Cooperation, risk, trust: A restatement of translator ethics Anthony Pym Rovira i Virgili University, Spain  and University of Melbourne, Australia A B ST RAC T Within the general approach known as translator ethics, complementary roles are played by the concepts of cooperation, risk, and trust. Cooperation, as a technical term, describes the attainment of mutual benefits as the desired outcome of an interaction, indeed as the foundation of social life. In translator ethics, the aim is more specifically to enhance long-term cooperation between cultures. The concept of risk is then used to think about the probabilities of that general aim not being obtained and what kinds of strategies and efforts can be employed to avert that outcome by increasing mutual benefits. Trust, finally, characterizes the relationship that translators must have with those around them in order for them to contribute to cooperation, such that the most critical risk they face is that of losing credibility. Together, these concepts are able to address some of the thornier issues in translator ethics and provide a frame for ongoing discussion and research. Keywords: translator ethics, cooperation, risk management, trust, translator decisions Sodelovanje, tveganje, zaupanje. Nova opredelitev prevajalčeve etike I Z V L EČ E K V okviru širšega pristopa, imenovanega prevajalčeva etika, se pojmi sodelovanja, tveganja in zau- panja med seboj dopolnjujejo. Sodelovanje kot tehnični termin pomeni, da je zaželeni cilj vsake interakcije, še več, da je temelj družbenega življenja doseganje vzajemnih ugodnosti. V okviru pre- vajalčeve etike se zasleduje še bolj specifični cilj, in sicer krepitev dolgoročnega sodelovanja med kulturami. Koncept tveganja se uporablja za presojo možnosti, da se ta splošni cilj ne doseže, in za razmislek o tem, kakšne strategije in napori se lahko uporabijo, da bi se takemu izteku izognili na način, da se vzajemne ugodnosti še povečajo. Zaupanje pa označuje odnos, ki ga morajo prevajalci imeti s tistimi, ki jih obkrožajo, da lahko prispevajo k sodelovanju, saj je največje tveganje, s katerim se soočajo, izguba kredibilnosti. Skupaj ti trije pojmi omogočajo obravnavo nekaterih težjih izzivov prevajalčeve etike in predstavljajo okvir za nadaljnjo razpravo in raziskovanje. Ključne besede: prevajalčeva etika, sodelovanje, obvladovanje tveganja, zaupanje, prevajalčeve od- ločitve 5Stridon. Journal of Studies in Translation and Interpreting, Volume 1 Issue 2, pp. 5–24 I beg forgiveness for the retro three-balls-in-the-air title. The concepts are indeed to be juggled, each in the air in its own time, yet going around together. My pragmatic purpose in using them is to formulate ethical guidelines that might help translators decide between alternatives. My more academic mission is to allay a few misunder- standings and refresh a framework for discussion.1 1. Introduction Over the past few decades I have been working on an ethics of the translator. The gen- eral approach might be called a “translator ethics”, focused on relations between peo- ple, possibly opposed to an “ethics of translation”, which would work from relations between texts. In the course of my work, I have used several related concepts as points of anchorage, mostly as ways of thinking about what translators should or should not do. To summarize the trajectory in very broad terms, I first talked about “cross-cul- tural cooperation” as an ideal that translators should strive for; I then spent several years trying to apply risk management to what translators do, generally claiming that translators have to manage the probability of non-cooperation; and more recently I have been working with concepts of trust, since translators cannot achieve coopera- tion without being trusted, which means that the major risk they have to manage is perhaps that of losing credibility. How those three concepts actually relate to each other is a little more complicated. My purpose here is to trace the ways in which cooperation, risk, and trust can be placed within a fairly unified approach to translator ethics. My presentation will be partly autobiographical, although not entirely in search of self-justification. I also hope to show what kinds of intellectual climates have provided groundings for theory. There was context then, and there is new context now. At the time of writing, surrounded by a pandemic, the basic concepts of cooperation, risk, and trust acquire renewed resonance. When members of a society take actions like wearing masks, respecting social distancing, washing hands, or being vaccinated, they perform acts of cooperation since these things are good not just for the person who does them, but also for the people around that person. There could be no clearer illustra- tion of the principle: the aim of cooperation is to produce mutual or win-win benefits; certain acts of self-interest also serve the interests of others. The purpose of healthcare 1 This text is based on the talk “Translator ethics: From cooperation to risk and trust” given online for Hong Kong Baptist University on 24 September 2020. The talk and additional responses to the subsequent questions can be seen at: http://hkbutube.lib.hkbu.edu.hk/st/ display.php?bibno=st969 and at: https://www.researchgate.net/publication/353333229_ Questions_on_ethics_with_a_few_answers. 6 Anthony Pym: Cooperation, risk, trust. A restatement of translator ethics translation in this context is consequently to foster such cooperation (going beyond health literacy as a general aim, which seeks to enable the individual to make optimal choices for their body). As for risks, all of our societies have been calculating, publicly or privately, the probabilities of negative consequences along several dimensions, mostly involving trade-offs between economic hardship and number of deaths. There could be few clearer examples of public policy as (good or bad) risk management. And then, in the world of a pandemic, trust is all: any concerted collective action requires public trust in specialized information, much of it translated, while public distrust is manifested in conspiracy theories and disruptive dissent, in some cases feeding off translations per- ceived as being defective and thus untrustworthy. Cooperation, risk, and trust are thus all very much at stake in pandemic communication and are closely interrelated. This is thus an appropriate time to think them through again. That said, I see the pandemic as a training exercise for an even greater public task. All three concepts also apply to the challenges of the climate emergency, which is even more clearly where communication has to connect with collective cooperative action. That is why cooperation, risk, and trust might be relevant now. So how did I get here? 2. What cooperation is better than One does not wake up one day and say: cooperation, what a great idea! In 1992 I pub- lished a book that was looking for an ethical principle but did not find one. Instead, I offered copious critique of the available ideas, but dissent is always the easy part. It is not hard to pick at faithfulness and equivalence as criteria for ethics. I was by no means alone in that: the 1980s and 1990s were the years of Skopos theory and what was becoming Descriptive Translation Studies, both of which pointed to the target side as the place where the game was to be played. Faithfulness and equivalence were looking backwards, the wrong way. For some, my subsequent disinterest in looking backwards has meant avoiding the proper subject of ethics. Meschonnic (2007, 13), extending his “police actions” (Sie- burth 2000, 323) based on the text to translate, concludes that “he [Pym] wanted an ethics, he only presents a social morality”, when apparently I should have seen that ethics is actually a poetics calqued on the rhythms of a source. Yes, it is good to listen very carefully to the cadences of the other, but is that really the whole show? With similar aplomb but rather less authority, Gao and Tian (2020, 327) regard faithfulness as the cornerstone of ethics and lament that “Pym avoids talking about it, which is not the correct way to deal with it”. So why not elaborate a poetics of faithfulness of some kind? Easy answer for those who look backwards: because of hermeneutics and deconstruction. I have long held 7Stridon. Journal of Studies in Translation and Interpreting, Volume 1 Issue 2, pp. 5–24 a view of sense as something that is construed in acts of reception. All texts have to be interpreted; none has imminent value, not even in the rhythms from which Mes- chonnic panted his poetics – if only because different cultures have different rhythms. If you take hermeneutics on board seriously, with whatever degree of deconstruction, there is nothing in a text that is solid enough to be faithful or equivalent to; we must work from the variable decisions actively made in the process of reception. So from that general position in the philosophy of language, I have long striven to exclude essentialist thinking about translation – and it has not been easy. That said, faithful- ness, especially in its historical avatar as equivalence, has by no means been excluded. Social or individual beliefs in the representative status of a text must be recognized as useful, operative, even necessary social fictions. They serve collective functions and we can analyse that. In fact, they were later to become the stuff of trust analysis (the repressed returns). But as a foundation for serious ethics, as an orientation for the way translators make decisions, they were never enough. For precisely the same reason, I have never tried to base ethical thought on the transla- tor’s or client’s supposedly unique purpose, on Skopos, at least not in the sense of blindly carrying out someone’s instructions. Why? Because purposes are just as much essen- tialist idealizations, transcendental signifieds if you will, as anything in the equivalence paradigm ever was. Purposes also have to be construed. And then, more obviously, an ethics of mercenary behaviour is never going to satisfy a thinking person. My 1992 book thus had some fairly powerful reasons for expressing discontent with the available professional codes of ethics, which made idealist assumptions about communication and then mostly said what translators should not do, with rarely an affirmative message about what they should do. I was searching for something more than inherited limits on action; I wanted to know how and why active communication decisions should be made. Similarly unappealing in those years was the idea that if we do what is expected of us, we are doing well, as seemed to be the argument in Chesterman (1993), as well as in some versions of norm theory and some usages of habitus. That would be a paroxysm of conservatism, philosophically justifying what the codes of ethics were stating: here is what we all agree on, so we must all agree on this. No, tradition cannot be reason. In the same vein, Nord’s principle of “loyalty” looked merely motherly: “Loyalty may oblige translators to reveal their translation purposes and justify their translational decisions” (Nord 2002, 37). So if you the translator are not doing what is expected, tell the reader about it – you can go out late at night, my mother used to say, “as long as I know where you are”. That is a very good general principle for all communication: own up to what you do and take responsibility for your decisions. But it is in no way specific to translation and it offers no guidelines at all as to how to decide. 8 Anthony Pym: Cooperation, risk, trust. A restatement of translator ethics In the early 1990s I was thus wandering across a small intellectual desert. I was look- ing for a principle that could guide a translator when choosing between alternative renditions. I sought something beyond essentialism, endemic expectations, and sin- cerity. Some eight years later, in a conference in Manchester (Pym 2000), I proposed what such a guideline might look like, in lapidary form: the goal of any translation pro- ject should be long-term cooperation between cultures. That seemed to offer something affirmative; it avoided essentialism, tradition, and assumed sincerity; it also avoided the huge binarisms that had been inherited from classical translation theory. My hope was that a translator in a particular situation, with a particular client, with a particular text and hopefully with some future text receivers, would be able to think about coop- eration as a way of relating all those ever-particular elements. And that thinking along these lines could inform decisions that would become actions in the world. You can describe norms, narratives, complexities, or language differences, but all you will ever find are that norms, narratives, or languages are different and complex. That knowledge can be useful to subvert official or monological views of the world. But it cannot offer a context-sensitive guide to action. Cooperation can. 3. What cooperation says How do translators decide between alternatives? One set of constraints belongs to the laws of the land, which we respect unless there is good reason to do otherwise: laws concerning privacy, defamation, fair pay, contracts, and so on. Within the laws of the land, other ideas concern professional conduct: respect for colleagues, timeliness, confidentiality, et cetera, all of which apply to any service profession whatsoever. And then, within the field of professional conduct, we might find a few principles that con- cern translation and interpreting as specific occupations (here I include both under the term “translation”): issues of copyright, where the name of the translator should appear, uses of translation technology, where dead labour becomes capital, and so on. Our various codes of ethics can be broken down into those levels: some principles adhere to the laws of the land, others reflect the norms of professional conduct, often leaving very few that intimately concern translation. Now, on a plane quite different from that analysis, we might place the translator as a person. That is, in addition to being a citizen, a service provider, and a translator, we have this person who can make good or bad decisions. That person might decide to act in order to help achieve universal equality, freedom, justice, diversity, inclu- sion, general respect for the other, and so on, which, stated as such, are principles so empty that few would try to disagree with them. Or that person might want to act in favour of specific oppressed minorities, maligned cultures, less-spoken languages, 9Stridon. Journal of Studies in Translation and Interpreting, Volume 1 Issue 2, pp. 5–24 non-violence, public health, climate action, and so on, acting in the interests of causes that are more specific to each historical moment and might thus be topics for debate. And still others will decide to spread the word of their god, the virtues of their nation, or the greatness of their author. There is no reason at all why individual translators should not subscribe to any or all such aims and seek to attain them. But they will do this on the basis of individual decisions, and then enact those decisions in all their actions, in all forms of communicating, not just when translating. That is, these causes are not in any way specific to translation and thus have no special reason to be in- cluded within any ethics that pretends to be specific to translation. They can certainly guide actions, they can justify decisions, but any discussion of them tends to be more on the level of universalist ethics, not with respect to translation as such. Cooperation is only deceptively like those general principles. Admittedly, if you use the word on that universalist level of discussion, it looks remarkably empty and ide- alistic. Cooperation can certainly be considered a good thing in and of itself; it can be seen as a particularly good thing in the face of a pandemic or climate change. And it is clearly a fact of innumerable types of communication, not just translation. The one difference with respect to most of the other good things is that cooperation becomes particularly crucial in the field of cross-cultural communication – that is, in a wide field within which we find translation. To put the argument in simple terms, cooper- ation is relatively easy when communication partners share the same language and culture; it becomes more difficult to achieve when different languages and cultures are involved. That is one reason why cooperation can be a privileged goal in translation, even though it is not rigorously specific to translation. Cooperation is also something that is very commonly misunderstood, and this has caused me a little frustration. Cooperation does not just mean being nice with each other, which is how some scholars seek to sideline the concept or write it off as simple naïveté. It does not involve any supposed neutrality; no one is called on to be an “hon- est broker”. On the contrary, in its neoclassical formulations, cooperation is a model of rational egoistic action. Cooperation is a technical term with a technical meaning. It is worth understanding that meaning before leaping to judgement – few things are more naïve than an ignorant accusation of naïveté. Here is the neo-classical model. In a cooperative interaction, all parties act in their own interests but do so in such a way that they all acquire more value than what they started with. Non-cooperation is a zero-sum game where if I win, you lose. In coop- eration, I win something and you win something, and that possibility gives us a very good reason to communicate. So in cooperative communication, you want to make sure not only that you win but also that the others do not lose. Note carefully, though: 10 Anthony Pym: Cooperation, risk, trust. A restatement of translator ethics the theory does not say that all parties make equal benefits, and it assumes that all parties can make decisions egoistically, in search of their own benefits. As long as all parties make gains on their initial positions, then the interaction can be considered ethically valid in terms of cooperation. And then, if there are non-egoistic or altruistic actions, there can certainly be more equitable distributions of benefits. Cooperation is not just a beautiful idea. It is something we do every day, with each purchase, each greeting, each morning glance of recognition of a partner, friend or companion: this is another day that is better spent together than apart. So where did that idea come from? References to cooperation are actually all around us. If you are doing pragmatics, it is in Grice’s “cooperative principle” (1975, 45): “Make your conversational contribution such as is required, at the stage at which it occurs, by the accepted purpose or direction of the talk exchange in which you are engaged.” I take this to be a definition of what a conversation is: it is very possible to be uncooperative, to mislead, insult, or abuse the other, but that would betray the initial assumptions of a conversation. Grice nevertheless seems not to tell us what conversations are good for; I would like to be able to suggest what they are good for: they enable cooperation. In translation studies, a notion of cooperation also figures prominently in Holz-Mänt- täri (1984), where the translator is seen as an expert in cross-cultural communication who cooperates with area experts in particular fields of activities. Again, this recog- nizes that cooperation is going on, but it does not say really why it should be there. Not until I encountered neoclassical economics did I become interested in concepts of cooperation that are more precise and more powerful in explanatory capacity. I had educated myself politically in a world where one side was good and the other side was bad, and many of my fellow scholars are still in that world. I nevertheless gradually began to see a real alternative to competition, right in the middle of ideol- ogies that I saw (and still see) as being mostly on the wrong side of history. The idea dates from Adam Smith’s passing note that when exporters pursue their own interests, the result may be beneficial for all ([1776] 2000, 4.2). That blossomed into an entire mathematics of win-win situations with multiple players – you might remember the Nash Equilibrium from the film A Beautiful Mind (Howard 2001). From the 1980s, that “mutual benefits” concept began to influence other disciplines. For me, one of the most important references was Robert Keohane’s After Hegemony (1984), which laid out the possibilities for international trade and diplomacy at the end of the Cold War. You then find cooperation in the extended game theoretics of Axelrod (e.g. 1997), who ran multiple-player prisoner’s dilemma games between computers and applied the model to the economic analysis of friendship, marriage, gang formation, trench 11Stridon. Journal of Studies in Translation and Interpreting, Volume 1 Issue 2, pp. 5–24 warfare, and much else. Since then, biology and sociology have developed a range of techniques for measuring the degrees of cooperation or competition that characterize particular societies. And sustainability is analysed as cooperation with future genera- tions, which is why I emphasize long-term cooperation here. Why was cooperation of interest in the 1980s and 1990s? In the United States, the apparent end of the Cold War opened up debates on ways of reorganizing interna- tional relations. In Europe, on the other hand, prolonged conflicts such as the painful disintegration of Yugoslavia showed the extent to which cultural differences could lead to competition over cooperation. There were good historical reasons for seeing the prime task of cross-cultural cooperation as being long-term cooperation between cultures. I pause to point out what cooperation is not: • No equality or symmetry: To restate: The people involved in cooperation can have very different starting positions and very different degrees of agency. As long as each party gains something, there is still cooperation. • No neutrality: Since mediators are active parties to cooperative interactions, they too seek gains and can be expected to act egoistically in that sense. There is no assumed neutrality. • No truth: Since communicative success is in the cooperative outcome of the com- munication act, there is no necessary assumption of an a priori truth. Truth can certainly enter the frame later, when viruses kill people, vaccines fail, and oceans rise, but those truths exist beyond the frame of human communication. • No full understanding: Since there is no assumption of an a priori truth, there is no basis for positing that the ethical aim is to have something ‘understood’. Instead, we might seek a series of ‘understandings’, in the sense of shared but transitory mutual expectations –what Davidson (1986) might call “passing theories”. • No clarity: Once you do not give priority to criteria of truth and understanding, there is little reason to subscribe to ideals of clear expression, as if there were a truth to which language can or should be transparent. The beauties and mysteries of difficult expression can also enter into calculations of cooperation. • No one-sided loyalty: As stated, the translator is here regarded as an active par- ticipant. If the translator systematically supports just one side to the systematic detriment of the other, mutual benefits are unlikely to result and the interaction could not be regarded as ethical. 12 Anthony Pym: Cooperation, risk, trust. A restatement of translator ethics These positions are not shared by many of the Enlightenment idealisms in Chester- man (2000), for example, and would seem puzzling when seen from the perspectives of most professional codes of ethics. Yet there is no adamant admonition here. If ap- peals to truth, understanding, clarity and loyalty can, in particular situations, enhance the probability of cooperation, then they should be considered positive values. But they are not ends in themselves. A final consequence of cooperation theory requires a little more explanation. Coop- eration can help address questions of how much effort should be put into mediated communication. If there are no great benefits to share, then it is not worth investing huge efforts in a translation, and vice versa. Indeed, by some calculations (Pym 1995), low-cost translations can allow a wider range of benefits, to be distributed across wid- er social groups. From this perspective, the use of online machine translation, with or without light post-editing, can be considered a potentially ethical mode of communi- cation. Now that less than one percent of the words translated in the world are done by professionals (see the calculation in Pym and Torres-Simón 2021), any ethics of translation has to be able to address questions of effort. But not every translation can be left to machines. And that, of course, raises the prob- lem of risks. 4. Risk If you can allow that cooperation describes a successful communication outcome (win-win), the step to risk analysis is easy. Once we have an idea of what success is, we can start calculating the probabilities of failure, which is minimally defined as non-cooperation. Mistakes still exist and they are still bad, but now we can say why they are bad: they can get in the way of cooperation. And now we can start to assess ways of dealing with the probability of that happening. The very rich literature on risk management gives us at least four ways of avoiding failure: • Risk aversion: Change your behaviour in order to lower the probability of a nega- tive outcome. • Risk transfer: Make someone else take on the risk. • Risk taking: Assume the risk in the hope of attaining high rewards. • Risk trade-offs: Take a minor risk to reduce a major risk. Let me explain each of these in turn. 13Stridon. Journal of Studies in Translation and Interpreting, Volume 1 Issue 2, pp. 5–24 There is an abundance of research that suggests translators are likely to be risk averse. All the “translation tendencies” announced by Levý ([1963] 2011) can be read in this sense: the language used in translations tends to be simpler, clearer, less rich, and less extreme than in non-translations. Keep it boring, keep it safe – translators tend not to take chances with language. Yet it is not all so clear-cut: if there is a high chance that your reader will not understand a reference, put in a piece of explicitation in order to help them. A Spanish text, for example, refers to “the last war”; the translator suspects the English-language reader will have doubts about which war is being referred to; the translator reduces that risk by explicitating the reference as “the Spanish Civil War”. This is risk aversion for as long as the translator is very sure that the explicitation is correct. So what would risk transfer be? Any action that moves the risk away from the translator would count as an instance of transfer. Most commonly, translators can check points of doubt with their project manager or client, as recommended by Gouadec (2007). In other situations they can refer to an authoritative glossary or draw on a client’s trans- lation memory. Even when they suspect there is a probability of error, they can always later say: It wasn’t me – I followed the material I was given! In many instances, simple literalism can work as risk transfer: Don’t blame me; I put what was in the text! In the case of the Spanish “last war”, for example, the translator may not be sure of which war is being referred to and therefore renders the reference literally as “the last war” in English. The risk of misinterpretation has thus been transferred both to the start text and to the readers, who are left to construe the reference for themselves. Risk taking is then when translators are very aware that their decisions may lead to non-cooperative outcomes but they decide to take their chances. To follow the same example, the translator may not be sure of which war is referred to but opts for “the Spanish Civil War” nevertheless. In order to justify taking a risk in this way, the trans- lator would have to envisage some major benefit being at stake somewhere down the line. If the reader does not have this particular information, for example, a whole series of similar historical references in the text might go off course. Trade-offs, finally, are when the translator takes a minor risk in order to mitigate a major risk. In Pym and Matsushita (2018) this strategy is actually called “risk miti- gation”, but “trade-off ” seems a clearer term. To continue with the same example, the translator may not want to take the absolute risk of specifying “the Spanish Civil War” and so will opt for literalism, but nevertheless take a minor risk by adding a footnote suggesting the nature of the reference. Most cases of trade-offs involve similar in- stances where two or more translation solutions are offered to solve the one problem. These risk management strategies give a fairly complex way of discussing translator decisions without entailing any dependence on essentialist notions of meaning or 14 Anthony Pym: Cooperation, risk, trust. A restatement of translator ethics reference. There are doubts at every turn, yet translators can make reasoned decisions in search of cooperation. The most worrying thing that ensues from these analyses, as might be predicted from studies on translation “tendencies” or “universals”, is that translators tend to be overwhelmingly risk averse. In our studies on COVID-19 com- munication in Melbourne in 2020 (Karidakis et al. forthcoming), we found that most official translations were extremely literalist, not wishing to take chances with poten- tially high-stakes information and effectively transferring risk to the authors of the start texts. In the various language communities, however, those translations were often not effective in changing behaviour: the risk transfer meant that the technical language was confusing. The official translations were thus reworked, simplified, dis- cussed, and put into multimedia formats by the many community associations, who adopted a far more diversified approach to risk management. So why were the official translators more reluctant to take chances than were the com- munity associations? It has to do with the nature of trust. 5. Trust Andrew Chesterman (2000, 182) states that translators “must be trusted by all parties involved, both as a profession and individually. […] Without this trust, the profes- sion would collapse, and so would its practice.” Why should trust be so important to translators? The most obvious reason is that the translator is representing a prior text and the person they are communicating with typically has no way of testing the linguistic validity of the representation – the reader of a translation normally does not understand the foreign language and is thus condemned to trust its representational validity, mostly on the basis of who the translator is and how much the translation corresponds to what is expected. Without that trust in the representation, the trans- lator cannot hope to contribute to cooperative interactions. In terms of risk manage- ment, we might therefore say that the greatest risk the translator faces is that of losing trust, or what might be termed “credibility risk” (Pym 2015). Beyond that simple logic, the concept of trust plays a key role in making translator ethics speak in terms that can have at least some psychological verisimilitude. Once we have dispensed with faithfulness and equivalence as criteria on which to judge a translation, we are nevertheless able to recuperate the presumption of those values downstream, from the perspective of the client or user of the translation. Trust here is initiated (or taken away) not by the translator – this is not the initial trust that Steiner (1975, 312) saw in the translator’s relation to the text to be translated – but by the users of translations. To talk about trust orients the translator’s view to the actions of people of the future, not back to the text in the past. 15Stridon. Journal of Studies in Translation and Interpreting, Volume 1 Issue 2, pp. 5–24 This kind of trust is to be distinguished from simple familiarity. True, we tend to trust the people we think we know best, and we may consider a person to be trustworthy on the basis of the repeated actions they have carried out in the past. But when we accept a translation as a substitute of a text to which we do not otherwise have access, the act of trusting is necessarily accompanied by complex factors that are beyond our control. We decide to trust a translator because it is a way of reducing that complexity (Luhmann 1968); trust, in this sense, is “a solution for specific problems of risk” (Lu- hmann 1988, 95); it can always flip into distrust, even in situations of great familiarity. The workings of trust are very clear in pandemic communication, where an ideal chain would see science being trusted by governments, who are trusted by profession- al writers of media communications, who are trusted by translators, who are trusted by users of translations, who adopt cooperative behaviour accordingly. At no point in this ideal chain can one party be assumed to fully ‘understand’ the previous link: this is not a model of relayed truth. However, when trust works, the end users may believe they are trusting science directly. Of course, trust tends not to be so linear but branch- es out in networks (since we tend to trust those who are trusted by people around us)2 and any link in the resulting network can be broken and active distrust may result: governments seek trade-offs between medical experts and the calculations of econo- mists; many people do not trust their governments in principle; professional writers address only the highly educated; translators follow suit; end users do not believe the translations; narratives of dissent give structure to instances of distrust. In practice, of course, many other factors can influence the workings of trust and dis- trust. In our study of pandemic communication in Melbourne (Karidakis et al. forth- coming), a pressured workflow meant that in two cases official translations actually mixed languages (Arabic and Farsi in one case, Indonesian and Turkish in the other). These became memes that spread across all media, leading to widespread reports that the official translations were not trustworthy. No matter how much I tried to argue (in Pym 2020) that the mistakes were not due to bad translators (they came from bad pro- ject management) and that the Australian translator certification system was actually one of the best in the world, distrust abounded. As a result, the Victorian government invested considerable additional funds into multilingual communication, with only a fraction of it earmarked for official translations. The revised government recommen- dations included an instruction that organizations should use not only certified trans- lators but also “a trusted, credible source to promote your message” (Victorian Gov- ernment 2020), for example “a health practitioner” or “a local elder as a messenger” 2 My thanks for the observation made by Piotr Blumczynski in the discussion following the talk on which this paper is based. 16 Anthony Pym: Cooperation, risk, trust. A restatement of translator ethics when reaching out to a specific language community. In terms of trust, this means mixing “thin trust” (we trust a translator because they have professional certification) with “thick trust” (we trust a local mediator because we know a lot about them) (see Hosking 2014, 46–49). A focus on trust thus invites us to move well beyond praise of any closed profession as the only way to achieve ethical communication. At the most general level, trust is involved in all acts of cooperation, if only because each party must trust the other in order for benefits to ensue over time. But trust is particularly important in cross-cultural communication, where familiarity levels are lower and the need for thin, risk-based trust is consequently higher. All the edu- cational qualifications and certification systems for translators address precisely this issue. The entire institutionalization of translation works to this end. As recognized by perhaps the foremost dismantler of scientific communication in our age, “facts remain robust only when they are supported by a common culture, by institutions that can be trusted, by a more or less decent public life, by more or less reliable media” (Latour 2018, 23). Translation must be placed within that wider view of social com- munication, as part of a kind of trust can extend beyond cultural borders. 6. Addressing problems An ethics based on cooperation, risk, and trust would be of little interest if it were unable to address some of the knottier problems ensuing from practice. This does not mean solving problems, as in mathematics. It does not mean calling the shots between ethical and non-ethical, as in a line decision in tennis. The more modest aim must be to provide considerations that might help translators decide for themselves how to choose between the very particular alternatives they confront. I select a few problems from recent debate and research. 6.1 Are translators in the sole service of their country? A growing strand of ethics in China takes issue with the very principle of differ- ence-based cooperation, which seems to contradict a “National Translation Program”. Ren and Gao (2015a) posit that the purpose of a national translation program is to further the one-sided interests of the state, and they further explain that “在国家翻 译实践的内部合作中，合作各方的利益是一致的，都是国家的利益，因此不 存在协调 各方利益的问题” (2015b, 108), which we might translate as follows: “In the internal cooperation of the national translation practice, the interests of the coop- erating parties are the same, since they are the interests of the country, and therefore there is no question of reconciling the interests of all parties.” So Pym’s win-win ethics 17Stridon. Journal of Studies in Translation and Interpreting, Volume 1 Issue 2, pp. 5–24 of cooperation is explicitly rejected (still, it’s nice to be noticed). This is quite logical: if all participants in the translation act are in the one country and agree on everything a priori, there is nothing to be negotiated and thus no basis for seeking a “win-win” out- come. The ethics of mutual benefits does not apply; there is no risk of communicative failure; there is no primal doubt or dialectics; trust is absolute. That is certainly an ideal for the national production of translations, on a level that is basically no different from the national production of tractors. But what about outgoing translation as a mode of communication across cultures? One is not surprised to see “cooperation” working as a key term in Xi Jinping’s thought, with “win-win coopera- tion” repeatedly used since 2014 as an ideological cornerstone of China’s foreign policy. One need only look at the titles of a few speeches: “Asia-Pacific Partnership of Mutual Trust, Inclusiveness, Cooperation and Win-Win Progress” or “Build a Win-Win, Equi- table and Balanced Governance Mechanism on Climate Change” (both in Xi 2017). So, as much as one would like to agree that all Chinese translation agents always agree on everything within their own country, that does not discount the search for win-win co- operation as an ethical purpose for translations between China and the rest of the world. 6.2 Should the interpreter reveal what Trump and Putin said? Our second problem is deceptively similar. In Helsinki in 2018, Donald Trump had a private meeting with Vladimir Putin. Present was the State Department interpret- er Marina Gross, who took notes. After the meeting, a US congressional committee called on her to tell them what was in her notes. So should she tell them? On the face of it, the question is easily answered by the standard codes of ethics, where “confidentiality” is a stock principle. The International Association of Conference Inter- preters (Field 2018) swiftly issued a statement saying that the interpreter should not tes- tify, and the American Translators Association was reported as taking the same position (Segal 2018). So how might we respond to this in terms of cooperation, risk and trust? The question to ask here is whether cross-cultural cooperation would be served by re- vealing what was said in that private meeting. Probably not. And a good argument can be made for the practice of private meetings as trust-building exercises between heads of state, exploiting the virtues of face-to-face spoken conversation. So we might agree with the principle of confidentiality in this case, but for a reason that goes beyond the fact that it just happens to be in the established codes of ethics. A more engaging question is whether Marina Gross was wholly a professional inter- preter in this case, since she was also an employee of the State Department and thus belonged to the same institutional system that was asking her to testify – this is basically 18 Anthony Pym: Cooperation, risk, trust. A restatement of translator ethics the identity problem of Sperthias and Bulis in Pym (2012). As role-identity analysis might tell us (Forde 2021), Gross could have assumed that one identity prevailed over the other and responded accordingly. In terms of translator ethics, she would be decid- ing in which network she was likely to achieve more long-term cooperation and trust. 6.3 Are interpreters of torture themselves torturers? Takeda (2021) examines the British trial transcripts of 39 interpreters who worked for the Japanese in the Second World War. The interpreters were civilians charged with “being concerned in” the ill-treatment of prisoners of war and local citizens in Japa- nese-occupied territories. So should the interpreters’ work be considered unethical? From the perspective of translator ethics, the first observation to make here is that the act of torture is very difficult to analyse in terms of mutual cooperation: the mediation is not likely to bring any particular benefit to the subaltern party. So we cannot offer justification on that count. Another consideration, however, is whether the interpret- ers were able to refuse to mediate. If they were able to refuse the task and proceeded nevertheless, then they are indeed liable for the consequences, both legal and ethical, of their non-cooperative actions (Pym 2012, 166). On this point, we once again reach a position that is in agreement with the standard codes of ethics (RedT, AIIC, and FIT in this case) when they insist that translators and interpreters should be able to refuse an assignment. An ethics based on cooperation can nevertheless point to something that unethical interpreters could be specifically responsible for: a radical mode of non-cooperative interaction. That said, one hesitates to condemn any mediator simply because they happened not to foresee which side was going to win. The same principle we apply to the interpreters working for the Japanese should also be applied to those working for the American, British, or Australian forces in any similar situations. By the same logic, an ethics based on cooperation cannot condemn the diplomat and interpreter Eugen Doll- mann, for example, because he facilitated exchanges between Hitler and Mussolini. Other kinds of ethics are more than capable of dealing with that problem. 6.4 Should translators work for free for profit-making companies? Zwischenberger (2021) looks in labour-value Marxist terms at the translators who work for free on Facebook sites. She correctly describes the company’s use of their work as “exploitation”. Along the way, though, she accepts that each individual trans- lator may rationally decide that their labour is more than compensated for by what they gain from the activity in terms of experience, social interaction, or the fact that 19Stridon. Journal of Studies in Translation and Interpreting, Volume 1 Issue 2, pp. 5–24 they are providing support for a particular language. Zwischenberger recognizes that the exploitation can indeed be “mutually beneficial” within the frame of that inter- action, but should nevertheless be considered reprehensible on a more global level. There seem to be two reasons for this. First, in terms of Marxist analysis (actually from Engels), the translators suffer from “false consciousness”, since if they knew the enor- mous profits being made from each language Facebook opens up into, they would not give their labour for free. And second, drawing on consequentialist ethics from Wertheimer (1999), cooperative exploitation “may be individually rational but col- lectively self-defeating” when a third party suffers as a consequence. Zwischenberger posits that in this case the third party would be professional translators, the market value of whose work is undermined by the labour given for free. So can a translator ethics support the condemnation on either of these counts? Once we abandon essentialist truth, unfortunately there is no certitude from which to distinguish between true and false consciousness – and I am not sure the world would be better if our activist academics’ presumably ‘true consciousness’ were universally in charge. Rather than assume we are right and the translators are wrong, we might want to act empirically (hence the move into consequentialist ethics). And then, if the translator is wholly aware of Facebook’s profits and translates nevertheless, what side is truth on then? Personally, I have interacted with the Facebook crowdsourcing site and I have contributed voluntarily and very knowingly to Google’s language assets, in both cases in the interests of developing electronic resources for Catalan, a language that I like very much and that needs support. Contrary to those who are outraged by the very mention of commerce, I see no reason to consider a company unethical sim- ply because of its profitability. The consequentialist argument is more interesting. To make it stick, though, Zwis- chenberger would have to identify not just how professional translators are unfairly affected by Facebook getting some translations for free (there are indeed some lan- guages where the company would otherwise pay for professional services), but also how that result has negative consequences for cross-cultural cooperation. Neither ar- gument is easy to make. In fact, the only evidence Zwischenberger presents on this score is that the translator associations seem peculiarly unperturbed by the threat posed by volunteer translators – perhaps because the associations need enlighten- ment from true consciousness, but perhaps also because the threat is not significant. So what evidence is there? One cannot assume that translations done for free are in any way inferior (that risk is taken care of by employing professional checkers any- way); one cannot say that they are not trusted (since they come from the community of users themselves); one cannot point to marked revenue loss among professionals: 20 Anthony Pym: Cooperation, risk, trust. A restatement of translator ethics superficial evidence suggests that the global market for translations is growing, not retracting, even despite the many instances of deprofessionalization (Pym and Torres 2021). And then, even if you do locate some way that less work for professional trans- lators diminishes cross-cultural cooperation, that result would have to be compared with the trade-off benefits of extending the range of less-spoken languages that are used in electronic space. The world has a long tail of smaller languages for which vol- unteer work or government subsidies are needed if profit-making companies are go- ing to operate in them (Catalan is on the edge of that space). But that is another story. An ethics should be able at least to address that kind of trade-off, prior to condemning out of hand everything that looks like unequal exchange. 7. A conclusion: How far should one look? It seems unreasonable to ask translators to save the world, as if they were prime rev- olutionary subjects. And yet it is quite reasonable to suggest that, confronted by al- ternatives between which an ethical decision is to be made, translators should at least look beyond the text in front of them. This means reflecting on the upstream prove- nance (How did this text get here? Why was effort invested in its presence? Can the text be improved?) and downstream effects (Who is seeking to cooperate with whom? In search of what potential benefits? With which lasting effects?). The work of ethical discourse should be to extend reflections in both those directions, to make translators think within wider frames, and hopefully, as a result, to give them the courage to take risks in search of rewards. As for the community of translation scholars, there can be little doubt that they are increasingly looking well beyond relations between texts. Ethics is these days a rich and exciting field of debate, as I hope the above few examples illustrate (and indeed as is made very clear in Pokorn and Koskinen 2021). At its best, our exchanges feed into empirical studies of the ways translators actually resolve problems, and the various rea- sons they give. At its worst, though, discussions of ethics slip into universalist certitudes about issues well removed from mediation between cultures, where good and bad are decided before any consideration of translation itself. One can look too far, too fast. I hope that the above concepts of cooperation, risk and trust can help provide some shared frames for continuing debate. 21Stridon. Journal of Studies in Translation and Interpreting, Volume 1 Issue 2, pp. 5–24 References Axelrod, Robert. 1997. The Complexity of Cooperation: Agent-Based Models of Competi- tion and Collaboration. Princeton NJ: Princeton University Press. Chesterman, Andrew. 1993. “From is to ought: Laws, Norms and Strategies in Translation Studies.” Target 5 (1): 1–12. Chesterman, Andrew. 2000. Memes of Translation. The Spread of Ideas in Translation The- ory. Amsterdam and Philadelphia: John Benjamins Publishing. Davidson, David. 1986. “A Nice Derangement of Epitaphs.” In Truth and Interpretation: Perspectives on the Philosophy of Donald Davidson, edited by Ernest LePore, 433–46. Oxford & New York: Blackwell. Field, Martin. 2018. “Confidentiality in the spotlight.” Communicate! 73, October 24, 2018. https://tinyurl.com/m52c49pd. Forde, Shane. 2021. “Role Identity Theory and Ethical Decision Making in Translation and Interpreting: The Case of Coordinators for International Relations.” Paper pre- sented at the conference Ethics and Professionalism in Translation and Interpreting, Ewha Womens University, Korea, July 15, 2021. Gao, Ning and Tian Chuanmao. 2020. “A Review of Anthony Pym’s Translator Eth- ics.” Journal of Humanities and Education Development 2 (5): 324–29. Gouadec, Daniel. 2007. Translation as a Profession. Amsterdam and Philadelphia: John Benjamins Publishing. Grice, H. Paul (1975) “Logic and Conversation.” In Syntax and Semantics 3: Speech Acts, edited by Peter Cole and Jerry L. Morgan, 41–58. New York: Academic Press. Holz-Mänttäri, Justa. 1984. Translatorisches Handeln. Theorie und Methode. Helsinki: Ac- ademia Scientiarum Fennica. Hosking, Geoffrey. 2014. Trust: A History. Oxford: Oxford University Press. Howard, John, director. 2001. A Beautiful Mind. Universal / Dreamworks. Karidakis, Maria, Robyn Woodward-Kron, Riccardo Amorati, Bei Hu, Anthony Pym, and John Hajek (forthcoming) “Enhancing COVID-19 public health communication for culturally and linguistically diverse communities: An Australian interview study with community representatives.” Qualitative Heath Communication. Keohane, Robert. 1984. After Hegemony: Cooperation and Discord in the World Political Economy. Princeton NJ: Princeton University Press. Latour, Bruno. 2018. Down to earth. Politics in the new climactic regime, translated by Catherine Porter. Cambridge UK and Medford MA: Polity Press. Levý, Jiří. (1963) 2011. Umění překladu. Translated by Patrick John Corness as The Art of Translation. Amsterdam and Philadelphia: John Benjamins Publishing. Luhmann, Niklas. 1968. Vertrauen. Ein Mechanismus der Reduktion sozialer Komplexität. Stuttgart: Ferdinand Enke. 22 Anthony Pym: Cooperation, risk, trust. A restatement of translator ethics Luhmann, Niklas. 1988. “Familiarity, Confidence, Trust: Problems and Alternatives.” In Trust Making and Breaking. Cooperative Relations, edited by Diego Gambetta, 94–108. New York: Blackwell. Meschonnic, Henri. 2007. Éthique et politique du traduire. Paris: Verdier. Nord, Christiane. 2002. “Manipulation and loyalty in functional translation.” Current Writing: Text and Reception in Southern Africa 14 (2): 32–44. https://doi.org/10.1080/ 1013929X.2002.9678123. Pokorn, Nike, and Kaisa Koskinen, eds. 2021. The Routledge Handbook of Translation and Ethics. Abingdon: Routledge. Pym, Anthony. 1992. Translation and Text Transfer. Frankfurt: Peter Lang. Pym, Anthony. 1995. “Translation as a Transaction Cost.” Meta 40 (4): 594–605. Pym, Anthony. 2000. “On Cooperation.” In Intercultural Faultlines: Research Models in Translation Studies I: Textual and Cognitive Aspects, edited by Maeve Olohan, 181–92. Manchester: St Jerome. Pym, Anthony. 2012. On translator ethics. Amsterdam and Philadelphia: John Benjamins Publishing. Pym, Anthony. 2015. “Translating as risk management.” Journal of Pragmatics 85: 67–80. https://doi.org/10.1016/j.pragma.2015.06.010. Pym, Anthony, and Kayo Matsushita. 2018. “Risk mitigation in translator decisions.” Across Languages and Cultures 19 (1): 1–18. Pym, Anthony. 2020. “When trust matters more than translation.” Pursuit, July 29, 2020. https://pursuit.unimelb.edu.au/articles/when-trust-matters-more-than-translation. Pym, Anthony, and Ester Torres-Simón. 2021. “Is automation changing the translation profession?” International Journal of the Sociology of Language 270: 39–57. https://doi. org/10.1515/ijsl-2020-0015. Ren, Dongsheng, and Gao Yuxia. 2015a. “A Preliminary Study on National Translation Practice.” Chinese Foreign Languages 3: 79–85. Ren, Dongsheng, and Gao Yuxia. 2015b. “National Translation Program: A New Con- cept.” Foreign Languages in China 3: 92–7. Segal, Corinne. 2018. “Lawmakers Want Trump’s Interpreter to Testify. That Might be Unethical.” PBS Newshour. Accessed July 19, 2021, https://www.pbs.org/newshour/ politics/putin-trump-russia-interpreter-marina-gross. Sieburth, Richard. 2000. Review of Antoine Berman, Pour une critique des traductions: John Donne. The Translator 6 (2): 321–27. Smith, Adam. (1776) 2000. The Wealth of Nations. New York: Random House. Steiner, George. 1975. After Babel. Aspects of Language and Translation. Oxford and New York: Oxford University Press. Takeda, Kayoko. 2021. Interpreters and war crimes. Abingdon and New York: Routledge. 23Stridon. Journal of Studies in Translation and Interpreting, Volume 1 Issue 2, pp. 5–24 Victorian Government. 2020. Best practice guide for translating content. Accessed July 20, 2021, https://www.vic.gov.au/interpreters-and-translation-guides Wertheimer, Alan. 1999. Exploitation. Princeton University Press. Xi Jinping. 2017. The Governance of China. Beijing: Foreign Language Press. Zwischenberger, Cornelia. 2021. “Online collaborative translation: its ethical, social, and conceptual conditions and consequences.” Perspectives https://doi.org/10.1080/09076 76X.2021.1872662. About the author Anthony Pym has been a professional translator and translator trainer in Spain, the United States and Australia for more than 20 years. He currently teaches at the Univer- sity of Melbourne and is Distinguished Professor at the Rovira i Virgili University in Tarragona in Spain. He was President of the European Society for Translation Studies from 2010 to 2016. He has authored, co-authored or edited 28 books and more than 200 articles in the general field of translation and intercultural communication. He holds a PhD in sociology from the École des Hautes Études en Sciences Sociales, Paris. 24 Anthony Pym: Cooperation, risk, trust. A restatement of translator ethics