted from SARIT: Search and Retrieval of Indic Texts. http://sarit. (accessed on 8/29/2014). Some sandhi has been dissolved here and in other cited passages.

(54.) I translate as if dehah (plural) and asti (singular) agree, because I am unable to make sense of this sentence otherwise.

(55.) [section] 10 discusses the five-fold typology of pitta and kapha. It is omitted from the translation because it is already represented in table 3 above.

(56.) On the lateness of the positive valuation of the tridosa, see section 2 above.

(57.) For observation of further discrepancies between these layers, see Muller, "Eigenwertungen," 86; "Uber Pitta und Galle," 88; Grundsatze altindischer Medizin, 128-29 n. 323; "Uber indische Vorstellungen von der Verdauung," 219; and "Uber verschiedene Ergebnisse indischer Textuntersuchungen durch Sprachwissenschaftler oder einen Medizingeschichtler," Rocznik Orientalisticzny 30 (1967); 103.1 would like to add to Muller's observations that the metrical couplets, which are interspersed throughout the prose, present a self-contained discussion of the relationship of pitta, kapha, and blood to tastes. The discussion of these metrical couplets stands apart from, and does not interact with, the prose text into which they are interspersed, which is yet another indication that the texts we are dealing with now were pieced together from different layers of text. A. K. Warder developed a chronological framework based on the metrical analysis in his Pali Metre: A Contribution to the History of Indian Literature (London: Pali Text Society, 1967).

(58.) The above excerpt from Susruta locates the pakvaiaya "below the navel (nabha) and above the srom (pelvis) and the guda (rectum)" and the amas'aya "above the "repository of bile (pittufaya)." References to navel, pelvis, and rectum are reasonably unambiguous, and allow us to conclude that the pakvaiaya is situated somewhere in the lower abdomen. By contrast, the reference to the "repository of bile" does not clarify anything, because we do not know where it was supposed to be located. Susruta simply states that the amasaya is the place where food is received after ingestion, heated by the fire of pittasaya from below, and broken down into chyle, while Caraka specifies elsewhere that the amasaya is "supposed to be located between the navel and the breast" nabhi-stanaantaram jantor amaSaya iti smrtah (Ca Sam vimanasthana 2).

(59.) Reiterated in [section][section] 6, 15, and 16.

(60.) Suvarnabhasottamasutra 16.12. See Johannes Nobel, ed., Suvarnabhasottamasutra: Das Goldglanz-siitra. Ein Sanskrillext des Mahayana-Buddhismus, nach den Handschriften und mit Hilfe der tibetischen und chinesischen (jbertragungen (Leipzig: Otto Harassowitz. 1937), 180, and Ein alter medizinischer Sanskrit-Text und seine Deutung, Supplement to JAOS 11 (1951), 11; Ronald Emmerick, The Sutra of Golden Light: Being a Translation of the Suvarnabhasottamasutra (Oxford: Pali Text Society, 2004), 75-76.

(61.) Acarya, Maharsina Punarvasunopadista, tacchisyen Agnivesena pranita, Caraka Drdhabalabhyam pratisamskrta Carakasamhita, sri Cakrapanidattaviracitaya Ayurvedadipikavyakhyaya samvalita (Mumbayyam: Nirnaya Sagara Press, 1941), quoted from SARIT: Search and Retrieval of Indic Texts, (accessed on 8/29/2014).

(62.) I translate this as uro-gatam as I cannot make sense of uro-gata.

(63.) Neither the Suvarnabhasottamasutra nor the Caraka Samhita discuss bile and phlegm in relation to digestion in these specific instances.

(64.) Acarya, CarakaDrdhabalabhyam pratisamskrta Carakasamhita.

(65.) Meulenbeld, History of Indian Medical Literature. IA: 191-94; IB; 295-300.

(66.) Vagbhata, Astangasamgraha. tr. Luise Hilgenberg and Willibald Kirfel (Leiden: Brill, 1941), su 20, 1 and 4; Vagbhata, Vagbhata's Astangasamgraha: The Compendium of Eight Branches of Ayurveda, Text and English Translation with Illustrations, ed. Vaidya Asha Ram (Delhi: Sri Satguru Publications, 1999), 285 and 287-88; Ray, Principles of Tridosa, 79.

(67.) Claus Vogel's transliteration. I have modified Vogel's transliteration scheme to make it coherent with the Sanskrit transliteration in other parts of this article. Vogel argues that kramena refers to the sequence wind- bilephlegm. Vogel, Vagbhata Astangahrdaya-samhita, 51-52.

(68.) Vagbhata, Astangahrdaya, su 25. Vogel, Vagbhata Astangahrdaya-samhita, 67.

(69.) Interestingly, the divergent development of therapeutic vs. physiological conceptions of the dosas continues in contemporary Ayurvedic theory. One faction of scholars, among them Dwarkanath and Murti, argues that the three dosas are best understood as immaterial functions. See Charles Leslie. "Interpretations of Illness: Syncretism in Modern Ayurveda," in Paths to Asian Medical Knowledge, ed. Charles Leslie and Allan Young (Berkeley and Los Angeles: Univ. of California Press. 1992), 189. Yet. despite these novel interpretations of the dosas in physiology, (he emphasis on their connection with digestion remains central to pathology and practice. A random example is supplied by the fieldwork account of Jean Langford, who observes that "heart pain ... is considered kaphaja, pittaja or vataja depending on whether it is experienced during the early stage of digestion associated with kapha, the middle stage associated with pitta, or the final stage associated with vata" and notes that "despite the assertion by many practitioners that dosa is a function and not a substance, it frequently behaves as a substance to great clinical purpose in narratives of disease and healing." Langford, Fluent Bodies: Ayurvedic Remedies for Postcolonial Imbalance (Durham: Duke Univ. Press, 2002), 173-74.

(70.) Zhang Zhongjing, Jingui yaoliie, "Tanyin kesou bing maizheng bing zhi," see section 1 of this paper.

(71.) Endo ct al., "Tan no kigen II," 544-46. The source is not a medical work, but Ge Hong's (284-364) Baopuzi. He Shuzhen, ed., Xinbian Baopuzi: Neipian (Taipei: Guoli bian yi guan, 2002). 441.

(72.) Endo et al., "Tan no kigen II," 546-47. The sources are Chen Yanzhi's Xiaopin fang, Tao Hongjing's Shennong bencao jing and Mingyi bielu, and Ge Hong's and Tao Hongjing's Zhouhou beiji fang

(73.) Ge Hong's Zhouhou beijifang also has three occurrences in juan 4, [TEXT NOT REPRODUCIBLE IN ASCII], and [TEXT NOT REPRODUCIBLE IN ASCII].

(74.) Taki Motoyasu also suggests that tanyin in Ge Hong's Zhouhou beijifang is simply tanyin. Taki, Yisho, juan shang, 32.

Table 3. Functions of pitta and kapha in Susruta Samhita (a)

pitta                  Susruta Samhita su 21, 9-13

pacakagni-"cooking     seated in between the repositories of the
fire"                  uncooked (amasaya) and the cooked (pakvasaya),
                       cooking the four kinds of food, separating
                       do?a, rasa, urine, and excrement, and, by
                       virtue of its own power, supporting the four
                       other seats of pitta in the body

ranjakagni-"dyeing     seated in the liver (yakrt) and spleen (plihan)
fire"                  dyeing the blood red

ranjakagni-            seated in the heart (hrdaya) bringing about
"enlightening fire"    realization of one's desires

alocakagni-"viewing    seated in the pupils of the eye (dr?ti)
fire"                  catching the form of images

bhrajakagni-           seated in the skin (tvac), creating a radiant
"irradiating fire"     complexion and digesting unguents that are
                       anointed on the body

Susruta Samhita su 15, 4
raga +krt--reddening
pakti +krt--cooking
ojah + krt--vitalizing
tejah +krt--making fiery
u?ma +krt--heating

kapha                  Susruta Samhita su 21, 16-17

[kledaka-              moistens and disintegrates food that arrives in
"moistener"]           the repository of raw food (amasaya). By virtue
                       of its own power it upholds the other seats of
                       s'le?man in the body with its watery essence.

[avalambaka-           seated in the chest (uras), supports the heart
"supporter"]           and together with its intrinsic potency and the
                       chyle, it supports the threefold (trika)

[bodhaka -             seated in the throat and at the root of the
"perceiver"]           tongue (jihvamulakantha), enabling the
                       perception of taste by maintaining the coolness
                       and moisture of the sense organ that is the

[tarpaka -             situated in the head (siras), upholding the
"cooler,               sense organs (indriya) by virtue of the
satiater"]             unctuousness and freshness; responsible for

[slesaka-              situated in the joints (samdhi), embracing and
"embracer"]            holding together all the joints

Susruta Samhita su 15, 5

sandhisamslesana-connecting joints together
snehana-making fatty, glossy, sticky
ropana-causing [sores] to heal, grow over
balasthairyakrt-strengthening and making firm

(a) These tables do not represent an exhaustive enumeration of all the
instances of tridosa typologies in the Suiruta and Caraka Samhitas.
For a comprehensive list of all instances, see Gerrit Meulenbeld, The
Madhavanidana and Its Chief Commentary (Leiden: Brill, 1974), 474, 452.

Table 4. Functions of pitta and kapha in Caraka Samhita

Caraka Samhita su 12, 11 agni in    Caraka Samhita sii 12,11-12 soma
pitta                               in kapha

cooking                             firmness
vision                              development
complexion                          enthusiasm, potency
heat                                knowledge, understanding

Caraka Samhita sii 18, 50 agni      Caraka Samhita su 18, 50 kapha

vision                              unctuousness, binding
cooking                             firmness
heat                                heaviness
hunger, thirst                      potency, strength
softness of the body                forbearance, restraint, absence
radiance of the body                of greed

Table 6. Pure and Impure Body Constituents in Caraka and Susruta (a,b)

The seven pure body     Corresponding waste     Corresponding waste
constituents            products in Caraka      products in Susruta

chyle (rasa)            phlegm (kapha)          phlegm (kapha)
blood (asrk/rudhira/    bile (pitta)            bile (pitta)
flesh (mamsa)           dirt from the           dirt from the
                        cavities of the body    cavities of the body
                        (khamala)               (matah khesu)

fat (meda)              sweat (sveda)           sweat (sveda)
bones (asthi)           hair and body hair      fingernails, body
                        (kes'aloman)            hair (nakha-roman)

marrow (majjan)         oiliness of the skin,   dirt of the eyes,
                        dirt of the eyes        oiliness of the skin
                        (tvacam snehah,         (netravit tvaksu ca
                        aksivit)                snehah)
semen (sukra)
COPYRIGHT 2016 American Oriental Society
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2016 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Kohle, Natalie
Publication:The Journal of the American Oriental Society
Article Type:Report
Geographic Code:9INDI
Date:Jul 1, 2016
Previous Article:The 'Alids: The First Family of Islam, 750-1200.
Next Article:Some notes on revenge and justice in the Mongol empire and the Il-Khanate of Iran.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |