Open Access With Chinese Characteristics: Understanding Recent History and Current Practice via Qualitative Interviews at a Large Chinese Research University

Chinese scholars, administrators, and librarians possess nuanced understandings of what defines open access (OA) in China and the barriers that make wider adoption of OA difficult. When we say “ OA ” in the United States, we imply a complex set of underlying assumptions tied to the history and practice of scholarship. Saying “ OA ” in China brings with it a similarly complex set of assumptions that may not be commensurate with the OA that we speak of, and such one-to-one translation may not be possible or desirable given the unique historical, political, and linguistic differences between the world ’ s two largest producers and consumers of schol-arship. Through a careful analysis of our participants ’ observations and a review of the history and context of Chinese academic institutions, we posit that “ OA with Chinese characteristics ” describes a set of possibilities and constraints that determine how Chinese academics experience both the theoretical project and the practical distribution method we commonly call OA. Although these multiple understandings of OA may not converge on a single shared meaning, we can endeavor to understand one another better in the service of creating and sharing knowledge.


INTRODUCTION
In the fall of 2020, sociologist Zeynep Tufekci described how Chinese scientists and health care workers gave the world the gifts of "time and information," which contributed to the speedy development of effective vaccines for COVID-19 (Tufekci, 2020). As Tufekci notes, open sharing of an early genome sequence for the virus kicked off a flurry of global research activity, culminating in the development and testing of vaccines that have saved many lives in the midst of a global pandemic. However, even in the years prior to the pandemic, the interdependent (and often contentious) relationship between China and the United States was deepening (Hass, 2021), with particular relevance for academics, as both countries are powerhouses of scholarly production.
Working on a research project on open access (OA) in China during this time, we have found ourselves frequently reflecting on the importance of open and free exchange of scholarship as well as the ways in which nationalist chauvinism and racism can be sources of material harm in our daily lives during a fraught moment in US-China relations. When we began this work in 2018, we were (and still are) motivated by a desire to understand how OA was practiced and what it meant for scholars "on the ground" in China. We approached our work as a "critical" ethnography (Bhavnani et al., 2014), aware that our findings would necessarily be partial and rooted in the particular experiences of our research subjects but also affected by larger economic, political, and historical forces.
As we describe later in this paper, our findings demonstrate that Chinese scholars, administrators, and librarians possess nuanced understandings of what defines OA in China and the barriers that make wider adoption of OA difficult. Despite these challenges, our participants expressed hopes for a future in which transformational change in the way scholarship is created, disseminated, and consumed might be possible. In our interviews with research participants, we found a number of themes that would be familiar for those who have studied OA publishing in North America, including the importance of external incentives for encouraging authors to choose OA publishing venues, the conservative tendencies of the academy, and successful marketing by publishers that has allowed "gold OA" to serve as the standard (in the views of some scholars) for all openly licensed distribution of scholarship.
A fuller understanding of the particular experiences of Chinese academics is critical for understanding how OA is practiced in an increasingly interdependent world because China is one of the largest producers of scholarship. Through a careful analysis of our participants' observations and a review of the history and context of Chinese academic institutions, we posit that "OA with Chinese characteristics" describes a set of possibilities and constraints that determine how Chinese academics experience both the theoretical project and the practical distribution method that we commonly call "OA" in the US. Whereas "OA with Chinese characteristics" and the unmarked category of "OA" in the US might appear quite similar at first glance-comparable technical infrastructure and stated goals to increase the reach of scholars' works and similar assumptions about OA among scholars, administrators, and librarians-important distinctions persist below the surface, including important implications regarding chosen language of publication and the scholarly autonomy of Chinese academics. As our interviews with participants and review of the current landscape of Chinese academic practice demonstrates, careful attention to these differences reveals a form of OA which is employed in the service of a powerful central government's control over scholarly production, funding, and dissemination.

LITERATURE REVIEW
Given our desire to focus on the contemporary practice of OA as well as the volume of writing on OA as a worldwide phenomenon, we limited our literature review to works published after 2010, with a particular emphasis on the most recent scholarship (published in the last 5 years) in both Chinese and English.
Chinese-language scholarship on OA OA is rendered into two different terms in Chinese: "kaifang cunqu 开放存取" and "kaifang huoqu 开放获取," with the first two characters ("kaifang") meaning "open." These terms are mostly used interchangeably but have different connotations depending on the context of the JLSC research project. The former indicates both depositing and retrieving, and the latter emphasizes only retrieving.
A considerable number of Chinese-language research articles feature case studies of OA policies, projects, initiatives, products, resources, and best practices in foreign countries, especially the US and Europe. This demonstrates that the OA-related scholarship in China has been internationally minded.
This body of scholarship portrays a comprehensive picture of the OA landscape. Whereas OA journals have been the center of scholarly attention, other OA phenomena such as institutional repositories, OA monographs, open educational resources, and research data are also extensively discussed. However, the focus has been disproportionately on OA in the science, technology, engineering, and mathematics (STEM) fields, which is similar to that of English-language scholarship, and only a small number of articles are dedicated to arts, humanities, and social sciences fields. The OA-related scholarship in China also indicates inclusivity and diversity in terms of the authors' backgrounds and their perspectives. Contributors are from communities of library professionals, publishing professionals, scholarpractitioners, library and information sciences scholars, and scholars from other fields (e.g., humanities, communication, computer science).
One notable trend is to investigate how the library interacts with and acts as an indispensable stakeholder of the OA movement in China. For example, Miao and Liu (2016) argue that the roles that Chinese libraries play in OA publishing are three-fold: 1) collection builder for better navigation and aggregation of OA resources; 2) OA content publisher through institutional repository management and the library's own OA journal publishing enterprise; and 3) sponsor and advocate for OA publishing efforts of academic community.
Besides mostly positive and welcoming views of OA as a concept, some scholars also put forward their cautions about and disapproval of contemporary OA practices. For example, based on case studies of well-known OA journals in STEM disciplines, Jiang and Mu (2018) strongly criticize that, on an international scale, OA journal publishing is an underwhelming "utopia" (乌托邦) that does not live up to its promise of radical changes. It is largely driven by commercial interests and has caused tremendous wasting of research funding of Chinese scholars. The current OA publishing model, if not revolutionized, is detrimental to the entirety of academic publishing and will finally meet its demise. Therefore, they suggest that it is not necessary for Chinese STEM journal publishing to engage in these OA practices (Jiang & Mu, 2018).
Generally, quantitative research methods are dominant in these research projects, with only a small number of studies using qualitative methods. Even fewer studies have adopted the methodology of in-depth interviews and incorporated viewpoints from librarians, administrators, academics, and publishing professionals. This article will address this methodological lacuna.

English-language scholarship on OA in China
Among recent publications detailing the current state of OA publishing practice in China, we find a number of important factors that characterize this scholarly environment and that might not be familiar to readers in North America who are unfamiliar with academic practice and publishing in China.
First, China is one of the largest producers of scholarly articles in the world (Tollefson, 2018). The growth of Chinese academic journals reflects larger, long-term trends in economic growth and investment in key areas of science and technology (Wang et al., 2018). Looking at OA scholarship specifically, China is the second-largest source of OA publications in a major international index for the sciences (Web of Science's Science Citation Index Expanded) but ranks seventh in contributions for the social sciences (Web of Science's Social Sciences Citation Index), reflecting a disciplinary imbalance in OA publishing output (Liu & Li, 2018). Despite the size of academic output in China overall and in English-language OA publishing specifically, OA scholarship published in Chinese is often not accessible via international indexes, which negatively affects Chinese-language journals' ability to attract both readers and submissions and possibly contributes to perceptions of low quality (Shen, 2017).
Second, the economic conditions of academic publishing in China differ significantly from what we experience in North America. China has no "serials crisis," as the existence of scholarly journals does not depend on a commercial market supported largely by institutional subscriptions sold by for-profit publishers (Montgomery & Ren, 2018). Whereas OA practice outside of China is heavily influenced by these large commercial publishers, there is no Chinese equivalent of Elsevier driving author activism, development of institutional publishing infrastructure, or vertical integration in the industry. Instead, OA journals in China face challenges unique to the Chinese academic environment, perhaps the most important of which is a lack of widespread investment in technical infrastructure for journals themselves, particularly in the social sciences. An investigation of OA journals in the Chinese Social Sciences Citation Index (CSSCI), an index of the country's most prestigious social science publications, found that only half of the 714 journals included in the index had their own websites, and fewer than 14% of the journals could be considered OA in the most generous sense (Guo et al., 2014). Although there has been investment in the creation of institutional repositories, these platforms often provide access to legacy literature rather than newly published scholarship or preprints (Zhong & Jiang, 2016). Third, publication incentives play an important role in influencing author behavior, particularly among authors at research-intensive universities. Publication incentives have been in use in various forms since the 1980s and have, in recent years, even found buy-in among humanities and social sciences disciplines (Xu et al., 2019). These incentives have typically rewarded publication in English-language journals, although they are changing in response to pushback from Chinese authors who wish to see domestic journals recognized for their contributions to scholarship (Xu, 2020).

METHODS
In the summer of 2018, we traveled to a major research university in Shanghai, China (hereafter referred to as "Chinese Research University"). At the time we conducted our research, the University of Michigan and Chinese Research University sponsored a program of mutual exchange to encourage research and collaboration between scholars at our respective universities. After traveling to Chinese Research University, we conducted semi-structured interviews with five librarians (two of whom also served as administrators), four faculty members (two of whom also served as administrators), and one publishing professional from the institution's university press. All participants were recruited with assistance from a scholar and administrator affiliated with the Chinese Research University, focusing on individuals in the social sciences who already had at least some knowledge of or interest in OA publishing. After our return to Ann Arbor, Michigan, we interviewed two additional participants (one librarian and one faculty member) from Chinese Research University who had traveled to the University of Michigan for an unrelated reason. It should be noted that one limitation of this study is that all participants were affiliated with the same major research university in Shanghai. The experiences of faculty, administrators, and librarians at this institution may not reflect those of smaller universities or educational institutions in less-affluent areas of China.
All interviews were conducted by the authors (Fu and Kahn), one of whom is a native Chinese speaker. Whereas the majority of interviews were conducted primarily in Chinese with some clarifying discussion and follow-up questions in English, two interviews were conducted primarily in English with some clarifying discussion in Chinese. Participants were free to answer in English, Chinese, or a mixture of both languages as desired. After interview recordings were transcribed by graduate student employees, the Chinese-language transcriptions were translated by Fu into English, and English-language transcriptions were edited by Kahn.
Once transcription and translation work were completed, we chose to use only the Englishlanguage transcriptions and translations for further analysis. We used Johnny Saldanã 's (2015) The Coding Manual for Qualitative Researchers (3rd edition) to explore various approaches to coding qualitative interviews and settled on using a mixture of descriptive, in vivo, and concept coding, focusing on major themes that we saw in the interview data, with particular emphasis on a thematic analysis of answers to our primary research questions. All interview transcripts were coded using Dedoose, a cloud-based qualitative analysis software.
This research project was supported with a generous grant from the University of Michigan's Lieberthal-Rogel Center for Chinese Studies. Grant funds were used to pay for travel to and from China and for qualitative research analysis software. Additional funds from the University of Michigan's University Library were used to pay for graduate students who transcribed recorded interviews. Study design was approved and deemed exempt from continuing review by the University of Michigan Health Sciences and Behavioral Sciences Institutional Review Board (HUM00128386).

RESULTS
Out of nine interviews we conducted, four included interviewees with hybrid identities (Table 1). Given the complexity of academic appointments in both the US and China, it is not unexpected to find participants who might be simultaneously a faculty member, administrator, and/or librarian. Two interviews were conducted in small groups, one of which featured participants of different types (Table 1).
In the coding process, we created 27 codes for application on all the interview transcripts ( Figure 1). The codes that could be applied to each interview vary considerably because,  although we used a standard set of questions for each participant type (see Appendix A), we also added customized questions suited to each participant's specific positions and roles. Therefore, the total number of codes appearing in each transcript reflects each participant's identity, breadth of the interview, and the interview questions.
Among the 27 codes, the top 5 with the most applications are "What is OA?," "academic evaluation and rewards," "imagining a boundless future," "individual barriers to OA," and "who should lead?" Whereas code frequency can tell us which topics were discussed at length and across different participants, we found code co-occurrence, a common tool of thematic analysis (Figure 2), to be a more-revealing indicator of participants' understanding of OA and its implications for their own scholarly practice. In addition, we discovered that some of the less-frequently applied codes offered meaningful opportunities for analysis, such as "academic freedom" and "language politics."

Examining co-occurring codes
In our analysis of the interviews, we found code co-occurrence to be a useful lens for examining how participants' understandings of OA affect their beliefs about how the larger ecosystem of academic publishing functions, how individual and collective choices might be constrained, and what they imagine is possible in the future. This reflects our methodological choice to use a mix of descriptive, in vivo, and concept coding (Saldaña, 2015) because we were interested in the broad themes of respondents' answers while keeping those themes rooted in the particularities of respondents' experiences and knowledge. As illustrated in Figure 2, code co-occurrence happens when multiple codes are applied to the same passage in a given interview. As a tool for qualitative data analysis, code co-occurrence can help us as researchers understand the complexity in respondents' answers to our interview questions. The quotations and discussion included later in this paper represent areas in which particular codes co-occurred numerous times and where these overlapping themes drew our attention to important insights.
"Individual barriers to OA" and external incentives for academics The "individual barriers to OA" code and the "academic evaluation and rewards" code cooccurred nine times throughout all interviews. Most faculty interviewees expressed the view that authors' individual behaviors go beyond personal control and are influenced by larger forces, especially the academic evaluation system. Managerially speaking, despite the long-existing academic ranking and performance evaluation system, academic positions in Chinese public universities have traditionally been associated with permanent employment internal to an institution or the so-called "iron rice bowl." In the face of global competition and recruitment, top-tier universities in China started to transform by adopting the Western-style tenure-track model in the early 2000s. In reality, this reform has developed into a unique system in which the legacy of permanent employment interweaves with the nascent tenure-track operation to evolve "up-or-out" into "up-or-transfer" inside the institution instead of direct dismissal from that institution (Wang & Jones, 2021). Although external incentives for faculty promotion in the contemporary Chinese higher-education context work differently from the tenure-track system prevalent in US universities, they both include the foundational principle of motivating scholars to publish. Therefore, changing authors' publishing behavior will likely require adjusting both short-and long-term incentives.
From our interviews, we see that promotion in academic rank is the top priority for junior scholars, and many senior scholars value further career advancement. Despite many differences, academics in China and in the US both have to conform their publication activities to already-existing evaluation criteria in order to get rewards they expect. Given the dependence of achieving career goals on the current academic evaluation system, scholars can be conservative and reluctant to try new publishing activities and models that go beyond the traditional without potentially contributing to their academic profiles: "Nowadays we are so busy at work and life, and we may not have energy and motivation to try new things. Trying new things really needs a lot of attention and energy. Instead of doing so, we are influenced by familiar things in an accumulative manner." (Interview H) Scholarly conservatism in choosing the OA model to publish is not solely an individual choice made by authors themselves. This is because the academy is itself risk-averse, and the widely adopted current systems of academic evaluation and promotion have not been adequately incentivizing risk-taking research and publishing activities. In this situation, scholars often have to face a choice between taking risks or following the traditional path up the academic ladder. This tension between changing attitudes about publishing and traditional systems of reward exists among both senior and junior scholars. However, from the interviews, we see that this tension is amplified in the community of junior scholars: "…If I were a student in STEM fields and I finish writing an article, my advisor who thinks my article is high-quality and creative may recommend that I try submitting to some journals which are traditional, high-quality with high impact factors, and recognized among the field. My advisor may not put OA journals as the top choice, I think." (Interview D) The aforementioned quotation is a hypothetical situation but based on a librarian's experiences working with faculty and students in their liaison departments. It demonstrates a librarian's concern that student researchers would receive less encouragement from experienced scholars to publish in OA venues. In addition to this hypothetical situation described earlier, we heard about this unintentional avoidance of emphasizing OA publishing from scholars themselves. For example, in an interview with a faculty member in a social sciences discipline, they mentioned that they, when teaching an undergraduate course on thesis writing and publishing, "didn't very much emphasize to my students to which kinds of journals they should submit, traditional or non-traditional." Instead, they asked students to select publishing venues based on their research topics: "After all, there are so many journals in [their discipline], and there are so many subfields in [their discipline]. So it's most important to choose a journal of which the topic fits yours the best" (Interview B).
OA, the "Author Pays" model, and the economics of publishing In a number of our interviews, participants discussed their understanding of OA as a model that requires payment on behalf of authors as being directly connected to both economic forces at play in the larger landscape of academic publishing and the particular challenges and benefits that individual authors face while navigating that landscape. The codes "What is open access?" and "challenges and benefits of author-pays model" appeared together eight times (Figure 2), as did the codes "economics of academic publishing" and "challenges and benefits of author-pays model." Given what we know from our anecdotal experience talking to scholars and library colleagues in the course of our work as liaison librarians and what the literature on OA tells us about author motivations and behaviors, this is not an entirely surprising finding. But it is notable that it holds true in the Chinese context in which we conducted this research.
In our conversations, participants conflated OA as a broader concept for the distribution of scholarship with the gold OA business model of many academic journals published by commercial presses, which would necessarily constrain authors' choices and behaviors: "I just know that if you want to publish a paper in an open access journal you have to pay the publishing fees." (Interview G) "There are many OA journals or other journals which set very high article processing charges for authors. So especially for junior scholars in developing countries such as China, they may not have that much funding because they just started their career." (Interview B) If one assumes that OA scholarship can only be achieved through the payment of expensive article processing charges (APCs), it would make sense to assume that only certain scholars would have the ability to publish and distribute their work in that manner. This conflation of OA as a distribution method with a particular business model used by many commercial publishers speaks to the success of those publishers' ability to market gold OA as the standard for all openly licensed distribution of scholarship.
Participants also described how OA understood as primarily an "author-pays" model fails to benefit authors or institutions: "The effect of OA we see now is a bit different from the original intention of OA. From my recent experience with a faculty member in the physics department, I learned that one of their students had to pay more than RMB [Renminbi] 10,000 for an APC to publish in an OA journal… I think [the university] put a lot of funding to support research, and especially spent much to pay for the OA fee, but the fee it pays doesn't generate benefit to the library, nor the scholars themselves. What is the meaning of OA?" (Interview D) "No matter what kinds of OA model the publishers have developed, be it golden OA or green OA, they were after all just some concepts publishers created. The bottom line is that their financial interests can never be touched." (Interview D) In the aforementioned quotations, we hear a librarian lamenting that OA, as implemented by many publishers, fails to expand access to research and instead reinforces a status quo in which publishers extract revenue from the academy in ways that might sound new but perhaps are not. Again, we see an example of a familiar complaint within our own US context showing up in a Chinese context as well. Perhaps not coincidentally, as we were writing this manuscript, librarians and libraries in the US expressed skepticism about the "transformative" nature of the University of California's "transformative agreement" with Elsevier (University of Virginia Library, 2021; MIT Libraries, 2021). In both Chinese and US contexts, we find individuals operating within the current system of scholarly communication who can readily understand how their choices might be constrained by that system. It is worth noting that an APC of Renminbi (RMB) 10,000 (at time of writing, approximately $1,600 USD) would represent a significant expense for a scholar in China based on a recent survey of Chinese academics' salaries (Hu & Yuan, 2021).

The status quo versus imagined futures
If participants understand how their individual choices and behaviors are constrained by larger forces, what do they imagine could be possible in the future? We can answer this question by looking at the areas in which participants discussed both imagined futures and transformational changes. Notably, participants discussed transformational changes for scholars far more often than they discussed it in an institutional context (Figure 1), and the codes "imagining a boundless future" and "transformational change for scholars" occurred together seven times throughout our interviews (Figure 2). One of our participants summed up the relationship between the scholarly communication system of today and possible futures quite succinctly: "The status quo limits our imagination for the future" (Interview A).
In short, the current system-in which external incentives are necessary to change author behavior and OA is understood as primarily a system oriented around publication charges paid by authors themselves-puts the onus on individuals to push for change but limits what changes might be imaginable or even possible.
Although it might be tempting to see this as hopeless, a number of our participants did express robust hopes for the future.
"The best way [forward into the future] would be breaking the boundaries of journal platforms" (Interview A).
"We hope in the future the OA model we are developing now will transform scholars from being chased by us [publishers] to publish OA, to [authors] spontaneously wanting to share their works in OA format" (Interview E).
"OA is really an opportunity because of its [potential for increasing] diversity and inclusivity… I very much support OA, because in the traditional distribution model, only scholars can see my research. It's too closed. But with OA, if scholars are willing to open and share, the consumers of our research can also be the public" (Interview H).
The aforementioned quotations indicate hopeful sentiments from a librarian, a publishing professional, and a faculty member for a future in which the traditional boundaries of academic production are less restrictive, authors embrace openly licensing their works, and scholarship can find readers beyond the academy.
How might we reach these imagined futures? Our participants recognized that incentive systems would likely need to be devised and that "top-down" solutions might be the most effective for lasting change. Somewhat humorously, one of the librarians we interviewed characterized the technical challenges involved in embracing OA as "not difficult to solve" (Interview D), insisting instead that the harder problem was one of political will and resource allocation. Therefore, although the status quo might indeed limit our imaginations, it does not preclude having hope for something different.

Infrequently used but important codes
Whereas analyzing codes by frequency and co-occurrence allowed us to explore notable insights from our interviews, we found ourselves interested in exploring the implications of two infrequently used codes that addressed two themes present in our interviews. The first is a phenomenon we termed "language politics," the role that language of publication (Chinese or English) plays in OA practices in the Chinese academic landscape, and the second is the complex interaction between academic freedom and OA. Our interest in and commitment to exploring these themes represents a methodological choice to approach our work with a "critical" (in the philosophical sense) eye, attendant to the partiality and positionality of our work, and with a desire to avoid overreliance on positivist, strictly empirical research practice (Bhavnani et al., 2014). Therefore, although these codes appeared less frequently than others, that does not make them less meaningful as we seek to understand our research participants' experiences, and these less-frequently used codes can also lead to notable insights.

Language politics
Although we did not explicitly ask our participants questions about language of publication, the topic did come up in our interviews, along with discussions of language style as a tool of scholarly communication. We coded these portions of our interviews as "language politics" to reflect the fact that linguistic and stylistic choices are rarely solely individual choices and instead are influenced by international information systems, government policies, and disciplinary practices. Three interviewees mentioned different perspectives on language politics regarding OA publishing. This section discusses two main questions: 1) what role language plays in scholars' decision-making regarding linguistic venues and format of publication; and 2) how language of publication and academic writing style could become barriers for scholarly communication in the OA environment.
English maintains its status as the de facto global language despite the growing importance of Chinese (Tollefson, 2018). In the global academic publishing landscape, anglicization prevails, and "the unified linguistic markets slanted towards English" stand in the way of unimpeded spread of research produced by non-English speaking Chinese scholars (Zheng & Guo, 2019, p. 126). The dominance of English as the standard language for international academic exchanges pushes scholars to pursue English-language publications in order to achieve more efficient global dissemination. As one of our participants straightforwardly described: "English is still the major language for scholarly communication. So I would choose to publish in English more." (Interview B).
Governmental and institutional incentives also influence scholars' decisions to publish in English. Recognizing that Chinese humanities and social sciences scholarship has been lagging behind STEM fields in terms of global visibility, Chinese policymakers prioritized investment in humanities and social sciences scholarship as part of a series of efforts to internationalize Chinese academia. Speaking at a high-profile symposium in 2016, Chinese President Xi Jinping pointed out that Chinese academia in the humanities and social sciences should not only "learn from international academia but also… promote Chinese discourses and achieve global impact" (Xu, 2020, p. 159). Research-intensive universities also devised and implemented incentive programs aimed at increasing international publications in humanities and social sciences disciplines (Xu et al., 2019).
Unlike in China, scholars in the US are not subject to formal, institutionally sponsored incentives to publish in a particular language. Instead, the dominance of English as a language of publication can be attributed to historical (and ongoing) trends in systems of tenure and promotion and the bibliometrics that have become increasingly important as tools of evaluation within those systems (Curry & Lillis, 2013). As a result, regardless of explicit institutional policy, US faculty can feel hesitant to publish in languages other than English owing to concerns about promotion (Fuentes & Gómez Soler, 2018). Outside of China and the US, the combination of explicit institutional (and sometimes national) policy and the dominance of English-privileging bibliometrics has resulted in faculty authors who feel publishing in English is a "taken-for-granted reality" (Pérez-Llantada, 2018).
Two notable trends in Chinese publishing are likely also influencing the English-language publication venues available to scholars in China. First, China's leading journal publishers have created their own English-language OA journals. With governmental and institutional subsidies, this enterprise has gained "competitive advantage in international markets" in terms of lowering or even eliminating APCs compared with English-language publishing venues outside China (Montgomery & Ren, 2018, p. 21). Second, in the last decade, there has been a surge in publication of English translations of Chinese-language research literature. In May 2013, China National Knowledge Infrastructure (CNKI), the largest commercial aggregator of Chinese-language publications, launched its "Journal Translation Project," a database featuring translations of selected Chinese-language journal articles, with 42% of the content from humanities and social sciences disciplines thus far (J. Zhang, personal communication, October 15, 2021). However, the effectiveness of this approach is unclear owing to the lack of competent translators with disciplinary expertise (Ren & Montgomery, 2015, p. 403). One of the unintended consequences of incentivizing publication in English is, not surprisingly, downplaying Chinese-language academic publishing for Chinese scholars.
Turning to the question of how language of publication and writing style enable an author to reach particular audiences, a participant shares insights into this question by connecting the mission of OA to the publication language. As one of few participants to explicitly mention public access to research, interviewee H stresses that the public "has the right to know and should be given opportunities to know" the results of scholarly research and "scholars should open [their] research to them." Believing OA to be the ideal mechanism for academia to achieve openness to the public, this participant suggests that "OA probably should include multiple languages in the future to accommodate people with different cultural backgrounds to get support from this public platform" (Interview H). According to their interpretation, the nature of the OA environment has the potential to lower the language barrier for both scholars and the global reading community because they could freely choose the languages they are more familiar with to read and write. In this regard, multilingualism could become a significant advantage of OA compared with traditional publishing. In this participant's view, privileging certain languages for publication would contradict OA's mission.
For humanities and social sciences scholars in China, publishing in English-language journals hosted in the West could come with cultural or ideological constraints. Interviewee H points out the realistic challenge: "Each journal has a core organization to run it. The editorial boards, editors-in-chief, and reviewers… are probably familiar with each other. They are in the same circle, even from the same countries, and speaking the same language. For us, it will be rather difficult to enter their circle and earn their recognition. People are usually more willing to accept ideas, expressions, and concepts that are similar to their own" (Interview H).
Provincialism on the part of editors and reviewers-if not outright prejudice-could create hurdles for those Chinese scholars who do not conform to Western norms to share their work with a larger audience. The interviewee acknowledged that this impression comes from traditional publishing practices in their own discipline. They held high hopes that the OA publishing environment would help improve the diversity of editorial boards of academic journals and allow for more-equitable contributions. JLSC Some interviewees also indicated that the OA environment is conducive to transforming scholarly writing styles and improving the accessibility of scholarly works. When discussing the value of a new OA publishing initiative that Interviewee E managed, they envisioned its potential impact on the presentation of academic writings: "[The platform] would lower the threshold for the public to come closer to the academic world but not drive them away from it. It's rather common that scholarly works are decorated with discourses and jargon, but they don't have to be. Academic authors can change their language to an easy-to-understand style when explaining their scholarship. This way it's easier to promote scholarship to more people" (Interview E).
Given that one of the critical missions of OA is to maximize the readership of scholarly works, academic writers need to be mindful that, in the OA environment, their audience could reach beyond a limited and rarefied few to also include the public. This raises questions regarding linguistic accessibility of the work: whether the scholarship is delivered, in another interviewee's words, "in a plain language style and with concrete, practical examples" (Interview H). By making wider distribution of scholarship possible, OA publishing invites scholars to reconsider how they might approach a more-diverse audience and fulfill the movement's promise of more-equitable access to research.

Academic freedom
Two of our interviewees mentioned academic freedom while discussing OA in China. In the US, academic freedom is generally understood as freedom from "institutional censorship" in the course of conducting one's research and teaching as long as the "special obligations" of the profession-speaking as "private persons" with "accuracy," "restraint," and "respect"-are upheld (AAUP, 2014, p. 14). The concept and its practice in the US date back to the early nineteenth century, and although it has not adequately and equitably protected all faculty everywhere since its codification in 1915 or its reaffirmation in updated statements since that time, it remains relevant and contested to this day. As we worked on this manuscript in the summer of 2021, a prominent African American scholar-a recipient of both a Mac-Arthur "genius" grant and a Pulitzer Prize-declined an offer of tenure from one university to join another in the wake of conservative objections to her scholarship (Wamsley, 2021).
How might we understand the relationship between academic freedom and OA in a US context? For those unfamiliar with changing labor patterns in American academia or the attitudes of those faculty who still benefit from the protections of tenure, it might be assumed that commitment to academic freedom and commitment to OA practices go hand in hand.
However, as many liaison librarians can report, faculty have not always seen these commitments as being in alignment, and they can perceive OA policies, particularly institutional ones, as potentially encroaching on their academic freedom (Johnston, 2017). In addition, the majority of the professoriate no longer enjoys the protections of tenure, which have traditionally been a key component of maintaining the academy's commitment to academic freedom. As of 2017, 70% of all instructional staff in American higher-education institutions are nontenure-track faculty in contingent or otherwise unprotected positions (Background facts on contingent faculty positions, n.d.).
Although faculty attitudes about OA are changing and a variety of institutional OA policies have proliferated (Johnston, 2017), it is important to remember that we cannot fall back on simplified understandings of the relationship between these concepts in the US, particularly if we seek to understand Chinese academic practice. It would not be true to simply say "but we have academic freedom in America" or "OA and academic freedom have a mutually supportive and uncomplicated relationship for American scholars." Taking care to remember the complexity and contested nature of academic freedom in our own academic environment prepares us to keep an open mind and critical eye as we endeavor to understand it in another environment.
Much like the US, in China, the concept of academic freedom can be traced back to the 1910s.
During the first few decades of the twentieth century, the concept was incorporated into the internal management of modern higher-educational institutions (Zhan & Kang, 2018). The practical definition of academic freedom and to what extent a scholar is protected by it have varied since the 1950s, depending on the intensity of the Chinese Communist Party (CCP) and the government's regulation of academia. For example, the Chinese government has long restricted the use of International Standard Serial Numbers (ISSNs) to authorized journal publishers, limiting the use of this fundamental piece of publishing infrastructure (Regulations of September 20, 2005, on Administration of Periodical Publications, n.d.). In the most recent decade, evidence has shown that academic freedom is shrinking for scholars and universities in China. For example, the CCP strengthened ideological constraints on the scope of scholarly research and publishing by discouraging scholars from including "sensitive topics" in their research agendas. In 2013, the list of these topics was expanded, with a ban reportedly issued by the CCP under President Xi Jinping on the so-called "Seven No's" or "seven speaknots" (Hao, 2015, pp. 116-117), including concepts such as "freedom of speech" and "universal values," which have particular relevance for journalism and academia (Carlson, 2013).
Challenges to the legitimacy of academic freedom continue, as a number of prominent universities recently removed intellectual freedom from their charters (Hernández & Zhang, 2019). Although on its face the "iron rice bowl" policy (or "up-or-transfer" model) might appear to protect academic freedom in a manner similar to tenure in the US given its implied job security, the foundation of the policy is managerial rather than a scholarly prerogative. As such, the significant pressure on academic freedom in China cannot be allayed by entitlements such as tenure, as practiced in the US.
In addition to explicit regulation of the academy and a shrinking space for practical expressions of academic freedom, state influence over research practice extends to the arena of research funding as well. The National Planning Office for Philosophy and Social Sciences (NPOPSS), a unit within the CCP's propaganda branch that administers the National Social Science Fund (NSSF), is one of the most important funding agencies for social sciences research in China. The NPOPSS has been adjusting their funding structure since the early 2010s to move away from an overemphasis on party theory in overall research agendas. However, in the meantime, even with an increasingly diversified agenda-setting effort, its newly realigned funding priorities continue to fit with the CCP's recent emphases on national history and socialist culture, ensuring funded projects are still in service of CCP's missions and goals (Holbig, 2014, pp. 26-28).
Addressing the relationship between academic freedom and OA in China, a publishing professional described a potential challenge in their interview: "Another thing I want to mention is that some scholars may think their words will be in the spotlight and easily exaggerated after sharing. So they don't want to share their opinions anywhere/everywhere. For them, maybe publishing a journal article is the safest. But other comments or speeches, even only a few sentences they speak somewhere, may be interpreted out of context or misunderstood. This is also a big obstacle for OA in present day China, which makes scholars not willing to freely express their views. In China before you make a remark, you will have to know its conse-quences… I think the implementation of OA also requires a very open environment for expressing one's opinion" (Interview E).
The publishing professional explicitly attributes scholars' hesitation to embrace OA to a fear of the consequences of expressing their ideas. As the publishing professional describes it, the author might be willing to take the risk of having their work misinterpreted or their conclusions challenged in a conventional, closed publication system of paywalls and subscriptions, but the equation changes as risks are multiplied by the effectiveness of OA as a distribution model.
A faculty member and high-level administrator discussed the impact of censorship on scholars and their work during an interview, with the implications ranging from practical difficulties in conducting research to fundamental challenges to understanding Chinese society. In a spirited and wide-ranging conversation, they described the realities of living behind the "Great Firewall," which can make even mundane tasks-emailing colleagues abroad, sharing files, or registering for conferences-difficult for Chinese scholars (Interview F). The administrator noted that, whereas the Great Firewall serves as an important political tool for the Chinese government, it creates "lots of micro-harm" for academics (Interview F) by limiting their free access to information and constraining how they can use the internet to conduct and disseminate their scholarship. They also noted that heavy-handed government control over media and the academy negatively affects the quality of data that is available about many aspects of daily life in China, including demographics, economics, and politics. Facetiously referencing the largest English-language news source published in China, the administrator asked, "How could American people know the Chinese? They just see China Daily." The administrator described how a lack of quality data about China discourages the study of social phenomena in China by both scholars abroad and Chinese scholars themselves because ready access to reliable information (rather than "propaganda," as they described it) forms the basis of much social sciences scholarship.
Our interviewees' uneasy sentiments toward the conditions of OA in China reflect two distinct trends that coexist in China. On the one hand, in recent years the government has sponsored many large-scale OA projects, which is arguably a welcome development. For example, in 2013, the Chinese Academy of Social Sciences (CASS) released the National Social Sciences Database (NSSD), which was approved by the aforementioned NPOPSS as an NSSF-funded special project. The NSSD was designed as a public-facing platform to promote open sharing, far-reaching dissemination, and internationalization of Chinese scholarship. It provides full text of major journals funded by the NSSF, CASS-administered journals, and over 2,000 other scholarly journals in philosophy, humanities, and social sciences (NSSD, n.d.). Government support has made the NSSD a leading source of Chinese-language scholarship.
However, on the other hand, evidence of censorship in the digital realm has become prevalent.
Besides the internet regulation that Interviewee F described earlier, digital surrogates of print academic resources are another target of digital policing. Scholar Glenn Tiffert investigated an alarming case in which the digitized copies of two significant Chinese law journals fell victim to content sanitation involving removal of articles from the period 1956 to 1958 on "rightists" and the Anti-Rightist Campaign. This was a formative period in the CCP's efforts to consolidate power and assert ideological control. The CCP's ideological campaign reached its climax in initiating the Anti-Rightist Campaign in 1957. The campaign significantly targeted intellectuals, but intellectuals also contributed to the formation of the campaign's culture (Wang, 2020, p. 188). The articles describing this important period of modern Chinese history were suppressed from the two major online platforms for Chinese scholarly journal articles. One of these platforms was precisely the OA NSSD. Tiffert argues that, by omitting these articles from digital existence, censors are "purposefully rewriting the history of this period" and are therefore causing contemporary researchers to "unwittingly promote the agendas of the censors" when faithfully using the digital version of these source materials (Tiffert, 2019, p. 554, p. 559 As Tiffert's investigation demonstrated, China's OA infrastructure can be deployed in the service of political goals decidedly out of step with the purported philosophy of OA or even the practical goals of unfettered access to scholarship. Our participants were not unaware of this contradiction, and, as information professionals, we are interested in the attendant implications of an OA that operates via coercive rather than persuasive measures.
This coexistence of robust infrastructure for the funding and dissemination of scholarship along with state-sponsored censorship leads us to wonder the following: is what we know as OA practiced in China? Can OA as we know it exist in an environment without freedom of expression, access to an unrestricted internet, an independent publishing industry, and educational institutions with some degree of autonomy from the political goals of a powerful central government?
As linguist Anna Wierzbicka cautions in her book Imprisoned in English, direct translation of meaning for complex concepts across linguistic boundaries is not easily attainable, and "the blinding power of English as the global language of science" can lull us into forgetting that English is as culturally contextual as any other language (Wierzbicka, 2013, p. 4, p. 185). When we say "OA" in North America, we bring with it a complex web of underlying assumptions tied to the history and practice of scholarship. Saying "OA" in China brings with it a similarly complex web of underlying assumptions, which may not be commensurate or coterminous with the OA of English-speaking North America. Our Chinese participants described experiences not unlike those of their American counterparts: a system in which authors' choices are constrained by forces beyond their individual control, that the assumed financial costs of OA publishing exceed its perceived benefits, and a status quo that limits imagined futures. However, beyond the surface-level resemblance, important differences persist.
Perhaps OA as we know it in the Budapest/Bethesda/Berlin sense of the term does not exist in China. However, obviously, a kind of OA does indeed exist and is thriving. We could call this "OA with Chinese characteristics," which we argue is a form of OA that relies on governmentsponsored infrastructure to enable public sharing of Chinese-language scholarship with the wider world. OA with Chinese characteristics intersects with other forms of OA by enabling individual Chinese authors to participate in OA publishing venues and systems outside of China via the standard practices of those venues and systems and their geographically informed norms. However, both the domestic practice of OA in China and the choices of Chinese scholars who publish abroad are subject to coercive measures of control, including international information systems that privilege the use of English, disciplinary trends toward monolingualism, government censorship, lack of speech protections, and funding mechanisms dominated by government entities. Inspired by Huang's (2008) description of Chinese economic practice, we could say that "OA with Chinese characteristics" is a form of OA that has many of the familiar hard infrastructures but few of the soft infrastructures we commonly associate with OA.
These multiple understandings of OA do not need to converge on a single shared meaning (and it is unclear whether that is possible or desirable), but we can endeavor to understand one another better in the service of creating and sharing knowledge. By describing and naming "OA with Chinese characteristics," we do not want to imply that this form of OA is somehow lesser than or insufficiently authentic when compared with OA as we know it in the US context or that scholars in the US are somehow exempt from other forms of coercive control. Instead, we ask that scholars of OA move beyond the idea of OA in the North American and European contexts serving as the "unmarked" OA, with the practices of the primarily Anglophone Global North serving as a baseline against which OA in other contexts is always measured. In our work, we have endeavored to understand OA in China on its own terms in order to offer those unfamiliar with the Chinese academy a nuanced picture of the beliefs and lived experiences of scholars, administrators, and information professionals in China. At the same time, we do not wish to be apologists for OA with Chinese characteristics. Rather, we hope that, as the landscape of OA is understood more fully in all its "complexity and contradiction," to borrow a phrase from architect and critic Robert Venturi (Stierli, 2016), practitioners, authors, and readers will continue to engage critically with OA as both a theoretical project and a practical distribution method.