Strategies and tactics in academic knowledge production by multilingual scholars

In the past decade, academic evaluation systems worldwide have markedly increased the use of mechanisms that privilege the use of English in journal publishing. In the context of these trends, this article highlights our findings from more than 12 years of research on the experiences and perspectives of 50 multilingual European scholars with writing for publication, particularly in English. We draw on de Certeau’s (1984) notions of strategies and tactics to explore key ways in which scholars manage often-competing demands and interests in writing for publication. Scholars both adopt strategies that align with official publication policies and use tactics that support scholars’ sometimes competing agendas. At different moments scholars embrace, accommodate, or resist the perceived dominance of English in knowledge production regimes and evaluation systems. We epaa aape Education Policy Analysis Archives Vol. 22 No. 32 2 conclude by summarizing the value of drawing on the notions of strategy and tactics in an era of increasing debates over evaluation systems.


Introduction
In the past decade, pressure has dramatically increased around the world for scholars to publish in high-status 'international' 1 journals-specifically, those listed in the citation indexes produced by the Institute for Scientific Information/Web of Knowledge (ISI). This pressure generates particular complexities and challenges for multilingual scholars, that is, scholars who regularly work, research, and publish in two or more languages. This pressure is articulated in policy documents that are formulated at the institutional, governmental, and supranational levels (e.g., Englander & Uzuner-Smith, 2013;Lee & Lee, 2013). Because these high-status 'international' journals are predominantly produced in Anglophone-center 2 countries and are mainly published in the linguistic medium of English, in many academic disciplines English has become naturalized, that is, unquestioningly accepted as the primary language of publishing. The growing policy push for publications in ISI-indexed journals is one result of the heightened competition that is taking place in higher education in the context of globalization (Lillis & Curry, 2013). In this context, many governments are aiming to increase their country's global share of research output as a marker of academic quality. Institutions of higher education around the world are therefore rewarding individual scholars and departments for publications appearing in journals ranked highly in ISI indexes and with high impact factors 3 -both counted in global measures of research productivity and both heavily implicating English (Feng, Beckett, & Huang, 2013;Lee & Lee, 2013;Lillis & Curry, 2013). We have contended, however, that these governmental and institutional policies are being created with little understanding of the consequences of shifting policy priorities on individual scholars, their local contexts, or knowledge production more broadly (Curry & Lillis, 2013b).
Academic evaluation regimes have resulted in two key tensions related to where and how scholars publish their work. First, as evaluation systems tend to value most highly publications that appear in ISI-indexed journals, competition to be published in the limited number of these journals increases even as the range of rewarded publications narrows. The effect is a potential disincentive for scholars to contribute to other types of journals and publications, including local journals (Salager-Meyer, 2008). Second, as scholars publish in 'international' academic journals (in English), concerns have arisen that local and regional contexts may be deprived of the relevant and important knowledge (Flowerdew & Li, 2009;Gibbs, 1995;Swales, 1997) that is central to the building of knowledge-based economies (Lillis & Curry, 2013). Responding to pressures for high status publications, therefore, multilingual scholars may choose English as the language of publication for their work although publications in English may not be accessible to members of their local or other communities for linguistic or material reasons (that is, such journals may not be available through local university libraries). In this way, knowledge published exclusively in 'international' Englishmedium journals tends to circulate away from the communities where it is produced (Gibbs, 1995) and where individuals and institutions often cannot afford to buy the journals and books that publish it (Canagarajah, 1996;Salager-Meyer, 1997). Pressure for high-status publications may also encourage scholars to use English to publish research findings in 'international' journals and use the local language to publish applied findings, a phenomenon Swales (1997) casts as publishing "their best in the west, offering more minor works for local publication" (p. 378; but see also discussion challenging this view in . Against this backdrop of the intensification of the imperative for high-status publications in English, in this article we draw on de Certeau's (1984) notions of strategies and tactics to explicate both officially legitimized (hereafter, official) research objectives-and the related strategies for achieving them-as well as the tactics scholars use to achieve their own objectives for distributing their workobjectives that sometimes come into conflict with official objectives, as we explain below. We next discuss our social practice theoretical framework, which we contend can usefully include de Certeau's metaphors of strategies and tactics. We then describe our longitudinal ethnographic study and summarize its key findings on multilingual scholars' publishing practices and perspectives. We illustrate the value of drawing on the lens of strategies and tactics to explore how scholars engage with competing demands and interests at different scale levels (Blommaert, Collins & Slembrouck, 2005, discussed more fully below) pointing to the use of strategies at the transnational (English-medium, global evaluation systems), national and institutional scales. We compare institutional policy articulations of objectives and attendant strategies with scholars' uses of both these same strategies (when their objectives align with official objectives) and the tactics they take up when their objectives conflict with official objectives. We close by discussing implications of our findings for understanding the writing and publishing practices of multilingual scholars and for supporting them-and contesting these trends in global academic knowledge production.

Social Practice Perspectives and Ideologies of English
Our research is grounded in social practice perspectives that see human activity as comprising patterns of interactions that occur within specific social spaces and relations of power (Bourdieu, 1990;Street, 1984;Wenger, 1998). Social practices, including those involved in writing for publication, are influenced by the evolving expectations of particular social contexts. They also inevitably entail the power dynamics of social institutions-in this case, the institutional and governmental apparatus that support and regulate knowledge production. Viewing academic writing for publication as a social practice shifts the predominant research focus on academic writers as isolated individuals to a view of academic writers as scholars working within social contexts and contending with the power relations of these contexts. As we discuss below, beyond what are conventionally considered as the practices of academic writing (drawing on the research literature; making notes, outlines, drafts, revisions, etc.), writing for publication also involves consequential social practices such as attending academic conferences and participating in research networks , 2013aLillis & Curry, 2006b, 2010, making decisions related to submitting papers to specific journals, and negotiating gatekeepers' feedback on papers submitted (Flowerdew, 2001;. Ideologies of language are centrally bound up in the global practices of academic writing and publishing. English is heavily implicated in academic evaluation practices, in particular because of the ways it is nested in the selection criteria for journals to be included in ISI indexes . The social practices of global academic journal publishing are currently grounded in powerful ideologies of English, particularly the dominant ideology of English as the presumed contemporary academic lingua franca (Lillis & Curry, 2013). This view is represented by comments such as these from a blog post at the American Institutes for Research: The ubiquity of English, like Latin before it, makes the spread of ideas -and the mobility of individuals -vastly easier. There's a strong case for viewing a linguistic common currency as empowering, whether to native speakers or to non-Englishspeaking teachers or students who want to participate in the global academic marketplace. (Wildavsky, 2013) This ideology that sees English as a purely functional and politically neutral academic lingua franca (EALF), however, does not account for a number of key issues related to access to "the global academic marketplace": 1) the historical legacy of English dominance as a result of colonization and geo-political power (Phillipson, 1992); 2) that while considerable academic publishing around the world takes place in other languages (as documented in our study and on www.Ulrichsweb.com), the presumption of EALF and the predominant monolingualism of Anglophone-center scholars contribute to making work published in other languages almost invisible in the Anglophone contexts where powerful evaluation mechanisms such as ISI indexes originate and are regulated; 3) that English, particularly the academic English used in publishing, is not equally accessible to everyone around the world (Swales, 1997); and 4) that despite limited tolerance for linguistic variation in spoken English in academic contexts, standard varieties of dominant Englishes (e.g., American and British) are used as both tacit and explicit criteria for judging the language of scholars' manuscripts. Indeed, markers of "non-native" uses of English often affect how journal gatekeepers evaluate multilingual scholars' submissions (Flowerdew, 2001;Hewings, 2006;, providing evidence for the critique that English does not, in fact, function as a lingua franca across global contexts. Multilingual scholars therefore enter the academic publishing "game" on a "field" (Bourdieu, 1990) that is not level, both in terms of access to English as well as to other social and material resources (research and travel funding, library access) needed for publication (Canagarajah, 1996;Mweru, 2010).
Bourdieu's metaphor of fields as the sites of competitive struggles connects with de Certeau's (1984) notion of strategies and tactics as a metaphor of military conflict and one that usefully signals the tensions and conflicts taking place in various fields of competition-if not of actual combat. de Certeau uses these metaphors as tools with which to analyze the struggles of the oppressed against the subjugated. He defines a strategy as: the calculation (or manipulation) of power relationships that becomes possible as soon as a subject with will and power (a business, an army, a city, a scientific institution) can be isolated. It postulates a place that can be delimited as its own and serve as the basis from which relations with an exteriority composed of targets or threats (customers or competitors, enemies, the country surrounding the city, objectives and objects of research, etc.) can be managed. (1984, pp. 35-36, emphasis original) Scientific institutions are one example de Certeau gives of the competitive "subject" that sets external "targets" for its activities. This subject is "isolated," or identified, as distinct from other subjects (army, business) that entail their own power relationships. For academic knowledge production, targets are thus the "objectives and objects of research" established by scientific institutions such as universities and, according to geopolitical context, by governments or supranational organizations (e.g., the European Union). Therefore, in the context of analyzing academic institutions and the governmental apparatus that regulates them, the notion of strategy suggests the need to identify how institutions not only set goals (e.g., targets for grants obtained, publications, numbers of students, etc.) but also how they attempt to achieve these goals. Below we provide examples of such strategies from our research data.
In contrast to a strategy, de Certeau posits that a tactic is a calculated action determined by the absence of a proper locus. . . . The space of a tactic is the space of the other. Thus it must play on and with a terrain imposed on it and organized by the law of a foreign power. It does not have the means to keep to itself, at a distance, in a position of withdrawal, foresight, and self-collection. . . . It operates in isolated actions, blow by blow. (pp. 36-37) Here de Certeau sees tactics as contingent, subversive moves wielded by the subordinate in combat with the official strategies of the dominant (scientific institutions, businesses, the army, etc.). In taking up this framing, which implies a view of scholars as "the other" in contradistinction to the scientific institution, it is important to recognize that around the world, to varying extents, many academics occupy a type of liminal space as both employees and management, signaled, for example, by perennial struggles over academic freedom, rights to unionize, and the casualization of academic labor. 4 Thus while scholars in the aggregate comprise the 'scientific institution', policy decisions related to setting goals and their attendant strategies for accomplishing research output goals are rarely made democratically.
Indeed, in the case of academic publishing, government and institutional policies create the official spaces within which scholars carry out their/our work, spaces increasingly influenced by evaluation systems generated at the global scale through the activities of multinational corporations such as Thomson Reuters/ISI. Within these official spaces, evaluation policies not only articulate institutions' specific research objectives, but in many cases also identify the institution's official strategies for achieving these objectives, as we illustrate below. By contrast, the "space of the other" is created when scholars choose to work toward their/our own interests and commitments for knowledge production and distribution-even when these interests may work against official policy objectives and their attendant strategies. 5 In these cases, individual scholars may realize these tactics in isolation ("blow by blow") or in concert with others. Broadly speaking, therefore, we understand strategies to be actions that support official, dominant institutional objectives for academic knowledge production, while tactics are actions aimed at achieving scholars' objectives that may not align with official objectives. It is important to underscore here that scholars use both strategies and tactics, because for example, institutional objectives quite often align with scholars' individual interests in having their work reach readers of high-status journals included in ISI indexes as well as other communities reachable through other types of publications. We return to this point below.
Moreover, in drawing on the theoretical metaphor of strategies and tactics, we do not mean to suggest that multilingual scholars lack power or agency in making decisions about publishing, or that power is located in one place. Rather, power is clearly operating at different scalar levels (Blommaert et al., 2005; see discussion in  and, in fact, we would emphasize that scholars actively consider their interests and commitments, the resources available to achieve them, and the consequences of their choices. Critically reflecting on our research findings, we see the notions of strategies and tactics as useful for describing and exploring the interplay between factors at a number of scale levels which impinge-in different ways and at different moments-on scholars' practices within the larger economy of global knowledge production, choices that scholars make contingently in accordance with, or in conflict with, official research objectives and strategies. Indeed, all scholars around the world daily confront the need to make choices about our work and allocating our resources-even as resources are unevenly distributed. In the rest of the article we explore the ways in which these choices are different, and often more complex, for multilingual scholars working outside contexts where English is commonly used, as compared with the choices 4 While clearly a much larger conversation than is possible to have here, this point is particularly salient for the ability of scholars to distribute the products of their/our labor according to their/our own interests rather than those of their/our employer or government, in ways that may be shaped by, for example, the availability of funding to pursue research that accords with particular interests of funding sources. 5 An example familiar to U.S. readers fortunate enough to be in tenure-track positions are the decisions that scholars make about the types of work to publish before tenure review and those they may reserve for after tenure, whether this work might be more political in nature or more oriented to practice rather than research audiences. Similar choices exist in contexts where scholars are being evaluated by national assessments that value research journal publications almost exclusively over other types of publications. faced by those of us who benefit, within the current global regime that values the use of English in higher education, by working in English-dominant contexts.

Research Methodology
When we began our study, Professional Academic Writing in a Global Context (PAW), in 2001, a small body of research had begun investigating writing for publication by multilingual scholars in the field of English for Academic Purposes/second language writing. 6 In the late 1980s, researchers began to document multilingual scholars' experiences and challenges with publishing (e.g., Baldauf & Jernudd, 1987;St. John, 1987). Because approximating the linguistic and rhetorical norms of Anglophone-center journals was generally seen as the key to achieving publication for multilingual scholars, the focus of much of this research was whether and how scholars and advanced graduate students mastered such norms. At the same time, however, Swales (1987Swales ( , 1997 raised early concerns about the effects of English on global knowledge production. In the 1990s research began to take into greater account the social context of multilingual scholars' writing, for instance, by considering the role of "non-discursive" resources in providing access to publishing (Canagarajah, 1996) and exploring scholars' interactions with journal editors and reviewers during the publishing trajectory (Flowerdew, 1999(Flowerdew, , 2000. Canagarajah (2002) traced the "coping strategies" used by multilingual scholars whom Swales (1987) calls "off-networked" (p. 47) to manage their publishing goals despite shortages of material resources and sometimes-conflicting rhetorical traditions. A strand of work complementing Swales's critical discussions of the growing global inequities in academic knowledge production (e.g., Ammon, 2001;Mauranen, 1993;Salager-Meyer, 1997) also informed our interest in how multilingual scholars respond to pressures to publish in English.
Our longitudinal ethnographic study has focused on the writing practices and publishing experiences of 50 scholars in education and psychology located in 12 institutions in Hungary, Slovakia, Spain, and Portugal. In these locations, scholars have varied and often limited access to the material resources that support doing and writing about research, including resources for learning English. We have been studying how scholars go about writing and submitting their work for publication, the tensions and conflicts they experience in response to pressures to publish in English (including tensions in their choices of how to allocate their time, energy, and resources), and their perspectives on these experiences. Through a range of academic networks, we contacted scholars in these sites and invited them to participate in our research; as a result, scholars who were writing for publication or presentation in English or were interested in doing so responded and participated. For our study we did not therefore seek to include participants who were not using English for publication, nor did we seek to investigate differences among scholars in these two groups. The general levels of proficiency in English of our research participants varied considerably, particularly according to linguistic register and mode (whether spoken or written discourse). As we note below, a key finding of the PAW study is that individual proficiency in English appears to be less salient to publishing success than is being part of academic research networks and working with "literacy brokers" (Lillis & Curry, 2006a).
In the main data collection phase of the study between 2001 and 2009, 60 visits were carried out by two researchers to scholars in 12 institutions within the four sites of our research. We have collected a wide range of ethnographic data: field notes made of our observations; approximately 6 Researchers in the sociology of science also focused on academic publishing but not specifically from the perspective of multilingualism (e.g., Bazerman [1988]; Knorr-Cetina [1981]; Latour & Woolgar [1986]; Myers [1990]). 1200 texts written by scholars-including multiple drafts as well as published versions of articles. In some cases, scholars began writing these texts in the local language and then used English to prepare them for submission to English-medium journals. 7 We also collected well over 500 pieces of correspondence about texts between scholars and others such as colleagues and journal gatekeepers (including reviewers' reports and editors' letters); conducted some 250 text-based individual interviews with scholars about their writing practices and publishing experiences and perspectives; and gathered documentary data from academic departments and institutions alongside national policy documents. Between our visits we maintained contact with these 50 participants by email and occasionally by telephone/Skype, also collecting these communications as data. The majority of these communications took place in English, with a few interviews in Spanish (thus the data extracts we include below are in their original language). As academic publication can take years from the moment when a text is begun, for example, as conference presentation slides, through drafts and revisions, to final publication in a journal (or abandonment), our longitudinal ethnographic methodology was particularly well suited to collecting this range of data across time as well as space.
A key way in which we analyzed this text-ethnographic data was to collect and/or reconstruct 'text histories' in order to trace the development and histories of particular texts in the trajectory toward submission and publication (or rejection). Each 'text history' was constituted minimally by three drafts of an article, plus at least one additional source of data such as interviews with scholars about the development of their texts and correspondence with journal gatekeepers. Some of the text histories involved considerably more data, such as 14 drafts of a text along with multiple rounds of journal correspondence (reviewers' reports, editors' letters, etc.), and email correspondence among co-authors, and interviews and emails with us. In analyzing our data, we (re)constructed 284 of these text histories (see Lillis & Curry, 2010, Ch. 1).
In analyzing our data across sites and participants, we also tallied scholars' publications records in different languages and genres, categorized the communities for which scholars write (discussed below, see Table 1), and created analytic tools including a "text-oriented heuristic" for tracking the influence of various people on specific changes across drafts, and network diagrams to map out scholars' connections with others and their collaborations. We also analyzed documentary data to construct an understanding of the policy and evaluation landscapes in which scholars work, which provided crucial context for our analyses. 8 Findings from our study can be briefly summarized as follows: multilingual scholars in our research sites have varied access to resources needed for academic research and publishing; scholars engage with various types of people we call "literacy brokers" who can support their publishing; they participate in local and transnational academic research networks; scholars publish in a range of communities, languages, and genres; they contribute to the production of English-medium national and regional journals, in part to develop capacity for local knowledge production; and they may engage in "equivalent" or "dual publication" to reach multiple audiences. Below we present "telling cases" (Mitchell, 1984, p. 239) of these salient findings. We also illustrate how multilingual scholars are, at different moments, engaging in both strategies and tactics of academic knowledge production, depending on their interests and commitments. 7 Contrary to popular beliefs about the role of translation in academic publishing, the scholars in our study tended not to use translators for reasons of availability, cost, and effectiveness, as explored in ; more often they self-translated or used both languages in the drafting of papers for publication. 8 For more detail on our text-ethnographic methodology, see .

Strategies at global, national, and institutional scale levels
As mentioned, strategies can be identified at different scale levels. We use the notion of scale following Blommaert et al. (2005) to explore how scholars' publishing practices are implicated in global power dynamics through relationships of hierarchy (see Lillis & Curry, 2010, Ch. 1). In codifying the criteria by which individual scholars are evaluated and rewarded, policy documents reify and instantiate these hierarchical relationships at different scale levels-here, institutional policy documents connect to local, national, and global scales of academic knowledge production which in turn influence institutional, national, and global evaluation systems. In articulating institutional priorities for research output, policy documents at the same time suggest official strategies for achieving objectives of institutions at various scales (university, government, etc.). Two examples of policy documents from our ethnographic data illustrate how policy documents are used both to evaluate research outputs and the ways in which different scale levels are built into the evaluation practices (Lillis & Curry, 2010, pp. 137-141). The example in Figure 1 comes from Spain's public university hiring criteria. These criteria specifically distinguish between an applicant's publications in journals with an impact factor and those without. Therefore, a key strategy at the global scale suggested by this document is for scholars to publish in journals with an impact factor (IF). The IF here acts as a proxy for other metrics of quality (e.g., the journal's article selection system, composition of the journal's editorial board, its inclusion in various databases) that are invoked only in cases where journals lack an IF. Consequently, by publishing in IF journals, a scholar does not have to explain why she decided to publish in journals that have not been codified as 'good' through indicators such as IF. This concern carries over to the evaluation of scholars once they have been hired, as articulated by Ernesto, 9 a Spanish full professor of psychology in our study, who discusses the role played by IF and ISI indexes in the evaluation procedures in his context: A national committee evaluates your curriculum [vitae] according to the impact of the journal so, for instance, if your publication has got into an important journal you don't need to justify it. . . . If you publish in a journal that nobody knows you have to justify the quality of your paper so it's more economic just to try to publish in a good journal. . . . In the defense of my curriculum [in the evaluation process], it's better if I can say [Journal A] has an impact, an index of such and such, or [Journal B] or so on.
(emphasis added) Scaling as a key phenomenon in academic publishing is indicated in Ernesto's comment in that while the evaluation happens at one scale level-the national-the journal's impact factor, coming from another scale level, currently operating at a global scale-is a key determiner of evaluation outcomes. Certain terms in Ernesto's comment illustrate the ways that ISI-indexed journals signal quality; for example, he equates "an important journal" and "a good journal" with a journal that "has an impact, an index." A second official strategy signaled in the Spanish hiring document is for scholars to publish in "international" journals, even those without an IF, which are perceived as having a global value. This strategy often means publishing in English-medium journals, whether because of the conflation of 'international' and 'English', discussed above, or because the inclusion criteria for some international indexes stipulate the use of English in a journal. In Ernesto's context, the appearance of a given journal in particular 'international' indexes is also consequential in relation to decisions about the internal distribution of institutional research funds: In my university we submit our research to a committee, they have a certain amount of money and they give research money according to the importance of this publication, for instance, and if your paper is published in a journal in the first half of the Social Science Citation Index you have three points and if not, then maybe one point. In the end they assemble all the points then divide the amount of money. (emphasis added) Ernesto's comment signals the interplay of practices taking place at different scalar levels: The evaluation at the local (university) level takes account of "the importance of this publication," which is partially equated to a journal's ranking in the SSCI. Therefore, in the Spanish system, official institutional strategies for publishing in high status journals-as indicated by the ways in which publications are evaluated and rewarded-have particular consequences for scholars at key career moments (hiring, promotion) as well as in receiving pay raises and obtaining research funds.
Similar policies and strategies are illustrated in Figure 2, which shows promotion criteria in the Slovak system. Unsurprisingly, journal article publications head the listing of data that a scholar needs to submit. Of additional interest is the prominence given in the document to the citations that a publication garners from other academic writings, with citation being an increasingly used metric of quality-at least for the scope of a journal's distribution (for more on citation practices, see Lillis, Hewings, Vladimirou, & Curry, 2010).
The scalar hierarchy for citations here mirrors the scales discussed above: 1) Citations appearing in the SCI, a measure legitimized as 'global', are most highly valued; 2) Citations appearing in other indexes, which may include regional or national journals, are less valued; and 3) Citations appearing in student texts are least valued. This hierarchy clearly signals a strategy for scholars to publish in ISI-indexed, English-medium journals-the journals most likely to secure these prized citations. 10 These examples illustrate the current global evaluation regime that, as discussed, values most highly English-medium 'international' publications in journals with an impact factor, thus included in ISI citation indexes, but also rewards scholars for publications which meet some of these or related criteria. The fact that rewards vary according to the relative perceived status of the journal in which an article appears is significant, as we explore next, for scholars' decisions about how, where, and to what ends to expend their energies and resources.

Multilingual scholars' engagement in strategies and tactics
As explored in the previous section, the policy documents structuring evaluation systems variably reward publishing at different scalar levels, from national to regional to global, while also suggesting strategies for meeting official objectives. In many cases, scholars' personal interests include publishing their work in high-status, indexed journals that they believe will reach the widest possible audience, or, a highly specialized audience not available in the local context (Curry & Lillis, 2004). Thus they adopt strategies that meet both their own and official objectives. Nonetheless, multilingual scholars also manage their resources in order to publish their work according to their own interests and commitments, which may include communities besides the English-medium 'international' research community (as in Table 1). In this section we explore the complexity of scholars' publishing practices by examining the interplay of the official strategies exemplified above and the sometimes-subversive tactics used by scholars, by discussing five of our key research findings in a particular order: starting with what we believe are strategies used both by institutions (which might be cast as the dominant) and scholars (who might be cast as the subordinate) and 10 For more on the complexities of citations in global publishing practices see  moving to tactics adopted by scholars as forms of resistance to the evaluation regime that pressures them to use English for publishing.

Interacting with literacy brokers.
In writing for publication, scholars interact with people we call "literacy brokers," that is, collaborators and gatekeepers who support or constrain access to publishing both before and after submitting to a journal (Lillis & Curry, 2006a, 2006b, 2010. We have identified two main types of literacy brokers-academic and language brokers-with academic brokers supporting access to resources, knowledge of the current "conversations" of academic disciplines and journals, and awareness of specific target publications, as well as offering assistance with crafting papers before submission and with responding to feedback post-submission. (Academic brokers also include journal gatekeepers who may constrain as well as support scholars' access to publishing.) The 284 text histories from our data analysis involved a total of 480 literacy brokers whom participating scholars identified as having played a direct role in the production of texts, much of which was documented in the text data we collected. We identified 83% of the brokers involved in these text histories as academic literacy brokers and 17% as language brokers. 11 Interactions with academic literacy brokers appear in the main to support multilingual scholars' efforts at publishing. In addition, enlisting the support of literacy brokers as a strategy for achieving publication aligns with official strategies suggested by institutional reward systems. Indeed, evaluation documents from some of our sites articulate specific rewards for scholars' collaborations with "international" or "foreign" colleagues. For instance, criteria from the Hungarian Academy of Sciences to obtain the Academic Doctorate, a prestigious post-PhD degree awarded to renowned scholars, include: Maintaining active relationships with distinguished international scientists, significant professional groups and associations in terms of joint publications, professional cooperation, mutual visits, or participation in foreign institutions. (Curry & Lillis, 2013a, p. 34, emphasis in Hungarian original) Implicated in this criterion is the strategy of "maintaining active relationships," presumably with people we would characterize as academic literacy brokers ("distinguished international scientists") who could write in a language other than Hungarian and engage in writing "joint publications." The preponderance of academic brokers in our data-and their mostly positive impact on publishing outcomes-indicates the crucial role they can play in supporting scholars in 'scale jumping' from 'national' to 'international' in terms of publishing outcomes. The involvement of academic literacy brokers takes various forms, including providing support to scholars in approximating in often subtle ways the kind of textual practices that journal reviewers and editors will recognize as 'international'-that is, the content and methodologies discussed, language used, and particular ideologies woven into texts submitted for publication (see . Literacy brokers are often situated in academic research networks, which we turn to next, that support publishing. 11 Using language brokers is another strategy that appears to align with both institutional objectives and scholars' interests. Indeed, language brokering is sometimes supported institutionally by the allocation of resources to pay for staff translators, for example, at central European institutions, and through departmental, institutional, and governmental budgets and research grants that may provide funds for language support (Lillis & Curry, 2006a). However, in our study the involvement of language brokers alone had little effect on achieving successful publishing outcomes for the scholars.

Participating in local and transnational academic research networks.
Academic research networks, both local and transnational, can provide scholars with access to the material and social resources that support publishing. Access to these material resources is a major concern for many multilingual scholars, who often lack funds to conduct research, to pay for bibliographic resources, and to travel to conferences where networks may be formed and sustained . Social resources include colleagues, students, and literacy brokers, discussed above. The involvement of literacy brokers and the importance of academic research networks help in explaining why individual competence in academic English is neither a prerequisite for, nor a guarantee of, publishing success, as noted earlier. For example, a Hungarian senior psychology scholar in our study, Istvan, who has one of the lowest levels of English proficiency, is among the most prolific in terms of publishing in high-status English-medium 'international' (EMI) journals. Istvan's success results from his longstanding collaborations with scholars in Anglophone contexts (Lillis & Curry, 2006b, 2010. These collaborations have resulted in a durable academic research network that includes Istvan's students and local colleagues as well as transnational collaborators and has resulted in some dozen EMI publications. Participating in transnational and local networks aligns with strategies emanating from the research dissemination goals of national governments as well as supranational bodies such as the European Union (EU). In fact, forging such connections is not only encouraged but even explicitly funded through, for example, specific EU programs .
At the same time, network participation becomes a tactic for scholars to achieve their own objectives. Our findings show that scholars' participation in transnational networks, including formal EU networks, had variable results in terms of joint EMI publications. Indeed, such involvement strengthened scholars' local research networks and supported their publications for local communities. In local networks, multilingual scholars can enact their interests in fostering local research cultures, supporting graduate students and colleagues, and contributing to the development of practice to benefit local contexts . For example, as Slovakia joined the EU, senior psychology scholar Olivia was invited to participate in an EU-sponsored research project involving five countries that aimed to adapt an English-medium literacy assessment for the partner countries. Her involvement in this network most obviously satisfied official objectives of her institution and government as well as the EU for collaborating on transnational grants. At the same time, Olivia used this collaboration to further her own objectives for the local research and practice communities. Figure 3 illustrates this transnational academic research network.

Figure 3. Olivia's European Union Network
While no EMI journal publications resulted from the project, participation in this network resulted in the strengthening of Olivia's local network of colleagues, students, and school teachers, and furthered her priority of supporting local practice: What I am much more interested in is publishing about [the EU project's evaluation instrument] in our home journals because [network members] abroad, they may be interested about doing these other things internationally or cross-linguistically or cross-culturally but what I would like to do is to publish the result and everything about it the project and the [evaluation] instrument here just to let teachers know. (Curry & Lillis, 2010, p. 290) In the example of Olivia's participation in a transnational, EU-sponsored network, an official strategy of 'international' collaboration also became a tactic in support of a scholar's interests and objectives of supporting the local research and practice communities.

Publishing for a number of communities.
Despite the pressure to achieve high-status EMI publications, multilingual scholars are publishing in a range of academic and practice-oriented communities, using multiple languages and writing in various genres (Curry & Lillis, 2004, 2010. Table 1 shows the language(s) used in the communities for which scholars are writing and the types of publications they contribute. Scholars contribute to their disciplines transnationally by publishing in English-medium 'international' journals (Community 7), which helps them build reputations, achieve promotion and tenure, and obtain research grants, and typically garners the highest rewards in evaluation systems, as discussed above. Most participating scholars are interested in these outcomes (as are many scholars around the world). In this case, therefore, while publishing in EMI journals most obviously aligns with institutional pressures and strategies for research outputs being evaluated as globally significant (i.e., as 'international'), it may also benefit scholars' individual careers and reputations.
In addition, multilingual scholars also contribute to English-medium journals published and distributed in local and 'intranational' contexts (Communities 3 and 5) in order to distribute their work to wider audiences and to achieve 'international' recognition. The distinction between these English-medium communities and the transnational English-medium community may be subtle, but publishing in these communities plays an important role in scholars' abilities to enact official strategies, although they may be less rewarded than for obtaining EMI publications. At the same time, such publishing can become a tactic by which scholars can build local or regional (while still transnational/ 'international') communities for research or practice, for example, where specific questions or issues are addressed that may be of particular interest to the region. We return to this point below in discussing the growth of English-medium national (EMN) journals.
Additionally, many scholars remain committed to publishing in research and applied journals in local languages (Communities 1 and 2) in order to share their work to interested research and practical audiences as well as to develop local knowledge-building capacity with their students and colleagues who may not have as much access to English (or value the local language more). To a lesser extent, scholars are also publishing in transnational communities that use local language (Community 4), or other national communities in other languages (Community 6). Scholars also produce these publications in order to maintain or sustain disciplinary conversations that may have developed from a shared local or regional history. Table 1 Academic communities to which multilingual scholars contribute Community/Language Type of Publications 1. National academic community in the local national/state language (e.g., Spanish, Hungarian) Scholarly publications aimed at national research community 2. National applied community in local national language Practitioner publications aimed at users of research findings such as teachers, health professionals, psychologists 3. National academic community in English medium Scholarly publications often aimed at a wider audience than the audience reached by publications in the local national language 4. 'International' academic community in the local national language Scholarly publications using local national languages aimed a wider audience than the local context (e.g., Spanish in Latin America, Portuguese in Brazil and Africa, Slovak in the Czech Republic)

'Intranational' academic community in English medium
Scholarly publications aimed at a transnational or transregional audience with strong political links, such as the European Union 6. Other national academic community in national languages Scholarly publications in languages that represent intellectual traditions related to particular (sub)fields, as well as geo-historical relations such as the role of German and Russian in central European contexts like Slovakia and Hungary 7. 'International' academic community in English Scholarly publications typically produced in Anglophone contexts and distributed worldwide, which increasingly often have higher status than journals published in other parts of the world (and sometimes indicated by inclusion in high status indexes and with an impact factor) Source: Adapted from Lillis & Curry (2010, Ch. 2) In sum, by publishing in English-medium international journals (some of which are highstatus ISI-indexed journals, while others are not) as well as English-medium national or regional journals that are often excluded from ISI indexes, 12 scholars enact strategies that align with the official strategies of their institutions or governments and may also support their personal objectives. At the same time, scholars publish in journals produced in local or global (non-English) languages, which may constitute a tactic that contravenes high-priority institutional objectives but serves scholars' own priorities and interests.
Nonetheless, alongside articulated pressure for high-status (EMI) publications, official pressures for publication in local communities also exist. These competing pressures are illustrated by the experiences of Hungarian associate professor of education, Julie. At the outset of her career, with the support of her PhD advisor who became a literacy broker for her to access high-status EMI publications, Julie published almost exclusively in such EMI publications. Later, she was encouraged to apply for the Academic Doctorate (mentioned above) for multiple reasons: Earlier I didn't publish a lot in Hungarian but now I have to, because our department is very young and we need somebody who could pass this academic doctorate, and for that I had to be known locally and much more known outside. . . . You cannot start your department, cannot start an MA program if you don't have somebody who has this Academy of Sciences doctorate, you cannot sit on a PhD examination, you cannot be the head of an examination board, in a PhD comprehensive exam if you don't have this, and I'm the person [in my department] who is closest. … I have everything I need for this but everybody told me that I'm too young […]. The only way I can get it so young is if I have a good record of publications. (Curry & Lillis, 2013b, p. 36) Thus multilingual scholars' priorities for publishing in different communities may shift over their careers based on their aims and interests at particular moments as well as official pressures. In response, their use of various strategies and tactics also shifts. In addition, besides writing journal articles, for a variety of reasons, scholars also produce other genres-conference proceedings, chapters in edited book, and books-for which they use multiple languages (Curry & Lillis, 2004;2013b;. 13 4. Developing English-medium national and regional journals. Many multilingual scholars are actively working to develop English-medium national (EMN) journals as editors and reviewers as well as contributors. They often do so in addition to contributing to local journals using the local national language as well as publishing in 'international' journals in English. Over the duration of our study, we have noted the growth in EMN journals in our research sites and other contexts and Lillis (2012) has investigated the perspectives of journal editors. Such journals occupy a complex position in global academic publishing and reflect the combined strategic/tactical engagement in which many scholars engage (see Figure 4 for a profile of one EMN journal). That is, such journals allow scholars to publish in English, thus fulfilling 12 Selection criteria for the listing of journals in ISI indexes include citations, peer review, composition of editorial board, frequency and timeliness of publication, including English-medium abstracts, and reputation of the publisher or sponsoring association. Notwithstanding these seemingly objective criteria, a heavy bias toward journals from the Anglophone center has been documented, manifested in the preponderance of journals publishing not only abstracts in English but entire articles/issues (Lillis & Curry, 2010, Ch. 1). 13 See Lillis & Curry (2010, Ch. 2) for a categorization of the publications by the scholars in our study according to these genres and languages.
institutional objectives for English-medium publications (strategies), yet at the same time addressing more specific local or regional audiences (tactics). In terms of the "economies of signs" of global publishing (Lillis, 2012), the very existence of such journals seems to be purely strategic, most obviously by using English rather than the local national or regional language. However, while the growth of these journals can be seen as aligning with official strategies for English-medium or 'international' publication, they also represent multilingual scholars' tactical engagement in a number of ways: 1) Their existence challenges the global evaluation regime which puts value only on journals included in ISI indexes and with impact factors. Thus scholars' decisions to edit such journals, when they know they will face an uphill battle to have these journals included in ISI indexes, or in fact have no expectation that the journals will be included, in part signals a tactic of operating outside the dominant evaluation system; 2) Scholars' decisions to use English reflects not only alignment with the dominant strategy of English as the global language of science but also a challenge to this strategy-in that English(es) are a semiotic resource increasingly used around the globe and, in the process of use, people are transforming the resource itself and the purposes for which the resource is being used; 3) The production of EMN journals alongside journals in local/national languages can be viewed as providing an additional intellectual resource both in scholars' own terms and as a bridge toward publishing in Anglophone-center journals, should writers so desire. In many contexts, EMN journals in the social sciences are not in fact replacing journals produced in other languages but exist alongside these; 4) EMN journals sometimes constitute a tactic at a national scale level by opening up new intellectual spaces and thus often challenge dominant local/national-as well as transnational-regimes of knowledge production.
An interdisciplinary social sciences journal established and located in Slovakia, this journal has been in existence since 1990 but has gone through three main stages in terms of disciplinary focus and overarching goals. In Stage One, the disciplinary focus was wide ranging, but with an emphasis on work focusing on Slovak culture, society, and history. Stage Two involved a change in editorship and the development of a specific epistemological orientation toward wide-ranging content and interest-that of an explicit inter/transdisciplinary orientation to key issues and topics in human sciences, including disciplinary areas such as philosophy, psychology, politics, and linguistics. Stage Three, the current stage, sees the journal continuing with its explicit transdisciplinary focus. In terms of location and publishing, the journal continues to be located in a Slovak academic institution, but, whilst during Stages One and Two it was published by a local academic publishing company, Stage Three involved accepting a contract with a multinational publishing company.
The journal is published in English and accepts papers in English and Slovak or Czech. A translator has been involved in the journal for some 20 years and works on translating papers into English and proofreading and 'correcting' papers submitted in English by authors who are "nonnative" users of English. Reviews are sometimes carried out in the language(s) in which articles are submitted (e.g., a Slovak-medium paper submitted in Slovak is reviewed in Slovak) and sometimes using translation (a Slovak-medium paper is translated into English and reviewed in English). The journal appears in a number of indexes, including Philosopher's Index, Sociological Abstracts, Linguistics & Language Behavior Abstracts, and The Central European Journal of Social Sciences and Humanities. The description of the journal in Figure 4 signals what we might consider a strongly tactical orientation by scholars (as editors and authors): linguistically, by using English as the medium of publication but enabling other languages to be used during at least part of the publishing process; publishing in English to locate the journal as a globally valuable product but at the same time forging a transdisciplinary space which meets the interest of local researchers and is of transnational significance; and seeking to be included in indexes that are legitimized as "global" whilst producing a journal over which local/regional scholars have considerable control.

Engaging in 'equivalent' publishing.
In addition to publishing different texts for multiple communities as discussed above, scholars may publish similar texts in multiple languages for a range of communities. They do so in order to make the knowledge and practical applications from their research available in a number of communities and to support the development of local research and practice communities. In many contexts (particularly the Anglophone-center) publishing different-language versions of translated or newly written texts has historically often been framed as self-plagiarism or as claiming novelty for previously presented work, even though the audiences for these different-language publications quite likely do not overlap. Indeed, the journals most valued in current institutional reward systems often explicitly prohibit the submission of material that has been previously published or is under consideration elsewhere.
The complexities of today's global publishing marketplace, however, are prompting multilingual scholars to adopt a tactic that we call "equivalent publishing"  and others term "dual publication" (Wen & Gao, 2007), that is, publishing closely related versions of academic texts in different languages for different communities. As an example of equivalent publishing, Figure 5 maps the publications by two senior scholars who have worked together for more than 40 years, and their research team, writing in both Portuguese and English. Source: Adapted from Lillis and Curry (2010, p. 132). Figure 5 shows that these scholars work hard to publish equivalent texts in both languages. Almost one-third of their publications are available in both English and Portuguese (as journal articles, book chapters, and conference proceedings). Of the publications in both languages, from our ethnographic data we know that 67% of these were written and published first in Portuguese and then in English; 15% were written first in English and subsequently in Portuguese; and 18% were published in both languages simultaneously, that is, in the same year. These scholars consider that equivalent publishing enables them to reach the communities to which they believe their research is relevant: They are committed to disseminating their work in Portuguese in the local national context to contribute to national research and policy; at the same time, they are passionate about building their research field transnationally and consider English to be the linguistic medium which enables them to do so (Curry & Lillis, 2013b).
However, journals published in the Anglophone center typically prohibit dual publication, viewing it as a form of self-plagiarism that contradicts the ethos of presenting new knowledge in academic journals (cf. Hamp-Lyons, 2009, and others in a TESOL Quarterly Forum, see also Curry & Lillis, 2013a). As discussed, for some scholars the realities of the global publishing marketplace create exigencies to which scholars respond with the tactic of equivalent publishing. This response is not motivated by any intention of what is often viewed by journal editors as "cheating the system" by garnering duplicate publications. In fact, as Figure 5 illustrates, scholars are keenly aware of the fine-grained distinctions among the communities to which they are targeting their publications and their reasons for doing so. The practice of equivalent or dual publishing could clearly help to resolve key tensions between publishing in 'international' (English-medium) journals and publishing in journals that support the interests of local and regional contexts. As we indicate below, we believe that the moment has arrived for Anglophone-center journal gatekeepers to reconsider their strict proscription against dual/equivalent publishing.

Conclusion and Implications
In the dozen years since we began our research into the experiences of multilingual scholars with writing for academic publication-particularly in response to the growing pressure to publish in English-pressures on scholars have intensified and the explicit codification of these pressures into policy documents has become more common. The official objective of publishing in ISI indexed journals has had the effect of narrowing the range of the most highly rewarded publishing outlets. In this article we have highlighted strategies attendant on these official policies for the distribution of academic knowledge: to publish in journals included in ISI indexes and with impact factors, thus signaling the need to publish in English; additionally, to publish in what are considered 'international' journals (though they may not be included in ISI indexes), which often signals English-medium; and to collaborate and co-author with 'international' scholars, often using English or other languages-by participating in academic research networks and working with literacy brokers. We have also articulated some of the tactics that multilingual scholars use in pursuit of their own interests and objectives, particularly in cases where these do not align with official objectives and strategies. These tactics include publishing in multiple languages and genres; publishing in and indeed developing English-medium national journals; and engaging in equivalent publishing.
As we have discussed, scholars are reacting to intensified publishing pressures in multiple and varied ways at different moments in their careers. Their responses include engaging in practices that enact both official strategies and individual tactics in relation to academic writing and knowledge production. It is important to emphasize that the scholars who participated in our research for nearly a decade are not passively or unquestioningly accepting the dominant pressure to write in what are, effectively, English-medium publications. When it suits their own interests and priorities to aim for publication in high-status EMI journals, they are doing so, often with success. Yet in cases where scholars' interests conflict with official institutional objectives, they may adopt tactics that enable them to pursue their own aims. Notably, some scholars are engaging in equivalent publishing of their findings and supporting the production of EMN journals by both contributing articles and working as editors and reviewers. These tactics support the development of local research cultures and the use of locally produced knowledge for the benefit of the contexts in which it is produced. As noted, the increased global competition to publish in ISI-indexed journals is likely to become untenable as the pressure to publish in a narrow set of journals reaches a greater number of contexts. Scholars' tactical responses including EMN and equivalent publishing attest to the shifts initiated by these pressures and are important to consider for the ways in which they represent scholars' commitments and priorities.
The theoretical metaphors of strategies and tactics borrowed from de Certeau serve well in framing scholars' different responses to the complexities of the global publishing landscape. While in ordinary usage these terms are often considered to be synonyms, the distinctions that de Certeau endows them with allow for a nuanced examination of how particular actions-practices-might serve different interests at different moments. Introducing the notion of scale (Blommaert et al.) additionally assists in analyzing the complexity of publishing in a globalized world, where the 'publishing context' cannot be viewed as one flat or horizontal space. Writing for publication takes place in many different contexts of production but is 'taken up' (or not) and evaluated (negatively or positively) in relation to a scalar hierarchy in which ideologies of language and locality play out in complex ways (see Lillis & Curry 2010, Ch. 6). The rapid shifts we have noted in knowledge production practices globally also call for a reexamination of the proscription against equivalent publishing, particularly for the complex situations of multilingual scholars aiming to contribute to a range of communities using multiple languages. To this end, we urge those Anglophone scholars working in gatekeeping roles such as journal reviewers and editors and/or in policymaking positions to help open access to the global knowledge production marketplace for scholars working outside the Anglophone context. While some journals, for example, in the field of applied linguistics, have articulated policies or programs to support the writing of multilingual scholars, these appear not to have taken much hold, not least because they rely on the volunteer labor of "literacy brokers" often from center-based contexts to support the efforts of individual authors. Anglophone scholars in gatekeeping positions can also lead efforts to expand the currently narrow conceptions of "acceptable" Englishes used in academic publishing. They can also help reviewers understand the constraints on multilingual scholars in terms of access to the current research literature and in-vogue methodologies that are often central to an article being deemed acceptable by gatekeepers.
In addition, we encourage Anglophone scholars to take up some of the tactics used by multilingual scholars to challenge the growing dominance of academic evaluation systems that privilege the use of questionable metrics of quality such as the impact factor and the dominance of a narrow range of ISI journal citation indexes. To start, center-based Anglophone scholars can resist the growing imposition of these policies (including the use of IF in evaluations) in our own universities by highlighting their biased and superficial nature. Anglophone scholars can also seek to publish in journals, including open-access journals that might be more accessible by scholars with limited bibliographic resources. Indeed, we might even publish in other languages.

About the Guest Editor
David Post UNESCO GMR & Pennsylvania State University. Post@psu.edu David Post is Senior Policy Analyst with the Global Monitoring Report of Education For All, and is currently based at UNESCO in Paris. He also is Professor of Comparative and International Education who is currently on leave from Penn State University in the USA. He has researched and published about educational stratification, about child labor issues, as well as the politics of educational mobilization. He also investigates the impact of concurrent employment on student academic achievement. He has been a visiting professor at the Colegio de México, at the Facultad Latinoamericana de Ciencias Sociales, and at the Hong Kong University of Science and Technology. Finally, David served for ten years as editor of Comparative Education Review, where he became concerned about the commodification of scholarship and the possible responses to it by intellectuals, for example through peer-reviewed, open access publication of studies like those in this special issue. Last year, the EEPA printed the Spanish version of his commentary, "Los Rankings Académicos." http://epaa.asu.edu/ojs/article/view/1347 Readers are free to copy, display, and distribute this article, as long as the work is attributed to the author(s) and Education Policy Analysis Archives, it is distributed for noncommercial purposes only, and no alteration or transformation is made in the work. More details of this Creative Commons license are available at http://creativecommons.org/licenses/by-nc-sa/3.0/. All other uses must be approved by the author (s)  Please contribute commentaries at http://epaa.info/wordpress/ and send errata notes to Gustavo E. Fischman fischman@asu.edu Join EPAA's Facebook community at https://www.facebook.com/EPAAAAPE and Twitter feed @epaa_aape.