A correlative STS: Lessons from a Chinese medical practice

Introduction

Science and Technology Studies (STS) treats knowledge as expressions of practice. But what does this mean in a post-colonial context? How might one think about the overlaps and intersections between ‘Western’ and ‘other’ ways of knowing and intervening in the world? To pose the question in this way is already to risk being tugged from the specificities of practice. Since both the so-called post-colonial and the ‘West’ are multiple and diverse and their locations of overlap are indefinitely complex, the quick lesson and the starting point is that while large categories may catch something important, they also have to be treated with extreme caution. So how should we think about difference and overlap without either reifying these or washing them away into a confusion of complexity?

Post-colonial scholars have tackled this problem in a range of different ways. Dipesh Chakrabarty (2000: 5) describes how scholarly conventions in history appear general but are in fact specific to Europe and potentially inappropriate to Indian history. Europe, he says, needs to be ‘provincialized’. As a counter to such unidirectional movements, Zhan (2009) has explored the modes of material and symbolic circulation of Chinese Medicine (CM) between China and the United States since Maoism, Vincanne Adams (2002) has charted the variable and often exploitative links between pharmaceutical research and traditional Tibetan medicine, and in a quite different context Peter Redfield (2002) has considered what he calls the ‘networks of empire’ in the discourses and practices of outer space in French Guyana. Others have attended to practices, treating these as enactments that are simultaneously epistemological and ontological. Helen Verran (2001) describes the consternation that arises when Yoruba and Western practices of numbering rub up together, ¹ and shows how these are artfully intertwined in Nigerian classrooms. Elsewhere (Verran, 1998), she has looked at how Aboriginal and settler versions of land and land ownership are enacted, again in creatively hybrid ways, and at practices of conservation (Verran, 2002). Arturo Escobar (2008) has shown that nature, the social, the political and the economic are all radically unlike their equivalents for of capital and technoscience in the practices of indigenous and mestizo people in Colombia. Marisol de la Cadena (2010) has considered how different ‘political’ entities – not simply people but, for instance, Pachamama (roughly speaking, ‘nature’) – have found a place in the constitution of Ecuador. Analogous arguments have been made for the encounters between first nations and governments or large corporations in North America (Feit, 2004; Noble, 2007).

The problems thrown up by asymmetrical moments of encounter, hybridization and the recognition of difference arise within STS in chronic form for a journal such as East Asian Science, Technology and Society: An International Journal where a number of authors have explored the difficulties and suggested a range of strategies. ² They are also thrown up in cultural anthropology in the form of debates about cosmological difference. Here the work of Eduardo Viveiros de Castro (1998, 2004a, 2004b) is a marker for a much larger discussion about the character of anthropological knowledge across difference. Viveiros de Castro is not afraid of large categories. Amerindian cosmology is distinctive, he says, because it is ‘perspectivist’ and ‘multinatural’. In the multiculturalism characteristic of ‘the West’ we recognize many cultures but only one reality, one nature. In Amerindian cosmology there is only one culture but there are multiple natures. So, for instance, jaguars see themselves as human. They drink manioc beer, just like us. But we see them as jaguars and see them drinking blood. Agents (people, jaguars) have the same kind of souls (‘culture’) but live in different bodies (‘natures’). The message is that there is a single (‘cultural’) perspective on a multiple nature. So how does knowing work?

The problem for indigenous perspectivism is not … of discovering the common referent (say, the planet Venus) to two different representations (say, ‘Morning Star’ and ‘Evening Star’). On the contrary, it is one of making explicit the equivocation implied in imagining that when the jaguar says ‘manioc beer’ he is referring to the same thing as us (i.e. a tasty, nutritious and heady brew). … the aim of perspectivist translation – translation being one of shamanism’s principal tasks … –, … is not that of finding a ‘synonym’ (a co-referential representation) in our human conceptual language for the representations which other species of subject use to speak about one and the same thing; rather, the aim is to avoid losing sight of the difference concealed within equivocal ‘homonyms’ between our language and that of other species, since we and they are never talking about the same things. (Viveiros de Castro, 2004b: 6–7)

In a nutshell, the problem is this. If homonyms hide ontological difference, then how can we hold onto that difference? And how might we learn from the shaman’s concern with equivocal translation between natures to inform anthropological (or STS) thinking about difference? Indeed, the argument is that any description bridges or translates ontological difference. Our translations are always hybrids because they draw analogies between different worlds. Viveiros de Castro’s message is that the art of anthropology is to dwell (as we must) in this space of equivocation, and then to control that equivocation by recognizing it. But how can we do this well? Despite the pragmatic value of a continuum of different translations, he tells us that a good translation is one that

betrays the destination language, not the source language. A good translation is one that allows the alien concepts to deform and subvert the translator’s conceptual toolbox so that the intentio of the original language can be expressed within the new one. (Viveiros de Castro, 2004b: 5)

The message is controversial but has been enthusiastically adopted by many in anthropology (Henare et al., 2007). Brought to (post-colonial) STS, it becomes clear that we are in the business of translating across ontological difference. Our STS words are hybrid and they necessarily ‘betray’ what is said both in the places we come from and the locations we are describing. But Viveiros de Castro’s take-home rule-of-thumb is that it is better to betray the stories told in the places we come from than those of other worlds. At the risk of sounding mechanistic, we might think of this by asking about the ‘balance of betrayal’ – and perhaps more interestingly, what might count in any particular context as a just or appropriate mode of betrayal.

There are many versions of STS, and in this way of thinking, they all count as hybrid equivocations. This suggests that we need to think about what counts as appropriate betrayal case by case and tradition by tradition. Even if we stick to post-colonial STS, we can see that there is no simple answer. Thus, the discipline’s analytical tools open up difference in important respects. So, for instance, the notion of ‘symmetry’ in Sociology of Scientific Knowledge (SSK) recognizes and validates scientific difference in a way that previous epistemologically based approaches did not. At the same time, however, STS tools efface other forms of difference, for instance in the context of post-coloniality (Anderson, 2002: 643). The balance or mode of betrayal shifts. This depends on question and situation. Indeed, to talk of ‘a post-colonial STS’ in the singular is to lose specificity because it homogenizes ‘the West’ (Mol, 2002) and ignores the endlessly many alternative knowledge traditions that make up the latter. It also fails to see that if we take controlled equivocation seriously, it follows that STS will be fragmented, diasporic and multiple; that indeed, there might not be an ‘it’ at all. ³

These observations provide the background for what follows. Working within the conventions of an STS that draws on post-Actor Network Theory (ANT) and feminist material semiotics (Haraway, 1991, 1997; Latour, 2005; Law, 2004, 2008), we follow Jensen, Morita and Blok by reworking this in a post-colonial context of equivocation (Jensen and Blok, 2013; Jensen and Morita, 2012). To do this, we characterize an encounter between a version of CM and its intersections with biomedicine by exploring a consultation between a patient and a CM practitioner in Taiwan. It comes as no surprise to discover that biomedicine and CM are being mixed together. We are not exactly in the world of manioc beer and blood, but the ‘same’ words certainly index different things. Our own description is an equivocation, too, balancing between the clinic on the one hand and our post-ANT conceptual toolbox on the other. Most of what we do is framed by that toolbox, but we attempt to push the balance between the two in the direction of CM by attending to the logic of hybridity. We do this by drawing on historical writing on CM and Chinese classical philosophy to argue that the version of CM that we have investigated is correlative rather than analytical.

To avoid misunderstanding, we need to make two crucial points. First, in the way we use the term in this article, ‘correlative’ is a term of art in Chinese philosophy that has nothing in common with statistical correlation. Briefly, while the latter seeks patterns of regularity across populations, sometimes in the hope of discovering causal relations, correlativity within (at least some) CM practices refracts a pattern of practice that hybridizes without purifying; gives priority to situated knowing; enacts a version of the body that is non-reductive even in principle; weaves patterns between what Chinese philosophical tradition calls ‘the ten thousand things’ (wàn wù, 萬物); and, finally, adopts a dynamic approach to propensities at work in specific situations. Correlativity in CM derives from its specific world view(s), and the situated enactment of the relationships between things with their shared propensity/qi alongside alternatives; for example, winter correlates with the kidney (meridian) that stores qi because ten thousand things hide and conserve themselves in this season, and water correlates with the season and the meridian because it is cold by nature and tends to go down; kidney (meridian), water or winter are correlated with each other in sharing the propensity of conserving. Recently CM has been under pressure to demonstrate its scientific status, and is increasingly analyzed through clinical and laboratory trials that rebuild its efficacy in terms of statistical correlations (Lei, 1999; 行政院農業委員會花蓮區農改場, 2009; 雷祥麟, 2010). We will have more to say about this below.

Second, it is also important to understand that to contrast statistical and ‘Chinese’ correlations in this way is itself to use a ‘Western’ STS-inflected analytical framework. This is for the fundamental reason that it is not possible to exemplify correlative reasoning within the conventions of an STS journal. Correlativity demands quite different institutional contexts and modes of authority. ⁴ Nevertheless, our argument is that this descriptive equivocation suggests the possibility of a different and correlative STS, and we conclude by briefly characterizing the form that this might take. In this paper this can be no more than a gesture, but our reasoning is that if Viveiros De Castro was in a position to learn from his post-structuralist adventures in Amerindian cosmology, there is no reason why STS cannot learn from a Taiwanese medical practice. Such a correlative STS would be a post-colonial STS, no doubt just one of many.

Our case comes from consulting room talk between a patient and a popular and respected Taiwanese CM practitioner, Dr Lee. The data are drawn from a continuing empirical study of CM practices by Lin who has undertaken weekly observations in Dr Lee’s clinic over a four-year period. Lin has also audited Dr Lee’s university CM course, consulted with Dr Lee about the CM classics and participated in training sessions held for local CM doctors. The data we offer in what follows are specific. We are neither seeking to make general empirical claims about CM nor to add to the body of knowledge about CM. Rather our object is to use a moment in a CM practice to illustrate the character of correlative reasoning, and show how the latter might be used as a controlled equivocation to rework the mode of betrayal within STS and so to reinterpret the character of Dr Lee’s practice.

Hybridities

Mr Wang has been on dialysis for twenty years. Three years before the consultation, he started visiting CM practitioners after a major health crisis that biomedicine failed to treat. Now he visits Dr Lee weekly.

He enters Dr Lee’s clinic, registers with his National Health Insurance (NHI) card and sits in the waiting room. There are around ten people there already. Dr Lee is popular, so patients often have a long wait. She is thorough, takes her time with each patient, and is one of best-known CM practitioners in the city. In the waiting room there is a display cabinet with rare and precious samples of Chinese medication. Some come from endangered species and are no longer used in the clinic. There are also copies of Ministry of Health licences and certificates for Dr Lee and her colleagues, and a certificate about the electronic patient record system (the latter is unusual in a CM clinic.) Other certificates reveal that Dr Lee and her colleagues are graduates of the Chinese Medical University. This means they have studied both biomedicine and CM. Then there are noticeboards with CM and biomedical posters about osteoporosis and diabetes, quality control reports about the safety of ‘scientific Chinese medication’ (kē xué zhōng yào, 科學中藥, hereafter SCM), and details of the blood tests done in the clinic.

Scientific evidence-based biomedicine is often distinguished from traditional experience-based CM, but this is a recent distinction and is not what is happening here. ⁵ People come for CM, but Dr Lee, with her double training, also offers biomedical examinations. ⁶ Traditional herbal and SCM are prescribed. CM is being mixed with biomedicine professionally and pharmaceutically. CM practice is being hybridized. How might we think about this?

One obvious response is that this is an expression of imperialism. STS has often described the hybridity and heterogeneity of imperialist science, technology and medicine. Actor-networks transported Newton’s laws of physics from England to Gabon (Latour, 1988), and the moving metropolis hybridized with the colonies to make tropical medicine (Arnold, 1993; Stoler and Cooper, 1997). The waiting room in the clinic fits with this. Indeed, it tells the subaltern side of the story, since its heterogeneity witnesses the weight of the colonial and post-colonial history of Taiwan and China, and the imposition of biomedicine in these countries. In Taiwan, this started in the 1860s with Western missionaries, doctors and gunboats; biomedicine was institutionalized during Japanese colonialism after 1895, when CM was nearly eliminated. After 1945, the post-colonial republic created a biomedical health system that included CM. The project was chaotic and controversial. There were debates about the division of labour between CM pharmacist and doctors, the legality of using biomedical instruments in CM clinics, the need to rebuild the administration and regulation for certificating existing CM practitioners, and most important of all, what a modern institution for training new generations of CM doctors should be like. The process was slow, and it was not until 1958 that CM practitioners sought to revive their tradition by establishing the first CM institution, the Chinese Medical College (where Dr Lee studied). Even so, CM needed to adapt to the administrative, epistemic and curricular practices of biomedicine (Hsu, 2001; Scheid and MacPherson, 2012; Taylor, 2005; Zaslawski and Soo, 2012; 王致譜 and 蔡景峰, 1999; 何小蓮, 2006; 林昭庚, 2004; 葉永文, 2013).

From this we learn that in Taiwan, CM knowledge and diagnosis have been hybridized with modern science and technology; the traditions of clinical mentoring and reputation-based practice have been replaced by curricular education and exam-based certification; and CM medication has been transformed by biomedical analysis and located within the political and economic concerns of biotechnology (Hsu, 1999, 2001; Kim, 2006, 2007; Lei, 1999; Scheid, 2002; Zhan, 2009). The message is that the waiting room reflects the fate of CM in a global world; it is a sign of its subordinated hybridity. Seen in this way, Western biomedicine mangles (Scheid, 2002), translates (Farquhar, 2012; Kim, 2006; Lei, 1999) or hybridizes its indigenous alternatives (Zhan, 2009). Perhaps it also seeks to purify the colonized body and turn it into a set of objective variables/valuables (Anderson, 2013; Arnold, 1993). Overall, however, it counts as another instance of modernism at work (Latour, 1993).

But how does this account look if we start to move the balance of betrayal? The answer is complicated. On the one hand, if we use the STS principle of symmetry to describe CM and biomedicine, then this shifts the balance in the direction of CM: the latter is being taken seriously rather than being dismissed as error. On the other hand, to do this is to work in ‘Western’ STS terms. Unless all descriptions qualify, perhaps it is not fair to call this equivocation ‘epistemic violence’ (Spivak, 1988). But what differences is it concealing? And how is it shifting STS’s own toolkit? The answer to the last question is not obvious, so let us think about how describing this encounter might move the equivocations of the STS repertoire. To do this, we ask what is distinctive about the hybridity in this clinic.

Correlations

Mr Wang goes into the consultation room. Dr Lee is sitting at a desk with a glass desktop and a computer underneath. On her right there is a small pillow and a torch. On the left she has an IC card reader. She welcomes Mr Wang, inserts his IC card and starts to take his pulse (bǎ mài, 把脈). He rests his left wrist, inner side up, on the pillow and Dr Lee gently presses the tips of the index, middle, and ring fingers of her right hand close together to his radial artery for around ten seconds. Then she does the same with his right wrist. She moves to and fro between his hands a few times and looks at the top and bottom of his tongue with the torch. Finally, she says: ‘As usual your pulsation is string and deep (chén xián, 沉弦), but it’s a bit faster and stronger’. She types a note about this and about Mr Wang’s tongue into her computer: ‘white, thin coating’. Then she asks: Have you been sleeping well recently?

A nurse comes with a haemadynamometer, while, Dr Lee feels his pulsation in both wrists again. Then the nurse speaks: ‘It’s 102 over 78’. Dr Lee repeats: ‘102 over 78’, and keys the data into her computer. Then she checks the records, looks up at Mr Wang and asks: ‘How about the problem with your leg? Have you brought the examination results?’ Mr Wang gives her his blood test result and a leg scan report.

This is hybridity again. Dr Lee does the ‘four diagnoses’ (sì zhěn, 四診). She smells and listens, looks, asks, and feels Mr Wang’s pulse. This is CM at work. But she also takes his blood pressure and looks at the scan and blood test results. These belong to biomedicine. The different logics and cultural contexts of pulse-taking in biomedicine and CM have been widely explored (Kuriyama, 2002), but how does this work here?

The haemadynamometer and blood tests are tools for specific ways of knowing the body. Their results may lead directly to a diagnosis because they indicate what is wrong with – and usually within – the body: they are used to tease out the underlying causes of ill health. But the four diagnoses work differently. First, they do not look for direct underlying causes. No individual sign leads to a specific diagnosis. Instead, CM explores the person in a specific and located composition of embodied, emotional and social correlations (Zhang, 2007). Particular signs and symptoms are associated with diet, sleep, excretion, lifestyle, the emotions and the practitioner’s own training and contexts. So the four diagnoses do not see symptoms or bodies in causal contexts, but in situations that are themselves complex and correlative (Farquhar, 1994). So Dr Lee looks for what might be causing bodily disturbance, but she is not looking for an individual cause but trying to locate this in a specific context of correlative composition. This extensive situated correlativity is the second feature of CM. It is hybrid but it is also correlative.

But what should we make of the fact that Dr Lee checks both pulsation and blood pressure? One answer is that there are varying degrees of hybridization between CM and biomedicine. At one end of the spectrum, some CM doctors insist on traditional forms of practice. They prescribe herbal medication, perform the four diagnoses, avoid using biomedical instruments and even write notes in classical language using classical calligraphy. At the same time, others modernize CM by working analytically, using experimental methods and the technologies of modern science and engineering (Lei, 1999; Lei et al., 2012; Ward, 2012; 黃進明, 2007). As we can see, Dr Lee works somewhere between these two extremes. She keeps her medical records on a computer, prescribes SCM, uses biomedical devices and reads biomedical reports, but she still performs the four diagnoses. She also talks of ‘circulation of the blood and qi’ and uses situated correlative reasoning to specify problems.

How should we understand this? Our suggestion is that we should neither worry about the presence of biomedicine, nor reify biomedicine and CM as two separate unities, since in practice both are more or less messy (Farquhar, 1994; Mol, 2002, 2008; Scheid, 2002). Instead, and adopting what one might think of as a ‘situated reification’, our tactic is to ask what elements of ‘biomedicine’ and ‘CM’ are actually doing in specific situations. It is almost impossible to avoid all reifications while moving between conceptual systems, because completely dissolving into complexifications loses something important about fairly systematic differences (as between ‘CM’ and ‘biomedicine’). Authors who write about ‘China’ from a ‘Western’ perspective all run into a version of this difficulty (Hall and Ames, 1995). Here the distinction between (‘Chinese’) ‘correlative’ and (‘Western’) ‘analytical’ practices is a ‘situated reification’. Our focus is on how hybridity might work in practice – and then on how this might help us to think about more appropriate modes of betrayal.

Modes of equivocation

Mr Wang complained that the backs of his legs were sore and that when he walked his right heel ached. His biomedical dialysis doctor suspected tendon injury, neuropathy caused by parathyroid hyperactivity, or arteriosclerosis and ordered the biomedical tests we mentioned above. Dr Lee encouraged Mr Wang to go for the tests, but also doubted whether the dialysis doctor was right. She thought that the problem lay with the kidney meridian. ‘It is the end of autumn’, she said, ‘and winter is coming. It is time to rest the body. It is time for the “yang qi” (yáng qì, 陽氣) to go into the body to be conserved and renewed in the spring. The kidney (meridian) is where this is stored. But you have been working too hard. You’ve been under pressure. This means that qi is flowing out of the kidney instead … Your pulsation tells me that. You are sore and you ache near the kidney meridian. And aching is also a sign of lack of qi to warm your body. You need’, she concluded, ‘to go to bed earlier and to work less hard’.

The biomedical scan revealed no sign of arteriosclerosis. The parathyroid hormone was far too low to cause neuropathy. The dialysis doctor said that there was no particular problem, so nothing could be done. He advised Mr Wang not to walk too much. Back in the CM clinic, Dr Lee disagrees when Mr Wang says that he thinks the tests were a waste of time: ‘The tests have eliminated some possibilities. We will stick to the original plan. We’ll stick with my previous diagnosis, the problem with your poo, and your back pain … The root of the problem, revealed in your pulsation; the pulsation at the “chi” position (chě, 尺) in both your wrists is always deep, but it is stronger on the left. This shows that you are constantly drawing out energy to keep your body going on a daily basis … The pulsation tells us about the overall dynamics and function of the meridians, but it doesn’t tell us about all somatic morbidity, particularly if you’ve got a problem that hasn’t yet caused any dysfunction. So we can also make good use of biomedical tests …’

We have already seen that the tests do different things in the two clinics. In the dialysis, clinic they work by finding out what is wrong inside the body. The assumption is that surface appearances conceal causal realities. The aim is to discover and to represent these, to analyse them and if possible to intervene to put them right. So scans look inside legs and search for signs of clotting; blood tests look for abnormal thyroxin increases to find out whether the parathyroid is malfunctioning. They work by finding relatively simple causal contexts to explain the symptoms. If they find nothing, then there is no causal link. Perhaps, then, Mr Wang is lucky. There is nothing detectably wrong with his biomedical legs, though the downside is that this means that nothing can be done.

The logic of practice in this CM clinic is different. Dr Lee adds the biomedical results to her findings. They supplement her diagnosis. Yes, there is a biomechanical body with its specific anatomy. But she also works with a body that has circulating qi and meridians. ⁷ This means that she is relating two kinds of bodies together. But how? The answer to this question is crucial. We want to suggest that Dr Lee is not wrestling with a colonizing body on the one hand, and a colonized body on the other. She is not reducing the qi body to a biomechanical alternative. She is not setting them up against one another in an attempt to generate relatively simple sets of causes. Instead, she is correlating them. She is putting at least parts of them alongside one another and relating them contextually and correlatively.

Despite our situated reification, though biomedicine tends to present itself as causal and analytical, in practice it is less reductive. There are too many tools, diagnostic systems and modes of representation for a ruthless reduction to work in practice. Indeed, much of biomedical practice is about tinkering different kinds of causes together (Mol, 2002, 2008; Mol et al., 2010). Even so, the aim of much diagnosis is to find – or rule out – direct causal explanations for symptoms. The body is decomposed into elements or processes in order to explore these. So how do qi and its flows fare in a world organized around biomedical causes? The answer is, not well at all. The meridians and their flows are not explicable in terms of biomechanical mechanisms. They do not belong within its understanding of the body. Neither are there technologies within biomedicine to detect meridians or their possible imbalances. The consequence is that there is no space for qi in the biomechanical universe. It simply does not exist.

As we have just seen, in Dr Lee’s CM practice, it works differently. It is not simply that there is room for meridians, flows of yang qi, and indeed thyroxin levels. It is also that she works the (hybrid) elements that she discovers together correlatively. The emphasis is not on hidden elements with fixed mechanisms. Neither is it on relatively simple causes. Instead it is on the ever-changing and situation-specific character of things and their relations. CM does not reduce things to essences. Instead the logic is processual and situational. Kaptchuk puts it thus:

The Chinese assume that the universe is continuously changing. Its movement is the result not of a first cause or creator, but of an inner dynamic of cyclical patterns. Just as the sun maps our four distinct seasons in its yearly round, so all biological organisms go through four seasons in a lifetime: birth, maturation, decline, and death. The constancy of the cosmos is in these patterns of change, which are regular. The cosmos itself is an integral whole, a web of interrelated things and events. Within this web of relationships and change, any entity can be defined only by its function, and has significance only as part of the whole pattern. (Kaptchuk, 2000: 15)

This has an interesting and possibly unexpected consequence for those schooled in analytical traditions. It means that entirely new kinds of things are able to enter the world – albeit with debate and controversy. They may come from or relate to the anatomical and physiological body, with all its elements and its pathways (Henderson, 1984; Lei, 2014; 皮國立, 2006, 2012). How does Dr Lee think of that body? The answer is that she handles it in a correlative manner by working processually, situationally and functionally. ⁸ So she takes the test results and relates these to CM’s body of meridians and flows. Those results do not challenge any fundamental CM commitments or understandings about the nature of the world or the body. And, as a consequence, they are not excluded. On the contrary, they are welcomed in and incorporated (馬光亞, 1998). This is correlativity at work. There are endless things. The issue is always how to relate them together in ways that work in the situation in question.

Although it comes in many forms, there is nothing new about correlative reasoning (Lei, 2014; 皮國立, 2006, 2012). Indeed at least one version of CM has been transforming itself correlatively since the Han dynasty. The Yellow Emperor’s Inner Canon (huáng dì nèi jīng, 黃帝內經) was one of the earliest collections that incorporated theories of yin-yang (yīn yang, 陰陽) and five phases (wǔ xíng, 五行) into an assemblage of therapeutic practices and principles. The Inner Canon probably first compiled records of the five schools of ancient medical practice of the Warring States Period (475 to 221 bc) (山田慶兒, 2003). Note, five schools. The five phases and meridians were added and changed during the Qin (221–206 bc) and the Han Dynasties (206 bc–24 ad) (徐復觀, 1993; 廖育群, 1994). Although they are loosely connected, different versions of the qi body and the meridians have thus been heterogeneous from the beginning (李建民, 2000). With the logic of correlativity in yin-yang and the five phases, different schools of practitioners developed and multiplied the CM legacy in the form of collections of medical case records (yī àn, 醫案), interpretations of classic texts and (more recently) new biomedical facts (Sivin, 1987).

Our situated reification thus suggests that both biomedicine and CM work by putting things together: by hybridizing. To the extent that they are descriptive, they both count as modes of equivocation. But biomedicine imagines a bodily coherence beneath the symptoms, albeit in practice often failing to find this. It assumes that symptoms are caused by particular mechanisms, which means that a major task of medicine is to reveal those mechanisms. The equivocations of CM exemplified in The Yellow Emperor’s Inner Canon work quite differently. Its practices exhibit some principles– yin and yang, the five phases and meridians, and even some kinds of anatomical structures – though it does not build on these, and neither does it seek to rebuild the body as a series of linear mechanisms. Instead, it works by correlating whatever is present within and beyond the body in situated practice. Adding the parathyroid gland? This is no problem either in theory or in Dr Lee’s practice. This is a mode of equivocation quite unlike that of biomedicine.

Propensities

In the CM assemblage of theories, the somatic structures of a qi body are not the point. As we noted above, the body is a set of processes such as digestion, breathing and thinking. Those processes are aggregated and associated with the circulation of two primary functional entities, qi and xue (blood), which move between the visceral systems of the five zang (wǔ zàng, 五臟) of heart, liver, spleen, lung and kidney and the six fu (liù fǔ, 六腑) of the gallbladder, stomach, large intestine, small intestine, bladder and three burners (sān jiāo, 三焦) ⁹ in the twelve meridians. ¹⁰ The relations between the visceral systems correlate with the dynamics of five phases (wood [mù, 木], fire [huǒ, 火], earth [tǔ, 土], metal [jīn, 金] and water [shuǐ, 水]) of qi dynamics. Indeed, all the phenomena in the universe and nature can be correlatively assimilated to these five elemental qualities, so lines of correspondence can be drawn between meridians, directions, colours, climates, musical notes, emotions, tastes, sense organs and parts of the body.

What should we make of this? One answer is that, like Viveiros de Castro’s jaguars and people, biomedicine and CM engage in different worlds. But the styles of those worlds or their modes of equivocation are also different. Unlike the biomedical analytical style, it is the ten thousand things following the correlative dynamic of yin-yang and five phases that constitute Chinese medical worlds. So how are correlations done in this encounter?

She goes back to her screen, types in the rest of the symptoms and starts to prescribe medicine. However, what she writes doesn’t have to do with qi or meridians. Instead she uses the World Health Organization’s International Classification of Diseases (guó jì fēn lèi, 國際分類) and types ‘30742, 5649, 7291’. ICD diagnostic entries get thrown up on her screen, entries such as ‘persistent disorder of initiating or maintaining sleep’, ‘unspecified functional disorder of intestine’ and ‘myalgia and myositis, unspecified’.

This can be read analytically, in which case it tells a colonizing story. This is a world in which there is no room in the International Classification of Diseases (ICD) for qi or liver fire or wood. As Sivin puts it, ‘modern scientific medicine replaces part of reality. It creates new facts, and destroys the facticity of the old ones’ (Sivin, 1987: 198). ¹² But it can also be read correlatively, and this is being done by thinking about propensities. But what are propensities?

Guang-ya Ma (mǎ guāng yǎ, 馬光亞) taught Dr Lee at the Chinese Medical University. In the Preface of his well-known textbook on diagnosis, he observes that

The difference between Western and Chinese medicine is the distinction between ‘examination’ (yàn, 驗) and ‘pattern’ (lǐ, 理 ¹³ ). The examination … investigates details with scientific methods … technologies and methods become ever more innovative and precise … and can examine everything in the greatest detail …. [By contrast] Chinese medicine … is all about ‘pattern’ … It is accumulated from experiences, implicitly building on and systematising the fundamental principle that ‘the full will empty and the depleted will grow’ (yíng xū xiāo zhǎng, 盈虛消長) in the study of yi (yì xué, 易學, i.e. The Book of Changes, 易經)’. (馬光亞, 2006: 3–4)

The key word here is pattern:

Yin and yang are the halves of two sides. You compare them and you elaborate them, and then you know the pattern of yin and yang; you understand depletion, excess, coldness and heat. What is ‘depletion’? To be ‘depleted’ means that the body is deficient, and that is yin. What is ‘excess’? ‘Excess’ means that pathogeny and evils predominate, and that is yang … Chinese medicine can improve any difficult disease with such methods of syndrome differentiation. Many problems that lie beyond the power of [Western medicine] can be resolved [by Chinese medicine]. (馬光亞, 2006: 7–8)

The yin-yang dynamics of the body and the ten thousand things are complex, and Ma’s words are little more than a gesture. At the same time, Ma is not the only CM practitioner who has attempted the equivocation of describing correlations analytically. The Fundamentals of Chinese Medicine, part of the Eleventh Five-Year National Plan from the People’s Republic of China, observes that the rhythms of the seasons are fundamental to yin-yang correlative propensities (王秀, 2011: 32–37). Then it adds,

So yin and yang are correlative propensities. Mutually dependent, in tension, included in each other, and in a dynamic and reciprocally transforming balance, they offer a rich set of correlative metaphors for thinking about the endless movements in the world and the body as creative and complementary tensions or propensities. When they are trained, CM doctors learn strategies for transforming propensities into interventions. They are taught (in different ways in different schools) to observe and explore the situation of the patient: ¹⁴

The analytical phase … must be seen as opening a range of possibilities that are variously deployed according to the conditions of the moment. These conditions naturally include the habit and the training of the doctors as well as the manifestation of illness with which he is dealing. (Farquhar, 1994: 134)

And this is how Dr Lee works. Mr Wang’s problem becomes a serious imbalance between the kidney, liver and stomach and spleen meridians, and she prescribes the specific prescription needed to tackle that imbalance. Her diagnosis, plan for treatment and the medication all have to do with re-arranging the unfolding of propensities. As she gathers biomedical tests results and asks Mr Wang about his daily life, his diet and emotions, and feels his pulse, she is locating relational propensities and imbalances. She is working correlatively. Those correlations include having been on dialysis for many years, a sore leg and a lack of yang qi in the kidney (shèn yáng xū, 腎陽虛); a general lack of yin and too much fire (yīn xū huǒ wang, 陰虛火旺). And there is more. She works by adding biomedical examination to her diagnosis, explaining the ambiguous meaning of the liver and gastric system to patients, and using ICD to register traditional practices. Thus, biomedical technologies, reports and the concepts in this encounter become parts of the pattern of propensity without unearthing a transcendent reality underpinning appearances. This is how an alternative balance of betrayal is done in a contemporary Chinese medical encounter.

A Correlative STS?

In this article, we have used the tools of STS to explore a standard post-colonial issue. While entering reservations about the terms, we have asked how to think about the intersections between ‘Western’ and ‘Other’ knowledge practices. In particular, we have asked how STS might make sense of that intersection. There is a straightforward way in which the STS commitment to symmetry puts right the assumption that ‘Western’ forms of knowledge and its practices are better grounded than those of ‘Others’. Each has to be taken seriously and understood in the same terms. But the attachment to symmetry opens up a second problem. If ‘Western’ and ‘Other’ knowledge practices are understood in the same terms, then the question is, on whose terms? STS is fragmented and comes in variants, but historically it is a fairly recent product of work by Western academics. It has not grown out of (say) Amerindian societies, Taiwan or The People’s Republic of China. In short, to understand post-colonial knowledge relations in STS’s terms is to understand these in a ‘Western’ manner. As we noted in the opening of this article, this has caused anxiety among non-Western, including East Asian communities committed to exploring ‘their’ STS (Fu, 2012). So, how might we think about this?

Our answer has been to draw on the work of anthropologist Eduardo Viveiros de Castro and his STS interpreters, including Casper Bruun Jensen. Like many, Viveiros de Castro argues that anthropological description is translation. Anthropologists tell stories in their language about other worlds, but translations are also betrayals, since equivalence is impossible. To do justice to the other knowledge traditions, we must do more than including them and must acknowledge the complexities and multiplicities. The issue becomes how to distribute betrayals. How far should we betray those whom we are describing? How far should we betray our own methodological and theoretical languages? And what might count as an appropriate form of betrayal? For Viveiros de Castro, a good translation is one in which the anthropologist betrays her own languages, methodological and conceptual toolkits, and practices. She shifts, she changes and she moves on.

As we suggested in the introduction, though it has its defenders within anthropology, as a general rule this seems too simple. Do we always want to betray ourselves? This is not obvious. It depends on what we are trying to do or, more subtly, on how we want to betray ourselves. But the larger point has to do with ontological difference. The problem with translations, and more generally with descriptions, is that they are equivocal. Homonyms – the same words – index different things but the problem is that those differences tend to get lost. The inspiration for Viveiros de Castro is Amerindian: cosmologies are multinatural, meaning that worlds are multiple and ontologically different. It is talking, subjectivity and culture that are singular. This means that descriptions are always equivocations. ¹⁵ And this dovetails with those STS traditions that discover ontological difference not only between the ‘Other’ and ‘the West’ but also within the latter, where STS homonyms index and conceal different realities. ¹⁶

How should we think about the issue of post-colonial encounters in STS? Our first response in this article has been characteristic of STS. We have worked with practices and specificities. In some sense, yes, ‘Taiwan’ and ‘the West’ or ‘Chinese Medicine’ and ‘biomedicine’ exist and inform our focus. We use these strategic reifications to explore the intersection of these big categories and realities. But we have also stayed with specificity by describing a particular consultation in a particular medical practice in Taiwan. Inflected by the STS commitment to symmetry and our own material semiotic sensibilities, we have argued that CM and biomedicine are combined in that consultation. We also touched on Taiwan’s colonial history. We observed that CM was pressed to the margins. The description is interesting, though scarcely surprising: the version of the assemblage of CM at which we have looked becomes a hybrid that is also being subordinated; such is the first step. At this point, we are firmly in the Western STS tradition. Our translation, equivocal though it is, takes CM seriously by pulling it into the STS apparatus. The balance of betrayal is no doubt complex, but the STS framework has barely shifted. But consider this. Unlike biomedicine, CM has always worked without presuming an ontology of the body in the form of a hidden reality. Instead, the body has opened itself and correlated with everything connected to it in an endless web of practices. In short, CM has always hybridized correlatively. So what happens if we use this mode of correlative equivocation to describe Dr Lee’s clinic? Do we still see subordination? This is an open question, but there is at least a case for saying that the answer is, no we do not. Correlatively, we are no longer looking at domination. We are no longer looking at subordination. Instead, what we are looking at is CM business as usual.

In a second move, we have attended to the style or mode of equivocation in the clinic and argued that this is distinctive: that it is correlative rather than analytical. Dr Lee does not seek specific causes for symptoms in determinate somatic structures, as would be common in a general practitioner (GP) consultation in (say) the United Kingdom. Instead, she works on a processual body-in-its-environment. For her, everything is related to everything else, and there are indefinitely many possible correlations. It becomes the job of the practitioner to sense out and influence imbalanced correlations by working upon the propensities. And since Dr Lee is working in terms of endlessly complex correlations relating to the body, we have shown that it is easy for her to add elements of biomedicine to the mix. There are no deterministic causal pathways or patches waiting to be undermined by such additions. While biomedicine keeps unearthing underlying mechanisms and entities and challenging CM, the latter manages ontological clashes time after time. This way of talking about a mode of equivocation still skews description in favour of STS. Nevertheless, this is the point where it becomes possible to see how that balance might shift. Indeed, since ANT has been more interested in associations than in causes or their equivalents, that tradition is helpful here (Jensen, 2014). Perhaps this renders it relatively open to the non-coherence of equivocation: to ontological difference. Certainly since ANT does not look for hidden or causal processes, and it is almost fiercely non-reductionist, the idea of correlation is not so difficult. However, this is where we need to remind ourselves that the correlativity of CM is not a matter of statistical co-variance. Instead, it draws on and exemplifies a radically different and rich correlative world. And this is what we have worked towards in the article. Using the CM consultation as material to think with, we have reached the point where we can ask what a correlative STS might look like. We have not got there yet, but this is the cusp, the place where the mode of betrayal starts to move. It is our particular version of Viveiros de Castro’s controlled equivocation. It is the move that starts, and unpredictably, to shift the toolkit of STS as it translates between post-ANT and CM.

So what might a correlative STS look like? The art, as Ma noted, is to grasp the pattern of things. This resonates with the teachings about propensity of the Tao (道) by Lao-tzu (lǎo zǐ, 老子):

In a more analytical version, sinologist Graham explains,

The great interest of system-building of the Ying-Yang type, odd as its results may seem, is that it tries to lay out explicitly the full range of comparisons and contrasts which other kinds of thinking leave implicit. Simply to apply a common name one has already to be classing as similar and distinguishing from the dissimilar. (Graham, 1986: 2)

This suggests that correlation is a rich toolkit for describing – and working upon – patterns of association. Such is the character of Dr Lee’s consultation: the mobilization of rich and complex relations, but now the post-colonial shift. If we draw on the correlative to inflect our own academic toolkits, we are moving towards a style of knowing or a mode of equivocation that is different – a potentially correlative STS. In such an inquiry, what counted as knowing – the mode of translation – would start to move. Recall our comment above that knowing in the CM tradition has rested on clinical apprenticeship rather than a formal curriculum. Practitioners have depended on reputation rather than examination and certification. The CM assemblage has also been disparate and multiple: there are endlessly many ways of working with and on the correlations that form its corpus. And as its history suggests, it has been syncretic and additive too. In her practice, Dr Lee absorbs biomedical tests and test results: ontological heterogeneity is no problem at all. If we were to bring this home, then STS – in this possible correlative version – would begin to change socially, organizationally, epistemically and didactically. What counted as an ‘explanation’ would be different. Of course, as we have noted, ‘Western’ STS is not a monolith and has to be re-situated as well. But the implications of this novel multiplicity would themselves be multiple.

So we see subordinated hybridity in Dr Lee’s clinic. But we also see managed ontological heterogeneity, the equivocal translation of syncretic patterns, propensities and the correlativity of ten thousand things. These are what are thrown into relief in an alternative and correlative STS. We do not necessarily want to make this move. There can be no general rules. There will be other locations or interventions that resonate with the story of domination, and equivocations may come in violent forms. In any case, we would need to think about how correlativity recognizes and handles hierarchy. But the very suggestion that the world looks different, indeed that the world is different, hints at the possibilities and uncertainties that follow once we start to rework the mode of betrayal. The descriptions offered in a correlative STS might look very different.